A TRAF-like E3 ubiquitin ligase TrafE coordinates ESCRT and autophagy in endolysosomal damage response and cell-autonomous immunity to Mycobacterium marinum
- Départment de Biochimie, Faculté des Sciences, University of Geneva, Switzerland
Figures
Figure 1 with 4 supplements
Upon infection TrafE is upregulated and recruited to MCVs.
(A) GFP-TrafE-expressing D. discoideum cells (green) were mock-infected or infected and then assessed at 1.5, 3, 6, or 24 hr with mCherry-expressing M. marinum (red). Representative maximum projections of live time-lapse spinning disk confocal images with arrowheads pointing at GFP-TrafE recruitment to MCVs/bacteria. Scale bars correspond to 10 µm. Images are representative of at least three independent experiments. (B) Quantification of the percentage of intracellular MCVs/bacteria positive for GFP-TrafE during the infection time-course. Each point is representative of a 6 random multi-position field with n=2 from N=3 independent experiments. Bars represent SEM. (C) Quantification of GFP-TrafE in the vicinity of M. marinum poles from random images in N>3 independent experiments. (D) Normalized mRNA levels of trafE in mock-infected or M. marinum-infected D. discoideum cells at 1.5, 6, or 24 hr. Shown are mean and SEM of the fold change (FC) representing three independent experiments. Statistical differences were calculated with an unpaired t test (n.s.: non-significant, *: p-value ≤0.05). (E) The intracellular growth of M. marinum-lux was monitored every 1 hr inside WT and TrafE overexpressing cells, for 72 hr. Shown are mean and SEM of the fold change (FC) from three independent experiments. Statistical differences were calculated using Bonferroni multiple comparison test after ANOVA (***: p-value ≤0.001). (F) Recruitment of the endogenous TrafE-GFP (green) to MCV/bacteria (red) after 1.5 hpi. Scale bars correspond to 10 µm. Images are representative of at least three independent experiments.
-
Figure 1—source data 1
Quantification data for Figure 1B.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig1-data1-v2.xlsx
-
Figure 1—source data 2
Quantification data for Figure 1C.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig1-data2-v2.xlsx
-
Figure 1—source data 3
Raw data for RT-qPCR in Figure 1D.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig1-data3-v2.xlsx
-
Figure 1—source data 4
Raw data for intracellular growth assay in Figure 1E.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig1-data4-v2.xlsx
Figure 1—figure supplement 1
Identification of D. discoideum TRAF-like proteins.
(A, B and C) UniProt generated Multiple Sequence Alignment, similarity matrix and phylogenetic tree of human TRAF6 (Q9Y4K3) and D. discoideum TrafA (AC: Q559R2), TrafB (AC: Q54NN4), TrafC (AC: Q54FG0), TrafD (AC: P11467), and TrafE (AC: Q54FB9) (The UniProt Consortium, 2022). (D) TrafE amino acid sequence NCBI Domain Enhanced Lookup Time Accelerated BLAST (Sayers et al., 2022). (E) TRAF6 and D. discoideum TrafA, TrafB, TrafC, TrafD, and TrafE AlphaFold protein structure prediction obtained from UniProt (The UniProt Consortium, 2022).
Figure 1—figure supplement 2
trafE is upregulated early after infection.
RNA-sequencing of non-infected and M. marinum-infected D. discoideum cells. RNA levels were compared and the fold change (FC) was plotted.
Figure 1—figure supplement 3
TrafA, TrafB, TrafC, and TrafD behaviour after infection with M. marinum.
(A) GFP-TrafA or (B) GFP-TrafB expressing D. discoideum cells (green) were mock-infected or infected for 1.5, 3, 6 or 24 hr with mCherry-expressing or DsRed-expressing M. marinum (red). Representative maximum projections of live images and quantification of at least three independent experiments. Scale bars correspond to 10 µm. (C) GFP-TrafC or (D) GFP-TrafD expressing D. discoideum cells (green) were mock-infected or infected for 1.5, 3, 6 or 24 hr with mCherry-expressing or DsRed-expressing M. marinum (red). Representative maximum projections of live images and quantification of at least three independent experiments. Scale bars correspond to 10 µm.
-
Figure 1—figure supplement 3—source data 1
Quantification data for Figure 1—figure supplement 3.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig1-figsupp3-data1-v2.xlsx
Figure 1—figure supplement 4
M. marinum growth is not affected by TrafC, TrafD overexpression and TrafE C-terminal GFP tagging.
(A) The intracellular growth of M. marinum-lux was monitored every hour in WT cells or cells expressing TrafC, TrafD, or TrafE GFP-fusions. Error bars represent SEM. Statistical differences from three independent experiments were calculated using Bonferroni multiple comparison test after ANOVA (n.s.: non-significant). (B) The intracellular growth of M. marinum-lux was monitored every hour inside WT or TrafE-GFP KI cells. Error bars represent SEM. Statistical differences from three independent experiments were calculated using Bonferroni multiple comparison test after ANOVA (n.s.: non-significant).
