A mitochondrial quality control mechanism reverses the phagosome maturation arrest caused by Mycobacterium tuberculosis

Cellular Immunology Group, International Centre for Genetic Engineering and Biotechnology, New Delhi, India
Department of Microbiology and Cell Biology (MCBL), Centre for Infectious Disease and Research (CIDR), Indian Institute of Science (IISc), Bangalore, India
A*STAR Infectious Diseases Labs (A*STAR ID labs), Agency for Science, Technology and Research (A*STAR), Singapore, Singapore
Singapore Immunology Network (SIgN), A*STAR, Singapore, Singapore
Cellular Immunology, International Centre for Genetic Engineering and Biotechnology, ICGEB, Trieste, Italy
Signal Transduction Lab, National Institute of Immunology, New Delhi, India
Department of Molecular and Cellular Medicine, Institute of Liver and Biliary Sciences, New Delhi, India
Department of Immunohematology and Transfusion Medicine, Institute of Liver and Biliary Sciences, New Delhi, India
Institute of Cancer and Genomic Sciences, Institute of Biomedical Research, College of Medical and Dental Sciences, University of Birmingham, Birmingham, United Kingdom
Centre for Cellular and Molecular Biology, Hyderabad, India
Lee Kong Chian School of Medicine, Nanyang Technological University, Singapore, Singapore
Department of Medicine, University of Massachusetts Medical School, Worcester, United States

https://doi.org/10.7554/eLife.112033.1

Open access
Copyright information

Figures and data

Selective recruitment of autophagy adaptors to Mtb phagosomes and mitochondria:
(A-E) Confocal images show the colocalisation of individually immunostained autophagy adaptors as indicated (green) with mCherry-H37Rv (red) respectively at indicated time points post-infection in THP-1 macrophages. The area in the box is zoomed in and shown across the individual colour channels at the right. The bar graphs at the right represent the percent co-localization calculated manually using Imaris software. Scale bar: 10 μm. Data show mean ± SEM, n >10 fields per group, 20-35 cells per field, from three independent experiments. (F) Representative images of the interaction between immunostained p62/SQSTM1 (green) and mitochondrial marker TOM20 (Red). Nucleus were stained with DAPI (blue) at 24 hpi in THP-1 macrophages. To the right, representative 3D reconstruction is shown for TOM20 structures and p62/SQSTM1 spots. Magenta spots: non-interacting, >0.2 μm from TOM20, blue spots: interacting, <0.2 μm from the TOM20 structure. The graph represents the percentage of indicated adaptor spots interacting with TOM20 structures, calculated by the shortest distance module of Imaris software. Scale bar: 10 μm. Data show mean ± SD, n=15 fields per group, 20-35 cells per field, from three independent experiments. Scale Bar: 10 μm.

Autophagy adaptors regulate intracellular Mtb survival:
(A) The bar graph represents the H37Rv CFU upon individual autophagy adaptor depletion using siRNAs in THP-1 macrophages 24 hpi. Data show mean ± SD, from four independent experiments. (B) Time course assay of H73Rv CFU in WT and p62^KD THP-1 macrophages. Data show mean ± SD from four independent experiments. (C) WT and p62^KD cells were infected with pBP10⁺-H37Rv, and CFU plating was done at designated time points on kan⁺ or Kan^- 7H11 plates. The total bacterial burden (i.e., CFU on Kan^- plates, left panel) and %pBP10⁺Mtb (ratio of CFU on Kan⁺ to Kan^- plates, right panel) for the WT (left) and p62^KD (right) cells are shown. Data is mean± SD, representative from two independent experiments. (D) The bar graph represents H37Rv CFU in hMDMs in control and p62 knockdown conditions at 24 hpi. Data show mean ± SD; individual data points for three different donors are shown. (E-G) Heatmap represents the mean z-scaled expression of mitochondrial-related DEGs (differentially expressed genes, absolute log2FC>=1 & q value <= 0.1) between infected versus uninfected in WT and p62^KD macrophages at 0-, 24- and 48-.hpi (H-I) Pie charts showing functional classes of mitochondrial-associated DEGs at different time points in infected WT and p62^KD macrophages.

