Clr4suv39h1 ubiquitination and non-coding rna mediate transcriptional silencing of heterochromatin via swi6 phase separation

ABSTRACT Transcriptional silencing by RNAi paradoxically relies on transcription, but how the transition from transcription to silencing is achieved has remained unclear. The Cryptic Loci

Regulator complex (CLRC) in _Schizosaccharomyces pombe_ is a cullin-ring E3 ligase required for silencing that is recruited by RNAi. We found that the E2 ubiquitin conjugating enzyme Ubc4

interacts with CLRC and mono-ubiquitinates the histone H3K9 methyltransferase Clr4SUV39H1, promoting the transition from co-transcriptional gene silencing (H3K9me2) to transcriptional gene

silencing (H3K9me3). Ubiquitination of Clr4 occurs in an intrinsically disordered region (Clr4IDR), which undergoes liquid droplet formation in vitro, along with Swi6HP1 the effector of

transcriptional gene silencing. Our data suggests that phase separation is exquisitely sensitive to non-coding RNA (ncRNA) which promotes self-association of Clr4, chromatin association, and

di-, but not tri- methylation instead. Ubc4-CLRC also targets the transcriptional co-activator Bdf2BRD4, down-regulating centromeric transcription and small RNA (sRNA) production. The

deubiquitinase Ubp3 counteracts both activities. SIMILAR CONTENT BEING VIEWED BY OTHERS RIXOSOMAL RNA DEGRADATION CONTRIBUTES TO SILENCING OF POLYCOMB TARGET GENES Article Open access 30

March 2022 A PROTEIN ASSEMBLY MEDIATES _XIST_ LOCALIZATION AND GENE SILENCING Article 09 September 2020 THE YEAST ISW1B ATP-DEPENDENT CHROMATIN REMODELER IS CRITICAL FOR NUCLEOSOME SPACING

AND DINUCLEOSOME RESOLUTION Article Open access 18 February 2021 INTRODUCTION Heterochromatin has crucial roles in genome stability by silencing repetitive DNA and by suppressing

recombination1,2,3,4,5,6,7. In the fission yeast _Schizosaccharomyces pombe_, the H3K9 methyltransferase Clr4, the homolog of mammalian SUV39H1 and SUVH39H2, methylates histone H3 lysine 9

(H3K9) to maintain centromeric heterochromatin during the cell cycle8,9,10. Centromeric non-coding RNA (ncRNA) is transcribed by RNA Pol II during S phase, and is processed to make

centromeric sRNA by the concerted action of the RNAi transcriptional silencing complex (RITS), the RNA-directed RNA Polymerase Complex (RDRC) and Dicer (Dcr1)7,11,12,13,14,15,16.

Di-methylation of H3K9 by Clr4, guided in part by sRNA, mediates co-transcriptional gene silencing (CTGS) of heterochromatin, which remains permissive to transcription during S phase. The

subsequent tri-methylation of H3K9 by Clr4 and strong binding of Heterochromatin Protein 1 (HP1) homologs Swi6 and Chp2, suppresses RNA Pol II-dependent ncRNA transcription via

transcriptional gene silencing (TGS), which is essential for the epigenetic inheritance of heterochromatin1,17,18,19. However, the mechanism that promotes the transition from RNAi-dependent

CTGS to TGS is unclear. Ubiquitination can regulate enzyme activity by adding a single ubiquitin to its substrate (mono-ubiquitination), or target protein degradation via the 26S proteasome

by the addition of a ubiquitin chain (poly-ubiquitination). Ubiquitination requires ubiquitin, an E1 ubiquitin activating enzyme, an E2 ubiquitin conjugating enzyme and an E3 ubiquitin

ligase20,21,22,23. E2 enzymes interact with E1 and E3 enzymes and have essential roles in determination of ubiquitin linkage specificity and ubiquitin chain length24. The ubiquitination

complex Cul4-Rik1Raf1Raf2 (CLRC) of _S. pombe_ is composed of cullin Cul4, the RING-box protein Rbx1, the DNA damage binding protein 1 (DDB1) homolog Recombination in K 1 (Rik1) and two

substrate receptors, Raf1 and Raf2, as well as the eponymous Clr4, which is only weakly associated in vitro25,26,27,28,29,30. Cul4 is a highly conserved E3 ligase and interacts with

Rik1/Raf1/Raf2 for substrate recognition. All subunits of CLRC are essential for heterochromatin formation but its E2 enzyme and ubiquitination targets are not well characterized25,26,27,28.

Liquid-liquid phase separation (LLPS) is the mechanism by which biomolecular condensates like cellular granules, including membraneless organelles such as the nucleolus, P granules, and

stress granules form without a lipid bilayer. Multivalent macromolecular interactions, especially between proteins harboring intrinsically disordered regions (IDRs) and nucleic acids, drive

formation of phase-separated liquid droplets. These biomolecular condensates have higher protein density and reduced internal diffusion, which contributes to a myriad of cellular functions,

including heterochromatin formation31,32,33,34,35,36,37,38. Nucleosomes are post-translationally modified and many chromatin-related factors, such as heterochromatin protein 1 (HP1) and

bromo-domain containing proteins (e.g. BRD4), which recognize these histone modifications, have an intrinsic propensity for LLPS and form phase-separated liquid droplets in vitro and in

cells39,40,41,42,43. Here we set out to identify how the E3 ubiquitin ligase complex CLRC mediates the transition from RNAi-dependent CTGS to TGS. We found that Ubc4-CLRC mono-ubiquitinates

Clr4 in the IDR, reducing its chromatin- and ncRNA-binding activities. Importantly, we found that Clr4 undergoes phase-separated liquid droplet formation along with Swi6 through its IDR.

However, liquid droplet formation of Clr4 is blocked by the presence of ncRNA, which instead promotes self-association of Clr4, chromatin association of Clr4 and di-methylation of H3K9 but

not tri-methylation of H3K9. We also found that the Ubc4-CLRC complex poly-ubiquitinates Bdf2 and inhibits the recruitment of Epe1 and Bdf2 to centromeric heterochromatin, reducing ncRNA

transcription and sRNA production via RNAi. Both H3K9 tri-methylation by mono-ubiquitinated Clr4 and reduced sRNA production by targeted Epe1-Bdf2 are crucial for the transition between CTGS

to TGS and spreading of heterochromatin, which can be reversed by overexpression of deubiquitinase Ubp3. These observations implicate phase separation as a critical step in the transition

from RNAi to transcriptional gene silencing during heterochromatin formation. RESULTS UBC4 AND CUL4 REGULATE HETEROCHROMATIN SILENCING We previously showed that a specific mutation in the E2

ubiquitin conjugating enzyme Ubc4 (_ubc4-G48D_) reduces heterochromatic silencing at centromeres, at the mating-type locus and at telomeres (Fig. 1a)44. Overexpression of Ubc4 also causes

chromosome mis-segregation45. We found that the _ubc4-G48D_ (_ubc4-1_) mutation results in Ubc4 protein instability leading to an overall lower Ubc4 protein level (Fig. 1b; Supplementary

Fig. 1c). Replacing _ubc4_ with its mouse homolog _UBE2D2_ fully complemented _ubc4-1_, while _mUBE2D2-G48D_ (_mUBE2D2-1_) had similar silencing defects to _ubc4-1_, showing that the

function of Ubc4 is likely conserved throughout evolution (Fig. 1a; Supplementary Fig. 1a, b). We also isolated a spontaneous second-site suppressor S137T, which re-stabilizes Ubc4-1 protein

levels, thus restoring silencing (Fig. 1b–d). The E3 ubiquitin ligase Cul4 is an essential subunit of the CLRC, and _cul4-GFP_ (_cul4-1_) is a hypomorphic mutant with reduced

heterochromatic silencing27. _ubc4-1_ and the _cul4-1_ mutants have very similar phenotypes, including loss of silencing of _ade6__+_ and _ura4__+_ reporter genes in centromeric

heterochromatin, as well as increased binding of Rpb1, the main subunit of RNA polymerase II, to heterochromatin (Figs. 1a, 2a) and increased production of siRNA (Figs. 1e, 2f, 5a).

Interestingly, while H3K9me3 was reduced from both _ubc4-1_ and _cul4-1_ mutants at all heterochromatic loci tested, H3K9me2 was increased at the centromeric _dg_ and _dh_ repeats (Figs. 1f,

2a; Supplementary Fig. 2a). Increased H3K9me2 was also observed at the centromeric _ade6__+_ reporter gene, which accumulates small RNA, but not at the _ura4__+_ reporter gene, which does

not (Fig. 1f; Supplementary Fig. 10a). The Swi6HP1 protein, which coats and maintains heterochromatin, displayed intermediate binding levels at centromeres in _ubc4-1_ and _cul4-1_ mutants

presumably because it recognizes both H3K9me2 and H3K9me3 (Supplementary Fig. 2b)18,46. In contrast, H3K9me2/3 and Swi6 were completely lost from both _ubc4-1_ and _cul4-1_ mutants at

telomeres or mating-type loci where RNAi-dependent sRNA production is limited (Supplementary Fig. 2c, d). To determine whether RNAi is important for maintaining H3K9me2 in _ubc4-1_ mutant

cells, we introduced the _ubc4-1_ mutation in _dcr1Δ_ mutant cells and found that the _dcr1Δ ubc4-1_ double-mutant lost both H3K9me2 and H3K9me3 and was hyper-sensitive to the microtubule

destabilizing agent thiabendazole (TBZ), showing that centromeric heterochromatin was completely disrupted similarly to the _clr4_ deletion mutant (Fig. 2a; Supplementary Fig. 2e). The more

severe alleles _ubc4-3xHA_ and _cul4Δ_ abolished H3K9me2 and H3K9me3 altogether indicating that _ubc4-1_ and _cul4-1_ are hypomorphic and they might have other targets important for

heterochromatin silencing (Supplementary Fig. 2f, g). In addition to their similar phenotypes, we found that Ubc4 and Cul4 physically interact and form a stable complex by

co-immunoprecipitation (Co-IP) (Supplementary Fig. 3a). Ubc4 also forms a complex with Raf1 and Raf2, two substrate receptors of CLRC (Supplementary Fig. 3b) that mediate recruitment by

siRNA-to-chromatin 1 (Stc1) and the RNA-induced transcriptional gene silencing (RITS) complex. The interaction between Ubc4-Cul4 and Ubc4-Raf2 was independent of nuclease treatment

(Supplementary Fig. 3c, d), which indicates that Ubc4 and Cul4 directly form the E2/E3 complex by which CLRC targets substrates important for H3K9 methylation, in a RNAi-dependent manner.

