Skip to main content

Function and dynamics of Lmo4-containing complexes during erythropoiesis

Final Report Summary - LMOFUNDYN (Function and dynamics of Lmo4-containing complexes during erythropoiesis)

Anemias and erythroid malignancies caused by impaired formation of erythrocytes (red blood cells) are common disorders. These diseases have a severe impact on the quality of life of the patients. It is essential to unravel the molecular mechanisms underlying erythropoiesis (the development of red blood cells) to be able to fully understand the aetiology of these diseases and to devise novel treatments. The most important part of understanding the molecular mechanisms of erythropoiesis is the unravelling of its transcription factor (TF) network.

A number of TFs appear to have essential functions during erythroid development/differentiation. They generally form large protein complexes involved in the repression of non-erythroid genes and/or in the activation of an erythroid specific program. One of such complexes, playing a crucial role in erythropoiesis, contains Ldb1, Gata1, the LIM domain only protein LMO2, and a number of other protein factors. This Ldb1 complex was shown to also contain, in addition to LMO2, another member of the family, namely LMO4. Whereas the essential role of LMO2 in hematopoiesis and angiogenesis is well characterised, almost nothing is known about the function of its family member, LMO4, in hematopoiesis. LMO4 appears to be essential for definitive erythropoiesis, as morpholino studies in zebrafish show that definitive erythroid cells are not formed in its absence. It has been shown, in an erythroid cell line, that LMO4 level increases during differentiation and is recruited to the Ldb1 complex. The aim of this project was to study the molecular functions of this essential actor of the erythroid process in the murine erythroleukemia cell line (or MEL cells).

We first expressed a tagged version of LMO4 in MEL and used an efficient tagging methodology for the direct capture of protein complexes and their identification by mass spectrometry. 52 LMO4 partners were characterised. We also used the tag-LMO4 MEL cell line to identify the genome-wide binding sites of LMO4, using the newly developed ChIP-Seq method (Chromatin immunoprecipitation followed by high throughput sequencing). To detect changes in protein complex composition and their targeted binding sites during the dynamic process of erythropoiesis, both studies were performed before and after induction of MEL cells differentiation. In addition to Ldb1 (known LMO4 interactor), we selected four new LMO4 partners identified by mass spectrometry (focusing on key factors from each novel complex/group of proteins found, playing a role in transcription with reported hematopoietic phenotype in mouse and/or in our tested morpholino-injected zebrafish) to perform another round of tagging and ChIP-Seq analysis in non-differentiated and differentiated MEL cells. Ongoing analyses (crossing genome-wide binding data of LMO4 and partners) will allow defining the composition of novel transcriptional complexes and identifying their respective target genes. Finally, we studied the effect of knock-downs of LMO4 and its protein partners (members of Ldb1 complex and four new partners) on gene expression by transient transfection of specific siRNA in non-differentiated and differentiated MEL cells. Micro-array analysis allowed to identify genes commonly mis-regulated by LMO4 and its interactors. The direct transcriptional functions of the different LMO4 complexes during erythropoiesis (transcriptional activator or repressor) as well as their biological function (related to the function of their target genes) are being analysed by crossing microarray to ChIP-Seq data.

We found new potential molecular mechanisms of transcription regulating the lineage-specific genetic programs associated with erythroid differentiation. This study gives, for the first time some evidence of a repressive function associated with LMO4 (and possibly Ldb1) in erythroid cells. The results of this work will be part of the ongoing modelling of erythroid TF networks.