Structural mechanism of H3K27 demethylation and crosstalk with heterochromatin markers

Chien-Chu Lin; Yani Zhao; Caroline A Foley; Aspen T Hawkins; Lindsey I James; Stephen V Frye; Robert K McGinty

doi:10.1016/j.molcel.2025.06.025

. Author manuscript; available in PMC: 2026 Jul 18.

Published in final edited form as: Mol Cell. 2025 Jul 18;85(15):2869–2884.e6. doi: 10.1016/j.molcel.2025.06.025

Structural mechanism of H3K27 demethylation and crosstalk with heterochromatin markers

Chien-Chu Lin ¹, Yani Zhao ¹, Caroline A Foley ¹, Aspen T Hawkins ², Lindsey I James ^1,³, Stephen V Frye ^1,³, Robert K McGinty ^1,^2,^3,^4,^*

PMCID: PMC12333602 NIHMSID: NIHMS2098826 PMID: 40683254

SUMMARY

Histone H3 lysine 27 trimethylation (H3K27me3) is a repressive histone modification that is a hallmark of facultative heterochromatin. H3K27me3 is installed by the Polycomb Repressive Complex 2 (PRC2) and removed by KDM6 family Jumonji C (JmjC) domain demethylases. Structural studies have elucidated how PRC2 functions on nucleosomes and its regulation by local histone modification signatures. However, the molecular mechanisms governing H3K27 demethylation to reactivate silenced chromatin remain poorly understood. Here, we report the cryo-EM structure of mouse KDM6B bound to the nucleosome. Our structure shows how KDM6B engages wrapped nucleosomal DNA together with both extranucleosomal DNA linkers to position its catalytic JmjC domain for H3K27 demethylation. KDM6B induces an overlapped linker DNA conformation consistent with function in a compact chromatin environment. We further show that linker histones and H2AK119ub1, both enriched in heterochromatin, antagonize KDM6B function, suggesting that linker histone eviction and H2A deubiquitylation precede H3K27 demethylation during heterochromatin activation.

Keywords: lysine demethylase, KDM6B, H3K27me3, nucleosome, histone, chromatin, epigenetics, cryo-EM, heterochromatin, post-translation modification

Grapgical Abstract

graphic file with name nihms-2098826-f0001.jpg

eTOC Blurb

Chien-Chu Lin et al. reveal the structure of the histone demethylase KDM6B bound to the nucleosome. The study shows how KDM6B is positioned for H3K27 demethylation and how heterochromatin features such as H2AK119ub1 and linker histones antagonize KDM6B function.

INTRODUCTION

Histone lysine post-translational modifications (PTMs), such as acetylation, methylation, and ubiquitylation, influence gene expression through the recruitment or exclusion of chromatin-interacting proteins and by altering local chromatin structure^1,2. Characteristic combinations of lysine PTMs demarcate distinct chromatin states. For example, facultative heterochromatin is defined by extended regions enriched in H3K27me3 and H2AK119ub1³. This histone PTM signature is associated with dense chromatin packing and transcriptional repression that is central to developmental gene regulation³. H3K27me3 is installed by the Polycomb Repressive Complex 2 (PRC2)⁴. PRC2 is activated by H2AK119ub1 and its H3K27me3 product allowing for local spreading and epigenetic maintenance^5–7. Reactivation of facultative heterochromatin requires reversal of H3K27me3 and H2AK119ub1⁸.

H3K27 is demethylated by the KDM6 family demethylases, comprising three members, KDM6A (UTX), KDM6B (JMJD3), and KDM6C (UTY). The KDM6 family demethylases are JmjC domain oxygenases that function through an Fe(II)-, α-ketoglutarate-dependent mechanism. All three KDM6 proteins catalyze the demethylation of H3K27 with a preference for tri- and dimethylated substrates⁹, but KDM6C has lower intrinsic activity due to substitutions within a conserved substrate recognition site¹⁰. KDM6A and KDM6B can associate with the KMT2D (MLL4) methyltransferase complex^11,12, linking H3K27 demethylation with H3K4 monomethylation and enhancer activation. Further supporting this connection, KMT2D and KDM6A mutations are causative in Kabuki syndrome, a rare multisystem congenital disorder with characteristic facial features, short stature, and varying levels of intellectual disability¹³.

Despite similar enzymatic functions, the KDM6 family demethylases have non-overlapping biological roles and disease associations. KDM6B functions in regulating genes associated with development¹⁴, immune responses¹⁵, and cell differentiation¹⁶. This activity promotes chromatin accessibility, which is essential for the activation of lineage-specific genes and cellular reprogramming¹⁷. Dysregulation of KDM6B has been implicated in various diseases, including cancer, neurodegenerative disorders, and inflammatory conditions¹⁸, highlighting its dual role in normal physiology and disease pathogenesis.

Structural analyses of KDM6 family demethylases in complex with H3 tail peptides have established the molecular determinants of H3K27 specificity based on local histone sequence^19–21. However, detailed mechanisms underlying KDM6 function on nucleosome substrates remain poorly characterized. To address this, we leveraged a nucleosome-conjugated JmjC inhibitor to stabilize the KDM6B-nucleosome complex, enabling high-resolution structural determination by cryogenic-electron microscopy (cryo-EM). Our structure demonstrates how KDM6B binds to nucleosome-wrapped DNA and both DNA linkers to position the catalytic JmjC domain to target H3K27. KDM6B recognizes the DNA linkers in a compact conformation, setting it up to function in a dense chromatin environment. We further demonstrate that H2AK119ub1 and linker histones inhibit H3K27 demethylation, establishing a biochemical logic for heterochromatin reactivation.

RESULTS

Structural insights into the KDM6B-nucleosome complex

All KDM6 family demethylases share a highly conserved C-terminus which forms a contiguous tertiary structure and includes a Jumonji C (JmjC) catalytic domain, a GATA-like zinc-finger (GATAL) domain, and two alpha-helical regions (αC), the first between the JmjC and GATAL domains and the second immediately following the GATAL domain (Figures 1A and S1A). In contrast, the N-terminal regions of KDM6 family demethylases are more divergent. KDM6A and KDM6C share a tetratricopeptide repeat (TPR) domain tethered to the catalytic C-terminus through a ~600 residue disordered sequence. The ~1100 residue KDM6B N-terminus is predicted to be mostly disordered and is enriched in prolines and polar and charged residues (Figures S1A and S1B). While structures of the JmjC-αC-GATAL regions of KDM6A and KDM6B in complex with H3 peptides have shed light on mechanisms of H3K27 specificity based on local sequence recognition^19,20, how KDM6 family demethylases function in a physiologically relevant nucleosome context is unclear. To investigate the structural basis of nucleosomal H3K27 demethylation, we selected KDM6B as a representative KDM6 family enzyme and prepared its conserved C-terminal region (mouse KDM6B residues 1130–1641, hereafter KDM6B) (Figures 1A, S1A, and S1C). We next used electrophoretic mobility shift assays (EMSA) to determine the optimal nucleosome composition for structural studies. KDM6B exhibited higher affinity for nucleosomes with symmetric 20 bp DNA linkers (20N20) compared to those with 1 bp DNA linkers (1N1) (Figure S1D). This finding suggests that in addition to direct recognition of the H3 N-terminal tail, KDM6B may also interact with linker DNA during nucleosome binding.

Figure 1. — (A) Domain organization of mouse KDM6B, region used in this study indicated with a dashed box. (B) Schematic representation of H3K27C-UNC8015 covalent conjugate bound to Fe(II) in KDM6B JmjC active site. (C) Three views of cryo-EM composite density map of KDM6B-nucleosome complex. (D) Identical views of atomic model of KDM6B-nuclesome complex. Map and model are colored as indicated with KDM6B coloring based on domain schematic in panel A. See also Figures S1–S4.

We recently reported a cryo-EM structure of the H3K36 demethylase KDM2A bound to the nucleosome using an inhibitor-histone conjugate to trap the JmjC domain in a catalytically relevant conformation²². This strategy involved installation of a unique cysteine at the H3K36 position, allowing site-specific conjugation of an α-ketoglutarate analog, UNC8015²². Given the expected promiscuity of UNC8015 for JmjC domain demethylases, we wondered whether this strategy could be adapted for KDM6B simply by changing the location of the H3 cysteine mutant. To test this hypothesis, we prepared an H3K27C/C110A mutant histone and alkylated it with UNC8015, yielding the H3K27C/C110A-UNC8015 conjugate with greater than 60% efficiency (Figures 1B, S2A, and S2B). H3K27C/C110A-UNC8015 was then assembled into a 20N20 nucleosome and subsequently reconstituted with KDM6B for cryo-EM analysis (Figure S2C). Following 2D classification of cryo-EM data, we observed particles with additional density near the DNA entry/exit site on the nucleosome and the location of the H3 N-terminal tail, consistent with the presence of KDM6B (Figures S3A and S3B).

We used three independent cryo-EM datasets to reconstruct a composite map of the 1:1 KDM6B–nucleosome complex (Figues 1C, S3A–S3G, and Table 1). The composite map was calculated using multibody refinement with masks around 1) the nucleosome core and one DNA linker (resolution range of 3.0–7.0 Å), 2) KDM6B and the other DNA linker (4.8–7.6 Å), together with the overall consensus map of complex (3.16 Å resolution) (Figures S3D–S3F). Our EM map contained sufficient structural features to model the nucleosome, including both 20 bp DNA linkers and one H3 N-terminal tail, and the KDM6B JmjC-αC-GATAL region (Figures 1D and S4A–S4H). The structure reveals that KDM6B engages the nucleosome through multivalent interactions involving both the cis-linker DNA (emerging from the same side of the nucleosome as the targeted H3 N-terminal tail and annotated with negative bp positions), the trans-linker DNA (emerging from the opposite side of the nucleosome and annotated with positive bp positions), wrapped DNA near the nucleosome dyad, and a single H3 N-terminal tail (Figures 1C–1D). In the KDM6B-nucleosome complex, the nucleosome is fully wrapped by DNA and the cis- and trans-linker DNAs are held in a crossed, closed conformation. The H3 N-terminal tail emerging from the nucleosome on the cis-linker DNA side, is directed toward the active site of the JmjC domain and loops back toward the cis-linker DNA in an outside-in hook conformation (Figures 1C, 1D, and S4I). Notably, no additional cryo-EM density was observed over the acidic patch – a known hotspot for nucleosome binding – suggesting that the structured KDM6B C-terminus binds the nucleosome in an acidic patch-independent manner (Figures 1C and S4J).

Table 1.

Cryo-EM data collection, refinement and validation statistics

	KDM6B-nucleosome EMDB-49676 PDB 9NQU
Data collection and processing
Microscope	Talos Arctica
Magnification	45,000x
Voltage (kV)	200
Electron exposure (e−/ Å²)	55
Defocus range (mm)	−0.5 to −3.0
Pixel size (Å)	0.876
Symmetry imposed	C1
Micrographs (no.)	24,562
Initial particle images (no.)	2,417,051
Final particle images (no.)	43,458
Map resolution (Å)	3.16 (overall), 3.2 (focused; NCP+trans-linker), 4.46 (focused; KDM6B+cis-linker)
FSC threshold	0.143

Refinement
Initial model used (PDB code)	6ESF, 6F6D
Model resolution (Å)	3.18
FSC threshold	0.5
Map sharpening B factors (Å²)	−51.7 (overall), −67.4 (focused; NCP), −118.7 (focused; KDM6B cis-linker), NA (composite)
Model composition
Non-hydrogen atoms	17359
Protein residues	1229
Nucleotides	370
Ligands	3
B factors (Å²)
Protein	37.55
Nucleotides	85.65
Ligand	65.74
R.m.s. deviations
Bond lengths (Å)	0.005
Bond angles (°)	0.755
Ramachandran plot
Favored (%)	92.31
Allowed (%)	7.69
Disallowed (%)	0.00

Open in a new tab

Linker DNA-dependent nucleosome binding and demethylation

The KDM6B-linker DNA interactions are mediated by positively charged surfaces of the KDM6B JmjC, αC, and GATAL domains (Figure 2A). Specifically, the αC and GATAL domains interact with the proximal region of the cis-linker DNA, immediately adjacent to super-helical location −7 (SHL −7). In this location, αC and GATAL function as a clamp that engages both the major and minor grooves of the cis-linker DNA (Figure S5A). The JmjC domain and the opposite end of the αC domain are positioned adjacent to a segment of the distal trans-linker DNA (Figure 2A). To further explore the role of linker DNA in nucleosome binding by KDM6B, we reconstituted an additional nucleosome with a single extended linker (19N1) and performed quantitative EMSA analysis using 20N20 (two extended linkers), 19N1 (one extended linker), and 1N1 (no extended linkers) nucleosomes. KDM6B exhibited a modestly higher affinity to the 20N20 nucleosome, with a dissociation constant (K_d) of 1.22 ± 0.05 μM (mean ± SEM), than the 19N1 (K_d = 1.83 ± 0.04 μM) or 1N1 nucleosomes (K_d = 2.90 ± 0.02 μM). (Figure 2B). The observation that KDM6B binds nucleosomes with and without linker DNA with similar affinities was surprising given the extensive interactions between KDM6B and the cis- and trans-linkers in our structure and suggests that KDM6B may bind nucleosomes heterogeneously in linker-dependent and -independent modes in our EMSA experiments. We next examined the role of linker DNA on H3K27 demethylation. We used expressed protein ligation to prepare H3K27me3/C110A (hereafter referred to as H3K27me3) (Figure S2D) and assembled the set of nucleosomes described above. KDM6B efficiently demethylated H3K27me3 in 20N20 nucleosomes, leaving less than 5% residual H3K27me3 (Figure 2C). In contrast, KDM6B showed minimal demethylase activity toward 1N1 substrates (90% residual H3K27me3) and intermediate demethylase activity with 19N1 substrates (50% residual H3K27me3). Interestingly, a head-to-head comparison of KDM6B activity on a peptide versus a 20N20 nucleosome revealed that H3K27me3 is more efficiently demethylated in the peptide context (Figure S6). Taken together, our binding and enzymatic assays demonstrate that linker DNA is critical for enzymatic function and overcoming nucleosome suppression of demethylation despite only contributing modestly to global nucleosome binding. This indicates that linker DNA allows KDM6B to bind the nucleosome in a productive conformation that orients the JmjC active site for H3K27me3 recognition. As each nucleosome contains two copies of H3, it is unclear whether the intermediate activity observed for KDM6B on nucleosomes with one extended linker DNA results from the requirement of two linkers for maximal demethylation efficiency of each H3K27me3 or if one linker is required for efficient demethylation of each matched H3 tail.

