OPA1 disease-causing mutants have domain-specific effects on mitochondrial ultrastructure and fusion

Abstract

Inner mitochondrial membrane fusion and cristae shape depend on optic atrophy protein 1, OPA1. Mutations in OPA1 lead to autosomal dominant optic atrophy (ADOA), an important cause of inherited blindness. The Guanosin Triphosphatase (GTPase) and GTPase effector domains (GEDs) of OPA1 are essential for mitochondrial fusion; yet, their specific roles remain elusive. Intriguingly, patients carrying OPA1 GTPase mutations have a higher risk of developing more severe multisystemic symptoms in addition to optic atrophy, suggesting pathogenic contributions for the GTPase and GED domains, respectively. We studied OPA1 GTPase and GED mutations to understand their domain-specific contribution to protein function by analyzing patient-derived cells and gain-of-function paradigms. Mitochondria from OPA1 GTPase (c.870+5G>A and c.889C>T) and GED (c.2713C>T and c.2818+5G>A) mutants display distinct aberrant cristae ultrastructure. While all OPA1 mutants inhibited mitochondrial fusion, some GTPase mutants resulted in elongated mitochondria, suggesting fission inhibition. We show that the GED is dispensable for fusion and OPA1 oligomer formation but necessary for GTPase activity. Finally, splicing defect mutants displayed a posttranslational haploinsufficiency-like phenotype but retained domain-specific dysfunctions. Thus, OPA1 domain-specific mutants result in distinct impairments in mitochondrial dynamics, providing insight into OPA1 function and its contribution to ADOA pathogenesis and severity.

Keywords: ADOA; OPA1; cristae; dynamics; mitochondria.

Fig. 1.

ADOA-causing mutants located at GTPase and GED domain differentially alter mitochondrial reshaping proteins in patient-derived fibroblasts. (A and B) Western blot analysis of whole-cell extracts from ADOA-derived fibroblast or myoblasts, representative images. (C) Representation of OPA1, primary structure indicating specific domains, based on the result of polypeptide sequence (isoform 1), showing the location of ADOA-causing mutations described in A. MTS, mitochondrial targeting sequence; TM, transmembrane domain; S1, processing site 1; GTPase, GTPase domain; GED, GTPase effector domain.

Fig. 2.

Mitochondrial ultrastructure of ADOA patient-derived cells carrying GTPase and GED mutations displays dissimilar aberrations in organelle morphology, and cristae shape. (A) Representative TEM images of mitochondria from ADOA-derived fibroblast. (Scale bar, 500 nm.) (B) Total mitochondrial area quantification. (Control skin, n = 2/19/345; preparations/cells/mitochondria), (OPA1c.889C>T, n = 2/30/411), (control SkM, n = 3/17/1,973), (OPA1c.870+5G>A, n = 3/25/982), (OPA1c.2713C>T, n = 3/28/1,223), (OPA1c.2818+5G>A. n = 2/18/1,128). (C) Schematic representation of an individual crista. The short dash line indicates how cristae junction width was measured. The large dash line indicates how cristae lumen width was calculated. (D) Cristae junction width analysis. (Control skin, n = 2/69/222; preparations/mitochondria/ CJ), (OPA1c.889C>T, n = 2/60/230), (control SkM, n = 3/74/192), (OPA1c.870+5G>A, n = 3/65/190), (OPA1c.2713C>T, n = 3/112/474), (OPA1c.2818+5G>A, n = 2/71/311). (E) Cristae lumen width analysis, (control skin, n = 2/113/697; preparations/mitochondria/CL), (OPA1c.889C>T, n = 2/102/484), (control SkM, n = 3/83/376), (OPA1c.870+5G>A, n = 3/60/335), (OPA1c.2713C>T, n = 3/184/1,320), (OPA1c.2818+5G>A, n = 2/89/766), Data are mean ± SEM. ****P < 0.0001 vs. respective control condition. Blue and red bars represent GTPase and GED mutants, respectively.

Fig. 3.

