Remodelling regulates the heterochromatin of retrotransposons in mouse embryonic stem cells
Chromatin remodellers slide, assemble, eject or edit nucleosomes influencing chromatin structure, DNA accessibility and transcriptional programmes. The SNF2-like remodeller SMARCAD1 is conserved from yeast to human cells and is highly expressed in mouse embryonic stem cells. Upon its loss cells lose...
|PDF Full Text
No Tags, Be the first to tag this record!
|Chromatin remodellers slide, assemble, eject or edit nucleosomes influencing chromatin structure, DNA accessibility and transcriptional programmes. The SNF2-like remodeller SMARCAD1 is conserved from yeast to human cells and is highly expressed in mouse embryonic stem cells. Upon its loss cells lose their pluripotent phenotype but its function in ES cells is not known.
In order to understand the role of SMARCAD1 in mouse embryonic stem cells a robust ChIP-seq protocol was developed for the tagged and endogenous protein in wild-type and knockdown cell lines. SMARCAD1 binding sites were found predominantly at intergenic sites genome-wide, and overlap with repressive histone modifications. Among the chromatin-bound proteins discovered enriched with SMARCAD1 binding sites is KAP1 (KRAB-associated protein 1; Krüppel-associated box), a critical factor for the silencing of endogenous retroviral elements (ERVs) in mouse ES cell, the histone methyltransferase SETDB1 (SET Domain Bifurcated Histone Lysine Methyltransferase I), and the histone variant H3.3. Taken together, the discovered binding sites provide new understanding of SMARCAD1 function in ES cells and illustrate that in the open chromatin environment characteristic of the pluripotent state, SMARCAD1 is associated with transcriptional repression.
An unresolved issue is how SMARCAD1 associates with its binding sites without a DNA-binding domain and no known domains for recruitment by histone modifications. Candidate SMARCAD1 targets were investigated and it was discovered that recruitment is dependent on the interaction with KAP1 via the CUE1 (Coupling of Ubiquitin conjugation to ER degradation) domain of SMARCAD1. Sequential ChIP experiments revealed that KAP1 and SMARCAD1 are co-enriched on their shared targets.
Among the discovered binding sites of the remodeller SMARCAD1 are endogenous retroviral elements (ERVs) an abundant type of transposable element derived from viral integrations in the germline. ERV expression is tightly controlled by repressive factors as they pose a threat for genome stability. A series of knockdowns and ChIP-qPCR experiments was performed to understand the cooperation and underlying mechanisms of how these factors shape and control ERV heterochromatin. SMARCAD1 was identified as a crucial component; it is required for the association of the KAP1-SETDB1 silencing machinery over class I and II ERVs, and consequently the maintenance of the histone modifications H3K9me3 and H4K20me3. The histone variant H3.3 has a controversial role in ERV control and is similarly reduced upon the loss of SMARCAD1, suggesting SMARCAD1 may be involved in the turn-over of this variant.
In summary, heterochromatin organisation is perturbed when SMARCAD1 is not present at ERVs. The presence of H3K9me3 on the other hand is not necessary for SMARCAD1 binding as shown in a SETDB1 knock-down. The assembly of KAP1 and H3K9me3 is rescued by the ectopic expression of wild-type SMARCAD1 but not by an ATPase mutant. Hence, it is the catalytic activity of SMARCAD1 and chromatin remodelling that is required for the silencing of ERVs. The KAP1 interaction mutant had no effect on the association of KAP1 itself but could similarly not restore H3K9me3 emphasizing that SMARCAD1 is required for successful heterochromatin formation on ERVs and identifying chromatin remodelling as a key mechanism of ERV control in mouse ES cells.