CLINICAL STUDIES

Both affected members (fig 1, IV.1 and IV.2) were the product of a normal pregnancy and birth and had normal developmental milestones and progress at school until disease onset at the age of 6 years. They developed an ataxic gait that was slowly progressive, and was associated with decreased school performance. At 8 years of age they developed inverted feet (calcaneovarus), which was associated with frequent falls. Around the age of 9 years, both started to have choreiform movements of all four limbs, more marked in the upper limbs than in the lower limbs. By the age of 10 years the condition had progressed and they were unable to walk without assistance, and they left school soon after. Their mother found difficulty in feeding them because of swallowing problems after the age of 10 years and they were unable to dress, bathe, or feed themselves by their mid-teens. There were no visual or auditory symptoms or history of epilepsy. Both parents and four older sibs, two brothers and two sisters, were in good health. The parents were first cousins and came from an inbred family (fig 1). The family lived in Karak, a town in southern Jordan.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

A simplified pedigree of the research family. Affected subjects with Karak syndrome are shown as shaded symbols. The closest link between family members is shown, although other more distant links also exist. The haplotype results for the polymorphic microsatellite markers tel-D20S906-D20S193-D20S116- D20S482-D20S895-cen are shown under each person who was genotyped. The PANK2 gene lies between markers D20S193 and D20S116. The box highlights the allele results of the two affected males and an unaffected sib, all of whom have an identical heterozygous haplotype.

On examination at the ages of 12 and 14 years respectively, the patients were not dysmorphic and were well grown (height, weight, and head circumference between the 10th-50th centiles). No telangiectasia was found on skin or mucus membrane, nor were skeletal abnormalities detected. Pursuit eye movements showed bilateral limitation of upward gaze in both children; saccadic eye movements were abnormal with bilateral hypometric saccades. Visual acuity was normal, as were the pupillary responses, fundoscopy of the optic discs and retina, and slit lamp examination. Both affected children had dysarthric scanning speech with dystonic features. There was dystonic movement of the tongue and facial muscles and choreiform movement was present in both upper and lower limbs, being more marked in the lower limbs, along with dystonic posture of the distal feet. Significant bradykinesia was present in both upper and lower limbs. Muscle bulk was normal with no wasting or fasciculation. Generalised hypertonia was evident in both upper and lower limbs of a mixed spastic-rigid type. Motor examination showed normal strength testing throughout. Evaluation of the different sensory modalities showed no abnormalities. Deep tendon reflexes were +2 and symmetrical, with bilateral extensor plantar responses. Dysmetria, past pointing, dysdiadochokinesia, and intentional tremor were bilateral and symmetrical. The gait was wide based, swayed in all directions, and had a spastic quality. Neurological examination showed evidence of impairment of cognitive functions and on the Wechsler Adult Intelligence Scale they both achieved less than 60 for verbal IQ.

The following investigations were normal in both children: blood count and film, haemoglobin electrophoresis, liver enzymes, lipid profile, urine testing, serum immunoglobulins, alpha-fetoprotein, serum B12, folate, copper, ceruloplasmin, creatinine phosphokinase, very long chain fatty acids, ferritin, urine organic acids (by gas chromatography and mass spectroscopy), audiogram, ECG, cardiac echo, chest radiographs, electromyography, motor and sensory conduction studies including studies of the late responses, and visual and brainstem evoked potentials. The potential differential diagnoses of PKAN, neuroferritinopathy, and aceruloplasminaemia were eliminated by these results. Muscle biopsy from the older affected child was normal with no evidence to suggest a mitochondrial myopathy. MRI studies of the brain of both affected sibs showed a moderate degree of pancerebellar atrophy (fig 2A). T2 weighted images showed markedly decreased signal intensity restricted to the substantia nigra (fig 2B) and the globus pallidus (fig 2C). In the centre of the globus pallidus a small high signal spot was seen (“eye of the tiger” sign) (fig 2C). There was no abnormal signal in the putamen, caudate nucleus, red nucleus, or dentate nucleus. T1 weighted inversion recovery images were not remarkable. The striking low attenuation on T2 weighted findings suggests an increased iron deposition in the globus pallidus, and is an obligatory radiological feature of PKAN caused by PANK2 gene mutations.⁷

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

MRI T2 weighted images of the brain of the older affected sib. (A) A parasagittal view showing a moderate degree of pancerebellar atrophy. (B) Decreased signal intensity in the substantia nigra, indicated by a black arrow. (C) The “eye of the tiger” sign with decreased signal intensity in the globus pallidus, indicated by a black arrow, in the centre of which a small high signal spot was seen, indicated by a white arrow.

