Portal de Periódicos da CAPES

The right place at the right time: chaperoning core histone variants

2015; Springer Nature; Volume: 16; Issue: 11 Linguagem: Inglês

10.15252/embr.201540840

1469-3178

Francesca Mattiroli, Sheena D’Arcy, Karolin Luger,

RNA and protein synthesis mechanisms

Review12 October 2015free access The right place at the right time: chaperoning core histone variants Francesca Mattiroli Corresponding Author Department of Molecular and Radiobiological Sciences, Howard Hughes Medical Institute, Colorado State University, Fort Collins, CO, USA Search for more papers by this author Sheena D'Arcy Department of Molecular and Radiobiological Sciences, Howard Hughes Medical Institute, Colorado State University, Fort Collins, CO, USA Search for more papers by this author Karolin Luger Department of Molecular and Radiobiological Sciences, Howard Hughes Medical Institute, Colorado State University, Fort Collins, CO, USA Search for more papers by this author Francesca Mattiroli Corresponding Author Department of Molecular and Radiobiological Sciences, Howard Hughes Medical Institute, Colorado State University, Fort Collins, CO, USA Search for more papers by this author Sheena D'Arcy Department of Molecular and Radiobiological Sciences, Howard Hughes Medical Institute, Colorado State University, Fort Collins, CO, USA Search for more papers by this author Karolin Luger Department of Molecular and Radiobiological Sciences, Howard Hughes Medical Institute, Colorado State University, Fort Collins, CO, USA Search for more papers by this author Author Information Francesca Mattiroli 1,†,‡, Sheena D'Arcy1,†,‡ and Karolin Luger1,† 1Department of Molecular and Radiobiological Sciences, Howard Hughes Medical Institute, Colorado State University, Fort Collins, CO, USA †Present address: Department of Biochemistry, University of Colorado Boulder, Howard Hughes Medical Institute, Boulder CO, USA †Present address: Department of Chemistry and Biochemistry, The University of Texas at Dallas, Richardson, TX, USA ‡These authors contributed equally *Corresponding author. Tel: +1 303 735 3224; E-mail: [email protected] EMBO Rep (2015)16:1454-1466https://doi.org/10.15252/embr.201540840 See the Glossary for abbreviations used in this article. PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info Abstract Histone proteins dynamically regulate chromatin structure and epigenetic signaling to maintain cell homeostasis. These processes require controlled spatial and temporal deposition and eviction of histones by their dedicated chaperones. With the evolution of histone variants, a network of functionally specific histone chaperones has emerged. Molecular details of the determinants of chaperone specificity for different histone variants are only slowly being resolved. A complete understanding of these processes is essential to shed light on the genuine biological roles of histone variants, their chaperones, and their impact on chromatin dynamics. Glossary ANP32E acidic leucine-rich nuclear phosphoprotein 32 family member E APLF aprataxin and PNK-like factor ART ADP-ribosyl-transferase ASF1 anti-silencing function 1 ATRX alpha thalassemia/mental retardation syndrome X-linked CABIN1 calcineurin-binding protein cabin-1 CAF-1 chromatin assembly factor 1 CATD CENP-A targeting domain CenH3 centromeric H3 variant CenH3CENP-A human centromeric H3 variant—centromere protein A CenH3CID fly centromeric H3 variant—centromere identifier CenH3Cse4 yeast centromeric H3 variant—chromosome segregation 4 CENP-B centromere protein B CENP-C centromere protein C CENP-N centromere protein N Chz1 chaperone for Htz1/H2A-H2B dimer 1 DAXX death domain-associated protein 6 FACT facilitates chromatin transcription HIRA histone cell cycle regulation-defective homolog A HJURP holliday junction recognition protein Nap1/NAP1 nucleosome assembly protein 1 p400/Tip60 histone acetyltransferase complex containing Tip60 and p400 PARP1 poly ADP-ribose polymerase 1 Scm3 suppressor of chromosome missegregation 3 SNF2 sucrose Non-fermenting 2—SWI/SNF catalytic subunit SNF2 SRCAP Snf2-related CREBBP activator protein Swr1 Swi2/Snf2-related protein 1 UBN1 ubinuclein-1 γH2A.X H2A.X S139ph Introduction The structural organization of eukaryotic DNA into chromatin is intricately linked to the regulation of many essential cellular processes. Nucleosomes are the structural unit of chromatin and are dynamically remodeled, assembled, and disassembled to allow DNA replication, transcription, and the different types of DNA repair. In this way, nucleosomes integrate a diverse array of cell signaling events to orchestrate the spatial arrangement of, and thus timely access to, the genetic information. At the center of this integration are the proteinaceous building