Mitochondrial diseases are a diverse group of inherited disorders that interfere with the ability of cells to produce and use energy. Because mitochondria are important in nearly every tissue, these disorders can affect the brain, muscles, heart, liver, and other organs. Symptoms and severity can also vary considerably from one patient to another.

Finding the genetic cause can be equally challenging. DNA sequencing has become an important diagnostic tool for mitochondrial disease, but as many as half of patients may remain without a definitive molecular diagnosis after genetic testing.

Researchers at the Department of Neurology at Beijing Children’s Hospital, Capital Medical University investigated whether RNA sequencing could provide another layer of information for children whose mitochondrial disease remained unexplained.

Looking beyond DNA sequencing

Whole exome sequencing, or WES, examines the protein-coding portions of the genome and can identify genetic variants that may cause disease. However, finding a variant does not necessarily reveal what that variant actually does inside a cell.

Some variants are classified as variants of uncertain significance because there is not enough evidence to determine whether they are harmful. Other disease-causing variants can occur outside the regions typically captured by exome sequencing.

RNA sequencing provides a complementary perspective because it measures the RNA molecules produced from genes. Researchers can therefore look for functional consequences of genetic variants, such as abnormal gene expression or improper RNA splicing.

RNA-based diagnostic workflow for WES inconclusive pediatric patients
with suspected mitochondrial disorders

(A) Clinical characteristics of 140 pediatric patients in the cohort. (B) Study design and outcome. Patients were categorized into two groups: candidate (n=28), in whom WES had prioritized variants and RNA-seq was used to assess the functional effect (molecular diagnosis in 20/28, 71.4%); and unsolved (n=112), with no prioritized variants, in whom RNA-seq was used to identify the candidate genes with aberrant RNA phenotypes followed by WES reanalysis or WGS (diagnosis in 15/112, 13.4%). The overall diagnostic yield of 25% (35/140) was achieved. WES: whole-exome sequencing, NMD: nonsense-mediated mRNA decay.

Using RNA sequencing in undiagnosed children

The researchers performed RNA sequencing on skin fibroblasts collected from 140 pediatric patients with suspected mitochondrial disease who remained undiagnosed after whole exome sequencing.

They analyzed the RNA using the Detection of RNA Outliers Pipeline, or DROP, which can identify unusual patterns of gene expression and RNA splicing.

The patients were divided into two groups. The first included 28 patients whose exome sequencing had already identified potentially relevant variants. For these patients, RNA sequencing was used to determine whether those variants actually disrupted RNA.

The second group consisted of 112 patients for whom WES had not provided a strong candidate. Here, RNA sequencing was used to search for abnormal RNA patterns that could point researchers toward previously overlooked genes or variants.

RNA sequencing increases the diagnostic yield

Combining RNA sequencing with DNA sequencing resulted in a genetic diagnosis for 35 of the 140 patients, an overall diagnostic yield of 25%.

The impact was particularly substantial among patients who already had candidate variants from exome sequencing. RNA analysis helped establish diagnoses for 20 of 28 patients in this group, or 71%.

Among the 112 previously unsolved patients, RNA sequencing contributed to diagnoses for 15 patients, or 13%. In six cases where RNA analysis detected an abnormal event but exome sequencing could not identify its genetic cause, the researchers added whole genome sequencing to investigate further.

Revealing hidden problems with RNA splicing

One of the most important contributions of RNA sequencing was detecting abnormal splicing.

Before RNA can be used to produce a protein, sections called introns generally must be removed and the remaining exons joined together. Genetic variants can interfere with this process, causing sections of RNA to be incorrectly included, excluded, or joined.

Some of the variants responsible for these abnormalities had not been accurately identified as damaging by computational prediction tools. Directly examining the RNA allowed the researchers to see the consequences of those variants inside patient-derived cells.

The analysis also identified several types of disease-associated variants among previously unsolved patients, including synonymous, missense, deep intronic, near-splice-site variants, and large deletions.

Finding disease effects beyond protein-coding variants

The findings also demonstrate why focusing only on protein-coding DNA can leave some genetic diseases unexplained.

When the researchers examined 233 pathogenic variants associated with abnormal RNA phenotypes from their work and previous reports, approximately half were coding variants and half were noncoding variants.

One notable example involved ECHS1. The researchers identified a recurrent synonymous variant associated with seven cases. Synonymous variants do not change the amino acid specified by the genetic code, so they can sometimes appear harmless when evaluated primarily by their predicted effect on protein sequence. RNA analysis can reveal whether such variants instead interfere with processes such as RNA splicing.

Testing predictions of nonsense-mediated decay

RNA sequencing also allowed the researchers to examine nonsense-mediated decay, or NMD.

NMD is a cellular quality-control mechanism that can recognize and remove abnormal RNA transcripts containing premature stop signals. Computational tools can predict whether a genetic variant will cause an RNA molecule to undergo this process.

However, the researchers found that 14% of protein-truncating variants predicted to undergo NMD escaped degradation. The result demonstrates an important limitation of relying entirely on computational predictions to determine how a genetic variant will behave in a patient’s cells.

Connecting genetic variants with their functional consequences

DNA sequencing identifies differences in the genetic code, while RNA sequencing can provide evidence about what those differences actually do to gene expression and RNA processing.

This distinction can be particularly valuable for rare genetic diseases. A patient may carry a suspicious DNA variant, but determining whether it disrupts normal biological function is often necessary before that variant can confidently be connected with disease.

For mitochondrial disorders, integrating RNA sequencing with whole exome and whole genome sequencing provides a way to connect genotype with functional consequences. The researchers conclude that transcriptome analysis can reveal cryptic splicing abnormalities, regulatory effects, and unexpected NMD behavior that may remain hidden when DNA is examined alone.