-
Figure 1—figure supplement 4—source data 1
Quantification data for Figure 1—figure supplement 4.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig1-figsupp4-data1-v2.xlsx
-
Figure 1—figure supplement 4—source data 2
Quantification data for Figure 1—figure supplement 4.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig1-figsupp4-data2-v2.xlsx
Figure 2 with 1 supplement
Loss of TrafE is detrimental for D. discoideum cells after infection with M. marinum.
(A) The intracellular growth of M. marinum-lux was monitored every 1 hr inside WT, atg1-KO, trafE-KO or GFP-TrafE complemented trafE-KO cells, for 72 hr. Shown are mean and SEM of the fold change (FC) representing three independent experiments. Statistical differences were calculated using Bonferroni multiple comparison test after ANOVA ***: p-value ≤0.001. (B) WT, atg1-KO or trafE-KO cells expressing mCherry-Plin were infected with GFP-expressing M. marinum. The percentage of cells with intracellular bacteria colocalizing with Plin was assessed manually at 6 and 24 hpi. Error bars indicate the SEM from 3≤n ≤ 4 relicates from N=3 independent experiments. Statistical differences are indicated with an asterisk and were calculated with an unpaired t test (n.s.: non-significant, *: p-value ≤0.05, **: p-value ≤0.01, ***: p-value ≤0.001). (C–D) WT or trafE-KO cells infected with DsRed-expressing M. marinum imaged by high-content microscopy every 1 hr. Number of infected cells (C) or number of extracellular bacteria (D) were monitored and the average of three replicates n≥200 cells per time point was plotted. Error bars represent SEM. Cell strain-time dependent statistical differences were calculated using Bonferroni multiple comparison test after two-way ANOVA (**: p-value ≤0.01). (E) WT, atg1-KO, trafE-KO or GFP-TrafE-complemented trafE-KO cells infected with mCherry-expressing M. marinum in microfluidic chip single-cell experiment were monitored every 1 hr. The number of live cells was counted at each time point for 20 hr and the counts from N=3 independent experiments were plotted as Kaplan-Meier probability of survival curves. Statistical differences were calculated using Bonferroni multiple comparison correction after two-way ANOVA (n.s.: non-significant, ***: p-value ≤0.001).
-
Figure 2—source data 1
Raw data for intracellular growth assay in Figure 2A.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig2-data1-v2.xlsx
-
Figure 2—source data 2
Quantification data for Figure 2B.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig2-data2-v2.xlsx
-
Figure 2—source data 3
Raw data for high-content segmentation in Figure 2C.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig2-data3-v2.xlsx
-
Figure 2—source data 4
Raw data for high-content segmentation in Figure 2D.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig2-data4-v2.xlsx
-
Figure 2—source data 5
Quantification data for single-cell experiment in Figure 2E.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig2-data5-v2.xlsx
Figure 2—figure supplement 1
Development of D. discoideum (A); M. marinum escapes early rom the MCV to the cytosol in trafE-KO cells (B).
(A) WT, atg1-KO, and trafE-KO cells were deposited on Soerensen-agar plates and visualized by microscopy after 24 hr of starvation. WT and trafE-KO cells aggregate and form fruiting bodies, whereas atg1-KO cells do not aggregate and do not progress to the formation of fruiting bodies. (B) WT, atg1-KO or trafE-KO cells expressing mCherry-Plin (red) infected with GFP-expressing M. marinum (green) at 6 hpi and 24 hpi.
Figure 3
TrafE recruitment to MCVs/bacteria is membrane damage-dependent.
(A) D. discoideum cells expressing endogenous TrafE-GFP (green) were infected and then assessed at 1.5, 3, or 24 hr with M. marinum WT (red) or M. marinum ∆RD1 (red). Representative maximum projections of live images with arrowheads pointing at TrafE-GFP recruitment to MCVs/bacteria. Scale bars correspond to 10 µm. Images are representative of three independent experiments. (B) Quantification of the percentage of intracellular MCV/bacteria positive for TrafE-GFP during the infection time-course. SEM from two to four independent experiments. Statistical differences were calculated with an unpaired t test (****: p-value ≤0.0001). (C) The intracellular growth of M. marinum-lux WT or M. marinum ∆RD1 was monitored every hour inside WT or trafE-KO cells, for 72 hr. Shown are mean and SEM of the fold change (FC) representing three independent experiments. Statistical differences were calculated using Bonferroni multiple comparison test after one-way ANOVA (***: p-value ≤0.001). (D) WT, atg1-KO, trafE-KO, or GFP-TrafE-complemented trafE-KO cells infected with M. marinum-lux ∆RD1 in microfluidic chip single-cell experiment were monitored every 1 hr. The number of live cells was counted at each time point for 20 hr and the counts from three independent experiments were plotted as Kaplan-Meier survival curves. Statistical difference were calculated using Bonferroni multiple comparison correction after ANOVA (n.s.: non-significant).
