Jeffrey A. Magee - US grants

PROJECT SUMMARY/ABSTRACT The goal of this proposal is to understand how normal developmental programs shape the genetic and epigenetic landscapes of acute myeloid leukemia (AML). AML can occur at any stage of life yet the mutations that cause AML differ between childhood and adulthood, especially when one compares young children to adults. For example, the Flt3 Internal Tandem Duplication (FLT3ITD) mutation is common in adolescent and adult AML, but it is rare in infant and early childhood AML. Likewise, MLL translocations are found in a majority of infant AML, yet they are rare in adult AML. These observations suggest that FLT3ITD may transform adult hematopoietic progenitors more efficiently than fetal progenitors, and MLL translocations may transform fetal progenitors more efficiently than adult progenitors. In preliminary studies, we discovered that Flt3ITD and cooperating Flt3ITD/Runx1 mutations caused hematopoietic stem cell (HSC) depletion and myeloid progenitor expansion in adult, but not fetal, stages of life. FLT3ITD activated STAT5 signal transduction in fetal, neonatal and adult progenitors, yet it did not induce changes in target gene expression until ~2 weeks after birth. The data suggest that fetal and neonatal progenitors are protected from transformation because they are not competent to express FLT3ITD target genes. Either they lack key transcriptional co-activators, or the epigenetic landscape of fetal progenitors suppresses FLT3ITD target gene activation. In parallel studies, we tested whether a tetracycline-inducible MLL-ENL allele transforms fetal progenitors more efficiently than adult progenitors. Fetal MLL-ENL induction caused AML in almost all mice tested. Adult induction did not cause AML in any of the mice tested (now at 6 months follow-up). These results suggest that adult progenitors resist transformation by MLL-ENL much like fetal progenitors resist transformation by FLT3ITD. Further work is needed to understand the cis- and trans-regulatory elements that determine when and how individual mutations are competent to transform. In Aim 1, we propose to precisely characterize the transition from fetal to adult transcriptional programs in developing HSCs and myeloid progenitors using Drop-seq and ATAC-seq technologies. In Aim 2, we propose to test whether Flt3ITD and cooperating Flt3ITD/Runx1 mutations have developmental context- specific effects on gene regulation and leukemogenesis. We will test whether enhancers for Flt3ITD and Flt3ITD;Runx1 target genes have age-specific patterns of accessibility and inaccessibility during development, and whether adult-specific, heterochronic transcription factors are necessary for AML formation. In Aim 3, we will test whether MLL-ENL has developmental context-specific effects on gene regulation and leukemogenesis. If we can understand how normal developmental programs interact with genetic mutations to cause malignancies, it may be possible to target these interactions therapeutically.

PROJECT SUMMARY The goal of this proposal is to understand how KMT2C/MLL3 regulates hematopoietic stem cell (HSC) self- renewal and why KMT2C mutations convey a selective advantage that can lead to leukemia. KMT2C encodes MLL3, a COMPASS family histone methyltransferase that binds enhancer elements and promotes transcription. KMT2C is mutated in human leukemias, both as part of large deletions of chromosome 7q and at specific domains such as the PHD zinc finger domains (which bind chromatin) or the SET methyltransferase domain (which can prime enhancer elements for activation). Prior murine studies have established that Kmt2c deletions enhance HSC self-renewal and promote leukemogenesis, but the mechanism is not clear. To better understand how Kmt2c regulates HSC self-renewal, we generated germline and conditional loss-of-function mice. Kmt2c deletions enhanced HSC self-renewal, consistent with prior observations, but the mutations did not alter cell cycle kinetics by themselves. Instead, Kmt2c deletions allowed serially transplanted or chemotherapy treated HSCs to retain self-renewal capacity after multiple division cycles. This allowed the mutant HSCs to outcompete wild type HSCs during marrow recovery. In the absence of stress, Kmt2c deletions did not convey a selective advantage. Altogether, our data suggest that Kmt2c mutations mitigate a phenomenon, called HSC exhaustion, in which HSCs lose self-renewal capacity after several cumulative divisions. Our mechanistic data suggest that MLL3 primes HSCs to differentiate in response to IL-1, and possibly other inflammatory cytokines, by either enhancing IL-1 signal transduction or by facilitating IL-1 target gene expression. The aims of this proposal are designed to extend these observations. Aim 1 will test whether Kmt2c/MLL3 deficiency conveys a selective advantage to dividing HSCs by reducing sensitivity to IL-1 and other inflammatory cytokines. Aim 2 will characterize the structure, regulation and IL-1 responsiveness of MLL3 target enhancers in HSCs with short and extensive division histories. Changes in enhancer priming may allow HSCs to archive their division histories and favor commitment, rather than self-renewal, after multiple division cycles. Aim 3 will test whether MLL3 requires functional SET or PHD domains to restrict HSC self-renewal capacity. This structure-function analysis will help us better understand how specific KMT2C mutations might convey a self-renewal advantage. If we can understand how HSCs change as they undergo cumulative self-renewing divisions, and how Kmt2c deletions convey a selective self-renewal advantage, we may ultimately be able to preserve HSC function through periods of stress, such as post-chemotherapy periods, without increasing leukemia risk.