CCBIO seminar: Jan Jacob Schuringa
Towards identification and targeting of leukemic stem cells and (epi)genetically distinct subclones using humanized niche xenograft mouse models
Jan Jacob Schuringa
Department of Experimental Hematology, Cancer Research Centre Groningen, University Medical Centre Groningen, The Netherlands
Over the past years it has become clear that multiple genetically distinct subclones can co-exist in patients that suffer from acute myeloid leukemia (AML). These subclones often carry similar founder mutations, but upon leukemic evolution different secondary mutations arise in distinct subclones. Tools to prospectively isolate viable subclones to functionally study them are currently lacking. We performed transcriptome and quantitative proteome analysis on large cohorts of primary AML CD34+ samples and healthy CD34+ controls. Integration of these datasets resulted in a list of 60 plasma membrane (PM) proteins that were upregulated in specific subtypes of AML. Many of these were validated in independent cohorts of patients, and expression of specific markers was correlated to mutation status, disease progression and minimal residual disease (MRD).
By using the flow cytometry software Infinicyt™, expression of multiple plasma membrane markers was linked to each other and this combinatorial approach was then used to perform principle component (PC) analyses in order to determine which (combination of) PM markers allows us to identify distinct subpopulations based on PM expression profiles. We have now performed these PC analyses on a first set of 20 AML patients, and while in 30-40% of investigated cases leukemias appear to be rather monoclonal, we have observed remarkable heterogeneity in clonal distributions in the remaining cases. Prospective sorting and subsequent targeted sequencing of these populations indeed confirmed the presence of different mutations in these subclones. Transcriptome studies revealed that gene expression profiles were also clearly distinct between subclones. We have begun to functionally analyse these subclones in vitro and in vivo. In order to further improve in vivo xenograft models we have implanted human mesenchymal stromal cells (MSCs) with a mixture of matrigel and scaffolds in order to create a human bone marrow-like environment in the mouse. A large cohort of patient samples was successfully engrafted in this model, covering all important genetic and risk subgroups. Furthermore, stem cell self-renewal properties were better maintained as determined by serial transplantation assays and genome-wide transcriptome studies, and less clonal drift was observed as determined by exome sequencing. The human leukemia xenograft mouse clinic that we have established will serve as an excellent resource for future studies aimed at exploring novel therapeutic approaches. For instance, we have recently utilized this model to investigate the role of polycomb group proteins in leukemia, and identified that the non-canonical polycomb repressive complex PRC1.1 is critically important for leukemia development. We have also utilized this model to follow the kinetics of prospectively isolated (epi)genetically distinct subclones from individual patients.
Chairperson: Bjørn Tore Gjertsen, CCBIO