By Monica Heger
By sequencing the exomes of 14 metastatic melanoma tumors and matched normal samples, researchers from the National Human Genome Research Institute have identified novel recurrent mutations that implicate the glutamate signaling pathway, which, if validated, could be targeted by therapeutics.
Prior to the study, published last week in Nature Genetics, only one whole genome of a melanoma tumor had been completed, so the study provides important insights into the driving mutations of the cancer and the pathways involved.
Some of these findings could have therapeutic implications. For instance, the glutamate signaling pathway was found to be important. The pathway had previously been implicated in melanoma, but the data was inconclusive. Additionally, there is already some evidence from other types of cancers that glutamate inhibitors can limit tumor growth.
"There are a lot of potentially clinically actionable things generated by the data," said David Smith, a professor of laboratory medicine and pathology at the Mayo Clinic who focuses on sequencing tumors from head and neck cancer patients, and who was not involved in the study. The glutamate signaling pathway "is an important pathway."
The researchers used Agilent's SureSelect Human All-Exon enrichment kit and the Illumina Genome Analyzer for sequencing. They sequenced metastatic tumors from 14 individuals who had not been treated, plus matched normal DNA. They sequenced to an average depth of 180-fold, generating around 12 gigabases of sequence per sample.
Applying a strict filtering system in order to look for novel recurrent mutations, they identified 68 genes that appeared to be mutated at significantly elevated frequencies. Of those, 16 were classified as driving mutations, including genes known to be involved in melanoma, such as the BRAF gene, as well as several novel ones. They also screened an additional 153 metastatic melanoma samples for mutations in seven genes that had occurred in at least two of the original samples.
Of potential clinical importance, the researchers found that the glutamate signaling pathway was significantly mutated, with variants appearing in seven different genes in the pathway.
"There are some [glutamate] inhibitors that have been developed previously that have been shown to limit [tumor] growth," said Yardena Samuels, head of molecular cancer genetics at the NHGRI, who led the study. However, "we'd need to investigate much more before jumping to clinical conclusions."
Now that relevant pathways and genes have been identified, "follow-up studies will focus on the prognostic values, functional relevance, and the development of preclinical models to see if we can target any of these," she added.
Smith added that the team would need to study a much larger number of patients. "Then, for those that have this pathway altered, determine what drugs we could give them and how effective those drugs could be at targeting" the pathway.
Aside from the glutamate signaling pathway, the team also identified several recurrent mutated genes.
Notably, they found a mutation in the TRRAP gene in two patients, plus an additional four tumors in the prevalence screen. All six patients harbored the exact same mutation. The finding suggests that the TRRAP gene may be a cancer driver and also that it may function as an oncogene.
The team also found mutations in a previously unidentified gene, GRIN2A, in six of the sequenced patients, plus an additional 11 in the prevalence screen — the most frequently mutated gene in the study other than BRAF. The researchers also searched two additional data sets comprising 71 individuals with melanoma, identifying 16 additional cases with a mutated version of GRIN2A, for an overall prevalence of 25 percent.
The GRIN2A gene falls within the glutamate signaling pathway, encoding a glutamate receptor subunit. Within the gene, the researchers found two mutation clusters, including three recurrent alterations. The mutations were found within evolutionarily highly conserved residues, and bioinformatics analysis predicted that more than half would affect protein function. Additionally, the team identified five nonsense mutations, which would also likely result in a non-functioning protein.
As part of the analysis, the researchers developed a statistical tool called Most Probable Genotype, or MPG, which uses sequencing error rates and the expected heterozygosity of a given sequence to calculate the reliability of genotype calls.
Melanoma has an extremely high mutation rate, making it difficult to distinguish between driver mutations and passenger mutations, Samuels said. Exome sequencing allowed the team to focus its analysis on the mutations that would likely be driving mutations, and MPG allowed them to further reduce the number of variants by 18 percent — further increasing their confidence in deciding which variants were driver mutations.
Samuels said that the next step is to look at additional samples to validate the findings. She said the team is also doing whole-genome sequencing of around 10 metastatic melanoma tumors, as well as exome sequencing of some primary melanoma tumors.
For the current study, she said they chose to sequence metastatic tumors as opposed to primary because they want to look at the "endpoint — the compilation of all the mutations" — and then work backwards, deriving information about which mutations occurred early and which occurred later.
Have topics you'd like to see covered by Clinical Sequencing News? Contact the editor at mheger [at] genomeweb.com.