KDM6B binding induces closed, compact linker DNA structure

In our 2D classification of the KDM6B-20N20 nucleosome complex, linker DNAs on unbound nucleosome classes occupied heterogeneous open conformations consistent with dynamic DNA end breathing (Figure S3B). In KDM6B-bound classes, however, the nucleosomes exhibited a single, defined linker DNA conformation characterized by overlapped ends of the two linker DNAs, thus creating a more compact structure reminiscent of the linker histone containing chromatosome complex (Figures S3B, S7B, and S7C)²³. To confirm that KDM6B binding stably induces this compact linker structure in the absence of crosslinking, we used Fluorescence Resonance Energy Transfer (FRET) to measure the relative distance between linker DNA ends in 20N20 nucleosomes labeled with Cy3 or Cy5 near the end of each of the two linkers, respectively. In the absence of KDM6B, we observed a Cy5/Cy3 ratio of ~3.1 (Figure 2D). Upon addition of KDM6B, the FRET ratio increased to ~4, indicating that KDM6B induces proximity of the DNA linkers, consistent with our EM analysis. In contrast, addition of KDM2A, which we previously reported binds to a nucleosome with unwrapped terminal DNA²², led to a decrease of the FRET ratio to ~1.9. These results demonstrate that KDM6B binding to the nucleosome induces a conformational change that brings the linker DNAs in close proximity, resulting in a more compact chromatin structure.

KDM6B-cis-linker DNA interaction

The KMD6B αC and GATAL domains interact with the cis-linker DNA between positions −1 to −10 bp (Figure 3A). Overall, KDM6B in the nucleosome-bound structure closely aligns with a previously reported crystal structure of mouse KDM6B in complex with an H3K27me3 peptide (PDB 4EZH)²⁰, with an average C-alpha RMSD (root mean square deviation) of 1.2 Å. However, cis-linker DNA binding is accompanied by a conformational rearrangement of one helix and three loops in the KDM6B αC-GATAL region to form the cis-linker DNA clamp (Figure S5A). This DNA clamp includes two bridges which span adjacent major and minor grooves and a major groove insertion loop (Figure 3A). First, a GATAL helical segment (residues 1601–1607) rotates ~5°, directing R1606 and R1607 toward the DNA phosphodiester backbone at cis-linker DNA positions −1’ and −2’. The helix rotation is accompanied by a rearrangement of loop residues 1608–1616 that places the side chain amide of Q1612 within 5 Å of the backbone phosphate on the complementary DNA strand at position −7. Together, the GATAL helix and the adjacent loop form a bridge spanning the major grove.

A nearby loop comprising αC domain residues 1554–1558 simultaneously rearranges upon cis-linker DNA binding to form a bridge across the adjacent minor groove (Figures 3A and S5A). Within this loop, the side chain amines of K1557 and K1558 are positioned to interact with the −6 and −10’ phosphates from complementary DNA strands on both sides of the minor groove. Finally, the GATAL loop formed by residues 1588–1596, which is partially unstructured in the absence of the nucleosome (Figure S5A), inserts into the DNA major groove adjacent to the major groove bridge, wedging in between the cis-linker DNA and the H3 N-terminal tail (Figure 3A). S1588, E1589, N1590, S1592, and N1594 side chains project into the major groove from positions −3 to ‒5. R1593 extends away from the cis-linker to interact with wrapped nucleosomal DNA near the dyad. Notably, mutation of the corresponding arginine in KDM6A (R1351Q) is causative in the craniofacial neurodevelopmental Kabuki syndrome²⁴.

To further explore the functional contributions of the KDM6B-cis-linker DNA interactions, we performed a series of KDM6B charge-swap, alanine, and disease-associated mutations and assessed 20N20 nucleosome binding, H3K27 demethylation, and linker DNA compaction (Figures 3B–3D). The K1557E/K1558E minor groove bridge mutation and K1557E/K1558E/R1607E minor plus major groove bridge mutation significantly blunted nucleosome binding and H3K27 demethylation (Figures 3B and 3C). In contrast, the R1606E/R1607E major groove bridge mutant was only modestly deficient in nucleosome binding but had strongly impaired demethylation. The Q1612A major groove bridge mutation and the E1589A/N1590A/S1592A major groove insertion loop mutation exhibited nucleosome binding and demethylation activities comparable to wild-type KDM6B. Finally, the R1593Q mutation led to a modest reduction in both nucleosome binding affinity and demethylation efficiency, but falling short of significance threshold. DNA linker compaction largely correlated with nucleosome binding, with K1557E/K1558E, K1557E/K1558E/R1607E, and R1593Q mutations abrogating DNA linker compaction as measured by FRET (Figure 3D). R1606E/R1607E, Q1612A, and E1589A/N1590A/S1592A mutations resulted in only slight reductions in FRET signals. Overall, these results indicate that the KDM6B αC-GATAL interactions with cis-linker DNA promote nucleosome binding that induces a compact DNA linker structure and is productive for H3K27 demethylation. The minor groove bridge is critical for nucleosome binding, demethylation, and linker DNA compaction. The major groove bridge is also critical for demethylation but, interestingly, contributes only minimally to nucleosome binding and linker DNA compaction. Finally, much of the major groove insertion loop is dispensable for KDM6B function and rather may be primarily important for positioning R1593 for wrapped DNA binding near the dyad. However, we cannot rule out DNA sequence-specific effects of major groove interactions that are not observed with the 601 nucleosome-positioning sequence used in our studies. Most of the KDM6B residues contributing to cis-linker DNA interactions are conserved in the KDM6 family across evolutionarily distant organisms (Figure S8).

KDM6B-trans-linker DNA interaction

The closed linker DNA conformation allows the compact KDM6B JmjC-αC-GATAL structure to simultaneously engage segments of both the cis- and trans-linker DNA. Unlike the KDM6B-cis-linker DNA interaction, we do not observe side chain resolution for the KMD6B-trans-linker DNA interaction in our composite map, in part due to segmentation of the trans-linker and KDM6B volumes into separate masks during multibody refinement. However, we can make testable predictions about KDM6B-trans-linker interactions based on our model. A concave surface of KDM6B with two prominent basic ridges is positioned adjacent to the trans-linker DNA (Figure 4A). The ridges include positively charged residues K1336, K1339 and R1353 of the JmjC domain (trans-linker interacting ridge 1) and R1521 and K1524 of the αC domain along with K1566 of the GATAL domain (trans-linker interacting ridge 2). These two ridges are poised to interact with a ~9 bp segment of trans-linker DNA from positions +11 to +19 primarily through electrostatic interactions with the DNA backbone (Figures 2A and 4A). To investigate the role of KDM6B-trans-linker DNA interactions in nucleosome binding and H3K27 demethylation, we generated charge-swap mutants of the two basic ridges on KDM6B. We found that K1336E/K1339E/R1353E (ridge 1) and R1521E/K1524E/K1566E (ridge 2) mutant binding to 20N20 nucleosomes was not detectable by EMSA (Figure 4B) and exhibited severely impaired demethylase function (Figure 4C). In linker DNA compaction assays, both ridge mutants showed intermediate FRET values as compared to nucleosomes in the absence and presence of WT KDM6B (Figure 4D). This would be expected if the mutants maintained a fixed cis-linker conformation but released the trans-linker, allowing unconstrained DNA breathing on one side of the nucleosome. All six trans-linker-interacting ridge residues are conserved across KDM6A and KDM6B orthologs (Figure S8), indicating that KDM6A is also likely to rely on trans-linker DNA binding for H3K27 demethylation.

Figure 4. — (A) Full and zoomed views of trans-linker DNA binding by KDM6B trans-DNA interacting ridges. Electrostatic surface (center) and cartoon representation shown (right). ‘ indicates the complementary DNA strand of trans-linker DNA. (B) EMSA using 20N20 nucleosomes and 0, 0.125, 0.25, 0.5, 1, 2, 3, 4, 5, 7.5, 10 and 15 μM concentrations of indicated wild-type (WT) or mutant KDM6B. Native gels stained with ethidium bromide. (C) Demethylase assays with H3K27me3 20N20 nucleosomes using indicated versions of KDM6B. Mean ± s.d. of n=3 demethylase assays. Statistical significance is denoted as ***P = 0.0002, ****P = 0.0001 using one-way ANOVA. (D) FRET assays showing Cy5/Cy3 FRET efficiency ratios for Cy5 and Cy3 labeled 20N20 nucleosome alone (Nuc only) or in the presence of WT or mutant KDM6B (n=3). See also Figures S1, S2, and S8.

KDM6B-wrapped DNA interaction

In addition to cis- and trans-linker DNA binding, we identified a third large interface between KDM6B and wrapped DNA. The side chains of KDM6B JmjC domain residues H1272, K1277, and K1325 form a basic patch that interacts with wrapped DNA at SHL +0.5 to +1, adjacent to the H3 N-terminal tail as it exits the nucleosome core (Figure 5A). To investigate the role of these residues in KDM6B function, we prepared the KDM6B charge-swap mutant H1272E/K1277E/K1325E. Similar to linker DNA charge swap mutants, this wrapped DNA mutant exhibited significantly reduced nucleosome binding affinity and H3K27 demethylation activity (Figures 5B and 5C). Interestingly, in the FRET assay, the H1272E/K1277E/K1325E mutant showed a reduced FRET signal, indicating that wrapped DNA binding by KDM6B is required to facilitate cis- and trans-linker DNA binding and thus linker DNA compaction. Unlike the conformationally flexible cis- and trans-linker DNA fragments, the wrapped DNA is in a fixed conformation prior to KDM6B binding. We hypothesize that the interaction of KDM6B with this stable, wrapped DNA acts as a pivot point, helping to anchor KDM6B to the nucleosome core, allowing binding of the dynamic H3 N-terminal tail and DNA linkers.

Figure 5. — (A) Full and zoomed views of wrapped DNA binding by KDM6B. Electrostatic surface (center) and cartoon representation shown (right). (B) EMSA using 20N20 nucleosomes and 0, 0.125, 0.25, 0.5, 1, 2, 3, 4, 5, 7.5, 10 and 15 μM concentrations of indicated wild-type (WT) or mutant KDM6B. Native gels stained with ethidium bromide. (C) Demethylase assays with H3K27me3 20N20 nucleosomes using indicated versions of KDM6B. Mean ± s.d. of n=3 demethylase assays. Statistical significance is denoted as ****P = 0.0001 using one-way ANOVA. (D) FRET assays showing Cy5/Cy3 FRET efficiency ratios for Cy5 and Cy3 labeled 20N20 nucleosome alone (Nuc only) or in the presence of WT or mutant KDM6B (n=3). See also Figures S1, S2, and S8.

Histone H1 and H2AK119 ubiquitylation modulate KDM6B activity

Given that H3K27me3 is a primary epigenetic marker for facultative heterochromatin, we next wanted to test how chromatin composition and epigenetic state influence KDM6B function. We focused on two additional important markers of facultative heterochromatin: the linker histone and H2AK119 monoubiquitylation (H2AK119ub1). This would allow us to define a biochemical logic of linker histone eviction, H2AK119 deubiquitylation, and H3K27 demethylation in the epigenetic reactivation of heterochromatin loci. Linker histones are enriched in both constitutive and facultative heterochromatin³ and linker histone occupancy is positively correlated with the size and density of nucleosome clusters, called clutches^25,26. The linker histone has been shown to promote H3K27me3 in reconstituted systems^27,28 and restoration of H3K27me3 during DNA replication²⁹. However, it is unclear if linker histone occupancy modulates H3K27 demethylation.

Structural and biochemical studies of linker histones bound to single nucleosomes (chromatosomes) and nucleosome arrays demonstrate that linker histones exhibit either on-dyad or off-dyad nucleosome binding modes that are specified by the sequence of the linker histone. The on-dyad mode is characterized by binding of the linker histone globular domain to nucleosomal DNA at the dyad along with interaction with both DNA linkers to stabilize a closed, compact linker configuration (Figure S7B). In the off-dyad mode, the linker histone is shifted to a position next to the dyad leading to a small reorientation of the linker DNA. At first glance, the linker DNA positioning in the on-dyad binding mode appears to closely resemble our KDM6B-nucleosome structure. However, alignment of an H1.0B-nucleosome structure with our KDM6B-nucleosome structure shows that KDM6B induces a more tightly crossed linker DNA conformation when viewing down onto the nucleosome disk (Figure 6A). Moreover, profile views of the nucleosome show that the linkers are splayed farther apart in the KDM6B structure, largely due to a kink in the cis-linker DNA (Figure 6B). The globular domain of H1.0B also sterically clashes with the outside-in hook conformation of the H3 N-terminal tail. In order to determine the effect of linker histones on KDM6B function, we prepared the globular domain of H1.5 (gH1.5) (Figure S9A), a ubiquitously expressed replication-dependent linker histone that has been shown to bind nucleosomes using an on-dyad mode²³. Titration of gH1.5 into H3K27me3 nucleosomes resulted in a dose-dependent inhibition of KDM6B demethylase function (Figure 6C). This result shows that linker histone binding antagonizes H3K27 demethylation by KDM6B and suggests that linker histone eviction must precede H3K27me3 removal.

Figure 6. — (A and B) Two views of superimposition of the cryo-EM structures of the KDM6B-nucleosome complex (this study, green DNA) and the linker histone H1.0B-nucleosome chromatosome complex (PDB: 5NL0, blue DNA) highlighting distinct linker DNA conformations. (C) Demethylase assays for H3K27me3 nucleosomes in the absence or presence of increasing concentrations of the globular domain of linker histone H1.5 (gH1.5). Representative western blots detecting H3K27me3 or H3 as a loading control, top. Mean ± s.d. of n=3 demethylase assays, bottom. (D and E) Two views of superimposition of the cryo-EM structures of the KDM6B-nucleosome complex (this study, green DNA) and the H2AK119ub1-nucleosome complex (PDB: 8G6S, red DNA) shows that H2AK119ub1 is positioned near DNA entry/exit site. (F) Demethylase assays for H3K27me3 nucleosomes with or without H2AK119C-DCA-ub1 (dichloroacetone linked monoubiquitin) in the absence or presence of KDM6B. Representative western blots detecting H3K27me3 or H3 as a loading control, top. Mean ± s.d. of n=3 demethylase assays, bottom. Statistical significance is denoted as **P = 0.0045 using unpaired t-test. (G) Single turnover demethylase assay time course using H3K27me3 nucleosomes with or without H2AK119C-DCA-ub1. Mean ± s.d. for n=3 assays shown. See also Figures S1, S7, and S9.

