Genetic diagnostics have undergone a revolution in the last decade, fueled by technological advances heralded by the development of massively parallel DNA sequencing. Within a remarkably short time, the new chemistry moved from the research laboratory into clinical practice, accelerating the pace of gene discovery and allowing the rapid diagnosis of genetic disorders on an unprecedented scale. There is a strong argument that so-called next or third generation sequencing will have the greatest effect in neurology, which is characterized by a seemingly endless list of discrete clinical syndromes, many thought to have a unique genetic basis. Until 2011, clinical neurogenetic practice has been frustrating. It has been difficult to screen more than a handful of the known genetic causes of a particular disorder, but we now face the real prospect of a reaching genetic diagnosis for every patient walking to the clinical door. Mitochondrial disorders provide a good illustration of the effect of this new technology in neurogenetic practice.
Mitochondrial disorders are the bane of the generalist. The myriad of clinical features overlap with common neurological and nonneurological diseases. Despite entering the differential diagnosis of the most common neurological diseases, mitochondrial disorders themselves are rare, and the diagnostic approach is complex, time consuming, and expensive. Following a detailed history, examination, and clinical investigations, many patients require an invasive biopsy of clinically affected tissue before samples are dispatched on ice to a limited number of laboratories worldwide. Although a positive biochemical result can substantiate a clinical diagnosis (muscle histochemistry or respiratory chain complex analysis), the results are often inconclusive. Subtle histochemical defects can occur as part of healthy aging, and individuals deconditioned from any cause can have low respiratory chain enzyme activities in skeletal muscle. However, pursuing a genetic diagnosis is important because similar clinical and biochemical phenotypes can be sporadic, maternally inherited (through mitochondrial DNA [mtDNA]), autosomal dominant, autosomal recessive, and occasionally X-linked. Thus, defining the underlying gene defect can sway the recurrence risks from zero to very high and everything in between. Targeted genetic analysis typically proceeds on a step-by-step basis, guided by the clinical picture and the biochemical profile. A thorough laboratory workup takes months or years and is still inconclusive in approximately one-third of cases. If ever there was a need for a genetic revolution, mitochondrial disorders have a strong case.
The first genetically defined mitochondrial disorders were identified in the late 1980s and early 1990s. Being only 16.5 kb, mtDNA analysis was experimentally tractable, and the next decade saw major progress. Mitochondrial DNA deletions and point mutations were defined in patients with a growing array of phenotypes, leading to the first epidemiological studies. In adults, mtDNA mutations emerged as a common cause of inherited neurological disease (1:11 000). In children, nuclear gene defects seemed to predominate but were very challenging to define at the molecular level. Progress in defining nuclear-genetic mitochondrial disorders was slow and limited by the available technology of positional cloning, chromosomal transfer, and candidate gene analysis. This changed dramatically in the late 2000s, with the advent of massively parallel sequencing, enabling the rapid, cost-effective screening of hundreds of genes simultaneously.
In the early stages, the greatest success was seen in patients with a presumed autosomal recessive disorder and a defined biochemical defect affecting a single respiratory chain enzyme complex. This pointed toward homozygous or compound heterozygous mutations in a known list of candidate genes that could be captured in a multigene panel. Advances in understanding the protein composition of the mitochondrion (the mitochondrial proteome) led to the development of mitochondria–specific capture arrays.1 This helped to identify new disorders, but progress was hampered by limitations in the efficiency of the capture system (which isolated segments of genomic DNA of interest) and what now seems like a rudimentary understanding of genetic variation in the population. Growing population genetic sequence databases have revealed the extraordinary genetic diversity in healthy humans, but with tumbling sequencing costs and near-complete (>98%) coverage of all known protein coding regions (the exome, approximately 1% of the entire human genome), exome sequencing became the platform of choice by 2012.
Exome sequencing dramatically improved our ability to diagnose mitochondrial disorders at the genetic level. A particularly challenging group to diagnose was patients with multiple respiratory chain complex defects who did not have an underlying mtDNA defect. In vitro translation studies implicated a defect of intramitochondrial protein synthesis thought to have a nuclear genetic basis. The limited number of candidate genes meant that, even in 2011, it was only possible to make molecular diagnosis in less than 5% of patients.2 However, exome sequencing immediately increased the diagnostic yield to greater than 60%, revealing a remarkable heterogeneity of underlying gene defects.3 These findings explained why it had been so difficult to identify the disease genes previously, using strategies based on pooling unrelated individuals with a similar biochemical profile. The widespread use of exome sequencing has seen an exponential increase in the number of different disease genes identified in patients with biochemically proven mitochondrial disorders. Although diverse at the genetic level, many affect related pathways. Nuclear-encoded disorders of mtDNA replication and repair have emerged as a common theme, where disruption of the nucleotide pools or the multiprotein mtDNA replisome leads to accumulation of secondary mtDNA mutations during life. Although these findings explain the phenotypic overlap with classic “primary” mtDNA disorders, it is still not clear why there are gene-specific features in the different “disorders of mtDNA maintenance” such as spastic paraplegia in patients with SPG7 mutations or optic atrophy in patients with OPA1 mutations.
