Reply: Biallelic POLR3A variants confirmed as a frequent cause of hereditary ataxia and spastic paraparesis

Sir,

We thank Rydning et al. (2019) for their letter. With great interest, we have read their results confirming those of our previous publication (Minnerop et al., 2017) in an independent study of patients with hereditary ataxia and spastic paraparesis (HSP) from Norway (Rydning et al., 2019). The authors investigated exome data form 95 Norwegian patients with the clinical diagnosis of ataxia or HSP without a molecular diagnosis. They identified 13 patients belonging to 10 families carrying biallelic and presumably disease-causing POLR3A variants, forming the second most common cause of recessive ataxia or HSP in the Norwegian cohort. The resulting frequency of 3.1% for biallelic POLR3A variants was remarkably identical to the frequency found in our patient cohort (Minnerop et al., 2017). Even more, the intronic variant c.1909+22G>A, a key finding in our patient cohort, was identified in almost all patients of the Norwegian sample (12/13); only one patient was not compound heterozygous for this intronic variant. Interestingly, this single case carried the c.1771–6C>G variant in a homozygous state (consanguinity), as described for this variant before (Azmanov et al., 2016; Minnerop et al., 2017).

The fact that the c.3655G>T variant was found in six of nine index patients compound heterozygous with the intronic variant c.1909+22G>A led the authors to hypothesize common ancestors among these patients. However, haplotype analysis did not point towards a founder effect—neither for the intronic c.1909+22G>A variant nor for the exonic c.3655G>T variant. However, additional genomic information would be necessary to explore cryptic relatedness between these patients.

As pointed out by the authors, the phenotype described for the Norwegian patients was also remarkably similar to the phenotype described in our patient cohort. This comprised mean age at onset, the presence of ataxia, frequently found action tremor, pyramidal signs of the lower limbs combined with paresis and muscle atrophy, as well as sensory disturbances. Interestingly, the authors reported the tremor to be alcohol-responsive. This is in accordance with the clinical findings in one patient of our previously published cohort (Minnerop et al., 2017), who even developed alcohol abuse in the meantime (now abstinent).

Thus, taking ours and Rydning et al. data together suggest that POLR3A mutations and in particular the intronic variant should be considered as differential diagnosis in patients with alcohol-responsive tremor, often classified as ‘essential tremor’, which might even occur in combination with ataxia symptoms. This also underscores the importance of a thorough neurological examination in this patient group, since it might reveal further hints to POLR3A mutations as the underlying cause (e.g. subclinical pyramidal signs, eventually only detected by motor evoked potentials). Most interestingly, Rydning et al. also reported abnormal brainstem evoked potentials in most of their patients, further supporting the impairment of central sensory pathways.

Regarding dental problems or hormonal dysfunction in patients carrying the intronic POLR3A variant, these signs were discovered only by specific questions on these aspects as we reported previously and now also confirmed by Rydning and colleagues. Gathering this information should, therefore, be included in the routine workup of patients with movement disorders. The finding of reduced p-folate (vitamin B9) in five of nine patients of the Norwegian cohort is interesting and warrants further investigation.

We are very grateful to Rydning et al. for the careful evaluation of MRI images in their patient group that further confirmed our report of bilateral hyperintensity along the superior cerebellar peduncles in patients carrying the c.1909+22G>A variant. This radiological finding has also been confirmed by Gauquelin et al. (2018) who re-evaluated magnetic resonance images of previously published patients with atypical POLR3-related disorders (La Piana et al., 2016) and found hyperintense superior cerebellar peduncles in four of eight previously published cases. In contrast to our patient cohort, Rydning et al. also found signs of mild hypomyelination, further underlining the variability of MRI findings observed in this patient group. Next to the evaluation of potential regional brain atrophy/hypoplasia (cervical cord, cerebellum, corpus callosum), the careful inspection of FLAIR images regarding hyperintensity of the superior cerebellar peduncles and signs of hypomyelination should be considered as further diagnostic key elements in suspected POLR3-related movement disorders.

Rydning et al. included patients with ataxia and HSP without refining the cohort to patients presenting typical symptoms and signs previously associated with POLR3A mutations and thus built a ‘hypothesis-free’ cohort (broader clinical spectrum). This strategy led the authors to confirm our previous observations in regard to the frequency of c.1909+22G>A and the overall rate of POLR3A mutations in the entire cohort. Conversely, Gauquelin et al. could not confirm our observations in a previous report (Gauquelin et al., 2018). Importantly, and as we previously stated in our reply (Minnerop et al., 2018), the authors in Gauquelin et al. (2018) have long-standing research experience on the genetics of POLR3A, which may have introduced bias in the selection of their study cohort (Daoud et al., 2013; Wolf et al., 2014; Cayami et al., 2015). The clinical awareness of the authors regarding features typically associated with POLR3A mutations may have produced a systematic depletion of potential POLR3A mutation carriers in both HSP and ataxia cohorts used by Gauquelin et al. (2018) resulting in the lower frequency of POLR3A mutation carriers in this study. Rydning et al. now independently confirm the relative frequency of POLR3A mutations in unsolved HSP and ataxia cases.

Together, the results presented by Rydning et al. and our previous report on POLR3A mutations, provide an additional example of the growing number of genes in which the clinical spectrum is changing due to findings from ‘hypothesis-free’ cohorts combined with genome-wide methods (Mefford, 2009). In fact, research following a ‘genotype-first’ approach in unbiased ‘hypothesis-free’ cohorts is expanding the spectrum of phenotypes associated with mutations in a particular gene (Dennert et al., 2017).

From a translational point of view, the results on POLR3A open important opportunities for molecular diagnostics and novel therapeutic options. In fact, based on the frequency of POLR3A mutations identified in sporadic or recessive forms of ataxia and HSP patients, this gene should be included by default in molecular diagnostic pipelines for the diagnostic work-up of movement disorders. Furthermore, the majority of the identified patients in our study and in Rydning et al. carry the identical intronic mutation. The activation of this cryptic splicing site can be hypothesized as common pathogenic mechanism in these patients. Thus, a therapeutic approach focusing on the modification of the aberrant splicing produced by c.1909+22G>A is likely to be beneficial for the great majority of the ataxia and HSP patients with POLR3A mutations. Herein, antisense oligonucleotide-based therapy offers an interesting therapeutic perspective as it has been proven to successfully modify splicing and clinical outcomes in other neuromuscular disorders, such as Duchenne muscular dystrophy and spinal muscular atrophy (Sardone et al., 2017).

In conclusion, the work of Rydning et al. provides further evidence supporting the relevant contribution of POLR3A to the disease mechanisms operating in ataxia and HSP, and that these mechanisms are shared with other leukodystrophy phenotypes, caused by mutations in POLR3A or POLR3B. Furthermore, detailed clinical characterization of these patients will allow the establishment of better criteria for the differential diagnosis. Finally, therapies targeting the aberrant splicing are promising strategies to treat HSP and ataxia caused by mutations in POLR3A.

Data availability

Data sharing is not applicable to this article as no new data were created or analysed in this study.

Funding

No funding was received towards this work.

Competing interests

The authors report no competing interests.

References