-
Figure 3—source data 1
Quantification data for Figure 3B.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig3-data1-v2.xlsx
-
Figure 3—source data 2
Raw data for intracellular growth assay in Figure 3C.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig3-data2-v2.xlsx
-
Figure 3—source data 3
Quantification data for single-cell experiment in Figure 3D.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig3-data3-v2.xlsx
Figure 4
K-63-linked polyubiquitination decrease in D. discoideum trafE-KO cells.
(A) Maximum projections showing colocalization (white arrowheads) of M. marinum (blue), K63-linked polyubiquitin chains (grey) and GFP-TrafE (green). Maximum projections showing (B) polyubiquitin (grey) or (D) K63-linked polyubiquitin (grey) colocalization (white arrowheads) with M. marinum. Scale bars correspond to 10 µm. Quantification of the percentage of (C) FK2-positive or (E) K63-positive intracellular MCV/bacteria at 6 hpi was carried out manually from 10 to 15 images per experiment n≥200 cells. Error bars indicate SEM. Statistical differences were calculated with an unpaired t test (n.s.: non-significant, **: p-value ≤0.01).
-
Figure 4—source data 1
Quantification data for immunofluorescence in Figure 4C.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig4-data1-v2.xlsx
-
Figure 4—source data 2
Quantification data for immunofluorescence in Figure 4E.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig4-data2-v2.xlsx
Figure 5 with 1 supplement
TrafE recruitment to MCV is RING domain-dependent.
(A) D. discoideum trafE-KO cells expressing GFP-TrafE, GFP-TrafE with truncated RING domain or GFP-TrafE with truncated N-terminal TRAF domain (green) were infected and then assessed at 1.5, 3, 6 or 24 hr with mCherry-expressing M. marinum (red). Representative maximum projections of live images with arrow heads pointing at GFP- TrafE recruitment to MCVs/bacteria. Scale bars correspond to 10 µm. (A) Images and (B) quantifications are representative of three independent experiments. (C) The intracellular growth of M. marinum-lux was monitored for 72 hr, every 1 hr inside WT, trafE-KO or D. discoideum cells expressing GFP-TrafE with truncated RING domain or GFP-TrafE with truncated N-terminal TRAF domain. Shown are mean and SEM of the fold change (FC) from three independent experiments. Statistical differences were calculated using Bonferroni multiple comparison correction after ANOVA (n.s. : non-significant, *** : p value≤ 0.001).
-
Figure 5—source data 1
Raw data for infection quantification in Figure 5B.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig5-data1-v2.xlsx
-
Figure 5—source data 2
Raw data for intracellular growth assay in Figure 5C.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig5-data2-v2.xlsx
Figure 5—figure supplement 1
Multiple sequence alignment (A) and domains and respective deletions (B).
(A) Multiple Sequence Alignment of D. discoideum TrafA (AC: Q559R2), TrafB (AC: Q54NN4), TrafC (AC: Q54FG0), TrafD (AC: P11467), and TrafE (AC: Q54FB9) with red rectangles that outline the predicted TrafE RING, TRAF, and MATH domains (The UniProt Consortium, 2022). The deleted regions of TrafE RING and TRAF domains are highlighted. (B) Table showing the predicted coordinates of RING, TRAF, and MATH domains with the TrafE RING and TRAF domain deletions coordinates in parenthesis.
Figure 6 with 2 supplements
LLOMe-induced damage triggers TrafE recruitment to endolysosomes.