blocks of the nucleosome, the core histone proteins. The core histones are small (< 20 kDa), positively charged polypeptides characterized by a histone fold domain. This domain is composed of three α-helices (α1-3) and assembles in the presence of a binding partner, particularly another histone, to form histone fold dimers 123 (Fig 1A). Nucleosomes contain stable heterodimers of histones H2A and H2B, as well as histones H3 and H4. Two copies of H2A–H2B pack on either side of a (H3–H4)2 tetramer, forming a histone octamer that wraps 147 bp of DNA 4 (Fig 1B). Histones also contain less-structured tails, which protrude outward and mediate interactions between nucleosomes, as well as with nuclear signaling proteins such as transcription factors and chromatin remodelers 5. The histone tails are heavily post-translationally modified and mediate a variety of epigenetic events. Modifications occur at virtually all amino acids and include mono-, di-, and tri-methylation, acetylation, mono-ubiquitination, and phosphorylation, among others. Figure 1. Canonical and variant nucleosomes(A) Elements of the histone fold and structures of Xenopus leavis H2A–H2B, H3–H4 and (H3–H4)2 (PDB ID: 1KX5). (B) Structure of the canonical Xenopus leavis nucleosome (PDB ID: 1KX5). Other nucleosome structures, such as the human nucleosome, are structurally similar. (C) Structure of the CenH3CENP-A-containing nucleosome (PDB ID: 3AN2). (D) Zoomed view of the αN helix of CenH3CENP-A (left) and H3 (right) involved in stabilizing the DNA ends. Histone H3 is blue, CenH3CENP-A is cyan, H4 is green, H2A is yellow, H2B is red, and DNA is white Download figure Download PowerPoint The bulk of the DNA is packaged by canonical histones H2A, H2B, H3 (also referred to as H3.1), and H4. Throughout evolution, however, an increase in complexity of chromatin-mediated signaling has led to a requirement for defined and localized changes in nucleosomes. A powerful mechanism to introduce such changes is the variation of histone primary sequence to produce the so-called histone variants (reviewed in 678). Histone variants resemble their canonical counterparts, but alter nucleosome function, by differing in sequence composition and in some cases by carrying additional domains (Fig 2). Phylogenetic studies have shown that histone variants have evolved in eukaryotic organisms in a non-concerted manner following a "birth-and-death" process 8910. The most pronounced sequence divergence is found in H2A and H2B variants, with H2A also having the greatest number of variants. This suggests some tolerance for H2A–H2B diversity within chromatin. In contrast, H4 has no variant identified and most H3 variants have only minimal sequence variation. This possibly reflects the evolutionary convergence required for H3–H4 deposition onto DNA, the first critical step in nucleosome formation 11. Figure 2. Schematic of select H3 and H2A variantsSchematic representation of select human H3 (A) and H2A (B) variants and the yeast homologs. The histone fold is shown in solid blue shades for H3 isoforms and in orange shades for H2A isoforms. The tails are shown as lines. Specific amino acid substitutions are reported. Different shades of color are used to indicate the degree of sequence divergence. Residues involved in forming the acidic patch on H2A isoforms are shown in red (where the acidic residues are present) and white (for H2A.B which lacks these residues). Sequence identity to the human canonical histone is shown in gray. Download figure Download PowerPoint Further to primary sequence, canonical and variant histones also differ in their expression patterns. Canonical histones are expressed from gene clusters in a cell cycle-dependent manner, peaking during S-phase to facilitate replication-dependent packaging of DNA (reviewed in 12). Most histone variants, however, are expressed throughout the cell cycle, meaning that their incorporation into chromatin is independent of DNA synthesis 1314. Histone variants can therefore fulfill specific functional niches that rely on tight temporally and spatially regulated alterations in chromatin. Such niches are required for transcriptional control of specific genes, maintenance of epigenetic information, repair of DNA damage, chromatin transition during spermatogenesis, and the formation of a single kinetochore on each chromosome 6151617. The challenge has been to determine the mechanistic basis of histone variant function and to decipher how histone variants are deposited in the right place at the right time. This review will focus on the specific crosstalk between histone variants and their chaperones, highlighting the regulation of their function and dynamics. Effect of histone variants on chromatin structure One of the initial hypotheses