Northwestern University Feinberg School of Medicine

Shilatifard Laboratory

Our Work

The regulation of gene transcription by RNA polymerase II is critical for development and differentiation, and its misregulation contributes to the pathogenesis of many cancers, including leukemia. The overall goal of our laboratory is to define the molecular mechanisms underlying leukemogenesis and to identify potential targets for therapy through detailed studies of proteins and protein complexes that regulate chromatin modifications, transcription initiation, and transcription elongation. For example, some of the homeotic (HOX) proteins are transcriptional regulators essential for normal hematopoiesis, and their misregulation is associated with hematological malignancies. Similarly, the mixed lineage leukemia (MLL) protein normally positively regulates multiple HOX genes, and several chromosomal rearrangements and translocations that result in the creation of MLL chimeric proteins cause various forms of leukemia.

Such malignancies presumably arise through changes in the hematopoiesis program, which results from the misregulation of the HOX genes by the MLL fusion proteins. Much of our understanding of the mechanisms—by which MLL, its target genes such as the HOX gene family, and its chimeras function—arises from studies on their close homologs in model organisms, such as yeast and Drosophila. Therefore, our laboratory takes full advantage of the powers of genetic, biochemistry, and cell biology in yeast, mammalian and Drosophila systems to decipher the roles of these factors during development and how their misregulation results in the pathogenesis of hematological malignancies.

Studies in Drosophila Systems

One fusion partner of MLL in acute myelogenous leukemia (AML) is the ELL protein. We show that human ELL functions as a transcription elongation factor. We have identified the Drosophila homolog of ELL and demonstrate it to be essential for development.  Drosophila ELL associates with elongating RNA polymerase II in vivo on chromosomes and is a regulator of the Notch signaling pathway.  This has suggested to us that human ELL might also participate in the same process.

In light of the identificaton of three ELL-related proteins in human, we show that they all share a conserved C- terminal domain, which is not required for the transcription elongation properties of the ELLs.  We show that in Drosophila, ELL's C-terminal domain is essential for development and the equivalent region of human ELL is critical for hematopoietic immortalization.  Therefore, defining the molecular role for ELL's C-terminal domain in Drosophila development is a major focus of our laboratory.

We have also taken advantage of RNAi technology in Drosophila to reduce the levels of the factors required for proper histone modifications to define their role in a living organism. We show that the components of the Rad6/Bre1, the Paf1 complex, and  other factors are required for histone methylation.  Furthermore, we have recently identified that the trithorax group gene in Drosophila, called little imaginal discs, encodes a histone trimethyl H3K4 demethylase. We are planning to follow through with our Drosophila studies to better define the molecular machinery involved in histone methylation and how the misregulation of their activities results in cellular immortalization.

Studies in Mammalian Model Systems

Chromosomal rearrangements resulting in alterations of gene expression are a major cause of hematological malignancies. Our goal is to advance the understanding of the biochemical and molecular mechanisms of rearrangement-based leukemia and to provide insights into how translocations affect cellular division by altering gene expression. Using mammalian model systems such as tissue culture and mouse genetics, we plan to explore the regulation of gene expression via the MLL gene and its translocation partners found in human leukemia. We are currently defining the molecular composition of the MLL complexes and how translocations alter its biochemical function and integrity, resulting in leukemic pathogenesis. We are also planning to define the mechanism of the targeting of the MLL complex and its histone methyltransferase activity to chromatin to determine its normal cellular functions and its mistargeting and disregulation in leukemogenesis.

Studies in Saccharomyces cerevisiae

The MLL gene was cloned over twenty years ago, yet the first molecular function for a MLL homolog was defined by our laboratory when we demonstrated that yeast Set1 (the MLL homolog) is a component of a large complex, Set1/COMPASS, which methylates histone H3 on its fourth lysine (H3K4) in the early transcribed regions of genes. Subsequent work by us and other labs demonstrated that the COMPASS family is highly conserved from yeast to Drosophila to mammalian cells functioning as H3K4 methylases. Although there is only one COMPASS in yeast, there are at least three COMPASS family members in Drosophila (dSet1, trx and Trr) and six COMPASS family members in human cells (Set1A-B, MLL1-4) with non-redundant functions.

To better define the molecular machinery required for proper histone H3K4 methylation by Set/1COMPASS in yeast and the COMPASS family in Drosophila and human cells, we devised a global functional proteomic screen, which we call Global Proteomic analysis in S. cerevisiae (GPS). With GPS, we tested extracts of each of the non-essential yeast gene deletion mutants in different mating types (~15,000 strains) for defects in modifications of histones by Western blotting. Employing an antibody specific to histone H3 methylated on its fourth lysine, GPS revealed that monoubiquitination of lysine 123 of histone H2B by Rad6 (the E2-conjugating enzyme) is required for histone methylation by COMPASS. We have taken advantage of GPS and have been able to put together a molecular pathway of factors required for proper histone methylation by COMPASS. This includes a role for Bre1 as Rad6’s E3 ligase, and a role for the Paf1 complex, the elongation factor, and the Bur1/Bur2 kinase for the proper regulation of H2B monoubiquitination. We now know that this H3K4 methylation machinery pathway is also highly conserved from yeast to Drosophila to human.

Given the extraordinary power of GPS in yeast, we are employing GPS to better define the molecular machinery required for COMPASS function and have also applied this screen to other posttranslational modifications of histones such as H3K36 methylation, H3K79 methylation and H3K56 acetylation. There is no doubt that the application of GPS will be extremely informative in defining such molecular pathways.