Mitochondrial quality is maintained during H37Rv infection in the absence of p62/SQSTM1:
(A-C) The bar graphs represent the MFI (% of control) of (A) MTG (MitoTracker Green), (B) TMRE (Tetramethylrhodamine, ethyl ester), and (C) MitoSOX, via flow cytometry at 0-, 24- and 48-hpi in H37Rv-infected and uninfected WT and p62^KD THP-1 macrophages. Data show mean ± SD, three independent experiments. (D) Mitostress (Seahorse extracellular flux assay) test in H37Rv-infected WT and p62^KD macrophages at 24 hpi. Data show mean ± SD, three independent experiments. (E) DCFDA MFI (% of control) in WT and p62^KD THP-1 macrophages at 0-, 24- and 48-hpi. Data show mean ± SD, three independent experiments. (F) IF analysis of CellROX staining (magenta) in WT and p62^KD THP-1 macrophages at 24 hpi. The bar graph represents CellROX MFI. Data show mean ± SD, n=5 from two independent experiments. (G) The ratio of emission MFI of Mrx1-roGFP reporter at 510nm as excited at 405nm to 488nm via flow cytometry in WT and p62^KD cells infected with the Mrx1-roGFP2 expressing H37Rv at 24 hpi. CHP and DTT were used as oxidising and reducing controls, respectively. Data show mean ± SD, three independent experiments. (H-I) Confocal microscopy images of H37Rv::Mrx1-roGFP2 in WT and p62^KDcells, at 24 hpi, when excited at 405nm and 488nm. The bar graph depicts the ratio of emission MFIs at 510nm obtained upon excitation at 405nm to that at 488nm. Data show mean ± SD, n>20 fields, 30-50 bacteria per field, from two independent experiments. (J) Time course CFU analysis of H37RvΔaosR and complemented strain H37RvΔaosR::WT in WT and p62^KDcells. Data show mean± SD, representative of two independent experiments (K) CFU analysis of H37Rv in the presence of N-acetyl cysteine (NAC, 2mM for 24 hours) in p62^KD macrophages at 24 hpi. Data show mean ± SD, three independent experiments. (L) Dot plot depicts the ratio of emission MFI of Mrx1-roGFP2::H37Rv at 510nm as excited at 405nm to 488nm via confocal microscopy in NAC treated and untreated WT and p62^KD at 24 hpi. Data show mean ± SD, n>=6 fields, 30-50 bacteria per field, from two independent experiments.

Mitochondrial Quality Control in p62/SQSTM1 KD cells:
(A) Immunofluorescence images for mCherry-H37Rv (Red) and TOM20 (green) interaction in WT and p62^KDTHP-1 macrophages at 24 hpi. The 3D reconstruction of H37Rv structures and TOM20 spots is shown at the right; the bar graph shows the average number of yellow TOM20 spots/bacteria. Data show mean ± SD, n> 10 fields, 20-35 cells per field, from three independent experiments, (B-C) Immunofluorescence images of TOM20 (green) and Parkin (red, B) or Ubiquitin (red, C) respectively. Representative TOM20 3D structures and Parkin or Ubiquitin spots are shown on the right. The bar graph shows per cent of Ubiquitin (B) or Parkin (C) spots interacting with the TOM20 structures. Data show mean± SD, n>6 fields, 20-35 cells per field, from two independent experiments. (D-E) Immunofluorescence images of Ubiquitin (D) and Parkin (E) respectively with mCherry-H37Rv, and 3D reconstructed images of H37Rv structures and Ubiquitin (D) or Parkin (E) spots. The bar graph shows per cent of spots <0.2 μm from Mtb. Data show mean ± SD, n≥5 fields, >20-35 cells per field, from two independent experiments. (F) Mitophagy flux assessed by the interaction of LC3B spots and TOM20 structures in the presence and absence of bafilomycin A1 (BafA₁, 100nM, 3 hours) in WT and p62^KD THP1 macrophages at 24 hpi. Data show mean ± SD, n≥6 fields, 20-35 cells per field, from two independent experiments. (G-H) Immunofluorescence images on the left show the interaction of BNIP3 spots with TOM20 (G) or H37Rv (H) structures, respectively. The bar graphs depict the number of BNIP3 spots <0.2 μm from TOM20 (G) or H37Rv (H) structures. Data show mean ± SD, n=5 fields, 20-35 cells per field, from two independent experiments. For all 3D reconstructions, yellow spots <0.2 μm from TOM20 or Mtb structures, magenta spots >0.2 μm from TOM20 or Mtb structures; Scale bar: 10 μm.

Enhanced TOM20⁺-Mitochondria-derived vesicles (MDVs) biogenesis in p62^KD cells:
(A) High-resolution images of immunostained TOM20 (green) in H37Rv-infected WT and p62^KDmacrophages at 24 hpi. The zoomed images on the right show encircled TOM20⁺-MDVs in the p62^KD cells. (B) High-resolution images of immunostained TOM20 (red) and MCU (green). The encircled area depicts TOM20⁺MDVs in the zoomed images at the right. (C) Quantification of number of TOM20⁺MDVs in uninfected and Mtb-infected WT and p62^KD cells. Data show mean ± SD, n=6 fields having 5-10 cells per field, from two independent experiments (D) Schematic depiction of the presence of different proteins on mitochondrial matrix, inner and outer membrane. (E) Bar graphs show quantitative interactions of matrix mitochondrial protein spots of PDHA1, HSP60 and SOD2 with TOM20 structures. (F) Bar graphs show the quantitative association of inner membrane mitochondrial protein spots of ECSIT and MCU with TOM20 structures. (G-K) Immunofluorescent images depict the interaction of mCherryH37Rv (red) with ECSIT (G), MCU (H), PDHA1 (I), SOD2 (J) and HSP60 (K) (green), respectively, in WT and p62^KD THP-1 macrophages at 24 hpi. The 3D reconstruction shows H37Rv structures and indicates mitochondrial protein spots. The yellow spots are <0.2 μm from H37Rv structures, and the magenta spots are >0.2 μm from bacterial structures. The bar graph on the right shows a number of mitochondrial protein spots on Mtb. Data show (E-K) mean ± SD, n≥5 fields having 20-35 cells per field, from two independent experiments, Scale bar =10 μm unless specified.

TOM20⁺-MDVs overcome phagosome maturation arrest imposed by Mycobacterium tuberculosis:
(A-C) Confocal images show the colocalisation of mCherry H37Rv (red) with RAB7, Lysotracker and Cathepsin D respectively (green) in WT and p62^KDTHP-1 macrophages at 24 hpi. The bar graph shows per cent colocalisation, calculated manually using Imaris software. Data show mean ± SD, n=5 fields, 20-35 cells per field, from two independent experiments. (D) The image shows the interaction of LAMP1 spots with TOM20 structures. The bar graph depicts the number of LAMP1 spots <0.2 um (yellow) from the bacterial structure. Data show means ± SD, n=5 fields, 20-35 cells per field, from two independent experiments. (E) Images show the interaction of LAMP1 spots with TOM20 structures in Mtb-infected WT and p62^KD cells. The plot shows LAMP1 spot quantification on TOM20 structure in Mtb-infected WT and p62^KD cells. Data show mean ± SD, n> 5 fields, 20-35 cells per field, from two independent experiments (F) H37Rv CFU in WT and p62^KD cells upon RAB7 knockdown at 24 hpi. Data show means ± SD, n=5. (G-H) Quantifying LAMP1 (G) and TOM20 (H) spots on H37Rv in RAB7 knockdown p62^KD cells at 24 hpi. Data show means ± SD, n=5 fields, > 20 cells per field, from two independent experiments. (I) The relative oxidation state of Mrx1-roGFP2 reporter as the ratios of emission MFI at 510nm when excited at 488nm to that of 405nm in p62^KD and RAB7 depleted p62^KD cells. Data show means ± SD, n=5 Fields, > 40 bacteria per field, from two independent experiments. Scale bar: 10 μm

