Novel Mutation m.10372A>G in MT-ND3 Causing Sensorimotor Axonal Polyneuropathy

See supplementary material (links.lww.com/NXG/A388).

See supplementary material (links.lww.com/NXG/A388).

ATP production rate and RC enzyme activities were measured in isolated mitochondria from both muscle biopsies as described earlier.²³ For determination of RC enzyme activities in cultured myoblasts, mitochondria were isolated by homogenization of cells in a homogenizing and washing buffer (320 mM sucrose, 1 mM ethylene glycol tetraacetic acid, and 20 mM Tris-HCl pH7.2 with 0.1% bovine serum albumin [BSA]), followed by differential centrifugation. The final mitochondrial pellet was resuspended in resuspension buffer (250 mM sucrose, 15 mM K₂HPO₄, 2 mM MgAc₂, 0.5 mM EDTA, and 0.5 g/L BSA, pH 7.2) before determination of RC enzyme activities as previously described.²³ The rate of mitochondrial ATP production was determined in permeabilized myoblasts. The permeabilization procedure was performed on an ice bath and using a refrigerated centrifuge. Two solutions were used; A: 0.1 g/L digitonin, 150 mM KCl, 25 mM Tris, 10 mM K₂HPO₄, 2 mM EDTA and 1 g/L BSA, pH 7.5); B: same as A but with digitonin omitted. Approximately 1 million myoblasts at a confluency of 80-90% were harvested for each experiment. The cell pellet was incubated for 10 minutes with 100 μL of solution A. After incubation cells were centrifuged for 3 minutes at 700g and washed twice with 100 μL of solution B. Finally, cells were resuspended in 100 μL of solution B. ATP production was measured in white 96-well microtiter plates. To each well, ATP monitoring reagent, adenosine diphosphate, and one of the following substrate combinations were added: glutamate + malate, pyruvate + malate, or succinate + rotenone. Each substrate combination was measured in the absence or presence of oligomycin (OM). Reaction was started by addition of approximately 1 × 10³ cells/well, and the increase in light emission was followed for 10 minutes at 25°C in a Perkin Elmer Victor² plate reader. Measurements were standardized by determining the subsequent increase in light emission after addition of 500 nM ATP to wells containing all the above-described combinations of substrates and OM, but without cells. Each measurement was repeated in 4 wells. In total, 24 wells were used per sample investigated, and 24 wells were used for the calibration procedure. Mitochondrial ATP production in myoblasts was calculated as ATP production rate (−OM) – ATP production rate (+OM) for each substrate combination. Final reagent concentrations and principles of the measurement procedure were as previously described.²⁴

See supplementary material (links.lww.com/NXG/A388).

See supplementary material (links.lww.com/NXG/A388).

See supplementary material (links.lww.com/NXG/A388).

This study has been approved by the regional ethical board in Stockholm, Sweden (2014/995-32). Written informed consent was obtained from the patient involved.

The data that support the findings of this study are available from the corresponding author on reasonable request.

