Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Biomedical Sciences


Cancer and Developmental Biology

Research Advisor

Kevin W. Freeman, Ph.D.


Suzanne J. Baker, Ph.D.; Andrew M. Davidoff, Ph.D.; Mark Hatley, M.D., Ph.D.; DavidR.Nelson, Ph.D.; Gerard P. Zambetti, Ph.D.


1p36, development, oncogene, pediatric cancer, stem cells, tumor suppressor


Mouse models of cancer are critical for developing therapeutic treatments for pediatric patients. Recent sequencing studies of neuroblastoma (NBL) patient tumors have uncovered precise deletions in the chromatin remodeler and tumor suppressor gene (TSG) ARID1A. Additional causal studies supported ARID1A’s candidacy as a putative 1p36 TSG in MYCN-driven NBL. This study aimed to causally test Arid1a loss during Mycn-driven NBL initiation through the development of a mouse model of high risk NBL.In this study, we modified a Mycn-driven mouse model of NBL to incorporate Cre mediated deletion of floxed Arid1a. Briefly, in freshly isolated embryonic day 9.5 primary trunk neural crest cells (NCCs), Arid1a was heterozygously and homozygously deleted in combination with Mycn overexpression. We injected these genetically modified primary cells into mice to determine if Arid1a loss collaborates with Mycn overexpression during NCC transformation. We evaluated tumor growth kinetics, characterized the pathological features and gene expression profiles of resulting tumors, and evaluated the Arid1a-dependent differentiation traits of tumors and primary NCCs in vivo and in vitro. We found that Arid1a loss caused a gene expression and phenotypic shift to an immature cell identity. Furthermore, heterozygous loss of Arid1a during transformation of primary murine NCCs resulted in tumors that pathologically and molecularly model human high-risk, MYCN amplified NBL. Gene set enrichment analysis demonstrated that gene expression profiles of Arid1a heterozygous tumor samples significantly correlated with previously established mesenchymal gene signatures found in high-risk NBL patients. Our model causally tested the clinical observation that 70% of high-risk MYCN amplified NBL patient tumors include 1p36 LOH. Our results support the hypothesis that ARID1A is a 1p36 tumor suppressor candidate that collaborates with MYCN to transform NCCs into high-risk NBL. Last, our model suggests that a shift in cell identity may be connected to NBL initiation.




2020-006-Wallace-DOA.pdf (471 kB)
Declaration of Authorship