Triggering loss of MQC in p62^KD cells aggravates bacterial killing
(A) TMRE per cent MFI in PRKN knockdown cells in WT and p62^KDTHP-1 macrophages. Data show mean ± SD from three independent experiments (B) The confocal images depict the circular depolarised mitochondria (immunostained with TOM20, green) in PRKN-depleted p62^KDcells. (C) MFI ratio of Mrx1-roGFP2::H37Rv at 405nm to 488nm in p62^KD and Parkin-depleted p62^KD cells. Data show mean ± SD, n=5 Fields, > 40 bacteria per field, from two independent experiments. (D) H37Rv CFU in WT and p62^KD THP-1 macrophages upon PRKN depletion at 24 hpi. Data show mean ± SD from three independent experiments. (E-F) TOM20 spots on H37Rv structures in p62^KD cells upon PRKN knockdown. Data show means ± SD, n=7 fields, each having 20-35 cells, from two independent experiments. (G) LAMP1 spots on H37Rv structures in p62^KD cells upon PRKN knockdown. Data show means ± SD, n=6 fields, 20-35 cells per field, from two independent experiments. (H) H37Rv CFU in siCtrl, siMIRO1 and siMIRO2 treated p62^KD THP-1 macrophages at 24 hpi. Data show mean ± SD from two independent experiments. (I) High-resolution images of immunostained TOM20 (green) in siCtrl, siMIRO1 and siMIRO2 treated p62^KD THP-1 macrophages at 24 hpi. The plot depicts the quantification of TOM20⁺ MDV in these cells. Data show means ± SD, n≥7 fields, 5-10 cells per field, from two independent experiments. (J) The images show the interaction of TOM20 spots with H37Rv structures in siCtrl, siMIRO1 and siMIRO2 treated p62^KD THP-1 macrophages at 24 hpi. The bar graph depicts the number of TOM20 spots <0.2 μm (blue) from the bacterial structure. Data show means ± SD, n=6 fields, 20-35 cells per field, from two independent experiments. (K) The image shows the interaction of RAB7 spots with H37Rv structures in siCtrl, siMIRO1 and siMIRO2 treated p62^KDTHP-1 macrophages at 24 hpi. The bar graph depicts the number of RAB7 spots <0.2 μm (blue) from the bacterial structure. Data show means ± SD, n=5 fields, 20-35 cells per field, from two independent experiments. Scale bar: 10 μm

Crosstalk between mitochondrial quality control mechanisms and lysosomal targeting of Mtb in macrophages.
Mitochondrial depolarisation triggers the mitophagy process by recruiting autophagy adaptor p62/SQSTM1 on mitochondria. p62/SQSTM1 binds the LC3B to form the mitophagosome, which fuses with lysosomes to clear the damaged mitochondria. In the p62^KD cells, the loss of p62/SQSTM1 impairs the mitophagy process and triggers an alternative mitochondrial quality control pathway, i.e., the biogenesis of MDVs. MIRO1 and MIRO2 drive the formation of these vesicles from the outer mitochondrial membrane enriched with TOM20 (TOM20+-MDVs). They bud off from the damaged areas of mitochondria to fuse with lysosomes in a RAB7-dependent process. In WT macrophages, virulent strains of Mtb inhibit the RAB5 to RAB7 conversion, thereby preventing the fusion of early bacterial phagosomes/autophagosomes with the lysosomes, a process known as phagosome maturation arrest. In the p62^KD condition, the TOM20⁺-MDVs get recruited on Mtb phagosomes/autophagosomes that inadvertently also help recruitment of RAB7 to the Mtb-phagosomes, thereby overcoming the phagosome maturation arrest. RAB7 recruitment to the phagosomes enhances the lysosomal targeting and Mtb killing in these cells. Both MDVs and damaged mitochondria contribute to the redox-mediated killing of Mtb and complement the lysosomal killing of Mtb.

Sign up for email alerts