among researchers was that histone variants would alter nucleosome structure. However, based on in vitro nucleosome reconstitution experiments and structural studies, it is clear that variants in general do not dramatically alter the overall structure or composition of nucleosomes (e.g. 18192021). This is perhaps not surprising, as the variant residues are typically located distal to the histone–histone and histone–DNA interfaces that hold together the histone octamer and wrap the DNA. While massive reorganization of variant-containing nucleosomes is not observed, they can display subtle differences that likely contribute to their functional specialization. A well-studied histone variant that induces a subtle conformational change is H3 variant CenH3. CenH3 is the most divergent of all H3 variants, functionally conserved from yeast (CenH3Cse4) to human (CenH3CENP-A) (Fig 2A). From a phylogenetic point of view, CenH3 does not seem to have a single origin. Similar to other variants, it has evolved multiple times and has been subject to strong positive selection, likely due to its specific function 8. This variant plays an essential role in defining the heritable chromosomal centromere that directs kinetochore assembly 22. The histone stoichiometry of CenH3 nucleosomes has been a matter of debate, with some groups proposing a dynamic sub-stoichiometric composition 23242526, while others suggest it is octameric 272829. In yeast, this debate has been partly clarified by a study that shows the dynamic behavior of the one CenH3Cse4 nucleosome throughout the cell cycle 30. CenH3Cse4 nucleosomes are assembled de novo during S-phase sampling sub-octameric intermediates, which eventually lead to the stable octameric form 30. Structural studies show that in humans CenH3CENP-A can form octameric nucleosomes, at least in vitro. These studies also reveal that CenH3CENP-A nucleosomes wrap only 121 bp of DNA, compared to 147 bp wrapped in canonical nucleosomes 31 (Fig 1C). This more "loose" wrapping of DNA is consistent with biophysical studies 3132 and may be attributed to a shorter helix, N-terminal to the histone fold (Fig 1D). This feature may contribute to the reduced stability of CenH3 nucleosomes as compared to canonical ones, when assembled on endogenous DNA templates 32. It is not yet known whether the change in DNA wrapping is essential for CenH3-directed centromere formation. It is known, however, that CenH3 incorporation into chromatin is sufficient for kinetochore formation 3334. A central CENP-A targeting domain (CATD) encompassing a variant loop containing an Arg-Gly insertion and the α2 helix of the histone fold confers CenH3CENP-A function and interacts with inner kinetochore component CENP-N 35363738 (Fig 3B). CenH3CENP-A N- and C-terminal tails also directly interact with CENP-B and CENP-C kinetochore proteins 3536394041. These multiple interactions contribute to the stability of the CenH3CENP-A nucleosome, centromere identity, and the initiation of kinetochore assembly. These studies highlight how a variant-containing nucleosome can induce a specific downstream signal by exposing the histone variant surface for recognition by select nuclear proteins. Figure 3. Structures of H3 variants and their chaperones(A) Structure of the H3-H4 (PDB ID: 1KX5) and of the complex with the histone-binding domain of ASF1 (PDB ID: 2HUE). (B) Structure of CenH3CENP-A-H4 and of the complex with the histone-binding domain of HJURP (PDB ID: 3R45) and Scm3 (PDB: 2YFV). The CATD domain of CenH3 is orange. (C) Structure of H3.3–H4 and of the complex with the histone-binding domain of DAXX (PDB ID: 4H9N). Shown in orange are the H3.3-specific residues, with G90 shown as spheres. (D) Zoomed view of the DAXX Glu225 and H3.3 Gly90 interface, the key determinant for specificity in vivo. (E) Structure of H3.3–H4 with the histone-binding peptide of UBN1 and ASF1 (PDB ID: 4ZBJ). Shown in orange are the H3.3-specific residues, with G90 shown as spheres. In all panels, H3 isoforms are blue, CenH3 is cyan, H4 is green, and ASF1 is gold. Specific chaperones HJURP, Scm3, UBN1, and DAXX are shown in maroon. Download figure Download PowerPoint Other histone variants can also induce changes in nucleosome stability. This has been shown for H2A.B (also referred to as H2A.Bbd), which is often associated with active transcription and is involved in the histone to protamine transition in spermatogenesis 424344. It has also been shown for nucleosomes with both H2A.Z and H3.3, which are typically associated with transcriptional start sites 4546. H2A.B and H2A.Z also influence interactions between nucleosomes in a multi-nucleosome template or array. H2A.B and H2A.Z have only 35 and 51% similarity to canonical H2A, respectively, with notable