Genetic and clinical features of Chinese patients with mitochondrial ataxia identified by targeted next‐generation sequencing

Summary

Aim

To characterize the mutations in mitochondrial DNA (mtDNA) and mitochondrion‐related nuclear genes (nDNA), and clinical features in Chinese patients with mitochondrial ataxia.

Methods

Targeted next‐generation sequencing (NGS) technology was performed to screen the whole mtDNA sequence and nDNA genes in a cohort of 33 unrelated ataxia patients.

Results

A total of 5 pedigrees were finally genetically diagnosed as mitochondrial ataxia, with 3 pathogenic mutations (m.8344A>G, m.9176T>C, and m.9185T>C), one likely pathogenic mutation (m.3995A>G) in mtDNA, and one pathogenic mutation (c.1159_1162dupAAGT, p.Ser388Terfs) in PDHA1. The prevalence of mitochondrial ataxia in our patient cohort is 15.2%. In addition, all 4 patients with mtDNA mutations experienced symptoms of ataxia with age at onset ranging from 12 to 39 years (21 ± 12.2) and developed extrapyramidal symptoms during the disease course. One male patient with pyruvate dehydrogenase deficiency showed an acute intermittent ataxia phenotype.

Conclusions

Our results implicate that mitochondrial ataxia might not be as rare in Chinese as previously assumed. This study firstly defines the mutations of mitochondrial ataxia in a Chinese population by targeted NGS, which broadens the clinical spectrum of mtDNA mutations and highlights the importance of screening mtDNA and nDNA mutations among undefined ataxia patients.

1. INTRODUCTION

Ataxia is a heterogeneous group of disorders with multiple types characterized by progressive incoordination of gait, eye movements, and speech, and mostly associated with atrophy of the cerebellum.1 According to the current classification, ataxia etiology can be divided into 3 categories, including acquired, hereditary, and nonhereditary degenerative.2 The hereditary ataxias can be inherited in autosomal dominant, autosomal recessive, and X‐linked mode. Thus far, more than 50 disease‐causing genes have been identified to be associated with ataxia of which the CAG repeat variant of ATXN3 is the most common cause of autosomal dominant ataxia worldwide.1

Previous research has demonstrated that mitochondrial DNA (mtDNA) or mitochondrion has been implicated in spinocerebellar ataxia either as primary or as predisposing factors.3, 4 Firstly, mitochondrial disorders are a group of genetic disorders affecting multisystem due to mutations in mtDNA or mitochondrion‐related nuclear genes (nDNA). Ataxia is one of the most common symptoms as a main feature or a part of syndromes in mitochondrial disorders, including myoclonic epilepsy with ragged red fibers (MERRF), neuropathy, ataxia, and retinitis pigmentosa (NARP) and Kearns‐Sayre syndrome (KSS).5, 6 It is usually in combination with multisystem symptoms and signs, present in isolation in a few cases. Two literatures have previously hypothesized that mtDNA analysis should be considered in unexplained sporadic or maternally inherited ataxia.7, 8 Pfeffer G et al8 demonstrated that mutations of ATP6 may be relatively common in previously unclassified spinocerebellar ataxia. However, Lee YC et al9 failed to survey any common mtDNA point mutations in a group of 265 Taiwanese patients to confirm the hypothesis. As only a few common mtDNA mutations were explored in adult‐onset ataxia, its prevalence is probably underestimated. In addition, many nuclear genes associated with mitochondrial function may be responsible for ataxia. For example, POLG, encoding the catalytic subunit of polymerase γ in the mtDNA replication and repair, has been demonstrated a common causative gene of adult‐onset ataxia in Europe.10, 11

Recently, the targeted next‐generation sequencing (NGS) technology approach has made it possible to parallel sequence the whole mtDNA and nuclear genomic regions of interest. Owing to its adequate sequencing depth, targeted NGS can characterize heteroplasmy variants with a very low mutant load. The aim of this study was to clarify the role of mtDNA and nDNA mutations in Chinese patients with ataxia and clinical features of patients with mitochondrial ataxia. Here, we searched for mtDNA and nDNA mutations using targeted NGS technology in 33 molecularly unassigned sporadic or hereditary ataxia patients. Finally, 5 mutations in mtDNA and nDNA were found, and the genetic and clinical features of individuals with mitochondrial ataxia were characterized.