GTPase and GED domain distinctly alter mitochondrial fusion/fission dynamics in ADOA patients-derived fibroblasts. (A) ADOA-derived fibroblasts coexpressing mt-PAGFP and mt-DsRed. Mitochondrial morphology was obtained using mt-DsRed by 560-nm laser excitation. Panels show representative images of the different mitochondrial morphologies observed (fragmented, intermediate, and elongated). (Scale bar, 10 μm.) (B) Quantification of mitochondrial morphology using mt-DsRed of control or ADOA-derived fibroblast carrying the mutants OPA1c.870+5G>A or OPA1c.2713C>T. (C) ADOA-derived fibroblast expressing mt-PAGFP and mt-DsRed 40 s after photoactivation of 5 × 5 μm ROIs (white squares) by 408-nm laser illumination, acquired upon 488- and 560-nm laser excitation; the Bottom panel shows Inset from the Middle panel, displaying one ROI highlighting the extent of mt-PAGFP diffusion. (D) Fluorescence decay quantification 40 s after mt-PAGFP photoactivation. (E) Fusion events frequency (control = 20 cells, OPA1c.870+5G>A = 22 cells, OPA1c.2713C>T = 24 cells; of ≥3 independent experiments). (F) Fission events frequency (control = 18 cells, OPA1c.870+5G>A = 17 cells, OPA1c.2713C>T = 17 cells; of ≥3 independent experiments). Data are mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0,001 vs. control condition. ##P < 0.01 vs. c.870+G>A. Blue and red bars represent GTPase and GED mutants, respectively.

Fig. 4.

Effect of OPA1 GTPase and GED domain-specific on mitochondrial dynamics proteins in Opa1^−/− and WT MEFs. (A and B) Opa1^−/− and WT MEFs were transfected with plasmids carrying the in vitro generated OPA1 ADOA mutants OPA1c.870+5G>A and OPA1c.2713C>T. Panels show representative western blot images of whole-cell extracts. Left and Right panels correspond to Opa1^−/− cells rescued with OPA1 mutants and WT cells overexpressing the OPA1 mutants, respectively.

Fig. 5.

Expression of OPA1 GTPase and GED domain-specific mutants differentially disrupts mitochondrial fusion dynamics in WT MEFs. (A) Cells acutely coexpressing OMM-targeting mCherry-OMP25, mt-PAGFP, and OPA1 mutants. Each row displays experiments before and 20 s after photoactivation of 3 × 3 μm ROIs (white squares) by two-photon laser illumination (760 nm), acquired upon 488- and 560-nm laser excitation; the Right panel shows Inset from the Middle panel, displaying one ROI (white dashed square) highlighting the extent of mt-PAGFP diffusion. (B) Quantification of mitochondrial network morphology of WT MEFs acutely overexpressing OPA1 mutants classified as elongated, intermediate, and fragmented. Data are at least 15 cells per condition from ≥3 independent experiments (WT = 23 cells, Opa1^−/− = 24 cells, WT+OPA1WT = 33 cells, WT+OPA1c.870+5G>A = 23 cells, WT+OPA1c.899G>A = 20 cells, WT+OPA1c.1334G>A = 20 cells, WT+OPA1c.2708delTTAG = 15 cells, WT+OPA1c.2713C>T = 21 cells, WT+OPA1c.2818+5G>A = 19 cells. (C) Quantification of mt-PAGFP decay after photoactivation. Lanes are means of individual ROIs. (D) The bar chart displays the mt-PAGFP fluorescence decay 20 s after PA. Data are mean ± SEM of cells used in B. (E) Mitochondrial fusion events frequency. Bar charts show mean ± SEM of the same cells used in B. (F) Quantification of mitochondrial fusion duration (time to completion of mt-PAGFP transference from a donor to an acceptor mitochondrion). Bar charts show mean ± SEM of the same cells used in B. (G) Mitochondrial morphology quantification on WT MEFs stably overexpressing OPA1 WT, OPA1 c.870+5G>A, or OPA1 c.2713C>T. Data are mean ± SEM. ***P < 0.001, ****P < 0.0001 vs. Opa1^−/−. #P < 0.05, ##P < 0.01, ###P < 0.001, ####P < 0.0001 vs. WT. Blue and red bars represent GTPase and GED mutants, respectively.

