Structural Basis for LMO2-Driven Recruitment of the SCL:E47bHLH Heterodimer to Hematopoietic-Specific Transcriptional Targets

摘要

class="head no_bottom_margin" id="sec1title">IntroductionMany, if not all, cellular processes are driven by multiprotein complexes. In the nucleus, control of gene expression is achieved through assembly of such molecular machines on genomic loci in a highly controlled manner. These complexes contain combinations of regulators such as transcription factors (TFs), cofactors, and chromatin-remodeling proteins () that assemble in a modular manner for rapid adaptation of gene expression programs. Although at the heart of transcription, little is known about the molecular mechanisms driving formation of multiprotein complexes and their interaction with DNA, the role of individual components in this process, and the relationship between TFs and chromatin-remodeling proteins. To address some of these questions, we have analyzed the associations between two evolutionarily conserved families of transcriptional regulators, namely basic-helix-loop-helix (bHLH) and LIM domain-containing proteins () at structural, functional, and biochemical levels.The hematopoietic system offers a well-characterized model to study bHLH/LIM protein interactions, such as those engaged by the tissue-specific class II bHLH TF SCL/TAL1 (hereafter called SCL) and the non-DNA-binding LIM-only protein LMO2. SCL and LMO2 were initially discovered through chromosomal translocations involved in T cell acute lymphoblastic leukemia (T-ALL) (). Indeed, their ectopic expression in T cell precursors occurs in up to 60% of childhood T-ALL cases of which 80% coexpresses SCL and LMO2 (or LMO1) (). In normal hematopoiesis, SCL and LMO2 are absolutely required for hematopoietic specification and terminal differentiation of specific hematopoietic lineages ().SCL forms obligate heterodimers through its HLH domain with ubiquitously expressed class I bHLH E proteins (such as the E2A gene products, E47 and E12) (). The SCL:E47 heterodimer binds through its basic regions to an E box DNA recognition sequence (CANNTG), each monomer recognizing one-half of the E box (). The SCL:E47 heterodimer then nucleates a “core” multiprotein complex by binding to the adaptor protein LMO2 and its interacting partner LDB1 (LIM-binding domain 1) ().The SCL core complex (SCL:E47:LMO2:LDB1) acquires further specificity through recruitment of additional protein partners. In erythroid cells, LMO2 recruits hematopoietic-specific TF GATA1, creating the so-called “pentameric” complex that binds a bipartite E box-GATA DNA sequence (). The complex regulates expression of important hematopoietic-specific genes () upon recruitment of various combinations of cofactors and chromatin remodelers, such as ETO2, mSin3A, P300, PCAF, and LSD1 ().In T-ALL, the prevailing model suggests that ectopically expressed SCL and LMO2 synergize to prevent the activity of E protein homodimers, essential for normal progression of T cell differentiation. Specifically, through a sequestration mechanism that is yet to be characterized at the molecular level, E proteins are locked into ectopic SCL core complexes that act by repressing apoptotic pathways and preventing the normal T cell transcriptional program ().Interestingly, for its functions in hematopoietic specification and leukemogenesis, SCL does not rely on direct DNA-binding activities (), suggesting that it may work off DNA or be tethered to DNA through other DNA-bound TFs. Additionally, we previously showed that one-fifth of SCL’s genomic targets in erythroid cells can recruit the factor independently of its DNA-binding activity (), suggesting a critical network of additional protein:protein and protein:DNA interactions for the nucleation of such complexes.Despite a wealth of studies defining the SCL core complex as a key transcriptional regulator in normal and malignant hematopoiesis, its mechanism of action remains undefined at a molecular level. Specifically, the molecular interactions governing protein:protein and protein:DNA associations are not understood. Moreover, the molecular details of the E protein sequestration model, which confers to ectopically expressed SCL and LMO2 their oncogenic properties, are unclear. Although functional, biochemical, and biophysical studies have suggested potential pathways to complex assembly (href="#bib28" rid="bib28 bib44 bib46 bib56" class=" bibr popnode">Lécuyer et al., 2007; Ryan et al., 2008; Schlaeger et al., 2004; Wadman et al., 1997), complete dissection of the molecular rules driving complex formation has been hindered by the lack of a full structural characterization. No structure exists for SCL, and whereas other bHLH proteins and LIM proteins, including E47 (href="#bib12" rid="bib12" class=" bibr popnode">Ellenberger et al., 1994) and LMO2 (href="#bib14" rid="bib14" class=" bibr popnode">El Omari et al., 2011), have been structurally characterized, no molecular account exists of their association. The mechanistic role of each complex component and their contribution to cofactor recruitment also remain unclear.Here, we report the structure of the SCL core complex (SCL:E47)bHLH:LMO2:LDB1LID bound to DNA. This structure, which details a bHLH:LIM protein association, provides an atomic description of the interactions driving the association of the SCL:E47 heterodimer and its interface with DNA in the presence and in absence of LMO2. Analysis of the structure uncovers unexpected roles for SCL, E47, and LMO2. Functional and biochemical analyses further reveal complex synergies between components of the complex, their cofactors, and DNA targets. Together, these results deepen our understanding of how higher-order, hematopoietic-specific complexes form and how tissue-specific gene expression programs might be regulated. Importantly, the structure, which delivers insights into the oncogenic properties of the complex, will drive the design of small molecules for targeted inhibition of oncogenic processes in T-ALL.