Date of Award

Spring 5-15-2019

Author's School

Graduate School of Arts and Sciences

Author's Department

Biology & Biomedical Sciences (Human & Statistical Genetics)

Degree Name

Doctor of Philosophy (PhD)

Degree Type



Cancer is a heterogeneous disease with myriad causes and outcomes. Many of the cancers that occur in adult populations have become increasingly well characterized with the advent of affordable high-throughput sequencing. These studies have revealed that cancer is largely a disease of somatic mutation in the adult population. In strong contrast to this, childhood cancers have an exceedingly low rate of somatic mutation. At the extreme end of this spectrum is Infant Leukemia (IL). Sequencing of IL has revealed that these tumors frequently have one or fewer somatic SNP. In the absence of a somatic explanation for IL, many other possible explanations have been put forth. To date, however, none of these has been able to fully explain the incidence of this disease. In this context, we hypothesized that inherited germline variation, rather than somatically acquired mutations, played a role in the development of IL. We showed that IL patients have an excess of rare, non-synonymous inherited variation in known-leukemia associated genes. We further showed that there are several genes that harbor far more putatively damaging variation in IL patients than either control exomes or population databases. These highly variant genes are intolerant of loss-of-function changes, and most fall into one of three critical cellular functions. Together, these data suggest that IL is indeed a result of predisposing genetic variation.

Obtaining a clear functional understanding of IL has been hindered by the lack of an appropriate model. The fact that this disease arises in utero makes it difficult to study in vivo, and no animal models have been able to recapitulate the rapid onset of disease. In recent years, several groups have developed in vitro models of hematopoiesis. While these are not yet able to fully capture all aspects of hematopoietic development, they do provide a system in which we can explore the effects of the genetic variation observed in IL patients in a controlled and developmentally relevant setting. Importantly, we are able to effectively separate the primitive and definitive hematopoietic programs and explore each independently, a necessary feature for any IL model. In this work, we present the first steps in the development of a model of IL that is consistent with our sequencing findings. While we do not achieve leukemic transformation, we do show that cells deficient in MLL3, a gene that was frequently variant in our IL cohorts, have a marked impairment in both primitive and definitive hematopoiesis. We find that this is evident both based on surface markers and colony forming ability. In addition to these functional characteristics, we show that the transcriptional and epigenetic profiles of the MLL3-knockout cells are greatly perturbed, consistent with the role of MLL3 as a transcriptional enhancer and epigenetic regulator. These results provide insight into the etiology of IL as a disease of aberrant development, and provide a basis for the establishment of an in vitro model of IL.


English (en)

Chair and Committee

Todd E. Druley

Committee Members

Grant Challen, Mary Dinauer, Jeffrey Magee, Christopher M. Sturgeon,


Permanent URL: https://doi.org/10.7936/4r6p-7919