Foxd3 controls heterochromatin-mediated repression of repeat elements and 2-cell state transcription.

Repeat element transcription plays a vital role in early embryonic development. The expression of repeats such as MERVL characterises mouse embryos at the 2-cell stage and defines a 2-cell-like cell (2CLC) population in a mouse embryonic stem cell culture. Repeat element sequences contain binding sites for numerous transcription factors. We identify the forkhead domain transcription factor FOXD3 as a regulator of major satellite repeats and MERVL transcription in mouse embryonic stem cells. FOXD3 binds to and recruits the histone methyltransferase SUV39H1 to MERVL and major satellite repeats, consequentially repressing the transcription of these repeats by the establishment of the H3K9me3 heterochromatin modification. Notably, depletion of FOXD3 leads to the de-repression of MERVL and major satellite repeats as well as a subset of genes expressed in the 2-cell state, shifting the balance between the stem cell and 2-cell-like population in culture. Thus, FOXD3 acts as a negative regulator of repeat transcription, ascribing a novel function to this transcription factor.

Complete loss of H3K9 methylation dissolves mouse heterochromatin organization.

Histone H3 lysine 9 (H3K9) methylation is a central epigenetic modification that defines heterochromatin from unicellular to multicellular organisms. In mammalian cells, H3K9 methylation can be catalyzed by at least six distinct SET domain enzymes: Suv39h1/Suv39h2, Eset1/Eset2 and G9a/Glp. We used mouse embryonic fibroblasts (MEFs) with a conditional mutation for Eset1 and introduced progressive deletions for the other SET domain genes by CRISPR/Cas9 technology. Compound mutant MEFs for all six SET domain lysine methyltransferase (KMT) genes lack all H3K9 methylation states, derepress nearly all families of repeat elements and display genomic instabilities. Strikingly, the 6KO H3K9 KMT MEF cells no longer maintain heterochromatin organization and have lost electron-dense heterochromatin. This is a compelling analysis of H3K9 methylation-deficient mammalian chromatin and reveals a definitive function for H3K9 methylation in protecting heterochromatin organization and genome integrity.

m6A RNA methylation of major satellite repeat transcripts facilitates chromatin association and RNA:DNA hybrid formation in mouse heterochromatin.

Heterochromatin has essential functions in maintaining chromosome structure, in protecting genome integrity and in stabilizing gene expression programs. Heterochromatin is often nucleated by underlying DNA repeat sequences, such as major satellite repeats (MSR) and long interspersed nuclear elements (LINE). In order to establish heterochromatin, MSR and LINE elements need to be transcriptionally competent and generate non-coding repeat RNA that remain chromatin associated. We explored whether these heterochromatic RNA, similar to DNA and histones, may be methylated, particularly for 5-methylcytosine (5mC) or methyl-6-adenosine (m6A). Our analysis in mouse ES cells identifies only background level of 5mC but significant enrichment for m6A on heterochromatic RNA. Moreover, MSR transcripts are a novel target for m6A RNA modification, and their m6A RNA enrichment is decreased in ES cells that are mutant for Mettl3 or Mettl14, which encode components of a central RNA methyltransferase complex. Importantly, MSR transcripts that are partially deficient in m6A RNA methylation display impaired chromatin association and have a reduced potential to form RNA:DNA hybrids. We propose that m6A modification of MSR RNA will enhance the functions of MSR repeat transcripts to stabilize mouse heterochromatin.

Major satellite repeat RNA stabilize heterochromatin retention of Suv39h enzymes by RNA-nucleosome association and RNA:DNA hybrid formation.

The Suv39h1 and Suv39h2 histone lysine methyltransferases are hallmark enzymes at mammalian heterochromatin. We show here that the mouse Suv39h2 enzyme differs from Suv39h1 by containing an N-terminal basic domain that facilitates retention at mitotic chromatin and provides an additional affinity for major satellite repeat RNA. To analyze an RNA-dependent interaction with chromatin, we purified native nucleosomes from mouse ES cells and detect that Suv39h1 and Suv39h2 exclusively associate with poly-nucleosomes. This association was attenuated upon RNaseH incubation and entirely lost upon RNaseA digestion of native chromatin. Major satellite repeat transcripts remain chromatin-associated and have a secondary structure that favors RNA:DNA hybrid formation. Together, these data reveal an RNA-mediated mechanism for the stable chromatin interaction of the Suv39h KMT and suggest a function for major satellite non-coding RNA in the organization of an RNA-nucleosome scaffold as the underlying structure of mouse heterochromatin.

Suv39h-dependent H3K9me3 marks intact retrotransposons and silences LINE elements in mouse embryonic stem cells.