Fig. 6.

Role of the OPA1 GED region in OMM–IMM fusion kinetics, OPA1 oligomerization, and GTPase activity. (A) Left panel: Representative example of a cell fusion assay of cells expressing mCherry-OMP25 (OMM) and COXIV8aGFP (IMM). Right panel: inset from left panel highlighting a sequential OMM and IMM fusion event. Live cell imaging was conducted at an LSM880 Zeiss confocal with Airyscan detectors. (B) Schematic representation of OMM, followed by IMM fusion, between two adjacent mitochondria. (C) Sequential OMM and IMM fusion mixing kinetics of mitochondrial fusion event indicated in white arrows in panel A. (D) OMM mixing completion time in Opa1^−/− MEFs expressing OPA1 mutants (WT = 10 fusion events, Opa1^−/− +OPA1 WT = 5 fusion events, Opa1^−/− +OPA1 c.2713C>T = 4 fusion events). Bar charts show mean ± SEM. (E) IMM mixing completion time in Opa1^−/− MEFs expressing OPA1 mutants. Bar charts show mean ± SEM. (F) GTPase activity assay measured in immunopurified proteins by malachite green assay. n = 3. Data are mean ± SEM. #P < 0.05, ##P < 0.01, ###P < 0.001 vs. WT. Blue and red bars represent GTPase and GED mutants, respectively.

Fig. 7.

OPA1 GTPase and GED mutant protein stability. (A) WT and Opa1^−/− MEFs expressing OPA1 GTPase c.870+5G>A or GED c.2713C>T mutants were incubated with FCCP 1 μM for 1, 3, and 6 h, representative OPA1 western blots of whole-cell extracts. (B) OPA1 processing was quantified as the ratio of the OPA1 short form vs. the OPA1 long form. n = 3. (C) OPA1 importation was quantified as OPA1 total (long + short form)/OPA1 preprotein. (D) WT and Opa1^−/− MEFs expressing OPA1 GTPase or GED mutant were incubated with 50 μM of cycloheximide (CHX) for 1, 3, and 6 h, representative OPA1 western blots. (E) OPA1 protein levels in cells expressing a GTPase point mutation (P.M.) or a GED splice defect (Sp.D.), upon incubation with CHX for 6 h. (F) The bar plot shows western blot quantification of WT or Opa1^−/− MEFs expressing Sp.D. or P.M. mutants for the GTPase domain or GE region after incubation with CHX for 6 h. n = 3. (G) WT and Opa1^−/− MEFs expressing OPA1 GTPase c.870+5G>A or OPA1 GED c.2713C>T mutants were incubated with 25 μM of MG132 to determine the ubiquitin–proteasome system role in OPA1 protein levels. (H) Cells expressing a GTPase P.M. or a GED Sp.D. were incubated with MG132 for 6 h, to test OPA1 protein levels by western blot. (I) Western blot quantification of WT or Opa1^−/− MEFs expressing splicing Sp.D. or P.M. for GTPase domain or GE domain after incubation with MG132 for 6 h. n = 3. Data are mean ± SEM *P < 0.05, **P < 0.01, ***P < 0.001 vs. t0. #P < 0.05 vs. Opa1^−/− +OPA1 WT t0. Blue and red bars represent GTPase and GED mutants, respectively.

Fig. 8.

Distinctive mitochondrial fusion dynamics and ultrastructure defects in OPA1 domain-specific mutants; summary of main findings. (A) Scheme of main common features for the OPA1 GTPase and GED ADOA-causing mutants studied in this work. The cartoon shows that OPA1 GTPase ADOA plus–causing mutants exhibit dissimilar mitochondrial morphology (elongated or fragmented), cristae discontinuities, and cristae junction loss. OPA1 GED ADOA-causing mutants exhibited fragmented mitochondria, swollen organelles, loss of cristae junction, and wider cristae lumen. (B) Table summarizing the domain-specific findings in ADOA patient-derived cells and acute expression of ADOA-causing mutants in a WT and Opa1 KO background. –, unchanged compared to the control condition; n.e., not evaluated. ^aData obtained from ref. .