CORDIS - EU research results

Systematic in-vivo analysis of chromatin-associated targets in leukemia

Final Report Summary - CHROMATINTARGETS (Systematic in-vivo analysis of chromatin-associated targets in leukemia)

Chromatin regulators maintaining aberrant self-renewal and cell identity programs have emerged as promising candidate therapeutic targets for the treatment of leukemia and other cancers. The overall aim of “ChromatinTargets” was to systematically identify and functionally characterize chromatin-associated dependencies various types of leukemia. To this end, we have developed advanced shRNAmir- and CRISPR screening methodology, which we applied to establish a comprehensive map of chromatin dependencies in diverse leukemia subtypes. Comparative screens in a panel of genetically engineered mouse models of AML, B-ALL, and T-ALL revealed that some of the most prominent selective chromatin dependencies are determined by the tissue and lineage context rather than by specific driver mutations. Surprisingly, lineage-restricted dependencies included several chromatin regulators that are also implicated as tumor suppressors, indicating that they have opposing functions in different lineages and leukemia subtypes (Roth et al, in prep). Among these, we identified and functionally characterized the histone-methyltransferase KMT2D as a strong and selective dependency in myeloid leukemias, while loss of KMT2D accelerates B-cell malignancies, in line with its known role as a tumor suppressor in this context. Additional context-specific dependencies we have identified and functionally investigated include SETD2 (Skucha et al, Nature Comm, 2018) and components of the chromatin complexes CAF1 (Cheloufi & Elling et al, Nature, 2015), SAGA and cohesin. In CRISPR synthetic lethality screens using a second-generation genome-wide sgRNA library (Michlits & Jude et al, Nature Methods, in press), we identified the cohesin subunit STAG1 as the most prominent and only selective dependency in leukemias harboring mutations in STAG2, which are commonly found in AML and other cancers (van der Lelij, LAS, in press). This hard-wired synthetic-lethal interaction and other identified chromatin-associated vulnerabilities may provide new entry points for the development of personalized leukemia therapies.

One of the most prominent tissue-specific dependencies we repeatedly identified in myeloid and B-cell malignancies is the histone acetyl reader BRD4 (Zuber et al, Nature 2011). While several BRD4/BET bromodomain inhibitors (BETi) have entered clinical evaluation, mechanisms underlying sensitivity and resistance to BETi treatment had remained elusive. Using an integrative approach combining functional-genetic screens, dynamic transcriptional and STARRseq-based enhancer profiling, we discovered that primary and acquired BETi resistance arises from compensatory mechanisms that rapidly restore the transcription of BETi target genes such as MYC (Rathert & Roth et al, Nature, 2015). Pathways and regulatory elements involved in this adaptation mechanism are conserved between different leukemia subtypes, suggesting that they can be exploited for combinatorial therapies. A major challenge for probing molecular and cellular functions of BRD4 and other chromatin regulators turned out to be the lack of scalable experimental strategies to define their direct transcriptional targets. To tackle this problem, we co-developed SLAMseq (Herzog et al, Nature Methods, 2017) as a simple and highly scalable method for quantifying direct transcriptional responses to drugs and other cell perturbations. In pioneering studies, we combined SLAMseq with emerging degron technologies to define primary regulatory functions of BRD4 and MYC, which had remained subject to a long-lasting debate. We found that BRD4 acts as broadly relevant co-factor of Pol2 pause release, while MYC is a selective transcriptional activator of a highly conserved set of target genes predominantly involved in basic biosynthesis processes. Therapeutically relevant doses of BETi selectively repress a small set of BETi hypersensitive genes that are conserved across diverse myeloid leukemias (Muhar et al, Science, 2018). Collectively, our project has developed and applied innovative transcriptome profiling and functional-genetic screening methodology to establish a comprehensive map of chromatin dependencies in leukemia, decipher their primary molecular and cellular functions, and guide the discovery and development of pharmacological approaches for exploiting these promising candidate targets for leukemia therapy.