Article Figures & Data
Figures
- Figure 1. Muscle pathology findings.
(A, B, C, D, E, F) Representative images of the muscle biopsy (A, B, C, D, E, F) including two formalin-fixed, paraffin-embedded slides (A, F) and four snap-frozen muscle sections (B, C, D, E). Distances are represented in the scale bars. (A) (haematoxylin–eosin; H&E) demonstrates extensive adipose tissue (“fatty replacement”) in the biopsy (black arrow). This was less obvious in the frozen tissue, which had been dissected away from obvious fat. (B) (H&E) illustrates the variation in myofibre size, as well as the increased connective tissue between myofibres (endomysial fibrosis; green arrow). (C) Biopsy had several necrotic fibres (panel (C), H&E; black arrow) and increased numbers of internalized nuclei (black arrowhead). In panel (D) (H&E) are two basophilic regenerating fibres (black arrows), internalized nuclei (black arrowheads), and increased endomysial connective tissue (green arrow). (E) Dystrophin immunoperoxidase stains the myofibre sarcolemma (Dys2 illustrated in panel (E)). Dys1 and Dys3 are similar (data not shown). (F) Desmin immunoperoxidase in longitudinal sections (panel (F)) illustrates the repetitive sarcolemmal units (stripes in each fibre) but does not stain significant sarcoplasmic deposits. (G) MRI image at the level of the legs demonstrating predominant atrophy in white arrows: gastrocnemius (i), soleus (ii), and tibialis anterior (iii) bilaterally. (H) MRI image of upper thighs demonstrating predominant atrophy of adductor longus and gluteus minimus.
- Figure 2. Initial characterization of the MFN2 Q367H variant.
(A) (i) Sanger sequencing results showing single nucleotide polymorphism in patient fibroblasts, transition from guanine to cytosine. (ii) Topology diagram of the MFN2 protein with domains and localization indicated. The site of the Q367H variant is shown in the diagram. (B) Amino acid sequence alignment using the T-COFFEE tool. (C) Western blot analysis comparing the expression of mitochondrial fusion protein in control versus patient fibroblasts, normalized to β-actin. (D, E, F, G) Comparison of protein expression levels (D) MFN2 (glucose media), (E) MFN2 (glucose-free/galactose-supplemented media), (F) MFN1, and (G) OPA1 levels in control versus patient fibroblasts. Data are indicative of the mean ± SD. Statistical comparisons were made by an unpaired t test (ns = not significant [P > 0.05]).
- Figure 3. Changes in mitochondrial morphology and respiratory function.
(A) Representative confocal images (top) with zoomed inset images (bottom) showing mitochondrial network morphology in control versus patient fibroblasts stained with TOMM20. (B) Qualitative scoring of proportion of cells showing fragmented/intermediate/fused mitochondrial network. Bars indicate the mean ± SD. (C) Quantitative analysis of mean mitochondrial branch length in control versus patient fibroblasts using glucose or glucose-free/galactose-supplemented media; bars show the mean ± SD, and colours indicate biological replicates. (D) Oxygen consumption rate (OCR) assay on control and patient fibroblasts using Seahorse XFe24; bars indicate the mean ± SD (n = 12 replicates). Statistical comparisons were made by an unpaired t test. (E) Mitochondrial membrane potential is reported as mean fluorescence intensity (MFI) of TMRE signal in live cells measured by flow cytometry. Bars indicate the mean ± SEM; points indicate biological replicates, unpaired t test (ns = not significant (P > 0.05); ****P ≤ 0.0001).
- Figure 4. Characterization of mtDNA nucleoids, mtDNA copy number, and mtDNA deletions.
(A) Representative confocal images showing mitochondria (TOMM20; grey) and mitochondrial nucleoids (dsDNA antibody; red) in patient and control fibroblasts under glucose and glucose-free/galactose-supplemented nutrient conditions. Circles indicate the presence of mtDNA nucleoids outside the mitochondrial network. (B) Number of nucleoids per cell; mean is shown, with colours indicating biological replicates. (C) Average size of mitochondrial nucleoids; violin plots show median and interquartile ranges. (D, E) qPCR data showing mitochondrial DNA copy number in (D) glucose and (E) glucose-free/galactose-supplemented media; bars indicate mean ± SD. (F) Agarose gel electrophoresis showing long-range PCR products used to detect full-length or deleted mtDNA and DNA ladder showing corresponding sizes. Statistical comparisons were made by an unpaired t test (ns = not significant (P > 0.05); *P ≤ 0.05; ***P ≤ 0.001; ****P ≤ 0.0001).