D. discoideum trafE-KO cells expressing (A) GFP-TrafE, (B) GFP-TrafE with truncated N-terminal TRAF domain or (C) GFP-TrafE with truncated RING domain (green) were monitored by spinning disk confocal microscopy each 1 min for 25 min, 5 min before (t = –5) and 20 min after addition of 4.5 mM LLOMe at (t=0). White rectangles indicate zoom of selected areas. Images are representative of at least three independent experiments. Scale bars correspond to 10 µm. (D) ImageJ SpotCounter quantification of GFP-TrafE number of structures from N=2 independent experiments. (E) Colocalization of GFP-TrafE (green) with Alexa Fluor 647 10 kDa dextran (red) 10 min after LLOMe treatment. Images are representative of N>3 independent experiments. Scale bars correspond to 10 µm. (F) ImageJ SpotCounter count of GFP-TrafE or Alexa Fluor 647 10 kDa dextran structures followed by manual count of their colocalization from N=3 independent experiments. (G) Maximum projection confocal images of live D. discoideum WT or trafE-KO cells incubated with Lysosensor Green DND-189 (green), imaged every 1 min, 5 min before and 30 min after addition of 4.5 mM LLOMe. The plot is a representation of n=3 technical replicates out of N=3 biological replicates performed. Scale bars correspond to 10 µm. (H) Quantification of pH tracer Lysosensor Green DND-189 (green) fluorescence average intensity using ImageJ Time Series Analyzer V3. Error bars represent SEM. Statistical differences were calculated using Bonferroni multiple comparison correction after two-way ANOVA (***: p-value ≤0.001).
-
Figure 6—source data 1
Raw data for GFP-TrafE dot number quantification in Figure 6D.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig6-data1-v2.xlsx
-
Figure 6—source data 2
Raw data for GFP-TrafE dextran colocalization in Figure 6F.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig6-data2-v2.xlsx
-
Figure 6—source data 3
Raw data for lysosensor intensity quantification in Figure 6H.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig6-data3-v2.xlsx
Figure 6—video 1
TrafE relocalization to endolysosomal compartments upon LLOMe treatment.
Live microscopy time-course of D. discoideum trafE-KO cells expressing GFP-TrafE, treated with 4.5 mM LLOMe. Images were taken every 1 min, 5 min before (t=0) and 55 min after LLOMe treatment. Images are representative of at least three independent experiments. Scale bars correspond to 10 µm.
Figure 6—video 2
TrafE colocalizes with 10 kDa dextran upon LLOMe treatment.
Colocalization of GFP-TrafE (green) with Alexa Fluor 647 10 kDa dextran (red) 10 min after LLOMe treatment. Images were taken every 5 min, for 13 min after LLOMe addition (t=0). Scale bars correspond to 10 µm.
Figure 7 with 2 supplements
TrafE is relocalized to endolysosomes following hypertonic shock.
(A) Maximum projection of confocal images of D. discoideum cells expressing GFP-TrafE (green) imaged every 1 min, 5 min before and 55 min after addition of 200 mM Sorbitol. Images are representative of at least three experiments. Scale bars correspond to 10 µm. (B) ImageJ SpotCounter quantification of GFP-TrafE number of structures. (C) Maximum projection of confocal images of live D. discoideum WT cells incubated with Lysosensor Green DND-189 (green) and imaged every 1 min, 5 min before and 30 min after addition of 4.5 mM LLOMe or 200 mM sorbitol. Images are representative of three independent experiments. Scale bars correspond to 10 µm. (D) Quantification of fluorescence average intensity of pH tracer Lysosensor Green DND-189 (green), before and after LLOMe or Lysosensor addition, using ImageJ Time Series Analyzer V3. Error bars represent SEM. Statistical difference were calculated using Bonferroni multiple comparison test after two-way ANOVA (***: p-value ≤0.001). (E) WT, atg1-KO or trafE-KO cells were incubated in propidium iodide (PI) and monitored by live confocal time-lapse microscopy every 1 min, 5 min before and 55 min after addition of 5 mM LLOMe. Number of PI-positive cells at 55 min were counted from 4 independent experiments. Error bars represent SEM. Statistical differences were calculated with an unpaired t test (n.s.: non-significant, *: p-value ≤0.05, **: p-value ≤0.01).
-
Figure 7—source data 1
Raw data for GFP-TrafE dot number quantification in Figure 7B.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig7-data1-v2.xlsx
-
Figure 7—source data 2
Raw data for lysosensor intensity quantification in Figure 7D.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig7-data2-v2.xlsx
-
Figure 7—source data 3
Quantification data for PI-positive cells in Figure 7E.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig7-data3-v2.xlsx
Figure 7—video 1
Vps32 is relocalized to endolysosomes following hypertonic shock.
Maximum projection of confocal images of D. discoideum cells expressing act5::GFP-Vps32 (green) imaged every 15 s, 1 min before and 9 min after addition of 200 mM Sorbitol. Images are representative of at least three experiments. Scale bar corresponds to 10 µm.
Figure 7—video 2
TrafE is required for optimal membrane damage response.
WT, atg1-KO or trafE-KO cells were stained with propidium iodide (PI) and monitored by live confocal time-lapse microscopy every 1 min, 5 min before and 55 min after addition of 5 mM LLOMe. No LLOMe control to assess the effect of the fluorescent excitation on the cell fitness.
Figure 8
RNA-seq results.