Ubiquitylation of histone H2A at lysine 119 (H2AK119ub1), catalyzed by the Polycomb Repressive Complex 1 (PRC1), is enriched in facultative heterochromatic and facilitates the recruitment of the Polycomb Repressive Complex 2 (PRC2) which installs H3K27me3^5,30,31. Moreover, H2AK119ub1 has been shown to promote linker histone binding³² and enhance H1-mediated chromatin condensation, leading to transcriptional repression³³. H2AK119 is located near the dyad and a recent structural analysis of H2AK119ub1 assembled into 1N1 nucleosomes shows that ubiquitin adopts multiple conformations including a position that would clash with linker DNA in the compact conformation observed in our KDM6B structure (Figures 6D and 6E)³⁴. To examine whether H2AK119ub1 interferes with KDM6B demethylation of nucleosomes, we prepared H2AK119ub1 using DCA (dichloroacetone) crosslinking and assembled 20N20 nucleosomes with both H3K27me3 and H2AK119ub1 (20N20-H2AK119C-DCA-ub1) (Figure S9B). H3K27 demethylation assays showed that H2AK119ub1 reduced the enzymatic activity of KDM6B (Figures 6F and S9C). This reduction in activity was also observed in nucleosomes enzymatically ubiquitylated at H2AK119 by PRC1 (Figures S9B and S9D). Furthermore, in a single-turnover demethylation assay using a large excess of KDM6B (15 μM), the catalytic rate of KDM6B was lower in the presence of H2A119ub1 compared to the unmodified 20N20 nucleosome, although both nucleosomes were ultimately demethylated completely (Figures 6G and S9E). These findings suggest that H2AK119ub1 may sterically hinder KDM6B binding at positions optimal for H3K27 demethylation, thereby reducing its catalytic efficiency. Altogether, our findings suggest that linker histones and H2AK119ub1 may cooperate to inhibit KDM6B-mediated H3K27 demethylation in heterochromatin by restricting KDM6B access to productive chromatin binding conformations.

DISCUSSION

Our KDM6B-nucleosome structure reveals that KDM6B binds to nucleosome-wrapped DNA and both linker DNAs to position the KDM6B catalytic site to interact with the H3 N-terminal tail for H3K27 demethylation (Figure 7A). The KDM6B αC and GATAL domains interact with the proximal half of the cis-linker DNA. The cis-linker interaction results in a conformational rearrangement of one helix and three loops in the nucleosome-bound KDM6B structure as compared to KDM6B-histone peptide structures²⁰. Notably, the GATAL domain of KDM6B adopts a distinct mode of DNA binding when compared to other GATA domains that interact with DNA outside of a nucleosome context (Figure S5B)³⁵. KDM6B simultaneously engages the distal half of the trans-linker DNA through two basic ridges on the JmjC-αC surface. Finally, KDM6B contacts wrapped nucleosomal DNA in two locations. First, a basic surface on the JmjC domain binds nucleosomal DNA near SHL +1.0, directly underneath the location where the H3 N-terminal tail exits the nucleosome core. Second, KDM6B R1593 from the GATAL domain major groove insertion loop contacts wrapped nucleosomal DNA adjacent to the dyad.

Figure 7. — (A) Schematic depicting interactions of KDM6B binding sites on nucleosomal and linker DNA through the αC, GATAL, and JmjC domains, top. Cis-linker DNA (−1 to −20 bp), trans-linker DNA (+1 to +20 bp) and wrapped DNA (superhelical location SHL −7 to +7) indicated. Model of key interactions facilitating KDM6B-mediated H3K27 demethylation in nucleosomes, bottom. (B) Linker histone H1 and H2AK119ub1 disrupt the interaction between KDM6B and linker DNA. However, H1 eviction and H2AK119ub1 deubiquitylation enable KDM6B to bind to a compact nucleosome structure. See also Figures S7 and S10.

Interestingly, we found that KDM6B can bind to nucleosomes with and without linker DNA, albeit with 2-fold higher affinity in the presence of both linkers. This suggests linker-dependent and -independent modes of KDM6B-nucleosome interaction in bulk nucleosome binding assays. Mutations of cis-linker, trans-linker, and wrapped DNA interaction surfaces on KDM6B largely eliminated detectable binding to 20N20 nucleosomes, indicating that each type of DNA binding is critical for nucleosome engagement regardless of binding mode. Despite only minor contributions to overall nucleosome binding affinity, both DNA linkers are required for efficient demethylation of H3K27 in a nucleosome context. Nucleosomes without linkers were poor substrates and intermediate activity was observed when using nucleosomes with a single linker DNA. This intermediate activity could result if demethylation of each histone tail requires a specific linker DNA or if binding both linkers is necessary for efficient demethylation of either tail. We favor the later model given our observations that mutations of the cis-linker or trans-linker interacting surfaces strongly disrupt H3K27 demethylation. In fact, we observed similar degrees of activity loss upon triple charge-swap mutations of wrapped, cis-linker, and trans-linker DNA binding surfaces, indicating that each type of interaction is equally critical for KDM6B function (Figures 3–5).

In binding both cis- and trans-linker DNA segments, KDM6B holds the nucleosome in a compact state, with the nucleosomal DNA fully wrapped and the two linkers crossed over one another. We validated this compact linker conformation using bulk FRET assays. Much like demethylase activity, wrapped, cis-linker, and trans-linker DNA interactions are each necessary for stabilizing the overlapped linker conformation. We previously showed that KDM2A and KDM2B, two H3K36 demethylases, bind nucleosomes with unwrapped terminal DNA (Figure S7E)²². The same has been observed with several H3K36 methyltransferases (Figures S7F and S7G)^36–40. This makes intuitive sense given the requirement for DNA unwrapping for access to the H3K36 site which is near the emergence of the H3 N-terminal tail from the nucleosome disk (Figures S7A). In contrast, H3K27 is much more accessible and does not require DNA unwrapping for demethylation. Interestingly, the PRC2-nucleosome complex adopts a compact linker structure similar to what we observed with KDM6B (Figures S7C and S7D)⁴¹ with the AEBP2 activating cofactor of PRC2 also binding to the cis-linker DNA⁵. In addition to lysine accessibility, the molecular recognition of distinct linker DNA conformations may reflect evolutionary adaptation to local chromatin structures. H3K36 methylation occurs co-transcriptionally and the methyltransferases for this position can target nucleosomes that are transiently unwrapped during the passage of RNA polymerase II³⁶. This process facilitates a more open and relaxed chromatin conformation. In contrast, H3K27 methylation proceeds through a feed-forward spreading mechanism and must be maintained in a more compact heterochromatin environment. This contrast highlights how chromatin-modifying enzymes involved in the regulation of gene expression must function within opposing chromatin states to mediate transcriptional switches. To initiate reactivation of heterochromatic loci, KDM6 family demethylases may need to invade and function within a similarly compact local environment. We showed that KDM6B can induce a compact chromatin state. However, this observation may also represent an evolved preference for more densely packaged heterochromatic structures.

Sequence alignments between KDM6A and KDM6B reveal strong conservation of residues involved in nucleosome-specific interactions essential for demethylation, with the only exceptions being the cis-linker interaction minor and major groove bridges (R1558 and Q1612) (Extended Data Fig. 7). This suggests that KDM6A likely binds and demethylates nucleosomes through a mechanism identical to that of KDM6B. In our KDM6B-nucleosome structure, R1593, located within the major groove insertion loop of the GATAL domain, interacts with the phosphodiester backbone of wrapped nucleosomal DNA near the dyad. An R-to-Q mutation at the corresponding position in KDM6A has been implicated in Kabuki syndrome²⁴. Our analysis indicates that the R1593Q mutation leads to a modest reduction in nucleosome binding affinity, linker DNA compaction, and H3K27 demethylation, suggesting that similar defects in KDM6A may contribute to Kabuki syndrome pathogenesis. A key region of sequence divergence between KDM6A and KDM6B is the major groove insertion loop in KDM6A (Figure S8). If this loop confers sequence-dependent properties, its divergence may result in distinct targeting or regulatory mechanisms between the two KDM6 family demethylases. However, a triple alanine mutation of this loop was well tolerated, suggesting that any sequence dependence of KDM6B-nucleosome binding or demethylation is not detectable using the artificial 601 nucleosome positioning sequence. Further investigation is required to determine whether linker DNA sequence preference influences KDM6B-mediated demethylation.

In the KDM6B-nucleosome complex, the H3 N-terminal tail emerges from the nucleosome core, extends toward the active site of the KDM6B JmjC domain and subsequently loops back in toward the nucleosome surface in a hook conformation (Figures 1D and S4I). Given that KDM6A and KDM6B are integrated into the H3K4 targeting MLL4 methyltransferase complex^11,42, and that KDM6B interacts with ASH2L, one core subunit of the MLL complex⁴³, we wondered whether this H3 tail conformation in the presence of KDM6B would also be compatible with simultaneous H3K4 methylation. Similar to KDM6B-peptide structures, we observed a defined conformation of H3 residues 15–34 in our nucleosome bound model. The homologous MLL1 core complex, sharing identical regulatory subunits with MLL4 (WDR5-RBBP5-ASH2L-DPY30), binds to the disk surface of the nucleosome with the catalytic SET domain of MLL1 above the DNA entry/exit site. In the MLL–nucleosome complex without KDM6B, the H3 tail remains flexible preventing modeling, potentially indicating inefficient tail engagement^44–46. Alignment of MLL1- and KDM6B-nucleosome structures suggests that KMD6B can both demethylate H3K27 and redirect the H3 N-terminal tail toward the MLL1 active site (Figure S10). These findings underscore the functional significance of linker DNA as a regulatory structure that facilitates not only chromatin compaction and enzyme recruitment and activity but may also contribute to histone tail trajectories to coordinate histone modification crosstalk.

In addition to H3K27me3, facultative heterochromatin is enriched in linker histones and H2AK119ub1³. As such, activation of heterochromatic loci requires H3K27 demethylation, H2AK119 deubiquitylation, and linker histone eviction. We showed that the linker histone globular domain and H2AK119ub1 both antagonize KDM6B-mediated H3K27 demethylation. Epigenetic signatures at heterochromatin are self-reinforcing. Linker histones and H2AK119ub1 activate PRC2 for H3K27me3 deposition^5,6,28,41. H2AK119ub1 promotes linker histone binding^32,33 and H3K27me3 allosterically activates PRC2 for H3K27me3 spreading^7,47,48. We now show that linker histones and H2AK119ub1 prevent H3K27 demethylation, further stabilizing the epigenetic landscape. We hypothesize that linker histone eviction and H2AK119 deubiquitylation occur prior to KDM6-mediated H3K27 demethylation during reversal of heterochromatin repression. These findings, consistent with the colocalization of H2AK119ub1 and H3K27me3⁴⁹, reveal that beyond its role in recruiting PRC2 for H3K27 methylation, H2AK119ub1 acts as a dual-function modulator of chromatin accessibility and enzymatic activity for KDM6B.

Previous studies have provided critical insights into how KDM6B interacts with histone H3 peptides and small-molecule inhibitors, elucidating the structural basis for the local sequence specificity and inhibitor binding of KDM6B^19,20. Expanding on this foundation, our study uncovers structural and mechanistic insights into how KDM6B binds nucleosomes in a heterochromatin environment. We highlight the essential roles of linker DNA and the functional domains of KDM6B in mediating nucleosome binding, enzymatic activity and nucleosome compaction. Additionally, we demonstrate that heterochromatin-associated factors, including linker histone H1 and H2AK119ub1, inhibit KDM6B activity by disrupting its interactions with linker DNA (Figure 7B). These findings advance our understanding of the molecular mechanisms governing KDM6 family function and provide new perspectives on its role in epigenetic regulation of the genome.

Limitations of the study

Due to sample heterogeneity and the intrinsic flexibility of the KDM6B-nucleosome complex, we were only able to reconstruct a 1:1 KDM6B-nuclesome model, despite observing 2:1 particles during 2D classification. In our EMSA assays, KDM6B-nucleosome complexes resulted in smeared bands, making it difficult to distinguish 1:1, 2:1, and higher binding stoichiometries. The heterogeneity of the complex also required us to use multiple focused classifications and the integration of these focused maps to interpret our cryo-EM data. Local resolution of the maps often limited confident placement of side chains. As such, molecular details of the model were thoroughly validated by KDM6B mutations paired with nucleosome binding and demethylase assays. Finally, our structural and functional studies were performed using a truncated KDM6B construct lacking the N-terminal unstructured region of the protein and thus only reflect interactions of the catalytic domain that is shared by all members of the KDM6 family.

RESOURECE AVAILABILITY

Lead contact

Correspondence and requests for resources and reagents should be directed to the lead contact, Robert K. McGinty (rmcginty@email.unc.edu).

Materials availability

Plasmids are available from Robert K. McGinty upon request.

Data and code availability

The cryo-EM composite map has been deposited in the Electron Microscopy Data Bank under the accession code EMD-49676 and the KDM6B-nucleosome model has been deposited in the Protein Data Bank under the accession code 9NQU and are publicly available as of the date of publication. Uncropped gel and blot images and raw data and calculations related to blot quantification and FRET experiments generated during this study have been deposited into the Mendeley database (https://doi.org/10.17632/gmgxzphgrc.1) and are publicly available as of the date of publication.
This paper does not report original code.
Any additional information required to reanalyze data reported in this paper is available from the lead contact upon request.

STAR METHDS

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

Cell culture

TG1, BL21(DE3)pLysS and HB101 E. coli cells were used for cloning, recombinant expression of proteins, and growth of nucleosomal 601 DNA repeat plasmids, respectively. Competent cells of these strains were prepared from cells restreaked from glycerol stocks originally obtained by Song Tan from Novagen and ATCC, respectively.

METHOD DETAILS

Plasmid construction

The DNA for mouse KDM6B was obtained from the pCS2-Jmjd3-F plasmid (Addgene, 17440). The C-terminal region of KDM6B (residues 1130–1641) was cloned into the pST50Tr expression vector⁶⁰ containing an N-terminal SUMO-10xHis tag and a tobacco etch virus (TEV) cleavage site using Gibson assembly, resulting in the pST50Tr-SMOHSTNmKDM6Bt1 construct. KDM6B mutations were introduced into this construct through site-directed mutagenesis. The DNA encoding the globular domain of histone H1.5 (residues 40–112) was obtained from the pST50Tr-STRaHSTNhH1.5 vector, a gift from Song Tan, and cloned into the pST50Tr vector with an N-terminal SUMO-10xHis tag and TEV cleavage site to generate the pST50Tr-SMOHSTNhH1.5t1 construct. The expression vector pST50Tr-STRaHSTNGS2hUbq, encoding human ubiquitin, was prepared as described previously⁵⁰. Plasmids containing tandem repeats of 165 bp (19N1), 147 bp (1N1), or 185 bp (20N20) Widom 601 DNA were also gifts from Song Tan. The expression vector pST50Tr-hH3.2 was used to generate the H3K27C/C110A mutant and an insertion mutant, H3(1–28)-ENLYFQGG-H3(29–135, A29C/C110A), for expressed protein ligation via site-directed mutagenesis.