sequence differences around a surface-exposed acidic patch (Fig 2B). This acidic patch mediates direct interactions with a vast number of proteins known to interact with the nucleosome, including other nucleosomes via the N-terminal tail of H4 in the neighboring particle 47. H2A.Z has an extended acidic region 18, while H2A.B has lost its acidic residues 44. Sedimentation studies show that these differences affect the ability of nucleosome arrays containing the variants to compact, with H2A.Z favoring compaction 48 and H2A.B reducing compaction 49. These cases again exemplify how histone variant incorporation into chromatin can alter the nucleosome surface to facilitate the engagement of variant-specific interactions responsible for the mediation of variant-specific functions. A role for histone chaperones An important part of variant and canonical histone biology is the dynamics of their incorporation into and eviction from chromatin. It is important to highlight that within the nucleosome, H2A–H2B dynamics have less impact on overall nucleosome stability than H3–H4 dynamics. In this light, incorporation or removal of H3 variants will require a more dramatic restructuring of chromatin as compared to H2A or H2B variants. H2A and H2B variants can be more easily exchanged without dismantling the central (H3–H4)2 tetramer–DNA complex. These dynamics are influenced by histone chaperones and ATP-dependent chromatin remodeling complexes. The latter primarily organize chromatin by sliding or ejecting assembled nucleosomes from chromatin. In this process, they can influence nucleosome composition by allowing the exchange of the more labile histone H2A–H2B dimers 50. Histone chaperones instead are a class of proteins that bind histones and facilitate their interactions to form the nucleosome 51. Chaperones are structurally diverse and are involved in histone trafficking between the nucleus and cytoplasm, as well as histone deposition into and eviction from chromatin 175253. With very few exceptions, it is important to realize that interactions between chaperones and histones are not always exclusive or binary. In fact, a particular histone can be bound by different chaperones. H2A–H2B, for example, is handled by numerous chaperones, including FACT and NAP1 5455. Variant H3.3–H4 can also bind chaperones ASF1, HIRA, and DAXX 56. On the flipside, a chaperone can also often bind more than one histone heterodimer. NAP1, for example, interacts with both H2A–H2B and H3–H4 as either a dimer or tetramer 57. ASF1 can also bind both H3.1–H4 and H3.3–H4 5859. This creates a redundancy that hinders the ability to fully understand the specific contributions of each chaperone in vivo. The structural underpinnings of chaperone redundancy stem from the fact that chaperones can bind histones in various manners. The most promiscuous chaperones, such as NAP1, interact primarily with the histone backbone, explaining their inability to discriminate between H2A–H2B and H3–H4, as well as many histone variants 54. Some more selective chaperones interact with histone surfaces that are conserved between canonical histones and their variant counterpart. ASF1, for example, binds the tetramerization H3 interface within a H3–H4 dimer 6061 (Fig 3A). This region is conserved among most H3 variants such that ASF1 can handle both H3.1–H4 and H3.3–H4 5859. Similarly, FACT mainly interfaces with H2B, explaining how it can recognize H2A–H2B containing different H2A variants 55 (Fig 4A). There are nonetheless chaperones specialized in handling specific histone variants (Table 1). These chaperones discriminate their substrate by recognizing specific side chains, unique to the variant. Figure 4. Structures of H2A variants and their chaperones(A) Structure of H2A-H2B (PDB ID: 1KX5) and of the complex with the histone-binding domain of the chaperone FACT (PDB ID: 4KHA). (B) Structure of H2A.Z–H2B and of the complex with the histone-binding region of Chz1 (PDB ID: 2JSS). (C) Structure of the complex between the Swr1 histone-binding region and the H2A.Z–H2B dimer (PDB ID: 4M6B). (D) Superposition of the H2A–H2B and H2A.Z–H2B dimers, close up view on the αC helix which is extended in H2A.Z due to a single amino acid deletion. (E) Structure of the ANP32E histone-binding region–H2A.Z–H2B complex (PDB ID: 4CAY). Shown in dark green is the region of the αC helix important for H2A.Z specificity. H2A is yellow, H2A.Z is gold, H4 is red, and FACT is light blue. Chaperones Chz1, Swr1, and ANP32E are purple. Download figure Download PowerPoint Table 1. Function of specific histone variant–chaperone pairs Histone variant Chaperone Function Experiments substantiating the specificity References CenH3 HJURP(Hs) Scm3(Sc) Cal1 (Dm) Centromere definition and maintenance