Ultrastructural Characterization of Flashing Mitochondria

Abstract

Mitochondria undergo spontaneous transient elevations in matrix pH associated with drops in mitochondrial membrane potential. These mitopHlashes require a functional respiratory chain and the profusion protein optic atrophy 1, but their mechanistic basis is unclear. To gain insight on the origin of these dynamic events, we resolved the ultrastructure of flashing mitochondria by correlative light and electron microscopy. HeLa cells expressing the matrix-targeted pH probe mitoSypHer were screened for mitopHlashes and fixed immediately after the occurrence of a flashing event. The cells were then processed for imaging by serial block face scanning electron microscopy using a focused ion beam to generate ~1,200 slices of 10 nm thickness from a 28 μm × 15 μm cellular volume. Correlation of live/fixed fluorescence and electron microscopy images allowed the unambiguous identification of flashing and nonflashing mitochondria. Three-dimensional reconstruction and surface mapping revealed that each tomogram contained two flashing mitochondria of unequal sizes, one being much larger than the average mitochondrial volume. Flashing mitochondria were 10-fold larger than silent mitochondria but with a surface to volume ratio and a cristae volume similar to nonflashing mitochondria. Flashing mitochondria were connected by tubular structures, formed more membrane contact sites, and a constriction was observed at a junction between a flashing mitochondrion and a nonflashing mitochondrion. These data indicate that flashing mitochondria are structurally preserved and bioenergetically competent but form numerous membrane contact sites and are connected by tubular structures, consistent with our earlier suggestion that mitopHlashes might be triggered by the opening of fusion pores between contiguous mitochondria.

Keywords: bioenergetics; correlative light and electron microscopy; electron microscopy; fluorescence microscopy; membrane contact sites; mitochondrial dynamics.

Figure 1. Mapping flashing mitochondria.

(a) Spontaneous matrix pH elevations recorded by confocal microscopy in HeLa cells expressing mitoSypHer (λ_ex/λ_em=488/520 nm). Dotted line indicates a cell exhibiting repetitive mitopHlashes. See also Supplementary Movie 1. (b) Time-resolved fluorescence recording of the cell highlighted in panel (a). Fixative agents were added immediately after the second flashing event (arrow). (c) Images acquired before (basal) and during the mitopHlash event (peak) were subtracted (peak–basal) and the active regions (orange, pointed by arrow) superimposed on the image acquired immediately after the addition of fixative agents (fixed). Bar: 10 μm.

Figure 2. Electron tomogram of flashing mitochondria.

(a) Tomogram generated by serial block face FIB-SEM from the active region in Figure 1. The top image of a 600-image stack from a 28 μm × 15 μm cellular volume in the x–z orientation at 5 nm × 5 nm × 10 nm resolution is shown. Full tomogram in Supplementary Movie 2. (b) Tomogram resampled orthogonally in the x–y orientation at 10 nm × 10 nm × 5 nm resolution to match the orientation of the fluorescence image. The top image of a 200-image stack is shown. Full tomogram in Supplementary Movie 3. Bars: 1 μm.

Figure 3. Correlative light and electron microscopy.

(a) Minimum intensity projection of the image stack shown in Figure 2 (b). Mitochondria appear as large interconnected electron-dense structures. (b) Overlay of the fluorescence image of fixed cells on the tomogram projection using MATLAB software. (c) Overlay of the flashing region (red), fluorescence of fixed cells (green), and tomogram projection (gray). Bar: 1 μm. (d) Overlay of a snapshot of the initial, uncurated 3D reconstruction (colored objects) and the flashing region (white).

Figure 4. Morphology of flashing and nonflashing mitochondria.

(a–c) Surface map of all mitochondria identified in three tomograms by the segmentation algorithm. Each color represents a single mitochondrion. The six flashing mitochondria, numbered sequentially from Cell 1 to Cell 3 are indicated by arrows. (d) Quantification of the volume of individual mitochondria (left) and of the average volume (center) and surface to volume ratio (right) of the different classes of mitochondria. Values in center and right panels are means ± standard deviation of the average values from the three tomograms comprising a total of 6 and 156 flashing and silent mitochondria, respectively, *p = .02, Welsh’s t test.

Figure 5. Ultrastructural features of the two flashing mitochondria from Cell 1.

(a) Surface map of the two flashing mitochondria from Cell 1. Flashing mitochondria are colored red (1) and pink (2), a juxtaposed nonflashing mitochondrion is colored green (silent). Dotted lines outline the tomograms sections shown in Supplementary Movies 4 and 5. (b) Tomogram slice illustrating a tubular structure (arrow) linking two parts of the Flashing Mitochondrion 1 from Cell 1, colored in red. Full tomogram in Supplementary Movie 4. (c) Tomogram slice illustrating a contact site (arrow) between Flashing Mitochondrion 2 from Cell 1 and its juxtaposed nonflashing mitochondrion. Full tomogram in Supplementary Movie 5. Bars: 200 nm.

Figure 6. Ultrastructural features of the two flashing mitochondria from Cell 2.

(a) Surface map of the two flashing mitochondria from Cell 2. Flashing mitochondria are colored red (3) and pink (4). Dotted lines outline the tomogram section shown in Supplementary Movie 6. (b) Tomogram slice illustrating a tubular structure (arrow) at the interface between Flashing Mitochondria 3 and 4 from Cell 2. Full tomogram in Supplementary Movie 6. Bar: 200 nm.

Figure 7. Ultrastructural features of the two flashing mitochondria from Cell 3.

(a) Surface map of the two flashing mitochondria from Cell 3. Flashing mitochondria are colored pink (5) and red (6), a juxtaposed nonflashing mitochondrion is colored green (silent). Dotted lines outline the tomogram section shown in Supplementary Movie 7. (b) Tomogram slice illustrating a junction (arrow) between Flashing Mitochondrion 6 from Cell 3 and a nonflashing mitochondrion. Note the fusion of the mitochondrial membranes. Full tomogram in Supplementary Movie 7. Bar: 200 nm.

Figure 8. Cristae of flashing and nonflashing mitochondria.

(a–c) EM pictures showing cristae segmentation within individual mitochondria. Each color corresponds to a single mitochondrion. (d) Quantification of the volume of cristae normalized to the total volume of each individual mitochondrion. The four largest silent mitochondria from each tomogram (n = 12) were quantified. (e) Quantification of the electron density, measured as IIUs. The left panel shows all flashing mitochondria (n = 6) compared with the four largest silent mitochondria from each tomogram (n = 12). The middle panel compared the average of all flashing mitochondria within a cell (n = 2/cell) to the average of all silent mitochondria within the same cell (n = 58/53/43 for cells 1/2/3). The right panel compares each flashing mitochondrion within a cell (n = 2/cell) to the silent mitochondrion within the same cells most closely matching it in volume (n = 2/cell). IIU = inverse intensity unit.

Figure 9. Membrane contacts within Flashing Mitochondrion 1 from Cell 1.

(a and b) EM pictures showing membrane contact sites (arrows) within Flashing Mitochondrion 1 from Cell 1, colored in blue. The contour of the mitochondrion was delineated using an internal morphological gradient (of three pixels) to define a homogenous thickness of membrane for all contact sites. (c) Surface mapping of Flashing Mitochondrion 1 with its internal membrane contact sites highlighted in light blue (arrows). See also Supplementary Movie 8. (d) Quantification of the fraction of mitochondrial membrane engaged in internal contact sites. The four largest silent mitochondria from each tomogram (n = 12) were analyzed. See also Tables 1 and 2 and Supplementary Movie 10.