Structure, Activity and Function of the Suv39h1 and Suv39h2 Protein Lysine Methyltransferases

SUV39H1 and SUV39H2 were the first protein lysine methyltransferases that were identified more than 20 years ago. Both enzymes introduce di- and trimethylation at histone H3 lysine 9 (H3K9) and have important roles in the maintenance of heterochromatin and gene repression. They consist of a catalytically active SET domain and a chromodomain, which binds H3K9me2/3 and has roles in enzyme targeting and regulation. The heterochromatic targeting of SUV39H enzymes is further enhanced by the interaction with HP1 proteins and repeat-associated RNA. SUV39H1 and SUV39H2 recognize an RKST motif with additional residues on both sides, mainly K4 in the case of SUV39H1 and G12 in the case of SUV39H2. Both SUV39H enzymes methylate different non-histone proteins including RAG2, DOT1L, SET8 and HupB in the case of SUV39H1 and LSD1 in the case of SUV39H2. Both enzymes are expressed in embryonic cells and have broad expression profiles in the adult body. SUV39H1 shows little tissue preference except thymus, while SUV39H2 is more highly expressed in the brain, testis and thymus. Both enzymes are connected to cancer, having oncogenic or tumor-suppressive roles depending on the tumor type. In addition, SUV39H2 has roles in the brain during early neurodevelopment.

The Drosophila HP1 family is associated with active gene expression across chromatin contexts

Drosophila Heterochromatin Protein 1a (HP1a) is essential for heterochromatin formation and is involved in transcriptional silencing. However, certain loci require HP1a to be transcribed. One model posits that HP1a acts as a transcriptional silencer within euchromatin while acting as an activator within heterochromatin. However, HP1a has been observed as an activator of a set of euchromatic genes. Therefore, it is not clear whether, or how, chromatin context informs the function of HP1 proteins. To understand the role of HP1 proteins in transcription, we examined the genome-wide binding profile of HP1a as well as two other Drosophila HP1 family members, HP1B and HP1C, to determine whether coordinated binding of these proteins is associated with specific transcriptional outcomes. We found that HP1 proteins share many of their endogenous binding targets. These genes are marked by active histone modifications and are expressed at higher levels than non-target genes in both heterochromatin and euchromatin. In addition, HP1 binding targets displayed increased RNA polymerase pausing compared to non-target genes. Specifically, co-localization of HP1B and HP1C was associated with the highest levels of polymerase pausing and gene expression. Analysis of HP1 null mutants suggests these proteins coordinate activity at transcription start sites to regulate transcription. Depletion of HP1B or HP1C alters expression of protein-coding genes bound by HP1 family members. Our data broaden understanding of the mechanism of transcriptional activation by HP1a and highlight the need to consider particular protein-protein interactions, rather than broader chromatin context, to predict impacts of HP1 at transcription start sites.

The Drosophila HP1 family is associated with active gene expression across chromatin contexts

Drosophila Heterochromatin Protein 1a (HP1a) is essential for heterochromatin formation and is involved in transcriptional silencing. However, certain loci require HP1a to be transcribed. One model posits that HP1a acts as a transcriptional silencer within euchromatin while acting as an activator within heterochromatin. However, HP1a has been observed as an activator of a set of euchromatic genes. Therefore, it is not clear whether, or how, chromatin context informs the function of HP1 proteins. To understand the role of HP1 proteins in transcription, we examined the genome-wide binding profile of HP1a as well as two other Drosophila HP1 family members, HP1B and HP1C, to determine whether coordinated binding of these proteins is associated with specific transcriptional outcomes. We found that HP1 proteins share many of their endogenous binding targets. These genes are marked by active histone modifications and are expressed at higher levels than non-target genes in both heterochromatin and euchromatin. In addition, HP1 binding targets displayed increased RNA polymerase pausing compared to non-target genes. Specifically, co-localization of HP1B and HP1C was associated with the highest levels of polymerase pausing and gene expression. Analysis of HP1 null mutants suggests these proteins coordinate activity at transcription start sites to regulate transcription. Depletion of HP1B or HP1C alters expression of protein-coding genes bound by HP1 family members. Our data broaden understanding of the mechanism of transcriptional activation by HP1a and highlight the need to consider particular protein-protein interactions, rather than broader chromatin context, to predict impacts of HP1 at transcription start sites.

HP1 proteins compact DNA into mechanically and positionally stable phase separated domains

In mammals, HP1-mediated heterochromatin forms positionally and mechanically stable genomic domains even though the component HP1 paralogs, HP1α, HP1β, and HP1γ, display rapid on-off dynamics. Here, we investigate whether phase-separation by HP1 proteins can explain these biological observations. Using bulk and single-molecule methods, we show that, within phase-separated HP1α-DNA condensates, HP1α acts as a dynamic liquid, while compacted DNA molecules are constrained in local territories. These condensates are resistant to large forces yet can be readily dissolved by HP1β. Finally, we find that differences in each HP1 paralog’s DNA compaction and phase-separation properties arise from their respective disordered regions. Our findings suggest a generalizable model for genome organization in which a pool of weakly bound proteins collectively capitalize on the polymer properties of DNA to produce self-organizing domains that are simultaneously resistant to large forces at the mesoscale and susceptible to competition at the molecular scale.