A 46-year-old woman with no family history of neuromuscular disease or other clinical symptoms typically associated with mitochondrial diseases presented with a 5-year history of painful paresthesia and numbness, which started in her toes but later progressed to her feet and hands. During the same period, she also experienced fluctuating muscular weakness in the lower and upper extremities. The initial neurologic examination showed normal walking capacity and a minor distal weakness in both upper and lower extremities, but no fasciculations in the extremities. The patient had reduced light touch sensibility in the lower legs, reduced pinprick sensibility in the feet, and lack of vibration sensibility in the ankles. The reflexes were generally reduced, with a total loss of ankle reflexes. The patient had no ataxia, cranial nerve affection, or Babinski sign. Nerve conduction studies and quantitative sensory testing showed a symmetric sensorimotor axonal polyneuropathy as well as signs of small fiber involvement (table e-1, links.lww.com/NXG/A388). The EMG showed a mostly chronic neurogenic pattern and no signs of myopathy (table e-2). To screen for known etiologic causes of her sensory-predominant neuropathy we performed a comprehensive work-up. The patient denied alcohol overconsumption. Routine basal blood samples including serum cobalamin, folate acid and homocysteine, oral glucose tolerance test and serum protein electrophoresis were normal. Antibodies for rheumatic disease (antinuclear antibodies, rheumatoid factor and anti-neutrophils cytoplasmic antibodies) were negative. Screening for hepatitis C, HIV, Lyme disease and syphilis was also negative. Transglutaminase antibodies (Celiac disease), abdominal fat biopsy and TTR gene sequencing (Amyloidosis) were normal. Urine trihexoside was normal and GLA gene sequencing (Fabry disease) did not reveal any disease-causing variants. To identify a possible inflammatory etiology, a muscle biopsy with a strict morphologic analysis was performed. Surprisingly, the muscle biopsy showed ragged red fibers indicating a mitochondrial disorder, but no signs of myositis or vasculitis. To expand the mitochondrial investigation, the patient was referred to the Centre for Inherited Metabolic Diseases for further investigation, as described below. During the following year, the patient experienced a gradual increase in the level of muscle weakness leading to inability to walk without using a walking aid. A follow-up EMG at this time revealed a generalized myopathic pattern (table e-2). Blood workup including creatine kinase levels was normal. During the following 3 years, the patient also experienced fluctuating bilateral blurred vision. The eye examination showed punctate retinal hemorrhages of unknown etiology in the right eye and no optic disc changes, although visual evoked potential revealed bilateral prolonged latencies indicating optic nerve affection, which was assumed to be related to her mitochondrial condition. Since then, repeated eye assessments have been performed. The patient has a Leber miliary aneurysm in the right retina and has received laser treatment. She has also undergone bilateral cataract surgery as well as epiretinal membrane removal and vitrectomy at her right eye. None of these conditions are considered to be associated with her mitochondrial disorder. At the time of diagnosis, investigation of other organ systems included normal echocardiography and an auditory test. MRI of the brain showed a slight global parenchymal atrophy, but no white matter changes. A repeated MRI of the brain at age 58 years showed similar results. The proximal muscle weakness has worsened during the years leading to the use of a power wheelchair outdoors at age 58 years, although the patient is still capable of walking short distances without aid indoors. She complains of muscle soreness, clumsiness in her hands, and fatigue, which worsen with physical activity. She has tried coenzyme Q10 and creatinine without effect. She has continuous severe neuropathic pain and uses gabapentin and sometimes tramadol. Earlier use of tricyclic antidepressant and duloxetine had no effect or troublesome side effects.

A muscle biopsy for mitochondrial investigation was conducted in 2008 when the patient was aged 47 years. Measurement of mitochondrial ATP production and the RC activity in mitochondria isolated from muscle was performed. Mitochondrial ATP production was decreased using all substrates involving CI. Only N,N,N',N'-Tetramethyl-p-phenylenediamine (TMPD) + ascorbate (complex IV) and succinate + rotenone (complex II + III + IV) showed normal activities (figure 1A). RC activity showed markedly decreased activities for CI and CI + complex III. Also, complex II and II + III were slightly reduced while the activity for complex IV was normal (figure 1B). The activities of the mitochondrial control enzymes glutamate dehydrogenase and citrate synthase were threefold increased compared with the controls, indicating mitochondrial proliferation in the muscle tissue (figure 1B). Morphologic analysis was performed on muscle tissue from the same biopsy. Gomori trichrome staining revealed numerous ragged red fibers (RRFs) indicating accumulation of abnormal mitochondria below the plasma membrane of the muscle fiber (figure 2A). All fibers, except 1 COX-negative fiber, were shown to be COX positive in cytochrome oxidase staining indicating normal complex IV function of the mitochondria (figure 2B). Electron microscopy demonstrated abnormal collection of mitochondria with paracrystalline inclusions (figure 2, C and D).

To search for the genetic cause of the disease, mtDNA isolated from muscle was analyzed by Sanger sequencing. The analysis detected a novel heteroplasmic mutation in MT-ND3, m.10372A>G (predicting p.(Glu105Gly)). MT-ND3 is one of the mtDNA-encoded subunits of CI in the RC. The identified mutation had a high heteroplasmic level in muscle according to the Sanger electropherogram, almost no wild-type mtDNA was detected (not shown). Sanger sequencing could not detect the mutation in DNA from blood, cultured fibroblasts, or urinary epithelial cells from the patient.