- Figure S1. Proximity ligation assay validation.
Pearson’s colocalization analysis comparing the overlap between proximity ligation assay signals with mitochondria (TOMM20) or ER (calnexin) in control and patient cells under glucose media. Violin plots indicate the median and interquartile range. Statistical analysis was performed via an unpaired t test.
- Figure 5. Mito-ER contact site alterations.
(A) Representative confocal images of mito-ER contact sites (MERCs) using proximity ligation assay (PLA) showing mitochondria (TOMM20; cyan), PLA probes (green), and ER (calnexin; magenta) in glucose and glucose-free/galactose-supplemented media. (B) Quantitative analysis of the number of MERCs per cell; lines indicate the mean ± SD, and colours indicate biological replicates. (C) Quantitative analysis of the average size of PLA probes; violin plots indicate the median and interquartile range. (D) Quantitative analysis of mitochondrial area in each cell type in μm2; lines show the mean ± SD, and colours indicate biological replicates. (E) Normalized number of MERCs to total mitochondrial area per cell; lines indicate the mean ± SD, and colours indicate biological replicates. P-values were determined by an unpaired t test (ns = not significant [P > 0.05]; *P ≤ 0.05; **P ≤ 0.01; ****P ≤ 0.0001).
- Figure 6. Changes to cellular lipid droplets.
(A, B) Representative confocal images showing lipid droplets stained with HCS LipidTox Green in (A) glucose and (B) glucose-free/galactose-supplemented media. (C) Quantitative analysis of the average number of lipid droplets per cell; bars indicate the mean ± SD, and colours indicate biological replicates. (D) Quantitative analysis of the average size of lipid droplets; violin plots show the median and interquartile range. (E) Representative live-cell confocal images showing mitochondria (MitoTracker Green) and fatty acids (BODIPY 558/568) at 0 and 24 h in standard or glucose-starved media. (F) Quantitative analysis of colocalization between fatty acid and mitochondrial network; colours indicate the cell type (red—control; and blue—MFN2 Q367H), and bars indicate the mean ± SD. P-values were determined by an unpaired t test (ns = not significant [P > 0.05]; **P ≤ 0.01; ****P ≤ 0.0001).
- Figure 7. MFN2 Q367H cells show activation of TLR9 and cGAS-STING inflammatory pathways, and colocalization of mtDNA with early endosomes.
(A) qRT-PCR data showing gene expression for TLR9/NF-κB targets (Asc, IL6, NLRP3, S100a9, TNF) and cGAS-STING targets (Ifn-α, Ifn-β) in control versus patient fibroblasts under glucose conditions. The graph also shows the effects of inhibitors for the TLR pathway (chloroquine) or cGAS-STING pathway (RU.521). Lines indicate the mean ± SD. (B) Representative confocal images showing mitochondria (TOMM20; cyan), mitochondrial nucleoids (anti-dsDNA antibody; green), and early endosomes (Rab5C; magenta). Images in the inset bottom right panel highlight the location of mitochondrial nucleoids outside the mitochondria and within the early endosomes, denoted by the orange arrowheads. (C) Quantitative analysis of signal colocalization between mitochondrial nucleoids and early endosomes in patient and control fibroblasts determined by Pearson’s coefficient; lines indicate the mean ± SD, and points indicate replicates. (D) Pathway model representing mitochondrial dysfunction leading to sterile inflammation. The diagram shows how the MFN2 Q367H variant can alter mitochondrial function, leading to the transfer of mtDNA from the mitochondria to the early endosomes (middle). This can lead to detection by TLR9 and subsequent activation of the NF-κB–mediated inflammation (right). In addition, mtDNA in the early endosomes could leak into the cytosol, leading to detection through the cGAS-STING pathway (middle-bottom), promoting type I interferon–mediated inflammation (right). Collectively, inflammation contributes to the distal myopathy disease outcome in the patient. P-values were determined by an unpaired t test (ns = not significant [P > 0.05]; *P ≤ 0.05; ****P ≤ 0.0001).
- Figure S2. Inflammatory response in glucose-free media.
RT–qPCR data showing gene expression for TLR9/NF-κB targets (Asc, IL6, NLRP3, S100a9, TNF) and cGAS-STING pathway targets (INF-α, INF-β) in control (CF) versus patient fibroblasts (Q367H) under glucose-free medium conditions. Colours indicate biological replicates. Error bars indicate the mean ± SD. Statistical analysis was performed via an unpaired t test with the relevant control.
- Figure 8. Altered MFN2 functions in transdifferentiated patient myoblasts.
