Copyright 2008 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2008
Although a molecular diagnosis is possible in most patients having Charcot-Marie-Tooth disease (CMT), recessively inherited and axonal neuropathies still present a diagnostic challenge.
To determine the cause of axonal CMT type 2 in 3 siblings.
Three siblings who subsequently developed profound cerebellar ataxia.
Main Outcome Measures
Muscle biopsy specimen molecular genetic analysis of the POLG1 (polymerase γ-1) gene, as well as screening of control subjects for POLG1 sequence variants.
Cytochrome c oxidase deficient fibers and multiple deletions of mitochondrial DNA were detected in skeletal muscle. Three compound heterozygous substitutions were detected in POLG1.
Even in the absence of classic features of mitochondrial disease, POLG1 should be considered in patients having axonal CMT that may be associated with tremor or ataxia.
The past decade has seen major advances in our understanding of the causes of inherited neuropathies. Duplication of the PMP22 (peripheral myelin protein 22) gene and point mutations in GJB1 (gap junction protein, beta 1) and MPZ (peripheral myelin protein zero) account for approximately 85% of unselected cases, but a growing number of genes have been implicated in the remaining 15% of patients, affecting a diverse group of cellular processes ranging from axonal transport to DNA repair.1 The central role of mitochondria in maintaining peripheral nerve integrity is an emerging theme, as illustrated by pathogenic mutations in MFN2 (mitofusin 2) and DNM2 (dynamin 2) that disrupt organelle fusion and fission.1 Although axonal neuropathies are well recognized in patients having classic multisystem mitochondrial disease, isolated neuropathy is uncommon.2 Herein, we describe a family manifesting an autosomal recessive axonal sensorimotor neuropathy that dominated the clinical phenotype for more than 2 decades. This was due to compound heterozygous mutations in the POLG1 gene, which codes for mitochondrial DNA (mtDNA) polymerase γ.
A 35-year-old man had manifested abnormal gait and pes cavus at age 10 years. When initially seen at age 22 years, his main symptoms were reduced dexterity and sensory disturbance affecting his face, hands, and feet. Findings on physical examination disclosed that he had marked bilateral distal weakness and wasting of the small muscles of the forearms, hands, calves, and feet. His proximal power was normal. There was clawing of the toes on the left foot and minimal tremor of the outstretched hands. Joint position and vibration sense were lost in the toes of both feet. Nerve conduction studies revealed an axonal sensorimotor neuropathy consistent with Charcot-Marie-Tooth disease type 2 (CMT2) (Table). He subsequently developed progressive, predominantly distal muscle wasting and a postural tremor at age 33 years, followed by dysarthria.
Neurological examination at age 35 years (a video is available here) disclosed that he had a prominent side-to-side (no-no) head tremor, slow saccades with upbeat nystagmus on upgaze, and pendular nystagmus on right lateral gaze, with a full range of external ocular movements. His visual fields, acuity, and fundi were normal, but he had cerebellar dysarthria and dysphagia. Limb examination revealed a postural and action tremor, marked distal wasting, dysmetria, and dysdiadochokinesis. He had predominantly distal symmetric limb weakness, absent tendon reflexes, and diminished sensation to all 4 modalities in all 4 limbs. He was unable to walk because of weakness and ataxia.
He was the product of a nonconsanguineous union, with healthy parents aged 63 and 61 years. His 40-year-old brother (II-2) was also initially diagnosed as having CMT2 (Table) but had a less severe clinical course. His 42-year-old sister (II-4) developed similar symptoms at age 9 years, was diagnosed as having CMT2 in her early 20s, and is now unable to walk, with severe ataxia (a video is available here). A 41-year-old brother (II-3) was clinically unaffected. Three half-siblings of the father (data not shown) remain unaffected.
Results of routine blood tests in the proband were normal. Cerebrospinal fluid examination results were unremarkable, except for an elevated lactate level (21.5 mg/dL [to convert to millimoles per liter, multiply by 0.111]). Brain magnetic resonance imaging, echocardiography, and pure-tone audiometry findings were normal. Neuropsychometry revealed no major deficits. Results of genetic testing for chromosome 17 duplication, PMP22, MPZ, and spinocerebellar ataxia types 1, 2, 3, 6, and 7 expansions were negative. A nerve biopsy specimen revealed epineural fibrosis but no evidence of inflammation or amyloid.
Muscle histochemistry findings were consistent with ongoing denervation and reinnervation, with fiber-type grouping. There was a mosaic defect of cytochrome c oxidase affecting 6% of muscle fibers (Figure 1A). Long-range polymerase chain reaction of skeletal muscle DNA revealed multiple deletions of mtDNA (Figure 1B). POLG1 sequencing revealed the following 3 heterozygous substitutions in the proband: c.191C>T in exon 2, c.695G>A in exon 3, and c.2209G>C in exon 13. Segregation analysis in the family (Figure 2) showed that the c.191C>T and c.695G>A substitutions were in cis and inherited from the mother and that the c.2209G>C substitution was inherited from the father. The affected brother (II-2) had the same genotype as that of the proband. The unaffected brother (II-3) had a wild-type sequence.