To further characterize the mutation, mtDNA from blood, cultured fibroblasts, and muscle was quantified by last-cycle hot PCR. Quantification showed 97% mutated mtDNA in muscle and 3% in fibroblasts, but undetectable in blood (figure 3, A and B).

To further characterize the muscle mitochondria carrying the m.10372A>G mutation, blue native polyacrylamide gel electrophoresis (BN-PAGE) of mitochondria isolated from the first biopsy was conducted. No deviation in the quantity or size of CI in the patient was seen compared with controls, indicating that the mutation in MT-ND3 does not affect assembly of CI (figure 4A).

The mutation was not found in 50,175 GenBank sequences according to MitoMap²⁴ and not in 2,704 sequences in the Human Mitochondrial Genome Database (mtDB).²⁵ The mutation was predicted to be pathogenic by the prediction tools PolyPhen2 (HumDiv: score 1.00, HumVar: score 0.99), SIFT (score: 0), and Combined Annotation Dependent Depletion (CADD, score 24.0), MutPred2 (altered ordered interface, p value = 0.003), PROVEAN (score: −6.767), and Rhapsody (score: 0.68). The residue is also highly conserved across different species down to the CI core subunit Nqo1 in Thermus thermophilus (figure 4B), and all 61 mammals in the mtSNP Database mtSAP²⁶ had glutamate in the corresponding position (not shown). The glutamate residue at position 105 is located close to the C-terminal tail of ND3 that faces the mitochondrial matrix (figure 4C). In a human CI structure obtained by cryogenic electron microscopy (PDB:5XTC),²⁷ the negatively charged glutamate does not establish ionic contacts with any of the predominantly hydrophobic residues in its proximity (figure 4D). Furthermore, modeling of mutant ND3 with SWISS-MODEL²⁸ did not cause structural changes to the protein compared with the canonical ND3 sequence (figure 4E), indicated by an root-mean-square deviation score of 0.001. This suggests that the mutation does not affect the stability of CI, but might rather have functional consequences.

To establish a cell type carrying the mutation for further experiments, the patient was asked to undergo another muscle biopsy in 2016 when she was aged 55 years. The second muscle biopsy confirmed the morphologic (not shown) and biochemical analysis (figure 1, A and B) as well as the high mutation levels in muscle tissue (figure 3C). Cultured myoblasts derived from the same biopsy, however, did not carry the m.10372A>G mutation (figure 3D). To ensure that the cells in the myoblast culture from the patient were myogenic cells, immunocytochemistry with the myogenic marker desmin was performed (figure 3E). Primary fibroblasts from the patient were used as control cells for the staining (figure 3F). Sanger sequencing of the complete mtDNA from the cultured myoblasts showed the same sequence of m.1-16569 as for the former sequenced muscle DNA, except for the position m.10372. The fact that mtDNA from cultured myoblasts vs muscle tissue from the same biopsy only differed in 1 position gave us the opportunity to investigate the functional consequences of this novel mutation. We therefore performed mitochondrial biochemical investigations of the myoblast cells. Analysis of the patient's myoblasts showed normal mitochondrial ATP production (figure 5A) and RC activity (figure 5B) compared with control myoblasts. This indicates that the previously identified CI deficiency in muscle tissue from the same biopsy was caused by the m.10372A>G mutation.

The m.10372A>G mutation had been identified before the era of massive parallel sequencing. A whole-genome sequencing (WGS) analysis of DNA from muscle targeting all disease-associated genes in Online Mendelian Inheritance in Man and all genes in Human MitoCarta 2.0 was thus performed later. This was performed to exclude additional mutations in nuclear genes that could potentially contribute to the CI deficiency or the clinical symptoms of the patient. As expected, the WGS analysis detected the former identified MT-ND3 mutation at a high heteroplasmic level, but no other disease-causing mutation in any known nuclear disease gene was found that could explain the phenotype.