The YEATS Domain Epigenetic Reader Proteins ENL and AF9 and Their Therapeutic Value in Leukemia

Abstract

Recent studies have uncovered similarities and differences between two highly homologous epigenetic reading proteins, ENL (MLLT1) and AF9 (MLLT3) with therapeutic implications. The importance of these proteins has traditionally been exemplified by their involvement in chromosomal translocations with the Mixed Lineage Leukemia gene (MLL; aka KMT2a). MLL-rearrangements occur in a subset of acute leukemias and generate potent oncogenic MLL-fusion proteins that impact epigenetic and transcriptional regulation. Leukemic patients with MLL-rearrangements display intermediate to poor prognoses, necessitating further mechanistic research. Several protein complexes involved in regulating RNA polymerase II transcription and the epigenetic landscape are hijacked in MLL-r leukemia that include ENL and AF9. Recent biochemical studies have defined a highly homologous YEATS domain in ENL and AF9 that binds acylated histones, which aids in the localization and retention of these proteins to transcriptional targets. In addition, detailed characterization of the homologous AHD on ENL and AF9 have revealed differential association with transcriptional activating and repressing complexes. Importantly, CRISPR knock out screens have demonstrated a unique role for wild type ENL in leukemic stem cell function, whereas AF9 appears important for normal hematopoietic stem cells. In this perspective, we examine the ENL and AF9 proteins with attention to recent work characterizing the epigenetic reading YEATS domains and AHD on both the wild type proteins and when fused to MLL. We summarize drug development efforts and their therapeutic potential and assess ongoing research that has refined our understanding of how these proteins function, which continues to reveal new therapeutic avenues.

ENL and AF9 in Transcriptional and Epigenetic Regulatory Complexes

The MLLT1 gene, which encodes the eleven-nineteen-leukemia (ENL) protein, is located on chromosome 19p13.3 and shares high sequence homology with its human paralog, MLLT3 (encoding the AF9 protein). Over thirty years ago, the discovery of the t(11;19) (q23;p13.3) (MLL1-MLLT1) and t(9;11) (p22;q23) (MLL1-MLLT3) translocations marked our introduction to these proteins as two of the most common MLL fusion partners, representing roughly 13% and 19% of the MLL-rearranged (MLL-r) cases across all ages [1–4] (Figure 1A). Given the poor prognosis of patients harboring MLL fusion proteins, significant effort has been made to understand the functions of ENL and AF9. Characterization of these two proteins revealed their essential roles in development, as ENL knockout is embryonically lethal at ~8.5 dpc [5] and AF9 null mice exhibit perinatal lethality and die within 2 weeks of birth [6]. Early biochemical studies, including several using ENL or AF9 as bait proteins, converged on the main finding that the most common MLL fusion partners exist in several overlapping transcriptional and epigenetic regulatory complexes: the Super Elongation Complex (SEC; used hereinafter), the AF4family/ENL family/P-Tefb complex (AEP), and the ENL Associated Proteins (EAPs)[7–9]. These regulatory complexes contain two core enzymatic units important for their function: the positive transcription elongation factor (P-TEFb), which phosphorylates RNA polymerase II (Pol II), and the DOT1-like (DOT1L) protein, which catalyzes H3K79me2/3 on histone H3 (Figure 1B). H3K79me2/3 is thought to associate with actively transcribed regions of the genome [10,11]. The question is how are these complexes impacted by ENL or AF9 fusion to MLL in human leukemias? We now understand that P-TEFb and DOT1L are hijacked to pro-leukemic target genes by these common MLL fusion proteins, which results in dysregulated gene expression[7,8,12–17]. The SEC is also implicated in Tat-mediated HIV transcription, as well as transcriptional regulation of developmental genes and pathways in Drosophila and in ES cells[18–21], highlighting their importance in both normal and disease settings.

Figure 1: Structure of MLL Fusion Proteins and ENL/AF9 Protein Complexes.