HP1 oligomerization compensates for low-affinity H3K9me recognition and provides a tunable mechanism for heterochromatin-specific localization

HP1 proteins bind with low affinity but high specificity to histone H3 lysine 9 methylation (H3K9me), forming transcriptionally inactive genomic compartments referred to as heterochromatin. How HP1 proteins traverse a complex and crowded chromatin landscape on the millisecond timescale to bind H3K9me chromatin remains paradoxical. Here, we apply single-molecule imaging to visualize an HP1 homolog, the fission yeast Swi6, in its native chromatin environment. By analyzing Swi6 motions, we identify individual mobility states that map to discrete biochemical intermediates. Using mutants that perturb Swi6 H3K9me recognition, oligomerization, or nucleic acid binding, we mechanistically parse how each biochemical property affects protein dynamics. While nucleic acid binding titrates Swi6 away from heterochromatin, as few as four tandem chromodomains are sufficient to restore H3K9me-dependent localization. Our studies propose a new paradigm where HP1 oligomerization stabilizes higher-order complexes to outcompete inhibitory nucleic acid and non-specific chromatin interactions, enabling high specificity H3K9me recognition in cells.

Structural mechanism of bivalent histone H3K4me3K9me3 recognition by the Spindlin1/C11orf84 complex in rRNA transcription activation

Spindlin1 is a unique multivalent epigenetic reader that facilitates ribosomal RNA transcription. In this study, we provide molecular and structural basis by which Spindlin1 acts in complex with C11orf84 to preferentially recognize non-canonical bivalent mark of trimethylated lysine 4 and lysine 9 present on the same histone H3 tail (H3K4me3K9me3). We demonstrate that C11orf84 binding stabilizes Spindlin1 and enhances its association with bivalent H3K4me3K9me3 mark. The functional analysis suggests that Spindlin1/C11orf84 complex can displace HP1 proteins from H3K4me3K9me3-enriched rDNA loci, thereby facilitating the conversion of these poised rDNA repeats from the repressed state to the active conformation, and the consequent recruitment of RNA Polymerase I for rRNA transcription. Our study uncovers a previously unappreciated mechanism of bivalent H3K4me3K9me3 recognition by Spindlin1/C11orf84 complex required for activation of rRNA transcription.

Mapping biochemical states associated with HP1 target recognition at sites of heterochromatin formation in living cells

ABSTRACTHP1 proteins bind with low affinity but high specificity to sites of histone H3 lysine 9 methylation (H3K9me) in the genome. HP1 binding to H3K9me compartmentalizes the genome into transcriptionally inactive heterochromatin and actively transcribed euchromatin. A characteristic feature of HP1 proteins is their dynamic and rapid turnover from sites of heterochromatin formation. How low-affinity H3K9me recognition enables HP1 proteins to rapidly and efficiently traverse a complex and crowded chromatin landscape on the millisecond timescale remains a paradox. Here, we visualize the real-time motions of an HP1 homolog, the fission yeast protein Swi6, in its native chromatin environment. By analyzing the motions of Swi6 with high spatial and temporal resolution, we map individual mobility states that are directly linked to discrete biochemical intermediates. We find that nucleic acid binding titrates Swi6 away from sites of heterochromatin formation, whereas increasing the valency of chromodomain-mediated H3K9me recognition promotes specific chromatin localization. We propose that Swi6 oligomerization compensates for low-affinity H3K9me recognition and provides a tunable mechanism for protein turnover. Our high-resolution biophysical studies provide a comprehensive framework for in vivo biochemistry and reveal how the competing biochemical properties of Swi6 affect H3K9me recognition in living cells.

HP1 proteins compact DNA into mechanically and positionally stable phase separated domains

In mammals HP1-mediated heterochromatin forms positionally and mechanically stable genomic domains even though the component HP1 paralogs, HP1α, HP1β, and HP1γ, display rapid on-off dynamics. Here we investigate whether phase-separation by HP1 proteins can explain these biological observations. Using bulk and single-molecule methods, we show that, within phase-separated HP1α-DNA condensates, HP1α acts as a dynamic liquid, while compacted DNA molecules are constrained in local territories. These condensates are resistant to large forces yet can be readily dissolved by HP1β. Finally, we find that differences in each HP1 paralog’s DNA compaction and phase-separation properties arise from their respective disordered regions. Our findings suggest a generalizable model for genome organization in which a pool of weakly bound proteins collectively capitalize on the polymer properties of DNA to produce self-organizing domains that are simultaneously resistant to large forces at the mesoscale and susceptible to competition at the molecular scale.

The Drosophila HP1 family is associated with active gene expression across chromatin contexts

ABSTRACTDrosophila Heterochromatin Protein 1a (HP1a) is essential for heterochromatin formation and is involved in transcriptional silencing. However, certain loci require HP1a to be transcribed properly. One model posits that HP1a acts as a transcriptional silencer within euchromatin while acting as an activator within heterochromatin. However, HP1a has been observed as an activator of a set of euchromatic genes. Therefore, it is not clear whether, or how, chromatin context informs the function of HP1 proteins. To understand the role of HP1 proteins in transcription, we examined the genome-wide binding profile of HP1a as well as two other Drosophila HP1 family members, HP1B and HP1C, to determine whether coordinated binding of these proteins is associated with specific transcriptional outcomes. We found that HP1 proteins share a majority of their endogenous binding targets. These genes are marked by active histone modifications and are expressed at higher levels than non-target genes in both heterochromatin and euchromatin. In addition, HP1 binding targets displayed increased RNA polymerase pausing compared to non-target genes. Specifically, co-localization of HP1B and HP1C was associated with the highest levels of polymerase pausing and gene expression. Analysis of HP1 null mutants suggests these proteins coordinate activity at transcription start sites (TSSs) to regulate transcription. Depletion of HP1B or HP1C alters expression of protein-coding genes bound by HP1 family members. Our data broadens understanding of the mechanism of transcriptional activation by HP1a and highlights the need to consider particular protein-protein interactions, rather than broader chromatin context, to predict impacts of HP1 at TSSs.