Interpretation challenges of novel dual-class missense and splice-impacting variant in POLR3A-related late-onset hereditary spastic ataxia

Uncovering recessive alleles in rare Mendelian disorders by genome sequencing of 174 individuals with monoallelic pathogenic variants

Clinical exome sequencing (ES) has facilitated genetic diagnosis in individuals with a rare genetic disorder by analysis of all protein-coding sequences in a single experiment. However, in 40-60% of patients, a conclusive diagnosis remains elusive. In 2-5% of these individuals, ES does identify a disease-associated monoallelic variant in a recessive disorder. We hypothesized that short-read genome sequencing (GS) might uncover a pathogenic variant on the second allele, thereby increasing diagnostic yield. We performed GS for 174 individuals in whom ES identified a monoallelic pathogenic variant in a gene associated with recessive disease related to their phenotype. GS interpretation was limited to the (non-)coding parts of the gene in which this first pathogenic variant was identified, focusing on splice-disrupting variants. Firstly, we uncovered a second pathogenic variant affecting coding sequence in five individuals, including two SNV/indel variants, two copy number variants, and one insertion. Secondly, for 24 individuals, we identified a total of 31 rare non-coding intronic SNV/indel variants, all predicted to disrupt splicing. Using functional follow-up assays, we confirmed an effect on splicing for three of these variants (in ABCA4, POLR3A and COL4A4) in three individuals. In summary, we identified a (likely) pathogenic second variant in 4.6% (8/174), and a possible diagnosis for 12.1% (21/174) of our cohort. Hence, when performing GS as first-tier diagnostic test, including the interpretation of SVs and rare intronic variants in known recessive disease genes, the overall diagnostic yield of rare disease will increase. The added diagnostic value of GS for recessive disease In our cohort of 174 individuals (84 males and 90 females) with a monoallelic pathogenic variant in genes associated with a wide and diverse range of recessive diseases (pie chart), using genome sequencing (GS) and a systematic approach (methods), we identified eight new diagnoses (4.6%). We identified a second likely pathogenic variant in eight individuals (results); In two a second coding variant was found, in three others, a rare non-coding SNV anticipated to disrupt splicing was uncovered, and in three individuals a structural rearrangement was identified (two copy number variants (CNV), and one structural variant (SV)).

POLR3A-related disorders: From spastic ataxia to generalised dystonia and long-term efficacy of deep brain stimulation

While biallelic POLR3A loss-of-function variants are traditionally linked to hypomyelinating leukodystrophy, patients with a specific splice variant c.1909+22G>A manifest as adolescent-onset spastic ataxia without overt leukodystrophy. In this study, we reported eight new cases, POLR3A-related disorder with c.1909+22 variant. One of these patients showed expanded phenotypic spectrum of generalised dystonia and her sister remained asymptomatic except for hypodontia. Two patients with dystonic arm tremor responded to deep brain stimulation. In our systemic literature review, we found that POLR3A-related disorder with c.1909+22 variant has attenuated disease severity but frequency of dystonia and upper limb tremor did not differ among genotypes.

Spinal cord-predominant neuropathology in an adult-onset case of POLR3A-related spastic ataxia

Biallelic mutations in POLR3A have been associated with childhood‐onset hypomyelinating leukodystrophies and adolescent‐to‐adult‐onset spastic ataxia, the latter of which has been linked to the intronic variant c.1909 + 22G>A. We report a case of adult‐onset spastic ataxia in a 75‐year‐old man, being a compound heterozygous carrier of this variant, whose brain and spinal cord were for the first time investigated by neuropathological examination. We describe prominent degeneration of the posterior columns, spinocerebellar tracts, and anterior corticospinal tracts of the spinal cord in a pattern resembling Friedreich's ataxia, with a notable lack of significant white matter pathology throughout the brain, in marked contrast with childhood‐onset cases. Immunohistochemical examination for the POLR3A protein demonstrated no apparent differences in localization or staining intensity between the proband and an age‐matched control subject. We demonstrate the clinicopathologic description of POLR3A‐related neurodegenerative disease and also mention the differential diagnosis of the childhood‐onset hypomyelinating leukodystrophy and late‐onset spastic ataxia phenotypes.

