By Julia Karow
Integrating whole-genome sequencing with epigenetic and gene expression analyses of retinoblastoma tumors, researchers at St. Jude Children's Research Hospital and the Genome Institute at Washington University School of Medicine have discovered a new potential treatment that genome sequencing alone would have missed.
"I think going forward, a lot of these cancer genome studies will really benefit from this kind of integrated, multi-platform type of analysis," said Michael Dyer, a researcher at the department of developmental neurobiology at St. Jude and the senor author of the study, which was published online in Nature last week.
Retinoblastoma is a rare pediatric cancer of the retina, and both copies of the RB1 gene are inactivated in almost all tumors of this type. The researchers set out to find additional mutations in the tumors that help transform normal cells into cancer cells. Earlier studies had suggested that the loss of RB1 leads to genomic instability, thus fostering new mutations.
For their study, which is part of the St. Jude/Washington University Pediatric Cancer Genome Project, the researchers sequenced the genomes of four retinoblastoma tumor/normal pairs and that of a mouse xenograft of one of the primary tumors. Using a paired-end sequencing approach on the Illumina Genome Analyzer, they generated about 29x average coverage and 24x average exon coverage for each sample. They also performed transcriptome sequencing of the four primary tumors to provide additional sequence coverage.
While the genome sequencing and alignment was done at WashU's Genome Institute, both the St. Jude and the WashU teams analyzed the data for mutations and compared their results, thus validating and complementing each other, Dyer said.
"We expected to find a lot of genetic lesions, either small lesions in particular genes or larger chromosomal rearrangements," he said. "But we actually found that there were very few genetic lesions, and that was a big surprise for us."
In fact, in each tumor, they found between zero and five amino-acid changing mutations, 15-fold less than what most whole-genome sequencing studies of adult tumors have found. Also, only one of the 11 genes with amino-acid changing mutations was recurrently mutated in other retinoblastoma tumors they tested.
Importantly, the number of mutations in the xenograft did not change much over nine months but remained very similar to that in the primary tumor.
In order to find out whether any genes were deregulated by epigenetic mechanisms rather than gene mutations, the researchers then performed a number of other studies, including array-based DNA methylation and chromatin immunoprecipitation, and transcriptome sequencing. Through an integrative analysis of the data, they found significant differences in 104 genes, including 15 known cancer genes.
One of those genes — the only upregulated kinase — is called SYK, and the scientists showed in further studies that it is required for tumor cell survival. Small-molecule inhibitors of SYK already exist, which were originally developed to treat autoimmune disorders but have shown promise in preclinical leukemia studies. These inhibitors killed retinoblastoma tumor cells both in vitro and in vivo, making SYK a promising new target to treat the disease.
The researchers are now working on developing a formulation of the drugs that can be delivered directly into the eye, and need to perform validation, preclinical testing, and toxicity work before they could enter into a phase I clinical trial, which could take a couple of years, Dyer said.
If the SYK kinase is also upregulated in metastatic retinoblastoma, the new treatment "could have a huge impact" on patients with metastatic disease, Dyer said, noting that these patients currently have a survival rate of only 10 percent. Many of those patients live in developing countries.
In developed countries, the main goal of a new retinoblastoma treatment would be to preserve vision and to reduce the side effects associated with current therapy, he said. Some patients with advanced disease still lose their eye today, he said, and conventional therapy can lead to secondary leukemias, neuropathies, and damage of surrounding tissue.
Besides developing SYK inhibitors to treat retinoblastoma, Dyer and his team are also following up on other results of their study. For example, they are validating genes that they found to be epigenetically deregulated and trying to figure out their role in tumorigenesis. They are also trying to explain how retinoblastomas express multiple differentiation programs that are incompatible during normal development.
A comprehensive analysis of the tumors, including not only whole-genome sequencing but also epigenetic studies, was key to uncovering SYK and the other genes, but this approach might not be needed for every cancer type. "I think that it's more warranted for tumors where, when you do the sequencing, you are not really uncovering those major cancer drivers that you might expect … so epigenetics would be the logical next step," Dyer said.
"I think it's going to be tumor-specific. For tumors where there is really not a lot of known cancer genes, and it's hard to sort out why they are so aggressive, it might be a really valuable approach."
Dyer said he believes it's "an unusual case" that a cancer genome study leads directly to a potential new drug treatment. One reason is that retinoblastoma is a simple tumor, which is rare. "It's really unusual to have a disease where a particular gene is either mutated or epigenetically deregulated in virtually all patients," he said. "Most of the childhood cancers are quite complex, there's different subgroups, and even within a clinical group, there's a lot of heterogeneity. It's going to be a lot more work to stratify patients and really customize the treatment," he said.
"But having said that, it's amazing that there are a lot of molecular targets out there that have small molecules and drugs developed for them, so I anticipate that there will certainly be some of this type of synergy" in future studies.
Also, to move genomic results forward in the clinic, a multidisciplinary approach is a real strength, he said. The retinoblastoma work, for example, required expertise in computational biology, pathology, clinical oncology, chemical biology to reformulate the drug, and will require pharmacology in the future. "I think that's a really nice model, bringing all those different disciplines together, very quickly making the most of these genomic types of discoveries, and moving them aggressively forward to clinical trials."
Have topics you'd like to see covered in Clinical Sequencing News? Contact the editor at jkarow [at] genomeweb [.] com.