Focus on Epigenetic Control of host defence
The MYSTerious MOZ, a histone acetyltransferase with a key role in haematopoiesis
Correspondence: Georges Lacaud, Cancer Research UK Stem Cell Biology Group, Paterson Institute for Cancer Research, The University of Manchester, Wilmslow Road, M20 4BX, Manchester, UK. Email: email@example.com Flor M. Perez-Campo, Cancer Research UK Stem Cell Biology Group, Paterson Institute for Cancer Research, The University of Manchester, Wilmslow Road,M20 4BX, Manchester, UK. Email: firstname.lastname@example.org
Senior author: Dr Georges Lacaud
The MOnocytic leukaemia Zing finger (MOZ; MYST3 or KAT6A1) gene is frequently found translocated in acute myeloid leukaemia. MOZ encodes a large multidomain protein that contains, besides others, a histone acetyl transferase catalytic domain. Several studies have now established the critical function of MOZ in haematopoiesis. In this review we summarize the recent findings that underscore the relevance of the different biological activities of MOZ in the regulation of haematopoiesis.
Cells of the immune system are generated from multipotent haematopoietic stem cells (HSCs) through an elaborate process of differentiation. Self-renewal of the HSCs coupled with a tightly controlled differentiation are therefore vital to replenish the blood and immune system throughout the lifespan of an organism. Proliferation of HSCs, as well as the successive stages of differentiation into more committed progenitors, are regulated at the transcriptional level through epigenetic modifications. These modifications include DNA methylation, histone modifications (such as acetylation, methylation, phosphorylation, SUMOylation and ubiquitylation) and chromatin remodelling. The enzymes responsible for these chromatin modifications are therefore critical for normal blood and immune system homeostasis. The finding that these chromatin modifiers are common targets of mutations or translocations in leukaemia further highlights their functional requirements.
MOZ is a histone acetyltransferase (HAT) frequently translocated in acute myeloid leukaemia, with a crucial role in controlling HSC proliferation.[2-4] Although the MOZ gene was originally identified more than 15 years ago in human leukaemia, the different biological activities of MOZ and their critical requirements are only starting to be unravelled. In this snapshot review, we will highlight the evidence demonstrating not only that MOZ functions as a histone acetyltransferase in HAT complexes, but also that MOZ acts as a transcriptional co-activator of diverse transcription factors in blood cells and other tissues. We will also present the findings suggesting that this novel epigenetic regulator plays a key role in blood homeostasis. In line with the critical function of MOZ in haematopoiesis, chromosomal translocations of MOZ with either CBP, p300 or TIF2 are recurrently found in acute myeloid leukaemia. As all these fusion partner proteins have, or are associated with, a HAT activity, it has been proposed that aberrant acetylation of MOZ transcriptional targets by these newly formed fusion proteins might participate in the process of leukaemogenesis.[5, 8]
1. MOZ, a histone acetyltransferase
MOZ is a founding member of the MYST family of lysine acetyltransferases (KATs), together with Ybf2/Sas3, Sas2 and TIP60.[5, 9] This protein family is defined by a MYST domain, comprising a C2HC nucleosome binding domain and a region with homology to an acetyl-coenzyme A binding site (HAT domain).
(a) MOZ has been shown to have intrinsic HAT activity and to acetylate both itself and lysine (K) residues on histone H2B, histone H3 (K14) and histone H4 (K5, K8, K12 and K16) in vitro[10-12] and H3K9 in vivo.
(b) In HeLa cells, MOZ was shown to be the catalytic subunit of a tetrameric INhibitor of Growth 5 (ING5) complex, which specifically acetylates nucleosomal histone H3K14. In addition to ING5 and MOZ, this complex contains BRomodomain PHD-Finger protein 1 (BRPF1; or its paralogues BRPF2 and BRPF3), which interacts with MOZ through the MYST domain, and serves as a scaffold subunit between MOZ and ING5 proteins and the fourth partner Esa1-Associated Factor 6 (hEAF6).[15, 16] Besides bridging the interaction between MOZ and the rest of the components of the complex, BRPF1 enhances the acetyltransferase and co-activator activities of MOZ.[15, 16] Supporting the physiological relevance of this complex, MOZ and ING5 have been shown to participate together in the in vivo regulation of T-box DNA binding Transcription Factor 1 gene (Tbx1) expression, a gene found hemizygously deleted in patients suffering from DiGeorge/velo-cardio-facial/22q11 deletion syndrome.[18-20] Although MOZ is able to acetylate BRPF1, the relevance of this inter-subunit acetylation is yet to be elucidated. Finally, a recent study investigating how MOZ and this complex are recruited to its binding sites has revealed that the tandem PHD finger of MOZ (PHD12) recognizes unmodified arginine residue 2 (R2) and acetylated lysine residue 14 (K14ac) on histone H3, promoting further acetylation of H3K14.
