Cerebellum DOI 10.1007/s12311-013-0538-z
ORIGINAL PAPER
A Novel Frameshift Mutation in the AFG3L2 Gene in a Patient with Spinocerebellar Ataxia Zuzana Musova & Michaela Kaiserova & Eva Kriegova & Regina Fillerova & Peter Vasovcak & Alena Santava & Katerina Mensikova & Alena Zumrova & Anna Krepelova & Zdenek Sedlacek & Petr Kanovsky
# Springer Science+Business Media New York 2013
Abstract Spinocerebellar ataxia type 28 (SCA28) is an autosomal dominant neurodegenerative disorder caused by missense AFG3L2 mutations. To examine the occurrence of SCA28 in the Czech Republic, we screened 288 unrelated ataxic patients with hereditary (N =49) and sporadic or unknown (N =239) form of ataxia for mutations in exons 15 and 16, the AFG3L2 mutation hotspots. A single significant variant, frameshift mutation c.1958dupT leading to a premature termination codon, was identified in a patient with slowly progressive speech and gait problems starting at the age of 68 years. Neurological examination showed cerebellar ataxia, mild Parkinsonian features with predominant bradykinesia, Z. Musova (*) : P. Vasovcak : A. Krepelova : Z. Sedlacek Department of Biology and Medical Genetics, Charles University 2nd Faculty of Medicine and University Hospital Motol, Prague, Czech Republic e-mail: [email protected] M. Kaiserova : K. Mensikova : P. Kanovsky Department of Neurology, Palacky University Olomouc Faculty of Medicine and Dentistry and University Hospital, Olomouc, Czech Republic E. Kriegova : R. Fillerova Department of Immunology, Palacky University Olomouc Faculty of Medicine and Dentistry and University Hospital, Olomouc, Czech Republic A. Santava Department of Medical Genetics and Foetal Medicine, Palacky University Olomouc Faculty of Medicine and Dentistry and University Hospital, Olomouc, Czech Republic A. Zumrova Department of Child Neurology, Charles University 2nd Faculty of Medicine and University Hospital Motol, Prague, Czech Republic Present Address: P. Vasovcak Gendiagnostica, Kosice, Slovak Republic
polyneuropathy of the lower limbs, and cognitive decline. However, other common SCA28 features like pyramidal tract signs (lower limb hyperreflexia, positive Babinski sign), ophthalmoparesis or ptosis were absent. The mutation was also found in a patient’s unaffected daughter in whom a targeted examination at 53 years of age revealed mild imbalance signs. RNA analysis showed a decreased ratio of the transcript from the mutated AFG3L2 allele relative to the normal transcript in the peripheral lymphocytes of both patients. The ratio was increased by puromycin treatment, indicating that the mutated transcript can be degraded via nonsense-mediated RNA decay. The causal link between the mutation and the phenotype of the patient is currently unclear but a pathogenic mechanism based on AFG3L2 haploinsufficiency rather than the usual dominantnegative effect of missense AFG3L2 mutations reported in SCA28, cannot be excluded. Keywords SCA28 . Spinocerebellar ataxia . AFG3L2 . Frameshift mutation
Introduction Spinocerebellar ataxias (SCAs) form a heterogeneous group of autosomal dominant disorders characterized by cerebellar ataxia frequently accompanied by additional neurological symptoms which differ among subtypes and also among and within affected families. Currently, at least 30 subtypes of SCA and 20 causative genes have been identified [1–3]. Most SCAs are caused by expansions of polyglutamine-coding CAG repeats (SCA types 1–3, 6–7, 17, and dentatorubropallidoluysian atrophy (DRPLA)). Rarer SCAs result from expansions of noncoding nucleotide repeats (SCA types 8, 10, 12, and 36), insertion of a pentanucleotide repeat (SCA 31) or conventional mutations (SCA types 5, 11, 13, 14, 15/ 16/29, 19/22, 20, 27, 28, and 35) [1–3].
Cerebellum
Missense mutations of one allele of the AFG3L2 gene cause SCA type 28 (SCA28, MIM#610246) [4–6] which represents about 1.5 % of SCAs in the European population [7, 8]. Clinical symptoms of SCA28 appear mostly in early adulthood but the age of onset varies between 3 and 60 years [4–8]. SCA28 is characterized by slowly progressive gait and limb ataxia, dysarthria, ptosis, ophthalmoparesis, gaze-evoked nystagmus, hyperreflexia of the lower limbs, and decreased vibration sense. Extrapyramidal signs (Parkinsonism or dystonia) and cognitive impairment are also occasionally present [4–8]. Homozygous missense AFG3L2 mutations lead to spastic ataxia-neuropathy syndrome [9]. The AFG3L2 gene located in 18p11 contains 17 exons [10] and encodes a subunit of m-AAA mitochondrial proteases. These proteases are involved in protein quality control, regulation, and chaperonelike activities in the mitochondrial proteome [11, 12]. The AFG3L2 subunits assemble into different m-AAA complexes: homo-oligomeric complexes consist solely of AFG3L2, and hetero-oligomeric complexes consist of AFG3L2 and paraplegin, which is encoded by the SPG7 gene [13]. SPG7 mutations cause spastic paraplegia type 7 (SPG7) which is mostly autosomal recessive, although some mutations show a possible dominant effect. SPG7 is frequently associated with cerebellar ataxia [14–16]. AFG3L2 and SPG7 are ubiquitously expressed [10] with high-selective expression in cerebellar Purkinje cells (PCs) and large neurons of the deep cerebellar nuclei; in addition, paraplegin is also enriched in spinal motor neurons [5]. AFG3L2 supports mitochondrial protein synthesis and PC survival [17], and AFG3L2 deficiency reduces mitochondrial calcium uptake capacity by fragmentation of the mitochondrial network [18]. We report here on a patient with a late onset of cerebellar ataxia with Parkinsonian features at the age of 68 years who carried a novel frameshift mutation in the AFG3L2 gene. Although the causality of the mutation for the phenotype is unclear, the observation of a reduced dose of the mutated transcript allows us to speculate that a mechanism based on AFG3L2 haploinsufficiency rather than the usual dominantnegative effect of missense AFG3L2 mutations could potentially play a role in the pathogenesis of the patient’s phenotype.
