Functional Analysis Of Genomic Variants Identified Through Whole Exome Sequencing Of Pediatric Lymphocytic Leukemia Kindreds

The underlying genetic basis of many childhood cancers remains largely unknown, which places a significant focus on the discovery and understanding of cancer susceptibility genes. Even genetic predisposition of the most common childhood cancer, acute lymphocytic leukemia (ALL), remains poorly understood. Our research focuses on the identification of novel high risk cancer susceptibility genes through the use of next generation sequencing methods, bioinformatics tools, and functional assays. We are analyzing a small cohort of ethnically diverse families with at least two first-degree relatives with childhood ALL or lymphoma transmitting in an autosomal dominant manner. Whole exome sequencing was performed on affected individuals9 normal lymphoblasts and tumor samples when available. In one particular Hispanic kindred, sequencing generated a list of over 1500 heterozygous missense variants shared among three affected individuals. Filtering out common variants found in databases such as dbSNP, reduced this list by 95% and filtering variants seen in a whole exome sequenced intrafamilial control narrowed down the list to 33 potential rare heterozygous cancer susceptibility variants. In order to select which rare variants to analyze by functional methods, various bioinformatics tools were utilized. Variants leading to the change of a conserved amino acid were given a higher priority. In addition, variants had to be predicted to be damaging or deleterious to protein structure or function by 3 out of 4 algorithms. These bioinformatics analyses left us with 8 candidate variant genes (PIK3R4, SUPT3H, NT5C1A, TRPM4, PRKAG3, C6orf136, GLT25D1, and LPIN2). All of these rare variants were retained in tumors from affected family members, but there was no evidence for second hits. RNA-seq data assessing expression in hematopoietic cell populations from murine bone marrow has been a valuable tool to select genes from this variant list for further analysis. Variants that are potential inactivating mutations in genes robustly expressed in murine bone marrow, such as PIK3R4, are prioritized for knock down in murine bone marrow transplantation studies. The p.M239V variant is located in the serine/threonine kinase domain of PIK3R4. PIK3R4 is needed to regulate and activate PIK3C3, which is involved in vesicle trafficking, autophagy, and nutrient sensing. We are currently developing a novel murine model in which we harvest the bone marrow from 5-FU treated CD45.2 mice. Harvested bone marrow is then transduced with murine stem cell virus (MSCV) retroviral vectors expressing GFP and miRNAi to target the knock down of our gene of interest (GOI). The transduced bone marrow is then transplanted into CD45.1 recipient mice that have had their bone marrow ablated by radiation. These mice are then observed for signs of leukemia and their blood is collected at different time points to detect any alterations in hematopoiesis caused by the knockdown of the GOI. Alternatively, if an activating mutation is predicted of a candidate gene, we can use the MSCV retroviral system to overexpress the mutant cDNA in donor bone marrow transplanted into a recipient mouse. Within our larger cohort of ALL kindreds, none of these 8 variants are present, which demonstrates genetic heterogeneity of familial childhood leukemia. Determining the underlying genetic cause of childhood cancers will require collaborative studies across many cancer centers along with multiple model systems for functional analysis. Citation Format: Jason Saliba, Ryan Zabriskie, Bradford Powell, Stephanie Hicks, Marek Kimmel, Hannah Cheung, Deborah Ritter, Donna M. Muzny, Jeffrey G. Reid, David A. Wheeler, Richard A. Gibbs, Sharon E. Plon. Functional analysis of genomic variants identified through whole exome sequencing of pediatric lymphocytic leukemia kindreds. [abstract]. In: Proceedings of the AACR Special Conference on Pediatric Cancer at the Crossroads: Translating Discovery into Improved Outcomes; Nov 3-6, 2013; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2013;74(20 Suppl):Abstract nr A8.