Long-read Genome Sequencing for the Molecular Diagnosis of Dystonia

Evaluation of the Value of Long-read Genome Sequencing for the Molecular Diagnosis of Dystonia: a Prospective Multicenter Study

Dystonia is a motor disorder caused by involuntary, intermittent, or sustained muscle contractions, leading to abnormal movements or postures. It can affect any body region and often results in significant functional disability and healthcare burden. Although its familial nature was recognized early on, the advent of high-throughput DNA sequencing has dramatically increased the identification of dystonia-associated genes. Dystonia now encompasses all modes of inheritance-autosomal dominant (e.g., TOR1A, KMT2B), autosomal recessive, X-linked, and mitochondrial-and over 100 genes have been implicated. Many forms involve structural variants (SVs) or copy number variations (CNVs), which are challenging to detect using standard short-read sequencing (srWGS).

Molecular diagnosis is essential, ending the diagnostic odyssey and enabling genetic counseling, prognosis, reproductive planning, and-in some cases-targeted therapies. For instance, GNAO1-related dystonia may respond to deep brain stimulation, while dopa-responsive dystonia benefits from levodopa.

Despite advances, srWGS has key limitations, especially for detecting repeat expansions, SVs, and phasing alleles. This likely explains the low diagnostic yield in dystonia compared to other neurological disorders, with over 70% of cases remaining unsolved.

Long-read sequencing (lrWGS), such as Oxford Nanopore technology, overcomes many of these challenges by reading native DNA fragments thousands of bases long. It enables comprehensive detection of SNVs, indels, SVs, CNVs, methylation changes, and repeat expansions-including known and newly discovered pathogenic expansions (e.g., in NOTCH2NLC). It also allows phasing without parental samples, which is crucial in recessive cases.

The investigators propose that lrWGS could significantly increase the diagnostic yield in dystonia, improving patient care, enabling appropriate genetic counseling, and paving the way for personalized treatment strategies.

Dystonia is a motor disorder characterized by abnormal movements and/or postures. It is caused by involuntary, intermittent, or sustained muscle contractions. Dystonia can affect all body segments, leading to highly variable and often severe functional impairment with a significant burden on healthcare systems.

The familial nature of some generalized dystonias was recognized as early as the 1940s. With the development and increased availability of high-throughput DNA sequencing technologies since the early 2010s, there has been an exponential rise in the discovery of genes associated with neurological phenotypes in which dystonia is either the main manifestation or part of a broader clinical picture. Dystonia is now known to be associated with all modes of inheritance, including autosomal dominant forms (e.g., TOR1A, KMT2B), autosomal recessive forms often presenting with complex phenotypes, X-linked dominant (e.g., WDR45) and recessive (e.g., TAF1) forms, and even mitochondrial inheritance.

To date, over 100 genes have been linked to dystonic phenotypes, highlighting the significant genetic heterogeneity of this condition. Many genetically determined dystonic syndromes are associated with structural variants (SVs) or copy number variations (CNVs), which are difficult to detect using routine short-read sequencing technologies.

Molecular diagnosis is crucial in the management of dystonia. It provides a definitive and precise diagnosis, bringing an end to the diagnostic odyssey for patients and their families. It also enables accurate genetic counseling, offering information on inheritance risks and the potential for other family members to be affected. In some cases, a molecular diagnosis directly influences therapeutic decisions. For example, pathogenic variants in GNAO1 are associated with early-onset dystonic syndromes that respond well to deep brain stimulation of the globus pallidus. Specific symptomatic treatments have been proposed for certain dystonias, such as levodopa in dopa-responsive dystonia. At the population level, identifying new genes or molecular causes can reveal novel pathogenic mechanisms and facilitate the development of animal or cellular models, paving the way for future gene therapies.

Over the past decade, the growing availability of short-read DNA sequencing technologies has enabled the simultaneous analysis of numerous dystonia-associated genes. In routine diagnostics, short-read sequencing involves fragmenting DNA into small pieces (usually a few hundred base pairs), which are then sequenced from both ends.

This approach has limitations, particularly for detecting repeat expansions that exceed the size of the sequenced fragments. These limitations may partly explain why a relatively small proportion of dystonia cases receive a molecular diagnosis compared to other neurological disorders like cerebellar ataxias or hereditary spastic paraplegias. Overall, more than 70% of dystonia patients sequenced via short-read WGS remain without a molecular diagnosis, despite strong clinical indicators of a genetic etiology.

To overcome these limitations, long-read DNA sequencing technologies have been developed. Unlike short-read methods, long-read sequencing preserves native DNA fragment sizes, which can span thousands of bases. One such approach (Oxford Nanopore) sequences DNA based on ionic current changes as the molecule passes through a nanometer-scale pore. Long-read sequencing offers several advantages. Like short-read methods, it detects small variants (SNVs and indels). However, it also excels at identifying structural variants, CNVs, methylation changes, and especially repeat expansions-key elements often missed by standard methods.

In neurogenetics, long-read sequencing has enabled the accurate detection and quantification of known repeat disorders (e.g., C9orf72 hexanucleotide expansions, RFC1), as well as the discovery of new disease-causing repeat expansions (e.g., GGC repeats in NOTCH2NLC).

It also allows phasing of alleles (cis or trans configurations) without requiring family segregation studies, which is particularly useful in recessive conditions.

Given the current diagnostic limitations of short-read sequencing, the emergence of newly identified repeat expansions, and the demonstrated advantages of long-read sequencing, the investigators hypothesize that a significant proportion of patients with undiagnosed dystonia could benefit from whole-genome long-read sequencing (lrWGS).

Such an approach could, in a single test, overcome many of the key diagnostic blind spots of short-read methods, particularly for detecting repeat expansions and structural variants. This would not only end the diagnostic odyssey for many patients, but also enable appropriate genetic counseling, access to reproductive options such as prenatal or preimplantation diagnosis, and even influence treatment strategies.

Inclusion criteria - Index case:

• Index case affected by familial dystonia (≥1 first-degree relative affected) and/or sporadic early-onset dystonia (symptom onset before age 50), meeting the criteria of the PFMG-2025 program.
• Index case who has undergone short-read genome sequencing, which did not lead to a molecular diagnosis.
• Ability to understand and sign informed consent by the index case and/or their parents or legal guardians for patients under 18 years of age.
• Availability of a blood sample from the index case and at least two relatives, either affected or unaffected.

Inclusion criteria - Relatives:

• Symptomatic or asymptomatic relative of an index case, who has also undergone short-read genome sequencing without a conclusive molecular diagnosis.
• Ability to understand and sign informed consent.

Exclusion criteria:

• Index case or relatives who are not affiliated with or not beneficiaries of a social security scheme.
• Index case and their parents presenting with a condition that, in the opinion of the investigator, would contraindicate participation in the study.
• Suspected non-genetic etiology (e.g., perinatal hypoxic-ischemic injury, kernicterus, history of severe head trauma or central nervous system infection).