Material and Methods Patient Cohort To establish the incidence of SCA28 in the Czech Republic, we screened 288 unrelated patients with ataxia for mutations in the AFG3L2 gene. The patients were previously negatively tested for SCA types 1–3, 6–8, 12, 17, DRPLA, and the fragile X-associated tremor/ataxia syndrome. The cohort included 49 patients with a positive family history (out of these, 21 patients
were from families with a clearly autosomal dominant pattern of inheritance), 96 sporadic patients, and 143 patients with unspecified family history. A control group of 104 unrelated healthy Czech individuals was also assessed. Clinical Description of Patients 1 and 2 Patient 1 noticed slowly progressive speech and gait problems at the age of 68 years; until then, she suffered only from glaucoma and coxarthrosis. The first neurological examination at the age of 70 showed dysarthria, bilateral limb ataxia (which was more pronounced on the lower limbs), gaze-evoked bidirectional nystagmus, and uncertain wide-base stand and gait. Neither extrapyramidal and pyramidal signs nor ophthalmoparesis were present. Brain magnetic resonance imaging (MRI) showed isolated cerebellar atrophy. At the age of 71, she developed mild Parkinsonian features with predominant bradykinesia, and gradually also, a mild cognitive decline. She showed no prominent postural instability with falls. The last detailed examination of patient 1 was performed at the age of 76. She showed progression in cerebellar symptoms, slow saccadic eye movements (but neither nystagmus nor ophthalmoparesis were observed), decreased reflexes, and vibration sense in the lower limbs and mild bradykinesia. The Parkinsonian symptoms did not respond to levodopa treatment. Electromyography revealed mild distal axonal sensorimotor polyneuropathy of the lower limbs. Extensive laboratory testing to reveal the cause of polyneuropathy yielded no explanation: there was no evidence of diabetes mellitus, hypothyroidism, vitamin B12 deficiency, or renal insufficiency; serum activities of liver enzymes and carbohydrate-deficient transferrin were normal and anti-ganglioside antibodies and borrelia antibodies in the cerebrospinal fluid were negative. Brain MRI showed cerebellar and diffuse supratentorial atrophy and small white matter hyperintensities (Fig. 1). The Mini Mental State Examination score was 21 points out of 30. A detailed neuropsychological examination revealed global deterioration of cognitive functions. The degree of dementia was mild. Depression was also present. Both the brain MRI findings and the results of neuropsychological testing were not typical for any common type of dementia (e.g., Alzheimer’s disease or vascular dementia). The patient did not suffer from urinary incontinence or urgency, and there was no significant orthostatic blood pressure decrease after 3 min of standing which would fulfill the criteria of orthostatic hypotension. Patient 1 had four siblings, and two of them (brother and sister) suffered from speech and gait problems. Unfortunately, they refused examination and no detailed clinical data were available. There was no evidence of similar problems in the parents of patient 1, but her mother died of breast carcinoma at the age of 45. The father died of heart failure at the age of 90. Patient 1 had three children. Two of them were not examined neurologically, but they did not complain of any speech or gait
Cerebellum
Fig. 1 Brain MRI images of Patient 1. Sagittal T1-weighted image (left) shows cerebellar atrophy. Transversal FLAIR image (middle) shows a mild diffuse brain atrophy and small white matter hyperintensities. Axial
T2 image through the pons (right) shows no signs typical for patients with multiple system atrophy or progressive supranuclear palsy
difficulties. The remaining daughter of patient 1 (patient 2) was examined at her own request, although she did not complain of any neurological symptoms. The examination performed at the age of 53 revealed a mild postural instability (positivity in Romberg standing test) and unstable tandem gait. The cranial nerves were intact. She had neither nystagmus nor any oculomotor abnormality. There were no extrapyramidal or pyramidal signs, and no clinical signs of polyneuropathy. Brain MRI did not show signs of cerebellar atrophy.
Mutation description was based on the reference sequences NM_006796.2 and NM_003119.2 with nucleotide 1 corresponding to A in the initiation codon. The nucleotide changes identified were compared to the NCBI dbSNP database (http://www.ncbi.nlm.nih.gov/snp) and the Exome Variant Server (http://evs.gs.washington.edu/EVS/).
DNA Analysis The AFG3L2 gene testing using high-resolution melting analysis (HRM) was focused on exons 15 and 16, which have been shown to be mutation hotspots [5, 7, 8]. Genomic DNA was isolated from blood using the Gentra Puregene Blood Core Kit (QIAGEN, Gaithersburg, MD, USA). PCR of the AFG3L2 exons 15 and 16 was performed using a published method [7] with adding LCGreen Plus Dye (Idaho Technology, Salt Lake City, UT, USA). HRM analysis of the PCR products was performed using LightScanner (Idaho Technology) according to the instrument manual. PCR products with aberrant melting profiles were purified using SureClean PCR Purification Kit (Bioline, London, UK) and directly sequenced using BigDye Terminator v3.1 Cycle Sequencing kit and an ABI 3130 Genetic Analyser (both Applied Biosystems, Foster City, CA, USA). The remaining 15 exons of the AFG3L2 gene as well as all 17 exons of the SPG7 gene were amplified according to a published protocol [7] and directly sequenced in patients 1 and 2 (in whom an AFG3L2 mutation was identified using HRM) to exclude the possible effects of additional AFG3L2 or SPG7 variants on their phenotype. Possible deletions or duplications in the SPG7 gene were analyzed in patients 1 and 2 using multiplex ligation-dependent probe amplification (MLPA) with SALSA MLPA Kit P213-B1 (MRC Holland, Amsterdam, The Netherlands) according to the manufacturer’s protocol.
RNA Analysis Peripheral blood samples of patients 1 and 2 and a control individual were collected into PAXgene Blood RNA tubes (PreAnalytiX, Hombrechtikon, Switzerland). Total lymphocyte RNA was isolated using PAX Gene Blood RNA Kit (QIAGEN) and reverse transcribed into cDNA using SuperScript One-Step RT-PCR with Platinum Taq System (Invitrogen, Carlsbad, CA, USA) and primers specific for AFG3L2 exons 13/14 and 17 (cAFG3L2-LF (TGGCTTAGAGAAGAAAACGCAG) and cAFG3L2-LR (CTAGTTGGCAACTTTCTCACCC)). PCR products were separated on a 1.8 % agarose gel and sequenced as above with nested primers cAFG3L2-14/15 F (CGCTTTTAAAGGTATCCATCATC) and cAFG3L2-B (CTTGCAGTGGCTTCACTGTAAG). To test the influence of different RNA isolation procedures on the AFG3L2 transcripts, RNA was also isolated from peripheral blood mononuclear cells separated with Ficoll from heparin-stabilized blood of patient 1 using mirVana miRNA kit (Ambion, Austin, TX, USA). An aliquot of the blood sample was pre-treated with puromycin (Sigma-Aldrich, St. Louis, MO, USA) at a concentration of 0.2 mg/ml for 3 h at 37 °C. To quantify the amount of normal and mutated transcripts, RT-PCR products obtained using primers cAFG3L2-B and 6-FAM labeled cAFG3L2-14/15F were analyzed on ABI 3130 Genetic Analyzer. The ratio of peak areas of normal and mutated RT-PCR products was adjusted to the peak area ratio of a PCR product obtained on DNA using primers AFG3L2-15F (6-FAM CCACTAAGGCTGATGAACT) [7] and AFG3L2-15RB (GCACAACTATATTGTTTCACAGCC).
Cerebellum
Results
genomic DNA genomic DNA
A A AG T A AC T C A G A G T G C A A A AG T T A AC T C A G A G T G C
mut:wt 1:1
cDNA (PAX)
A A AG T A AC T C A G A G T G C A A A AG T T A AC T C A G A G T G C
0.33:1
A A AG T A AC T C A G A G T G C A A A AG T T A AC T C A G A G T G C
0.33:1
A A AG T A AC T C A G A G T G C A A A AG T T A AC T C A G A G T G C
0.33:1
cDNA (H0)
A A AG T A AC T C A G A G T G C A A A AG T A AC T C A G A G T G C A
cDNA (H3P)
cDNA (H3)
Patient 1
normal control
HRM analysis of AFG3L2 exons 15 and 16 in 288 ataxic patients and 104 controls revealed the frameshift mutation c.1958dupT in exon 15 (Fig. 2) in one patient (patient 1), and subsequent analysis identified the same mutation also in her daughter (patient 2). The mutation resulted in a premature termination codon (p.Thr654Asnfs*15). The HRM analysis also revealed the c.2175 + 18G>A variant in intron 16 (rs117096851) in seven patients and three controls. The
analysis of the remaining exons of AFG3L2 and all exons of SPG7 in patients 1 and 2 identified several additional known variants which were unlikely to affect the phenotype. The AFG3L2 gene harbored heterozygous variants c.553-95G>A (rs2298542), c.753-55T>C (rs7407640), c.752 + 6C>T (rs8097342), c.1389G>A (p.Leu463=, rs11080572), c.1650A>G (p.Glu550=, rs11553521), and c.*28G>C (rs1129115) in patient 1, and homozygous variants c.1389G>A and c.1650A>G in patient 2. The SPG7 gene harbored heterozygous variants c.286+46C>T (rs11553521), c.618 + 12T>C (rs3803679), c.619-47G>A (rs3935626), c.862-34G>T (rs4785690), c.987 + 5A>G (rs4785691), c.1779+47G>C (rs3794632), c.2063G>A (p.Arg688Gln, rs12960), and c.2292C>T (rs61747711) in patient 1, and a homozygous variant c.862-34G>T in patient 2. No SPG7 deletions or duplications were found in patients 1 and 2 using the MLPA analysis (data not shown). The RT-PCR of AFG3L2 exon 15 in patients 1 and 2 and the unaffected control showed a major band of the expected length in all individuals. Sequencing and fragment analysis showed a decreased ratio of the RT-PCR product from the allele carrying the c.1958dupT mutation relative to the normal allele (0.33:1, Fig. 2) in both patients. The signal from the mutated allele was stronger in the cDNA from a blood sample treated with puromycin (0.87:1, Fig. 2). Real-time RT-PCR experiments failed to show a significantly reduced amount of total AFG3L2 transcripts in peripheral lymphocytes of patients 1 and 2, although their AFG3L2 mRNA levels fell in the lower range of AFG3L2 levels observed in seven controls (data not shown). The analysis of the AFG3L2 protein in peripheral blood mononuclear cells using Western blot was unsuccessful.