(A) Structure of the wild type MLL protein is shown at top with domains labeled. The ENL protein, MLL-ENL fusion protein, AF9 and MLL-AF9 fusion proteins are shown below. The green dashed line indicates an approximate breakpoint within the various proteins following translocation. Approximately 84% of MLL-ENL leukemia patients harbor breakpoints at or before amino acid 5 of ENL following translocation with MLL, resulting in inclusion of the YEATS domain, IDR and AHD. Approximately 92% of MLL-AF9 leukemia patients harbor breakpoints at or after amino acid 376 of AF9 and exclude the YEATS domain but retain the AHD. (B) Model of ENL/AF9 protein with domains labeled. The YEATS domain is shown bound to acetylated histones H3K9/K18/K27. The YEATS domain is also shown bound to the PAF1c and MOZ. The AHD binds to the SEC and DOT1L. DOT1L is shown methylating histone H3K79. The AHD also binds to CBX8 and BCOR, which are components of the Polycomb Repressive Complex 1 (PRC1) and derivatives.

Studies have now begun to map out the protein-protein interactions within these transcriptional and epigenetic regulatory complexes. In ENL and AF9, we now understand that the C-terminal ANC-1 homology domain (AHD) interacts with the DOT1L protein directly as well as the SEC anchoring proteins AFF1, AFF4 and MLLT10 [7,22–24] (Figure 1B). At the N-terminus of ENL and AF9 is a conserved YEATS domain capable of binding and “reading” epigenetic modifications like histone acylation [25,26] (discussed in detail below). The ENL and AF9 YEATS domain also interacts with the Polymerase Associated Factor 1 Complex (PAF1c)[27,28]. The PAF1c is a protein complex that associates with Pol II and regulates both epigenetic deposition and transcriptional activation (for extensive reviews of the PAF1c, see [29,30]). In addition, the histone acetyltransferase MOZ (aka MYST3, KAT6A) complex associates with the YEATS domain of ENL and is recruited to target genes through interactions with MLL[31]. Further, recent studies proposed a model where the YEATS domain within the MOZ/ENL complex is needed for recruitment of AF10 fusion proteins (MLL-AF10, CALM-AF10 and others) to the HOXA gene cluster, critical for leukemogenesis[32]. These studies place ENL and AF9 at the center of the epigenetic and transcriptional control of target genes.

The biochemical characterization of ENL and AF9 has: 1) built our foundational knowledge on the roles of these epigenetic and transcriptional regulatory complexes in disease mechanisms and 2) revealed these transcriptional and epigenetic complexes as potential targets for therapeutic development. For instance, both the PAF1c and DOT1L have been demonstrated to be essential factors for MLL-r leukemogenesis [13,14,16,33–37]. However, PAF1c components have also been shown to be essential for proper hematopoietic stem cell function[38]. Further, direct chemical targeting of DOT1L has shown only modest effectiveness in clinical trials[39]. A more complete understanding of these disease mechanisms is needed to optimize therapy. Here we investigate how recent studies investigating ENL and AF9 interactions could impact therapeutic development. For example, important functional differences between ENL and AF9 have recently been described in leukemic cells and hematopoietic stem cells[25,40–42]. Work from multiple investigators have demonstrated the existence of different SEC and SEC-like complexes, all of which involve ENL and/or AF9 [27,43,44]. However, questions remain on how these SECs and derivatives work in concert or opposition under normal and pathological conditions. Given the high sequence homology between ENL and AF9 in the N-terminal YEATS domain (~;88%), the C-terminal ANC-1 Homology Domain (AHD) (~80%), and ~74% overall, recent reports defining functional differences are particularly interesting[45–47]. Recent structural studies of the YEATS domain and the AHD of ENL and AF9 have identified unique characteristics that better define their roles in transcriptional regulation in normal and diseased cells.

The ENL YEATS Domain: An Epigenetic Binder with Therapeutic Potential

Two back-to-back studies in 2017, published by Wan et al. and Erb et al., reported the importance of wild type ENL, but not wild type AF9, for leukemic cell growth using several AML models [25,40]. A more detailed analysis revealed that the YEATS domain was required for ENL functionality in AML cells [25,40]. The name “YEATS” comes from 5 protein members containing this conserved domain (Yaf9, ENL, AF9, Taf14, and Sas5). This domain is highly conserved and generally thought of as an epigenetic reader domain that can recognize histone acylation with the strongest preference for crotonylated histone marks[48]. In humans, there are 4 conserved YEATS domain proteins: ENL (YEATS1; encoded by the MLLT1 locus), YEATS2, AF9 (YEATS3; encoded by the MLLT3 locus), and GAS41 (YEATS4) [49]. The YEATS domain adopts a β-sandwich conformation, with an aromatic cage that binds to acetylated lysine residues[25,26,50–52]. Structural analyses of the YEATS aromatic cage have identified key amino acid residues contributing to the epigenetic reader function [25,26,50–52]. All 4 human YEATS proteins have confirmed acetylated reader function with varying degrees of affinity for acetylated H3 lysines and weak to no interactions with acetylated H4 or H2B lysines[25,26,50,52]. For example, both ENL and AF9 YEATS domains show affinity for H3K9ac, 18ac and 27ac and have a much weaker interaction with H3K14ac[25,26]. It is likely that the selective affinity has functional consequences, but further research is required. All human YEATS domain proteins are associated with protein complexes that regulate epigenetic and transcriptional mechanisms, which have been linked to the progression of several different cancers (summarized in Table 1).