Skeletal muscle histochemistry and mitochondrial DNA (mtDNA) analysis in the proband. A, Histochemical demonstration of cytochrome c oxidase activity in muscle highlighting a cytochrome c oxidase–deficient muscle fiber (asterisk) among fibers with normal enzyme activity. B, Long-range polymerase chain reaction of skeletal muscle DNA amplifying a 9.9-kilobase (kb) fragment across the major mtDNA arc (nucleotide positions 6249-16215). Lane 1 indicates DNA size marker; lanes 2 and 3, control muscle; and lane 4, muscle from the proband showing multiple mtDNA deletions.
Pedigree showing segregation of POLG1 (polymerase γ-1) gene mutations. N indicates heterozygous substitution; P, the proband; squares, men; circles, women; shaded symbols, affected; and open symbols, unaffected.
The frequency of the POLG1 substitutions was determined by primer extension and by matrix-assisted laser desorption ionization–time-of-flight mass spectrometry(MALDITOF; Sequenom, San Diego, California). The c.191C>T substitution was detected in 90 of 294 control alleles (30.6% [95% confidence interval, 25.4%-36.2%]). The c.695G>A substitution was not detected in 282 control alleles (95% confidence interval, 0%-1.06%). The c.2209G>C substitution was detected in 2 of 666 control alleles (0.3% [95% confidence interval, 0.04%-1.08%]).
Mitochondrial DNA codes for 13 essential components of the respiratory chain that is linked to oxidative phosphorylation, the principal source of adenosine triphosphate within the cell. POLG1 mutations affect the maintenance of mtDNA, causing deletions, point mutations, and depletion of mtDNA. Secondary mtDNA defects lead to a biochemical defect of the respiratory chain in affected tissues, organ dysfunction, and the clinical phenotype.3
The c.695G>A and c.2209G>C substitutions are the likely cause of the autosomal recessive axonal sensorimotor neuropathy in this family. The c.695G>A substitution is predicted to alter the highly conserved arginine 232 to histidine in the exonuclease (proofreading) domain of polymerase γ, and the c.2209G>C substitution is predicted to change the highly conserved glycine 737 to arginine in the linker region of polymerase γ. Both have previously been described in compound heterozygotes having other POLG1 mutations and having multiple mtDNA deletions4,5; they did not cause disease in the heterozygous parents and are uncommon in the general population. The c.191C>T substitution is predicted to cause a serine to leucine substitution at position 64, but this amino acid is located outside the functional domain of the protein, is not as conserved across species, is not positioned close to any reported pathogenic mutations, and was present in approximately 30% of control alleles.3 Therefore, this substitution is unlikely to be directly responsible for the disorder in this family, although we cannot exclude the possibility that the serine to leucine substitution at position 64 contributes to the phenotype, as has been described for other polymerase γ substitutions present at high frequencies in control subjects.4
Axonal neuropathies have been described in patients having dominant and recessive POLG1 mutations,4,6- 10 often causing profound sensory ataxia; however, the neuropathy in each case was associated with additional neurological features, pointing toward a multisystem mitochondrial disorder. By contrast, in all 3 siblings in the present family, peripheral neuropathy dominated the clinical picture for 2 decades. Only later did they develop tremor and ataxia, but without the ophthalmoplegia or dysphagia that characterizes the well-described mitochondrial phenotype of SANDO (sensory ataxic neuropathy with dysphagia and ophthalmoplegia).11 Therefore, our observations add to the broad phenotypic spectrum of POLG1-related disease and suggest that POLG1 should be sequenced in patients having unexplained axonal hereditary neuropathy. Although muscle biopsy may reveal the presence of cytochrome c oxidase–negative fibers, this is not universally the case,12 and POLG1 should be sequenced if there is a strong clinical suspicion.
Correspondence: Patrick F. Chinnery, PhD, FRCP, Mitochondrial Research Group, The Medical School, Newcastle University, Room M41014, Framlington Place, Newcastle upon Tyne NE2 4HH, England (firstname.lastname@example.org).
Accepted for Publication: August 21, 2007.
Author Contributions:Study concept and design: Stewart, Findlay, De Silva, and Chinnery. Acquisition of data: Harrower, Stewart, Hudson, Houlden, Warner, O’Donovan, Findlay, Taylor, and De Silva. Analysis and interpretation of data: Stewart, Hudson, Taylor, and Chinnery. Drafting of the manuscript: Harrower, Stewart, Houlden, and Chinnery. Critical revision of the manuscript for important intellectual content: Harrower, Stewart, Hudson, Warner, O’Donovan, Findlay, Taylor, and De Silva. Obtained funding: Taylor and Chinnery. Administrative, technical, and material support: Harrower, Stewart, Hudson, Houlden, O’Donovan, Findlay, and De Silva. Study supervision: De Silva and Chinnery.
Financial Disclosure: Dr Houlden is a Medical Research Council Clinician Scientist. Dr Taylor receives support from the Wellcome Trust. Dr Chinnery is a Wellcome Trust Senior Fellow in Clinical Science.
Funding/Support: Dr Chinnery receives funding from the United Mitochondrial Diseases Foundation and from the European Union Research Framework Programme EUmitocombat and MITOCIRCLE.
Harrower T, Stewart JD, Hudson G, Houlden H, Warner G, O’Donovan DG, Findlay LJ, Taylor RW, De Silva R, Chinnery PF. POLG1 Mutations Manifesting as Autosomal Recessive Axonal Charcot-Marie-Tooth Disease. Arch Neurol. 2008;65(1):133-136. doi:10.1001/archneurol.2007.4