2. MATERIALS AND METHODS

2.1. Subjects

The medical records of ataxia patients referred to Department of Neurology in Second Hospital affiliated to Zhejiang University School of Medicine and Huashan Hospital affiliated to Fudan University between September 2008 and September 2016, serving primarily the population of Southeastern China, were reviewed and patients with suspected genetic causes of unknown etiology were identified. Inclusion/exclusion criteria were as follows: First, all patients were initially required to have classical ataxia findings (such as limb or truncal ataxia, dysmetria, or ocular findings), subacute or chronic duration of symptoms, and progressive time course. Second, routine clinical investigations (history of drinking, smoking, and drug abuse; history of intake of anticonvulsant drugs; history of venereal disease and malignancies; laboratory tests in thyroid function, folic acid, vitamin B12 and E levels; no ischemia, hemorrhage, or tumor of cerebellum and spinal cord) were collected to exclude causes of acquired ataxia. Third, genetic testing for CAG repeat‐associated genes was performed to exclude SCA1, 2, 3, 6, 7, 12,17, DRPLA as our previous reports.12, 13 Fourth, the remaining patients who met the criteria of pedigree analysis were included as follows: (i) recessive transmission: independent pedigrees with single generation affected (n = 13); (ii) dominant transmission: independent pedigrees with no paternal transmission but at least one documented maternal transmission (n = 7); and (iii) sporadic cases: independent pedigrees with early‐onset age before 40 years without similar disorders in first‐ and second‐degree relatives (n = 13). Each participant signed a written informed consent and the study was approved by the local Ethics Committees.

2.2. Genetic analysis

DNA was extracted from the peripheral blood using Blood Genomic Extraction Kit (Qiagen, Hilden, Germany). A customized panel was designed to cover 46 known causative genes, including 16 autosomal dominant ataxia genes, 24 autosomal recessive ataxia genes, and 6 X‐linked hereditary ataxia genes previously reported (Table S1). For those subjects without mutations in hereditary ataxia genes, mtDNA and nDNA genes were screened with a panel covering mitochondrial 16 569 base pairs (NC_012920.1) and 56 nuclear genes involved in mitochondrial disorders (Table S2). The targeted NGS and data analysis were performed using our previously reported protocol.14, 15, 16 Sanger sequencing was used to validate the variants that passed the filtering criteria in the patients and their family members. The potential sequence variants in nuclear genome identified were classified according to the American College of Medical Genetics and Genomics (ACMG) standards and guidelines.17

For classification of mtDNA variants, we uploaded a serial of mtDNA variants into MITOMASTER, a mtDNA variant query system in MITOMAP (http://www.MITOMAP.org/MITOMAP), to annotate all mtDNA variants, their gene locations, variant frequency, phenotype reports, haplogroup, and evolutionary conservation according to the standard protocol.18 To identify disease‐causing or disease‐associated variants from polymorphisms, these variants were further filtered using the following criteria19: (i) finding out mtDNA variants reported in the literature as mutation or listed in MITOMAP as disease‐causing mutation, further classifying those reported in multiple patients with clinical correlation and/or functional studies into confirmed disease‐causing mutation category; (ii) finding out mtDNA variants reported in the genetic association studies or listed in MITOMAP with influence on phenotypes such as penetrance of certain disease, further excluding high‐frequency (≥1% in MITOMAP) variants and noncoding or synonymous variants, classifying those rare nonsynonymous variants in protein‐coding regions and mitochondrial rRNA, tRNA variants into disease‐associated variant category. MitImpact 2 (version 2.4) is used to extract pathogenicity predictions of mitochondrial nonsynonymous variants.20 PON‐mt‐tRNA helps us to classify mitochondrial tRNA variants.21

3. RESULTS

3.1. Clinical features of 33 patients with ataxia

The study patients consisted of 18 males and 15 females. The personal and medical histories of all patients are summarized in Table 1. The age at onset ranged from 2 to 65 years, with the average of 29 years. Characteristics of ataxia were observed in all patients. Cerebellar atrophy based on brain magnetic resonance imaging (MRI) was observed in 81.82% (n = 27/33) while pure cerebellar atrophy in 48.48% (n = 16/33) of patients. Participants with other manifestations of MRI account for 33.33% (n = 11/33), including brain stem or cerebral atrophy in 5 patients, leukoencephalopathy in 3 patients, and cerebral infarction in 3 patients. Among noncerebellar features, cognitive disorders or dementia were observed in 21.21% (n = 7/33), epilepsy in 6.06% (n = 2/33), and extrapyramidal symptom in 15.15% (n = 5/33) of patients.