Table 1:

List of human YEATS proteins, their participation in epigenetic and transcriptional regulatory complexes, and their association to human cancers.

Member	Protein Complex	Cancer Association
YEATS1 (ENL)	AEP/EAP/SEC and similar complexes[7–9,43], DotComm[21], EBAFb[86]	Translocation in MLL-r leukemia[3,4,87], Wilm’s tumor[58]
YEATS2	ATAC (metazoan HAT)[52]	Amplified in lung squamous cell carcinoma, ovarian serous cystadenocarcinoma, and head & neck squamous carcinoma[52]
YEATS3 (AF9)	AEP/EAP/SEC and similar complexes[7–9,43], DotComm[21]	Translocation in MLL-r leukemia[2,4]
YEATS4 (GAS41)	SRCAP, Tip60/p400[50]	Amplified in sarcoma, NSCLC, lung cancer, bladder cancer, uterine cancer, and glioblastoma[50,88]

The YEATS domains of all but AF9 have been linked with oncogenic functions[25,40,50,52,53]. In the case of YEATS2 and GAS41, the YEATS epigenetic reader function is linked with maintaining an epigenetic landscape conducive for cancer cell proliferation and survival[50,52]. As mentioned above, the YEATS domain is necessary for wild type ENL to promote AML growth[25,40]. In the context of MLL rearrangements, we and others found the YEATS domain is included in the majority of MLL-ENL fusion proteins in leukemic patients with (11;19) translocations (Figure 1A) [53–55]. Further, the YEATS domain impacts the leukemic stem cell frequency of MLL-ENL leukemia by helping to stabilize the MLL-ENL fusion protein at target loci [53]. Of note, while this study found the YEATS domain was necessary for MLL-ENL mediated leukemia in vivo [53], Yokoyama and colleagues determined a similar MLL-ENL fusion protein lacking the YEATS domain is capable of inducing leukemia in vivo[56]. Further, an artificial construct that fuses the PWWP domain of LEDGF, the CXXC domain of MLL, and the AHD of ENL induces leukemia in vivo[57]. Conversely, leukemia patients harboring a (9;11) translocation almost exclusively exclude the YEATS domain from resultant MLL-AF9 fusion proteins[53] (Figure 1A). This finding is striking given the high sequence homology and functionality between the YEATS domains of ENL and AF9. Why is the YEATS domain differentially retained in leukemia patients harboring MLL-ENL vs MLL-AF9 fusion proteins? Additional research is needed to more fully understand how these domains impact MLL fusion protein functionality.

In addition to the role of the YEATS domain in both wild type ENL and MLL-ENL fusion proteins in leukemia, somatic mutations have been found in the YEATS domain of ENL in a small subset (~5%) of Wilm’s tumor patients [58]. Wilm’s tumor, also known as nephroblastoma, is the most common form of childhood kidney cancer[59]. Patients with in-frame indels in the YEATS domain display dysregulated MYC and HOX gene expression[58]. Mechanistically, the small indels in the YEATS domain promote a pathological level of ENL self-association on chromatin, which is dependent upon its epigenetic reader function[60,61]. These findings expose the ENL YEATS domain as a potential target for therapeutic development for multiple malignancies.