What about mtDNA in the third-generation sequencing era? First, because each diploid cell contains many copies of mtDNA (eg, approximately 500 in lymphocytes), even the most efficient nuclear chromosomal capture techniques are contaminated by mtDNA. The off-target mtDNA sequences can be identified bioinformatically, allowing reliable sequencing of mtDNA with standard exome approaches. Many pathogenic mtDNA point mutations are heteroplasmic, with a mixture of mutated and wild-type mtDNA in varying proportions, but the off-target sequence coverage is usually sufficiently high to allow the detection of heteroplasmic variants down to approximately the 10% level.4 There is 1 caveat: it should be remembered that common mtDNA mutations may not be detectable in blood (such as large-scale mtDNA deletions and the common heteroplasmic m.3243A>G mutation in adults). This means a tissue biopsy may still be required to make the genetic diagnosis (Figure). However, with a greater understanding of genes known to cause mitochondrial disorders, we are moving to a point where a diagnosis will be possible for most patients using a single blood sample. There will be exceptions, so for the immediately foreseeable future, it is important that diagnostic laboratories retain their biochemical expertise, not least to validate the mechanism implicated by a novel genetic finding, a so-called variant of uncertain significance (Figure).
In 2016, patients with progressive external ophthalmoplegia or other nonspecific phenotypes required a muscle biopsy for mitochondrial DNA and histochemical/biochemical analysis. Patients with specific clinical syndromes (such as Leber hereditary optic neuropathy, a POLG disorder, or autosomal-dominant optic atrophy) point toward a specific genetic blood test. If results are negative, the next step is a next-generation sequencing multigene panel. If a variant of unknown significance is detected, then a muscle biopsy is indicated to determine whether the biochemical defect corresponds to the molecular defect. If the multigene panel results are negative, exome sequencing is indicated (which may detect a variant of unknown significance), and if results of this are negative, whole-genome sequencing is indicated. The interpretation of whole-genome sequence is challenging at present and may be helped by biochemical analyses, transcriptomics, metabolomics, proteomics, and other functional studies. With whole-genome sequencing costs falling to approximately $1000 (plus bioinformatics and validation costs), several of these steps will be bypassed beyond 2016, leading to a single genetic blood test for most patients.
Despite massive progress, clinical exome sequencing does not solve all cases. There are several reasons for this. Specific regions of the genome are resistant to exome capture (typically those with a high content of G and C residues, often found in the first exon of a gene), large-scale deletions are easily missed, and 99% of the human genome sequence is noncoding space where important gene regulatory regions exist.5 Several whole-genome sequencing projects are under way including the 100 000 Genomes Project in the United Kingdom, which already includes approximately 200 patients with suspected mitochondrial disease with no molecular diagnosis. International efforts will build our understanding of genetic variation in the noncoding space, facilitating our interpretation of novel genetic findings, but much will remain uncertain until time-consuming functional work validates a putative pathogenic mechanism. RNA sequencing provides an unbiased survey of transcript isoform diversity caused by varied splicing and may provide further evidence for the pathogenic role of noncoding variants. However, splicing patterns are largely tissue specific and their interpretation in mitochondrial disease is still difficult. The role of other ‘omics technologies, such as proteomics and metabolomics, is gaining importance in defining biomarkers and understanding disease mechanisms, but their role in the diagnostic process is less clear.6
These are exciting times, and it may seem that the clinical discipline of neurogenetics is under threat. However, nothing could be further from the truth. Genetic neurology has emerged as a major superspecialty, with a vast array of disorders affecting approximately 1 in 500 of the population.7 The field is in transition. Although major challenges need to be resolved in the near future, not least the correct interpretation of novel variants, with improved prediction tools and a strong commitment to data sharing, achieving a molecular diagnosis will soon be straightforward. This is good news for patients and neurologists. But it is now time to harness this knowledge to develop new treatments based on our understanding of the genetic mechanisms. It is time for precision medicine in neurogenetics, based on a comprehensive understanding of our patient’s genome.
Corresponding Author: Patrick F. Chinnery, FRCP, FMedSci, Department of Clinical Neurosciences, University of Cambridge, Cambridge Biomedical Campus, Cambridge CB2 0QQ, United Kingdom (firstname.lastname@example.org).
Published Online: November 14, 2016. doi:10.1001/jamaneurol.2016.4506
Conflict of Interest Disclosures: Dr Chinnery is a Wellcome Trust Senior Fellow in Clinical Science and a UK National Institute for Health Research Senior Investigator who receives support from the Medical Research Council Mitochondrial Biology Unit, the Wellcome Trust Centre for Mitochondrial Research, the UK Medical Research Council Centre for Translational Muscle Disease, EU FP7 TIRCON, and the National Institute for Health Research Biomedical Research Centre based at Cambridge University Hospitals National Health Services Foundation Trust and the University of Cambridge. Dr Horvath is a Wellcome Trust Investigator who receives support from the UK Medical Research Council, the European Research Council, the Wellcome Trust Pathfinder Scheme, the Newton Fund, and the FP7-PEOPLE-ITN.
Disclaimer: The views expressed are those of the authors and not necessarily those of the National Health Services, the National Institute for Health Research, or the Department of Health.
Horvath R, Chinnery PF. The Effect of Neurological Genomics and Personalized Mitochondrial Medicine. JAMA Neurol. 2017;74(1):11-13. doi:10.1001/jamaneurol.2016.4506