Table 1.

Clinical features of 33 patients with ataxia in this study

Demographic information
Gender (n; male/female)	33 (18/15)
Age at onset (y)	29 (2‐65)
Family history(n; inherited/sporadic)	33 (20/13)
Symptom onset (%)
Gait imbalance	84.9
Slurred speech	24.2
Upper limb trembling or clumsiness	18.2
Impaired vision	9.1
Epilepsy	3.0
Clinical features (%)
Ataxia	100
Dysarthria	84.9
Dysphagia	42.4
Nystagmus	45.5
Extrapyramidal symptom	15.2
Cognitive impair	21.2
Epilepsy	6.1
Autonomic nervous dysfunction	6.1
Optic atrophy	6.1
MRI (%)
Normal	18.2
Pure cerebellar atrophy	48.5
Cerebellar atrophy with other signs	33.3

MRI, magnetic resonance imaging.

3.2. Genetic findings

Targeted sequencing covered 93.9% of the target bases with at least 30× per individual, with a mean coverage of 160×. None pathogenic mutation in hereditary ataxia genes was identified in the included 33 patients. A total of 4 disease‐causing variants in mtDNA (Table 2, Figure 1), including 3 confirmed variants (tRNA ^Lys: m.8344A>G, ATP6: m.9176T>C, m.9185T>C) and one reported in a single literature as pathogenic (ND1: m.3995A>G),22 were detected in 4 unrelated patients. In addition, one known pathogenic mutation, c.1159_1162dupAAGT (p.Ser388Terfs) in a nuclear gene (PDHA1, NM_ 000284.3), was identified in an intermittent ataxia patient (Figure 2). This is a de novo mutation. Table 3 summarizes 7 different reported disease‐associated variants found in 5 of 33 patients, including 4 variants in tRNA genes and 3 in structural genes.

Table 2.

Clinical features and molecular findings in 4 patients with mtDNA mutations

Proband (gender)	AAO (y)	Family history	First symptom	Clinical features	MRI	Pathogenic variants	Haplogroup	Previously reported disease
29(F)	39	Affected siblings	Gait imbalance	Gait imbalance, dysarthria, nystagmus, tendon areflexia	Cerebellar atrophy	m.8344A>G	C4a	MERRF
3(M)	12	Affected mother	Gait imbalance	Cerebellar ataxia, nystagmus, muscle weakness, tendon areflexia, Babinski sign (+)	Normal	m.9176T>C	B4c	LS, FBSN, CMT
26(M)	15	Sporadic	Gait imbalance	Cerebellar ataxia, cognitive impairment, spasticity, tendon hyperreflexia, Babinski sign (+)	Normal	m.9185T>C	A13	LS, NARP
21(M)	18	Affected siblings	Gait imbalance	Gait imbalance, cognitive impairment, tendon hyperreflexia	Cerebellar vermis agenesis	m.3995A>G	C7a	MELAS

F, female; M, male; AAO, age at onset; MRI, magnetic resonance imaging; MERRF, myoclonic epilepsy with ragged red fibers; CMT, Charcot‐Marie‐Tooth disease; LS, Leigh syndrome; NARP, neuropathy, ataxia, and retinitis pigmentosa; FBSN, familiar bilateral striatal necrosis; MELAS, mitochondrial encephalomyopathy with lactic acidosis and stroke‐like episodes.

Figure 1.

Pedigrees and chromatograms of mtDNA variants. A‐D, Sequencing chromatograms of variants in tRNA^{L ys}, ATP6, and ND1, respectively. E‐H, Pedigree of 4 families with known mtDNA mutations

Figure 2.