(A) Volcano plot of Differential Expression (DE) genes comparing trafE-KO to WT. After filtering for remaining ribosomal gene reads and low expressed genes, 8744 genes were left for the DE genes analysis, of which 283 were DE (absolute log2fc ≥ 0.585, FDR ≤ 0.05). From these, 144 were upregulated (red) and 139 were downregulated (blue). For illustrational purposes, trafE was included, which is DE and downregulated (blue dot on the far left), which is a confirmation of the trafE clean KO on transcriptional level. (B) Heatmap of normalized read counts. Filtered and normalized read counts were interrogated for a previously defined list of genes of interest. This included a set associated with autophagy (upper segment) and a set associated with the ESCRT machinery (lower segment). Biological replicates are depicted as three columns per condition (WT and trafE-KO). A clear pattern of upregulation is visible in autophagy-related genes, whereas no clear overall pattern is visible for ESCRT-related genes. However, key players in the ESCRT machinery such as alxA show a striking pattern and most notably, Ca2+ dependent regulators pefA and pefB are regulated inversely. None of the genes depicted are DE by the statistical criteria defined previously. (C) Gene set enrichment analysis (GSEA) of log2 fold change of filtered genes. For GSEA, genes were ranked by log2 fold change, resulting groups were restricted by minimum and maximum group size of respectively 2 and 200 and considered significant at a p value ≤ 0.05 and a q-value ≤ 0.1. Depicted are the log2 fold change density distributions for enriched core genes in each respective term. The significant signature included several metabolic terms including Proteasome, Endocytosis and most notably Autophagy as being upregulated.
-
Figure 8—source data 1
RNA-seq mapped reads, DEG and GSEA OTs for Figure 8A, B and C.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig8-data1-v2.xlsx
Figure 9
TrafE absence promotes Atg8a accumulation or formation of aberrant structures in starved or LLOMe-treated cells.
(A) Live confocal microscopy images of D. discoideum WT or trafE-KO cells expressing act5::GFP-Atg8a (green) incubated for 12 hr with Alexa Fluor 647 10 kDa dextran (red) and imaged 90 min after a shift to SorMC-Sorbitol starvation medium. (B) Live high-content imaging quantification of D. discoideum WT or trafE-KO cells expressing act5::GFP-Atg8a 90 min after a shift to SorMC-Sorbitol starvation medium. GFP-Atg8a dot average number per cell was collected from 27 images with 20≤n ≤ 40 cells per image. (C) Quantification of live time-lapse high-content confocal images of D. discoideum WT or trafE-KO cells expressing act5::GFP-Atg8a, 5 min before (t = –5) and after addition of 4.5 mM LLOMe at t=0 for a total of 40 min. Images were taken every 5 min and GFP-Atg8a dot count per image was collected from 3 wells with 2 image fields per well, each image contains 20≤n ≤ 40 cells. The plot is a representation of n=3 technical replicates out of N=3 biological replicates performed. (D) Live time-lapse spinning disk confocal images of D. discoideum WT or trafE-KO cells expressing act5::GFP-Atg8a, 5 min before (t = –5) and after addition of 4.5 mM LLOMe at t=0. White arrowheads point at aberrant structures, zoomed within white rectangles. All statistical differences were calculated using Bonferroni multiple comparison correction after two-way ANOVA (n.s.: non-significant and ****: p-value ≤0.0001).
-
Figure 9—source data 1
Raw data for high-content segmentation in Figure 9B.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig9-data1-v2.xlsx
-
Figure 9—source data 2
Raw data for high-content segmentation in Figure 9C.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig9-data2-v2.xlsx
Figure 10 with 1 supplement
TrafE is necessary for proper ESCRT turnover during membrane repair.
(A and B) Live time-lapse spinning disk confocal images with dedicated x100 objective of D. discoideum WT or trafE-KO cells expressing act5::ALIX-GFP or act5::GFP-Vps32, 5 min before (t = –5) and after addition of 4.5 mM LLOMe at t=0. (C and D) Quantification of live time-lapse high-content confocal images with dedicated x40 objective of D. discoideum WT or trafE-KO cells expressing act5::ALIX-GFP or act5::GFP-Vps32, 5 min before (t = –5) and after addition of 4.5 mM LLOMe at t=0, every 5 min for a total of 40 min. Number of GFP-positive structures was measured from three wells with two image fields per well, each image contains 20≤n ≤ 80 cells. The plot is a representation of n=3 technical replicates out of N=3 biological replicates performed. All statistical differences were calculated using Bonferroni multiple comparison correction after two-way ANOVA (**: p-value ≤0.01).
