NEW YORK (GenomeWeb) – Rare genetic diseases are increasingly diagnosed by gene sequencing panels or clinical exome tests, but the number of variants of unknown significance (VOUS or VUS) emerging from these tests is also growing.
To help characterize their function — and potentially reclassify them as disease-causing variants — researchers at Emory University have been developing a transcriptome sequencing test, which Emory Genetics Laboratory plans to launch clinically this summer, initially for neuromuscular disorders. Similar tests might be useful for other types of genetic diseases.
Madhuri Hegde, executive director and CSO of EGL, whose team developed the test, presented the assay during a workshop at the American College of Medical Genetics and Genomics annual meeting in Tampa last month and spoke with GenomeWeb last week.
EGL currently offers close to 200 gene panels, a medical exome test, and is soon going to launch a whole-genome test to diagnose rare genetic diseases. But according to Hegde, a large percentage of variants called by tests like these are VOUS, leaving patients with unclear results. According to Hegde, 41 percent of variants in EGL's database are unclassified, falling into the categories of missense, silent, and splice site mutations where there is no functional evidence, as well as loss-of-function changes in new genes where there is no proven disease mechanism.
One way to resolve these variants is to submit them to databases such as ClinVar, allowing laboratories who may have seen the same or similar variants in other patients to discuss them and classify them according to the ACMG's variant interpretation guidelines, "and we have done that very successfully," she said.
But in many cases, even though a lab believes a VOUS, based on bioinformatic predictions, is probably disease causing, there is not enough functional evidence to call it pathogenic. For those variants, "we should be looking at functional assays to determine an effect of the variant on the transcript and the downstream product of that gene," Hedge said. The new transcriptome test for neuromuscular disorders promises to be able to do that.
EGL has been involved in projects with the Muscular Dystrophy Association and the Jain Foundation to provide free genetic testing to individuals with muscle weakness suggesting limb-girdle muscular dystrophy (LGMD) who have not received a genetic diagnosis. The MDA project, which is funded by a grant from Genzyme and for which EGL provides all genetic testing, was announced in March of 2015.
EGL uses a 35-gene panel to test these patients for LGMD and has generated reports for about 1,800 of the 2,800 samples it has received to date. Emory continues to improve the panel, for example, by adding baits in intronic regions where potentially functional effects have been reported in the literature. So far, the test has yielded a definite diagnosis for about 720 cases, translating to a diagnostic yield of 40 percent.
In another 20 percent to 25 percent of cases, the EGL team found one pathogenic variant and one VOUS, leaving patients with a likely but not a certain diagnosis. For these patients, a functional assay might be able to reclassify the VOUS as a pathogenic variant. "The idea is, through transcriptome analysis, to prove that this does have an effect on the transcript, causing a splice site defect, an inclusion of a pseudoexon, or loss of expression — there are a variety of things that can happen," Hegde said.
Emory has already been sequencing individual cDNAs for years — usually in a research setting — to prove functional effects of a VOUS in a specific gene, but this will be the first high-throughput clinical transcriptome assay, she said.
EGL designed two RNA-seq assays, which both run on the Illumina HiSeq: a full transcriptome assay and a targeted 89-gene assay that includes all known neuromuscular disorder genes and uses Agilent's SureSelect for target enrichment. Both assays use tissue from muscle biopsies as input material, which is already available for many patients because it is part of the diagnostic algorithm for neuromuscular disorders, Hegde said.
Based on a side-by-side comparison of the two assays, they appear to perform equally well, she said, and EGL will likely start validating the targeted assay first because the amount of data is smaller, with the goal to launch it as a clinical test this summer. "As we gather more experience, we might then go to the full transcriptome," she said. The cost of the targeted assay will be in the same ballpark as that of gene panel tests.
The test has already helped to reclassify VOUS as pathogenic mutations in several patients. In the case of a boy with muscular dystrophy, for example, the Emory team was able to show that a silent variant in a muscular dystrophy gene created a new splice site that impacted the transcript.
However, the functional results cannot be used in isolation. "The data is not standalone data, you have to use it in the full picture of what you know about the patient, his or her family history, the gene, the clinical presentation, segregation data, and any additional studies done," Hegde said.
Hegde estimated that the test could resolve at least half the cases where there is one pathogenic variant and one likely pathogenic VOUS. "There will be some variants where we will not be able to prove a functional defect — it is not actually causing a splicing defect or creating a new exon, but it might be a defect which affects the folding of a protein" or has some other uncharacterized effect, she said.
Clinical transcriptome assays like this could also be useful in other disorders, and EGL is looking into developing similar assays for bone, eye, and liver disorders. While it may not be possible to obtain samples of the target tissue for many diseases — for example, intellectual disability or other disorders that affect the brain — a transcriptome assay could run on a blood sample instead, provided the gene transcript carrying the VOUS is present in blood, she said.