Mitochondrial fission and fusion factors reciprocally orchestrate mitophagic culling in mouse hearts and cultured fibroblasts

Abstract

How mitochondrial dynamism (fission and fusion) affects mitochondrial quality control is unclear. We uncovered distinct effects on mitophagy of inhibiting Drp1-mediated mitochondrial fission versus mitofusin-mediated mitochondrial fusion. Conditional cardiomyocyte-specific Drp1 ablation evoked mitochondrial enlargement, lethal dilated cardiomyopathy, and cardiomyocyte necrosis. Conditionally ablating cardiomyocyte mitofusins (Mfn) caused mitochondrial fragmentation with eccentric remodeling and no cardiomyocyte dropout. Parallel studies in cultured murine embryonic fibroblasts (MEFs) and in vivo mouse hearts revealed that Mfn1/Mfn2 deletion provoked accumulation of defective mitochondria exhibiting an unfolded protein response, without appropriately increasing mitophagy. Conversely, interrupting mitochondrial fission by Drp1 ablation increased mitophagy and caused a generalized loss of mitochondria. Mitochondrial permeability transition pore (MPTP) opening in Drp1 null mitochondria was associated with mitophagy in MEFs and contributed to cardiomyocyte necrosis and dilated cardiomyopathy in mice. Drp1, MPTP, and cardiomyocyte mitophagy are functionally integrated. Mitochondrial fission and fusion have opposing roles during in vivo cardiac mitochondrial quality control.

Figure 1. Conditional deletion of Drp1 or Mfn1/Mfn2 in adult mice evokes different cardiomyopathies

A. Schematic of tamoxifen-inducible myh6-Cre mediated Drp1 or Mfn1/Mfn2 gene deletion. B-C. Immunobloting of the targeted factors at 3 and 6 weeks after cardiac Drp1 (B) or Mfn1/Mfn2 (C) gene ablation. D–E. Cardiac phenotypes induced by Drp1 (D) or Mfn1/Mfn2 (E) gene deletion. Upper left: Four-chamber view of hearts. Upper right: Ratio of left ventricular (LV) end-diastolic radius (r) to wall thickness (h). Lower left: Serial M-mode echocardiographic analyses. Lower right: Quantitative group data showing % change of LV end-diastolic dimension (EDD) and fractional shortening (FS) (Raw data in Figure S2). Mean ± SEM of numbers of mice indicated in the upper right panel of Fig 1D and 1E, respectively; * = P<0.05 vs identically treated controls. See also Figures S1 and S2.

Figure 2. Inhibition of mitochondrial fission, but not fusion, in adult hearts provokes cardiomyocyte necrosis

A–B. Mitochondria dysmorphometry evoked by Drp1 (A) or Mfn1/Mfn2 (B) deficiency. Upper: Transmission electron microscopic (TEM) images (5,000x); representative cardiomyocyte mitochondria are outlined in white. Lower left: Mitochondrial area. Lower middle and right: Flow cytometry of mitochondrial size (forward scatter). Representative size distribution curves (black line) are compared to paired controls (grey solid); quantitative data are to the right. C–D. Masson’s trichrome-stained (200x original magnification) Drp1 KO (C) or Mfn DKO (D) hearts. Quantification of blue-staining replacement fibrosis is to the right. E–F. Fluorescence microscopy of Evans blue-perfused Drp1 KO (E) or Mfn DKO (F) hearts. Evans blue dye fluoresces red; green is wheat germ agglutinin staining of cell membranes. Quantitative group data for Evans blue positive cells (%) and cardiomyocyte cross-sectional area (CSA) are to the right. Data are mean ± SEM on indicated numbers of mice; * = P<0.05 vs identically treated controls. See also Figure S3.

Figure 3. Induction of mitophagy in Drp1 null hearts and the mitochondrial unfolded protein response (UPR^mt) in Mfn1/Mfn2 DKO hearts

A. Immunoblot analysis of autophagy markers p62 and LC3-II in mitochondrial fractions (top) and the mitochondrial stress marker Fgf21 in cardiac homogenates (bottom). B. Immunoblot analysis of UPR^mt markers LONP1, AFG3L2, and Hsp60 in cardiac homogenates. Each lane is a different mouse heart. p62 is Sequestosome1; LC3-II is microtubule associated protein 1 light chain 3, processed form; Fgf21 is fibroblast growth factor 21; LONP1 is Lon peptidase 1; AFG3L2 is ATPase family member 3-like 2; Hsp60 is heat shock protein chaperone 60; GAPDH is loading control. Quantitative data comparing Drp1 KO to Mfn1/Mfn2 DKO are at the bottom; black columns are 6 weeks, and grey columns 3 weeks, after tamoxifen. Horizontal dashed line represents normal value from Cre-negative controls. Data are mean ± SEM of 4 hearts per group; # = P<0.05 vs identically treated Drp1 KO hearts. See also Figure S4.

