Bcl-2 family members in microenvironmental support in Chronic Lymphocytic Leukemia

Introduction Chronic lymphocytic leukemia (CLL) is an incurable disease of the elderly which is increasing in numbers and shows high levels of heterogeneity, as reflected in its clinical behaviour: Some patients have a very indolent type of disease, while others show a very progressive form that may lead to death within a few years. This heterogeneity in prognosis is also reflected in the underlying biology, as specific genetic changes are only found in small minorities of patients and thus cannot explain the aggressiveness of the disease alone. By contrast, increasing evidence suggest that signals derived from the surrounding non- tumor cells (the so called microenvironment) are at least equally important in CLL progression than genetic aberrations in the tumorcells themselves. It may thus be more attractive for the treatment of CLL to find unmutated signaling events or protein families, which transmit the signals from microenvironmental interactions and are essential for CLL proliferation and survival. In this respect, the Bcl-2 family of proteins represents an optimal candidate target. Hypothesis, Methods and Content of the research project In the previous funding period we established mouse models for CLL with genetic modifications in the Bcl- 2 family, which lack proapoptotic members (Bim, Puma, Bmf) or gain antiapoptotic members (Mcl1, Bcl-2). Mice from all of these models succumb to CLL quicker than conventional CLL tumors. In the previous funding period we could show that overexpression of Mcl1 rendered CLL tumors more resistant to deprivation of microenvironmental signals in vivo and in vitro. These findings will be extended by transplantation studies of Bim, Puma, Bmf and Bcl-2 modified tumors into recipients with defined microenvironmental defects and complemented by in vitro approaches including assays to study potential therapy resistance mechanisms. Moreover, we will study the relation of Bcl-2 modification to cell cycle associated cell death. To ensure that all studies in the murine system remain relevant for the human disease, we will perform protein and RNA expression measurements of Bcl-2 family members in human CLL, combined to analysis on the genetic level and to in vitro studies mimicking different types of microenvironmental interactions. Conclusion The primary aim of this proposal is thus, to overcome the current lack of systematic and relevantly complete information on Bcl-2 family gene and protein expression in CLL and to better characterize the interplay of microenvironmental and genetic factors in CLL. The overall goal of the program is to improve our knowledge of the role of Bcl-2 family proteins within complex microenvironmental and genetic interactions in order to predict optimal treatment combinations for the elimination of CLL in clinical contexts. All aims include critical collaboration within the FOR 2036 group.

This project investigates the contribution of Bcl2 family members to pathogenesis of and resistance phenotypes in chronic lymphocytic leukemia (CLL). The Bcl-2 family protein Bmf, a so-called BH3-only protein is a cellular stress sensor and transmits death signals within the cell. BMF has demonstrated superficially defined functions within the development of immune tolerance in normal mice. Regarding human CLL, a variant of this gene locus has been associated with a predisposition to developing the disease. Using CLL mice with an engineered lack of Bmf we were able to investigate the mechanisms behind this predisposition. Before a CLL clone can develop, there needs to be a selection of a BCR that is essential for the leukemia cells. If the BCR signal is too strong (as could be that case in autoimmune BCRs) this will lead to death of the cells. Indeed, leukemia cells have relatively weak BCR signals, likely associated with the slow growth of the disease. We now show that BMF controls this signal under normal circumstances and that loss of Bmf (in analogy to the lower expression in the human variant) leads to accelerated development of leukemia, the selection of more active BCRs and stronger BCR signalling. These changes in signal transduction in the model are further associated with resistance against current, as well as developing, clinical therapeutics: Analyse on data from human CLL suggest the presence of similar phenomena and therapeutic resistances. A second BH3-only protein, Puma, is also a critical regulator of cell death in the immune system. Puma is mainly responsible for the transduction of exogenous cellular stresses, e.g. through chemotherapeutic drugs. In our genetic model we find that, contrary to the case of Bmf, loss of puma did not accelerate disease development. Furthermore, we could not observe altered selection of BCRs or BCR signals, suggesting that the primary CLL selection is independent from Puma. However, Puma deficient CLL clones show increased dependency on microenvironmental stimulation, as evidenced in co-cultures with supporting cells. In addition, we find that loss of Puma also leads to specific resistance patterns against clinically relevant drugs. First analyses of human CLL samples support these results. Our work thus shows that specific BH3-only proteins have highly divergent effects on the development of CLL and lead to predictable resistance patterns against therapeutic drugs. Next to uncovering a mechanistic explanation behind a known CLL predisposition, we thus help to make predictions for an optimised use of therapeutics.