MONO-UBIQUITINATION OF CLR4 BY UBC4-CLRC IS REQUIRED FOR H3K9ME2 TO H3K9ME3 TRANSITION Next, we set out to identify the substrates of this novel E2/E3 complex, Ubc4-CLRC, for the regulation

of H3K9 methylation. Interestingly, Clr4 co-purifies with the CLRC in vivo, but does not behave like a structural component of CLRC in vitro25,26,27,28. This led us to hypothesize that Clr4

is itself a target of the Ubc4-CLRC. Indeed, another Clr4 human homolog, SETDB1, is mono-ubiquitinated at its SET-Insertion domain and this modification is important for its H3K9

methyltransferase activity47,48. To test whether Clr4 is a substrate of Ubc4-CLRC, we purified 3xFlag-Clr4 protein from _S. pombe_ cells and detected a single higher molecular weight form

that reacted with ubiquitin antibodies in a Ubc4-dependent fashion (Fig. 2b; Supplementary Fig. 4a, b). A pulldown of ubiquitinated proteins using Ubiquitination Affinity Beads (UBA)

revealed a similar higher molecular weight Clr4 which is also dependent on Ubc4 (Supplementary Fig. 4c). In addition to Ubc4, the presence of this higher molecular weight Clr4 depends on

Cul4, Rik1, Raf1 and Raf2, the other components of CLRC, but not on the proteasome inhibitor MG-132 (Supplementary Fig. 4d, e). These results strongly suggest that Ubc4, together with CLRC,

mono-ubiquitinates Clr4 in vivo. We were unable to recover sufficient peptides in vivo for the identification of Clr4 ubiquitination sites by Mass Spectrometry because of the very low

abundance of Clr4 in _S. pombe_ cells49, so instead we ubiquitinated 6xHis-Clr4 recombinant protein in vitro. Clr4 was mono-ubiquitinated efficiently by UBE2D3, the human homolog of Ubc4,

independently of an E3 ligase complex (Fig. 3d; Supplementary Fig. 5a–c), similarly to SETDB148,50. This approach uncovered 9 possible ubiquitination sites, which mostly reside in the

intrinsically disordered region (IDR) between the chromodomain (CD) and the PreSET domain (Fig. 2c and Supplementary Fig. 5a–d) which is involved in nucleosome binding51. To identify which

of these represent functional sites in vivo, we partitioned the 36 lysines of Clr4 into 6 subgroups, which we then systematically mutated into arginines (R). We also included a mutant of all

36 lysines (_AllKtoR_). The all-lysine mutant (_AllKtoR_) and the second block mutant (_2-8KtoR_), corresponding to the middle of the IDR, completely abolished Clr4 ubiquitination in vivo

(Fig. 2c, d; Supplementary Fig. 5d, e), consistent with the results of in vitro ubiquitination and Mass Spectrometry. By this means, we could narrow down the possible sites to 4 lysines

(_clr4-2-4KtoR_; K109, K110, K113, and K114) (Fig. 2c, d; Supplementary Fig. 5f–i). These lysines are organized in tandem and are therefore very likely to compensate for each other. Using

ChIP-seq, we found that _clr4-AllKtoR_ lost both H3K9me2 and H3K9me3, presumably by reduced enzymatic activity52. _clr4-2-8KtoR_ and _clr4-2-4KtoR_ mutants had the same increase of H3K9me2

and decrease of H3K9me3 seen in _ubc4-1_ and _cul4-1_ hypomorphic mutants but loss of silencing of centromeric _ade6__+_ and _ura4__+_ reporter genes in _clr4-2-4KtoR_ mutant was weaker than

_clr4-AllKtoR_ and _cul4-1_ mutants (Fig. 2e; Supplementary Fig. 6a–c). In _ubc4-1_ and _cul4-1_ mutants, we observed more than 2-fold increase of centromeric sRNA, as previously observed

for _cul4-1_, but _clr4-2-8KtoR_ and _clr4-2-4KtoR_ mutants did not change centromeric sRNA production (Fig. 2f)53. These results indicate that Clr4 mono-ubiquitination by Ubc4-CLRC complex

mediates the transition from H3K9me2 to H3K9me3 (Fig. 2g). CLR4 MONO-UBIQUITINATION INDUCES ITS DISSOCIATION FROM CHROMATIN AND FROM CENTROMERIC NCRNA Remarkably, the lysines in Clr4-2-8

co-incided precisely with two adjacent patches in the IDR that were previously shown to bind to the nucleosome core, independently of H3K9me3 (KKVFS and KRQSRK)51. Consistently, Clr4 binding

to chromatin was increased in _ubc4-1_ and _cul4-1_ mutants, especially at _dh_ centromeric repeats and at the _ade6__+_ reporter gene, where sRNAs are over-produced (Fig. 3a; Supplementary

Figs. 7a, b, 10a). This increased binding of Clr4 to chromatin is not dependent on its chromodomain because there was no change in combination with a chromodomain mutant (_clr4-W31G

ubc4-_1, Fig. 3a; Supplementary Fig. 7a)18,53. Note that the _clr4-2-4KtoR_ mutant has much weaker binding than _ubc4-1_ or _cul4-1_ mutants presumably because _clr4-2-4KtoR_ mutation does

not affect ncRNA transcription and sRNA production (Figs. 2f, 3a; Supplementary Fig. 6c). Mutating all lysines (_AllKtoR_) resulted in an even stronger affinity of Clr4 to chromatin (Fig. 

3a; Supplementary Fig. 7a). In contrast, Flag-Clr4 binding to centromeric heterochromatin decreased when H3K9me2/3 was completely abolished by _rik1Δ_, _raf1Δ_ and _raf2Δ_ (Fig. 3b)26. Clr4

also binds centromeric ncRNA54 and this activity was strongly increased in _ubc4-1_ and _cul4-1_ mutants and slightly but significantly increased in the _clr4-2-4KtoR_ mutant (Fig. 3c;

Supplementary Fig. 6d). These results indicate that both chromatin and ncRNA binding activities of Clr4 are down regulated by Clr4 mono-ubiquitination. To further explore the effect of Clr4

mono-ubiquitination on chromatin binding, we carried out nucleosome binding assays using in vitro ubiquitinated 6xHis-Clr4 recombinant protein and 3xFlag-Clr4 purified from _S. pombe_ cells,

with nucleosomes with either unmodified histone H3 (me0) or H3K9me3 (me3) tethered to streptavidin beads. In contrast to unmodified Clr4 protein, which had strong binding affinity to the

nucleosome, ubiquitinated Clr4 did not bind to nucleosomes (Fig. 3d; Supplementary Fig. 7c). These results suggest that Clr4 mono-ubiquitination promotes dissociation of Clr4 from chromatin

in agreement with the in vivo Clr4 ChIP-seq data. Next, we tested whether Ubc4-Cul4 ubiquitination activities are important in vivo in the context of a minimal silencing system, relying on

tethering Clr4 to a specific chromatin locus. To do this, we generated a reporter system in which _clr4-ΔCD_ lacking the chromodomain is fused to Gal4 DNA binding domain (GBD) and tethered

to 5 Gal4-binding sites (5xUAS) upstream of _ade6__+_, inserted at the endogenous _ura4_ locus, and with additional intact WT _clr4__+_ available55. Artificial tethering of GBD-_Clr4-ΔCD_

resulted in silencing of the _ade6__+_ reporter gene as indicated by red colony formation on low-adenine medium and accumulation of H3K9me2/3 by ChIP-qPCR (Fig. 3e), in accordance with

previous studies55,56,57,58,59,60. In contrast, _ade6__+_ silencing was greatly decreased by introduction of _ubc4-1_ and _cul4-1_ mutations, demonstrating that the ubiquitination activity

of Ubc4-CLRC is required for spreading of silencing into the _ade6__+_ reporter (Fig. 3e). PHASE-SEPARATED LIQUID DROPLET FORMATION OF CLR4 IS INHIBITED BY NON-CODING RNA IN VITRO Clr4 and

its mouse homolog, mSUV39H1, have an intrinsically disordered region (IDR), also known as a low-complexity domain (LCD), between its chromodomain (CD) and PreSET domain (Supplementary Fig. 

8a, b)51, and IDR/LCD are typically responsible for phase separation31,33,35,37,38,61. We found that recombinant GFP-Clr4Full and GFP-Clr4IDR (51-220 amino acids of Clr4) proteins readily

form liquid droplets in vitro in a salt-dependent manner even without addition of crowding agents (Fig. 4a; Supplementary Fig. 8c)62. The formation of liquid droplets by GFP-Clr4IDR was

reduced by adding 1,6-hexanediol, which dissolves phase-separated liquid droplets by disrupting the hydrophobic interactions of protein-protein or protein-RNA (Supplementary Fig. 8d)37,63.

Furthermore, photobleaching of inside part of liquid droplets of GFP-Clr4IDR resulted in slow but steady recovery of fluorescence indicative of dynamic exchange of GFP-Clr4IDR protein inside

of liquid droplets (Fig. 4b and Supplementary Fig. 8e)31,39,43. These data indicate an involvement of phase separation for liquid droplet formation of Clr4. As the oligomerization of HP1 is

important to form phase-separated liquid droplets as well as during chromatin compaction, and as SUV39H1 is also able to tether together to form dimers40,42,64,65,66, we reasoned that Clr4

may self-associate to form phase-separated liquid droplets. To do this, we mixed recombinant GFP-Clr4 and mCherry-Clr4 and found that these proteins readily associate together to form

phase-separated liquid droplets, supporting the idea of Clr4 self-association (Fig. 4c). CTGS during G1/S phase of the cell cycle is accompanied by transcription of ncRNAs and accumulation

of sRNAs and the interaction of RNA-binding proteins with RNA are important for LLPS17,19,31,34. To investigate the possible role of centromeric ncRNA in Clr4 phase-separated liquid droplet

formation, we added 120 nt, synthetic centromeric ncRNA to GFP-Clr4/mCherry-Clr4. Surprisingly, ncRNA inhibited phase-separated liquid droplet formation of Clr4 and Clr4IDR in a dosage

dependent manner (Fig. 4c and Supplementary Fig. 8c), reminiscent of other IDR-containing proteins that interact with RNA67,68,69. We next examined Clr4 self-association via

immunoprecipitation using recombinant Flag-Clr4 and Myc-Clr4 proteins in the presence or absence of ncRNA. The weak interaction of Flag-Clr4 with Myc-Clr4 in the absence of ncRNA is

consistent with multivalent weak interactions that form phase-separated liquid droplets (Fig. 4d). In contrast, by addition of ncRNA, Clr4 self-association was greatly increased with

increasing concentration of ncRNA, indicating that Clr4 self-association induced by ncRNA actually inhibits its phase-separated liquid droplet formation (Fig. 4c, d). Furthermore,

mono-ubiquitinated Clr4 did not bind to the centromeric ncRNA compared to unmodified Clr4 protein from ncRNA binding assays using 3xFlag-Clr4 purified from _S. pombe_ cells (Supplementary