POLR3A variants in hereditary spastic paraparesis and ataxia: clinical, genetic, and neuroradiological findings in a cohort of Italian patients

Mutations in POLR3A are characterized by high phenotypic heterogeneity, with manifestations ranging from severe childhood-onset hypomyelinating leukodystrophic syndromes to milder and later-onset gait disorders with central hypomyelination, with or without additional non-neurological signs. Recently, a milder phenotype consisting of late-onset spastic ataxia without hypomyelinating leukodystrophy has been suggested to be specific to the intronic c.1909 + 22G > A mutation in POLR3A. Here, we present 10 patients from 8 unrelated families with POLR3A-related late-onset spastic ataxia, all harboring the c.1909 + 22G > A variant. Most of them showed an ataxic-spastic picture, two a "pure" cerebellar phenotype, and one a "pure" spastic presentation. The non-neurological findings typically associated with POLR3A mutations were absent in all the patients. The main findings on brain MRI were bilateral hyperintensity along the superior cerebellar peduncles on FLAIR sequences, observed in most of the patients, and cerebellar and/or spinal cord atrophy, found in half of the patients. Only one patient exhibited central hypomyelination. The POLR3A mutations present in this cohort were the c.1909 + 22G > A splice site variant found in compound heterozygosity with six additional variants (three missense, two nonsense, one splice) and, in one patient, with a novel large deletion involving exons 14-18. Interestingly, this patient had the most "complex" presentation among those observed in our cohort; it included some neurological and non-neurological features, such as seizures, neurosensory deafness, and lipomas, that have not previously been reported in association with late-onset POLR3A-related disorders, and therefore further expand the phenotype.

Two intronic cis-acting variants in both alleles of the POLR3A gene cause progressive spastic ataxia with hypodontia

POLR3A encodes the largest subunit of the DNA-dependent RNA polymerase III. Pathogenic variants in this gene are associated with dysregulation of tRNA production and other non-coding RNAs. POLR3A-related disorders include variable phenotypes. The genotype-phenotype correlation is still unclear. Phenotypic analysis and exome sequencing were performed in four affected siblings diagnosed clinically with hereditary spastic ataxia, two healthy siblings and their unaffected mother. All four affected siblings (ages 46-55) had similar clinical features of early childhood-onset hypodontia and adolescent-onset progressive spastic ataxia. None had progeria, gonadal dysfunction or dysmorphism. All affected individuals had biallelic POLR3A pathogenic variants composed by two cis-acting intronic splicing-altering variants, c.1909 + 22G > A and c.3337-11 T > C. The two healthy siblings had wild-type alleles. The mother and another unaffected sibling were heterozygous for the allele containing both variants. This is the first report addressing the clinical consequence associated with homozygosity for a unique pathogenic intronic allele in the POLR3A gene. This allele was previously reported in compound heterozygous combinations in patients with Wiedemann-Rautenstrauch syndrome, a severe progeroid POLR3A-associated phenotype. We show that homozygosity for this allele is associated with spastic ataxia with hypodontia, and not with progeroid features. These findings contribute to the characterization of genotype-phenotype correlation in POLR3A-related disorders.

Interpretation challenges of novel dual-class missense and splice-impacting variant in POLR3A-related late-onset hereditary spastic ataxia

Background

RNA polymerase III (Pol III)‐related disorders are autosomal recessive neurodegenerative disorders caused by variants in POLR3A or POLR3B. Recently, a novel phenotype of adult‐onset spastic ataxia was identified in individuals with the c.1909+22G>A POLR3A variant in compound heterozygosity.

Methods

Whole‐exome sequencing was performed in the proband and parents. Variants were confirmed by Sanger sequencing. RNA sequencing was performed to evaluate splicing implications.

Results

A 42‐year‐old female was evaluated for unexplained neurological findings with a slow progressive decline in gait and walking speed since adolescence. WES revealed a novel missense variant (c.3593A>C, p.Lys1198Arg) in exon 27 of POLR3A in compound heterozygosity with the c.1909+22G>A variant. Summary of previously reported clinical features from individuals with pathogenic biallelic alterations in POLR3A and adult‐onset phenotype is consistent with our findings. RNA analysis revealed c.3593A>G drives the production of four RNA transcript products each with different functional impacts.

Conclusion

The novel dual‐class c.3593A>C variant in POLR3A causes an amino acid substitution and complex disruption of splicing. Our report supports the need to investigate variants near splice junctions for proper interpretation. Current interpretation guidelines need to address best practices for inclusion of predicted or measured transcriptional disruption pending functional activity or reliable transcript abundance estimates.