Discussion
A A AG T A AC T C A G A G T G C A A A AG T T A AC T C A G A G T G C
0.87:1
Fig. 2 Sequencing electropherograms and fragment analysis curves of exon 15 of the AGF3L2 gene in a normal control and patient 1 showing the heterozygous c.1958dupT mutation in genomic DNA and different preparations of lymphocyte cDNA (PAX PAXgene-stabilized blood; H heparin-stabilized blood; 0 RNA isolation without incubation; 3 RNA isolation after 3-h incubation at 37 °C; P incubation with puromycin). Arrows indicate the duplicated base in sequencing electropherograms and the mutated allele in fragment analysis curves. The ratio of the mutated transcript relative to the normal transcript calculated from peak areas adjusted to the 1:1 ratio in DNA is also shown. The decreased amount of the mutated transcript is clearly evident. The ratio was increased after puromycin treatment
In this study, we screened a heterogeneous group of 288 unrelated patients with ataxia for mutations in exons 15 and 16 of the AFG3L2 gene. A single significant variant, frameshift mutation c.1958dupT (p.Thr654Asnfs*15), was identified in patient 1, a currently 76-year-old female from a family suggestive of autosomal dominant pattern of inheritance. This novel mutation was found neither in 104 unrelated Czech controls, nor in any of the previous studies focused on AFG3L2 mutations in SCA patients and controls [5–8]. The variant was also absent in dbSNP and the Exome Variant Server. The same variant was identified in the daughter of patient 1 who had mild imbalance signs upon examination at the age of 53 years but currently cannot be considered affected (patient 2). The level of the transcript from the mutated allele was reduced in blood lymphocytes of both patients relative to the transcript from the normal allele.
Cerebellum
The c.1958dupT mutation is located in the peptidase M14 domain of the AFG3L2 protein. It has been shown that this domain is critical for the pathogenesis of SCA28. With the exception of the p.N432T mutation in the AAA domain [5], all previously described pathogenic AFG3L2 variants were clustered in the M14 domain [5–9]. However, in contrast to the c.1958dupT mutation in patient 1, all mutations described so far were missense and were proposed to exert a dominantnegative effect [5–9]. The absence of the c.1958dupT mutation in control cohorts and its influence on the relative ratio of transcripts from the mutated and normal AFG3L2 alleles do not represent a clearcut proof that the mutation is pathogenic and causal for the phenotype observed in our patient. We cannot exclude that the phenotype can be completely unrelated to this AFG3L2 mutation or perhaps just modulated by this variant. However, our finding is remarkable as two different, not mutually exclusive mechanisms cannot be excluded in association with this mutation: a reduced dose of the AFG3L2 transcript and protein, and production of a truncated non-functional protein. Our RNA analysis showed a 0.33:1 ratio of the mutated transcript relative to the transcript from the normal allele. This decrease could be partially reverted (to 0.87:1) by treatment of blood with translational inhibitor puromycin prior to RNA preparation, indicating that the mutated transcript was likely degraded via the nonsense-mediated RNA decay (NMD) pathway [19, 20]. Varying NMD efficiency was observed in different cell types and tissues [19], and we can only speculate about the actual level of aberrant transcripts in the site of the highest AFG3L2 expression, the cerebellar PCs [5]. It could be even lower than that observed in blood as NMD in leukocytes has been shown to be less complete compared to other tissues [21]. A reduced dose of AFG3L2 has never been described in SCA28 patients and its phenotypic effect is unclear, but haploinsufficient mouse models exhibit progressive late onset symptoms similar to those in SCA28 patients [22]. Using real-time PCR experiments, we could not prove a significantly reduced total level of AFG3L2 transcripts in lymphocytes of patients 1 and 2, although their transcript levels fell in the lower range of levels observed in the controls. Considering the 33 % output from the mutated allele and 100 % output from normal alleles (i.e., with no compensatory upregulation of the normal allele in the patients), the decrease in the patients should theoretically reach 67 % of the normal level, which can be difficult to prove experimentally. Simultaneously, the dominant-negative effect could also play a role if some residual amount of the truncated AFG3L2 protein lacking a significant part of the M14 domain is present in the cerebellum. The sole replacement of the first amino acid affected by the c.1958dupT mutation has been shown to be critical for the function of AFG3L2 in a patient with missense mutation p.Thr654Ile [7]. The truncated protein could associate with other components of the m-AAA complexes
and render them non-functional. Unfortunately, our efforts to analyze the AFG3L2 protein using Western blot failed. The phenotype of patient 1 was different from the typical SCA28 picture. Especially the late onset of symptoms in patient 1 (68 years) has not been described yet. Cerebellar ataxia, mild Parkinsonism unresponsive to levodopa, and polyneuropathy of the lower limbs in patient 1 can be a part of the clinical spectrum of SCA28, and extensive laboratory testing did not reveal any other etiology. Dementia has been also seen in several isolated cases of SCA28. In contrast, common features of SCA28 like pyramidal tract signs (hyperreflexia of the lower limbs, positive Babinski sign), ophthalmoparesis, or ptosis were absent in patient 1. Her daughter carrying the same mutation, patient 2, did not complain of ataxia at the age of 53, but her neurological examination revealed mild postural instability and unstable gait. Multiple system atrophy (MSA-C) should also be considered in the differential diagnosis of cerebellar ataxia with mild Parkinsonism. However, in patient 1, no autonomic failure (i.e., urinary problems or orthostatic hypotension) was present, and therefore, the clinical criteria of MSA-C were not met. The absence of both vertical supranuclear gaze palsy and prominent postural instability with falls in the first year of onset argued also against the diagnosis of another atypical Parkinsonian syndrome, the progressive supranuclear palsy (PSP). Moreover, brain MRI in patient 1 showed “only” cerebellar and diffuse supratentorial atrophy and small white matter hyperintensities. There were no typical signs of MSA (“hot cross bun” sign in the pons) or PSP (midbrain atrophy with the “standing penguin silhouette” sign). The AFG3L2 protein is present together with the SPG7 gene product paraplegin in m-AAA protease complexes in the mitochondrial membrane [13]. A clear connection between different genetic defects in the AFG3L2 and SPG7 genes is evident also on the clinical level. Mutations of both alleles of SPG7 usually present with spastic paraparesis, optic neuropathy (or abnormalities in optical coherence tomography), and cerebellar ataxia. Cerebellar ataxia occurs mostly in patients with biallelic null SPG7 mutations [16]. Recently, several patients with compound heterozygous SPG7 mutations (p.Arg485_Glu487del, p.Ala510Val) have been described, who showed cerebellar signs and cerebellar atrophy on MRI, peripheral neuropathy, and no spasticity of the lower limbs [15]. Homozygous AFG3L2 mutations cause the spasticataxia-neuropathy syndrome [9]. To exclude possible biallelic or digenic effects in patients 1 and 2, we sequenced their complete AFG3L2 and SPG7 genes, but we did not reveal any additional significant variants which could have an impact on their phenotype. A reduced dose of the AFG3L2 gene could also be expected in patients with large 18p deletions involving the AFG3L2 locus. These patients showed different phenotypes and various malformations; however, cerebellar ataxia was not
Cerebellum
described among the symptoms [23]. Up to now, just a single exonic AFG3L2 deletion (involving exons 10–11), which may affect the expression of AFG3L2, has been reported in a healthy Korean individual [24]. A single occurrence of two rare variants in the AFG3L2 gene, which could possibly have a similar effect like the frameshift mutation in our patients, was also reported in healthy controls in dbSNP and the Exome Variant Server. The c.1651C>T variant (rs149121681) leads to a premature stop codon (p.Arg551*), and the c.2028_2029insA variant causes frameshift and premature termination (p.Leu677Thrfs*15). Considering the late onset of the disease in the family identified in this study, it is likely that the patients with large 18p deletions as well as the presumably control individuals with the rare nucleotide variants were too young to present the symptoms. Alternatively, there is also the possibility that these variants may be nonpathogenic. Concerning the incidence of SCA28, it was reported to range between 0.7–1.5 % among autosomal dominant cerebellar ataxia patients in Europe [7, 8]. We cannot directly compare our results (1/288 patients) with these data, because our sample contained mostly patients with unspecified family history. In conclusion, we report for the first time a frameshift mutation in the AFG3L2 gene in a patient with late onset of cerebellar ataxia, mild Parkinsonism, and polyneuropathy. If the mutation is indeed pathogenic and causal for the disease, it may point to another possible mechanism contributing to the pathogenesis of SCA28, which is different from the dominant-negative effect of missense mutations described previously, and in which a decreased level of the AFG3L2 transcript could lead to the atypical milder presentation of SCA28. Acknowledgments We thank the patients for their participation in the study, two anonymous reviewers for very valuable comments and Michal Krupka, Ph.D., for his help with Western blot analysis. This work was supported by grant NT-12221 and grant for conceptual development of research organization University Hospital Motol 00064203 from the Ministry of Health of the Czech Republic, and by grant IGA-LF_13_13 from the Internal grant agency of Palacky University. Conflict of interest There are no potential conflicts of interest that might bias this work.