Fig. 8f). Thus, the RNA binding activity of Clr4 is key for its self-association and phase-separated liquid droplet formation, and is inhibited by mono-ubiquitination. Previously, ncRNA has

been shown to inhibit Clr4 enzyme activity51, but as nucleosomes were used as substrate, this was thought to be due to competition with nucleosome binding. To characterize the effect of

ncRNA-mediated self-association on Clr4 H3K9 methylation activity, we fused the N terminal tail of _S. pombe_ H3 (A1-K36) with GST (H3N-GST) and used it as a substrate of Clr4 in the absence

or presence of ncRNA70. Interestingly, Clr4 enzymatic activity for H3K9me3 decreased greatly as the concentration of ncRNA increases, while Clr4 enzymatic activity for H3K9me2 did not

change with increasing ncRNA and self-association of Clr4 (Fig. 4d, e). These results indicate that ncRNA binding and Clr4 self-association has an inhibitory effect on Clr4 enzymatic

activity for H3K9me3, independent of nucleosome binding. SUV39H1 forms a complex with HP1, and SUV39H1 and HP1 together colocalize to distinct heterochromatic subnuclear domains

(chromocenters) in human cells43,66,71,72. HP1α readily forms phase-separated liquid droplets without addition of DNA or nucleosomes in vitro, but HP1β only forms liquid droplets with

addition of the N-terminal chromo domain (CD) of SUV39H1 and nucleosomes modified on histone H3K9me343. Like other HP1 proteins, _S. pombe_ Swi6 has been shown to undergo liquid-liquid phase

separation (LLPS) which contributes to the highly dynamic features of heterochromatin40,42,73. Swi6 forms phase-separated liquid droplets only in combination with nucleosomes73. As in human

cells, we found that recombinant mCherry-Swi6 could not form phase-separated liquid droplets in vitro even in combination with full-length GFP-Clr4. However, GFP-Clr4IDR and mCherry-Swi6

could form phase-separated liquid droplets when mixed together (Fig. 4f). These droplets dissolved upon the addition of a much smaller amount of ncRNA compared to that required to dissolve

droplets formed by mixing GFP-Clr4 with mCherry-Clr4 (Fig. 4c, f). Thus, these results suggest the possibility that phase-separated liquid droplet formation of Swi6 can be promoted by

Clr4IDR independently of modified nucleosomes, but only during transcriptional silencing when RNA is absent. GFP-Swi6 localizes to heterochromatin in vivo in a H3K9me-dependent manner and we

used this to assess heterochromatin condensation in _ubc4_ mutant backgrounds (Fig. 4g). Compared to WT and _dcr1Δ_ strains which have up to 4 condensed foci representing the centromeric

chromocenter, telomere clusters and the mating type locus, _ubc4-1_ mutants retained only one visible chromocenter, presumably due to the complete loss of H3K9me2/3 and Swi6 at telomeres and

mating-type loci detected by ChIP (Supplementary Fig. 2c, d). The remaining GFP-Swi6 foci were abolished in _dcr1Δ ubc4-_1, _ubc4-3xHA_, _cul4Δ_ and _clr4Δ_ mutants in accordance with

complete loss of H3K9me2/3 from all heterochromatin in these mutants (Fig. 4g). When a strain which has both GFP-Clr4 and mCherry-Swi6 was used, GFP-Clr4 and mCherry-Swi6 co-localized to a

condensed focus in vivo in WT cells. Only one chromocenter was visible because the N-terminal GFP fusion to Clr4 causes a silencing defect (Supplementary Fig. 9a, b). GFP-Clr4 and

mCherry-Swi6 co-localization was significantly reduced in _ubc4-1_ and _dcr1Δ_ and completely abolished in _dcr1Δ ubc4-1_ double mutants (Supplementary Fig. 9a). These results indicate that

Clr4 and Swi6 form condensates in vitro and in vivo, which is promoted in the absence of RNA during the G2 phase of the cell cycle. After mitosis, phosphorylation of H3S10 evicts Swi6 from

heterochromatin, resulting in transient transcription in G1 and S phase19. RNA promotes self-association of Clr4 and rapid dissolution of heterochromatic condensates, consistent with

previously observed de-clustering of GFP-Swi6 chromocenters during the cell cycle74. Subsequent mono-ubiquitination by CLRC releases Clr4 from RNA and promotes phase-separated liquid droplet

formation together with H3K9me3 and Swi6 to drive the transition to TGS in G2. UBIQUITINATION AND DE-UBIQUITINATION OF EPE1 AND BDF2BRD4 REGULATES CENTROMERIC SRNA AND CLR4 FOR

HETEROCHROMATIN SILENCING It has recently been shown that Ubc4 and Cul4 are strongly enriched in Swi6-associated chromatin, along with other CLRC subunits, as well as in chromatin associated

with the BET family double bromodomain protein Bdf2 (a homolog of mammalian BRD4)49. Bdf2 is recruited by the putative H3K9 demethylase, Epe1, which localizes to centromeric heterochromatin

in a Swi6-dependent manner and contributes to sRNA production75,76,77, which also occurs mainly in S phase19. Bdf2 binds to RNA Pol II promoters at the centromeric heterochromatin

boundary78 and interacts with the TFIID transcription factor complex79,80 while Epe1 is confined to these boundaries by Cul4-Ddb1Cdt2 E3 ligase-dependent poly-ubiquitination75. We found that

the increased centromeric sRNA production in _ubc4-1_ and _cul4-1_ mutants was accompanied by spreading of Epe1 and Bdf2 to pericentromeric repeats, as well as Ago1 and Chp1, two components

of the RITS siRNA complex (Figs. 2f, 5a, b; Supplementary Fig. 10a–c). Moreover, _epe1Δ_ and _bdf2Δ_ deletion mutants suppressed the increased sRNA production in _ubc4-1_ and _cul4-1_

mutants back to wild-type levels (Fig. 5a), by reducing recruitment of RNA Pol II (Supplementary Fig. 10d). Consistent with the regulation of Epe1 by Cul475, we found that Ubc4 and CLRC

induced poly-ubiquitination and degradation of Bdf2 (Fig. 5c and Supplementary Fig. 10e, f). Bdf2BRD4 recruits TFIID and RNA Pol II to centromeric heterochromatin, an important step in

precursor ncRNA transcription for sRNA production79,80. To mimic the effect of _ubc4-1_ and _cul4-1_ on sRNA production, we overexpressed Epe1 (_epe1-OE_, Supplementary Fig. 10g). As

expected, _epe1-OE_ resulted in increased sRNA production by increased recruitment of Epe1 and Bdf2, but unlike _ubc4-1_ and _cul4-1_, it also resulted in decrease of both H3K9me2 and

H3K9me3 at centromeric heterochromatin (Fig. 5a, b; Supplementary Fig. 10h–j). Thus, Ubc4-CLRC regulates sRNA production separately through Epe1 and Bdf2, which in combination with Clr4,

regulates the proper transition from CTGS to TGS (Supplementary Fig. 10k). Ubiquitination is a reversible, highly dynamic and transient modification, and its removal is catalyzed by

deubiquitinating enzymes21,81,82. A deletion mutant of the deubiquitinating enzyme Ubp3 rescues the heterochromatic silencing defects of _ago1Δ_, and _ubp3Δ_ shows positive epistatic genetic

interactions with CLRC component mutants83,84. While deletion of Ubp3 did not increase Clr4 mono-ubiquitination most likely because of redundancy between de-ubiquitinating enzymes

(Supplementary Fig. 11a)85, the overexpression of Ubp3 (_ubp3-OE_) resulted in a silencing defect at the centromeric _ura4__+_ reporter gene and reduced mono-ubiquitination of Clr4,

supporting the idea that Ubp3 counteracts Clr4 mono-ubiquitination by Ubc4-CLRC (Fig. 5d and Supplementary Fig. 11b, c). The overexpression of Ubp3 also resulted in an increase of H3K9me2

and decrease of H3K9me3, as well as increased sRNA production, with increased binding of Epe1 and Bdf2 to centromeric heterochromatin, resembling the phenotypes of _ubc4-1_ or _cul4-1_

mutants at centromeric heterochromatin (Fig. 5e, left; Supplementary Fig. 11d). Similarly, at telomeres, _ubp3-OE_ resulted in increased H3K9me2 but decreased H3K9me3 as well as greatly

increased recruitment of Epe1 and Bdf2, resulting in ectopic sRNA production (Fig. 5e, right; Supplementary Fig. 11e). The 5’-nucleotide of these sRNAs from centromere and subtelomere

repeats had a strong U/A bias and the typical sRNA size was 23±1 nt, as is expected for bona fide RNAi products (Fig. 5f, g)12,86. These results indicate that Ubp3 counteracts the activities

of Ubc4-CLRC, both in terms of H3K9me2/3 transitions and sRNA production, with overexpression of Ubp3, making the chromatin structure of telomeres resemble that of the centromere.

DISCUSSION Our findings indicate that, while CTGS and H3K9 di-methylation depend strictly on RNAi, mono-ubiquitination of Clr4 promotes the transition from CTGS to TGS by first releasing

Clr4 from heterochromatin. The key lysine residues in the IDR form a β-sheet on binding of the nucleosome core51, which is likely disrupted by ubiquitination. Deubiquitination by Ubp3 then

allows re-engagement and spreading of H3K9 tri-methylation to adjacent chromatin, as demonstrated by ectopic heterochromatin formation at the _ade6__+_ reporter gene mediated by tethered

Clr4 (Fig. 3e). Spreading plays a similar role at the silent mating type locus and at telomeres where heterochromatic silencing is completely lost in _ubc4-1_ and _cul4-1_ mutants (Fig. 1c;

Supplementary Fig. 2c, d). The ability to induce the spreading of silencing by this mechanism may be critical at these genomic loci as these chromosomal locations have only limited capacity

for ncRNA transcription and RNAi. Ubc4-CLRC likely has additional functions, such as H3K14 ubiquitination, which was recently shown to enhance H3K9 methylation in vitro70. But H3K9me2 and

H3K9me3 are both lost at centromeres in H3K14R mutants, which suggests that H3K14 ubiquitination is essential for histone H3K9 methylation but not for transition of H3K9me2 to H3K9me3

(Supplementary Fig. 12a–c). Centromeric sRNA production was not affected by mutants in the chromo- (_clr4_-_W31G_), and SET domains (-_F449Y_ and -_I418P_) which are also defective in the

H3K9me2-to-H3K9me3 transition (Supplementary Fig. 12d, e)18. Interestingly, we found that _clr4_-_W31G_, -_F449Y_ and -_I418P_ mutants also have reduced Clr4 mono-ubiquitination indicating a

strong correlation between Clr4 mono-ubiquitination and transition of H3K9me2 to me3 (Supplementary Fig. 12f, g). Recently, phosphorylation of the Clr4 SET domain by Cdk1 was also shown to

be required for H3K9me3 during meiosis, and for proper gametogenesis, although the role in silencing, if any, was not examined87. The hierarchical relationship between phosphorylation,

auto-methylation and mono-ubiquitination of Clr4 remains to be explored52. Like Swi6, Clr4 forms biomolecular condensates in vitro and this property is mediated by the IDR.

