Methylation-dependent SUMOylation of the architectural transcription factor HMGA2.

High mobility group A2 (HMGA2) is a chromatin-associated protein involved in the regulation of stem cell function, embryogenesis and cancer development. Although the protein does not contain a consensus SUMOylation site, it is shown to be SUMOylated. In this study, we demonstrate that the first lysine residue in the reported K66KAE SUMOylation motif in HMGA2 can be methylated in vitro and in vivo by the Set7/9 methyltransferase. By editing the lysine, the increased hydrophobicity of the resulting 6-N-methyl-lysine transforms the sequence into a consensus SUMO motif. This post-translational editing dramatically increases the subsequent SUMOylation of this site. Furthermore, similar putative methylation-dependent SUMO motifs are found in a number of other chromatin factors, and we confirm methylation-dependent SUMOylation of a site in one such protein, the Polyhomeotic complex 1 homolog (PHC1). Together, these results suggest that crosstalk between methylation and SUMOylation is a general mode for regulation of chromatin function.

Reexpression of LSAMP inhibits tumor growth in a preclinical osteosarcoma model.

BACKGROUND: Osteosarcomas are the most common primary malignant tumors of bone, showing complex chromosomal rearrangements with multiple gains and losses. A frequent deletion within the chromosomal region 3q13.31 has been identified by us and others, and is mainly reported to be present in osteosarcomas. The purpose of the study was to further characterize the frequency and the extent of the deletion in an extended panel of osteosarcoma samples, and the expression level of the affected genes within the region. We have identified LSAMP as the target gene for the deletion, and have studied the functional implications of LSAMP-reexpression. METHODS: LSAMP copy number, expression level and protein level were investigated by quantitative PCR and western blotting in an osteosarcoma panel. The expression of LSAMP was restored in an osteosarcoma cell line, and differences in proliferation rate, tumor formation, gene expression, migration rate, differentiation capabilities, cell cycle distribution and apoptosis were investigated by metabolic dyes, tumor formation in vivo, gene expression profiling, time-lapse photography, differentiation techniques and flow cytometry, respectively. RESULTS: We found reduced copy number of LSAMP in 45/76 osteosarcoma samples, reduced expression level in 25/42 samples and protein expression in 9/42 samples. By restoring the expression of LSAMP in a cell line with a homozygous deletion of the gene, the proliferation rate in vitro was significantly reduced and tumor growth in vivo was significantly delayed. In response to reexpression of LSAMP, mRNA expression profiling revealed consistent upregulation of the genes hairy and enhancer of split 1 (HES1), cancer/testis antigen 2 (CTAG2) and kruppel-like factor 10 (KLF10). CONCLUSIONS: The high frequency and the specificity of the deletion indicate that it is important for the development of osteosarcomas. The deletion targets the tumor suppressor LSAMP, and based on the functional evidence, the tumor suppressor function of LSAMP is most likely exerted by reducing the proliferation rate of the tumor cells, possibly by indirectly upregulating one or more of the genes HES1, CTAG2 or KLF10. To our knowledge, this study describes novel functions of LSAMP, a first step to understanding the functional role of this specific deletion in osteosarcomas.

Identification of target genes for wild type and truncated HMGA2 in mesenchymal stem-like cells.

BACKGROUND: The HMGA2 gene, coding for an architectural transcription factor involved in mesenchymal embryogenesis, is frequently deranged by translocation and/or amplification in mesenchymal tumours, generally leading to over-expression of shortened transcripts and a truncated protein. METHODS: To identify pathways that are affected by sarcoma-associated variants of HMGA2, we have over-expressed wild type and truncated HMGA2 protein in an immortalized mesenchymal stem-like cell (MSC) line, and investigated the localisation of these proteins and their effects on differentiation and gene expression patterns. RESULTS: Over-expression of both transgenes blocked adipogenic differentiation of these cells, and microarray analysis revealed clear changes in gene expression patterns, more pronounced for the truncated protein. Most of the genes that showed altered expression in the HMGA2-overexpressing cells fell into the group of NF-kappaB-target genes, suggesting a central role for HMGA2 in this pathway. Of particular interest was the pronounced up-regulation of SSX1, already implicated in mesenchymal oncogenesis and stem cell functions, only in cells expressing the truncated protein. Furthermore, over-expression of both HMGA2 forms was associated with a strong repression of the epithelial marker CD24, consistent with the reported low level of CD24 in cancer stem cells. CONCLUSIONS: We conclude that the c-terminal part of HMGA2 has important functions at least in mesenchymal cells, and the changes in gene expression resulting from overexpressing a protein lacking this domain may add to the malignant potential of sarcomas.

The Drosophila G9a gene encodes a multi-catalytic histone methyltransferase required for normal development.

