Date of Award

Spring 5-15-2021

Author's School

Graduate School of Arts and Sciences

Author's Department

Biology & Biomedical Sciences (Immunology)

Degree Name

Doctor of Philosophy (PhD)

Degree Type



Myelodysplastic syndromes (MDS) are a heterogeneous group of clonal bone marrow disorders characterized by ineffective hematopoiesis. Approximately 30% of MDS patients progress to secondary acute myeloid leukemia (AML). MDS is caused by somatic mutations in hematopoietic stem/progenitor cells and progression to secondary AML is associated with the acquisition and/or expansion of at least one subclone. We hypothesized that specific gene mutations would be enriched in subclones compared to founding clones, and that the order of mutation acquisition would be critical for clonal evolution and progression from MDS to secondary AML. Sequencing of paired MDS and secondary AML samples from 44 patients revealed that signaling and transcription factor gene mutations were more frequently identified at secondary AML than MDS. While secondary AML signaling gene mutations were often absent in the MDS sample, transcription factor gene mutations were typically present at the time of MDS diagnosis, suggesting that transcription factor gene mutations are acquired prior to mutations in signaling genes. To directly address this possibility, we deciphered the clonal architecture of bone marrow samples from a subset of these patients using whole genome and single-cell DNA sequencing. Using this approach, transcription factor and signaling gene mutations were identified in subclones. Moreover, we found that when these mutations co-occur in the same subclone, the transcription factor mutation is almost always acquired before the signaling gene mutation based on the clonal architecture. Finally, we observed that subclones carrying these mutations expanded during progression to secondary AML, further implicating them as potential drivers of disease progression. Multiple co-occurring signaling gene mutations are often observed in secondary AML samples. To address if multiple signaling gene mutations occur in the same clone, we analyzed whole genome and single cell sequencing data. We observed that signaling gene mutations rarely co-occur in the same cell and typically exist in parallel subclones. Given the presence of multiple signaling gene mutations in distinct subclones at secondary AML, and the association of signaling gene mutations with progression, we then asked if previously undetected signaling gene mutations were present at MDS below the level of detection for standard sequencing. Using error-corrected sequencing, we found that signaling gene mutations, and in particular mutations in RAS family members, are present in up to 46% of MDS samples, three times more common than previously identified using standard sequencing. However, only one-third of these MDS signaling gene mutations expand and become detectable using standard sequencing at secondary AML. The diverse patterns of signaling gene mutation expansion and collapse, often involving multiple mutations within a single patient, suggest that MDS cells are under strong selective pressure to acquire a signaling gene mutation, but mutations must occur in the correct context to contribute to disease progression. Finally, we observed that detection of a signaling gene mutation at MDS, regardless of its persistence at secondary AML, may be associated with an increased risk of progressing to secondary AML. Future studies are needed to determine the potential clinical utility and prognostic importance of detecting these low-level signaling gene mutations at MDS. To further investigate the impact of subclonal gene mutations on progression from MDS to secondary AML, we developed preclinical mouse models expressing transcription factor (e.g., Runx1) and signaling gene (e.g., Nras) mutations, along with a founding clone mutation (e.g., U2AF1). Mice bearing mutant U2AF1(S34F) and NRAS(G12D) had altered hematopoiesis in the peripheral blood, bone marrow, and spleen compared to wild-type and single-mutant mice. However, despite altered hematopoiesis, mice expressing U2AF1(S34F) and NRAS(G12D) mice did not develop overt leukemia and showed similar survival compared to NRAS(G12D) alone. Therefore, we created mice expressing a founding clone mutation (e.g., U2AF1) and two subclone mutations (e.g., Runx1, Nras). These mice showed further perturbations in hematopoiesis and are currently being monitored for development of MDS or AML. Together, the data illustrate that mutations that frequently co-occur in patients cooperate to alter hematopoiesis in pre-clinical models. These models will be valuable to study the mechanisms of mutation cooperation, the importance of the order of mutation acquisition, and to test novel therapies with the goal of improving patient outcomes.


English (en)

Chair and Committee

Matthew J. Walter

Committee Members

Jeffrey Bednarski, Timothy Ley, Daniel Link, Eugene Oltz,