ER-mitochondria contacts promote mtDNA nucleoids active transportation via mitochondrial dynamic tubulation

Abstract

A human cell contains hundreds to thousands of mitochondrial DNA (mtDNA) packaged into nucleoids. Currently, the segregation and allocation of nucleoids are thought to be passively determined by mitochondrial fusion and division. Here we provide evidence, using live-cell super-resolution imaging, that nucleoids can be actively transported via KIF5B-driven mitochondrial dynamic tubulation (MDT) activities that predominantly occur at the ER-mitochondria contact sites (EMCS). We further demonstrate that a mitochondrial inner membrane protein complex MICOS links nucleoids to Miro1, a KIF5B receptor on mitochondria, at the EMCS. We show that such active transportation is a mechanism essential for the proper distribution of nucleoids in the peripheral zone of the cell. Together, our work identifies an active transportation mechanism of nucleoids, with EMCS serving as a key platform for the interplay of nucleoids, MICOS, Miro1, and KIF5B to coordinate nucleoids segregation and transportation.

Fig. 1. Mitochondrial DNA nucleoids are transported via MDT.

a Representative time-lapse images of MDT in a Cos-7 cell expressing Tom20-GFP and TFAM-mCherry demonstrate the transportation of TFAM-labeled nucleoids spatially linked to the MDT process. White arrowheads mark the tips of tubules generated by the MDT processes. Orange arrowheads indicate the sites of nucleoids. Additional examples are shown in Supplementary Fig. 2a–c. b Percentage of nucleoids near the MDT initiation sites that get transported into the thin tubule (blue box) or remain at the MDT initiation sites (orange box) for over 2 min, n = 42 events. c The trajectory and velocity of the nucleoid and that of the dynamic tubule during the tubulation process in a. d Spatial cross-correlation analysis of the tubulation process (Tom20-GFP) and nucleoid (TFAM-mCherry) trajectories, n = 17 events examined over three independent experiments. Data are shown as mean ± SEM. e Percentage of nucleoids in dynamic tubules transported to the tip of the tubules (orange box) or not the tip of the tubules (blue box) during the MDT process, n = 17 events. f Representative time-lapse images of MDT in a Cos-7 cell expressing Tom20-GFP and TFAM-mCherry demonstrate unsynchronized motility of TFAM-labeled nucleoids and the MDT process. White arrowheads mark the tips of tubules generated by the MDT processes. Orange arrowheads indicate the sites of nucleoids. g The trajectory and velocity of the mitochondrial DNA nucleoid and that of the dynamic tubule during the tubulation process in f. Scale bars: a, 1 μm and f, 2 μm. Source data are provided as a Source data file.

Fig. 2. ER tubules mark the initiation sites of MDT.

a, b Examples of MDT at ER–mitochondria contacts. Left-hand images show Cos-7 cells expressing a mEmerald-Sec61β and mito-DsRed, and b mCherry-KDEL and Tom20-GFP, merged as indicated. Right graphs are linescans drawn through the mitochondria (dashed box), and show the relative fluorescence intensity of mitochondria (green) and ER (purple) along their lengths. White arrowheads at constrictions in the images correspond to the black arrows shown on the corresponding linescan graphs. Scale bars: 2 μm. Additional examples are shown in Supplementary Fig. 5a–d.

Fig. 3. Miro1 is enriched at EMCS and spatially associated with mtDNA.

a A multicolor SIM image of a Cos-7 cell expressing mEmerald-Sec61β and mito-DsRed, with Miro1 labeled with α-Miro1 antibody. b Higher magnifications of the boxes in a showing three examples of Miro1 enriched at EMCS. White arrowheads indicate Miro1 punctae. Graphs to the right show relative pixel intensities of mito-DsRed, anti-Miro1, and mEmerald-Sec61β from a linescan drawn on each corresponding image (dashed line). The white arrowheads correspond to the black arrows on the linescans. c The percentage of Miro1 punctae that colocalized with the ER–mitochondria contacts in fixed Cos-7 cells. Data are shown as mean ± SEM (n = 10 cells examined over three independent experiments). d Cos-7 cells were transfected with mito-YFP, and immunostained by anti-Miro1 and anti-mtDNA. Arrowheads show mtDNA foci with adjacent Miro1 foci. e Spatial cross-correlation analysis of mtDNA and Miro1 fluorescence intensity along linescans of mitochondria (n = 91 mtDNA from 17 images) in fixed Cos-7 cells. Blue dotted lines indicate 95% confidence interval cutoffs, corresponding to the position of 224 nm. f Percentage of mtDNA that colocalized with Miro1 in Cos-7 cells (distance < 224 nm), corresponding to the 95% confidence interval marked by the blue dotted lines in e. Significantly more mtDNA is spatially (74%) associated with Miro1 than that expected by a randomly scrambled distribution (19%; n = 91 mtDNA from 17 images examined over three independent experiments, P < 0.0001, ***P < 0.001, two-tailed, unpaired Student’s t-test). Scale bars: a, 5 µm; b and d, 2 µm. Source data are provided as a Source data file.