Protein preparation

6xHis tagged mouse Uba1 (HismUba1), human UbcH5c (E2), human BMI1–RING1B heterodimer complex, and Strep-10xHIS-TEV-Gly-Ser-Gly-Ser tagged ubiquitin (HSTNGS2hUbq) were prepared identically as previously described⁵⁰. Ubiquitin with a G76C mutation was prepared as previously described⁵¹. KDM6B (residues 1130–1641), histone H1.5 (residues 40–112, gH1.5), and ubiquitin(G76C) were expressed in BL21(DE3)pLysS E. Coli. Cells were grown at 37 °C in 2xYT broth medium, and protein expression was induced with 0.1 mM isopropyl-thio-β-D-galactoside (IPTG) when the culture reached an OD600 of 0.6–0.7 for 16–18 hours at 18 °C. Cell pellets were resuspended in lysis buffer (50 mM sodium phosphate pH 7.0, 300 mM NaCl, 1 mM benzamidine, and 5 mM 2-mercaptoethanol) and lysed by sonication. Cleared lysates were applied to TALON affinity resin (Takara Bio) and proteins were eluted using 200 mM imidazole.

For KDM6B, the eluted protein was treated with TEV protease to remove the SUMO-10xHis tag and dialyzed into a buffer containing 10 mM MES pH 6.5, 100 mM NaCl and 10 mM 2-mercaptoethanol. The cleaved protein was further purified by cation-exchange chromatography using Source S resin (GE Healthcare, 17094405) and size exclusion chromatography on a Superdex 75 increase 10/300 GL column (Cytiva, 29148721) pre-equilibrated with a buffer containing 20 mM HEPES pH 7.5, 200 mM NaCl and 1 mM DTT. The final purified protein was supplemented with 20% glycerol, stored at −80 °C, and subsequently used in the reconstitution of the KDM6B-nucleosome complex for cryo-EM studies. Wild-type and mutant KDM6B proteins were purified following the same protocol, except that the TEV protease digestion step was omitted for proteins used in electrophoretic mobility shift assays, nucleosome demethylation assays, and nucleosomal linker DNA FRET assays. The purity of all proteins was confirmed by SDS-PAGE analysis (Figures S1C and S1E).

For gH1.5, the imidazole-eluted protein from TALON affinity resin was digested with TEV protease and dialyzed into a buffer containing 10 mM HEPES pH 7.5, 100 mM NaCl and 10 mM 2-mercaptoethanol. The digested protein was further purified using Source S resin (GE Healthcare, 17094405) and a Superdex 75 Increase 10/300 GL column (Cytiva, 29148721) pre-equilibrated with a buffer containing 20 mM HEPES pH 7.5, 200 mM NaCl and 1 mM DTT. The purified protein was concentrated, supplemented with 20% glycerol, and stored at −80 °C. The purity of the gH1.5 was confirmed by SDS-PAGE analysis (Figure S9A)

Nucleosome preparation

1N1, 19N1, and 20N20 nucleosomes were reconstituted using recombinant human histones (H2A, H2B, H3.2, and H4) and 601 nucleosome positioning sequences of varying lengths (1N1, 19N1, or 20N20). Briefly, histones were expressed in BL21(DE3)pLysS E. coli, extracted from inclusion bodies, and refolded into H2A/H2B dimers and H3.2/H4 tetramers as previously described⁵². H2A/H2B dimers and H3.2/H4 tetramers were reconstituted by dialysis into refolding buffer (10 mM HEPES pH 7.5, 100 mM NaCl, 10 mM 2-mercaptoethanol) and purified by cation-exchange chromatography using Source S resin (GE Healthcare, 17094405). Nucleosome reconstitution was performed by salt gradient dialysis, transitioning from a buffer containing 10 mM Tris-HCl pH 7.5, 0.5 mM EDTA, and 2000 mM KCl to a buffer containing 10 mM Tris-HCl pH 7.5, 0.5 mM EDTA, and 250 mM KCl, from H2A/H2B dimers, H3.2/H4 tetramers and 601 DNA fragments and then nucleosomes were purified by anion-exchange chromatography using Source Q resin (GE Healthcare, 17094705) and dialyzed into a NCP storage buffer (10 mM potassium cacodylate pH 6.5 and 0,1 mM EDTA), following previously established protocols⁵².

To produce the H3K27C/C110A-UNC8015/20N20 nucleosome for cryo-EM analysis, the H3K27C/C110A histone mutant was alkylated with UNC8015. UNC8015 was prepared identically as previously reported²². Briefly, the H3K27C/C110A histone was dissolved in an alkylation buffer (1 M HEPES pH 7.8, 4 M guanidine-HCl, 10 mM D/L-methionine and 20 mM DTT) and incubated for 1 hour at 37 °C. UNC8015 was added to the reaction at a 50:1 molar ratio to the histone H3K27C/C110A and incubated for 2.5 hours at 37 °C. An additional aliquot of UNC8015 (50:1 molar ratio) was then added and the reaction was allowed to proceed for another 2.5 hours. Following the reaction, the alkylated H3 histone was dialyzed into 2 mM 2-mercaptoethanol and lyophilized. The purity of H3K27C/C110A-UNC8015 was confirmed via LC-MS, with >60% completion of the alkylation reaction (Figure S2B). H3K27C/C110A-UNC8015 was used to reconstitute the H3K27C/C110A-UNC8015/20N20 nucleosome as described above.

H3K27me3/C110A was prepared for nucleosome demethylation assays using a method as previously described⁶¹ with minor modifications. Briefly, H3(1–28)-ENLYFQGG-H3(29–135, A29C/C110A) was expressed from pST50Tr-hH3.2×52 plasmid in BL21(DE3)pLysS E. coli, extracted from inclusion bodies, and purified by reverse phase HPLC using a preparative Vydac C18 column (218TP, 10 μm, 250×22 mm, Mac-Mod Analytical). After lyophilization, the H3 protein was resuspended at a concentration of 6 mg/mL in digestion buffer (50 mM Tris-HCl pH 8.0, and 4 mM DTT). The protein was incubated with 1/100 molar equivalents of TEV protease at room temperature overnight to generate H3(29–135, GG-A29C/C110A) and the truncated H3 was purified by reverse phase HPLC as above. The GlyGly (GG) dipeptide was removed using the Qiagen TAGzyme kit (Qiagen, 34362) following the manufacturer’s protocol for 2 h. Digestion completion was confirmed by LC-MS. Following the digestion, 400 mM methoxylamine-HCl and 20 mM TCEP were added to the reaction mixture, and the truncated H3 was further purified by reverse phase HPLC as above and lyophilized. H3K27me3/C110A was prepared by expressed protein ligation between H3(29–135, A29C/C110A) and a synthetic H3K27me3 peptide (residues 1–28) peptide hydrazide following the established protocol⁶² followed by desulfurization to convert the ligation site cysteine to the native alanine. Briefly, the synthetic H3K27me3 peptide (2 mg, 0.65 μmol) was dissolved in 130 μl of 0.2 M phosphate containing 6 M guanidine-HCl (pH 3.0–3.1). The peptide hydrazide was converted to an azide by addition of 10 molar equivalents of sodium nitrite using a 0.5 M sodium nitrate stock. This reaction was allowed to proceed for 15–20 min at −15 °C in an ice/salt bath. The H3(29–135, A29C/C110A) protein (6 mg, 0.49 μmol) was dissolved in 130 μl of 0.2 M phosphate containing 6 M guanidine-HCl (pH 6.8–7.0) containing 100 molar eq of 4-mercaptophenylacetic acid (MPAA, Sigma, 653152) and added to the peptide. The pH of the ligation mixture was adjusted to 6.8–7.0 and the ligation was allowed to proceed for 24 h at room temperature, affording H3K27me3/A29C/C110A. The H3K27me3/A29C/C110A ligation was analyzed by SDS-PAGE, purified using reverse phase HPLC using a semi-preparative Vydac C18 column (218TP, 5 μm, 250×10 mm, Mac-Mod Analytical), with a 43–53% B gradient (A: 0.1% TFA in water; B: 90% acetonitrile, 0.1% TFA in water) over 45 min, yielding 3.2 mg, confirmed by LC-MS, and lyophilized. The ligation product was desulfurized as previously reported⁶³. In a typical desulfurization reaction, the ligation mixture described above was diluted to 1000 μl with desulfurization buffer (6 M guanidinium-HCl, 200 mM sodium phosphate, pH 7.0). To this solution was added an equal volume of 1 M TCEP in desulfurization buffer with pH adjusted to 5, 200 μL 2-methyl-2-propanethiol, and 100 μl 0.1 M VA-044 in desulfurization buffer. The reaction mixture was incubated overnight at 37 °C. The reaction progress was monitored by LC-MS. After completion, the final product was purified by semipreparative reverse phase HPLC using a Vydac C18 column (218TP, 5 μm, 250×10 mm, Mac-Mod Analytical using the same gradient, yielding approximately 2.5 mg of H3K27me3/C110A. The identity and purity of the product was characterized by LC-MS (Figure S2D) and pool fractions were lyophilized. H3K27me3/C110A was assembled into H3.2/H4 tetramers and subsequently reconstituted into nucleosomes with varying DNA lengths as described above.

To prepare nucleosomes with Cy3 and Cy5 labeled linker DNA for FRET assays, labeled 20N20 601 DNA fragments were generated by PCR using the 185 bp 601 nucleosome positioning sequence as a template. A large-scale PCR reaction (10 mL total volume) was conducted using primers (ATCGC/iCY3-dT/GTTCAATACATGC and ATCCCTA/iCY5-dT/ACGCGGCC, GenScript), where Cy3 and Cy5 fluorophores were conjugated to amino-modified C6-dT at internal positions to prevent end-stacking of fluorophores. The reaction mixture was distributed into a 96-well PCR plate for amplification. Following the PCR reaction, the product was extracted by ethanol precipitation and further purified using anion-exchange chromatography with Source Q resin (GE Healthcare, 17094705). The purified Cy3 and Cy5 labeled 185 bp 601 DNA was then used for nucleosome reconstitution as described above.

The H2AK119ub1-H3K27me3/C110A/20N20 nucleosomes were prepared by both enzymatic nucleosome ubiquitylation and DCA (1,3-dichloracetone) crosslinking. For enzymatic ubiquitylation, 1 μM H3K27me3-C110A/20N20 nucleosome was incubated with 30 nM HismUba1 (E1), 0.2 μM human UbcH5C(E2), 0.2 μM human BMI1–RING1B heterodimer complex (E3), and 5 μM HSTNGS2hUbq (Strep-10xHIS-TEV-Gly-Ser-Gly-Ser tagged ubiquitin) in ubiquitylation buffer (50 mM HEPES pH 7.5, 75 mM NaCl, 2 mM MgCl₂, 10 μM ZnSO₄, 3 mM ATP, and 1 mM DTT). The ubiquitylation reaction was allowed to proceed at 30 °C for 45 min. Following ubiquitylation, the H2AK119ub1-H3K27me3/C110A/20N20 nucleosome was purified by anion-exchange as described above. For DCA crosslinking, a lyophilized H2AK119C/H2B histone dimer and STRHSTNGS2hUbq (G76C) were dissolved in crosslinking buffer (33.3 mM sodium borate, 6 M urea, 4.75 mM TCEP) and incubated at room temperature for 30 minutes. 0.5 μM histone dimer was then mixed with 1 μM STRHSTNGS2hUbq(G76C) and supplemented with 3.25 μM DCA at RT for 1 hour. The reaction was quenched by adding 50 mM 2-mercaptoethanol. The reaction mixture was dialyzed overnight into H100 buffer (10 mM HEPES pH 7.5, 100 mM NaCl) to refold the histone dimer. The H2AK119C-DCA-ub1/H2B dimer was further purified via TALON affinity resin, digested with TEV protease to remove the Strep-10xHis tag, and further purified by cation-exchange chromatography using Source S resin (GE Healthcare, 17094405). The purified histone dimer was then used for 20N20 nucleosome reconstitution with H3K27me3/C110 as described above.

Cryo-EM sample preparation

To prepare the KDM6B-nucleosome complex for cryo-EM studies, 7 molar equivalents of KDM6B (final concentration 22 μM) were incrementally added to the H3K27C/C110A-UNC8015/20N20 nucleosome (final concentration 3.1 μM) in four steps at 5 min intervals in 300 μL reconstitution buffer (10 mM HEPES pH 7.5, 150 mM NaCl and 1 mM DTT). The complex was then gradually diluted with 800 μL of low-salt buffer (10 mM HEPES pH 7.5, 20 mM NaCl and 1 mM DTT) and concentrated using a Vivaspin 500 centrifugal concentrator with a 10 kDa molecular weight cutoff (Vivascience, VS0101). The complex was purified by size exclusion chromatography using a Superdex 200 increase 10/300 GL column (Cytiva, 28990944) equilibrated in low salt buffer. Fractions were analyzed by SDS-PAGE (Figure S2D) and pooled fractions were concentrated to 3.1 mg/mL. For vitrification, the complex was diluted to 0.75 mg/ml in low-salt buffer, and 3 μL was applied to a plasma-cleaned Quantifoil R1.2/1.3, 300-mesh grids. Plasma cleaning was performed with a Tergeo-EM plasma cleaner for 60 seconds. Cryo-EM grids were prepared using a Vitrobot Mark IV (Thermo Fisher Scientific) at 4 °C and 100% humidity, with a 4 s blot time and −10 to 0 blot force using Whatman 595 filter paper. The grids were plunge-frozen into a 40/60% ethane/propane mixture and stored in liquid nitrogen.

Cryo-EM data collection and image processing

Cryo-EM grids were imaged using a Thermo Fisher Scientific Talos Arctica G3 operating at 200 kV, equipped with a Gatan K3 direct electron detector, at the Cryo-EM Core at UNC⁶⁴. Data were collected as 60-frame movies using SerialEM with a multi-shot acquisition in a 5 × 5 regular pattern. Images were recorded at a nominal magnification of 45,000x, corresponding to a pixel size of 0.876 Å, with a defocus range of −0.5 to −3.0 μm and a total dose of 55 e⁻/Å². Three grids were imaged, resulting in three datasets for the KDM6B-nucleosome complex.