Mammalian G9a is a histone H3 Lys-9 (H3-K9) methyltransferase localized in euchromatin and acts as a co-regulator for specific transcription factors. G9a is required for proper development in mammals as g9a-/g9a- mice show growth retardation and early lethality. Here we describe the cloning, the biochemical and genetical analyses of the Drosophila homolog dG9a. We show that dG9a shares the structural organization of mammalian G9a, and that it is a multi-catalytic histone methyltransferase with specificity not only for lysines 9 and 27 on H3 but also for H4. Surprisingly, it is not the H4-K20 residue that is the target for this methylation. Spatiotemporal expression analyses reveal that dG9a is abundantly expressed in the gonads of both sexes, with no detectable expression in gonadectomized adults. In addition we find a low but clearly observable level of dG9a transcript in developing embryos, larvae and pupae. Genetic and RNAi experiments reveal that dG9a is involved in ecdysone regulatory pathways.

The Drosophila SET domain encoding gene dEset is essential for proper development.

We have identified dEset, the fly homolog of human SETDB1 and mouse ESET histone lysine methyltransferases (HKMTases) that methylates the lysine 9 residue of histone 3 (H3-K9) and negatively regulates transcription of target genes. By using spatio-temporal RNA interference we show that dEset is required at several stages of development coinciding with ecdysone pulses, possibly as a repressor of transcription of target genes. Several interacting partners, for example USP, spire, and cut up were identified in a yeast two-hybrid screen. The spatio-temporal expression profiles of dEset and its potential partners suggest that they may act together or even in a larger complex.

Drosophila dSet2 functions in H3-K36 methylation and is required for development.

Lysine methylation has important functions in biological processes that range from heterochromatin formation to transcription regulation. Here, we demonstrate that Drosophila dSet2 encodes a developmentally essential histone H3 lysine 36 (K36) methyltransferase. Larvae subjected to RNA interference-mediated (RNAi) suppression of dSet2 lack dSet2 expression and H3-K36 methylation, indicating that dSet2 is the sole enzyme responsible for this modification in Drosophila melanogaster. dSet2 RNAi blocks puparium formation and adult development, and causes partial (blister) separation of the dorsal and ventral wing epithelia, defects suggesting a failure of the ecdysone-controlled genetic program. A transheterozygous EcR null mutation/dSet2 RNAi combination produces a complete (balloon) separation of the wing surfaces, revealing a genetic interaction between EcR and dSet2. Using immunoprecipitation, we demonstrate that dSet2 associates with the hyperphosphorylated form of RNA polymerase II (RNAPII).

The acetyltransferase activity of Drosophila CBP is dispensable for regulation of the Dpp pathway in the early embryo.

The CBP protein is a transcriptional co-activator and histone acetyltransferase. Reduced expression of Drosophila CBP (dCBP) in the early embryo specifically impairs signaling by the TGF-beta molecules Dpp and Screw (Scw). This occurs by a failure to activate transcription of the tolloid (tld) gene, which codes for a protease that generates active Dpp and Scw ligands. We show that dCBP directly regulates this gene by binding to the tld enhancer, and that tld expression can be partially rescued with a dCBP transgene. At a slightly later stage of development, Dpp/Scw signaling recovers in mutant embryos, but is unable to turn on expression of the Dpp/Scw-target gene rhomboid (rho). Interestingly, an acetyltransferase (AT)-defective dCBP transgene rescued tld and rho gene expression to an extent comparable to the wild-type transgene, whereas a transgene containing a 130 amino acid deletion rescued tld but not late rho expression. A tracheal phenotype caused by the reduced dCBP levels was also rescued more efficiently with the wild-type dCBP transgene than with this mutant transgene. Our results indicate that separate parts of the dCBP protein are required on different promoters, and that the AT activity of dCBP is dispensable for certain aspects of Dpp signaling. We discuss the similarity of these results to the role of p300/CBP in TGF-beta signaling in the mouse.

The CBP coactivator functions both upstream and downstream of Dpp/Screw signaling in the early Drosophila embryo.

The CBP histone acetyltransferase plays important roles in development and disease by acting as a transcriptional coregulator. A small reduction in the amount of Drosophila CBP (dCBP) leads to a specific loss of signaling by the TGF-beta molecules Dpp and Screw in the early embryo. We show that the expression of Screw itself, and that of two regulators of Dpp/Screw activity, Twisted-gastrulation and the Tolloid protease, is compromised in dCBP mutant embryos. This prevents Dpp/Screw from initiating a signal transduction event in the receiving cell. Smad proteins, the intracellular transducers of the signal, fail to become activated by phosphorylation in dCBP mutants, leading to diminished Dpp/Screw-target gene expression. At a slightly later stage of development, Dpp/Screw-signaling recovers in dCBP mutants, but without a restoration of Dpp/Screw-target gene expression. In this situation, dCBP acts downstream of Smad protein phosphorylation, presumably via direct interactions with the Drosophila Smad protein Mad. It appears that a major function of dCBP in the embryo is to regulate upstream components of the Dpp/Screw pathway by Smad-independent mechanisms, as well as acting as a Smad coactivator on downstream target genes. These results highlight the exceptional sensitivity of components in the TGF-beta signaling pathway to a decline in CBP concentration.