References 1. Duarri A, Jezierska J, Fokkens M, Meijer M, Schelhaas HJ, den Dunnen WF, et al. Mutations in potassium channel kcnd3 cause spinocerebellar ataxia type 19. Ann Neurol. 2012;72:870–80.
2. Lee YC, Durr A, Majczenko K, Huang YH, Liu YC, Lien CC, et al. Mutations in KCND3 cause spinocerebellar ataxia type 22. Ann Neurol. 2012;72:859–69. 3. Matilla-Duenas A. The ever expanding spinocerebellar ataxias. Editorial. Cerebellum. 2012;11:821–7. 4. Cagnoli C, Mariotti C, Taroni F, Seri M, Brussino A, Michielotto C, et al. SCA28, a novel form of autosomal dominant cerebellar ataxia on chromosome 18p11.22-q11.2. Brain. 2006;129:235–42. 5. Di Bella D, Lazzaro F, Brusco A, Plumari M, Battaglia G, Pastore A, et al. Mutations in the mitochondrial protease gene AFG3L2 cause dominant hereditary ataxia SCA28. Nat Genet. 2010;42:313–21. 6. Mariotti C, Brusco A, Di Bella D, Cagnoli C, Seri M, Gellera C, et al. Spinocerebellar ataxia type 28: a novel autosomal dominant cerebellar ataxia characterized by slow progression and ophthalmoparesis. Cerebellum. 2008;7:184–8. 7. Cagnoli C, Stevanin G, Brussino A, Barberis M, Mancini C, Margolis RL, et al. Missense mutations in the AFG3L2 proteolytic domain account for approximately 1.5 % of European autosomal dominant cerebellar ataxias. Hum Mutat. 2010;31:1117–24. 8. Edener U, Wollner J, Hehr U, Kohl Z, Schilling S, Kreuz F, et al. Early onset and slow progression of SCA28, a rare dominant ataxia in a large four-generation family with a novel AFG3L2 mutation. Eur J Hum Genet. 2010;18:965–8. 9. Pierson TM, Adams D, Bonn F, Martinelli P, Cherukuri PF, Teer JK, et al. Whole-exome sequencing identifies homozygous AFG3L2 mutations in a spastic ataxia-neuropathy syndrome linked to mitochondrial m-AAA proteases. PLoS Genet. 2011;7:e1002325. 10. Banfi S, Bassi MT, Andolfi G, Marchitiello A, Zanotta S, Ballabio A, et al. Identification and characterization of AFG3L2, a novel paraplegin-related gene. Genomics. 1999;59:51–8. 11. Nolden M, Ehses S, Koppen M, Bernacchia A, Rugarli EI, Langer T. The m-AAA protease defective in hereditary spastic paraplegia controls ribosome assembly in mitochondria. Cell. 2005;123:277–89. 12. Tatsuta T, Langer T. Quality control of mitochondria: protection against neurodegeneration and ageing. Embo J. 2008;27:306–14. 13. Koppen M, Metodiev MD, Casari G, Rugarli EI, Langer T. Variable and tissue-specific subunit composition of mitochondrial m-AAA protease complexes linked to hereditary spastic paraplegia. Mol Cell Biol. 2007;27:758–67. 14. Casari G, De Fusco M, Ciarmatori S, Zeviani M, Mora M, Fernandez P, et al. Spastic paraplegia and OXPHOS impairment caused by mutations in paraplegin, a nuclear-encoded mitochondrial metalloprotease. Cell. 1998;93:973–83. 15. Klebe S, Depienne C, Gerber S, Challe G, Anheim M, Charles P, et al. Spastic paraplegia gene 7 in patients with spasticity and/or optic neuropathy. Brain. 2012;135:2980–93. 16. van Gassen KL, van der Heijden CD, de Bot ST, den Dunnen WF, van den Berg LH, Verschuuren-Bemelmans CC, et al. Genotypephenotype correlations in spastic paraplegia type 7: a study in a large Dutch cohort. Brain. 2012;135:2994–3004. 17. Almajan ER, Richter R, Paeger L, Martinelli P, Barth E, Decker T, et al. AFG3L2 supports mitochondrial protein synthesis and Purkinje cell survival. J Clin Invest. 2012;122:4048–58. 18. Maltecca F, De Stefani D, Cassina L, Consolato F, Wasilewski M, Scorrano L, et al. Respiratory dysfunction by AFG3L2 deficiency causes decreased mitochondrial calcium uptake via organellar network fragmentation. Hum Mol Genet. 2012;21:3858–70. 19. Hwang J, Kim YK. When a ribosome encounters a premature termination codon. BMB Rep. 2013;46:9–16. 20. Kervestin S, Jacobson A. NMD: a multifaceted response to premature translational termination. Nat Rev Mol Cell Biol. 2012;13:700–12.
Cerebellum 21. Magyar I, Colman D, Arnold E, Baumgartner D, Bottani A, Fokstuen S, et al. Quantitative sequence analysis of FBN1 premature termination codons provides evidence for incomplete NMD in leukocytes. Hum Mutat. 2009;30:1355–64. 22. Maltecca F, Magnoni R, Cerri F, Cox GA, Quattrini A, Casari G. Haploinsufficiency of AFG3L2, the gene responsible for
spinocerebellar ataxia type 28, causes mitochondria-mediated Purkinje cell dark degeneration. J Neurosci. 2009;29:9244–54. 23. Turleau C. Monosomy 18p. Orphanet J Rare Dis. 2008;3:4. 24. Kim JI, Ju YS, Park H, Kim S, Lee S, Yi JH, et al. A highly annotated whole-genome sequence of a Korean individual. Nature. 2009;460: 1011–5.
ORIGINAL PAPER
A Novel Frameshift Mutation in the AFG3L2 Gene in a Patient with Spinocerebellar Ataxia Zuzana Musova & Michaela Kaiserova & Eva Kriegova & Regina Fillerova & Peter Vasovcak & Alena Santava & Katerina Mensikova & Alena Zumrova & Anna Krepelova & Zdenek Sedlacek & Petr Kanovsky
# Springer Science+Business Media New York 2013
Abstract Spinocerebellar ataxia type 28 (SCA28) is an autosomal dominant neurodegenerative disorder caused by missense AFG3L2 mutations. To examine the occurrence of SCA28 in the Czech Republic, we screened 288 unrelated ataxic patients with hereditary (N =49) and sporadic or unknown (N =239) form of ataxia for mutations in exons 15 and 16, the AFG3L2 mutation hotspots. A single significant variant, frameshift mutation c.1958dupT leading to a premature termination codon, was identified in a patient with slowly progressive speech and gait problems starting at the age of 68 years. Neurological examination showed cerebellar ataxia, mild Parkinsonian features with predominant bradykinesia, Z. Musova (*) : P. Vasovcak : A. Krepelova : Z. Sedlacek Department of Biology and Medical Genetics, Charles University 2nd Faculty of Medicine and University Hospital Motol, Prague, Czech Republic e-mail: [email protected] M. Kaiserova : K. Mensikova : P. Kanovsky Department of Neurology, Palacky University Olomouc Faculty of Medicine and Dentistry and University Hospital, Olomouc, Czech Republic E. Kriegova : R. Fillerova Department of Immunology, Palacky University Olomouc Faculty of Medicine and Dentistry and University Hospital, Olomouc, Czech Republic A. Santava Department of Medical Genetics and Foetal Medicine, Palacky University Olomouc Faculty of Medicine and Dentistry and University Hospital, Olomouc, Czech Republic A. Zumrova Department of Child Neurology, Charles University 2nd Faculty of Medicine and University Hospital Motol, Prague, Czech Republic Present Address: P. Vasovcak Gendiagnostica, Kosice, Slovak Republic
polyneuropathy of the lower limbs, and cognitive decline. However, other common SCA28 features like pyramidal tract signs (lower limb hyperreflexia, positive Babinski sign), ophthalmoparesis or ptosis were absent. The mutation was also found in a patient’s unaffected daughter in whom a targeted examination at 53 years of age revealed mild imbalance signs. RNA analysis showed a decreased ratio of the transcript from the mutated AFG3L2 allele relative to the normal transcript in the peripheral lymphocytes of both patients. The ratio was increased by puromycin treatment, indicating that the mutated transcript can be degraded via nonsense-mediated RNA decay. The causal link between the mutation and the phenotype of the patient is currently unclear but a pathogenic mechanism based on AFG3L2 haploinsufficiency rather than the usual dominantnegative effect of missense AFG3L2 mutations reported in SCA28, cannot be excluded. Keywords SCA28 . Spinocerebellar ataxia . AFG3L2 . Frameshift mutation
Introduction Spinocerebellar ataxias (SCAs) form a heterogeneous group of autosomal dominant disorders characterized by cerebellar ataxia frequently accompanied by additional neurological symptoms which differ among subtypes and also among and within affected families. Currently, at least 30 subtypes of SCA and 20 causative genes have been identified [1–3]. Most SCAs are caused by expansions of polyglutamine-coding CAG repeats (SCA types 1–3, 6–7, 17, and dentatorubropallidoluysian atrophy (DRPLA)). Rarer SCAs result from expansions of noncoding nucleotide repeats (SCA types 8, 10, 12, and 36), insertion of a pentanucleotide repeat (SCA 31) or conventional mutations (SCA types 5, 11, 13, 14, 15/ 16/29, 19/22, 20, 27, 28, and 35) [1–3].
Cerebellum
Missense mutations of one allele of the AFG3L2 gene cause SCA type 28 (SCA28, MIM#610246) [4–6] which represents about 1.5 % of SCAs in the European population [7, 8]. Clinical symptoms of SCA28 appear mostly in early adulthood but the age of onset varies between 3 and 60 years [4–8]. SCA28 is characterized by slowly progressive gait and limb ataxia, dysarthria, ptosis, ophthalmoparesis, gaze-evoked nystagmus, hyperreflexia of the lower limbs, and decreased vibration sense. Extrapyramidal signs (Parkinsonism or dystonia) and cognitive impairment are also occasionally present [4–8]. Homozygous missense AFG3L2 mutations lead to spastic ataxia-neuropathy syndrome [9]. The AFG3L2 gene located in 18p11 contains 17 exons [10] and encodes a subunit of m-AAA mitochondrial proteases. These proteases are involved in protein quality control, regulation, and chaperonelike activities in the mitochondrial proteome [11, 12]. The AFG3L2 subunits assemble into different m-AAA complexes: homo-oligomeric complexes consist solely of AFG3L2, and hetero-oligomeric complexes consist of AFG3L2 and paraplegin, which is encoded by the SPG7 gene [13]. SPG7 mutations cause spastic paraplegia type 7 (SPG7) which is mostly autosomal recessive, although some mutations show a possible dominant effect. SPG7 is frequently associated with cerebellar ataxia [14–16]. AFG3L2 and SPG7 are ubiquitously expressed [10] with high-selective expression in cerebellar Purkinje cells (PCs) and large neurons of the deep cerebellar nuclei; in addition, paraplegin is also enriched in spinal motor neurons [5]. AFG3L2 supports mitochondrial protein synthesis and PC survival [17], and AFG3L2 deficiency reduces mitochondrial calcium uptake capacity by fragmentation of the mitochondrial network [18]. We report here on a patient with a late onset of cerebellar ataxia with Parkinsonian features at the age of 68 years who carried a novel frameshift mutation in the AFG3L2 gene. Although the causality of the mutation for the phenotype is unclear, the observation of a reduced dose of the mutated transcript allows us to speculate that a mechanism based on AFG3L2 haploinsufficiency rather than the usual dominantnegative effect of missense AFG3L2 mutations could potentially play a role in the pathogenesis of the patient’s phenotype.
Material and Methods Patient Cohort To establish the incidence of SCA28 in the Czech Republic, we screened 288 unrelated patients with ataxia for mutations in the AFG3L2 gene. The patients were previously negatively tested for SCA types 1–3, 6–8, 12, 17, DRPLA, and the fragile X-associated tremor/ataxia syndrome. The cohort included 49 patients with a positive family history (out of these, 21 patients
were from families with a clearly autosomal dominant pattern of inheritance), 96 sporadic patients, and 143 patients with unspecified family history. A control group of 104 unrelated healthy Czech individuals was also assessed. Clinical Description of Patients 1 and 2 Patient 1 noticed slowly progressive speech and gait problems at the age of 68 years; until then, she suffered only from glaucoma and coxarthrosis. The first neurological examination at the age of 70 showed dysarthria, bilateral limb ataxia (which was more pronounced on the lower limbs), gaze-evoked bidirectional nystagmus, and uncertain wide-base stand and gait. Neither extrapyramidal and pyramidal signs nor ophthalmoparesis were present. Brain magnetic resonance imaging (MRI) showed isolated cerebellar atrophy. At the age of 71, she developed mild Parkinsonian features with predominant bradykinesia, and gradually also, a mild cognitive decline. She showed no prominent postural instability with falls. The last detailed examination of patient 1 was performed at the age of 76. She showed progression in cerebellar symptoms, slow saccadic eye movements (but neither nystagmus nor ophthalmoparesis were observed), decreased reflexes, and vibration sense in the lower limbs and mild bradykinesia. The Parkinsonian symptoms did not respond to levodopa treatment. Electromyography revealed mild distal axonal sensorimotor polyneuropathy of the lower limbs. Extensive laboratory testing to reveal the cause of polyneuropathy yielded no explanation: there was no evidence of diabetes mellitus, hypothyroidism, vitamin B12 deficiency, or renal insufficiency; serum activities of liver enzymes and carbohydrate-deficient transferrin were normal and anti-ganglioside antibodies and borrelia antibodies in the cerebrospinal fluid were negative. Brain MRI showed cerebellar and diffuse supratentorial atrophy and small white matter hyperintensities (Fig. 1). The Mini Mental State Examination score was 21 points out of 30. A detailed neuropsychological examination revealed global deterioration of cognitive functions. The degree of dementia was mild. Depression was also present. Both the brain MRI findings and the results of neuropsychological testing were not typical for any common type of dementia (e.g., Alzheimer’s disease or vascular dementia). The patient did not suffer from urinary incontinence or urgency, and there was no significant orthostatic blood pressure decrease after 3 min of standing which would fulfill the criteria of orthostatic hypotension. Patient 1 had four siblings, and two of them (brother and sister) suffered from speech and gait problems. Unfortunately, they refused examination and no detailed clinical data were available. There was no evidence of similar problems in the parents of patient 1, but her mother died of breast carcinoma at the age of 45. The father died of heart failure at the age of 90. Patient 1 had three children. Two of them were not examined neurologically, but they did not complain of any speech or gait
Cerebellum
Fig. 1 Brain MRI images of Patient 1. Sagittal T1-weighted image (left) shows cerebellar atrophy. Transversal FLAIR image (middle) shows a mild diffuse brain atrophy and small white matter hyperintensities. Axial
T2 image through the pons (right) shows no signs typical for patients with multiple system atrophy or progressive supranuclear palsy
difficulties. The remaining daughter of patient 1 (patient 2) was examined at her own request, although she did not complain of any neurological symptoms. The examination performed at the age of 53 revealed a mild postural instability (positivity in Romberg standing test) and unstable tandem gait. The cranial nerves were intact. She had neither nystagmus nor any oculomotor abnormality. There were no extrapyramidal or pyramidal signs, and no clinical signs of polyneuropathy. Brain MRI did not show signs of cerebellar atrophy.
Mutation description was based on the reference sequences NM_006796.2 and NM_003119.2 with nucleotide 1 corresponding to A in the initiation codon. The nucleotide changes identified were compared to the NCBI dbSNP database (http://www.ncbi.nlm.nih.gov/snp) and the Exome Variant Server (http://evs.gs.washington.edu/EVS/).
DNA Analysis The AFG3L2 gene testing using high-resolution melting analysis (HRM) was focused on exons 15 and 16, which have been shown to be mutation hotspots [5, 7, 8]. Genomic DNA was isolated from blood using the Gentra Puregene Blood Core Kit (QIAGEN, Gaithersburg, MD, USA). PCR of the AFG3L2 exons 15 and 16 was performed using a published method [7] with adding LCGreen Plus Dye (Idaho Technology, Salt Lake City, UT, USA). HRM analysis of the PCR products was performed using LightScanner (Idaho Technology) according to the instrument manual. PCR products with aberrant melting profiles were purified using SureClean PCR Purification Kit (Bioline, London, UK) and directly sequenced using BigDye Terminator v3.1 Cycle Sequencing kit and an ABI 3130 Genetic Analyser (both Applied Biosystems, Foster City, CA, USA). The remaining 15 exons of the AFG3L2 gene as well as all 17 exons of the SPG7 gene were amplified according to a published protocol [7] and directly sequenced in patients 1 and 2 (in whom an AFG3L2 mutation was identified using HRM) to exclude the possible effects of additional AFG3L2 or SPG7 variants on their phenotype. Possible deletions or duplications in the SPG7 gene were analyzed in patients 1 and 2 using multiplex ligation-dependent probe amplification (MLPA) with SALSA MLPA Kit P213-B1 (MRC Holland, Amsterdam, The Netherlands) according to the manufacturer’s protocol.
RNA Analysis Peripheral blood samples of patients 1 and 2 and a control individual were collected into PAXgene Blood RNA tubes (PreAnalytiX, Hombrechtikon, Switzerland). Total lymphocyte RNA was isolated using PAX Gene Blood RNA Kit (QIAGEN) and reverse transcribed into cDNA using SuperScript One-Step RT-PCR with Platinum Taq System (Invitrogen, Carlsbad, CA, USA) and primers specific for AFG3L2 exons 13/14 and 17 (cAFG3L2-LF (TGGCTTAGAGAAGAAAACGCAG) and cAFG3L2-LR (CTAGTTGGCAACTTTCTCACCC)). PCR products were separated on a 1.8 % agarose gel and sequenced as above with nested primers cAFG3L2-14/15 F (CGCTTTTAAAGGTATCCATCATC) and cAFG3L2-B (CTTGCAGTGGCTTCACTGTAAG). To test the influence of different RNA isolation procedures on the AFG3L2 transcripts, RNA was also isolated from peripheral blood mononuclear cells separated with Ficoll from heparin-stabilized blood of patient 1 using mirVana miRNA kit (Ambion, Austin, TX, USA). An aliquot of the blood sample was pre-treated with puromycin (Sigma-Aldrich, St. Louis, MO, USA) at a concentration of 0.2 mg/ml for 3 h at 37 °C. To quantify the amount of normal and mutated transcripts, RT-PCR products obtained using primers cAFG3L2-B and 6-FAM labeled cAFG3L2-14/15F were analyzed on ABI 3130 Genetic Analyzer. The ratio of peak areas of normal and mutated RT-PCR products was adjusted to the peak area ratio of a PCR product obtained on DNA using primers AFG3L2-15F (6-FAM CCACTAAGGCTGATGAACT) [7] and AFG3L2-15RB (GCACAACTATATTGTTTCACAGCC).
Cerebellum
Results
genomic DNA genomic DNA
A A AG T A AC T C A G A G T G C A A A AG T T A AC T C A G A G T G C
mut:wt 1:1
cDNA (PAX)
A A AG T A AC T C A G A G T G C A A A AG T T A AC T C A G A G T G C
0.33:1
A A AG T A AC T C A G A G T G C A A A AG T T A AC T C A G A G T G C
0.33:1
A A AG T A AC T C A G A G T G C A A A AG T T A AC T C A G A G T G C
0.33:1
cDNA (H0)
A A AG T A AC T C A G A G T G C A A A AG T A AC T C A G A G T G C A
cDNA (H3P)
cDNA (H3)
Patient 1
normal control
HRM analysis of AFG3L2 exons 15 and 16 in 288 ataxic patients and 104 controls revealed the frameshift mutation c.1958dupT in exon 15 (Fig. 2) in one patient (patient 1), and subsequent analysis identified the same mutation also in her daughter (patient 2). The mutation resulted in a premature termination codon (p.Thr654Asnfs*15). The HRM analysis also revealed the c.2175 + 18G>A variant in intron 16 (rs117096851) in seven patients and three controls. The
analysis of the remaining exons of AFG3L2 and all exons of SPG7 in patients 1 and 2 identified several additional known variants which were unlikely to affect the phenotype. The AFG3L2 gene harbored heterozygous variants c.553-95G>A (rs2298542), c.753-55T>C (rs7407640), c.752 + 6C>T (rs8097342), c.1389G>A (p.Leu463=, rs11080572), c.1650A>G (p.Glu550=, rs11553521), and c.*28G>C (rs1129115) in patient 1, and homozygous variants c.1389G>A and c.1650A>G in patient 2. The SPG7 gene harbored heterozygous variants c.286+46C>T (rs11553521), c.618 + 12T>C (rs3803679), c.619-47G>A (rs3935626), c.862-34G>T (rs4785690), c.987 + 5A>G (rs4785691), c.1779+47G>C (rs3794632), c.2063G>A (p.Arg688Gln, rs12960), and c.2292C>T (rs61747711) in patient 1, and a homozygous variant c.862-34G>T in patient 2. No SPG7 deletions or duplications were found in patients 1 and 2 using the MLPA analysis (data not shown). The RT-PCR of AFG3L2 exon 15 in patients 1 and 2 and the unaffected control showed a major band of the expected length in all individuals. Sequencing and fragment analysis showed a decreased ratio of the RT-PCR product from the allele carrying the c.1958dupT mutation relative to the normal allele (0.33:1, Fig. 2) in both patients. The signal from the mutated allele was stronger in the cDNA from a blood sample treated with puromycin (0.87:1, Fig. 2). Real-time RT-PCR experiments failed to show a significantly reduced amount of total AFG3L2 transcripts in peripheral lymphocytes of patients 1 and 2, although their AFG3L2 mRNA levels fell in the lower range of AFG3L2 levels observed in seven controls (data not shown). The analysis of the AFG3L2 protein in peripheral blood mononuclear cells using Western blot was unsuccessful.
Discussion
A A AG T A AC T C A G A G T G C A A A AG T T A AC T C A G A G T G C
0.87:1
Fig. 2 Sequencing electropherograms and fragment analysis curves of exon 15 of the AGF3L2 gene in a normal control and patient 1 showing the heterozygous c.1958dupT mutation in genomic DNA and different preparations of lymphocyte cDNA (PAX PAXgene-stabilized blood; H heparin-stabilized blood; 0 RNA isolation without incubation; 3 RNA isolation after 3-h incubation at 37 °C; P incubation with puromycin). Arrows indicate the duplicated base in sequencing electropherograms and the mutated allele in fragment analysis curves. The ratio of the mutated transcript relative to the normal transcript calculated from peak areas adjusted to the 1:1 ratio in DNA is also shown. The decreased amount of the mutated transcript is clearly evident. The ratio was increased after puromycin treatment
In this study, we screened a heterogeneous group of 288 unrelated patients with ataxia for mutations in exons 15 and 16 of the AFG3L2 gene. A single significant variant, frameshift mutation c.1958dupT (p.Thr654Asnfs*15), was identified in patient 1, a currently 76-year-old female from a family suggestive of autosomal dominant pattern of inheritance. This novel mutation was found neither in 104 unrelated Czech controls, nor in any of the previous studies focused on AFG3L2 mutations in SCA patients and controls [5–8]. The variant was also absent in dbSNP and the Exome Variant Server. The same variant was identified in the daughter of patient 1 who had mild imbalance signs upon examination at the age of 53 years but currently cannot be considered affected (patient 2). The level of the transcript from the mutated allele was reduced in blood lymphocytes of both patients relative to the transcript from the normal allele.
Cerebellum
The c.1958dupT mutation is located in the peptidase M14 domain of the AFG3L2 protein. It has been shown that this domain is critical for the pathogenesis of SCA28. With the exception of the p.N432T mutation in the AAA domain [5], all previously described pathogenic AFG3L2 variants were clustered in the M14 domain [5–9]. However, in contrast to the c.1958dupT mutation in patient 1, all mutations described so far were missense and were proposed to exert a dominantnegative effect [5–9]. The absence of the c.1958dupT mutation in control cohorts and its influence on the relative ratio of transcripts from the mutated and normal AFG3L2 alleles do not represent a clearcut proof that the mutation is pathogenic and causal for the phenotype observed in our patient. We cannot exclude that the phenotype can be completely unrelated to this AFG3L2 mutation or perhaps just modulated by this variant. However, our finding is remarkable as two different, not mutually exclusive mechanisms cannot be excluded in association with this mutation: a reduced dose of the AFG3L2 transcript and protein, and production of a truncated non-functional protein. Our RNA analysis showed a 0.33:1 ratio of the mutated transcript relative to the transcript from the normal allele. This decrease could be partially reverted (to 0.87:1) by treatment of blood with translational inhibitor puromycin prior to RNA preparation, indicating that the mutated transcript was likely degraded via the nonsense-mediated RNA decay (NMD) pathway [19, 20]. Varying NMD efficiency was observed in different cell types and tissues [19], and we can only speculate about the actual level of aberrant transcripts in the site of the highest AFG3L2 expression, the cerebellar PCs [5]. It could be even lower than that observed in blood as NMD in leukocytes has been shown to be less complete compared to other tissues [21]. A reduced dose of AFG3L2 has never been described in SCA28 patients and its phenotypic effect is unclear, but haploinsufficient mouse models exhibit progressive late onset symptoms similar to those in SCA28 patients [22]. Using real-time PCR experiments, we could not prove a significantly reduced total level of AFG3L2 transcripts in lymphocytes of patients 1 and 2, although their transcript levels fell in the lower range of levels observed in the controls. Considering the 33 % output from the mutated allele and 100 % output from normal alleles (i.e., with no compensatory upregulation of the normal allele in the patients), the decrease in the patients should theoretically reach 67 % of the normal level, which can be difficult to prove experimentally. Simultaneously, the dominant-negative effect could also play a role if some residual amount of the truncated AFG3L2 protein lacking a significant part of the M14 domain is present in the cerebellum. The sole replacement of the first amino acid affected by the c.1958dupT mutation has been shown to be critical for the function of AFG3L2 in a patient with missense mutation p.Thr654Ile [7]. The truncated protein could associate with other components of the m-AAA complexes
and render them non-functional. Unfortunately, our efforts to analyze the AFG3L2 protein using Western blot failed. The phenotype of patient 1 was different from the typical SCA28 picture. Especially the late onset of symptoms in patient 1 (68 years) has not been described yet. Cerebellar ataxia, mild Parkinsonism unresponsive to levodopa, and polyneuropathy of the lower limbs in patient 1 can be a part of the clinical spectrum of SCA28, and extensive laboratory testing did not reveal any other etiology. Dementia has been also seen in several isolated cases of SCA28. In contrast, common features of SCA28 like pyramidal tract signs (hyperreflexia of the lower limbs, positive Babinski sign), ophthalmoparesis, or ptosis were absent in patient 1. Her daughter carrying the same mutation, patient 2, did not complain of ataxia at the age of 53, but her neurological examination revealed mild postural instability and unstable gait. Multiple system atrophy (MSA-C) should also be considered in the differential diagnosis of cerebellar ataxia with mild Parkinsonism. However, in patient 1, no autonomic failure (i.e., urinary problems or orthostatic hypotension) was present, and therefore, the clinical criteria of MSA-C were not met. The absence of both vertical supranuclear gaze palsy and prominent postural instability with falls in the first year of onset argued also against the diagnosis of another atypical Parkinsonian syndrome, the progressive supranuclear palsy (PSP). Moreover, brain MRI in patient 1 showed “only” cerebellar and diffuse supratentorial atrophy and small white matter hyperintensities. There were no typical signs of MSA (“hot cross bun” sign in the pons) or PSP (midbrain atrophy with the “standing penguin silhouette” sign). The AFG3L2 protein is present together with the SPG7 gene product paraplegin in m-AAA protease complexes in the mitochondrial membrane [13]. A clear connection between different genetic defects in the AFG3L2 and SPG7 genes is evident also on the clinical level. Mutations of both alleles of SPG7 usually present with spastic paraparesis, optic neuropathy (or abnormalities in optical coherence tomography), and cerebellar ataxia. Cerebellar ataxia occurs mostly in patients with biallelic null SPG7 mutations [16]. Recently, several patients with compound heterozygous SPG7 mutations (p.Arg485_Glu487del, p.Ala510Val) have been described, who showed cerebellar signs and cerebellar atrophy on MRI, peripheral neuropathy, and no spasticity of the lower limbs [15]. Homozygous AFG3L2 mutations cause the spasticataxia-neuropathy syndrome [9]. To exclude possible biallelic or digenic effects in patients 1 and 2, we sequenced their complete AFG3L2 and SPG7 genes, but we did not reveal any additional significant variants which could have an impact on their phenotype. A reduced dose of the AFG3L2 gene could also be expected in patients with large 18p deletions involving the AFG3L2 locus. These patients showed different phenotypes and various malformations; however, cerebellar ataxia was not
Cerebellum
described among the symptoms [23]. Up to now, just a single exonic AFG3L2 deletion (involving exons 10–11), which may affect the expression of AFG3L2, has been reported in a healthy Korean individual [24]. A single occurrence of two rare variants in the AFG3L2 gene, which could possibly have a similar effect like the frameshift mutation in our patients, was also reported in healthy controls in dbSNP and the Exome Variant Server. The c.1651C>T variant (rs149121681) leads to a premature stop codon (p.Arg551*), and the c.2028_2029insA variant causes frameshift and premature termination (p.Leu677Thrfs*15). Considering the late onset of the disease in the family identified in this study, it is likely that the patients with large 18p deletions as well as the presumably control individuals with the rare nucleotide variants were too young to present the symptoms. Alternatively, there is also the possibility that these variants may be nonpathogenic. Concerning the incidence of SCA28, it was reported to range between 0.7–1.5 % among autosomal dominant cerebellar ataxia patients in Europe [7, 8]. We cannot directly compare our results (1/288 patients) with these data, because our sample contained mostly patients with unspecified family history. In conclusion, we report for the first time a frameshift mutation in the AFG3L2 gene in a patient with late onset of cerebellar ataxia, mild Parkinsonism, and polyneuropathy. If the mutation is indeed pathogenic and causal for the disease, it may point to another possible mechanism contributing to the pathogenesis of SCA28, which is different from the dominant-negative effect of missense mutations described previously, and in which a decreased level of the AFG3L2 transcript could lead to the atypical milder presentation of SCA28. Acknowledgments We thank the patients for their participation in the study, two anonymous reviewers for very valuable comments and Michal Krupka, Ph.D., for his help with Western blot analysis. This work was supported by grant NT-12221 and grant for conceptual development of research organization University Hospital Motol 00064203 from the Ministry of Health of the Czech Republic, and by grant IGA-LF_13_13 from the Internal grant agency of Palacky University. Conflict of interest There are no potential conflicts of interest that might bias this work.