Figure 4. Conditional ablation of Drp1 in cultured MEFs recapitulates mitochondrial phenotypes from conditional Drp1 KO hearts

A–B. Schematic depiction of time course experimental design in conditional Drp1 KO MEFs (Drp1^fl/fl MEFs treated with adeno-Cre) and Drp1 KO adult mice, respectively. C–D. Mitochondria of MEFs co-stained with MitoTracker Green and TMRE (C) and hearts by TEM (10,000x) (D). Red TMRE stains fully polarized mitochondria that appear yellow in merged images; inset is quantitative group data. E. Quantitative mitochondrial aspect ratio (left), mitochondrial content (middle), and cell membrane permeability (right) in Drp1 KO MEFs. F. Quantitative data for mitochondrial aspect ratio (left) and mitochondrial content (right) in Drp1 KO hearts based on TEM images. G. Flow cytometry of isolated cardiac mitochondrial transmembrane electrical potential measured with DiOC6; inset is group data. H. Cardiac mitochondrial respiration; inset is quantitative data for state 3 (ADP stimulated)/state 2. I. Mitochondrial ROS production: Left, O₂⁻ measured with MitoSOX Red; Right, H₂O₂ measured with Amplex Red. Figure 4G-4I: white bar (black line) is ctrl, blue is 3 wks Drp1 KO, and red is 6 wks Drp1 KO. Data are mean ± SEM of indicated number of experiments; * = P<0.05 vs Cre negative controls. See also Figure S5.

Figure 5. Conditional ablation of Mfn1/Mfn2 in cultured MEFs recapitulates mitochondrial phenotypes from Mfn DKO heart

A–F, exactly as described for Figure 4 except using Mfn1/Mfn2 floxed allele MEFs and mice. G–I: white bar (black line) is ctrl, blue is 3 wks Mfn DKO (16,000g fragmented mitochondria) or 6 wks Mfn DKO (8,000g normal mitochondria), and red is 6 wks Mfn DKO (16,000g fragmented mitochondria). See also Figure S5.

Figure 6. Conditional ablation of Drp1 in MEFs increases MPTP-dependent mitochondrial autophagy

A–B. Parkin aggregation in conditional Drp1 KO (A) or Mfn DKO (B) MEFs infected with adeno-mcherry-parkin (mcParkin) before and 6 days after adeno-Cre mediated gene ablation. Quantitative group data are to the right. C–D. Confocal co-localization of lysosomes (LysoTracker Red) and mitochondria (MitoTracker Green) in conditional Drp1 KO (C) or Mfn DKO (D) MEFs before and 6 days after adeno-Cre mediated gene ablation. Mitochondria engulfed in lysosomes appear yellow (arrows). E–F. Parkin aggregation (E) and co-localization of lysosome and mitochondria (F) in conditional Drp1 KO MEFs in the absence or presence of 2 μM cyclosporinA (CsA) at 2 days after adeno-Cre. Quantitative data are to the right. G–H. Group data for mitochondria aspect ratio (G) and content (H) of Drp1 KO MEFs in the absence or presence of CsA. I. Cell membrane permeability assessed by Trypan blue exclusion in Drp1 KO MEFs in the absence or presence of CsA. Horizontal dashed line represents normal value from Cre-negative controls. Data are mean ± SEM of indicated number of experiments; * = P<0.05 vs identically treated Cre-negative controls; # = P<0.05 vs conditional Drp1 KO MEFs (no CsA). See also Figure S6.

Figure 7. Inhibition of MPTP opening improves the cardiomyopathy provoked by conditional Drp1 ablation

A. Representative four-chamber view and echocardiographic images of hearts 6weeks after Drp1 gene deletion in the presence or absence of MPTP regulatory protein, cyclophilin D (CypD). B. Ratio of heart weight to body weight. C–D. Echocardiographic LV r/h (C) and %FS (D). E. Masson’s trichrome-staining of myocardial fibrosis; group data are to the right. F. Fluorescence microscopy of Evans blue-perfused hearts co-stained with wheat germ agglutinin; quantitative group data are to the right. G. Group quantitative data for cardiomyocyte cross-sectional area (CSA). H. Ratio of mitochondrial protein to LV weight. I–K. Group data for myocardial mitochondrial content (I), individual mitochondrial area (J), and mitochondrial aspect ratio (K) derived from TEM images. Data are mean ± SEM; * = P<0.05 vs identically treated controls (Drp1 f/f ± CypD KO); # = P<0.05 vs Drp1 KO.