Pedigree and chromatograms of PDHA1 mutation. A, The pedigree of a family with PDHA1 mutation. B, Chromatograms showing the mutation in PDHA1

Table 3.

Previously reported disease‐associated variants identified in our ataxia patients

Proband (gender)	AAO (y)	Family history	Clinical features	Variant	Locus	Frequency MITOMAP	Predicted Impacta	Previously reported disease
5(F)	3	Affected siblings	Cerebellar ataxia, tendon areflexia, cavus foot	m.4386T>C	tRNA Gln	0.37 %	N, NA, NA	CPEO, MM, CHD
15(F)	39	Affected siblings	Gait imbalance, cerebellar atrophy	m.5628T>C	tRNA Ala	0.19 %	LP, NA, NA	CPEO
19(M)	9	Sporadic	Cerebellar ataxia, seizure, tendon hyperreflexia	m.5802T>C	tRNA Cys	0	N, NA, NA	Deaf
15(F)	39	Affected siblings	Gait imbalance, cerebellar atrophy	m.5821G>A	tRNA Cys	0.60%	N, NA, NA	CHD
7(F)	52	Affected siblings	Gait imbalance, dysarthria, nystagmus, cerebellar atrophy	m.12397A>G	ND5	0.50 %	NA, UN, T	PD
22(F)	43	Affected siblings and mother	Gait imbalance, dysarthria, blurred vision	m.13135G>A	ND5	0.97 %	NA, B, T	HCM
19(M)	9	Sporadic	Cerebellar ataxia, seizure, tendon hyperreflexia	m.15497G>A	Cytb	0.43 %	NA, B, T	Obesity

F, female; M, male; AAO, age at onset; NA, unavailable; CPEO, chronic progressive external ophthalmoplegia; MM, mitochondrial myopathy; CHD, coronary heart disease; PD, Parkinson disease; HCM, hypertrophic cardiomyopathy.

^aThe impact of tRNA and nonsynonymous protein‐coding region variants was determined using prediction software (see Materials and methods section) and in the order indicated includes PON‐mt‐tRNA results as neutral (N), likely neutral (LN), pathogenic (P), and likely pathogenic (LP); Polymorphism Phenotyping, version 2 (PolyPhen‐2), results as unknown (UN), benign (B), possibly Damaging (PSD), or probably damaging (PRD); and sorting intolerant from tolerant (SIFT) results as tolerated (T) or deleterious (D).

3.3. Clinical features of patients with mtDNA mutations

Mutations in mtDNA were detected in 3 familial cases and one sporadic case (Table 2). All 4 patients experienced symptoms of ataxia with age at onset ranging from 12 to 39 years (21 ± 12.2) and developed extrapyramidal symptoms during the disease course. Cerebellar atrophy or dysplasia was observed in the proband of family 29 and family 21 but absent in the other 2 cases.

The proband (II1) in family 29 carrying the m.8344A>G mutation is a 49‐year‐old woman who initially suffered gait imbalance 10 years ago. She gradually presented with mild dysarthria and movement retardation of upper limbs in the past 2 years. On the neurological examinations, she displayed horizontal nystagmus, and abnormal finger‐to‐nose and heel‐knee‐shin tests. The muscle strength of 4 limbs was grade 4. Deep tendon reflexes were decreased in the lower limbs. Brain MRI revealed cerebellar atrophy (Figure 3A). She had a Scale for the Assessment and Rating of Ataxia (SARA) score of 8 and International Cooperative Ataxia Rating Scale (ICARS) score of 27. Her younger sister (II3) and younger brother (II4) also presented similar, but more severe symptoms. Her younger brother (II4) developed ataxia and dysarthria in his 20s and then died in his 40s. Unfortunately, her younger sister and parents refused genetic investigation.

Figure 3.

