Background Leukodystrophies are a heterogeneous group of inherited neurodegenerative disorders characterised by abnormal central nervous system white matter. Mutations in POLR3A and POLR3B genes were recently reported to cause four clinically overlapping hypomyelinating leukodystrophy phenotypes. Our aim was to investigate the presence and frequency of POLR3A and POLR3B mutations in patients with genetically unexplained hypomyelinating leukodystrophies with typical clinical and/or radiologic features of Pol III-related leukodystrophies.
Methods The entire coding region and the flanking exon/intron boundaries of POLR3A and/or POLR3B genes were amplified and sequenced in 14 patients.
Results Recessive mutations in POLR3A or POLR3B were uncovered in all 14 patients. Eight novel mutations were identified in POLR3A: six missenses, one nonsense, and one frameshift mutation. Seven patients carried compound heterozygous mutations in POLR3B, of whom six shared the common mutation in exon 15 (p.V523E). Seven novel mutations were identified in POLR3B: four missenses, two splice sites, and one intronic mutation.
Conclusions To date, our group has described 37 patients, of whom 27 have mutations in POLR3A and 10 in POLR3B, respectively. Altogether, our results further support the proposal that POLR3A and POLR3B mutations are a major cause of hypomyelinating leukodystrophies and suggest that POLR3A mutations are more frequent.
Statistics from Altmetric.com
Leukodystrophies are a heterogeneous group of inherited neurodegenerative disorders characterised by abnormal central nervous system white matter, affecting mainly the synthesis and maintenance of cerebral myelin.1 ,2 It is estimated that despite extensive investigations, at least 30–40% of patients remain without a molecular diagnostic. Leukodystrophies are further classified into demyelinating and hypomyelinating according to their MRI characteristics, and several of them have now been genetically described.3 Recently a group of four clinically overlapping hypomyelinating leukodystrophies (MIM 607694, MIM 614381) has been associated with mutations in polymerase III enzyme (Pol III) subunits.4–8 Clinically, the patients affected by Pol III-related leukodystrophies present between early childhood to adolescence, have motor regression with upper motor neuron signs, cerebellar features, and some milder degree of cognitive dysfunction and/or regression. In most cases dental abnormalities and/or hypogonadotropic hypogonadism are also present.9 The MRI features show partial deficits in the deposition of myelin in the brain referred to as hypomyelination.10
Pol III is an enzyme responsible for the transcription of a group of more than 200 small untranslated RNAs (eg, tRNAs, 5S, 7SK RNA, and U6 RNA) involved in the regulation of essential cellular processes, such as transcription, RNA processing and translation.11 The enzyme is composed of 17 subunits. POLR3A (MIM 614258) and POLR3B (MIM 614366) are the two largest and together form the enzyme's catalytic centre. It has been suggested that mutations in these two subunits cause hypomyelinating leukodystrophy by disturbing the assembly of the pol III complex or by leading to an abnormal enzymatic activity, thus affecting the level of small RNAs important for the development of white matter.5
In this study, we report a mutational analysis of a group of 14 patients presenting clinical and imaging characteristics of Pol III-related leukodystrophies.10 POLR3A and/or POLR3B were sequenced in all cases and mutations were uncovered in all 14 patients. These results further support POLR3A and POLR3B as a major cause of hypomyelinating leukodystrophies.
All patients included in this study were selected according to a combination of clinical findings and brain MRI features supporting the existence of Pol III-related hypomyelinating leukodystrophies.10 All patients and family members signed an informed consent from at least one of the following four institutions who had approved this research: the Centre de Recherche du CHUM (Montreal, Quebec, Canada), the Montreal Children's Hospital/McGill University Health Center Research Institute (Montreal, Quebec, Canada), the National Institute of Neurological Disorders and Stroke (NIH Bethesda, Maryland, USA), and the Children's National Medical Center (Washington DC, USA).
Clinical and MRI selection criteria
Patients were selected for sequencing if they had compatible clinical and/or radiological features of Pol III-related leukodystrophies.10 ,12 More specifically, compatible clinical features were: motor delay/regression, pyramidal and cerebellar features, dental abnormalities, and hypogonadotropic hypogonadism. The patients did not need to have all the features. These clinical features are summarised by Bernard and Vanderver.12 The online supplementary table S1 summarises the clinical features of the patients. As for the MRI features, hypomyelination was a necessary criteria,3 that is, hyperintense signal of the white matter compared with grey matter structures in T2-weighted sequences and variable signal of the white matter (ie, hypo-, iso- or hyperintense) compared to grey matter structures in T1-weighted images. Compatible MRI features included those reported by Steenweg et al10 and summarised by Bernard and Vanderver12: T2 hypointensities (relative preservation of the myelin) of the following structures: dentate nuclei, anterolateral nuclei of the thalami, globi pallidi, optic radiations and/or corticospinal tracts at the level of the internal capsules, with or without the following: cerebellar atrophy, thin corpus callosum, white matter atrophy most prominent posteriorly. The online supplementary figure S1 illustrates the MRI features of patient 3. Of note, MRI was not available for patient P1 as she died before MRI became available. Her brain pathology was, however, compatible the diagnosis of leukodystrophy.
Genomic DNA was extracted from peripheral blood cells using standard methods. The entire coding sequence and intron–exon boundaries of POLR3A (NM_007055.3) and/or POLR3B (NM_018082) genes were amplified by PCR as previously described.4 ,5 Oligonucleotide primers were designed using the exon primer software from the UCSC (University of California, Santa Cruz) genome browser (http://genome.ucsc.edu) or the Primer 3 software (http://frodo.wi.mit.edu/primer3/). PCR products were sequenced at the Genome Quebec Innovation Center and McGill University (Montreal, Quebec, Canada) using an 3730XL Genetic Analyzer (Applied Biosystems, Foster City, California). All mutations were confirmed in independent reactions by sequencing both strands. SeqMan v4.03 (DNASTAR, Inc, Wisconsin, USA) and Mutation Surveyor v4.0.6 (SoftGenetics, Pennsylvania, USA) softwares were used for mutation detection analyses. Co-segregation analysis of the identified mutations was performed by sequencing the corresponding amplicons in available family members.
Total RNA was extracted from lymphoblastoid cells derived from the patients with the predicted splice site mutations. cDNA was generated from RNA using the Superscript III enzyme (Invitrogen). Reverse transcription PCR (RT-PCR) was performed using POLR3B specific primers and RT-PCR products were subsequently purified and sequenced.
The entire coding region and splice junctions of POLR3A and/or POLR3B genes were sequenced in 14 patients with a hypomyelinating leukodystrophy suggestive of Pol III-related leukodystrophies, according to typical clinical and MRI features described previously.10 ,12 All 14 patients were found to carry mutations in one of the two genes: seven had mutations in POLR3A and seven in POLR3B (table 1).
In POLR3A, 10 mutations were identified, eight of which are novel (table 1). Three patients had homozygous missense mutation: p.Val166Ile was found in patient P1 and p.Trp671Arg in two siblings (P2 and P3). The remaining four patients were compound heterozygotes. Among these, one patient (P4) carried an insertion of one base pair (c.1741insA) in exon 13 leading to a frameshift mutation and a premature stop codon (p.Lys581Serfs*28) in addition to a missense mutation p.Ser602Arg; one patient (P6) had an insertion of adenine at position 1302 leading to a premature stop codon (p.Tyr434*) and a missense mutation p.Met852Val; the two remaining patients carried two missense mutations each—P5 had p.Glu1261Lys and p.Ala387Gly mutations, and P7 had p.Pro91Leu and p.Arg1005Cys mutations. As previously reported,4 none of these patients is homozygote for null mutations. Segregation analyses in families for which DNA was available in parents showed that the unaffected parents carry one heterozygous mutation each. None of these mutations was found in the Single Nucleotide Polymorphism database (dbSNP 135). Only the p.Glu1261Lys and p.Val166Ile missense mutations were found each in one individual out of 6503 from the National Heart, Lung, and Blood Institute Exome Variant Server in the heterozygous state. Finally, all the missense mutations identified in POLR3A affect evolutionarily conserved amino acids (see online supplementary figure S2A).