Targeting the ENL YEATS Domain in Hematological Malignancies

Given the roles of YEATS domain proteins in human disease, identifying, and improving small molecule inhibitors targeting the YEATS domain has garnered recent interest. Genetic mutations perturbing the YEATS epigenetic reader pocket dislodge ENL and MLL-ENL fusion proteins from its transcriptional targets most notably around the transcriptional start sites (TSS) [25,53]. These genetic experiments provided proof of principle for small molecule screens in search of selective inhibitors to the ENL YEATS domain. The chemical probe SGC-iMLLT, discovered by Brennan, Fedorov and colleagues, was the first small molecule compound specifically targeting the ENL and AF9 YEATS domains with significantly less affinity for GAS41 and YEATS2[47]. Presumably, this reflects the different epigenetic reader conformations of the ENL/AF9 YEATS domains and the GAS41/YEATS2 YEATS domains[50]. Indeed, a panel of diverse mouse and human leukemia cell lines all displayed variable sensitivity to SGC-iMLLT [53]. This is consistent with a role for wild type ENL in maintaining growth of leukemic cells [25,40]. Notably, a human B-ALL cell line harboring an MLL-ENL fusion protein containing the YEATS domain, HB11;19 cells, (Figure 1), were amongst the most sensitive [53]. The in vivo half-life of the SGC-iMLLT compound was optimized by the work of Wan and colleagues resulting in the small molecule inhibitor TDI-11055 while retaining target selectivity and potency [41]. TDI-11055 successfully dislodges wild type ENL from its genomic targets and inhibits cellular proliferation and colony formation of several AML cell lines and primary leukemic samples[41].

Given the role of ENL as an epigenetic reader linking transcriptional and epigenetic complexes to chromatin, YEATS inhibitors should disrupt the recruitment of the SEC and DOT1L. Indeed, TDI-11055 treatment reduces binding of the SEC scaffold protein AFF1 and DOT1L at ENL targets such as ZEB2, MYC and HOXA9/10[41]. This finding is consistent with sgRNA mediated genetic depletion of ENL, which resulted in reduced SEC occupancy at ENL targets [25,40]. In the context of MLL-ENL fusion proteins, a subset of direct MLL-ENL target genes appear to rely more heavily on the YEATS domain, most notably Eya1 [53]. Consistent with a role for the YEATS domain in targeting epigenetic and transcriptional complexes to the Eya1 locus, H3K79me2, H3K9ac and H3K4me3 are all markedly reduced following mutation of the YEATS domain on MLL-ENL fusion proteins [53]. SGC-iMLLT treatment of MLL-ENL mouse cell lines, which would target both the MLL-ENL fusion protein and wild type ENL, also significantly reduces Eya1 expression[53]. Together, both genetic and small molecule approaches clearly demonstrate the importance of the ENL YEATS epigenetic reader function in maintaining the epigenetic landscape and transcriptional expression of proleukemic target genes in leukemic cells[25,40,53].

Intriguingly, it was revealed that NPM1-mutant AMLs also respond to TD1-11055 treatment[41]. Two recent reports on NPM-1 mutant AMLs elucidated mechanisms into NPM-1 mutant leukemogenesis that may provide further insight into the role of ENL and the YEATS domain[62,63]. NPM-1 mutant proteins bind to genomic regions demarcated by H3K27ac. There it forms molecular condensates with the MLL-Menin complex and SEC to dysregulate pro-leukemic targets[62,63]. The presence of the SEC directly implicates ENL and, consequently, NPM-1 mutant AMLs are sensitive to both Menin inhibitors [63] and YEATS inhibitors [41].

It is intriguing that genetic perturbation of the MLL-ENL YEATS epigenetic reader function reduces leukemic stem cell (LSC) frequency[53]. In a separate report, Liu et al. showed that TDI-11055 treatment of the MV4;11 AML cell line impacts an LSC gene signature[41]. Several epigenetic proteins have been associated with AML LSCs. LSD1, a histone H3K4 and H3K9 demethylase [64], is highly expressed in a subset of AMLs[65]. Targeted LSD1 knockdown destroys LSCs in a mouse AML model, demonstrated by a lack of leukemia in secondary transplants [66]. Currently, several compounds targeting LSD1 are in phase I/II clinical trials[67]. The therapeutic targeting of epigenetic proteins to target LSCs is intriguing and continues to develop. In addition to ENL and LSD1, more epigenetic writers, readers, and erasers are continually being identified that contribute to LSC gene programs: such as DOT1L, HBO1, G9A, SUV39H1, SETD2 and others[16,68–71]. As the molecular and cellular characterization of these proteins continues, it must be considered how therapeutic targeting may impact normal hematopoietic stem cells and other tissues thereby narrowing a therapeutic window.