A total of 24,562 movies from three datasets of the KDM6B-nucleosome complex were imported into RELION 5.0 beta⁵⁴ and grouped into optics groups based on beam image shifts used during data collection⁶⁵. Motion correction with dose-weighting was performed using MotionCor2⁶⁶, and contrast transfer function (CTF) estimation was carried out with CTFFIND-4.1⁶⁷. Micrographs with an estimated resolution lower than 5 Å were excluded from further processing. Particle templates generated from RELION’s Laplacian-of-Gaussian particle picking on dataset #2 were utilized for templated particle picking across all three datasets. After four rounds of 2D classification and one round of 3D classification, 947,666 particles belonging to the KDM6B-nucleosome complex (from class 2 of the 3D classification) were selected. These particles were further processed using 3D refinement, CTF refinement, and Bayesian particle polishing in RELION. To enhance the map densities of KDM6B and the linker DNA, two rounds of masked classification focusing on KDM6B and the linker DNA were performed, resulting in improved resolution of these regions. Multibody refinement was conducted by masking the 1) nucleosome + trans-linker DNA and 2) KDM6B + cis-linker DNA, yielding better-resolved secondary structure features of KDM6B. Finally, a composite map was generated using the Phenix software package (v1.21)⁵⁸ by combining densities from the consensus map with the multibody refinement maps of the nucleosome + trans-linker DNA and KDM6B + cis-linker DNA. The reported resolution was determined based on the gold-standard Fourier shell correlation (FSC) curve at the 0.143 criterion.

KDM6B–nucleosome complex model building

An initial model of the KDM6B-nucleosome complex was constructed by manually docking the catalytic domain of human KDM6B in complex with H3 (residues 17–33, K18I/A21M; PDB: 6F6D) and the nucleosome (PDB: 6ESF) into the composite maps using Chimera⁵⁵. Subsequent modifications were performed in Coot⁵⁷, including adjustments to the DNA sequence and length, replacement of histone sequences with human histone sequences, and substitution of the human KDM6B sequence with the corresponding mouse KDM6B sequence. The model was then subjected to real-space refinement against the composite map using Phenix software package (v1.21).

Electrophoretic mobility shift assay

Electrophoretic mobility shift assays (EMSA) were conducted by incubating nucleosomes (50 nM) with serial dilutions of KDM6B in 10 μL of binding buffer (20 mM HEPES pH 7.5, 50 mM NaCl, 5% sucrose and 1 mM DTT) for 15–30 minutes on ice. Samples were analyzed on 5% native polyacrylamide gels run in 0.2x TBE buffer at 150 V for 75 minutes at 4 °C. The gels were stained with ethidium bromide and scanned using a Bio-Rad ChemiDoc imager. All EMSA experiments were performed with SUMO-10xHis tagged KDM6B excepted for the experiment in Extended Data Fig. 1d, which used TEV cleaved, untagged KDM6B. Dissociation constants were calculated using specific binding with Hill slope function in GraphPad Prism.

Nucleosome demethylation assay

Nucleosome demethylation assays were performed using 4 μM H3K27me3 nucleosomes and 2 μM KDM6B/SUMO-10xHis-KDM6B or its mutants in demethylation buffer (20 mM HEPES pH 7.5, 50 mM NaCl, 1 mM α-ketoglutarate, 2 mM L-ascorbic acid, 50 μM iron(II) sulfate, and 1 mM DTT) at 30 °C for 1.5 hours. To investigate the effect of linker histone H1 on nucleosome demethylation, histone H1.5 (globular domain residues 40–112, gH1.5) was added to the reaction at indicated concentrations. For each reaction, 4 μL of the reaction was stopped by adding 6 μL of 2x SDS-PAGE sample loading buffer. When using H2AK119ub1/H3K27me3 nucleosomes as substrates, reactions were stopped at specific time points using 2x SDS-PAGE sample loading buffer. A 2.5 μL aliquot of each sample was analyzed on a 4–20% Mini-PROTEAN TGX gel (Bio-Rad) and immunoblotted with primary antibodies: mouse monoclonal histone H3 (1:5000, Cell Signaling Technology, 3638) and rabbit monoclonal H3K27me3 (1:5000, Invitrogen, MA5–11198), followed by secondary antibodies IRDye 680RD goat anti-rabbit IgG (1:10,000, LI-COR) and IRDye 800CW goat anti-mouse IgG (1:10,000, LI-COR). Blots were imaged on an Odyssey imager (LI-COR) and quantified using ImageJ software⁵⁹, with H3K27me3 levels normalized to the wild-type H3 signal from the corresponding sample. Statistical significance was determined using GraphPad Prism with ordinary one-way ANOVA for KDM6B mutants and unpaired t-test for nucleosomes with different DNA lengths or ubiquitylated H2A.

Dot Blot assay

Demethylation reactions were performed by incubating 2 μM KDM6B with either 4 μM H3K27me3 20N20 nucleosome or 8 μM synthetic H3K27me3 peptide (residues 14–34) in demethylation buffer at 30 °C. For nucleosome-based assays, 2.5 μl of each reaction was quenched at defined time points by adding to 3.75 μL quench buffer (4% SDS, 2 mM EDTA). For peptide-based assays, 9 μL of reaction mixture was quenched with 1 μL of 10 mM EDTA. For detection, 2 μL of each nucleosome reaction or 10 μL of each peptide reaction was spotted onto an activated PVDF membrane. A standard titration series of nucleosome or peptide was also spotted to allow estimatation of demethylation activity. Membranes were air-dried, blocked in 5% nonfat milk in TBST for 1 hour, and immunoblotted with rabbit monoclonal anti-H3K27me3 antibody (Invitrogen, MA5–11198; 1:5000 for nucleosomes, 1:2000 for peptides) for 1.5 hours. Blots were then incubated with IRDye 680RD goat anti-rabbit IgG secondary antibody (1:10,000, LI-COR) and imaged using an Odyssey imager (LI-COR). Assays were performed in triplicate.

Nucleosomal linker DNA FRET assay

FRET assays were performed by adding buffer (20 mM HEPES pH 7.5, 50 mM NaCl, 1 mM DTT and 0.1 mg/ml BSA) along with 2 μM KDM6B, its mutants, or KDM2A to 25 nM Cy3- and Cy5-labeled 20N20 nucleosomes in a 384-well plate. The plate was vortexed for 1 minute, briefly centrifuged, and analyzed using an EnVision 2103 Multilabel Plate Reader (Perkin Elmer). FRET efficiency was calculated as the ratio of the Cy5 emission signal at 685 nm to the Cy3 emission signal at 590 nm following excitation at 535 nm.

QUANTIFICATION AND STATISTICAL ANALYSIS

All statistical analysis and quantifications are described in the figure legends or STAR Methods section. Lysine demethylase assays, KDM6B-nucleosome binding measurements, and FRET measurement were collected using n=3 independent biochemical assay replicates. Lysine demethylase assays and KDM6B-nucleosome binding measurements are shown as mean ± s.d. for each condition. For demethylase assays, statistical significance was determined using GraphPad Prism with ordinary one-way ANOVA for KDM6B mutants and unpaired t-test for nucleosomes with different DNA lengths or ubiquitylated H2A. KDM6B-nuclesome dissociation constants were calculated using specific binding with Hill slope function in GraphPad Prism and are represented as mean ± s.e.m. FRET assays are shown as n=3 data points without additional statistical analysis.

Supplementary Material

NIHMS2098826-supplement-1.pdf^{(7.2MB, pdf)}

Document S1. Figures S1–S10.

KEY RESOURCES TABLE

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
Mouse monoclonal anti-Histone H3	Cell Signaling Technology	Cat#3638S; RRID: AB_1642229_
Rabbit monoclonal anti-H3K27me3	Thermo Fisher Scientific	Cat#MA5-11198; RRID: AB_11000749
IRDye 680RD Goat anti-Rabbit IgG	LI-COR Biosciences	Cat# 926-68071; RRID: AB_10956166
IRDye 800CW Goat anti-Mouse IgG	LI-COR Biosciences	Cat# 925-32210; RRID: AB_2687825
Bacterial and virus strains
E. coli BL21(DE3)pLysS	Novagen	Cat# 69388-3
E. coli TG1	Agilent	Cat# 200123
E. coli HB101	ATCC	Cat# 67593
Chemicals, peptides, and recombinant proteins
Q5 High-Fidelity DNA polymerase	New England Biolabs	Cat# M0491S
Gibson Assembly Master Mix	New England Biolabs	Cat# E2611L
HismUba1	McGinty et al.⁵⁰	N/A
UbcH5C	McGinty et al.⁵⁰	N/A
BMI1-RING1B complex	McGinty et al.⁵⁰	N/A
Benzamidine Hydrochloride	TCI	Cat# B0013
2-mercaptoethanol	Avantor	Cat# M131-250ML
TEV protease	Anderson et al.⁵¹	N/A
UNC8015	Spangler et al.²²	N/A
Guanidine Hydrochloride	Fisher Scientific	Cat# BP178-1
D-methionine	Sigma-Aldrich	Cat# M9375
L-methionine	Sigma-Aldrich	Cat# M9625
4-mercaptophenylacetic acid	Sigma-Aldrich	Cat# 653152
Sodium phosphate monobasic monodhydrate	Fisher Scientific	Cat# S369-3
Sodium nitrite	Sigma-Aldrich	Cat# 237213
VA-044	Wako	Cat# 27776-21-6
2-methyl-2-propanethiol	Sigma-Aldrich	Cat# 109207
Methoxylamine hydrochloride	Fisher Scientific	Cat# AC210490250
TCEP	UBP Bio	Cat# P1020-25
1,3-dichloracetone	Fisher Scientific	Cat# AC173630250
L-Ascorbic acid	Sigma-Aldrich	Cat# A92902
Ammonium iron(II) sulfate hexahydrate	Sigma-Aldrich	Cat# 203505
α-Ketoglutaric acid	Sigma-Aldrich	Cat# K1750
ATP	UBP Bio	Cat# P1040-25
Critical commercial assays
TAGZyme DAPase Enzyme	Qiagen	Cat#34362
Deposited data
KDM6B-nucleosome complex	This study	PDB: 9NQU
KDM6B-nucleosome complex map	This study	EMD-49676
Uncropped gel and blot images, western blot quantitation, and raw FRET data	This study	Mendeley Data: https://doi.org/10.17632/gmgxzphgrc.1
Oligonucleotides
185 bp 601 Cy3 forward primer (ATCGC/iCY3-dT/GTTCAATACATGC)	This study	N/A
185 bp 601 Cy5 reverse primer (ATCCCTA/iCY5-dT/ACGCGGCC)	This study	N/A
Recombinant DNA
pCS2-Jmjd3-F	Addgene	Plasmid # 17440
pST50Tr-SMOHSTNmKDM6Bt1	This study	N/A
pST50Tr-STRaHSTNhH1.5	Gift from Song Tan	N/A
pST50Tr-STRaHSTNGS2hUbq	Gift from Song Tan	N/A
pST100-20xNCP601a	Gift from Song Tan	N/A
pST103-16xNCP601a185M	Gift from Song Tan	N/A
pST83-16xNCP601a165L	Gift from Song Tan	N/A
pST50Trc3-hH2A.D	Anderson et al.⁵¹	N/A
pST50Trc4-hH2B.C	Anderson et al.⁵¹	N/A
pST50Tr-hH3.2	Anderson et al.⁵¹	N/A
pST50Trc2-hH4	Skrajna et al.⁵²	N/A
pST50Tr-hH3.2×31 H3.2(K27C/C110A)	This study	N/A
pST50Tr-hH3.2×52 H3.2(1–28)-ENLYFQGG-H3(29–135, A29C/C110A)	This study	N/A
Software and algorithms
SerialEM	Mastronarde⁵³	https://bio3d.colorado.edu/SerialEM/
RELION 5.0 beta	Scheres et al.⁵⁴	https://relion.readthedocs.io/en/release-5.0/
UCSF Chimera	Pettersen et al.⁵⁵	https://www.cgl.ucsf.edu/chimera/
UCSF Chimera X	Pettersen et al.⁵⁶	https://www.cgl.ucsf.edu/chimerax
Coot	Emsley et al.⁵⁷	https://www2.mrc-lmb.cam.ac.uk/personal/pemsley/coot/
Phenix v1.21	Liebschner et al.⁵⁸	https://phenix-online.org/
ImageJ	Schneider et al.⁵⁹	https://imagej.nih.gov/ij/
GraphPad Prism (version 10.2.3)	GraphPad Software Inc.	https://www.graphpad.com
Other
TALON Metal Affinity Resin	Takara Bio	Cat#635669
Source S resin	GE Healthcare	Cat#17094405
Superdex 75 increase 10/300 GL column	Cytiva	Cat#29148721
Source Q resin	GE Healthcare	Cat#17094705
Superdex 200 increase 10/300 GL column	Cytiva	Cat# 28990944
Vydac 218TP 10 μ M C18 column 250×22mm	Mac-Mod Analytical	Cat# 218TP1022
Vydac 218TP 5 μ M C18 column 250×10mm	Mac-Mod Analytical	Cat# 218TP510
Vivaspin 500 centrifugal concentrator MWCO:10 kDA	SARTORIUS	Cat# VS0101
Quantifoil Holey Carbon, Cu 300 mesh grid, R1.2/1.3	Quantifoil Micro Tools	Cat# Q350CR1.3
Vitrobot Mark IV	Thermo Fisher Scientific	N/A
ChemiDoc imaging system	BIO-RAD	Cat# 12003153

Open in a new tab

Highlights.

Cryo-EM structure of KDM6B bound to the nucleosome
KDM6B engages nucleosomal DNA and both DNA linkers for H3K27me3 demethylation
Structure provides mechanism for KDM6B function in compact heterochromatin
H2AK119ub1 and linker histones antagonize H3K27 demethylation

ACKNOWLEDGEMENTS

Single particle cryo-EM data were collected with assistance from Dr. Joshua Strauss at the University of North Carolina at Chapel Hill School of Medicine Cryo-Electron Microscopy Facility and peptide synthesis was performed by Dr. Krzysztof Krajewski in the UNC High-Throughput Peptide Synthesis and Array Facility, both of which are partially supported by NIH P30CA016086. Plasmids containing 601 nucleosome positioning DNA fragments were gifts from Song Tan. We thank members of the McGinty laboratory for critical comments on the manuscript. Molecular graphics and analyses were performed with UCSF ChimeraX, developed by the Resource for Biocomputing, Visualization, and Informatics at the University of California, San Francisco, with support from NIH P41-GM103311.