Eight mutations were identified in POLR3B, including seven that had not been previously described (table 1). All patients were found to be compound heterozygotes. Six patients carried the common missense mutation p.Val523Glu in exon 15, in addition to another mutation (table 1). Haplotypes of the carrier chromosome for this shared mutation were extracted from sequencing results. Five surrounding single nucleotide polymorphisms (SNPs) in these six patients, as well as three previously reported patients carrying this common missense mutation,5 shows that all carriers share a common haplotype, suggesting that the p.Val523Glu mutation derives from a single ancestral mutation (see online supplementary table S1). This haplotype was only observed in individuals carrying the p.Val523Glu mutation. In addition to the common mutation, two unrelated patients P8 and P9 carried a substitution of an adenine to a guanine (c.2084-6A>G) in intron 19, six base pairs before exon 20.
To determine the impact of this mutation on splicing, we performed RT-PCR targeting exons flanking exon 20. We demonstrated that the c.2084-6A>G mutation caused the creation of a cryptic acceptor splice site, leading to a frameshift and a premature stop codon (p.Gly695Valfs*5) five amino acids downstream (figure 1A). One patient (P13) had a missense mutation p.Leu104Phe in exon 6 in addition to a mutation affecting the consensus donor splice site of intron 22 (c.2570+1G>A). Similarly, RT-PCR analysis using total RNA extracted from this patient showed that this mutation cause the skipping of exon 22, leading to a frameshift and a premature truncated protein (p.Gly818Alafs*13) (figure 1B). One patient (P11) carried the common mutation in exon 15 in addition to a one base pair substitution (c.2817+30T>A) in intron 24, predicted to create a cryptic acceptor splice site in intron 24 by the BDGP (Berkeley Drosophila Genome Project) Splice Site Prediction by Neural Network (http://www.fruitfly.org/seq_tools/splice.html). The remaining four patients carried the common missense mutation p.Val523Glu in addition to another missense mutation each, as p.Cys527Arg was found in P10, p.Arg442Cys in P12, and p.Ser268Gly in P14. As for POLR3A, none of these mutations was found in the Single Nucleotide Polymorphism database (dbSNP 135), except the common mutation (p.Val523Glu) which was found in 2/374 (0.5%) of control chromosomes, and the intronic variant (c.2817+30T>A) which was observed in the 1000 Genomes Project with a minor allele frequency of 0.001. Finally, all missense mutations appear to be located at regions conserved across species (see online supplementary figure S2B).
In this study, we undertook a mutational analysis of POLR3A and POLR3B genes to evaluate the type and frequency of mutations in a cohort of 14 patients with hypomyelinating leukodystrophies, carefully selected according to a combination of diagnostic clinical and brain MRI features as described by Steenweg et al in 201010 and reviewed by Bernard and Vanderver in 2012,12 respectively. Recessive mutations were uncovered in all 14 patients. Six unrelated patients carried 10 mutations in POLR3A, eight of which were novel. These mutations included eight missenses, one nonsense, and one frameshift mutation (table 1), and they are distributed throughout the gene and not clustered in a specific region or domain of the encoded protein. In POLR3B, seven novel mutations as well as one previously described mutation were identified in seven patients. Among these, six patients carried the common missense mutation p.Val523Glu in exon 15. While these patients are of different ethnic backgrounds, we show for the first time that this common mutation results from an ancestral haplotype shared by these six individuals as well as three additional patients previously reported to carry this mutation.5 The seven novel mutations in POLR3B included four missenses, two splice sites, and one intronic mutation (table 1). The splice site mutations cause a frameshift leading to predicted truncated proteins of different sizes. To the best of our knowledge, this is the first description of POLR3B splicing mutations leading to premature stop codons. On the other hand, the intronic mutation in patient P11 is located 30 base pairs after the intron 24 donor splice site. Although quite far from the consensus canonical splice sites, this mutation was predicted by bioinformatic analysis to create a cryptic acceptor splice site in intron 24 that would probably lead to a frameshift in the encoded protein. Unfortunately, we could not verify this hypothesis as cell lines from this patient were not available.
