Frazer Lab Department of Pediatrics, Genome Information Sciences


Driver Mutations of Regulatory Elements in Breast Cancer

It is known that driver mutations are largely responsible for the development of tumors, however one of the challenges of cancer genomics is identifying these driver mutations within the milieu of mutations and inherited variants that are present in a cancer’s genome. To date, only the coding sequences, which comprise just 3% of the human genome, have been investigated for driver mutations. Recently large datasets such as the TCGA have become publicly available, which now enables us to scan the 97% of the genome composed of non-coding sequences. These non-coding sequences include regulatory elements with established roles in controlling the expression of genes, including known cancer driver genes. We use an innovative computational bioinformatic pipeline to identify and annotate driver mutations in regulatory elements in a group of 200 breast cancer samples.

Investigators associated with this research project include:
Andrea L. Richardson, MD, PhD

Ovarian Tumor-Specific mRNA Isoforms for Early Detection and Vaccine Design

New Approaches for Early Detection

Ovarian cancer is the fifth-leading cause of cancer death among women in the United States. Unfortunately, ovarian cancer does not commonly present with obvious symptoms and therefore often goes undetected until it is at an advanced stage. The vast majority of ovarian cancer deaths (~70%) are of patients with advanced-stage ovarian cancer. However, the cure rates for patients diagnosed with Stage 0/1 ovarian cancer approaches 95%. Our hope is that through the creation of an early stage detection test we can detect ovarian cancer at this very early stage of pathogenesis, allowing its eradication before it becomes life threatening.

To date, there are no early detection tests are available for ovarian cancer. Common to almost all detection strategies that have been attempted is their reliance on a blood-based biomarker. Our strategy is fundamentally different. It is now known that ovarian tumor cells disseminate at a low rate to the cervix and can be collected in low numbers with a Pap test—a procedure that is part of routine gynecological exams. We intend to detect the presence of ovarian tumor cells in Pap tests by the presence of mRNA isoforms that are only expressed in tumor cells. The implementation of this strategy requires, first, the discovery of mRNA isoforms that are only expressed in ovarian tumor cells.

Over the last three years we have developed a custom RNA-seq computational pipeline that maximizes the capabilities of forefront technologies and data sources. Our pipeline is specifically designed for the discovery of tumor-specific isoforms using the RNA-seq data being produced from The Cancer Genome Atlas and the Genotype-Tissue Expression databases at the NIH—comprising more than 2,000 data sets overall. The most important feature of our approach is our focus on mRNA isoforms and not “genes”. The term “gene” is a collective term for all of the mRNA isoforms expressed from a genomic locus, and few (if any) “genes” are exclusively expressed in tumors. Thus, tumor specificity must be sought at the isoform level. To date we have identified hundreds of mRNA isoforms that are predicted by RNA-seq to be only expressed in ovarian tumors. Using software that we have developed for isoform-specific primer design, we have performed hundreds of validation RT-qPCR experiments using primary tumor and normal tissue samples. In our current work, we are evaluating validated candidates with increasingly stringent criteria to confirm their absolute tumor specificity.

New Approaches for Highly Specific Cancer Vaccines

A cancer vaccine is broadly composed of two components: an adjuvant to activate the immune system and antigens to guide the immune system’s destructive capabilities. To date our collaborators have developed powerful adjuvants that are highly effective activators of the immune system’s viral response mechanisms. The adjuvant is designed such that we can directly physically couple it to numerous “viral” peptide antigens, with the effect that the immune system will eradicate any cells on which these antigens are

Investigators associated with this research project include:
Cheryl Saenz, MD
Dennis A. Carson, MD

Genetics and Genomics of Chronic Lymphocytic Leukemia

We have multiple projects investigating the pathogenesis of chronic lymphocytic leukemia (CLL). The course of CLL is variable and its development is poorly understood: while some patients have long-term indolent disease prior to progression, others progress rapidly requiring therapy within a relatively short time after diagnosis. We are conducting a retrospective matched case-control study comparing an early-progressing group of CLL patients with a later-progressing group. We have performed exome sequencing, methylation microarrays, and targeted deep sequencing of DNA mutations to characterize the nature of CLL progression in these two groups.

We are also using transcriptome sequencing of CLL cases with and without mutations in splicing factor SF3B1 to identify aberrant mRNA splicing and determine its role in the pathogenesis of CLL. We are integrating our data with public data from the TCGA and others to extend our results to other cancers with SF3B1 mutation.

Lastly, we are investigating the transcriptome of a rare class of CLL patients with low cell surface expression of the oncofetal protein ROR1. These cases have distinct transcriptional profiles that may indicate patients lacking ROR1 expression are a distinct subclass of CLL whose disease pathogenesis and outcome may differ from the larger CLL population with typical ROR1 expression

Investigators associated with this research project include:
Thomas Kipps, MD, PhD

Page 'Breadcrumb' Navigation:

External Resources: