Autosomal recessive cerebellar ataxia caused by a homozygous mutation in<i>PMPCA</i>

Sir, Recently, Jobling et al. (2015) reported the identification of mutations in PMPCA in 17 patients from four families affected with autosomal recessive non-progressive cerebellar ataxia. A homozygous missense mutation in PMPCA (c.1129G>A, p.Ala377Thr) was uncovered in three families of Christian Lebanese Maronite origin, while compound heterozygous mutations (c.287C>T, p.Ser96Leu; c.1543G>A, p.Gly515Arg) were identified in a French patient ( Jobling et al. , 2015 ). PMPCA encodes the alpha subunit of the mitochondrial processing peptidase (MPP), which cleaves the targeting peptide of nuclear-encoded mitochondrial precursor proteins upon their import into mitochondria ( Teixeira and Glaser, 2013 ). Jobling et al. (2015) showed compelling evidence for the causality of PMPCA variants. First, they demonstrated co-segregation of the variants with the disease among the four families, including a large consanguineous family of 32 individuals. Moreover, they observed decreased levels of MPPα in lymphoblasts from two affected members compared to heterozygote carriers and healthy controls. Interestingly, they also showed impaired processing of frataxin (FXN), a mitochondrial protein that is a known substrate of MPP ( Cavadini et al. , 2000 ; Schmucker et al. , 2008 ). The paper by Jobling et al. (2015) is the first to associate defects in PMPCA with a human disease, which suggests a new mechanism for the pathogenesis of non-progressive cerebellar ataxias. We wish to complement this study with our own identification of two brothers affected by a juvenile-onset recessive cerebellar ataxia caused by a homozygous mutation in PMPCA . These two patients were born of French Canadian parents who are distantly related ( Fig. 1 A). Considerable clinical variability was present between the two affected cases. Patient II.2 started developing impaired gait, dysarthria, dysmetria and mild distal atrophy during adolescence. He was diagnosed with a slowly progressive spinocerebellar ataxia. He did not have intellectual deficiency but had learning difficulties in school. Patient II.3 had an earlier onset at ∼5 years of age and a more severe form of the disease than his older brother. He presented for consultation at the age of 15 with ataxia, hemidystonia, dysarthria and neurosensorial hearing loss. The MRI exams of both siblings ( Fig. 1 D–G) documented the presence of global cerebellar atrophy, involving both the vermis and the cerebellar hemispheres. In addition, MRI of one of the two brothers (Patient II.2) demonstrated a bilateral and symmetric T 2 hyperintense signal at the level of the deep cerebellar white matter. No structural or signal abnormalities were noted in the supratentorial structures. Genetic and MRI analysis. ( A ) Pedigree of the family. The red arrow indicates the individual sent for whole-exome sequencing. ( B ) Genomic sequence chromatograms of one affected, one heterozygote carrier and one unrelated control. The arrow indicates the substitution of the G for an A at position 266 of the coding sequence of PMPCA . ( C ) Amino acid conservation for the PMPCA p.Val256Met variant. MRI findings in subjects with PMPCA mutations. Sagittal T 1 - ( D ) and T 2 - ( E ) weighted images of Patients II.3 and II.2, respectively showing the presence of cerebellar atrophy and cerebellar white matter abnormalities ( E ). Axial T 2 -weighted images showing no supratentorial abnormalities ( F ) and hyperintense signal at the level of the peri-dentate cerebellar white matter ( G ) in Patient II.2. To uncover the genetic cause of disease in this family, we performed whole-exome sequencing on Patient II.3 (SureSelect All Exon v.5 and Illumina HiSeq 2000 with 100-bp paired-end sequencing reads). We filtered for non-synonymous, splicing and coding indel variants that were present at a minor allele frequency <5% in 1000 Genomes, EVS and ExAC or present in less than 30 in-house exomes. Using this filtering strategy, we identified nine rare homozygous variants. According to its gene function, the homozygous missense variant c.766G>A (p.Val256Met) in PMPCA (NM_015160) stood out as a good candidate. This variant is predicted to be pathogenic and is present at a conserved position ( Fig. 2 C). In addition, it is located in a large homozygous region of 6.5 Mb and Sanger sequencing showed that it co-segregates with disease status in the family ( Fig. 2 B). Functional analysis of mitochondrial processing peptidase in patient lymphoblasts . Sodium dodecyl sulphate polyacrylamide gel electrophoresis analysis of whole cell protein extracts in lymphoblast from controls (Control-1 and Control-2), carriers (Patients I-1, I-2 and II-1) and patients (Patients II-2 and II-3). Immunoblots show the protein steady state levels of the MPP alpha (MPPα) and beta (MPPβ) subunits, frataxin isoform-1 (FXN42-210) and frataxin isoform-2 (FXN81-210), malate dehydrogenase (MDH2) and actin (loading control). To further link this variant to the ataxia in this family, we assessed the functional impact of the p.Val256Met mutation on MPP function. We performed immunoblots on protein extracts from immortalized lymphoblasts of the two patients, their parents, a sibling and two unrelated healthy controls. Steady-state levels of MPPα were comparable amongst patients, carriers and controls. Using a commercial antibody (14147-1-AP, Proteintech) to detect the steady state levels of both human frataxin isoforms (FXN81-210, FXN42-210), we showed that all family members accumulate FXN42-210 compared to controls. Moreover, both patients presented a slight increase in FXN42-210 compared to heterozygous carriers ( Fig. 2 ). These results were further validated using a second antibody (Dr Isaya PAC2518, Supplementary Fig. 1A ). In addition, we did not observe a