We successfully generated mouse models for different bone marrow failure syndromes and for syndromes predisposing to leukemia. In addition we were able to analyze primary patient-derived samples.
The most important findings were:
1) For the telomeropathy dyskeratosis congenita (DC), we generated and optimized a mouse model that closely resembled human blood disease. Approx. 45% of mice succumbed due to hematopoietic failure. In line with our hypothesis, we were able to prevent death of half of the animals by inhibiting p53-induced apoptosis by genetic deletion of its pro-apopotic target PUMA (21% vs. 45%). Surviving mice had much better blood counts when Puma was deleted, and telomere attrition was delayed. We thus were able to break the vicious circle and prevent the development of bone marrow failure. Most importantly, our treatment maintained genomic stability and did not increase risk of malignant transformation. In parallel to the mouse experiments performed, we analyzed trephine biopsies from DC patients and found increased levels of p53 and PUMA expression indicating a conserved role of the p53/PUMA axis in murine and human telomeropathies.
2) For the DNA damage repair Fanconi anemia, we decided to work with primary patient material rather than mouse models. Patient samples from different disease stages (bone marrow failure, secondary MDS or AML) could be included and were subjected to RNAseq and WES. We identified few mutations in oncogenes and tumor suprressor genes but significant transcriptomic changes, both in non-malignant and malignant disease stages. While inflammatory signals were strongly increased in the pre-malignant stage (bone marrow failure), these signals were reduced during malignant transformation. The underlying mechanisms still need to be investigated.
3) During the last years, GATA2 deficiency gained strong relevance as predisposition syndrome for MDS and AML. In our ApoptoMDS project, we hypothesized that apoptosis deregulation represents a major pathogenetic driver in GATA2 deficiency. It was thus our aim to manipulate apoptosis signaling (similar as in dyskeratosis congenita). However, we realized that all available mouse models were not appropriate to study these research questions. We therefore first established a mouse model, in which mice first developed bone marrow failure (in 40% of all mice) and then leukemia (in 30% of mice which earlier had developed bone marrow failure). None of the mice that did not develop bone marrow failure showed signs of leukemia which indicates that leukemogenesis is a secondary event after bone marrow failure. We analyzed mechanisms contributing to bone marrow failure and identified increased apoptotic susceptibility as well as reduced proliferation potential of Gata2-haploinsufficient stem and progenitor cells. To understand why some but not all mice succumbed to bone marrow failure, we performed transcriptomic analyses and found two very different transcriptomic programs. Whether this is the cause of the variable outcomes needs to be tested. In addition, we performed WES to identify which mutations contribute to leukemia development. We found especially genes in the NOTCH signaling pathways and DNA damage repair genes to be mutated in Gata2+/- leukemias.