Brain magnetic resonance imaging (MRI) features of patients with mitochondrial ataxia. A, For proband II1 in family 29, sagittal T2‐weighted image shows mild atrophy of cerebellum. B‐C, For proband II2 in family 21, axial and sagittal T2‐weighted images show cerebellar vermis atrophy with enlargement of the fourth ventricle. D, For proband II1 in family 21, sagittal T2‐weighted image shows evident atrophy of cerebellum. E‐H, For proband II2 in family 33, axial T2‐weighted images show symmetric high signal in the bilateral pallidal (E) and dentate nuclei (F), and the signal recovered to normal nearly between episodes (G and H)

The proband (II2) from family 26 carrying the m.9185T>C mutation experienced involuntary trembling of upper limbs and unsteady gait at the age of 15 years. His parents complained that his learning and communication ability lagged behind typically developing peers. Physical examinations manifested bilaterally increased deep tendon reflex in the lower limbs but decreased deep tendon reflex in the upper limbs and bilaterally positive Babinski sign. Finger‐to‐nose and heel‐knee‐shin tests were abnormal on the left. No remarkable changes were found in the brain MRI. No history of neurological disease was documented in his family.

The index case (II1) of family 3, carrying the m.9176T>C, is a 35‐year‐old man who started the disease with gait imbalance and dysarthria at the age of 12 years. Neurological examinations revealed obvious horizontal and vertical nystagmus, decreased muscle strength, and hyporeflexia in the lower limbs. Babinski signs were bilaterally positive. Brain MRI was normal. His affected mother (I1) developed progressive gait unsteadiness, rigidity, and was diagnosed with Parkinson's disease (PD) in the local hospital.

In family 21, the m.3995A>G mutation was detected in the proband (II2) who had a positive family history of ataxia. His symptoms began initially with incoordination of gait and involuntary movement of upper limbs at the age of 18 years. Two months later, the tremor symptoms progressed to the lower limbs and chin. He became unwilling to communicate with people. Examination of cranial nerves showed normal. Strength was 4/5 in the upper limbs and 5/5 in the lower limbs. Deep tendon reflexes were hyperactive in the lower limbs. Ataxia signs of limbs and trunk were noted, including positive finger‐to‐nose and heel‐knee‐shin tests. As shown in Figure 3B‐C, the brain MRI showed the cerebellar vermis agenesis. Basic lactic acid level from the blood was 1.4 mmol/L (reference range 0.7‐ 2.1 mmol/L), immediate after exercise was 21.6 mmol/L (reference range 0.7‐2.1 mmol/L), and after 10 minutes rest was 1.7 mmol/L (reference range 0.7‐2.1 mmol/L). His sister (II1) suffered similar gait imbalance and cerebellar atrophy (Figure 3D).

3.4. Clinical features of patients with PDHA1 mutation

The proband (II1) from family 33 identified as a patient of the p.Ser388Terfs mutation within PDHA1 is a 6‐year‐old boy (Figure 2). He presented delayed motor development after birth and could not walk by himself until 20 months. He suffered 3 episodes of ataxia since the age of 3 years. Most attacks followed acute infection, lasting a few hours to days, and he was normal between the episodes. A mild elevation of blood lactate was noted during the episode (6.0‐7.8 mmol/L; reference range 0.7‐2.1 mmol/L). Examinations showed an ataxic gait and dysmetria on finger‐to‐nose and heel‐knee‐shin tests. Brain MRI demonstrated hyperintense signal in the dentate nuclei and bilateral pallidal in axial T2‐weighted images with preserved cerebral and cerebellar structure during the episode, and the signal recovered to normal nearly between episodes (Figure 3E‐H).

4. DISCUSSION

Mitochondrial mutations have been rarely reported in patients with ataxia. Chinnery PF et al7 screened the two most common mtDNA mutations, m.3243A>G and m.8344A>G, in 29 hereditary and 54 sporadic spinocerebellar ataxia cases, and found 2 patients carrying m.3243A>G and m.8344A>G, respectively. Pfeffer G et al8 reported 2 families carrying mutations in the mitochondrially encoded ATP6 gene after genetic screening of ATP6 gene in 64 ataxia pedigrees. As only the most common mutations were investigated in these studies, some other deleterious mutations in mtDNA or nuclear genome may be ignored. Therefore, we suggest to use targeted NGS technology to screen the whole mtDNA and nuclear variants simultaneously in unexplained hereditary ataxia patients or sporadic patients with early‐onset ages.