2. MOZ, a transcriptional co-activator
Besides its activity as an HAT, MOZ acts as a co-activator for several DNA-binding transcription factors, particularly with haematopoietic specificity, such as acute myeloid leukaemia/Runt-related transcription factor 1 (AML1/RUNX1),[11, 22] Mixed Lineage Leukaemia 1 (MLL1) and PU.1 (Spi-1).
(a) The gene encoding the transcription factor AML1/RUNX1 is the most frequent target of chromosomal translocations leading to leukaemia in humans, and RUNX1, together with its binding partner Core binding Factor-β, is critical for haematopoietic development in vivo and in vitro.[24-26] MOZ has been shown to interact with RUNX1 and to stimulate in vitro RUNX1-dependent transcription of haematopoietic genes,[11, 22] such as the Macrophage Inflammatory Protein 1-α (MIP1-α) and Myeloperoxidase (MPO) genes. Interestingly, the HAT domain of MOZ seems to be dispensable for this stimulation of RUNX1 transcriptional activity. In contrast, phosphorylation of MOZ and RUNX1 by the Homeodomain Interacting Protein Kinase 2 (HIPK2) strongly reinforces this interaction and the stimulation of transcription. In addition to RUNX1, MOZ physically interacts with MLL (KMT2A), a histone methyltransferase that catalyses the dimethylation and trimethylation of lysine 4 in histone H3 (H3K4me2 and H3K4me3 respectively). MOZ also associates with WDR5, an adaptor protein that stabilizes the binding of the MLL complexes at gene promoters, therefore facilitating the trimethylation of H3K4. MOZ and MLL were shown to be co-ordinately recruited to HOX promoters and to co-operate in the epigenetic control of HOXA5, HOXA7 and HOXA9 genes. Finally, MOZ interacts with PU.1, a transcription factor belonging to the ETS family that is required for the normal development of the lymphoid and myeloid lineages,[30, 31] as well as for the maintenance of HSCs and their differentiation.[32, 33] MOZ is able to interact with PU.1 and to activate PU.1-dependent transcription. Interestingly the fusion proteins MOZ-TIF2 and MOZ-CBP induce a stronger PU.1-dependent activation of the expression of macrophage-colony stimulating factor receptor CSF1R (M-CSFR or CSF1R/c- FMS/CD115) than the normal MOZ protein.
(b) In addition to RUNX1, MOZ co-operates with RUNX2, another member of the Runt domain family of proteins, with a crucial role in osteogenesis. MOZ also interacts with the tumour suppressor p53 upon DNA damage to induce p21 expression and G1 phase cell cycle arrest in mouse embryonic fibroblasts. Moreover, during hepatocarcinogenesis MOZ has been shown to immunoprecipitate with the transcription factor v-maf musculoaponeurotic fibrosarcoma oncogene homolog K (MAFK) and to stimulate the activity of the NRF2-MAFK heterodimer. MOZ and its fusion protein with CBP (MOZ-CBP) are capable of interacting with the p65 subunit of the nuclear factor-κb (NF-κB) and to enhance expression of NF-κB target promoters. Finally, interactions of MOZ with c-JUN, the ETL transcription factor TEL and CBP have also been described.
3. MOZ, a key epigenetic regulator of haematopoiesis
To analyse the function of MOZ and of its HAT activity in normal haematopoiesis, several groups have generated mice with distinct targeted alleles of MOZ. In the mouse model generated by Thomas et al. the C-terminal part of MOZ had been deleted (MOZΔ), which unexpectedly resulted in the absence of the anticipated truncated MOZ protein. Katsumoto et al. directly deleted the complete coding sequence of MOZ (MOZ−) and the resulting MOZ−/− mice suffer from embryonic lethality by mid-gestation whereas MOZΔ/Δ mice die at birth. In both models, the cellularity of haematopoietic organs is strongly decreased. Although precursors for all haematopoietic lineages and HSCs are present, their frequencies, as detected by phenotypic analyses or replating assays, and their functionalities are dramatically reduced. In addition, perturbations of erythroid development are observed in both cases whereas decreased numbers of T cells (MOZΔ/Δ) or defects in B cells and myeloid development (MOZ−/−) are more specifically observed. To more specifically investigate the specific importance of the HAT activity of MOZ, Perez-Campo et al. developed a mouse line exclusively lacking the HAT activity of MOZ (MOZHAT−/−). The MOZHAT−/− mice exhibit, similarly to the previous models, significant defects in the number and functionality of HSCs and haematopoietic committed precursors, as well as less immature B-cell development, supporting the critical function of MOZ acetylation on HSCs. Furthermore, the failure to maintain a normal number of haematopoietic precursors was shown to directly result from their inability to proliferate. These findings suggest a specific role of MOZ-driven acetylation in controlling a desirable balance between proliferation and differentiation in haematopoietic progenitor and stem cells.
The work of the author is funded by Leukaemia & Lymphoma Research (LLR), Cancer Research UK (CRUK) and Biotechnology and Biological Sciences Research Council (BBSRC).