Post-translational modifications of IDRs, and their association with RNA, can affect protein function significantly; including the induction of disorder-to-order transitions, alternative

complex formation and differential subcellular localization88,89,90,91,92,93,94. We found that ubiquitination of the IDR of Clr4 prevents RNA binding and protein self-association. This

promotes phase-separated liquid droplet formation, and is critical for the CTGS to TGS transition in vivo. These phase-separated liquid droplets of Clr4 and Clr4IDR/Swi6 are rapidly

dissolved in the presence of ncRNA and this RNA-dependent interaction promotes Clr4 self-association instead (Fig. 4c, d, f; Supplementary Fig. 8c). Intriguingly, binding of centromeric

ncRNA to the Swi6 hinge region, which is also disordered and lysine-rich, has previously been shown to prevent Swi6 binding to chromatin95, consistent with our results. In the absence of

mono-ubiquitination, Clr4 recruitment to genomic loci where ncRNA and sRNA are overproduced greatly increased in agreement with previous studies that mammalian SUV39H1 chromodomain binds to

pericentromeric RNA for the recruitment of SUV39H1 to heterochromatin, H3K9me3 deposition and heterochromatin assembly (Fig. 3a–c)96,97,98. We found that ncRNA binding inhibits Clr4

enzymatic activity for H3K9me3 likely through binding of ncRNA to the chromodomain and IDR of Clr4 and subsequent self-association (Fig. 4d, e and Supplementary Fig. 8c)54. In summary (Fig. 

6), we have found that ncRNA transcription and sRNA production during G1/S promote CTGS and H3K9 di-methylation through recruitment of CLRC. The production of sRNA depends on the

transcriptional co-activator Epe1-Bdf2, which licenses recruitment of RNA Pol II to transcriptionally permissive heterochromatin13,14,15,18,46,53,59,99,100. Subsequently, removal of

Epe1-Bdf2 by Ubc4-CLRC-mediated poly-ubiquitination silences transcription, while Clr4 mono-ubiquitination is required for its dissociation from chromatin, and from ncRNA, triggering Clr4

dependent H3K9me3 and subsequent binding of HP1 proteins, Swi6 and Chp2 important for RNAi-independent TGS18,53,101. In the absence of RNAi, variable amounts of H3K9me2 are maintained at the

centromere, indicating that Clr4 must be recruited, at least in part independently of RNAi102. Based on our results, one explanation is that ncRNA, in the presence of existing H3K9

methylation, initially recruits Clr4 and promotes self-association and chromatin binding. In S phase, RNAi recruits the Ubc4-Cul4-Rik1Raf1Raf2 E3 ligase59, which mono-ubiquitinates Clr4

preventing RNA binding, and promoting H3K9me3, potentially by phase-separated liquid droplet formation and three-dimensional “spreading” along with Swi6 (Fig. 6). Romoval of ubiquitin

moieties by Ubp3 allows rapid recycling of Epe1-Bdf2 and of Clr4, ready for the next cell cycle. Given the high conservation of Ubc4, Clr4, Brd4 and Cul4, similar transitions between

RNAi-dependent gene silencing and heterochromatin assembly may be regulated by ubiquitination and RNA in other organisms, including mammals, where TRIM28, HP1 and SUVH39 form condensates

with H3K9me3 nucleosome arrays, similar to those reported here43. METHODS YEAST STRAINS AND PLASMIDS MANIPULATION Yeast strains, plasmids, oligonucleotide sequences and antibodies are listed

in Supplemental Tables S1–S4. All yeast strains were constructed by the PCR-based targeting method103 and lithium-acetate transformation method104. For N-terminus tagging of 3xFlag to

_clr4_, _rik1_ and _raf1_, endogenous promoters and terminators of each gene were used. For C-terminus tagging of 3xFlag, 3xFlag sequence and _adh1_ terminator with Kan resistant (KanR) gene

was inserted at the end of open reading frames (ORF). For some 3xFlag-Clr4 lysine mutants, _adh1_ terminator with Kan resistant (KanR) gene was inserted at the end of _clr4_ open reading

frames (ORF). For overexpression of _ubp3_, _ubp8_, and _epe1_, 500 bp of _adh1_ promoter was inserted between their promoter and ORF. To construct pET-14b-Clr4, the coding DNA sequence of

full-length Clr4 (amino acids 1-490) was amplified from _S. pombe_ genomic DNA, digested with _NdeI_ and cloned into _NdeI_-digested pET-14b. For pHis-parallel-GFP-Clr4Full and

pHis-parallel-GFP-Clr4IDR construction, the coding DNA sequence of full-length Clr4Full (amino acids 1-490) and Clr4IDR (amino acids 51-220) were amplified from _S. pombe_ genomic DNA,

digested with _BamHI_/_XhoI_ and cloned into _BamHI_/_XhoI_-digested pHis-parallel-GFP. For pHis-parallel-mCherry-Clr4, mCherry-Clr4 DNA sequence was synthesized (IDT), digested with

_NdeI_/_XhoI_ and cloned into _NdeI_/_XhoI_-digested pHis-parallel-GFP. For pHis-Clr4IDR-GFP, DNA sequence was synthesized (IDT) and cloned into pET-14b vector using In-Fusion HD cloning kit

(Takara). To construct pET-14b-Flag-Clr4 and pET-14b-Myc-Clr4, the coding DNA sequence of full-length Clr4 (amino acids 1-490) was amplified from _S. pombe_ genomic DNA using oligomers

containing 6xHis-1xFlag or 6xHis-1xMyc sequences respectively and cloned into pET-14b vector using In-Fusion HD cloning kit (Takara). SILENCING SPOT ASSAY Cells grown freshly on YES plate

were picked and resuspended in water to make a concentration of ~5 × 104 cells per μl. 8 μl of 10-fold serial dilutions were spotted on appropriate plates like YES, low adenine (Low Ade),

EMM-Ura (-Ura), and YES+TBZ (TBZ), and incubated 3 days at 30 °C for photography. _ura4__+_ reporter gene silencing was also tested on 5-fluoro-orotic acid (FOA) plates, which kills cells

expressing Ura4. IODINE STAINING Freshly grown yeast cells were spotted on SPA plates to induce sporulation, incubated at room temperature for 3 days, and then stained with iodine vapor for

2–3 min. Staining of starch which is abundant in mature spore produces a dark brown color. EMS MUTAGENESIS AND WHOLE-GENOME SEQUENCING FOR _UBC4-G48D SUPPRESSOR_ EMS mutagenesis was carried

out as previously described in ref. 105. Briefly, freshly grown 1 x 109 yeast cells were washed with distilled water, resuspended in 1.5 ml of 0.1 M sodium phosphate buffer (pH 7.0) and

treated with 50 μl EMS (Sigma M0880) for 1 h at 30 °C. 0.2 ml of treated cells was moved to 8 ml of sterile 5% sodium thiosulfate to inactivate EMS. Cells were washed, resuspended in sterile

water and plated on YES plate. Single colonies from plates were patched on SPA plates to induce sporulation, incubated at room temperature for 3 days, and then stained with iodine vapor for

2–3 min. For whole-genome sequencing for _ubc4-G48D_ suppressors, genomic DNA was purified from freshly grown yeast cells using Genomic-tip 20/G column (QIAGEN 10223). DNA library was

constructed using the TruSeq DNA PCR-Free Library Prep Kit (Illumina 20015962). Barcoded DNA library was sequenced using Illumina MiSeq platform, which generates paired-end 151 nt reads and

analyzed as previously described in ref. 106. Briefly, reads were adapter-trimmed and quality-filtered using Sickle (paired-end mode), then mapped to the _S. pombe_ genome using Bowtie2107.

Duplicate reads were removed using Samtools (≤ 0.14% per library). SNPs were called using FreeBayes to find SNPs present in suppressor strain but absent in the parental strain108. This

uncovered a G→C transversion at position chr2:716,192, a second-site mutation in the _ubc4_ gene. The mutation was verified by Sanger sequencing and confirmed to be causative by constructing

the same fresh mutation in WT and _ubc4-1_ strains. LIVE CELL IMAGING Images of freshly growing cells were collected using Zeiss Axio Imager M2 with DIC and EGFP channels and processed

using Zeiss Zen software. QUANTITATIVE REVERSE TRANSCRIPTION PCR (RT-QPCR) Total RNAs were purified from freshly grown yeast cells using _Quick_-RNA Fungal/Bacterial Miniprep Kit

(Zymoresearch R2014). Genomic DNAs were removed using DNase I (Roche 04716728001) and RNAs were cleaned using RNA Clean & Concentrator (Zymoresearch R1013). Reverse transcription was

carried out using SuperScript IV First-Strand Synthesis System (ThermoFisher 18091050) with random hexamer or oligo dT as primers. For quantitative PCR (qPCR), iQ SYBR Green Supermix

(Bio-rad 1708882) was used with the primers listed in Supplemental Table S3 and samples were run on CFX96 Real-Time PCR Detection System (Bio-rad). CHROMATIN IMMUNOPRECIPITATION (CHIP)

Chromatin Immunoprecipitation (ChIP) experiments were performed as previously described in ref. 79. Briefly, freshly grown 40 ml cells in YES were fixed with 1% formaldehyde from 37% stock

(Sigma F8775) for 20 min at room temperature for H3K9me2, H3K9me3, RNA Pol II, Ago1 and Chp1 ChIP. For Swi6 and 3xFlag-Clr4 ChIP, 40 ml cells were pre-incubated at 18 °C for 2 h and fixed

with 1.45% Formaldehyde from 16% stock (ThermoFisher 28908) for 30 min at 18 °C. Cells were quenched with 360 mM glycine and 2.4 mM Tris for 5 min. Whole-cell extracts (WCE) were prepared

using FastPrep-24 and sonicated using Covaris. 1~2 μg of antibodies like anti-H3K9me2 (Abcam ab1220), anti-H3K9me3 (Absolute Antibody Ab00700-1.26), anti-RNA polymerase II (BioLegend

904001), Swi6 (Abcam ab188276), Ago1 (Abcam Ab18190), Chp1 (Abcam ab18191) and anti-Flag (Sigma F1804) were pre-incubated with 50 μl Protein A and Protein G sepharose (GE 17513801 and

17061801) mix for 3~4 h and then incubated with sheared chromatin overnight at 4 °C. After washing, bead-antibody-chromatin mixture were treated with Proteinase K (ThermoFisher 100005393)

for 1 h at 42 °C and incubated 5~6 h at 65 °C to reverse cross-links. DNA was cleaned up using ChIP DNA Clean & Concentrator (Zymo Research D5205) and subsequently used for qPCR or

ChIP-Seq. CHIP-SEQ ChIP-Seq libraries were constructed using NEBNext Ultra II DNA Library Prep Kit for Illumina (NEB E7645) and barcoded multiplex libraries were sequenced using Illumina

MiSeq platform, which generates paired-end 101 bp reads. Raw reads were trimmed using Trimmomatic, mapped to _S. pombe_ genome using Bowtie2 and visualized in IGV genome browser. Total

ChIP-seq read counts for defined region were analyzed using MultiBigwigSummary or MultiBamSummary. SRNA-SEQ For sRNA-seq, total RNAs were purified from freshly grown yeast cells using

