Posted: February 21, 2013
CASE STUDY: Reinvigorating GBM Research with Hope for Future Therapies
Emma Jane Spaulding
“Ultimately, our job is to try to save some lives," says Antonio Iavarone, M.D., speaking from the perspective of a clinician and a researcher, about the deadly form of brain cancer, glioblastoma (GBM). An effective long-term treatment has not yet been found, and Dr. Iavarone notes, “Things have not changed in the last few years.” However, Dr. Iavarone and his research group at the Institute for Cancer Genetics at Columbia University are on the forefront of brain cancer research, having recently discovered the first recurrent gene fusion in GBM. Their novel findings were published in the journal Science in July 2012.
Prior to beginning his career in brain cancer research, Dr. Iavarone grew up and attended medical school in Italy. Following graduation, he joined the molecular neuro-oncology lab of Peter Preuss, Ph.D., at the University of California at San Francisco’s Brain Tumor Research Center as a research fellow before planning to return to Italy to accept an assistant professorship at the Catholic University of Rome. However, once he began research, his plans changed. “I actually decided to continue in research because I saw that there is so much to be done,” he explains. “The type of opportunities and new projects in cancer research were very exciting.”
The Landscape of GBM Research
Dr. Iavarone remembers that for a time there was little interest in studying GBM. The cancer was perceived uninteresting with few research opportunities. Describing the state of GBM research in the past, he says, “There were no recurrent chromosomal translocations, meaning present in more than one tumor. Sometimes, a description of sporadic alterations in just one tumor was published, but the significance of these papers wasn’t strong because there was no indication that the alterations were present in multiple patients.” In spite of these challenges, Dr. Iavarone began his research in pediatric and adult GBM.
Most recently his work has focused on gene fusions, a term for a hybrid gene that occurs when there is a break in one gene and rejoining in a different gene, sometimes due to problems in cell division. If the genes join in just the right way, the fusion gene may express a fusion protein. The fusion protein may maintain part of the function of each of the individual genes or express an entirely new function. Sometimes, the fusion protein’s new function is oncogenic, or cancer-causing. The most well-known fusion gene is called the BCR-ABL gene fusion. This gene fusion occurs across two different chromosomes in the BCR and Abl1 genes and results in a fusion protein thought to cause a type of blood cancer, chronic myelogenous leukemia. The fusion protein accelerates cell division and inhibits RNA repair, which can cause genomic instability. However, in a great biological success story, a drug called imantinib or known by its brand name Gleevec, was developed to inhibit the function of the fusion protein, slowing or stopping proliferation of cancer cells.
Dr. Iavarone points out that Gleevec “is probably the most successful example of targeted therapy in cancer that has been seen so far.” He continues, explaining that though gene fusions were found to be causative in blood cancers such as leukemia and lymphoma, they weren’t considered to have potential as drug targets in GBM research. However, supported by the success story of Gleevec and hope for future clinical therapies, Dr. Iavarone decided to examine gene fusions in brain cancer.
From Hope of Success to Research Plan
Starting with just nine GBM samples, Dr. Iavarone and his group looked for gene fusions that might serve as potential drug targets. If a gene fusion occurs in GBM in the same way it does in Gleevec-treatable leukemia, a fusion protein will be expressed. To find these potential fusion proteins, Dr. Iavarone used an algorithm called TX-Fuse developed by his colleague and collaborator, Raul Rabadan, Ph.D., also at Columbia University. TX-Fuse finds splits in the RNA sequence, which suggest the presence of gene fusions. Using this tool, five gene fusions were located and confirmed. Dr. Iavarone and his group decided to focus on the gene fusion most strongly supported by their sequence data – between FGFR, a growth factor gene, and TACC, a gene essential for cell division. To further elucidate this fusion, the scientists would need to include more biospecimens in their analysis.
However, as GBM is a rare brain tumor, Dr. Iavarone and his group did not have access to additional samples at their university. From previous research experience, Dr. Iavarone knew that The Cancer Genome Atlas (TCGA) had comprehensively characterized the genomes of rare brain tumor samples, among them GBM. “So, the very first thing we did after we saw the gene fusion was go to TCGA.” He exclaims, “We simply went to the TCGA expression data and there were over 500 samples!” Among these, they found four samples with markedly higher expression levels of FGFR and TACC, a suggestion that these genes may have become improperly regulated and that the researchers were on the right track. Because TCGA has generated data on many available platforms from a single tissue sample, the scientists were able to find other signs as well.
When a gene breaks and comes back together at a different spot on the chromosome, a very small part of the chromosome can be multiplied (a microamplification) or deleted (a microdeletion). These situations as a class are called microalterations, and they often go unnoticed. “Without knowing these samples could be candidates for fusion, we never would have seen those microamplifications or microdeletions because they are so small,” says Dr. Iavarone. He further describes the situation, saying that without a potential location to search, they would not have been able to find the microalterations.
Considering previous research, Dr. Iavarone reflects, “That’s why these fusions have not been found in the past. The alterations are too small and the genes are too close on the chromosome, so you really cannot find them using the traditional analysis.” Even though these changes are small, TCGA had the data. Dr. Rabadan and his bioinformatics team modified the TX-Fuse algorithm to suit DNA exome sequences, naming it Exome-Fuse. The algorithm successfully identified the microalterations at the FGFR and TACC genes.
Discovering the Meaning of the Recurrent Alteration
The data from both TCGA and Dr. Iavarone’s tissue samples suggested that the FGFR and TACC fusion gene recurrently occurred in GBM. However, Dr. Iavarone understands that correlation does not mean causation, and frames the next challenge in his research succinctly. “Simply knowing that [FGFR/TACC fusion proteins] exist, unfortunately, doesn’t mean that they are important and that they can be targeted therapeutically.” Dr. Iavarone wanted to understand how the fusion protein could be producing these oncogenic effects.
Through experiments with cell samples developed to express the fusion protein, mouse models, and other research platforms, which can be read about in this Research Brief, Dr. Iavarone and his team found that the FGFR-TACC fusion protein was disturbing chromosome segregation during cell division. As a result, cells were ending up with too many or too few chromosomes, leading to chromosomal instability, which has been suggested to contribute to oncogenesis. Dr. Iavarone says, “This is really unprecedented… This gene fusion is something that no other oncogene in glioblastoma before has been shown to do - to oncogenically transform brain cells into a glioblastoma.”
Looking forward, Dr. Iavarone speculates that one day all cancer patients with GBM will be assigned to different therapeutic plans based on the specific alterations in their genomes. “What we see more and more is that there is not a therapy that is good for all, but for a small fraction of patients with the specific driver alterations.” Dr. Iavarone sees TCGA as a starting point for developing clinical trials. “The most exciting experiments will be those that you do based on what you find in TCGA data.”