Defective nucleotide-dependent assembly and membrane fusion in Mfn2 CMT2A variants improved by Bax

Abstract

Mitofusins are members of the dynamin-related protein family of large GTPases that harness the energy from nucleotide hydrolysis to remodel membranes. Mitofusins possess four structural domains, including a GTPase domain, two extended helical bundles (HB1 and HB2), and a transmembrane region. We have characterized four Charcot-Marie-Tooth type 2A-associated variants with amino acid substitutions in Mfn2 that are proximal to the hinge that connects HB1 and HB2. A functional defect was not apparent in cells as the mitochondrial morphology of Mfn2-null cells was restored by expression of any of these variants. However, a significant fusion deficiency was observed in vitro, which was improved by the addition of crude cytosol extract or soluble Bax. All four variants had reduced nucleotide-dependent assembly in cis, but not trans, and this was also improved by the addition of Bax. Together, our data demonstrate an important role for this region in Mfn2 GTP-dependent oligomerization and membrane fusion and is consistent with a model where cytosolic factors such as Bax are masking molecular defects associated with Mfn2 disease variants in cells.

Figure 1.. Structural model of the positions of four hinge 1-proximal amino acid substitutions associated with CMT2A.

(A) Structural model of the predicted extended structure of Mfn2 based on the crystal structure of the structurally related protein BDLP with 5′-Guanylyl imidodiphosphate (GMPPNP) (protein data base [PDB] 2W6D). The GTPase domain is green, HB1 is blue, HB2 is red, the transmembrane (TM) domain is grey, and Loops 1/2 are purple. Structural prediction performed by I-TASSER server (Zhang, 2009; Yang & Zhang, 2015). (B) Enlarged view of hinge 1 showing the positions of relevant CMT2A-related amino acids in pink with side chains. (C) Structural model of the predicted closed structure of Mfn2 based on the crystal structure of the structurally related protein BDLP with GDP (PDB 2J69). (A) Domains are colored as described in (A). Structural prediction performed by I-TASSER server (Zhang, 2009; Yang & Zhang, 2015). (D) Enlarged view of Loop 1 from hinge 1 showing the positions of the CMT2A-related amino acids in pink with side chains.

Figure S1.. Mfn2_IM structure with Mfn2 CMT2A–associated variant positions in α helix 3.

The Mfn2_IM crystal structure with the GTPase domain in green, HB1 in blue and S378, A383, Q386, and C390F in dark pink. The positions of the N-terminal truncation (R400) and the start of the C-terminal helix (R707) are shown with arrows (PDB 6JFK).

Figure S2.. Mfn2 protein expression in MEF clonal populations.

Whole-cell lysates prepared from the indicated cell lines were subject to SDS–PAGE and immunoblotting with α-Mfn2 and α-tubulin. Molecular weight markers are indicated in kD on the left.

Figure 2.. Mfn2 hinge 1-proximal variants support mitochondrial fusion when expressed in Mfn2-null cells.

(A) Representative images of mitochondrial networks in wild-type (Mfn1^+/+Mfn2^+/+) or Mfn2-null (Mfn1^+/+Mfn2^−/−) MEFs expressing the indicated Mfn2 variant. Mitochondria were stained with MitoTracker Red CMXRos and visualized by fluorescence microscopy. Images represent a maximum intensity projection. Scale bars = 5 μm. (A, B) Quantification of mitochondrial morphology in cells represented in (A). Error bars indicate mean + SD from three independent, blinded experiments (n ≧ 100 cells per population per experiment).

Figure S3.. Mitochondrial connectivity and fusion as measured by redistribution of GFP in cells.

Wild-type (Mfn1^+/+Mfn2^+/+) or clonal populations of Mfn2-null MEFs expressing the indicated Mfn2 variant were transduced with the mitochondrial matrix-targeted photoactivatable-GFP (mito-PAGFP). These cells were labeled with MitoTracker Red CMXRos before imaging. A small region of the cells that contained mitochondria (∼1 μm²) was activated using a 405-nm laser, and the cell was imaged immediately after the activation and 50 min later. The area of green pixels and red pixels was determined for each image, where red pixels represent the area of the mitochondrial network. The proportion of the mitochondrial network that possessed green pixels was determined at time = 0 and time = 50 min. The increase in this proportion of t = 50 relative to t = 0 is shown. Paired t test analysis indicated that the decreased overlap observed was not statistically significant.

Figure S4.. Mitochondrial morphology in cells after oxidative stress.

Wild-type (Mfn1^+/+Mfn2^+/+) or Mfn2-null MEFs expressing the indicated Mfn2 variant were treated with 100 μM diamide for 1 h. Mitochondria were stained with MitoTracker Red CMXRos and visualized by fluorescence microscopy. The graph represents the mitochondrial morphology as scored in three independent experiments. Error bars indicate mean + SD (n ≧ 100 cells per population per experiment).

Figure 3.. Mitochondrial in vitro fusion assay reveals a defect for all Mfn2 hinge 1-proximal variants.

(A) Mitochondria isolated from wild-type cells or clonal populations of Mfn2-null MEFs either transduced with empty vector or expressing the indicated Mfn2 variant were subject to in vitro fusion conditions at 37°C for 60 min. The data are represented as relative to wild-type controls performed in parallel. Error bars indicate mean + SD from at least three independent experiments. The average values of each is indicated in white within the individual bars. (B) Mitochondrial in vitro fusion assay performed as in (A) except with the addition of cytosol-enriched fraction to the reaction buffer. Data are represented as proportion of control reactions performed in parallel without cytosol. Error bars indicate mean + SD from at least three independent experiments. The average values of each is indicated in white within the individual bars. (C) Mitochondrial in vitro fusion assay performed as in (A) except with the addition of 300 nM purified Bax protein to the reaction buffer. Data are represented as proportion of control reactions performed in parallel without Bax. Error bars indicate mean + SD from at least three independent experiments. The average values of each is indicated in white within the individual bars.