Fig. 4. Mic60 links mtDNA to Miro1 and contributes to EMCS stability.

a Co-immunoprecipitation of Mic60 with endogenous TFAM and Miro1 in HEK293T cells, and the immunoprecipitates were blotted as indicated. Input: 2%. b Cos-7 cells immunostained by anti-mtDNA and anti-Mic60. Arrowheads show mtDNA foci with adjacent Mic60 foci. (Right) the graph is linescan drawn through the mitochondria, and show the relative fluorescence intensity of mtDNA (green) and Mic60 (purple) along their length. White arrow positions correspond to black arrows on the linescan. c As in b for cells immunostained by anti-Miro1 (green on linescan) and anti-Mic60 (purple on linescan). d Dual-color confocal image of a Cos-7 cell expressing mito-DsRed and mEmerald-Sec61β (left), and the corresponding mitochondrial perimeter profile (right). Mitochondrial perimeter profiles are presented by red pixels, and the contacts between the mitochondrial perimeter and the ER are presented by yellow pixels. e ER–mitochondria colocalization index (R), calculated as the ratio between the perimeter pixels of mitochondria colocalized with the ER and the total perimeter pixels of mitochondria for the control and Mic60 RNAi cells (n = 10 cells were used per condition from three independent experiments). f Fluorescence image (modified as described in “Methods” section) of a Mic60 RNAi cell expressing mito-DsRed and mEmerald-Sec61β. The boxed region is enlarged and shows time-lapse images. White arrowheads indicate unstable ER–mitochondria interactions. g Representative kinetics of the ratio (R) on the entire Cos-7 cell shown in f. ΔR represents the difference between the R of one image (frame) and that of the previous one. h Mean ΔR/R0 values (where R0 is the R of the first frame of each pair; n = 110 frames from ten control and ten Mic60 RNAi cells examined over three independent experiments). Scale bars: b, c, inset in d, and inset in f, 2 μm; d and f, 10 μm. Data are presented as mean ± SEM. P value from top to bottom: p = 0.4981, p = 0.0089. NS no significant difference; **p < 0.01, two-tailed, unpaired Student’s t-test. Source data are provided as a Source data file.

Fig. 5. MDT contributes to nucleoids distribution in peripheral mitochondrial network.

a Confocal images of KIF5B^−/− NRK cells expressing mito-YFP, TFAM-mCherry, and TET-ON-KIF5B and treated with 0.5 µg/ml tetracycline. Higher magnifications of the white dashed boxes are in the lower panels. Red dashed lines indicate the mitochondrial network boundary before tubulation. b–d KIF5B^−/− NRK cells expressing mito-YFP, TFAM-mCherry, and TET-ON-KIF5B were treated with 0.5 µg/ml tetracycline for 1 h (b, c) and 2 h (d). White arrows indicate a dynamic tubule. Orange arrowheads indicate nucleoids. e Kymographs of mtDNA labeled by DNA dye picogreen in a control cell and a Mic60 RNAi cell. f Time-lapse images of Mic60 RNAi cells expressing mito-DsRed and stained by picogreen. White arrowheads indicate MDT processes. Orange arrowheads mark the sites of mtDNA. g Percentage of MDT events that drive mtDNA transportation to total MDT events in control and in Mic60 RNAi cells over a 5-min course (n = 10 cells were used per condition from three independent experiments). h KIF5B^−/− NRK cells expressing mito-YFP, TFAM-mCherry, and TET-ON-KIF5B contain control or Mic60 siRNA. Both control and Mic60 RNAi cells were treated with 0.5 µg/ml tetracycline for 4 h. Yellow dashed lines indicate the boundary of the mitochondrial network. Red dashed lines indicate the boundary of the perinuclear mitochondria. Inset regions are magnified from the boxed areas. i Western blots of Mic60 indicate depletion of Mic60 in lysates of cells transfected with siRNA against Mic60. j Quantification of nucleoids density in the perinuclear zone and peripheral zone of control or Mic60 RNAi cells after adding 0.5 µg/ml tetracycline for 4 h (n = 20 cells were used per condition from three independent experiments). Scale bars: a, h, 10 μm; inset in a, 5 μm; b–d, f and inset in h, 2 μm; e, 1 μm. Data are presented as mean ± SEM. P value from top to bottom and left to right: p < 0.0001, p = 0.7101, p = 0.0006. NS no significant difference; ***p < 0.001, two-tailed, unpaired Student’s t-test. Source data are provided as a Source data file.

Fig. 6. The role of MDT in the partitioning of mitochondrial DNA nucleoids.

At the ER–mitochondria contact sites, nucleoids interact with the inner membrane protein complex MICOS. MICOS is associated with the outer membrane protein Miro1, which is concentrated at the ER–mitochondria contact sites and recruits KIF5B motors on the mitochondria. When KIF5B-driven dynamic tubulation happens, nucleoids are transported via the interaction axis of nucleoid/MICOS/Miro1/KIF5B.