-
Figure 10—source data 1
Raw data for high-content segmentation in Figure 10C.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig10-data1-v2.xlsx
-
Figure 10—source data 2
Raw data for high-content segmentation in Figure 10D.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig10-data2-v2.xlsx
Figure 10—figure supplement 1
ESCRT machinery components redistribution upon starvation.
(A and B) Live spinning-disk confocal microscopy images of D. discoideum WT cells expressing act5::ALIX-GFP or act5::GFP-Vps32 (green) incubated for 12 hr with Alexa Fluor 647 10 kDa dextran (red) and imaged 90 min after a shift to SorMC-Sorbitol starvation medium. (C–G) D. discoideum WT, atg1-KO or trafE-KO cells expressing act5-driven GFP-fusions of ALIX, VPS32 or Vps4 were imaged by high-content confocal microscopy 90 min after a shift to SorMC-Sorbitol starvation media. ALIX, Vps32, or Vps4 structures count per image was collected from 9 to 27 images with 20≤n ≤ 40 cells per image. Bars represent SEM. Statistical differences were calculated using Bonferroni multiple comparison test after two-way ANOVA (n.s.: non-significant, *: p-value ≤0.05, ****: p-value ≤0.0001).
-
Figure 10—figure supplement 1—source data 1
Quantification data for Figure 10—figure supplement 1.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig10-figsupp1-data1-v2.xlsx
-
Figure 10—figure supplement 1—source data 2
Quantification data for Figure 10—figure supplement 1.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig10-figsupp1-data2-v2.xlsx
-
Figure 10—figure supplement 1—source data 3
Quantification data for Figure 10—figure supplement 1.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig10-figsupp1-data3-v2.xlsx
-
Figure 10—figure supplement 1—source data 4
Quantification data for Figure 10—figure supplement 1.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig10-figsupp1-data4-v2.xlsx
Figure 11 with 1 supplement
TrafE regulates Vps4 recruitment to endolysosomal membrane damage sites.
(A and B) Live time-lapse spinning disk confocal images of D. discoideum WT or trafE-KO cells expressing act5::GFP-Vps4, 5 min before (t = –5) and after addition of 4.5 mM LLOMe at t=0. (B) Quantification of live time-lapse high-content confocal images of D. discoideum WT or trafE-KO cells expressing act5::GFP-Vps4, 5 min before (t = –5) and after addition of 4.5 mM LLOMe at t=0, every 5 min for a total of 40 min. Number of GFP-positive structures was measured from three wells with two image fields per well, each image contains 20≤n ≤ 80 cells. GFP-Vps4 puncta tend to be small and difficult to detect, therefore the represented numbers are underestimation. The plot is a representation of n=3 technical replicates out of N=3 biological replicates performed. All statistical differences were calculated using Bonferroni multiple comparison correction after two-way ANOVA (**: p-value ≤0.01). (C) Live time-lapse spinning disk confocal images of D. discoideum complemented trafE-KO cells expressing act5::GFP-Vps4 (green) and dsRed-TrafE (red), 5 min before (t = –5) and after addition of 4.5 mM LLOMe at t=0. (D) Live time-lapse spinning disk confocal images showing (white arrowheads) colocalization of GFP-Vps4 (green) and dsRed-TrafE (red) in LLOMe-treated trafE-KO cells.
-
Figure 11—source data 1
Raw data for high-content segmentation in Figure 11B.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig11-data1-v2.xlsx
Figure 11—figure supplement 1
ImageJ SpotCounter quantification of live time-lapse spinning disk confocal images of D. discoideum complemented trafE-KO cells expressing act5::GFP-Vps4 and dsRed-TrafE from N=3 independent biological replicates.
-
Figure 11—figure supplement 1—source data 1
Quantification data for Figure 11—figure supplement 1.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig11-figsupp1-data1-v2.xlsx
Figure 12 with 1 supplement
In trafE-KO cells, the absence of Vps4 correlates with low K63-polyubiquitination despite the accumulation of ALIX-GFP and GFP-Vps32.
(A, B and C) Immunofluorescence images of mock-treated (mock) or 10 min after addition of LLOMe in WT, trafE-KO or atg1-KO cells expressing act5::ALIX-GFP (green) stained with antibodies specific to K63-polyubiquitin (red). (D and E) Average number of ALIX- or K63-Ub-positive structures per cells compared between mock-treated (mock) or cells treated with LLOMe for 10 min. Quantification of ALIX-GFP or K63-Ub structures from N=4 independent experiments comprised of 5≥n ≥ 45 cells per image was performed using ImageJ ValelabUtils SpotCounter. Statistical differences were calculated using Bonferroni multiple comparison correction after two-way ANOVA (n.s.: non-significant and ****: p-value ≤0.0001).
-
Figure 12—source data 1
Quantification data for number of ALIX-GFP- and K63-Ub-positive structures in Figure 12D and E.