_Quick_-RNA Fungal/Bacterial Miniprep Kit (Zymoresearch R2014) and sRNA were enriched using RNA Clean & Concentrator (Zymoresearch R1013). sRNA libraries were constructed using NEBNext

Small RNA Library Prep Set for Illumina (NEB E7330) and barcoded multiplex libraries were sequenced using Illumina MiSeq platform, which generates single-end 36 nt reads. Reads were

adapter-trimmed using Cutadapt se, keeping a minimum read size of 15 nt, and mapped to the _S. pombe_ genome using Bowtie109 allowing up to 1 mismatch. The strand-specific genomic coverage

was calculated using Bedtools110. CO-IMMUNOPRECIPITATION (CO-IP) AND WESTERN BLOTTING (WB) Whole-cell extracts (WCE) were prepared from freshly grown yeast cells in lysis buffer (40mM

Tris-HCl, pH 7.5, 1 mM EDTA, 10% Glycerol, 0.1% NP-40, 300 mM NaCl) with Protease inhibitor cocktail (Roche 11836170001). For Co-IP of Cul4-GFP with Ubc4, 2 μg of GFP antibody (Roche

11814460001) was pre-incubated with 50 μl Protein A and Protein G Sepharose (GE Healthcare 17513801 and 17061801) mix for 3~4 h and then incubated with WCE overnight at 4 °C. After washing,

purified proteins were boiled and loaded on 4–20% Mini-PROTEAN TGX Gel (Bio-rad 456-1093) and Western blotting was performed with anti-GFP (Abcam 290) and anti-mUBE2D3 (Proteintech

11677-1-AP) antibodies to detect Cul4-GFP and Ubc4 using ECL (Thermo Fisher 32209) assay according to the manufacturer’s instruction. For Co-IP of Flag-Rik1, Flag-Dos1 or Dos2-Flag with

Ubc4, 50 μl of Anti-Flag M2 Affinity Gel (Sigma A2220) was incubated with WCE overnight at 4 °C. For detection of Flag tagged protein and Ubc4, anti-Flag (Sigma F7425) and anti-mUBE2D3

(Proteintech 11677-1-AP) antibodies were used. PURIFICATION OF UBIQUITINATED FLAG-CLR4 AND BDF2-FLAG PROTEINS For purification of ubiquitinated Flag-Clr4, WCE prepared from freshly grown

cells in lysis buffer (50 mM Hepes, pH 7.6, 150 mM NaCl, 300 mM KAc, 5 mM MgAc, 10% Glycerol, 20 mM β-GP, 1 mM EDTA, 1 mM EGTA, 0.1% NP-40, 1 mM DTT, 250 U/ml Dnase I (Roche 04716728001),

25~50 μg/ml Rnase A (Sigma R4642), 1 mM PMSF, 0.3 mM PR-619 (Sigma 662141)) with Protease inhibitor cocktail (Roche 11836170001) and De-ubiquitination/sumoylation Inhibitor (Cytoskeleton

NEM09BB) was incubated with Anti-Flag M2 Affinity Gel or Ubiquitination Affinity Beads (UAB, Cytoskeleton UBA01- Beads) overnight at 4 °C. After washing, purified proteins were loaded on

4–20% Mini-PROTEAN TGX Gel (Bio-rad 456-1093). For detection of ubiquitinated Clr4 protein, anti-Flag (Sigma F7425) or anti-ubiquitin (Cell Signalling 3936T) antibodies were used. Cells were

also treated with proteasome inhibitor 100 μM MG-132 (APExBIO A2585) for 4 h before harvesting to detect poly-ubiquitinated Bdf2-Flag or to see the effect of blocking proteasome pathway by

treating MG-132 on mono-ubiquitinated Flag-Clr4. CLRC COMPLEX PURIFICATION FROM _S. POMBE_ Protein purifications were performed as described previously15. Briefly, Rik1-TAP were purified

from 6 to 12 g of cells. Cells were lysed in 1 volume of lysis buffer (50 mM HEPES-KOH, pH 7.6, 300 mM KAc, 10% glycerol, 1 mM EGTA, 1 mM EDTA, 0.1% NP-40, 1 mM DTT, 5 mM MgAc, 1 mM NaF, 20 

mM β-GP, 1 mM PMSF, 1 mM Benzamidine and 1 μg/ml of Leupeptin, Aprotinin, Bestatin and Pepstatin). 1 volume lysis buffer was added and the lysate was centrifuged at 16,500 rpm for 25 min at

4 °C in a SA600 rotor. The first affinity purification was performed using Ig-G Sepharose beads at 4 °C for 2 h. The beads were washed with lysis buffer and TEV-C buffer (10 mM Tris-HCl, pH

8.0, 150 mM KAc, 0.1% NP-40, 0.5 mM EDTA, 1 mM DTT) and bound protein was eluted using a GST-TEV protease. The eluate was diluted in 2 volumes of CAM-B buffer (10 mM Tris-HCL, pH 8.0, 150 mM

NaCl, 1 mM MgAc, 1 mM Imidazole, 2 mM CaCl2, 10 mM βME) and the second affinity purification was performed by adding Calmodulin-Sepharose at 4 °C for 1 h. The beads lysate was washed with

CAM-B buffer containing 0.1 % NP-40 and bound protein was eluted with CAM-E buffer (10 mM Tris-HCL, pH 8.0, 150 mM NaCl, 1 mM MgAc, 1 mM Imidazole, 10 mM EGTA, 10 mM βME). Purified CLRC

complex was added to in vitro ubiquitination assay of 6xHis-Clr4. RECOMBINANT PROTEIN PURIFICATION FROM E. COLI, IN VITRO UBIQUITINATION AND NUCLEOSOME BINDING ASSAY BL21-CodonPlus (DE3)-RIL

strains (Agilent Technologies 230245) that express recombinant Clr4 proteins were grown with 0.4 mM IPTG for 20 h at 18 °C. Induced proteins were purified on HisPur Ni-NTA Superflow Agarose

(ThermoFisher 25214) and eluted with elution buffer containing 150 mM Imidazole. The recombinant proteins purified from E. coli were further dialyzed against 1x TBS using Slide-A-Lyzer

Dialysis Cassettes (ThermoFisher 66003). For in vitro ubiquitination, recombinant 6xHis-Clr4 protein or 3xFlag-Clr4 purified form _S. pombe_ cells were mixed with human Ubiquitin-activating

Enzyme (UBE1, R&D systems E-305), human UbcH5c/UBE2D3 (R&D systems E2-627), human Ubiquitin (R&D systems U-100H) or human HA-Ubiquitin (R&D systems U-110), Mg-ATP (R&D

systems B-20) and 10x E3 Ligase Reaction Buffer (R&D systems B-71) with or without CLRC complex purified from _S. pombe_ cells. For nucleosome or ncRNA binding assay, biotin labeled

Recombinant Mononucleosomes (H3.1, Active Motif 31467) and Recombinant Mononucleosomes H3K9me3 (H3.2, Active Motif 31555) or biotin labeled ncRNA (synthesized from IDT) were pre-incubated

with Streptavidin beads (GE Healthcare 17-5113-01) and incubated in vitro ubiquitinated 6xHis-Clr4 or Flag-Clr4 proteins purified from _S. pombe_ cells. Samples were washed, boiled and

loaded on 4–20% Mini-PROTEAN TGX Gel (Bio-rad 456-1093) and WB was performed with anti-Flag (Abcam ab1791), anti-H3K9me3 (Abcam ab8898) and anti-H3 (Abcam ab1791), and using ECL (Thermo

Fisher 32209) assay according to the manufacturer’s instruction. MASS SPECTROMETRY The gel band sample was reduced and alkylated with _N_-ethylmaleimide (NEM), digested with trypsin for 1.5

h at 37 °C and washed with ammonium bicarbonate/acetonitrile to remove stain, SDS and other reagents. Peptides were extracted from the gel pieces, dried down and re-dissolved in 2.5%

acetonitrile, 0.1% formic acid. Each digest was run by nanoLC-MS/MS using a 2 h gradient on a 0.075 mm x 250 mm C18 column feeding into a Q-Exactive HF mass spectrometer. All MS/MS samples

were analyzed using Mascot (Matrix Science, London, UK; version 2.6.1). Mascot was set up to search the cRAP_20150130.fasta; custom4_20190215.fasta; SwissProt_2019_02 database (selected for

_Schizosaccharomyces pombe_, unknown version, 5269 entries) assuming the digestion enzyme trypsin. Mascot was searched with a fragment ion mass tolerance of 0.060 Da and a parent ion

tolerance of 10.0 PPM. Deamidated of asparagine and glutamine, oxidation of methionine, phospho of serine, threonine and tyrosine, GG of lysine, N-ethylmaleimide of cysteine were specified

in Mascot as variable modifications. Scaffold (version Scaffold_4.8.9, Proteome Software Inc., Portland, OR) was used to validate MS/MS based peptide and protein identifications. Peptide

identifications were accepted if they could be established at greater than 80.0% probability by the Peptide Prophet algorithm111 with Scaffold delta-mass correction. Protein identifications

were accepted if they could be established at greater than 99.0% probability and contained at least 2 identified peptides. Protein probabilities were assigned by the Protein Prophet

algorithm112. Proteins that contained similar peptides and could not be differentiated based on MS/MS analysis alone were grouped to satisfy the principles of parsimony. Proteins sharing

significant peptide evidence were grouped into clusters. LIQUID-LIQUID PHASE SEPARATION (LLPS) LLPS of GFP-Clr4, mCherry-Clr4, GFP-Clr4IDR and Clr4IDR-GFP proteins were induced by diluting

proteins to 100–500 nM range in 10 mM HEPES buffer (pH 7.6). To observe the effect of ncRNA on Clr4 LLPS, 120nt synthesized centromeric ncRNA (IDT) was added to the Clr4 proteins. To observe

the effect of 1,6-Hexanediol (Sigma H11807) on GFP-Clr4IDR LLPS, GFP-Clr4IDR protein was diluted to 10 mM HEPES buffers (pH 7.6) containing 5% Ethanol or 5% Hexanediol. LLPS was imaged on

Zeiss Axio Imager M2 with differential interference contrast (DIC) and EGFP channels and processed using Zeiss Zen software. FLUORESCENCE RECOVERY AFTER PHOTOBLEACHING (FRAP) EXPERIMENTS

FRAP experiments were performed as previously described in ref. 113. Briefly, 500 nM of GFP-Clr4IDR protein were diluted in 3.6% PEG8000. PEG8000 was used to stabilize phase-separated liquid

droplets during the experiments. FRAP experiments were carried out using Zeiss LSM710 Confocal Laser Scanning Microscope with 63x objective. All Bleaching experiments were performed with

488 nm argon laser at 100% intensity. Images before and after photobleaching were taken at 5 s interval and analyzed using standard methods. IN VITRO PULL DOWN ASSAY 4 μg of Flag antibody