The targeted NGS technology has been successfully performed in molecular diagnosis of several hereditary diseases in our previous studies.14, 15, 16 This technology allows us to find hundreds of mitochondrial variants, and each patient may harbor an average of 40 mtDNA variants. It is a great challenge to interpret potential pathogenicity of these variants. Because there are no golden criteria on classifying mtDNA variants, we have classified our variants based on database search and silico prediction in the present study.

Here, we used this technology to identify the mtDNA and nDNA variants in a cohort of 33 ataxia patients of Chinese descent. We have demonstrated that mtDNA mutations detected in the present study (12.1%,4/33) are more common than the previous reports (0%‐10%).7, 8, 9, 23 We detected 4 different mtDNA mutations including 2 mutations, m.8344A>G and m.9185T>C, which had been previously reported in patients with spinocerebellar ataxia.7, 8 The m.9176T>C in ATP6 gene which was present in family 3 had been described previously in Leigh syndrome or Leigh‐like syndrome mainly,24 less frequently in familiar bilateral striatal necrosis,25 hereditary spastic paraplegia (HSP)‐like disorder,26 and Charcot‐Marie‐Tooth (CMT) disease.27 Leigh syndrome caused by m.9176T>C mutation has primarily manifested typical, early‐onset progressive neurodegenerative syndrome with characteristic bilateral symmetric necrotic lesions of gray matter nuclei in the basal ganglia, thalami, brainstem, or cerebellum. The major neurological features include mental retardation, abnormal motor function (hypotonia, dystonia, ataxia, ophthalmoparesis) accompanied by seizures, optic atrophy, and polyneuropathy.28 Leigh syndrome in adolescence or adult is rare and tend to present with atypical features or milder phenotype.29 Symptoms prominently include slowly progressive ataxia, abnormal ocular findings, and extrapyramidal features such as spasticity and dystonia, which remain more stable than early‐onset Leigh patients.29, 30 Our patient from family 3 carrying the m.9176T>C mutation had onset of disease in adulthood, with ataxia as the predominant feature. Although our proband presented clinical features compatible with adult‐onset Leigh syndrome partly, routine clinical diagnostic tests and brain MRI indicated a low level of suspicion for a mitochondrial disease. Follow‐up investigation of characteristic clinical, neuroradioimaging features and respiratory chain enzyme analysis should be performed to help define the diagnosis. The remaining variant, m.3995A>G in ND1 gene, had been reported in a patient with phenotypes mimic mitochondrial encephalomyopathy with lactic acidosis and stroke‐like episodes (MELAS).22 However, no functional study was performed in that study to determine the pathogenicity of this variant. In the present study, the m.3995A>G was found in 2 siblings associated with ataxia. It was noted that their mother carried the same mutation; however, she did not present symptoms of ataxia or other mitochondrial disorders. Nuclear modifying genes or the environmental factors may be involved in occurrence and progression of disease. In vitro functional study should be further performed to determine the consequence of this likely pathogenic mutation.

In addition, we have identified 7 previously documented disease‐associated variants in the rest of the uncertain patients. A homoplasmic variant m.4386T>C in tRNA ^Gln was identified in a female Chinese patient, and this variant has been considered neutral and described in patients with chronic progressive external ophthalmoplegia (CPEO),31 mitochondrial myopathy,32 and coronary heart disease (CHD).33 Two variants (m.5628T>C and m.5821G>A) in the tRNA ^Ala and tRNA ^Cys were observed in a female patient. The m.5628T>C variant was reported to be associated with CPEO and considered pathogenic;34 however, it was also reported as neutral polymorphism in another CPEO patient.35 The second variation in this patient was m.5821G>A, also described in Chinese subjects with CHD and controls.33 The fourth disease‐associated variation occurring in tRNA region was an m.5802T>C variant in a conserved area of the anticodon stem in the tRNA ^Cys gene. It may disturb the tRNA metabolism and increase the penetrance and expressivity of the deafness‐associated 12S rRNA m.1555A>G mutation.36 Two variants (m.12397A>G and m.13135G>A) in the ND5 gene of complex I were observed in 2 female patients, respectively. The m.12397A>G was recently reported to be associated with early‐onset PD and change the production of reactive oxygen species (ROS) and complex I enzyme activity.37 The m.13135G>A was previously detected with a significantly higher frequency in hypertrophic cardiomyopathy (HCM) individuals than in unaffected family individuals.38 Finally, one missense variant in the Cytb gene, m.15497G>A, was significantly associated with obesity, and A allele of this polymorphism may be one of the important determinants of obesity.39 We believe that these variants are more or less associated with ataxia investigated in the present study, and the effect of each variant should be determined in a larger‐scale mtDNA mutation screening among different populations and regions.