(A) qRT-PCR data showing gene expression for TLR9/NF-κB targets (Asc, IL6, NLRP3, S100a9, TNF) and cGAS-STING targets (Ifn-α, Ifn-β) in control versus patient myoblasts; lines indicate the mean ± SD.(B) Representative confocal images showing mitochondrial network morphology (TOMM20; cyan) in control and patient myoblasts. (C) Quantitative analysis of mean mitochondrial branch length in control and patient myoblasts; violin plots show the median and interquartile ranges. (D) Representative confocal images of control and patient myoblasts showing PLA probes (representing MERCs; green), mitochondria (TOMM20; cyan), and ER (calnexin; magenta). (E) Quantitative analysis of the number of MERCs per cell in control and patient myoblasts; violin plots show the median and interquartile ranges. (F) Quantitative analysis of the average size of MERCs per cell in control and patient myoblasts; violin plots show the median and interquartile ranges. (G) Representative confocal images showing mitochondria (TOMM20) and mitochondrial nucleoids (dsDNA antibody). Circles indicate mtDNA staining outside the mitochondrial network. (H) Quantification of the average number of nucleoids per cell in control and patient myoblasts; violin plots show the median and interquartile ranges. (I) Quantification of the average size of mtDNA nucleoids in control and patient myoblasts. Violin plots show the median and interquartile ranges. P-values were determined by an unpaired t test (ns = not significant [P > 0.05]; ***P ≤ 0.001; ****P ≤ 0.0001).
- Figure 9. Re-Expression of WT MFN2 and MFN2 Q367H in KO cells.
(A) Representative confocal images showing mitochondria (TOMM20; grey), mtDNA (dsDNA; green), and early endosomes (Rab5; magenta) in U2OS WT, U2OS MFN2 KO, and U2OS MFN2 KO cells with re-expression of WT MFN2 and U2OS MFN2 KO cells with re-expression of Q367H MFN2 (left to right). Insets are of the boxed regions in above pictures, dotted circles show mtDNA outside the mitochondrial network, whereas arrows show mtDNA outside the mitochondrial network and colocalized with early endosomes. (B) Western blot analyses confirming the re-expression of WT MFN2 and MFN2 Q367H in U2OS MFN2 KO cells. (C) Quantitative analyses of mitochondrial network morphology in the four cell lines. (D) Quantitative analyses of Pearson’s colocalization coefficient between mtDNA and early endosomes in the four cell lines. Colours indicate biological replicates; violins show the median and interquartile range. Statistics: two-way ANOVA (ns = not significant [P > 0.05]; *P ≤ 0.05; **P ≤ 0.01; ****P ≤ 0.0001).
- Figure S3. Re-Expression of WT MFN2 and MFN2 Q367H in KO cells.
Quantification of additional cellular phenotypes for the indicated cell lines from Fig 8. (A, B, C, D, E) Characterization of mitochondrial DNA including (A) nucleoid copy number; (B) nucleoid size; (C) average number of extramitochondrial nucleoids; (D) average number of extramitochondrial nucleoids within Rab5-positive early endosomes (D); and (E) percentage of extramitochondrial nucleoids within Rab5-positive endosomes. (F, G, H) Quantification of Rab5-positive endosomes, including (F) average number of endosomes; (G) average endosome size; and (H) average distance between mitochondria and endosomes. Colours indicate biological replicates; violin plots show the median and interquartile range, whereas error bars show the mean ± SD. P-values are all determined by two-way ANOVA (ns = not significant [P > 0.05]; *P ≤ 0.05; **P ≤ 0.01; ****P ≤ 0.0001).
Tables
Gene Disease/Inheritance pattern Coding variant Amino acid change rsID Max reported MAF ClinVar Zygosity PolyPhen2 prediction MutationTaster prediction MFN2 CMT neuropathy 2A; AD c.1101G>C p.Gln367His rs373211062 0.00024 VUS, conflicting interpretations Heterozygous B D DYSF LGMD 2B; AR c.5180G>C p.Gly1727Ala rs146153532 1.16E-04 VUS Heterozygous P D SYNE1 EDMD 4; AD/AR c.9934G>A p.Asp3312Asn rs147281213 2.33E-04 VUS Heterozygous D D AD = autosomal dominant; AR = autosomal recessive; B = benign; CMT = Charcot–Marie–Tooth; D = deleterious; EDMD = Emery–Dreifuss muscular dystrophy; LGMD = limb–girdle muscular dystrophy; MAF = minor allele frequency; P = possibly tolerated; VUS = variant of uncertain significance.
In this Issue
Subjects
Related Articles
Cited By...
- No citing articles found.