(Sigma F1804) was pre-incubated with 50 μl Protein A and Protein G Sepharose (GE Healthcare 17513801 and 17061801) mix in 1 mL lysis buffer (40 mM Tris-HCl, pH 7.5, 1 mM EDTA, 10% Glycerol,

0.1% NP-40, 250 mM NaCl) for 1 h and then incunbated with recombinant Flag-Clr4 and Myc-Clr4 proteins (500 nM each) and different amount of ncRNA (0, 50, 100 and 200 nM) in 500 μL lysis

buffer at 4 °C for 3 h. After washing, purified proteins were boiled and loaded on 4–20% Mini-PROTEAN TGX Gel (Bio-rad 456-1093) and Western blotting was performed with anti-Myc (Abcam

ab9106), and anti-Flag (Sigma, F7425) using ECL (Thermo Fisher 32209) assay according to the manufacturer’s instruction. IN VITRO HISTONE METHYLATION (HMT) ASSAY Recombinant Flag-Clr4

protein (5 μM) were mixed with recombinant H3N-GST (10 μM), 200 μM _S_-(5′-Adenosyl)-_L_-methionine chloride dihydrochloride (SAM, Sigma 7007) and different amoount of ncRNA (0, 0.5, 1 and 2

 μM) in 50 μL HMT buffer (50mM Tris, pH 8.0, 20 mM KCl, 10 mM MgCl2, 5% Glycerol, 1 mM DTT and 1 mM PMSF) and incubated at 30 °C for 1.25 h. Samples were boiled and loaded on 4–20%

Mini-PROTEAN TGX Gel (Bio-rad 456-1093) and Western blotting was performed with anti-H3K9me2 (Abcam ab1220), and anti-H3K9me3 (Absolute Antibody Ab00700-1.26) using ECL (Thermo Fisher 32209)

assay according to the manufacturer’s instruction. STATISTICS AND REPRODUCIBILITY The experiments for Figs. 1b, e, 2b, d, 3d, 4a, c–f, 5c–d and Supplementary Figs. 1c, 3, 4, 5e–I, 7c, 8d,

f, 10e, f, 11a, 12f, g were performed at least twice and similar results were obtained. The representative results were shown in the figures. SOURCE DATA FILE For an example of presentation

of full scan blots, see the Source Data file of https://www.nature.com/articles/s41467-020-16984-1#Sec35 and for more information, please refer to

https://www.nature.com/nature-research/editorial-policies/image-integrity. REPORTING SUMMARY Further information on research design is available in the Nature Portfolio Reporting Summary

linked to this article. DATA AVAILABILITY Genome-wide datasets are deposited in the Gene Expression Omnibus (GEO) under the accession number GSE156069. Source Data is provided. REFERENCES *

Castel, S. E. & Martienssen, R. A. RNA interference in the nucleus: roles for small RNAs in transcription, epigenetics and beyond. _Nat. Rev. Genet_ 14, 100–112 (2013). Article  CAS 

PubMed  Google Scholar  * Janssen, A., Colmenares, S. U. & Karpen, G. H. Heterochromatin: Guardian of the Genome. _Annu Rev. Cell Dev. Biol._ 34, 265–288 (2018). Article  CAS  PubMed 

Google Scholar  * Allshire, R. C. & Madhani, H. D. Ten principles of heterochromatin formation and function. _Nat. Rev. Mol. Cell Biol._ 19, 229–244 (2018). Article  CAS  PubMed  Google

Scholar  * Martienssen, R. & Moazed, D. RNAi and heterochromatin assembly. _Cold Spring Harb. Perspect. Biol._ 7, a019323 (2015). Article  PubMed  Google Scholar  * Gutbrod, M. J. et al.

Dicer promotes genome stability via the bromodomain transcriptional co-activator BRD4. _Nat. Commun._ 13, 1001 (2022). Article  ADS  CAS  PubMed  Google Scholar  * Hall, I. M., Noma, K.

& Grewal, S. I. RNA interference machinery regulates chromosome dynamics during mitosis and meiosis in fission yeast. _Proc. Natl Acad. Sci. USA_ 100, 193–198 (2003). Article  ADS  CAS 

PubMed  Google Scholar  * Provost, P. et al. Dicer is required for chromosome segregation and gene silencing in fission yeast cells. _Proc. Natl Acad. Sci. USA_ 99, 16648–16653 (2002).

Article  ADS  CAS  PubMed  Google Scholar  * Bannister, A. J. et al. Selective recognition of methylated lysine 9 on histone H3 by the HP1 chromo domain. _Nature_ 410, 120–124 (2001).

Article  ADS  CAS  PubMed  Google Scholar  * Nakayama, J., Rice, J. C., Strahl, B. D., Allis, C. D. & Grewal, S. I. Role of histone H3 lysine 9 methylation in epigenetic control of

heterochromatin assembly. _Science_ 292, 110–113 (2001). Article  ADS  CAS  PubMed  Google Scholar  * Rea, S. et al. Regulation of chromatin structure by site-specific histone H3

methyltransferases. _Nature_ 406, 593–599 (2000). Article  ADS  CAS  PubMed  Google Scholar  * Colmenares, S. U., Buker, S. M., Buhler, M., Dlakić, M. & Moazed, D. Coupling of

double-stranded RNA synthesis and siRNA generation in fission yeast RNAi. _Mol. Cell_ 27, 449–461 (2007). Article  CAS  PubMed  Google Scholar  * Djupedal, I. et al. Analysis of small RNA in

fission yeast; centromeric siRNAs are potentially generated through a structured RNA. _Embo J._ 28, 3832–3844 (2009). Article  CAS  PubMed  Google Scholar  * Motamedi, M. R. et al. Two RNAi

complexes, RITS and RDRC, physically interact and localize to noncoding centromeric RNAs. _Cell_ 119, 789–802 (2004). Article  CAS  PubMed  Google Scholar  * Volpe, T. A. et al. Regulation

of heterochromatic silencing and histone H3 lysine-9 methylation by RNAi. _Science_ 297, 1833–1837 (2002). Article  ADS  CAS  PubMed  Google Scholar  * Verdel, A. et al. RNAi-mediated

targeting of heterochromatin by the RITS complex. _Science_ 303, 672–676 (2004). Article  ADS  CAS  PubMed  Google Scholar  * Gutbrod, M. J. & Martienssen, R. A. Conserved chromosomal

functions of RNA interference. _Nat. Rev. Genet_ 21, 311–331 (2020). Article  CAS  PubMed  Google Scholar  * Chen, E. S. et al. Cell cycle control of centromeric repeat transcription and

heterochromatin assembly. _Nature_ 451, 734–737 (2008). Article  ADS  CAS  PubMed  Google Scholar  * Jih, G. et al. Unique roles for histone H3K9me states in RNAi and heritable silencing of

transcription. _Nature_ 547, 463–467 (2017). Article  CAS  PubMed  Google Scholar  * Kloc, A., Zaratiegui, M., Nora, E. & Martienssen, R. RNA interference guides histone modification

during the S phase of chromosomal replication. _Curr. Biol._ 18, 490–495 (2008). Article  CAS  PubMed  Google Scholar  * Cappadocia, L. & Lima, C. D. Ubiquitin-like Protein Conjugation:

Structures, Chemistry, and Mechanism. _Chem. Rev._ 118, 889–918 (2018). Article  CAS  PubMed  Google Scholar  * Hochstrasser, M. Origin and function of ubiquitin-like proteins. _Nature_ 458,

422–429 (2009). Article  ADS  CAS  PubMed  Google Scholar  * Rape, M. Ubiquitylation at the crossroads of development and disease. _Nat. Rev. Mol. Cell Biol._ 19, 59–70 (2018). Article  CAS

  PubMed  Google Scholar  * Swatek, K. N. & Komander, D. Ubiquitin modifications. _Cell Res_. 26, 399–422 (2016). Article  CAS  PubMed  Google Scholar  * Ye, Y. & Rape, M. Building

ubiquitin chains: E2 enzymes at work. _Nat. Rev. Mol. Cell Biol._ 10, 755–764 (2009). Article  CAS  PubMed  Google Scholar  * Hong, E. J., Villén, J., Gerace, E. L., Gygi, S. P. &

Moazed, D. A cullin E3 ubiquitin ligase complex associates with Rik1 and the Clr4 histone H3-K9 methyltransferase and is required for RNAi-mediated heterochromatin formation. _RNA Biol._ 2,

106–111 (2005). Article  CAS  PubMed  Google Scholar  * Horn, P. J., Bastie, J. N. & Peterson, C. L. A Rik1-associated, cullin-dependent E3 ubiquitin ligase is essential for

heterochromatin formation. _Genes Dev._ 19, 1705–1714 (2005). Article  CAS  PubMed  Google Scholar  * Jia, S., Kobayashi, R. & Grewal, S. I. Ubiquitin ligase component Cul4 associates

with Clr4 histone methyltransferase to assemble heterochromatin. _Nat. Cell Biol._ 7, 1007–1013 (2005). Article  CAS  PubMed  Google Scholar  * Kuscu, C. et al. CRL4-like Clr4 complex in

Schizosaccharomyces pombe depends on an exposed surface of Dos1 for heterochromatin silencing. _Proc. Natl Acad. Sci. USA_ 111, 1795–1800 (2014). Article  ADS  CAS  PubMed  Google Scholar  *

Thon, G. et al. The Clr7 and Clr8 directionality factors and the Pcu4 cullin mediate heterochromatin formation in the fission yeast Schizosaccharomyces pombe. _Genetics_ 171, 1583–1595

(2005). Article  CAS  PubMed  Google Scholar  * Li, F. et al. Two novel proteins, dos1 and dos2, interact with rik1 to regulate heterochromatic RNA interference and histone modification.

_Curr. Biol._ 15, 1448–1457 (2005). Article  CAS  PubMed  Google Scholar  * Banani, S. F., Lee, H. O., Hyman, A. A. & Rosen, M. K. Biomolecular condensates: organizers of cellular

biochemistry. _Nat. Rev. Mol. Cell Biol._ 18, 285–298 (2017). Article  CAS  PubMed  Google Scholar  * Brangwynne, C. P. et al. Germline P granules are liquid droplets that localize by

controlled dissolution/condensation. _Science_ 324, 1729–1732 (2009). Article  ADS  CAS  PubMed  Google Scholar  * Kato, M. et al. Cell-free formation of RNA granules: low complexity

sequence domains form dynamic fibers within hydrogels. _Cell_ 149, 753–767 (2012). Article  CAS  PubMed  Google Scholar  * Rhine, K., Vidaurre, V. & Myong, S. RNA Droplets. _Annu Rev.