What are the potential side effects of using a YEATS inhibitor to treat AMLs, such as those driven by MLL-r and NPM1 mutations? Developmental studies on ENL in mice report that ENL is essential for embryonic development [5]. ENL knockdown in hematopoietic progenitor cells minimally impacts their proliferation and differentiation[25], although this observation should be further expanded to wholistically examine the hematopoietic system in an in vivo model. Current YEATS inhibitors selectively target the YEATS domains of both ENL/AF9 but not GAS41 and YEATS2[41,47]. AF9 is required for proper embryonic development, hematopoietic stem cell function, and proper erythroid/megakaryocytic lineage commitment[42]. Given the role of the YEATS domain in directing ENL and AF9 to their genomic targets, YEATS inhibitors may dislodge AF9 from genomic targets and profoundly affect the hematopoietic system. While inhibition of the YEATS epigenetic reader function is an attractive therapeutic strategy, additional mechanistic characterization is needed to address these concerns.

ENL in Liquid-Liquid Phase Separation and Condensate Formation

In 2015, Perlman et al. reported a cluster of ENL indel mutations in approximately 5% of Wilm’s tumor patients[58]. These mutations reside within the YEATS domain and it was speculated that they would impact YEATS epigenetic reader function or PAF1 interaction. While biochemical analyses ruled out alterations to the YEATS-PAF1 interaction[28,60], it was unclear how these mutations may impact the YEATS epigenetic reader function [28,58,60]. One model proposed that these mutations may drive increased ENL occupancy at target loci independent of the YEATS domain’s epigenetic reader function[60]. In a follow up study, Wan and colleagues demonstrated that YEATS domain indel mutations alter the secondary structure of the YEATS domain, triggering a pathogenic level of ENL self-association and ENL condensate formation with the SEC at its genomic targets[61].

Commonly referred to as condensates or puncta, this phenomenon of regulatory protein assembly on chromatin has recently been more extensively characterized. These biomolecular “hubs” of regulatory proteins on the chromatin are thought to associate through multivalent and non-stoichiometric interactions mediated in part by intrinsically disordered regions (IDR) of transcriptional factors and co-activators that induce liquid-liquid phase separation and biocondensate formation[72,73]. In ENL, an intrinsically disordered region was identified between the YEATS domain and the AHD (Figure 1). Among MLL fusion partners, two IDR-containing proteins (ENL and AF4) have been shown to initiate biocondensate formation with the SEC/P-TEFb [74]. In a study by Wan and colleagues, it was further clarified that the ENL IDR contributes to phase separation through its regional net charges and serine-rich sequence[61]. These pathogenic models allow us to appreciate the role of the ENL IDR in SEC nuclear biocondensate formation, but many questions remain. For instance, how does the expression level of ENL impact the IDR’s contribution to SEC nuclear biocondensate under normal physiological conditions? How does the ENL IDR-mediated condensate formation and protein interactions work in concert with AHD-mediated protein-protein interactions to connect ENL with SEC?

Our studies found a minor subset of MLL-ENL patients that lack the ENL YEATS domain and part of the IDR [53]. In experimental mouse models, these truncated forms of the MLL-ENL fusion protein fail to induce leukemia in vivo. Do MLL-ENL fusion proteins with an intact IDR also form biocondensates with the SEC, and if so, how does this mechanism contribute to MLL-ENL leukemogenesis? Considering our recent understanding of the ENL IDR and phase separation, these are important questions to consider in the context of MLL-ENL leukemias. These mechanistic studies will help us understand how the SEC is hijacked in MLL-ENL leukemias and why MLL-r leukemias are sensitive to loss of ENL, which may aid in therapeutic design.

The role of the ANC-1 Homology Domain (AHD) in Disease Mechanisms

The ENL and AF9 ANC1 Homology Domain (AHD) is invariably retained in all MLL-ENL and MLL-AF9 fusion proteins (Figure 1). The AHD of ENL and AF9 are intrinsically disordered domains that interact with the SEC and DOT1L[8,24] (Figure 1). Indeed, the ENL and AF9 AHDs are required in MLL-ENL and MLL-AF9 mediated transformation[8,12,75]. Importantly, point mutations perturbing the AHD-DOT1L or AHD-SEC interactions abrogate MLL-ENL and MLL-AF9 ex vivo colony formation[8], highlighting the importance of these interactions in both MLL-ENL and MLL-AF9 leukemias. Thus, targeting of the AHD-DOT1L protein-protein interaction was tested as a therapeutic target for leukemia treatment. Nikolovska-Coleska and colleagues reported the first peptidomimetic inhibitor disrupting the interaction of DOT1L with the AHD of AF9 and/or ENL [76]. Targeting of DOT1L has been demonstrated with Pinometostat, a small molecule inhibitor that blocks DOT1L enzymatic activity and demonstrates anti-leukemic effect against MLL-r models[77,78]. However, in clinical trials, this small molecule inhibitor saw only modest effects due to drug resistance imparted through upregulation of drug efflux transporters like ABCB1 [39,79]. Armstrong and colleagues have recently shown that dual inhibition of Menin and DOT1L completely dislodges the MLL-AF9 fusion protein from its targets compared to single agent treatment[80]. Multiagent design for MLL-r leukemias may benefit from the recent findings on ENL and AF9 and development of YEATS and AHD inhibitors.