MOLECULAR GENETIC STUDIES

We performed linkage analysis between the family’s disease and the PANK2 gene that causes PKAN.² DNA was extracted using standard techniques from blood samples obtained from nuclear family members. Polymorphic microsatellite markers were sought that flanked the PANK2 gene by scrutiny of the Marshfield Linkage Maps, the DeCode genetic map,⁸ and the draft Human Genome Browser: D20S906-D20S193-PANK2 gene-D20S116-D20S482-D20S895 (at the following distances from 20pter: 1.5 Mb-3.3 Mb-3.85 Mb- 4.05 Mb-4.5 Mb-5.05 Mb). PCR amplification of all markers was performed using a Roboseq 4200 (MWG BioTech Ltd). Amplified markers were pooled and electrophoresed on an ABI Prism 377 gene sequencer (Applied Biosystems) on 4.2% polyacrylamide gels, at 3000 V and 52°C, for 2.5 hours. Fragment length analysis was undertaken using the ABI Prism Genescan and Genotyper 1.1.1 analysis packages. Fig 1 shows the family genotyping results. The two affected subjects and one of the unaffected sibs all had the same heterozygous results for the informative markers surrounding the PANK2 gene. This suggests that linkage of the disease to the PANK2 gene is very unlikely. A further implication of the result is that the affected sibs do not have two separate conditions, one PANK2 related and the other causing cerebellar ataxia. The results do not help determine if the condition is X linked or autosomal recessive. We also eliminated linkage to the mitochondrial ferritin gene on chromosome 5q23⁹ (data not shown).

DISCUSSION

The disorder documented here presented at 6 years of age with cerebellar ataxia. Later, extra pyramidal motor features typical of classical PKAN due to PANK2 gene mutation developed and intellectual ability declined. There have been a few case reports in which ataxia is associated with PKAN, but none with ataxia as a presenting feature.^10–12 However, in a large study of PKAN two groups were differentiated; firstly, cases with typical clinical symptoms (of what would have been once called Hallervorden-Spatz syndrome¹³) or atypical features with acanthocytosis, hypoprebetalipoproteinaemia and/or orofacial dyskinesia (known as HARP syndrome^14–16) but all with the “eye of the tiger” sign on MRI brains scan and PANK2 mutations; secondly, a group of clinically atypical cases which did not have PANK2 mutations nor the “eye of the tiger sign” but did have hypodensity in the globus pallidus on T2 weighted MRI brain scans, cerebellar atrophy, and often evidence of iron deposition in the red nucleus and dentate nucleus.⁷ The family we report appears to have a demonstrably different condition from these two phenotypes because of the clinical presentation and the neuroradiological findings. The “eye of the tiger” sign is present as is hypodensity in the substantia nigra, but with no abnormal signal in the putamen, caudate nucleus, red nucleus, and dentate nucleus. Exclusion of linkage to the PANK2 gene provides further support that Karak syndrome is a distinct disorder. In the original paper mapping PKAN due to PANK2, family HS9 was the only unlinked family presented. However, insufficient clinical details are given to determine if this single Japanese male case was similar to the family we report.^17,18

The highest concentration of iron in the brain is found within the basal ganglia (substantia nigra, putamen, caudate nucleus, and globus pallidus), red nucleus, and dentate nucleus. Cerebral iron accumulates with age, particularly in the basal ganglia.¹⁹ While iron is essential for normal brain function, free iron ions impose an oxidative stress on cells, owing to the production of oxygen free radicals, leading to cell damage. Cerebral diseases in which excess iron accumulation is shown on MRI scan are summarised in table 1. All cause clinical features of basal ganglia disease and all exhibit excess iron deposition in some or all of the sites where iron concentrations are the highest in the disease free state. Iron is stored predominantly complexed with ferritin within brain cells, and dysfunction of the ferritin light chain causes neuroferritinopathy.³ Ceruloplasmin is a ferroxidase with roles including ferric iron uptake by transferrin (which transports iron in the ferric state in the circulation) and dysfunction leads to aceruloplasminaemia.⁶ How mutations in the PANK2 gene, a brain specific CoA biosynthesis regulatory enzyme, cause pathological iron accumulation is unclear, but may relate to an accumulation of cysteine, which can chelate iron. A potential common pathogenic mechanism for neurological disease involving excess cerebral iron accumulation would be the release of free iron ions, either in excess or in a non-physiological situation or both.²⁰ Because of this we sought, but did not find, linkage between the mitochondrial specific ferritin gene and Karak syndrome.

The inheritance pattern of Karak syndrome may be autosomal recessive or X linked recessive. The lack of other affected males in the matrilineal line and parental consanguinity favour recessive inheritance but clarification awaits further reports of the condition. We hypothesise that the disorder is novel and a member of the growing family of neurological disease involving excess pathological cerebral iron accumulation. Potentially, the gene mutated in Karak syndrome could be involved in iron sequestration, transport, storage, ferrous/ferric status control, intercalation of iron into enzymes, and oxidative respiration.