- https://cdn.elifesciences.org/articles/85727/elife-85727-fig12-data1-v2.xlsx
Figure 12—figure supplement 1
ALIX likely comprises a UBD.
Multiple sequence alignment of D. discoideum ALIX with a group of structurally defined but evolutionary distant coiled-coil UBD-containing proteins D. discoideum ALIX (AC: Q8T7K0) comparison with NEMO UBD (Mouse AC: O88522), ALIX (Human AC: Q8WUM4), ALIX (Mouse AC: Q9WU78), ALIX (Rat AC: Q9QZA2), ALIX (African clawed frog AC: Q9W6C5), and ALIX (Giant tiger prawn AC: A5XB12).
Figure 13
Working model.
(A) In WT cells, ALIX, Vps32, and TrafE are recruited to small size damage sites. TrafE-dependent ubiquitination of unknown endolysosomal membrane protein(s) with K63-polyubiquitin promotes K63-Ub-ALIX interaction which enhances ALIX ability to stimulate Vps4 activity, a step necessary for its functional recruitment to damage sites resulting in proper ESCRT-III turnover. In case of substantial damage, it is possible that TrafE-deposited K63-Ub serves as a signal for the recruitment of the autophagy adaptor p62 and the autophagy machinery, leading to formation of a phagophore-like structure which seals the damage site by membrane fusion. (B) In atg1-KO cells, the ESCRT-dependent damage repair is functional; however, the cells are autophagy deficient and therefore there is no formation of phagophore, hence no autophagy-dependent fusion-type membrane repair. (C) In trafE-KO cells, ALIX and Vps32 are recruited to damage sites where the lack of ALIX interaction with TrafE-deposited K63-Ub hinders Vps4 activity stimulation and as a consequence, ESCRT-III turnover and functional ESCRT-dependent repair are impaired. In parallel, the absence of K63-Ub also impairs the autophagy-dependent fusion-type repair. (D) Based on recent studies revealing the function of the ESCRT machinery in phagophore closure, we speculate that TrafE may also be involved in this process.
Author response image 1
Author response image 2
Tables
Appendix 1—key resources table
| Reagent type (species) or resource | Designation | Source or reference | Identifiers | Additional information |
|---|---|---|---|---|
| Gene (Dictiostelium discoideum) | trafA | Dictybase | DDB_G0272454 | |
| Gene (Dictiostelium discoideum) | trafB | Dictybase | DDB_G0285149 | |
| Gene (Dictiostelium discoideum) | trafC | Dictybase | DDB_G0290883 | |
| Gene (Dictiostelium discoideum) | trafD | Dictybase | DDB_G0290961 | |
| Gene (Dictiostelium discoideum) | trafE | Dictybase | DDB_G0290971 | |
| Cell line (Dictiostelium discoideum) | Ax2(Ka) | Dictybase | WT, parental strain of the trafE KO, atg1 KO and GFP-trafE KI | |
| Cell line (M. marinum) | M | L.Ramakrishnan (University of Cambridge) | WT | |
| Cell line (Dictiostelium discoideum) | Ax2(Ka) | This paper | trafE KO | |
| Cell line (Dictiostelium discoideum) | Ax2(Ka) | This paper | atg1 KO | |
| Cell line (Dictiostelium discoideum) | Ax2(Ka) | This paper | GFP-trafE KI | Cell line maintained in T. Soldati lab |
| Cell line (Dictiostelium discoideum) | Ax2(Ka) | This paper | WT; act5::Atg8a | Cell line maintained in T. Soldati lab |
| Cell line (Dictiostelium discoideum) | Ax2(Ka) | This paper | WT; act5::ALIX | Cell line maintained in T. Soldati lab |
| Cell line (Dictiostelium discoideum) | Ax2(Ka) | This paper | WT; act5::Vps32 | Cell line maintained in T. Soldati lab |
| Cell line (Dictiostelium discoideum) | Ax2(Ka) | This paper | WT; act5::Vps4 | Cell line maintained in T. Soldati lab |
| Cell line (Dictiostelium discoideum) | Ax2(Ka) | This paper | trafE KO; act5::Atg8a | Cell line maintained in T. Soldati lab |
| Cell line (Dictiostelium discoideum) | Ax2(Ka) | This paper | trafE KO; act5::ALIX | Cell line maintained in T. Soldati lab |
| Cell line (Dictiostelium discoideum) | Ax2(Ka) | This paper | trafE KO; act5::Vps32 | Cell line maintained in T. Soldati lab |