In addition to their interactions with the SEC and DOT1L, the ENL and AF9 AHDs also interact with the PRC1 polycomb group proteins (PcGs) CBX8 and BCOR[7,81–84]. Interestingly, these PcGs contribute to the proper growth of MLL-ENL and MLL-AF9 leukemia cells[82–85]. At least three models have been proposed to address the binding of the AHD to both activating and repressive regulatory complexes: 1) the ENL AHD-CBX8 interaction neutralizes PRC1’s repressive influences [83], 2) the AF9 AHD-CBX8 interaction recruits additional activating complexes including Tip60 [84] and 3) the AF9 AHD-BCOR interaction is essential for regulating distinct transcriptional targets[82]. In particular, Bushweller and colleagues recently reported a difference in biochemical affinity between the AHD of ENL and AF9 to BCOR and CBX8 while maintaining similar affinity to DOT1L and the SEC scaffold protein AF4[82]. Specifically, the ENL AHD displayed much higher affinity for BCOR and CBX8 compared to the AF9 AHD pointing to important biochemical differences[82]. CBX8 is required in MLL-ENL and MLL-AF9 leukemias[83,84], while BCOR interaction with AF9 (on the MLL-AF9 fusion protein) has been shown to be required for MLL-AF9 leukemia[82]. In support of gene specific complexes regulating subsets of direct MLL-AF9 targets, disruption of the MLL-AF9-BCOR interaction disrupts Eya1 but not Hoxa9 and Meis1 expression[82]. This is remarkably similar to our studies that identified Eya1, but not Hoxa9 and Meis1, as a direct MLL-ENL target gene dependent on the ENL YEATS epigenetic reader function [53]. From these observations, several questions specific to MLL-ENL and MLL-AF9 fusion proteins arise. Are subsets of MLL-fusion target genes regulated by specific protein complexes? What are the mechanisms that mediate gene specific assembly of unique MLL-fusion protein complexes? Additional studies are needed to interrogate these questions and others as we explore the possibilities of targeting multiple activities within the MLL-fusion super complex.

Concluding Remarks and Translational Outlooks

The biomedical importance of ENL and AF9 originated from their fusion to MLL in MLL-r leukemias. Indeed, they are two of the most common MLL fusion partners in patients with MLL-r acute myeloid and lymphoid leukemia. Structure/function studies focusing on MLL-ENL and MLL-AF9 began to identify the protein domains important for leukemogenesis. These found the C-terminal ANC-1 homology domain (AHD) as absolutely necessary for leukemogenesis. However, recent studies now show an important role for the N-terminal YEATS domain of both wild type ENL and when fused to MLL on MLL-ENL fusion proteins. Recent studies that identified wild type ENL, but not wild type AF9, as a necessary transcriptional and epigenetic factor in a broad range of acute myeloid leukemias has renewed interest in ENL as a therapeutic target. In particular, the YEATS domain of wild type ENL has emerged as an ideal target to disrupts its H3ac binding activity and genomic localization. As described above, the YEATS domain may also prove to be a point of susceptibility for directly targeting a subset of MLL-ENL fusion proteins. Recent discoveries focusing on the various domains of ENL and AF9 have all contributed to a more precise understanding of the similarities and differences between ENL and AF9 including their biochemical nature and role in development and disease. These studies refine our knowledge of ENL and AF9 and expose their potential as drug targets in hematological malignancies and other diseases.

Highlights.

• The YEATS domain is an epigenetic reader for histone acylation.

• ENL associates with several protein complexes that regulate transcription.

• The YEATS domain is differentially retained in MLL-ENL and MLL-AF9 fusion proteins in leukemia patients.

• Chemically targeting epigenetic readers, such as the ENL YEATS domain, represents an attractive target for treating leukemia.