Open Collections

Abstract

Acute myeloid leukemia (AML) is the most common form of acute leukemia in adults, characterized by clonal expansion of immature blood cells that leads to bone marrow failure. AML is a heterogeneous disease despite its low mutational burden, leading to a relapse rate of over 60\% within five years of diagnosis. Relapse is attributed to the persistence of leukemic stem cells which emerge from clonal molecular heterogeneity, eventually repopulating the marrow. Understanding how clonal structure evolves is essential for comprehending the factors involved in relapse. Recently, a novel method for assaying clonal structure using mitochondrial variants was reported. Mitochondrial DNA is reported to have a 10- to 100-fold increased mutation rate relative to nuclear DNA, and these mutations are inherited in a stochastic manner with each cell division. These properties make mitochondrial variants particularly useful as endogenous barcodes from which clonal relationships can be reconstructed. In this thesis, I first reproduce this method in a synthetic experiment to validate the findings and assess the sensitivity of the assay. I then develop a single-cell DNA-sequencing panel that can infer clonal structure using a combination of nuclear and mitochondrial variants. I apply this method to longitudinal bone marrow samples collected from AML patients before treatment and after relapse. Assessment of changes to clone frequency between timepoints revealed that 40\% of cases were clonally stable based on mutations in nuclear DNA. However, reassessment of the clonal structure using mitochondrial DNA reveals that half of these stable cases exhibit a subclonal change in structure that is not linked to a mutational change. This work suggests that non-genetic factors play a role in modifying clone fitness and should be examined in combination with genetic changes to fully understand how clones respond to and evade treatment, ultimately resulting in a high rate of relapse.