Funding:

This work was supported by NIH grants R35GM133498 to R.K.M., R35GM139514 to S.V.F., and R01CA010305 to L.I.J.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

DECLARATION OF INTERESTS

The authors declare no competing interests.

DECLARATION OF GENERATIVE AI AND AI-ASSISTED TECHNOLOGIES

During preparation of this work, the authors used ChatGPT in order to revise the text for clarity. After using this tool, the authors reviewed and edited the manuscript and take full responsibility for the contents of the publication.

REFERENCES

1.Kouzarides T (2007). Chromatin Modifications and Their Function. Cell, 128, 693–705. 10.1016/j.cell.2007.02.005. [DOI] [PubMed] [Google Scholar]
2.Allis CD, and Jenuwein T (2016). The molecular hallmarks of epigenetic control. Nature Reviews Genetics 17, 487–500. 10.1038/nrg.2016.59. [DOI] [PubMed] [Google Scholar]
3.Trojer P, and Reinberg D (2007). Facultative Heterochromatin: Is There a Distinctive Molecular Signature? Mol Cell 28, 1–13. 10.1016/j.molcel.2007.09.011. [DOI] [PubMed] [Google Scholar]
4.Cao R, and Zhang Y (2004). The functions of E(Z)/EZH2-mediated methylation of lysine 27 in histone H3. Current Opinion in Genetics and Development 14, 155–164. 10.1016/j.gde.2004.02.001. [DOI] [PubMed] [Google Scholar]
5.Kasinath V, Beck C, Sauer P, Poepsel S, Kosmatka J, Faini M, Toso D, Aebersold R, and Nogales E (2021). JARID2 and AEBP2 regulate PRC2 in the presence of H2AK119ub1 and other histone modifications. Science 371. 10.1126/science.abc3393. [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Laugesen A, Højfeldt JW, and Helin K (2019). Molecular Mechanisms Directing PRC2 Recruitment and H3K27 Methylation. Mol Cell 74, 8–18. 10.1016/j.molcel.2019.03.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Poepsel S, Kasinath V, and Nogales E (2018). Cryo-EM structures of PRC2 simultaneously engaged with two functionally distinct nucleosomes. Nat Struct Mol Biol 25, 154–162. 10.1038/s41594-018-0023-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
8.Dochnal SA, Francois AK, and Cliffe AR (2021). De novo polycomb recruitment: Lessons from latent herpesviruses. Viruses 13. 10.3390/v13081470. [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Hua C, Chen J, Li S, Zhou J, Fu J, Sun W, and Wang W (2021). KDM6 Demethylases and Their Roles in Human Cancers. Frontiers in Oncology 11. 10.3389/fonc.2021.779918. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Walport LJ, Hopkinson RJ, Vollmar M, Madden SK, Gileadi C, Oppermann U, Schofield CJ, and Johansson C (2014). Human UTY(KDM6C) is a male-specific N∈-methyl lysyl demethylase. Journal of Biological Chemistry 289, 18302–18313. 10.1074/jbc.M114.555052. [DOI] [PMC free article] [PubMed] [Google Scholar]
11.De Santa F, Totaro MG, Prosperini E, Notarbartolo S, Testa G, and Natoli G (2007). The Histone H3 Lysine-27 Demethylase Jmjd3 Links Inflammation to Inhibition of Polycomb-Mediated Gene Silencing. Cell 130, 1083–1094. 10.1016/j.cell.2007.08.019. [DOI] [PubMed] [Google Scholar]
12.Tran N, Broun A, and Ge K (2020). Lysine Demethylase KDM6A in Differentiation, Development, and Cancer. Mol Cell Biol 40. 10.1128/mcb.00341-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Bögershausen N, Gatinois V, Riehmer V, Kayserili H, Becker J, Thoenes M, Simsek-Kiper PÖ, Barat-Houari M, Elcioglu NH, Wieczorek D, et al. (2016). Mutation Update for Kabuki Syndrome Genes KMT2D and KDM6A and Further Delineation of X-Linked Kabuki Syndrome Subtype 2. Hum Mutat 37, 847–864. 10.1002/humu.23026. [DOI] [PubMed] [Google Scholar]
14.Wang W, Cho H, Lee JW, and Lee SK (2022). The histone demethylase Kdm6b regulates subtype diversification of mouse spinal motor neurons during development. Nat Commun 13, 958. 10.1038/s41467-022-28636-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
15.Zhang H, Hu Y, Liu D, Liu Z, Xie N, Liu S, Zhang J, Jiang Y, Li C, Wang Q, et al. (2022). The histone demethylase Kdm6b regulates the maturation and cytotoxicity of TCRαβ+CD8αα+ intestinal intraepithelial lymphocytes. Cell Death Differ 29, 1349–1363. 10.1038/s41418-021-00921-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Ye L, Fan Z, Yu B, Chang J, Al Hezaimi K, Zhou X, Park NH, and Wang CY (2012). Histone demethylases KDM4B and KDM6B promotes osteogenic differentiation of human MSCs. Cell Stem Cell 11, 50–61. 10.1016/j.stem.2012.04.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Li J, Hardy K, Olshansky M, Barugahare A, Gearing LJ, Prier JE, Sng XYX, Nguyen MLT, Piovesan D, Russ BE, et al. (2021). KDM6B-dependent chromatin remodeling underpins effective virus-specific CD8+ T cell differentiation. Cell Rep 34, 108839. 10.1016/j.celrep.2021.108839. [DOI] [PubMed] [Google Scholar]
18.Zhang X, Liu L, Yuan X, Wei Y, and Wei X (2019). JMJD3 in the regulation of human diseases. Protein and Cell 10, 864–882. 10.1007/s13238-019-0653-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Jones SE, Olsen L, and Gajhede M (2018). Structural Basis of Histone Demethylase KDM6B Histone 3 Lysine 27 Specificity. Biochemistry 57, 585–592. 10.1021/acs.biochem.7b01152. [DOI] [PubMed] [Google Scholar]
20.Kruidenier L, Chung CW, Cheng Z, Liddle J, Che K, Joberty G, Bantscheff M, Bountra C, Bridges A, Diallo H, et al. (2012). A selective jumonji H3K27 demethylase inhibitor modulates the proinflammatory macrophage response. Nature 488, 404–408. 10.1038/nature11262. [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Sengoku T, and Yokoyama S (2011). Structural basis for histone H3 lys 27 demethylation by UTX/KDM6A. Genes Dev 25, 2266–2277. 10.1101/gad.172296.111. [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Spangler CJ, Skrajna A, Foley CA, Nguyen A, Budziszewski GR, Azzam DN, Arteaga EC, Simmons HC, Smith CB, Wesley NA, et al. (2023). Structural basis of paralog-specific KDM2A/B nucleosome recognition. Nat Chem Biol 19, 624–632. 10.1038/s41589-023-01256-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Bednar J, Garcia-Saez I, Boopathi R, Cutter AR, Papai G, Reymer A, Syed SH, Lone IN, Tonchev O, Crucifix C, et al. (2017). Structure and Dynamics of a 197 bp Nucleosome in Complex with Linker Histone H1. Mol Cell 66, 384–397. 10.1016/j.molcel.2017.04.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
24.Chi YI, Stodola TJ, De Assuncao TM, Levrence EN, Tripathi S, Dsouza NR, Mathison AJ, Basel DG, Volkman BF, Smith BC, et al. (2021). Molecular mechanics and dynamic simulations of well-known Kabuki syndrome-associated KDM6A variants reveal putative mechanisms of dysfunction. Orphanet J Rare Dis 16, 66. 10.1186/s13023-021-01692-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Fyodorov DV, Zhou BR, Skoultchi AI, and Bai Y (2018). Emerging roles of linker histones in regulating chromatin structure and function. Nature Reviews Molecular Cell Biology 19,192–206. 10.1038/nrm.2017.94. [DOI] [PMC free article] [PubMed] [Google Scholar]
26.Ricci MA, Manzo C, García-Parajo MF, Lakadamyali M, and Cosma MP (2015). Chromatin fibers are formed by heterogeneous groups of nucleosomes in vivo. Cell 160, 1145–1158. 10.1016/j.cell.2015.01.054. [DOI] [PubMed] [Google Scholar]
27.Yuan W, Wu T, Fu H, Dai C, Wu H, Liu N, Li X, Xu M, Zhang Z, Niu T, et al. (2012). Dense chromatin activates polycomb repressive complex 2 to regulate H3 lysine 27 methylation. Science 337, 971–975. 10.1126/science.1225237. [DOI] [PubMed] [Google Scholar]
28.Willcockson MA, Healton SE, Weiss CN, Bartholdy BA, Botbol Y, Mishra LN, Sidhwani DS, Wilson TJ, Pinto HB, Maron MI, et al. (2021). H1 histones control the epigenetic landscape by local chromatin compaction. Nature 589, 293–298. 10.1038/s41586-020-3032-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Liu C, Yu J, Song A, Wang M, Hu J, Chen P, Zhao J, and Li G (2023). Histone H1 facilitates restoration of H3K27me3 during DNA replication by chromatin compaction. Nat Commun 14, 4081. 10.1038/s41467-023-39846-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Cooper S, Grijzenhout A, Underwood E, Ancelin K, Zhang T, Nesterova TB, Anil-Kirmizitas B, Bassett A, Kooistra SM, Agger K, et al. (2016). Jarid2 binds monoubiquitylated H2A lysine 119 to mediate crosstalk between Polycomb complexes PRC1 and PRC2. Nat Commun 7, 13661. 10.1038/ncomms13661. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Kalb R, Latwiel S, Baymaz HI, Jansen PWTC, Müller CW, Vermeulen M, and Müller J (2014). Histone H2A monoubiquitination promotes histone H3 methylation in Polycomb repression. Nat Struct Mol Biol 21, 569–571. 10.1038/nsmb.2833. [DOI] [PubMed] [Google Scholar]
32.Jason LJM, Finn RM, Lindsey G, and Ausió J (2005). Histone H2A ubiquitination does not preclude histone H1 binding, but it facilitates its association with the nucleosome. Journal of Biological Chemistry 280, 4975–4982. 10.1074/jbc.M410203200. [DOI] [PubMed] [Google Scholar]
33.Zhao J, Lan J, Wang M, Liu C, Fang Z, Song A, Zhang T, Wang L, Zhu B, Chen P, et al. (2024). H2AK119ub1 differentially fine-tunes gene expression by modulating canonical PRC1- and H1-dependent chromatin compaction. Mol Cell 84, 1191–1205. 10.1016/j.molcel.2024.02.017. [DOI] [PubMed] [Google Scholar]
34.Hicks CW, Rahman S, Gloor SL, Fields JK, Husby NL, Vaidya A, Maier KE, Morgan M, Keogh MC, and Wolberger C (2024). Ubiquitinated histone H2B as gatekeeper of the nucleosome acidic patch. Nucleic Acids Res 52, 9978–9995. 10.1093/nar/gkae698. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Chen Y, Bates DL, Dey R, Chen PH, Machado ACD, Laird-Offringa IA, Rohs R, and Chen L (2012). DNA Binding by GATA Transcription Factor Suggests Mechanisms of DNA Looping and Long-Range Gene Regulation. Cell Rep 2, 1197–1206. 10.1016/j.celrep.2012.10.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Markert JW, Soffers JH, and Farnung L (2025). Structural basis of H3K36 trimethylation by SETD2 during chromatin transcription. Science 387, 528–533. 10.1126/science.adn6319. [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Liu Y, Zhang Y, Xue H, Cao M, Bai G, Mu Z, Yao Y, Sun S, Fang D, and Huang J (2021). Cryo-EM structure of SETD2/Set2 methyltransferase bound to a nucleosome containing oncohistone mutations. Cell Discov 7, 32. 10.1038/s41421-021-00261-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Li W, Tian W, Yuan G, Deng P, Sengupta D, Cheng Z, Cao Y, Ren J, Qin Y, Zhou Y, et al. (2021). Molecular basis of nucleosomal H3K36 methylation by NSD methyltransferases. Nature 590, 498–503. 10.1038/s41586-020-03069-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Sato K, Kumar A, Hamada K, Okada C, Oguni A, Machiyama A, Sakuraba S, Nishizawa T, Nureki O, Kono H, et al. (2021). Structural basis of the regulation of the normal and oncogenic methylation of nucleosomal histone H3 Lys36 by NSD2. Nat Commun 12, 6605. 10.1038/s41467-021-26913-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Bilokapic S, and Halic M (2019). Nucleosome and ubiquitin position Set2 to methylate H3K36. Nat Commun 10, 3795. 10.1038/s41467-019-11726-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Sauer PV, Pavlenko E, Cookis T, Zirden LC, Renn J, Singhal A, Hunold P, Hoehne-Wiechmann MN, van Ray O, Kaschani F, et al. (2024). Activation of automethylated PRC2 by dimerization on chromatin. Mol Cell 84, 3885–3898.e8. 10.1016/j.molcel.2024.08.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Ntziachristos P, Tsirigos A, Welstead GG, Trimarchi T, Bakogianni S, Xu L, Loizou E, Holmfeldt L, Strikoudis A, King B, et al. (2014). Contrasting roles of histone 3 lysine 27 demethylases in acute lymphoblastic leukaemia. Nature 514, 513–517. 10.1038/nature13605. [DOI] [PMC free article] [PubMed] [Google Scholar]
43.Li Q, Zou J, Wang M, Ding X, Chepelev I, Zhou X, Zhao W, Wei G, Cui J, Zhao K, et al. (2014). Critical role of histone demethylase Jmjd3 in the regulation of CD4 + T-cell differentiation. Nat Commun 5. 10.1038/ncomms6780. [DOI] [PMC free article] [PubMed] [Google Scholar]
44.Park SH, Ayoub A, Lee YT, Xu J, Kim H, Zheng W, Zhang B, Sha L, An S, Zhang Y, et al. (2019). Cryo-EM structure of the human MLL1 core complex bound to the nucleosome. Nat Commun 10, 1165. 10.1038/s41467-019-13550-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Xue H, Yao T, Cao M, Zhu G, Li Y, Yuan G, Chen Y, Lei M, and Huang J (2019). Structural basis of nucleosome recognition and modification by MLL methyltransferases. Nature 573, 445–449. 10.1038/s41586-019-1528-1. [DOI] [PubMed] [Google Scholar]
46.Rahman S, Hoffmann NA, Worden EJ, Smith ML, Namitz KEW, Knutson BA, Cosgrove MS, and Wolberger C (2022). Multistate structures of the MLL1-WRAD complex bound to H2B-ubiquitinated nucleosome. Proc Natl Acad Sci U S A 119, e2205691119. 10.1073/pnas.2205691119. [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Jiao L, and Liu X (2015). Structural basis of histone H3K27 trimethylation by an active polycomb repressive complex 2. Science 350, aac4383. 10.1126/science.aac4383. [DOI] [PMC free article] [PubMed] [Google Scholar]
48.Margueron R, Justin N, Ohno K, Sharpe ML, Son J, Drury WJ, Voigt P, Martin SR, Taylor WR, De Marco V, et al. (2009). Role of the polycomb protein EED in the propagation of repressive histone marks. Nature 461, 762–767. 10.1038/nature08398. [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Tamburri S, Lavarone E, Fernández-Pérez D, Conway E, Zanotti M, Manganaro D, and Pasini D (2020). Histone H2AK119 Mono-Ubiquitination Is Essential for Polycomb-Mediated Transcriptional Repression. Mol Cell 77, 840–856.e5. 10.1016/j.molcel.2019.11.021. [DOI] [PMC free article] [PubMed] [Google Scholar]
50.McGinty RK, Henrici RC, and Tan S (2014). Crystal structure of the PRC1 ubiquitylation module bound to the nucleosome. Nature 514, 591–596. 10.1038/nature13890. [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Anderson CJ, Baird MR, Hsu A, Barbour EH, Koyama Y, Borgnia MJ, and McGinty RK (2019). Structural Basis for Recognition of Ubiquitylated Nucleosome by Dot1L Methyltransferase. Cell Rep 26, 1681–1690.e5. 10.1016/j.celrep.2019.01.058. [DOI] [PMC free article] [PubMed] [Google Scholar]
52.Skrajna A, Goldfarb D, Kedziora KM, Cousins EM, Grant GD, Spangler CJ, Barbour EH, Yan X, Hathaway NA, Brown NG, et al. (2020). Comprehensive nucleosome interactome screen establishes fundamental principles of nucleosome binding. Nucleic Acids Res 48, 9415–9432. 10.1093/nar/gkaa544. [DOI] [PMC free article] [PubMed] [Google Scholar]
53.Mastronarde DN (2005). Automated electron microscope tomography using robust prediction of specimen movements. J Struct Biol 152. 10.1016/j.jsb.2005.07.007. [DOI] [PubMed] [Google Scholar]
54.Scheres SHW (2012). RELION: Implementation of a Bayesian approach to cryo-EM structure determination. J Struct Biol 180, 519–530. 10.1016/j.jsb.2012.09.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
55.Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, and Ferrin TE (2004). UCSF Chimera - A visualization system for exploratory research and analysis. J Comput Chem 25, 1605–1612. 10.1002/jcc.20084. [DOI] [PubMed] [Google Scholar]
56.Pettersen EF, Goddard TD, Huang CC, Meng EC, Couch GS, Croll TI, Morris JH, and Ferrin TE (2021). UCSF ChimeraX: Structure visualization for researchers, educators, and developers. Protein Science 30. 10.1002/pro.3943. [DOI] [PMC free article] [PubMed] [Google Scholar]
57.Emsley P, Lohkamp B, Scott WG, and Cowtan K (2010). Features and development of Coot. Acta Crystallogr D Biol Crystallogr 66, 486–501. 10.1107/S0907444910007493. [DOI] [PMC free article] [PubMed] [Google Scholar]
58.Liebschner D, Afonine PV, Baker ML, Bunkoczi G, Chen VB, Croll TI, Hintze B, Hung LW, Jain S, McCoy AJ, et al. (2019). Macromolecular structure determination using X-rays, neutrons and electrons: Recent developments in Phenix. Acta Crystallogr D Struct Biol 75, 861–877. 10.1107/S2059798319011471. [DOI] [PMC free article] [PubMed] [Google Scholar]
59.Schneider CA, Rasband WS, and Eliceiri KW (2012). NIH Image to ImageJ: 25 years of image analysis. Nature Methods 9, 671–675. 10.1038/nmeth.2089. [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Tan S, Kern RC, and Selleck W (2005). The pST44 polycistronic expression system for producing protein complexes in Escherichia coli. Protein Expr Purif 40, 385–395. 10.1016/j.pep.2004.12.002. [DOI] [PubMed] [Google Scholar]
61.Chen Z, Grzybowski AT, and Ruthenburg AJ (2015). Traceless semisynthesis of a set of histone 3 species bearing specific lysine methylation marks. ChemBioChem 16, 2071–2075. 10.1002/cbic.201402313. [DOI] [PMC free article] [PubMed] [Google Scholar]
62.Zheng JS, Tang S, Qi YK, Wang ZP, and Liu L (2013). Chemical synthesis of proteins using peptide hydrazides as thioester surrogates. Nat Protoc 8, 2483–2495. 10.1038/nprot.2013.152. [DOI] [PubMed] [Google Scholar]
63.McGinty RK, Köhn M, Chatterjee C, Chiang KP, Pratt MR, and Muir TW (2009). Structure-activity analysis of semisynthetic nucleosomes: Mechanistic insights into the stimulation of Dot1L by ubiquitylated histone H2B. ACS Chem Biol 4, 958–968. 10.1021/cb9002255. [DOI] [PMC free article] [PubMed] [Google Scholar]
64.Peck JV, Fay JF, and Strauss JD (2022). High-speed high-resolution data collection on a 200 keV cryo-TEM. IUCrJ 9, 243–252. 10.1107/S2052252522000069. [DOI] [PMC free article] [PubMed] [Google Scholar]
65.Zivanov J, Nakane T, and Scheres SHW (2020). Estimation of high-order aberrations and anisotropic magnification from cryo-EM data sets in RELION-3.1. IUCrJ 7, 253–267. 10.1107/S2052252520000081. [DOI] [PMC free article] [PubMed] [Google Scholar]
66.Zheng SQ, Palovcak E, Armache JP, Verba KA, Cheng Y, and Agard DA (2017). MotionCor2: Anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nature Methods 14, 331–332. 10.1038/nmeth.4193. [DOI] [PMC free article] [PubMed] [Google Scholar]
67.Rohou A, and Grigorieff N (2015). CTFFIND4: Fast and accurate defocus estimation from electron micrographs. J Struct Biol 192, 216–221. 10.1016/j.jsb.2015.08.008. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