specific pattern of accumulation in the steady state levels of FXN1-210 or FXN81-210 ( Fig. 2 and Supplementary Fig. 1A ). We also evaluated the processing profile of the malate dehydrogenase (MDH2), an established substrate of the MPP ( Ogishima et al. , 1995 ), which remained unaffected in all samples. To further investigate the effect of MPPα depletion, we used a small interfering RNA construct to knock-down MPPα for 2, 3 and 4 days in a control fibroblast cell line. Western blot analysis of these cell lines showed that the complete loss of MPPα at 4 days resulted in a significant accumulation of FXN42-210 and depletion of FXN81-210 ( Supplementary Fig. 1B ). As mentioned above, recessive mutations in PMPCA have recently been associated with non-progressive cerebellar ataxia. Considering the clinical and MRI similarities between our patients and the ones reported in Jobling et al. (2015) , this homozygous variant in PMPCA appears to be the cause of cerebellar ataxia in this family. Although inter- and intra-familial variability is observed, there is a clear clinical overlap between the previously reported patients and ours. Gait ataxia, dysarthria and dysmetria are observed in all PMPCA patients. However, intellectual disability was not observed in our patients, although it is a frequent phenotype in the previously reported patients. In contrast to the majority of patients described by Jobling et al. , our patients seem to have a milder course of disease, resembling Patient F4-II.1 in their study. It is important to point out that the cerebellar ataxia phenotype is slowly progressing in our patients in comparison to the report from Jobling et al. , extending the phenotype associated with PMPCA mutations to slowly progressive cerebellar ataxia. The variability in phenotype could potentially be explained by a different functional impact of the mutations. The previously reported p.Ala377Thr mutation is in close proximity to the glycine-rich loop, which is the most conserved part of MPPα and is crucial for initial interaction with and binding to the substrate ( Nagao et al. , 2000 ; Dvorakova-Hola et al. , 2010 ; Kucera et al. , 2013 ). Jobling et al. (2015) showed that the mutation resulted in a decreased amount of MPPα in patients and that it altered the processing of frataxin. The other two mutations (p.Ser96Leu and p.Gly515Arg) are located in two other conserved regions of the peptidase. The variant p.Val256Met identified in the two affected siblings reported here also localizes to a conserved region away from the glycine-rich loop. Immunoblot analysis in our patients did not reveal decreased levels of MPPα; however, we observed a consistent increase in FXN42-210 with no obvious changes of FXN81-210. Together, these results imply that the initial cleavage event in the processing of frataxin proceeds normally until the generation of FXN42-210, which cannot be efficiently recognized for the second cleavage step, resulting in its accumulation. As suggested for the two missense variants found in Patient F4-II.1, the p.Val256Met may affect protein folding or stability, given the longer side chain of methionine compared to valine. The different functional impact could potentially explain the milder phenotype observed in Patient F4-II.1 and ours. In this and in other studies where the MPP activity is impaired ( Greene et al. , 2012 ; Jobling et al. , 2015 ), the analysis of the processing profile of several mitochondrial proteins, other than frataxin, remained normal. Based on the evidence presented here and in Jobling et al. (2015) , the kinetics of processing are likely to be substrate-specific. Thus in the context of a mutant form of the protein, some substrates may be more easily processed than others. The challenge that remains is to identify the substrates that are poorly processed in affected cell types that are relevant to the clinical phenotype. In fact, as suggested by Horvath and Chinnery (2015) , the biochemical defect caused by the PMPCA mutation p.Val256Met may be mild enough that it only affects selected cell types or tissues. For instance, decreased levels of MPPα and processing of frataxin and/or other mitochondrial proteins may be more severely altered in the cerebellum. This mild impact may also be consequential only during earlier brain development, when the demand for mitochondrial processing is higher ( Horvath and Chinnery, 2015 ). Both alternatives would explain the absence of a strong biochemical phenotype in adult patient lymphoblasts. In conclusion, we report an additional family, from an independent study, with a mutation in PMPCA, confirming the implication of this gene in cerebellar ataxias. Our report highlights key clinical and radiological features associated with PMPCA -related ataxias and also the presence of clinical heterogeneity by extending the phenotype to slowly progressive cerebellar ataxias. We believe these findings will help clinicians to make an accurate diagnosis and that PMPCA should be considered for both non-progressive and slowly progressive cerebellar ataxias. We thank the patients and their relatives who accepted to partake in this study. We would also like to thank Dr Grazia Isaya for kindly providing us with the frataxin antibody PAC2518. This work was supported by the Fondation Groupe Monaco and the Care4Rare Canada Consortium, funded in part by Genome Canada, the Canadian Institutes of Health Research (CIHR), the Ontario Genomics Institute, Ontario Research Fund, Genome Quebec and the Children’s Hospital of Eastern Ontario Foundation. KC and RLP received a Doctoral Award from the Fonds de recherche du Québec - Santé (FRQS). MT received a post-doctoral award from the Réseau de Médecine Génétique Appliquée, FRQS and CIHR. Supplementary material is available at Brain online.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.