The Function and Molecular Mechanism of Asxl2 in the Mammalian Heart
thesisposted on 31.10.2013, 00:00 by Hsiao Lei Lai
Polycomb Group (PcG) and Trithorax Group (TrxG) genes were originally identified in Drosophila as repressors and activators of Hox genes, respectively (Schuettengruber et al., 2007). In PcG and TrxG mutants, the expression of Hox genes is mis-regulated, leading to homeotic transformation phenotypes of body segments (Lewis, 1978, 1982). In addition, PcG and TrxG proteins are found to be regulators of numerous genes involved in cell fate decision, stem cell identity and cancer (Aloia et al., 2013; Kohler and Hennig, 2010; Martinez et al., 2006; Sparmann and van Lohuizen, 2006); At the molecular level, PcG and TrxG proteins form multiple complexes that silence and activate chromatin, respectively (Pirrotta, 1998). A hallmark of PcG proteins activity is tri-methylation of histone H3 lysine 27 (H3K27me3), which is associated with a repressed transcriptional state (Cao et al., 2002). In contrast, tri-methylation of histone H3 lysine 4 (H3K27me4) is a product of TrxG activity, and is associated with transcriptional activation (Byrd and Shearn, 2003; Cao et al., 2002). A class of genes known as Enhancers of Trithorax and Polycomb (ETP) genetically interacts with both PcG and TrxG. Additional sex combs (Asx) of Drosophila is a member of the ETP group because Asx mutations enhance both PcG and TrxG mutant phenotypes (Milne et al., 1999). A recent study has shown that a newly characterized Drosophila PcG protein, Calypso, forms a complex with Asx (Scheuermann et al., 2010). This new PcG complex is named Polycomb Repressive Deubiquitinase (PR-DUB). Biochemical analysis has shown that recombinant Calypso interacts with the Asx-N terminus (aa2-337), and this interaction enhances deubiquitination of nucleosomal H2A by Calypso in vitro. The enzyme activity of Calypso is required for repression of the Hox gene Ubx in Drosophila. Three mammalian homologs of Asx, named Asx-like 1, 2, and 3, have been identified in mice and humans (Fisher et al., 2003; Katoh, 2003, 2004). Like Drosophila Asx and Calypso, mammalian ASXL1 and BAP1, the mammalian homolog of Calypso, form a stable PR-DUB complex in vitro (Scheuermann et al., 2010). We have generated an Asxl2 -/- mouse line to study Asxl in a mammalian system by taking advantage of a gene-trapped ES cell line from the Gene Trap Consortium [www.genetrap.org] (Baskind et al., 2009). The ES cell gene trap line contains a β-geo cassette with a polyadenylation site at the 3’ end which is integrated into the first intron of Asxl2 (Baskind et al., 2009). The truncated mRNA encodes a protein product that is missing all the conserved domains of ASXL2. Asxl2 is highly expressed in embryonic and adult hearts (Baskind et al., 2009). In collaboration with Dr. David Geenen’s laboratory, we have conducted a series of physiological and biochemical assays of Asxl2-/- mice in the B6/129 F1 background to assess the role of ASXL2 in heart function. Our results show that Asxl2-/- mice can survive to adulthood but gradually develop ventricle dysfunction. Microarray analysis shows more than 753 cardiac genes are mis-regulated in the absence of Asxl2. These results suggest that Asxl2 functions to maintain normal cardiac function and gene expression in postnatal stages. The axial skeletons of Asxl2-/- mice exhibit both posterior and anterior transformations, which are classic PcG/TrxG phenotypes, respectively (Baskind et al., 2009). This indicates that Asxl2 has ETP function. Furthermore, Asxl2 deficiency results in a reduction in the level of bulk H3K27me3, a repressive mark generated by the Polycomb Repressive Complex 2 (PRC2) (Baskind et al., 2009). Our goal is to determine the mechanism by which ASXL2 regulates H3K27me3 level. I show that ASXL2 interacts with PRC2 in the adult heart. The loss of Asxl2 results in loss of H3K27me3 as well as loss of PRC2 enrichment surrounding the de-repressed target promoters. The loss of PRC2 and H3K27me3 enrichment at these loci is not due to degradation of PRC2 core components or a failure of these components to form the PRC2 complex in Asxl2-/- hearts. In addition, we tested whether PR-DUB function is conserved in ASXL2. Our results show that ASXL2 interacts with BAP1 and is required for mono-deubiquitination of H2A at lysine 119 (H2AK119ub1) in the heart. In summary, I have made several contributions to further our understanding of ASXL2’s biological and functional mechanisms. Specifically, I have shown that (1) ASXL2 is required for normal ventricular function in adult heart; (2) ASXL2 is required for the homeostasis of two important histone marks, H3K27me3 and uH2A; (3) ASXL2 specifically affects the conversion of H3K27me2 to H3K27me3; (4) ASXL2 maintains the repression of select cardiac genes by facilitating the binding of PRC2 to target loci. Taken together, these data show that ASXL2 is a novel epigenetic regulator in the adult heart. In-depth studies of mouse ASXL2 will provide valuable insight on the diagnostic and/or therapeutic value of human ASXL2 in heart disease.