Heterochromatin is required to restrict aberrant expression of retrotransposons, but it remains poorly defined due to the underlying repeat-rich sequences. We dissected Suv39h-dependent histone H3 lysine 9 trimethylation (H3K9me3) by genome-wide ChIP sequencing in mouse embryonic stem cells (ESCs). Refined bioinformatic analyses of repeat subfamilies indicated selective accumulation of Suv39h-dependent H3K9me3 at interspersed repetitive elements that cover ∼5% of the ESC epigenome. The majority of the ∼8,150 intact long interspersed nuclear elements (LINEs) and endogenous retroviruses (ERVs), but only a minor fraction of the >1.8 million degenerate and truncated LINEs/ERVs, are enriched for Suv39h-dependent H3K9me3. Transcriptional repression of intact LINEs and ERVs is differentially regulated by Suv39h and other chromatin modifiers in ESCs but governed by DNA methylation in committed cells. These data provide a function for Suv39h-dependent H3K9me3 chromatin to specifically repress intact LINE elements in the ESC epigenome.

Two modes of transcriptional activation at native promoters by NF-kappaB p65.

The NF-kappaB family of transcription factors is crucial for the expression of multiple genes involved in cell survival, proliferation, differentiation, and inflammation. The molecular basis by which NF-kappaB activates endogenous promoters is largely unknown, but it seems likely that it should include the means to tailor transcriptional output to match the wide functional range of its target genes. To dissect NF-kappaB-driven transcription at native promoters, we disrupted the interaction between NF-kappaB p65 and the Mediator complex. We found that expression of many endogenous NF-kappaB target genes depends on direct contact between p65 and Mediator, and that this occurs through the Trap-80 subunit and the TA1 and TA2 regions of p65. Unexpectedly, however, a subset of p65-dependent genes are transcribed normally even when the interaction of p65 with Mediator is abolished. Moreover, a mutant form of p65 lacking all transcription activation domains previously identified in vitro can still activate such promoters in vivo. We found that without p65, native NF-kappaB target promoters cannot be bound by secondary transcription factors. Artificial recruitment of a secondary transcription factor was able to restore transcription of an otherwise NF-kappaB-dependent target gene in the absence of p65, showing that the control of promoter occupancy constitutes a second, independent mode of transcriptional activation by p65. This mode enables a subset of promoters to utilize a wide choice of transcription factors, with the potential to regulate their expression accordingly, whilst remaining dependent for their activation on NF-kappaB.

Six2 functions redundantly immediately downstream of Hoxa2.

Hox transcription factors control morphogenesis along the head-tail axis of bilaterians. Because their direct functional targets are still poorly understood in vertebrates, it remains unclear how the positional information encoded by Hox genes is translated into morphogenetic changes. Here, we conclusively demonstrate that Six2 is a direct downstream target of Hoxa2 in vivo and show that the ectopic expression of Six2, observed in the absence of Hoxa2, contributes to the Hoxa2 mouse mutant phenotype. We propose that Six2 acts to mediate Hoxa2 control over the insulin-like growth factor pathway during branchial arch development.

IGFBP5 is a potential regulator of craniofacial skeletogenesis.

Six known proteins bind to the insulin-like growth factor (IGF) with high affinity. Igfbp5 encodes one of these proteins, which regulates the activity of IGF, but also exerts IGF-independent actions. Using in situ hybridization to detect cells expressing Igfbp5 mRNA, we show that Igfbp5 is expressed in a dynamic pattern in the mouse embryonic craniofacial region. At early stages corresponding to the completion of neural crest migration, Igfbp5 mRNA was found predominantly in the epithelia, whereas when the craniofacial mesenchyme has begun its differentiation into skeletal tissue, Igfbp5-expressing cells surrounded the developing cartilages and bones. Embryos transgenically expressing Igfbp5 in restricted areas of the mesenchyme fated to form craniofacial bones revealed decreased ossification and even deletion of head bones areas. Transgenic expression of a mutant Igfbp5, encoding a product with reduced binding affinity for IGF, led to no skeletal abnormalities, suggesting that the observed negative effects on skeletal development rely on a mechanism that depends on binding to IGF.

Hoxa2 downregulates Six2 in the neural crest-derived mesenchyme.

The Hoxa2 transcription factor acts during development of the second branchial arch. As for most of the developmental processes controlled by Hox proteins, the mechanism by which Hoxa2 regulates the morphology of second branchial arch derivatives is unclear. We show that Six2, another transcription factor, is genetically downstream of Hoxa2. High levels of Six2 are observed in the Hoxa2 loss-of-function mutant. By using a transgenic approach to overexpress Six2 in the embryonic area controlled by Hoxa2, we observed a phenotype that is reminiscent of the Hoxa2 mutant phenotype. Furthermore, we demonstrate that Hoxa2 regulation of Six2 is confined to a 0.9 kb fragment of the Six2 promoter and that Hoxa2 binds to this promoter region. These results strongly suggest that Six2 is a direct target of Hoxa2.