NIHMS2098826-supplement-1.pdf^{(7.2MB, pdf)}

Data Availability Statement

The cryo-EM composite map has been deposited in the Electron Microscopy Data Bank under the accession code EMD-49676 and the KDM6B-nucleosome model has been deposited in the Protein Data Bank under the accession code 9NQU and are publicly available as of the date of publication. Uncropped gel and blot images and raw data and calculations related to blot quantification and FRET experiments generated during this study have been deposited into the Mendeley database (https://doi.org/10.17632/gmgxzphgrc.1) and are publicly available as of the date of publication.
This paper does not report original code.
Any additional information required to reanalyze data reported in this paper is available from the lead contact upon request.

[R1] 1.Kouzarides T (2007). Chromatin Modifications and Their Function. Cell, 128, 693–705. 10.1016/j.cell.2007.02.005. [DOI] [PubMed] [Google Scholar]

[R2] 2.Allis CD, and Jenuwein T (2016). The molecular hallmarks of epigenetic control. Nature Reviews Genetics 17, 487–500. 10.1038/nrg.2016.59. [DOI] [PubMed] [Google Scholar]

[R3] 3.Trojer P, and Reinberg D (2007). Facultative Heterochromatin: Is There a Distinctive Molecular Signature? Mol Cell 28, 1–13. 10.1016/j.molcel.2007.09.011. [DOI] [PubMed] [Google Scholar]

[R4] 4.Cao R, and Zhang Y (2004). The functions of E(Z)/EZH2-mediated methylation of lysine 27 in histone H3. Current Opinion in Genetics and Development 14, 155–164. 10.1016/j.gde.2004.02.001. [DOI] [PubMed] [Google Scholar]

[R5] 5.Kasinath V, Beck C, Sauer P, Poepsel S, Kosmatka J, Faini M, Toso D, Aebersold R, and Nogales E (2021). JARID2 and AEBP2 regulate PRC2 in the presence of H2AK119ub1 and other histone modifications. Science 371. 10.1126/science.abc3393. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R6] 6.Laugesen A, Højfeldt JW, and Helin K (2019). Molecular Mechanisms Directing PRC2 Recruitment and H3K27 Methylation. Mol Cell 74, 8–18. 10.1016/j.molcel.2019.03.011. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R7] 7.Poepsel S, Kasinath V, and Nogales E (2018). Cryo-EM structures of PRC2 simultaneously engaged with two functionally distinct nucleosomes. Nat Struct Mol Biol 25, 154–162. 10.1038/s41594-018-0023-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R8] 8.Dochnal SA, Francois AK, and Cliffe AR (2021). De novo polycomb recruitment: Lessons from latent herpesviruses. Viruses 13. 10.3390/v13081470. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R9] 9.Hua C, Chen J, Li S, Zhou J, Fu J, Sun W, and Wang W (2021). KDM6 Demethylases and Their Roles in Human Cancers. Frontiers in Oncology 11. 10.3389/fonc.2021.779918. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R10] 10.Walport LJ, Hopkinson RJ, Vollmar M, Madden SK, Gileadi C, Oppermann U, Schofield CJ, and Johansson C (2014). Human UTY(KDM6C) is a male-specific N∈-methyl lysyl demethylase. Journal of Biological Chemistry 289, 18302–18313. 10.1074/jbc.M114.555052. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R11] 11.De Santa F, Totaro MG, Prosperini E, Notarbartolo S, Testa G, and Natoli G (2007). The Histone H3 Lysine-27 Demethylase Jmjd3 Links Inflammation to Inhibition of Polycomb-Mediated Gene Silencing. Cell 130, 1083–1094. 10.1016/j.cell.2007.08.019. [DOI] [PubMed] [Google Scholar]

[R12] 12.Tran N, Broun A, and Ge K (2020). Lysine Demethylase KDM6A in Differentiation, Development, and Cancer. Mol Cell Biol 40. 10.1128/mcb.00341-20. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R13] 13.Bögershausen N, Gatinois V, Riehmer V, Kayserili H, Becker J, Thoenes M, Simsek-Kiper PÖ, Barat-Houari M, Elcioglu NH, Wieczorek D, et al. (2016). Mutation Update for Kabuki Syndrome Genes KMT2D and KDM6A and Further Delineation of X-Linked Kabuki Syndrome Subtype 2. Hum Mutat 37, 847–864. 10.1002/humu.23026. [DOI] [PubMed] [Google Scholar]

[R14] 14.Wang W, Cho H, Lee JW, and Lee SK (2022). The histone demethylase Kdm6b regulates subtype diversification of mouse spinal motor neurons during development. Nat Commun 13, 958. 10.1038/s41467-022-28636-7. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R15] 15.Zhang H, Hu Y, Liu D, Liu Z, Xie N, Liu S, Zhang J, Jiang Y, Li C, Wang Q, et al. (2022). The histone demethylase Kdm6b regulates the maturation and cytotoxicity of TCRαβ+CD8αα+ intestinal intraepithelial lymphocytes. Cell Death Differ 29, 1349–1363. 10.1038/s41418-021-00921-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R16] 16.Ye L, Fan Z, Yu B, Chang J, Al Hezaimi K, Zhou X, Park NH, and Wang CY (2012). Histone demethylases KDM4B and KDM6B promotes osteogenic differentiation of human MSCs. Cell Stem Cell 11, 50–61. 10.1016/j.stem.2012.04.009. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R17] 17.Li J, Hardy K, Olshansky M, Barugahare A, Gearing LJ, Prier JE, Sng XYX, Nguyen MLT, Piovesan D, Russ BE, et al. (2021). KDM6B-dependent chromatin remodeling underpins effective virus-specific CD8+ T cell differentiation. Cell Rep 34, 108839. 10.1016/j.celrep.2021.108839. [DOI] [PubMed] [Google Scholar]

[R18] 18.Zhang X, Liu L, Yuan X, Wei Y, and Wei X (2019). JMJD3 in the regulation of human diseases. Protein and Cell 10, 864–882. 10.1007/s13238-019-0653-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R19] 19.Jones SE, Olsen L, and Gajhede M (2018). Structural Basis of Histone Demethylase KDM6B Histone 3 Lysine 27 Specificity. Biochemistry 57, 585–592. 10.1021/acs.biochem.7b01152. [DOI] [PubMed] [Google Scholar]

[R20] 20.Kruidenier L, Chung CW, Cheng Z, Liddle J, Che K, Joberty G, Bantscheff M, Bountra C, Bridges A, Diallo H, et al. (2012). A selective jumonji H3K27 demethylase inhibitor modulates the proinflammatory macrophage response. Nature 488, 404–408. 10.1038/nature11262. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R21] 21.Sengoku T, and Yokoyama S (2011). Structural basis for histone H3 lys 27 demethylation by UTX/KDM6A. Genes Dev 25, 2266–2277. 10.1101/gad.172296.111. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R22] 22.Spangler CJ, Skrajna A, Foley CA, Nguyen A, Budziszewski GR, Azzam DN, Arteaga EC, Simmons HC, Smith CB, Wesley NA, et al. (2023). Structural basis of paralog-specific KDM2A/B nucleosome recognition. Nat Chem Biol 19, 624–632. 10.1038/s41589-023-01256-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R23] 23.Bednar J, Garcia-Saez I, Boopathi R, Cutter AR, Papai G, Reymer A, Syed SH, Lone IN, Tonchev O, Crucifix C, et al. (2017). Structure and Dynamics of a 197 bp Nucleosome in Complex with Linker Histone H1. Mol Cell 66, 384–397. 10.1016/j.molcel.2017.04.012. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R24] 24.Chi YI, Stodola TJ, De Assuncao TM, Levrence EN, Tripathi S, Dsouza NR, Mathison AJ, Basel DG, Volkman BF, Smith BC, et al. (2021). Molecular mechanics and dynamic simulations of well-known Kabuki syndrome-associated KDM6A variants reveal putative mechanisms of dysfunction. Orphanet J Rare Dis 16, 66. 10.1186/s13023-021-01692-w. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R25] 25.Fyodorov DV, Zhou BR, Skoultchi AI, and Bai Y (2018). Emerging roles of linker histones in regulating chromatin structure and function. Nature Reviews Molecular Cell Biology 19,192–206. 10.1038/nrm.2017.94. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R26] 26.Ricci MA, Manzo C, García-Parajo MF, Lakadamyali M, and Cosma MP (2015). Chromatin fibers are formed by heterogeneous groups of nucleosomes in vivo. Cell 160, 1145–1158. 10.1016/j.cell.2015.01.054. [DOI] [PubMed] [Google Scholar]

[R27] 27.Yuan W, Wu T, Fu H, Dai C, Wu H, Liu N, Li X, Xu M, Zhang Z, Niu T, et al. (2012). Dense chromatin activates polycomb repressive complex 2 to regulate H3 lysine 27 methylation. Science 337, 971–975. 10.1126/science.1225237. [DOI] [PubMed] [Google Scholar]

[R28] 28.Willcockson MA, Healton SE, Weiss CN, Bartholdy BA, Botbol Y, Mishra LN, Sidhwani DS, Wilson TJ, Pinto HB, Maron MI, et al. (2021). H1 histones control the epigenetic landscape by local chromatin compaction. Nature 589, 293–298. 10.1038/s41586-020-3032-z. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R29] 29.Liu C, Yu J, Song A, Wang M, Hu J, Chen P, Zhao J, and Li G (2023). Histone H1 facilitates restoration of H3K27me3 during DNA replication by chromatin compaction. Nat Commun 14, 4081. 10.1038/s41467-023-39846-y. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R30] 30.Cooper S, Grijzenhout A, Underwood E, Ancelin K, Zhang T, Nesterova TB, Anil-Kirmizitas B, Bassett A, Kooistra SM, Agger K, et al. (2016). Jarid2 binds monoubiquitylated H2A lysine 119 to mediate crosstalk between Polycomb complexes PRC1 and PRC2. Nat Commun 7, 13661. 10.1038/ncomms13661. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R31] 31.Kalb R, Latwiel S, Baymaz HI, Jansen PWTC, Müller CW, Vermeulen M, and Müller J (2014). Histone H2A monoubiquitination promotes histone H3 methylation in Polycomb repression. Nat Struct Mol Biol 21, 569–571. 10.1038/nsmb.2833. [DOI] [PubMed] [Google Scholar]

[R32] 32.Jason LJM, Finn RM, Lindsey G, and Ausió J (2005). Histone H2A ubiquitination does not preclude histone H1 binding, but it facilitates its association with the nucleosome. Journal of Biological Chemistry 280, 4975–4982. 10.1074/jbc.M410203200. [DOI] [PubMed] [Google Scholar]

[R33] 33.Zhao J, Lan J, Wang M, Liu C, Fang Z, Song A, Zhang T, Wang L, Zhu B, Chen P, et al. (2024). H2AK119ub1 differentially fine-tunes gene expression by modulating canonical PRC1- and H1-dependent chromatin compaction. Mol Cell 84, 1191–1205. 10.1016/j.molcel.2024.02.017. [DOI] [PubMed] [Google Scholar]

[R34] 34.Hicks CW, Rahman S, Gloor SL, Fields JK, Husby NL, Vaidya A, Maier KE, Morgan M, Keogh MC, and Wolberger C (2024). Ubiquitinated histone H2B as gatekeeper of the nucleosome acidic patch. Nucleic Acids Res 52, 9978–9995. 10.1093/nar/gkae698. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R35] 35.Chen Y, Bates DL, Dey R, Chen PH, Machado ACD, Laird-Offringa IA, Rohs R, and Chen L (2012). DNA Binding by GATA Transcription Factor Suggests Mechanisms of DNA Looping and Long-Range Gene Regulation. Cell Rep 2, 1197–1206. 10.1016/j.celrep.2012.10.012. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R36] 36.Markert JW, Soffers JH, and Farnung L (2025). Structural basis of H3K36 trimethylation by SETD2 during chromatin transcription. Science 387, 528–533. 10.1126/science.adn6319. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R37] 37.Liu Y, Zhang Y, Xue H, Cao M, Bai G, Mu Z, Yao Y, Sun S, Fang D, and Huang J (2021). Cryo-EM structure of SETD2/Set2 methyltransferase bound to a nucleosome containing oncohistone mutations. Cell Discov 7, 32. 10.1038/s41421-021-00261-6. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R38] 38.Li W, Tian W, Yuan G, Deng P, Sengupta D, Cheng Z, Cao Y, Ren J, Qin Y, Zhou Y, et al. (2021). Molecular basis of nucleosomal H3K36 methylation by NSD methyltransferases. Nature 590, 498–503. 10.1038/s41586-020-03069-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R39] 39.Sato K, Kumar A, Hamada K, Okada C, Oguni A, Machiyama A, Sakuraba S, Nishizawa T, Nureki O, Kono H, et al. (2021). Structural basis of the regulation of the normal and oncogenic methylation of nucleosomal histone H3 Lys36 by NSD2. Nat Commun 12, 6605. 10.1038/s41467-021-26913-5. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R40] 40.Bilokapic S, and Halic M (2019). Nucleosome and ubiquitin position Set2 to methylate H3K36. Nat Commun 10, 3795. 10.1038/s41467-019-11726-4. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R41] 41.Sauer PV, Pavlenko E, Cookis T, Zirden LC, Renn J, Singhal A, Hunold P, Hoehne-Wiechmann MN, van Ray O, Kaschani F, et al. (2024). Activation of automethylated PRC2 by dimerization on chromatin. Mol Cell 84, 3885–3898.e8. 10.1016/j.molcel.2024.08.025. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R42] 42.Ntziachristos P, Tsirigos A, Welstead GG, Trimarchi T, Bakogianni S, Xu L, Loizou E, Holmfeldt L, Strikoudis A, King B, et al. (2014). Contrasting roles of histone 3 lysine 27 demethylases in acute lymphoblastic leukaemia. Nature 514, 513–517. 10.1038/nature13605. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R43] 43.Li Q, Zou J, Wang M, Ding X, Chepelev I, Zhou X, Zhao W, Wei G, Cui J, Zhao K, et al. (2014). Critical role of histone demethylase Jmjd3 in the regulation of CD4 + T-cell differentiation. Nat Commun 5. 10.1038/ncomms6780. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R44] 44.Park SH, Ayoub A, Lee YT, Xu J, Kim H, Zheng W, Zhang B, Sha L, An S, Zhang Y, et al. (2019). Cryo-EM structure of the human MLL1 core complex bound to the nucleosome. Nat Commun 10, 1165. 10.1038/s41467-019-13550-2. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R45] 45.Xue H, Yao T, Cao M, Zhu G, Li Y, Yuan G, Chen Y, Lei M, and Huang J (2019). Structural basis of nucleosome recognition and modification by MLL methyltransferases. Nature 573, 445–449. 10.1038/s41586-019-1528-1. [DOI] [PubMed] [Google Scholar]

[R46] 46.Rahman S, Hoffmann NA, Worden EJ, Smith ML, Namitz KEW, Knutson BA, Cosgrove MS, and Wolberger C (2022). Multistate structures of the MLL1-WRAD complex bound to H2B-ubiquitinated nucleosome. Proc Natl Acad Sci U S A 119, e2205691119. 10.1073/pnas.2205691119. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R47] 47.Jiao L, and Liu X (2015). Structural basis of histone H3K27 trimethylation by an active polycomb repressive complex 2. Science 350, aac4383. 10.1126/science.aac4383. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R48] 48.Margueron R, Justin N, Ohno K, Sharpe ML, Son J, Drury WJ, Voigt P, Martin SR, Taylor WR, De Marco V, et al. (2009). Role of the polycomb protein EED in the propagation of repressive histone marks. Nature 461, 762–767. 10.1038/nature08398. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R49] 49.Tamburri S, Lavarone E, Fernández-Pérez D, Conway E, Zanotti M, Manganaro D, and Pasini D (2020). Histone H2AK119 Mono-Ubiquitination Is Essential for Polycomb-Mediated Transcriptional Repression. Mol Cell 77, 840–856.e5. 10.1016/j.molcel.2019.11.021. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R50] 50.McGinty RK, Henrici RC, and Tan S (2014). Crystal structure of the PRC1 ubiquitylation module bound to the nucleosome. Nature 514, 591–596. 10.1038/nature13890. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R51] 51.Anderson CJ, Baird MR, Hsu A, Barbour EH, Koyama Y, Borgnia MJ, and McGinty RK (2019). Structural Basis for Recognition of Ubiquitylated Nucleosome by Dot1L Methyltransferase. Cell Rep 26, 1681–1690.e5. 10.1016/j.celrep.2019.01.058. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R52] 52.Skrajna A, Goldfarb D, Kedziora KM, Cousins EM, Grant GD, Spangler CJ, Barbour EH, Yan X, Hathaway NA, Brown NG, et al. (2020). Comprehensive nucleosome interactome screen establishes fundamental principles of nucleosome binding. Nucleic Acids Res 48, 9415–9432. 10.1093/nar/gkaa544. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R53] 53.Mastronarde DN (2005). Automated electron microscope tomography using robust prediction of specimen movements. J Struct Biol 152. 10.1016/j.jsb.2005.07.007. [DOI] [PubMed] [Google Scholar]

[R54] 54.Scheres SHW (2012). RELION: Implementation of a Bayesian approach to cryo-EM structure determination. J Struct Biol 180, 519–530. 10.1016/j.jsb.2012.09.006. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R55] 55.Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, and Ferrin TE (2004). UCSF Chimera - A visualization system for exploratory research and analysis. J Comput Chem 25, 1605–1612. 10.1002/jcc.20084. [DOI] [PubMed] [Google Scholar]

[R56] 56.Pettersen EF, Goddard TD, Huang CC, Meng EC, Couch GS, Croll TI, Morris JH, and Ferrin TE (2021). UCSF ChimeraX: Structure visualization for researchers, educators, and developers. Protein Science 30. 10.1002/pro.3943. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R57] 57.Emsley P, Lohkamp B, Scott WG, and Cowtan K (2010). Features and development of Coot. Acta Crystallogr D Biol Crystallogr 66, 486–501. 10.1107/S0907444910007493. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R58] 58.Liebschner D, Afonine PV, Baker ML, Bunkoczi G, Chen VB, Croll TI, Hintze B, Hung LW, Jain S, McCoy AJ, et al. (2019). Macromolecular structure determination using X-rays, neutrons and electrons: Recent developments in Phenix. Acta Crystallogr D Struct Biol 75, 861–877. 10.1107/S2059798319011471. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R59] 59.Schneider CA, Rasband WS, and Eliceiri KW (2012). NIH Image to ImageJ: 25 years of image analysis. Nature Methods 9, 671–675. 10.1038/nmeth.2089. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R60] 60.Tan S, Kern RC, and Selleck W (2005). The pST44 polycistronic expression system for producing protein complexes in Escherichia coli. Protein Expr Purif 40, 385–395. 10.1016/j.pep.2004.12.002. [DOI] [PubMed] [Google Scholar]

[R61] 61.Chen Z, Grzybowski AT, and Ruthenburg AJ (2015). Traceless semisynthesis of a set of histone 3 species bearing specific lysine methylation marks. ChemBioChem 16, 2071–2075. 10.1002/cbic.201402313. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R62] 62.Zheng JS, Tang S, Qi YK, Wang ZP, and Liu L (2013). Chemical synthesis of proteins using peptide hydrazides as thioester surrogates. Nat Protoc 8, 2483–2495. 10.1038/nprot.2013.152. [DOI] [PubMed] [Google Scholar]

[R63] 63.McGinty RK, Köhn M, Chatterjee C, Chiang KP, Pratt MR, and Muir TW (2009). Structure-activity analysis of semisynthetic nucleosomes: Mechanistic insights into the stimulation of Dot1L by ubiquitylated histone H2B. ACS Chem Biol 4, 958–968. 10.1021/cb9002255. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R64] 64.Peck JV, Fay JF, and Strauss JD (2022). High-speed high-resolution data collection on a 200 keV cryo-TEM. IUCrJ 9, 243–252. 10.1107/S2052252522000069. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R65] 65.Zivanov J, Nakane T, and Scheres SHW (2020). Estimation of high-order aberrations and anisotropic magnification from cryo-EM data sets in RELION-3.1. IUCrJ 7, 253–267. 10.1107/S2052252520000081. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R66] 66.Zheng SQ, Palovcak E, Armache JP, Verba KA, Cheng Y, and Agard DA (2017). MotionCor2: Anisotropic correction of beam-induced motion for improved cryo-electron microscopy. Nature Methods 14, 331–332. 10.1038/nmeth.4193. [DOI] [PMC free article] [PubMed] [Google Scholar]

[R67] 67.Rohou A, and Grigorieff N (2015). CTFFIND4: Fast and accurate defocus estimation from electron micrographs. J Struct Biol 192, 216–221. 10.1016/j.jsb.2015.08.008. [DOI] [PMC free article] [PubMed] [Google Scholar]

PERMALINK

Structural mechanism of H3K27 demethylation and crosstalk with heterochromatin markers

Chien-Chu Lin

Yani Zhao

Caroline A Foley

Aspen T Hawkins

Lindsey I James

Stephen V Frye

Robert K McGinty

SUMMARY

Grapgical Abstract

eTOC Blurb

INTRODUCTION

RESULTS

Structural insights into the KDM6B-nucleosome complex

Figure 1. Cryo-EM structure of the KDM6B-nucleosome complex.

Table 1.

Linker DNA-dependent nucleosome binding and demethylation

Figure 2. Linker DNA-dependent KDM6B activity.

KDM6B binding induces closed, compact linker DNA structure

KDM6B-cis-linker DNA interaction

Figure 3. KDM6B interaction with cis-linker DNA.

KDM6B-trans-linker DNA interaction

Figure 4. KDM6B interaction with trans-linker DNA.

KDM6B-wrapped DNA interaction

Figure 5. KDM6B interaction with wrapped DNA.

Histone H1 and H2AK119 ubiquitylation modulate KDM6B activity

Figure 6. Linker histone H1 and H2AK119ub1 modification antagonize KDM6B-mediated nucleosome demethylation.

DISCUSSION

Figure 7. Model of linker DNA-dependent KDM6B demethylation in heterochromatin.

Limitations of the study

RESOURECE AVAILABILITY

Lead contact

Materials availability

Data and code availability

STAR METHDS

EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS

Cell culture

METHOD DETAILS

Plasmid construction

Protein preparation

Nucleosome preparation

Cryo-EM sample preparation

Cryo-EM data collection and image processing

KDM6B–nucleosome complex model building

Electrophoretic mobility shift assay

Nucleosome demethylation assay

Dot Blot assay

Nucleosomal linker DNA FRET assay

QUANTIFICATION AND STATISTICAL ANALYSIS

Supplementary Material

Highlights.

ACKNOWLEDGEMENTS

Funding:

Footnotes

REFERENCES

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases