A multi-center team of researchers has sequenced exons from a large receptor family to identify potentially druggable melanoma-associated mutations that activate a signaling pathway already implicated as a potential therapeutic target for melanoma and other cancers.
Rather than taking the whole-genome or whole-exome approach used in some other cancer-sequencing studies, the researchers from the National Human Genome Research Institute and elsewhere relied on targeted exon capture to specifically sequence the exons of more than 700 G protein-receptor-coupled genes.
These GPCRs, the largest gene family in humans, are prime suspects in melanoma and other cancers due to their central role in regulating many cellular signaling pathways. Consequently, they are also common targets for drug development.
"The importance these molecules play in human diseases is evident by the fact that [50 to 60 percent] of the US Food and Drug Administration-approved therapeutics target GPCRs," the study authors wrote in their findings, which appear in the Sept. 25 online edition of Nature Genetics. "As GPCRs regulate pathways that affect cell growth, their genetic analysis in cancer is warranted."
Such considerations, combined with past studies hinting that members of the GPCR family might contribute to melanoma tumorigenesis, prompted the team to examine the family in more detail, said one of the study's researchers.
Co-first author Todd Prickett, a research fellow in senior author Yardena Samuels' molecular cancer genetics lab at NHGRI, said that when his team started the study methods for capturing and sequencing specific exons were just starting to become more common. This made it an appealing alternative to sequencing a small number of candidate genes or taking a more costly approach such as whole-genome or whole-exome sequencing.
"At the time, the exon capture sequencing method was sort of an up-and-coming way to sequence large amounts of genetic material, especially large genes, but also to look at a large gene family," Prickett told Clinical Sequencing News. "And we know that GPCRs are the largest gene family in the human genome."
Other researchers said they believe the results of Prickett's study highlight the validity of performing broad genetic surveys that focus on a specific set of candidate genes.
One is Washington University pediatrics and genetics researcher Todd Druley, who was not involved in the new study but who has performed pooled, targeted DNA sequencing on dozens of candidate genes to look for acute lymphoblastic leukemia-related mutations in 450 children with the disease (In Sequence 2/1/2011).
Druley said that Prickett and colleagues "asked a very specific question here: '[Is] an aggregation of different genetic variants in these G protein-coupled receptors associated with melanoma?' And they were able to demonstrate that."
Had they chosen "a broader approach — looking at whole-exome or whole-genome sequencing — they potentially would have found some other information, but it would have been far afield of what their original hypothesis would have been," he added.
Columbia University melanoma researcher Julide Celebi, who also was not involved in the study, agreed. Prickett and his colleagues "were very focused in their approach," she told CSN. "That gave them their success."
A New Driver
As part of the Nature Genetics study, Prickett's team sequenced GPCR genes in 11 individuals with melanoma and matched control samples. They uncovered recurrent mutations affecting the metabotropic glutamate receptor-coding gene GRM3.
The GRM3 mutations also turned up in more than 16 percent of 80 melanoma tumor samples the researchers screened in follow-up experiments, while subsequent analyses suggest the mutations lead to MEK phosphorylation and signaling — a process activated by other recurrent melanoma-associated mutations, including the BRAF mutation V600E.
The BRAF V600E mutation, which swaps out valine for glutamic acid at position 600 in the resulting BRAF protein, has been linked to melanoma, papillary thyroid cancers, and hairy-cell leukemia (CSN 6/15/2011).
BRAF mutation status is increasingly being considered to determine melanoma treatment. For instance, the US Food and Drug Administration recently approved the Roche/Plexxikon drug Zelboraf and its companion BRAF diagnostic because studies have suggested that patients with melanomas that carry BRAF V600E mutations are more apt to respond to the drug (PGx Reporter 8/17/2011).
Notably, Prickett said, many melanoma samples containing GRM3 mutations were also known to carry BRAF mutations. This suggests that there might be a benefit to using a combination of therapies or treatments that target hubs such as MEK/MAP kinase in which pathways downstream of BRAF and GRM3 merge.
To Prickett, the MEK/MAP kinase pathway is a possible "node of interest in terms of clinical relevance."
"The majority of our melanoma samples either have a BRAF mutation and/or NRAS mutations, which are both known to activate the MAP kinase pathway," he explained.
According to Prickett, the GRM3 mutation data described in his team's Nature Genetics paper suggests that there may be other targetable driver mutations in melanoma that activate overlapping pathways.
"We need to target not only upstream [of MEK/MAP kinase], but also at the major nodes where they all come in together," he said.
Prickett and his co-authors also claimed in their study that the new findings may help explain why treatments targeting components of the MEK-signaling pathway are more effective in some melanoma cases involving known BRAF mutations than others.
"The prior failure of MEK inhibitors to obtain significant tumor responses in many BRAF p.Val600Glu melanomas may have resulted in part from the absence of additional mutations that activate the MEK pathway, such as those in GRM3," the team wrote in their paper.
[ pagebreak ]
A Focused Approach
The researchers were able to home in on the MEK-activating GRM3 mutations by using thousands of molecular-inversion probes to capture 2,400 exons representing 734 GPCR genes in tumor and matched normal samples from 11 melanoma patients. Then they sequenced the GPCR exons with the Illumina GAII.
From the 755 possible non-synonymous mutations they detected, the team then used computational approaches, including a bioinformatics program called Sorting Intolerant From Tolerant, or SIFT, to narrow in on somatic mutations that affected the same gene in at least two melanoma samples, and that were predicted to have negative consequences for protein function.
They then sequenced the coding regions of 11 candidate genes in 80 additional tumor samples before screening for mutations in one of the most commonly affected GPCR genes — GRM3 — in dozens more samples.
In this panel of 57 melanoma samples, the team identified non-synonymous GRM3 mutations in nine samples, or 15.7 percent of those tested.
Many of these mutations clustered in a mutation hotspot in the gene, Prickett noted. Four individuals had the same amino acid change at position 870 of GRM3, for instance, which trades the amino acid lysine in the place of glutamic acid.
And in a series of experiments in a melanoma cell line and in mice, the researchers found that all of the melanoma-associated GRM3 mutations examined appeared to phosphorylate MEK, which activated MEK signaling and led to downstream changes in cell growth and migration.
"Taken together, our data highlight a model for melanoma pathogenesis in which activation of MEK by GRM3 alterations promotes the proliferation and migration of melanoma cells," the team concluded.
"Although further investigation into the mechanism of GRM3 activation of the MEK pathway is required, our study suggests that the presence of GRM3 mutations is expected to indicate sub-populations of individuals whose tumors are dependent on MEK signaling," they added.
Reversing this MEK activation — for instance using short hairpin RNAs targeting GRM3 transcripts — curbed this excess growth and movement, they reported. So did treatment with Array Biopharma/AstraZeneca's selective MEK inhibitor AZD-6244, which is being tested in phase 2 clinical trials as a treatment for malignant melanoma and other cancers.
"By limiting themselves to the G protein-coupled receptors, they increase the likelihood that they would find variation within a gene or a few genes where there might already be existing agents that they could use in potential therapeutic strategies," Washington University's Druley said. "And, in fact, they were able to show that."
Given its findings so far, Prickett's team argues that melanomas harboring genetic glitches in GRM3 and/or other glutamate receptors that kick-start MEK signaling might be vulnerable to MEK inhibitors, or drugs that target pathways downstream of MEK.
The results also hint that mutations in other glutamate-receptor genes may activate melanoma-related pathways, Prickett explained. He noted, for example, that the metabotropic glutamate receptor coding genes GRM1 and GRM5 have previously been implicated in melanoma development or etiology.
"It would be, I think, important if we could potentially show that all metabotropic glutamate receptors are important [in melanoma]," Prickett noted.
He said the team has no immediate plans to follow up on any of the other mutated genes they identified in their Nature Genetics study. Instead, they are trying to learn more about the signaling pathways, and the interactions between them, that contribute to melanoma.
"Instead of looking at it as one-dimensional signaling pathway, we're trying to find many different pathways that might funnel into one or two major nodes that we can then clinically target," said Prickett.
Columbia University's Celebi praised the study for adding a new player to the repertoire of genes involved in melanoma, but noted that follow-up studies will be needed to understand whether GRM3 status can be used to help classify melanoma, and to look in more detail at the relationship between GRM3 and other key melanoma genes.
"It requires some more studies to see the relationship of GRM3 with the most commonly mutated melanoma genes like NRAS and BRAF and KIT," she said.
Based on his experience with pediatric cancer patients, Druley said there may also be interesting patterns in the germline variation data generated in the Nature Genetics study.
Though Prickett's team focused on somatic mutations in the paper, there might be some benefit to comparing germline GPCR patterns in melanoma patients and unaffected controls to explore the possibility that some individuals are born with GPCR mutations that predispose them to cancer, perhaps through an interaction with somatic mutations, Druley added.
"If these sequence variants and mutations are present at the time of birth, what's happening throughout these people's lives? Are they acquiring somatic mutations that act in synergy with these germline mutations, or are the germline mutations sufficient to cause the melanoma?" he said. "I certainly don't think they are irrelevant."
Despite his team's success with the candidate pathway-sequencing approach, Prickett noted that he and his colleagues are starting to rely more heavily on whole-genome and whole-exome sequencing methods.
For example, a group published a study in Nature Genetics this spring describing how it identified recurrent glutamate signaling gene mutations via whole-exome sequencing of 14 metastatic melanoma and matched samples (CSN, 4/9/2011).
According to Prickett, such approaches were extremely expensive when the his team started exploring the pathways involved in melanoma, but are now comparable or less expensive than the candidate pathway method he and his collagues used in their Nature Genetics paper.
He did not provide a firm figure for the current cost for performing the type of targeted GPCR-sequencing described in the paper because the price of sequencing has gone down since the project began. However, he estimated that at the time the study was being performed the cost was between $30,000 and $40,000 per exon set.
Have topics you'd like to see covered in Clinical Sequencing News? Contact the editor at anderson [at] genomeweb [.] com.