Biophys._ 49, 247–265 (2020). Article  CAS  PubMed  Google Scholar  * Rippe, K. Liquid-Liquid Phase Separation in Chromatin. _Cold Spring Harb Perspect Biol_ 14,

https://doi.org/10.1101/cshperspect.a040683 (2022). * Weber, S. C. & Brangwynne, C. P. Getting RNA and protein in phase. _Cell_ 149, 1188–1191 (2012). Article  CAS  PubMed  Google

Scholar  * Molliex, A. et al. Phase separation by low complexity domains promotes stress granule assembly and drives pathological fibrillization. _Cell_ 163, 123–133 (2015). Article  CAS 

PubMed  Google Scholar  * Mittag, T. & Parker, R. Multiple Modes of Protein-Protein Interactions Promote RNP Granule Assembly. _J. Mol. Biol._ 430, 4636–4649 (2018). Article  CAS  PubMed

  Google Scholar  * Gibson, B. A. et al. Organization of Chromatin by Intrinsic and Regulated Phase Separation. _Cell_ 179, 470–484.e421 (2019). Article  CAS  PubMed  Google Scholar  *

Larson, A. G. et al. Liquid droplet formation by HP1α suggests a role for phase separation in heterochromatin. _Nature_ 547, 236–240 (2017). Article  ADS  CAS  PubMed  Google Scholar  *

Sabari, B. R. et al. Coactivator condensation at super-enhancers links phase separation and gene control. _Science_ 361, https://doi.org/10.1126/science.aar3958 (2018). * Strom, A. R. et al.

Phase separation drives heterochromatin domain formation. _Nature_ 547, 241–245 (2017). Article  ADS  CAS  PubMed  Google Scholar  * Wang, L. et al. Histone Modifications Regulate Chromatin

Compartmentalization by Contributing to a Phase Separation Mechanism. _Mol. Cell_ 76, 646–659.e646 (2019). Article  CAS  PubMed  Google Scholar  * Irvine, D. V. et al. Mapping epigenetic

mutations in fission yeast using whole-genome next-generation sequencing. _Genome Res_. 19, 1077–1083 (2009). Article  CAS  PubMed  Google Scholar  * Javerzat, J. P., Cranston, G. &

Allshire, R. C. Fission yeast genes which disrupt mitotic chromosome segregation when overexpressed. _Nucleic Acids Res_. 24, 4676–4683 (1996). Article  CAS  PubMed  Google Scholar  *

Schalch, T. et al. High-affinity binding of Chp1 chromodomain to K9 methylated histone H3 is required to establish centromeric heterochromatin. _Mol. Cell_ 34, 36–46 (2009). Article  CAS 

PubMed  Google Scholar  * Ishimoto, K. et al. Ubiquitination of Lysine 867 of the Human SETDB1 Protein Upregulates Its Histone H3 Lysine 9 (H3K9) Methyltransferase Activity. _PLoS One_ 11,

e0165766 (2016). Article  PubMed  Google Scholar  * Sun, L. & Fang, J. E3-Independent Constitutive Monoubiquitination Complements Histone Methyltransferase Activity of SETDB1. _Mol.

Cell_ 62, 958–966 (2016). Article  CAS  PubMed  Google Scholar  * Iglesias, N. et al. Native Chromatin Proteomics Reveals a Role for Specific Nucleoporins in Heterochromatin Organization and

Maintenance. _Mol. Cell_ 77, 51–66.e58 (2020). Article  CAS  PubMed  Google Scholar  * Hoeller, D. et al. E3-independent monoubiquitination of ubiquitin-binding proteins. _Mol. Cell_ 26,

891–898 (2007). Article  CAS  PubMed  Google Scholar  * Akoury, E. et al. Disordered region of H3K9 methyltransferase Clr4 binds the nucleosome and contributes to its activity. _Nucleic

Acids Res_. 47, 6726–6736 (2019). Article  CAS  PubMed  Google Scholar  * Iglesias, N. et al. Automethylation-induced conformational switch in Clr4 (Suv39h) maintains epigenetic stability.

_Nature_ 560, 504–508 (2018). Article  ADS  CAS  PubMed  Google Scholar  * Zhang, K., Mosch, K., Fischle, W. & Grewal, S. I. Roles of the Clr4 methyltransferase complex in nucleation,

spreading and maintenance of heterochromatin. _Nat. Struct. Mol. Biol._ 15, 381–388 (2008). Article  CAS  PubMed  Google Scholar  * Ishida, M. et al. Intrinsic nucleic acid-binding activity

of Chp1 chromodomain is required for heterochromatic gene silencing. _Mol. Cell_ 47, 228–241 (2012). Article  CAS  PubMed  Google Scholar  * Kagansky, A. et al. Synthetic heterochromatin

bypasses RNAi and centromeric repeats to establish functional centromeres. _Science_ 324, 1716–1719 (2009). Article  ADS  CAS  PubMed  Google Scholar  * Holla, S. et al. Positioning

Heterochromatin at the Nuclear Periphery Suppresses Histone Turnover to Promote Epigenetic Inheritance. _Cell_ 180, 150–164.e115 (2020). Article  CAS  PubMed  Google Scholar  * Ragunathan,

K., Jih, G. & Moazed, D. Epigenetics. Epigenetic inheritance uncoupled from sequence-specific recruitment. _Science_ 348, 1258699 (2015). Article  PubMed  Google Scholar  * Shan, C. M.

et al. A histone H3K9M mutation traps histone methyltransferase Clr4 to prevent heterochromatin spreading. _Elife_ 5, https://doi.org/10.7554/eLife.17903 (2016). * Bayne, E. H. et al. Stc1:

a critical link between RNAi and chromatin modification required for heterochromatin integrity. _Cell_ 140, 666–677 (2010). Article  CAS  PubMed  Google Scholar  * Ban, H., Sun, W., Chen, Y.

H., Chen, Y. & Li, F. Dri1 mediates heterochromatin assembly via RNAi and histone deacetylation. _Genetics_ 218, https://doi.org/10.1093/genetics/iyab032 (2021). * Shin, Y. &

Brangwynne, C. P. Liquid phase condensation in cell physiology and disease. _Science_ 357, https://doi.org/10.1126/science.aaf4382 (2017). * Alberti, S., Gladfelter, A. & Mittag, T.

Considerations and Challenges in Studying Liquid-Liquid Phase Separation and Biomolecular Condensates. _Cell_ 176, 419–434 (2019). Article  CAS  PubMed  Google Scholar  * Kroschwald, S. et

al. Promiscuous interactions and protein disaggregases determine the material state of stress-inducible RNP granules. _Elife_ 4, e06807 (2015). Article  PubMed  PubMed Central  Google

Scholar  * Canzio, D. et al. Chromodomain-mediated oligomerization of HP1 suggests a nucleosome-bridging mechanism for heterochromatin assembly. _Mol. Cell_ 41, 67–81 (2011). Article  CAS 

PubMed  Google Scholar  * Hinde, E., Cardarelli, F. & Gratton, E. Spatiotemporal regulation of Heterochromatin Protein 1-alpha oligomerization and dynamics in live cells. _Sci. Rep._ 5,

12001 (2015). Article  ADS  CAS  PubMed  Google Scholar  * Müller-Ott, K. et al. Specificity, propagation, and memory of pericentric heterochromatin. _Mol. Syst. Biol._ 10, 746 (2014).

Article  PubMed  Google Scholar  * Elbaum-Garfinkle, S. et al. The disordered P granule protein LAF-1 drives phase separation into droplets with tunable viscosity and dynamics. _Proc. Natl

Acad. Sci. USA_ 112, 7189–7194 (2015). Article  ADS  CAS  PubMed  Google Scholar  * Maharana, S. et al. RNA buffers the phase separation behavior of prion-like RNA binding proteins.

_Science_ 360, 918–921 (2018). Article  ADS  CAS  PubMed  Google Scholar  * Van Lindt, J. et al. F/YGG-motif is an intrinsically disordered nucleic-acid binding motif. _RNA Biol._ 19,

622–635 (2022). Article  PubMed  Google Scholar  * Oya, E. et al. H3K14 ubiquitylation promotes H3K9 methylation for heterochromatin assembly. _EMBO Rep._ 20, e48111 (2019). Article  CAS 

PubMed  Google Scholar  * Aagaard, L. et al. Functional mammalian homologues of the Drosophila PEV-modifier Su(var)3-9 encode centromere-associated proteins which complex with the

heterochromatin component M31. _Embo J._ 18, 1923–1938 (1999). Article  CAS  PubMed  Google Scholar  * Yamamoto, K. & Sonoda, M. Self-interaction of heterochromatin protein 1 is required

for direct binding to histone methyltransferase, SUV39H1. _Biochem Biophys. Res. Commun._ 301, 287–292 (2003). Article  CAS  PubMed  Google Scholar  * Sanulli, S. et al. HP1 reshapes

nucleosome core to promote phase separation of heterochromatin. _Nature_ 575, 390–394 (2019). Article  ADS  CAS  PubMed  Google Scholar  * Pidoux, A. L., Uzawa, S., Perry, P. E., Cande, W.

Z. & Allshire, R. C. Live analysis of lagging chromosomes during anaphase and their effect on spindle elongation rate in fission yeast. _J. Cell Sci._ 113, 4177–4191 (2000). Article  CAS

  PubMed  Google Scholar  * Braun, S. et al. The Cul4-Ddb1(Cdt)² ubiquitin ligase inhibits invasion of a boundary-associated antisilencing factor into heterochromatin. _Cell_ 144, 41–54

(2011). Article  CAS  PubMed  Google Scholar  * Trewick, S. C., Minc, E., Antonelli, R., Urano, T. & Allshire, R. C. The JmjC domain protein Epe1 prevents unregulated assembly and

disassembly of heterochromatin. _Embo j._ 26, 4670–4682 (2007). Article  CAS  PubMed  Google Scholar  * Zofall, M. & Grewal, S. I. Swi6/HP1 recruits a JmjC domain protein to facilitate

transcription of heterochromatic repeats. _Mol. Cell_ 22, 681–692 (2006). Article  CAS  PubMed  Google Scholar  * Keller, C., Kulasegaran-Shylini, R., Shimada, Y., Hotz, H. R. & Bühler,

M. Noncoding RNAs prevent spreading of a repressive histone mark. _Nat. Struct. Mol. Biol._ 20, 994–1000 (2013). Article  CAS  PubMed  Google Scholar  * Kim, H. S. et al. Identification of a

BET family bromodomain/casein kinase II/TAF-containing complex as a regulator of mitotic condensin function. _Cell Rep._ 6, 892–905 (2014). Article  CAS  PubMed  Google Scholar  * Wang, J.

et al. Epe1 recruits BET family bromodomain protein Bdf2 to establish heterochromatin boundaries. _Genes Dev._ 27, 1886–1902 (2013). Article  CAS  PubMed  Google Scholar  * Komander, D.,

Clague, M. J. & Urbé, S. Breaking the chains: structure and function of the deubiquitinases. _Nat. Rev. Mol. Cell Biol._ 10, 550–563 (2009). Article  CAS  PubMed  Google Scholar  *

Leznicki, P. & Kulathu, Y. Mechanisms of regulation and diversification of deubiquitylating enzyme function. _J. Cell Sci._ 130, 1997–2006 (2017). Article  CAS  PubMed  Google Scholar  *

Reyes-Turcu, F. E., Zhang, K., Zofall, M., Chen, E. & Grewal, S. I. Defects in RNA quality control factors reveal RNAi-independent nucleation of heterochromatin. _Nat. Struct. Mol.