Figure 4.. Mfn2 hinge 1-proximal variants interact with Mfn1 in cis and trans.

(A) Schematic of the differential epitope labeling used in the co-immunoprecipitation assay. Interactions tested are Mfn2-FLAG with Mfn1 in cis (top left), Mfn2-FLAG with Mfn1-eGFP in trans (top center), and Mfn2-FLAG with Mfn2 in trans (bottom center). (B) Mitochondria were isolated from a clonal population of Mfn1-null cells expressing Mfn1^WT-eGFP at endogenous levels and clonal populations of Mfn2-null MEFs expressing the indicated Mfn2-FLAG variant. Mitochondria that possess Mfn1-eGFP and Mfn2 were combined with mitochondria that possess Mfn1 and Mfn2-FLAG; these mixtures were incubated with BeF₃ in the absence or presence of GDP. After lysis, immunoprecipitation was performed with α-FLAG magnetic beads. Proteins eluted from the beads were subjected to SDS–PAGE and immunoblotting with α-Mfn1 and α-Mfn2, as indicated. Arrows indicate endogenous Mfn1 (black) and Mfn2 (white); arrowheads indicate Mfn1-eGFP (white) and Mfn2-FLAG (black). Input represents 3% of the input and elution represents 37.5% of the immunoprecipitated protein. (C) Quantification of the percentage of Mfn1-eGFP in the elution compared with Mfn2-FLAG is shown as the mean + SD of three independent experiments.

Figure S5.. Immunoprecipitation of Mfn1-eGFP requires Mfn2-FLAG.

Mitochondria were isolated from a clonal population of Mfn1-null cells expressing Mfn1^WT-eGFP at endogenous levels (Mfn1-eGFP) or a clonal population of Mfn2-null cells expressing Mfn2-FLAG, or a clonal population of Mfn1-null cells expressing Mfn1-FLAG. Mitochondria were combined in the indicated combinations and incubated with BeF₃ in the presence or absence of GDP. After lysis, immunoprecipitation was performed with α-FLAG magnetic beads. Proteins eluted from the beads were subjected to SDS–PAGE and immunoblotting with α-Mfn1, α-FLAG, α-VDAC, or α-Hsp60, as indicated. Black arrow indicates endogenous Mfn1, white arrowhead indicates Mfn1-Egfp, black arrowhead indicates Mfn2-FLAG, and line indicates Mfn1-FLAG. Input represents 3% of the input and elution represents 37.5% of the immunoprecipitated protein. The molecular weight markers are shown in kD on the left.

Figure 5.. Mfn2 hinge 1-proximal variants have altered nucleotide-dependent assembly.

(A) Mitochondria were isolated from clonal populations of Mfn2-null MEFs expressing the indicated Mfn2 variant. Mitochondria were either untreated or incubated with 2 mM GTP or 2 mM GMPPNP as indicated before lysis and separation by BN-PAGE followed by immunoblotting with anti-FLAG antibody. Arrow indicates predicted dimer, closed arrowhead indicates ∼320-kD species, and open arrowhead indicates ∼450-kD species. Asterisk (*) indicates nonspecific signal. Approximate molecular weights are in kD on the left. (B) Quantification of proportion of the total protein observed in the ~320 or ~450-kD band after treatment with GMPPNP (filled arrowhead). Error bars indicate mean + SD from at least three independent experiments and the statistical significance were determined by paired t test analysis between the indicated data and wild type (*P < 0.05). (C) Mitochondria were prepared as in (A) and incubated with 2 mM GMPPNP with or without 1 μM recombinant purified Bax. Arrow indicates predicted dimer, closed arrowhead indicates ∼320-kD species, and open arrowhead indicates ∼450-kD species. Asterisk (*) indicates nonspecific signal. Approximate molecular weights are in kD on the left. (D) Quantification of proportion of the total protein observed in the ∼320-kD band. Error bars indicate mean + SD from at least three independent experiments and the statistical significance were determined by paired t test analysis between the indicated data and wild type (*P < 0.05).

Figure S6.. Structural model of the positions of hinge amino acid substitutions associated with CMT2A.

Structural models of the predicted closed structure of Mfn2 based on the crystal structure of the structurally related protein BDLP with GDP (PDB 2J69). The GTPase domain is green, HB1 is blue, HB2 is red, the transmembrane (TM) domain is grey, and Loops 1/2 are purple. (A) The positions of C390 and L710 are indicated in orange and labeled. (B) The positions of S378 and L710 are indicated in orange and labeled. Structural prediction performed by I-TASSER server (Zhang, 2009; Yang & Zhang, 2015).

Figure 6.. Mfn2 variants with substitutions in both HB1 and HB2 are defective for fusion in Mfn2-null cells.

(A) Representative images of mitochondrial networks in Mfn2-null MEFs expressing the indicated Mfn2-mNeonGreen variant. Mitochondria were stained with MitoTracker Red CMXRos and visualized by fluorescence microscopy. Images represent a maximum intensity projection. Scale bars = 5 μm. (B) Quantification of mitochondrial morphology of cells represented in (A). Error bars indicate mean + SD from at least three independent experiments and the statistical significance was determined by paired t test analysis between the indicated data and wild type (*P < 0.05) or between the indicated data and Mfn2^S378P/L710P (†P < 0.05).