Here, we show that mutations in POLR3A and POLR3B genes may occur at any given coding exon or splicing junctions of these genes, highlighting the high degree of mutational heterogeneity of Pol III-related hypomyelinating leukodystrophies. Yet, POLR3A sequencing should be prioritised for genetic testing since 73% (27/37) of all patients in our cohort carry mutations in POLR3A and 27% (10/37) of patients have mutations in POLR3B.4 ,5 ,7 On the other hand, when mutations in POLR3A are ruled out, POLR3B should be sequenced. The exon 15 of POLR3B appears to be a ‘hotspot’ for mutations as four different mutations (the common missense mutation as well as three other mutations) were identified in this specific exon, which correspond to a third of all the mutations uncovered by our group so far in this gene. Consequently, the exon 15 of POLR3B could be prioritised for sequencing in patients for whom no mutation is identified in POLR3A.
Taken together, mutations in either POLR3A or POLR3B genes were uncovered in 14 patients with a hypomyelinating leukodystrophy phenotype in this study. Combining these numbers with our previous reports4 ,5 brings the number of patients with mutations in these two genes to 37—27 in POLR3A and 10 in POLR3B, respectively. This very high yield of mutation detection is likely due to our stringent selection criteria, but it strongly suggests that mutations in these two genes represent a major cause of hypomyelinating leukodystrophies with or without dental abnormalities and/or hypogonadotropic hypogonadism.
We wish to thank all family members for their participation. We would also like to thank for their services the Genome Quebec Innovation Center and McGill University. Dr Bernard wishes to thank the Montreal Children's Hospital Foundation, the MSSA (Medical Staff Service Association), and the Montreal Children's Hospital Associates in Neurology. The project was funded by Fondation sur les Leucodystrophies, the Montreal Children's Hospital and McGill University Health Center Research Institutes, and the Myelin Disorders Bioregistry project, with special thanks to Johanna L Schmidt. HD received a postdoctoral fellowship from the Canadian Institutes of Health Research (CIHR). MT received the Frederick Banting and Charles Best Doctoral scholarship from the CIHR. GB received a Research Scholar Junior 1 of the Fonds de Recherche du Québec en Santé (FRQS).
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
HD and MT contributed equally.
Contributors HD, MT, and KG performed the molecular genetic analyses. HD and MT wrote the manuscript. WG, AC, JGA, MS, BB, CS, RSC, CG, SN, AV and GB referred the patients, provided the patient's clinical informations, and revised the manuscript. GB supervised the study, wrote and revised the manuscript. GB and AV reviewed the patients' MRIs.
Funding The study was funded by Fondation sur les Leucodystrophies, the Montreal Children's Hospital and McGill University Health Center Research Institutes, and the Myelin Disorders Bioregistry project.
Patient consent Obtained.
Competing interests None.
Ethics approval The Centre de Recherche du CHUM (Montreal, Quebec, Canada), the Montreal Children's Hospital/McGill University Health Center Research Institute (Montreal, Quebec, Canada), the National Institute of Neurological Disorders and Stroke (NIH Bethesda, Maryland, USA), and the Children's National Medical Center (Washington DC, USA).
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.