Biol._ 18, 1132–1138 (2011). Article  CAS  PubMed  Google Scholar  * Ryan, C. J. et al. Hierarchical modularity and the evolution of genetic interactomes across species. _Mol. Cell_ 46,

691–704 (2012). Article  CAS  PubMed  Google Scholar  * Liu, Y. et al. Functional characterization of the Arabidopsis ubiquitin-specific protease gene family reveals specific role and

redundancy of individual members in development. _Plant J._ 55, 844–856 (2008). Article  CAS  PubMed  Google Scholar  * Bühler, M., Spies, N., Bartel, D. P. & Moazed, D. TRAMP-mediated

RNA surveillance prevents spurious entry of RNAs into the Schizosaccharomyces pombe siRNA pathway. _Nat. Struct. Mol. Biol._ 15, 1015–1023 (2008). Article  PubMed  Google Scholar  * Kuzdere,

T. et al. Differential phosphorylation of Clr4(SUV39H) by Cdk1 accompanies a histone H3 methylation switch that is essential for gametogenesis. _EMBO Rep_. e55928,

https://doi.org/10.15252/embr.202255928 (2022). * Babu, M. M. The contribution of intrinsically disordered regions to protein function, cellular complexity, and human disease. _Biochem Soc.

Trans._ 44, 1185–1200 (2016). Article  CAS  PubMed  Google Scholar  * Bah, A. & Forman-Kay, J. D. Modulation of Intrinsically Disordered Protein Function by Post-translational

Modifications. _J. Biol. Chem._ 291, 6696–6705 (2016). Article  CAS  PubMed  Google Scholar  * Bah, A. et al. Folding of an intrinsically disordered protein by phosphorylation as a

regulatory switch. _Nature_ 519, 106–109 (2015). Article  ADS  CAS  PubMed  Google Scholar  * Oldfield, C. J. & Dunker, A. K. Intrinsically disordered proteins and intrinsically

disordered protein regions. _Annu Rev. Biochem_ 83, 553–584 (2014). Article  CAS  PubMed  Google Scholar  * Qamar, S. et al. FUS Phase Separation Is Modulated by a Molecular Chaperone and

Methylation of Arginine Cation-π Interactions. _Cell_ 173, 720–734.e715 (2018). Article  CAS  PubMed  Google Scholar  * Stowell, J. A. W. et al. A low-complexity region in the YTH domain

protein Mmi1 enhances RNA binding. _J. Biol. Chem._ 293, 9210–9222 (2018). Article  CAS  PubMed  Google Scholar  * Yang, P. et al. G3BP1 Is a Tunable Switch that Triggers Phase Separation to

Assemble Stress Granules. _Cell_ 181, 325–345.e328 (2020). Article  CAS  PubMed  Google Scholar  * Keller, C. et al. HP1(Swi6) mediates the recognition and destruction of heterochromatic

RNA transcripts. _Mol. Cell_ 47, 215–227 (2012). Article  CAS  PubMed  Google Scholar  * Johnson, W. L. et al. RNA-dependent stabilization of SUV39H1 at constitutive heterochromatin. _Elife_

6, https://doi.org/10.7554/eLife.25299 (2017). * Shirai, A. et al. Impact of nucleic acid and methylated H3K9 binding activities of Suv39h1 on its heterochromatin assembly. _Elife_ 6,

https://doi.org/10.7554/eLife.25317 (2017). * Velazquez Camacho, O. et al. Major satellite repeat RNA stabilize heterochromatin retention of Suv39h enzymes by RNA-nucleosome association and

RNA:DNA hybrid formation. _Elife_ 6, https://doi.org/10.7554/eLife.25293 (2017). * Gerace, E. L., Halic, M. & Moazed, D. The methyltransferase activity of Clr4Suv39h triggers RNAi

independently of histone H3K9 methylation. _Mol. Cell_ 39, 360–372 (2010). Article  CAS  PubMed  Google Scholar  * Bühler, M., Verdel, A. & Moazed, D. Tethering RITS to a nascent

transcript initiates RNAi- and heterochromatin-dependent gene silencing. _Cell_ 125, 873–886 (2006). Article  PubMed  Google Scholar  * Al-Sady, B., Madhani, H. D. & Narlikar, G. J.

Division of labor between the chromodomains of HP1 and Suv39 methylase enables coordination of heterochromatin spread. _Mol. Cell_ 51, 80–91 (2013). Article  CAS  PubMed  Google Scholar  *

Chang, A. Y., Castel, S. E., Ernst, E., Kim, H. S. & Martienssen, R. A. The Conserved RNA Binding Cyclophilin, Rct1, Regulates Small RNA Biogenesis and Splicing Independent of

Heterochromatin Assembly. _Cell Rep._ 19, 2477–2489 (2017). Article  CAS  PubMed  Google Scholar  * Bähler, J. et al. Heterologous modules for efficient and versatile PCR-based gene

targeting in Schizosaccharomyces pombe. _Yeast_ 14, 943–951 (1998). Article  PubMed  Google Scholar  * Murray, J. M., Watson, A. T. & Carr, A. M. Transformation of Schizosaccharomyces

pombe: Lithium Acetate/ Dimethyl Sulfoxide Procedure. _Cold Spring Harb. Protoc._ 2016, https://doi.org/10.1101/pdb.prot090969 (2016). Article  PubMed  Google Scholar  * Winston, F. EMS and

UV mutagenesis in yeast. _Curr. Protoc. Mol. Biol._ CHAPTER 13, Unit 13.13B (2008). Google Scholar  * Roche, B., Arcangioli, B. & Martienssen, R. A. RNA interference is essential for

cellular quiescence. _Science_ 354, https://doi.org/10.1126/science.aah5651 (2016). * Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. _Nat. Methods_ 9, 357–359

(2012). Article  CAS  PubMed  Google Scholar  * Garrison, E. & Marth, G. Haplotype-based variant detection from short-read sequencing. _arXiv e-prints_, arXiv:1207.3907 (2012).

<https://ui.adsabs.harvard.edu/abs/2012arXiv1207.3907G>. * Langmead, B., Trapnell, C., Pop, M. & Salzberg, S. L. Ultrafast and memory-efficient alignment of short DNA sequences to

the human genome. _Genome Biol._ 10, R25 (2009). Article  PubMed  PubMed Central  Google Scholar  * Quinlan, A. R. & Hall, I. M. BEDTools: a flexible suite of utilities for comparing

genomic features. _Bioinformatics_ 26, 841–842 (2010). Article  CAS  PubMed  Google Scholar  * Keller, A., Nesvizhskii, A. I., Kolker, E. & Aebersold, R. Empirical statistical model to

estimate the accuracy of peptide identifications made by MS/MS and database search. _Anal. Chem._ 74, 5383–5392 (2002). Article  CAS  PubMed  Google Scholar  * Nesvizhskii, A. I., Keller,

A., Kolker, E. & Aebersold, R. A statistical model for identifying proteins by tandem mass spectrometry. _Anal. Chem._ 75, 4646–4658 (2003). Article  CAS  PubMed  Google Scholar  * Wang,

H. et al. Rubisco condensate formation by CcmM in β-carboxysome biogenesis. _Nature_ 566, 131–135 (2019). Article  ADS  CAS  PubMed  Google Scholar  Download references ACKNOWLEDGEMENTS We

thank all Martienssen lab members as well as Danesh Moazed and Mo Motamedi for valuable discussions and suggestions. The authors acknowledge assistance from the Cold Spring Harbor Laboratory

Shared Resources, which are funded in part by the Cancer Center Support Grant (5PP30CA045508). This work was supported by the Howard Hughes Medical Institute, and a grant from the National

Institute of Health (R35GM144206) (to R.A.M.). AUTHOR INFORMATION Author notes * Benjamin Roche Present address: University of North Dakota, School of Medicine & Health Sciences, 1301 N

Columbia Rd. Stop 9037, Grand Forks, ND, 58202, USA AUTHORS AND AFFILIATIONS * Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 11724, USA Hyun-Soo Kim, Benjamin Roche, Sonali

Bhattacharjee, An-Yun Chang, Christopher Hammell & Robert A. Martienssen * Howard Hughes Medical Institute, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, NY, 11724, USA

Hyun-Soo Kim & Robert A. Martienssen * Institute for Advanced Biosciences, UMR InsermU1209/CNRS5309/UGA, University of Grenoble Alpes, Grenoble, France Leila Todeschini & André

Verdel Authors * Hyun-Soo Kim View author publications You can also search for this author inPubMed Google Scholar * Benjamin Roche View author publications You can also search for this

author inPubMed Google Scholar * Sonali Bhattacharjee View author publications You can also search for this author inPubMed Google Scholar * Leila Todeschini View author publications You can

also search for this author inPubMed Google Scholar * An-Yun Chang View author publications You can also search for this author inPubMed Google Scholar * Christopher Hammell View author

publications You can also search for this author inPubMed Google Scholar * André Verdel View author publications You can also search for this author inPubMed Google Scholar * Robert A.

Martienssen View author publications You can also search for this author inPubMed Google Scholar CONTRIBUTIONS H.S.K. performed experiments and bioinformatics. B.R. contributed sRNA-seq

experiments and bioinformatics. A.Y.C. contributed to EMS mutagenesis, C.H. contributed to LLPS. S.B. LT and AV contributed to protein purification. H.S.K. and R.A.M. designed the

experiments and wrote the manuscript. CORRESPONDING AUTHOR Correspondence to Robert A. Martienssen. ETHICS DECLARATIONS COMPETING INTERESTS The authors declare no competing interests. PEER

REVIEW PEER REVIEW INFORMATION _Nature Communications_ thanks the anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available. ADDITIONAL

INFORMATION PUBLISHER’S NOTE Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. SUPPLEMENTARY INFORMATION SUPPLEMENTARY

INFORMATION PEER REVIEW FILE REPORTING SUMMARY RIGHTS AND PERMISSIONS OPEN ACCESS This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use,

sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative

Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated

otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds

the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Reprints and

permissions ABOUT THIS ARTICLE CITE THIS ARTICLE Kim, HS., Roche, B., Bhattacharjee, S. _et al._ Clr4SUV39H1 ubiquitination and non-coding RNA mediate transcriptional silencing of

heterochromatin via Swi6 phase separation. _Nat Commun_ 15, 9384 (2024). https://doi.org/10.1038/s41467-024-53417-9 Download citation * Received: 21 March 2023 * Accepted: 02 October 2024 *

Published: 30 October 2024 * DOI: https://doi.org/10.1038/s41467-024-53417-9 SHARE THIS ARTICLE Anyone you share the following link with will be able to read this content: Get shareable link

Sorry, a shareable link is not currently available for this article. Copy to clipboard Provided by the Springer Nature SharedIt content-sharing initiative