| Cell line (Dictiostelium discoideum) | Ax2(Ka) | This paper | trafE KO; act5::Vps4 | Cell line maintained in T. Soldati lab |
| Cell line (M. marinum) | ∆RD1 | L.Ramakrishnan (University of Cambridge) Volkman et al., 2004 | RD1 locus ablation mutant | |
| Antibody | anti-GFP (rabbit polyclonal) | MBL | 598 | IF(1:1000), WB (1:1000) |
| Antibody | anti-Ub (FK2) (Mouse monoclonal) | Enzo Life Sciences | BML-PW8810 | IF(1:1000) |
| Antibody | anti-K63-linkage-specific (Mouse monoclonal) | Enzo Life Sciences | HWA4C4 | IF(1:50) |
| Recombinant DNA reagent | pDM317 (plasmid) | Veltman et al., 2009 | Dictybase | GFP (N-terminal on backbone) |
| Recombinant DNA reagent | pDM1513 (plasmid) | Paschke et al., 2018 | 108998 | GFP (N-terminal on backbone) |
| Recombinant DNA reagent | pDM1515 (plasmid) | Paschke et al., 2018 | 109000 | GFP (C-terminal on backbone) |
| Recombinant DNA reagent | pDM318 (plasmid) | Veltman et al., 2009 | Dictybase | dsRed (N-terminal on backbone) |
| Sequence-based reagent | LR32F | This paper | PCR primers; trafE 5’ | CAGGATCCAAAATGACAGTAAAATATTCAATTAATG |
| Sequence-based reagent | LR32R | This paper | PCR primers; trafE 3’ | CAACTAGTTGGTAAAACTTGAATTCTAAG |
| Sequence-based reagent | LR63F | This paper | PCR primers; trafE; qPCR | GAGTCTTGTAAAAAATCATTCCCAAG |
| Sequence-based reagent | LR63R | This paper | PCR primers; trafE; qPCR | GTTGGTTATTTATAACTTTGTCCATC |
| Sequence-based reagent | LR7F | This paper | PCR primers; trafA 5’ | CAGGATCCAAAATGGATATTTCTCAAATCC |
| Sequence-based reagent | LR7R | This paper | PCR primers; trafA 3’ | CAACTAGTATGTTTATCACATTGAGAC |
| Sequence-based reagent | LR8F | This paper | PCR primers; trafB 5’ | CAGGATCCAAAATGACAGAGTTTAAAATTAG |
| Sequence-based reagent | LR8R | This paper | PCR primers; trafB 3’ | CAACTAGTTTTAGTAGTTAAAGGATC |
| Sequence-based reagent | LR9/10 F | This paper | PCR primers; trafC 5’ | CAGGATCCAAAATGTCAATTGATATAAAATTTAC |
| Sequence-based reagent | LR9R | This paper | PCR primers; trafC 3’ | CAACTAGTAGACTCCAATGGTTCATATTC |
| Sequence-based reagent | LR9/10 F | This paper | PCR primers; trafD 5’ | CAGGATCCAAAATGTCAATTGATATAAAATTTAC |
| Sequence-based reagent | LR10R | This paper | PCR primers; trafD 3’ | CAACTAGTAGACTCCAATGGTTCATATTC |
| Commercial kit | RNA extraction kit Direct-zol | Zymo research | R2062 | |
| Other | Lysosensor Green DND-189 | Invitrogen | L7535 | Fluorescent Dye;1 µM |
| Other | 10 kDa Alexa Fluor 647 Dextran | Invitrogen | D22914 | Alexa fluorophore labelled dextran;10 µg/mL |
| Other | DAPI stain | Invitrogen | D1306 | 1:50 |
| Other | PI stain | ThermoFisher | R37108 | 1 µg/mL |
| Chemical compound | LLOMe | Bachem | 16689-14-8 | 4.5–5 mM |
Additional files
-
Supplementary file 1
(a) D. discoideum material used in this study. The Supplementary file includes table with the D. discoideum strains used in this study, the overexpression plasmids and the plasmids used for generation of trafE knock-out and GFP knock-in. (b) M. marinum material used in this study. The Supplementary file includes table with the M. marinum strains and M. marinum plasmids used in this study. (c) Primers used in this study. The Supplementary file includes table with primers used to amplify trafA, trafB, trafC and trafE CDSs, primers used for knock-out, GFP knock-in generation, act5 locus integration and screenings.
- https://cdn.elifesciences.org/articles/85727/elife-85727-supp1-v2.docx
-
MDAR checklist
- https://cdn.elifesciences.org/articles/85727/elife-85727-mdarchecklist1-v2.pdf
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
A TRAF-like E3 ubiquitin ligase TrafE coordinates ESCRT and autophagy in endolysosomal damage response and cell-autonomous immunity to Mycobacterium marinum
eLife 12:e85727.
https://doi.org/10.7554/eLife.85727
