Michael Dyer, PhD, St. Jude Children’s Research Hospital

Integrated Epigenetic Analysis Advances Retinoblastoma Treatment

Children with retinoblastoma are among the youngest pediatric oncology patients. This tumor of the developing retina can often go undetected for months or even years before it fills the eye. If the tumor cells metastasize beyond the eye, the probability of survival plummets because the cancer aggressively invades the brain and bone marrow.

Retinoblastoma, therefore, is not surgically removed from the eye because of the high risk of tumor dissemination during the procedure. Radiation therapy is also rarely used to treat retinoblastoma because it can damage the growing bone surrounding the eye and the underlying brain regions. Without these two important tools, pediatric oncologists must rely heavily on chemotherapy to treat retinoblastoma and save vision in these very young children.

Historically, children with retinoblastoma have been treated with broad-spectrum combination chemotherapy delivered systemically over many months. However, this regimen can have debilitating side effects such as permanent hearing loss and neuropathy. Conventional chemotherapy also does not always halt retinoblastoma growth; thus, many children with advanced-stage disease must undergo surgical enucleation of the diseased eye(s).

Michael A. Dyer, PhD (Developmental Neurobiology), is leading a multidisciplinary translational research team focused on identifying more selective drugs that can be delivered directly to the eye to treat retinoblastoma. One of the major challenges his team faces is identifying cellular pathways in retinoblastoma that are deregulated and druggable. Virtually all retinoblastomas begin with biallelic inactivation of the retinoblastoma-susceptibility gene RB1. The signaling pathway controlled by the RB1 protein is deregulated in nearly all human cancers, and there are no drugs available that can restore normal regulation of this pathway.

To overcome this challenge, Dr. Dyer and his team partnered with Jinghui Zhang, PhD (Computational Biology), and colleagues leading the PCGP to sequence the retinoblastoma genome. During the first year of the PCGP, four retinoblastoma tumors were sequenced with matched normal DNA. Also included in that set was an orthotopic xenograft (i.e., human retinoblastoma grown in the eye of a mouse) that was developed in Dr. Dyer’s laboratory.

The xenograft proved to be particularly important, because it allowed the investigators to overturn conventional dogma in the field about retinoblastoma progression and to identify novel therapeutics for this debilitating childhood cancer. The results from this work, which represents the first genomic and epigenetic analyses of a childhood cancer, were published in the journal Nature.

Until the sequencing of the retinoblastoma genome through the PCGP, it was widely believed that retinoblastoma progression is rapid because RB1 inactivation causes genome instability, which in turn increases the rate of mutation and dramatically accelerates tumorigenesis. Dr. Dyer and colleagues learned from the initial analysis of the whole-genome sequencing that shockingly few mutations are present in the retinoblastoma genomes. The xenograft also had few new genomic changes or mutations, despite being continually grown for nearly a year. These data show that genomic instability is not a universal hallmark of retinoblastoma and is not required for tumor progression.

One clue from the genomic data led Dr. Dyer and his team to consider an alternative hypothesis about retinoblastoma progression. The epigenetic regulator BCOR was the only gene, other than RB1, that was recurrently mutated in retinoblastoma. Two members of the Dyer lab, Drs. Justina McEvoy and Claudia Benavente, were exploring epigenetic processes that regulate normal retinal development and retinoblastoma. They questioned whether RB1 loss leads to epigenetic deregulation of key cancer pathways that contribute to rapid disease progression. They also considered the possibility that deregulated pathways that are essential for tumor growth are druggable. To test their ideas, the researchers worked with Dr. Zhang to perform the first integrated epigenetic analysis of a childhood cancer.

As reported in Nature, the investigators found that more than a dozen known cancer genes are epigenetically deregulated in retinoblastoma, and among those is the spleen tyrosine kinase (SYK) oncogene. Through a phosphorylation cascade, SYK promotes the stabilization of the prosurvival protein myeloid cell leukemia sequence 1 (MCL1). Joseph T. Opferman, PhD (Biochemistry), who studies the role of MCL1 in cellular homeostasis, provided valuable guidance and reagents to explore the role of the SYK/MCL1 pathway in retinoblastoma. SYK was upregulated in virtually all retinoblastomas and required for tumor cell survival. The team also showed that R406, a rheumatoid arthritis drug that specifically inhibits SYK, can kill retinoblastoma cells.

Through the PCGP, Dr. Dyer and colleagues not only overturned the long-held belief in the retinoblastoma field that genomic instability is required for rapid tumor progression but also identified a promising new signaling pathway that can be targeted to treat retinoblastoma. Efforts are ongoing at St. Jude to develop retinoblastoma eye drops that contain an SYK inhibitor.