In addition to point mutations, mitochondrial diseases also arise from large‐scale mtDNA fragment rearrangement or from nuclear genes defects involved in mtDNA maintenance predisposing to mtDNA depletion. Single large‐scale deletions of mtDNA are typically documented in cases of CPEO, KSS, and Pearson's marrow‐pancreas syndrome (PS).40, 41, 42 The mtDNA depletion, reduction in cellular mtDNA copy number, is characterized in mtDNA depletion syndromes (MDDSs), a genetically and clinically heterogeneous group of disorders caused by molecular defects in at least 9 nuclear genes involved in mtDNA biosynthesis and the maintenance of the deoxynucleotide (dNTP) pools.43 Although rare, ataxia has been reported in patients with large‐scale mitochondrial DNA deletions or mtDNA depletion syndromes caused by mutations in POLG, C10orf2, and TYMP genes.44, 45, 46 Mitochondrial DNA deletion/depletion load varies widely among different tissues, and there will be a challenge for molecular analysis if mutations manifest mainly or solely in a specific organ. In generally, DNA from urinary sediment and skeletal muscle had the higher proportion of mutant genomes than blood, which has been proven not sensitive for detecting mtDNA depletion.47 As mutated mtDNA may be undetectable in blood cells, muscle biopsy may be necessary. Owing to our limitations in targeted NGS technique and unavailability of muscle DNA, we did not perform direct detection of mtDNA large‐scale structure rearrangement and mtDNA depletion. Nevertheless, our NGS panel included 8 nuclear genes (TK2, POLG, C10orf2, TYMP, DGUOK, RRM2B, SUCLA2, and SUCLG1.) reported to cause mtDNA depletion syndrome, and no mutations were detected among the scope of this panel.

Besides mtDNA mutations, we also found one mutation within nDNA. One known PDHA1 mutation, c.1159_1162dupAAGT (p.Ser388Terfs), was found in a case with intermittent ataxia. The p.Ser388Terfs variant was included in Human Gene Mutation Database (HGMD, Accession Number CI941903) and regarded as pathogenic according to ACMG standards and guidelines (one piece of very strong pathogenic evidence—PVS1—and one piece of strong pathogenic evidence—PS2).17 PDHA1, encoding the E1‐alpha polypeptide of the pyruvate dehydrogenase (PDH) complex in mitochondrion, is the disease‐causing gene of pyruvate dehydrogenase E1‐alpha deficiency (PDHAD), which has a broad clinical spectrum, ranging from metabolic acidosis in the newborn to slowly progressive Leigh syndrome. There are also rare neurological problems such as intermittent ataxia, episodic weakness, exercise‐induced dystonia, and recurrent demyelination.48 This mutation was first reported in a child with a mild metabolic acidosis and seizures,49 but involved in intermittent ataxia in our study. To our knowledge, pyruvate dehydrogenase deficiency presenting as intermittent ataxia is rarely reported. Our findings expand the phenotype of pyruvate dehydrogenase deficiency.

In summary, this study identified 4 different mtDNA mutations and one known duplication mutation in PDHA1 gene in 5 of 33 Chinese ataxia patients, which broaden the phenotype of mtDNA mutations. The results suggest that screening of mtDNA and nDNA genes should be performed via targeted NGS technology in maternally inherited or sporadic ataxia patients without mutations of ataxia genes.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

Supporting information

 

 

Dong H‐L, Ma Y, Li Q‐F, et al. Genetic and clinical features of Chinese patients with mitochondrial ataxia identified by targeted next‐generation sequencing. CNS Neurosci Ther. 2019;25:21–29. 10.1111/cns.12972