References 1. Duarri A, Jezierska J, Fokkens M, Meijer M, Schelhaas HJ, den Dunnen WF, et al. Mutations in potassium channel kcnd3 cause spinocerebellar ataxia type 19. Ann Neurol. 2012;72:870–80.
2. Lee YC, Durr A, Majczenko K, Huang YH, Liu YC, Lien CC, et al. Mutations in KCND3 cause spinocerebellar ataxia type 22. Ann Neurol. 2012;72:859–69. 3. Matilla-Duenas A. The ever expanding spinocerebellar ataxias. Editorial. Cerebellum. 2012;11:821–7. 4. Cagnoli C, Mariotti C, Taroni F, Seri M, Brussino A, Michielotto C, et al. SCA28, a novel form of autosomal dominant cerebellar ataxia on chromosome 18p11.22-q11.2. Brain. 2006;129:235–42. 5. Di Bella D, Lazzaro F, Brusco A, Plumari M, Battaglia G, Pastore A, et al. Mutations in the mitochondrial protease gene AFG3L2 cause dominant hereditary ataxia SCA28. Nat Genet. 2010;42:313–21. 6. Mariotti C, Brusco A, Di Bella D, Cagnoli C, Seri M, Gellera C, et al. Spinocerebellar ataxia type 28: a novel autosomal dominant cerebellar ataxia characterized by slow progression and ophthalmoparesis. Cerebellum. 2008;7:184–8. 7. Cagnoli C, Stevanin G, Brussino A, Barberis M, Mancini C, Margolis RL, et al. Missense mutations in the AFG3L2 proteolytic domain account for approximately 1.5 % of European autosomal dominant cerebellar ataxias. Hum Mutat. 2010;31:1117–24. 8. Edener U, Wollner J, Hehr U, Kohl Z, Schilling S, Kreuz F, et al. Early onset and slow progression of SCA28, a rare dominant ataxia in a large four-generation family with a novel AFG3L2 mutation. Eur J Hum Genet. 2010;18:965–8. 9. Pierson TM, Adams D, Bonn F, Martinelli P, Cherukuri PF, Teer JK, et al. Whole-exome sequencing identifies homozygous AFG3L2 mutations in a spastic ataxia-neuropathy syndrome linked to mitochondrial m-AAA proteases. PLoS Genet. 2011;7:e1002325. 10. Banfi S, Bassi MT, Andolfi G, Marchitiello A, Zanotta S, Ballabio A, et al. Identification and characterization of AFG3L2, a novel paraplegin-related gene. Genomics. 1999;59:51–8. 11. Nolden M, Ehses S, Koppen M, Bernacchia A, Rugarli EI, Langer T. The m-AAA protease defective in hereditary spastic paraplegia controls ribosome assembly in mitochondria. Cell. 2005;123:277–89. 12. Tatsuta T, Langer T. Quality control of mitochondria: protection against neurodegeneration and ageing. Embo J. 2008;27:306–14. 13. Koppen M, Metodiev MD, Casari G, Rugarli EI, Langer T. Variable and tissue-specific subunit composition of mitochondrial m-AAA protease complexes linked to hereditary spastic paraplegia. Mol Cell Biol. 2007;27:758–67. 14. Casari G, De Fusco M, Ciarmatori S, Zeviani M, Mora M, Fernandez P, et al. Spastic paraplegia and OXPHOS impairment caused by mutations in paraplegin, a nuclear-encoded mitochondrial metalloprotease. Cell. 1998;93:973–83. 15. Klebe S, Depienne C, Gerber S, Challe G, Anheim M, Charles P, et al. Spastic paraplegia gene 7 in patients with spasticity and/or optic neuropathy. Brain. 2012;135:2980–93. 16. van Gassen KL, van der Heijden CD, de Bot ST, den Dunnen WF, van den Berg LH, Verschuuren-Bemelmans CC, et al. Genotypephenotype correlations in spastic paraplegia type 7: a study in a large Dutch cohort. Brain. 2012;135:2994–3004. 17. Almajan ER, Richter R, Paeger L, Martinelli P, Barth E, Decker T, et al. AFG3L2 supports mitochondrial protein synthesis and Purkinje cell survival. J Clin Invest. 2012;122:4048–58. 18. Maltecca F, De Stefani D, Cassina L, Consolato F, Wasilewski M, Scorrano L, et al. Respiratory dysfunction by AFG3L2 deficiency causes decreased mitochondrial calcium uptake via organellar network fragmentation. Hum Mol Genet. 2012;21:3858–70. 19. Hwang J, Kim YK. When a ribosome encounters a premature termination codon. BMB Rep. 2013;46:9–16. 20. Kervestin S, Jacobson A. NMD: a multifaceted response to premature translational termination. Nat Rev Mol Cell Biol. 2012;13:700–12.
Cerebellum 21. Magyar I, Colman D, Arnold E, Baumgartner D, Bottani A, Fokstuen S, et al. Quantitative sequence analysis of FBN1 premature termination codons provides evidence for incomplete NMD in leukocytes. Hum Mutat. 2009;30:1355–64. 22. Maltecca F, Magnoni R, Cerri F, Cox GA, Quattrini A, Casari G. Haploinsufficiency of AFG3L2, the gene responsible for
spinocerebellar ataxia type 28, causes mitochondria-mediated Purkinje cell dark degeneration. J Neurosci. 2009;29:9244–54. 23. Turleau C. Monosomy 18p. Orphanet J Rare Dis. 2008;3:4. 24. Kim JI, Ju YS, Park H, Kim S, Lee S, Yi JH, et al. A highly annotated whole-genome sequence of a Korean individual. Nature. 2009;460: 1011–5.