Co-purification, co-localization, reduced variant incorporation upon chaperone mutation, correlated increased expression in tumors associated with lower survival rate, pull-down using amino acid swap mutants, crystal structure, and in vitro assembly assay 30 32 34 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 H3.3 DAXX (Hs) Pericentromeric and telomeric deposition of H3.3 Co-purification, co-localization, correlated increased expression in tumors, pull-down using amino acid swap mutants, crystal structure, and in vitro assembly assay 56 88 89 90 91 92 93 94 HIRA (Hs) Hir (Sc) H3.3 deposition in bulk DNA—Transcription Co-purification, co-localization, reduced variant incorporation upon chaperone mutation, pull-down using amino acid swap mutants, and crystal structure 56 59 83 95 96 97 H2A.Z Chz1 (Sc) H2A.Z incorporation—Transcription Co-purification, preferential association compared to canonical histone 115 116 118 Swr1 (Sc) H2A.Z incorporation—Transcription Co-purification, co-localization, functional overlap upon individual knock-down, reduced variant incorporation upon chaperone mutation, preferential association compared to canonical histone, and crystal structure 111 114 119 120 121 122 123 ANP32E (Hs) Removal of H2A.Z—DNA damage response Co-purification, co-localization, increased variant incorporation upon chaperone mutation, crystal structure 127 128 129 130 It is important to note that while specificity in chaperone recognition can be critical in determining the timely and localized dynamics of histone variants, it is by far not the only controlling factor. Regulation of histone chaperones themselves, in particular their expression pattern, modification, sub-cellular and sub-nuclear localization, and interacting partners, can greatly impact these dynamics. Understanding how the cell integrates these different layers of regulation to control histone variants has proven more complex than initially predicted. CenH3 and HJURP: a paradigm for chaperone specificity As introduced above, CenH3 is a striking example of a functionally specific histone variant. Such specificity requires a CenH3-specific chaperone to orchestrate the timely and accurate deposition of CenH3 into chromatin 22. CenH3-specific chaperones have been identified in a number of organisms. In humans, HJURP handles CenH3CENP-A 62636465, in yeast, Scm3 binds CenH3Cse4 666768, and in fly, CAL1 is responsible for deposition of CenH3CID 69 (Table 1). Although these chaperones have low sequence similarity and largely diverge in domain composition, they fulfill the same functional niche with regards to CenH3. Structures of the N-terminal region of yeast Scm3 and human HJURP bound to their cognate CenH3-H4 orthologues reveal a conserved mode of binding 707172 (Fig 3B). The structures, combined with biochemical studies, show that the aforementioned CATD in CenH3 is the primary interface with the chaperone 6270. Additional studies with HJURP further show that Ser68, a residue in the α1 helix, plays a role in chaperone specificity. A CenH3CENP-A point mutant replacing Ser68 with the H3 Gln residue shows loss of HJURP binding 72. It seems that Ser68 and the CATD work in concert to ensure accurate and faithful deposition of CenH3CENP-A by HJURP. The CATD was initially identified as it was sufficient to cause an H3 swap mutant to be incorporated into centromeric chromatin 3862. More recent studies highlight the complexity of this deposition in human cells, where phosphorylation at Ser68 and ubiquitination at Lys124 also coordinate CenH3CENP-A binding and deposition 7374. The complexity of CenH3 deposition is enhanced by the regulation of the chaperone itself. The necessity of HJURP for CenH3CENP-A deposition is obvious from the dramatic reduction of CenH3CENP-A incorporation in HJURP knockdown cells 65. Artificial tethering of HJURP to non-centromeric loci is also sufficient to induce kinetochore assembly 64. In yeast, Scm3 remains associated with the centromeric nucleosome, likely ensuring stability of the particle 30. HJURP function at centromeres relies on its dimerization 75 and DNA-binding capacity 76. It is also modulated by post-translational modifications, in particular phosphorylation that coordinates its action in a cell cycle-dependent manner 76. More recently, an additional role for HJURP in assembling de novo centromere was described and shown to be dependent on a direct interaction with CENP-C, a member of the kinetochore complex 34. Regulation of CenH3CENP-A by HJURP is of crucial importance for fidelity in chromosome segregation and cell division. Aberrant functions of the chaperone, as well as the histone variant, have been associated with chromosomal instability and ultimately cancer. Overexpression of HJURP and CenH3CENP-A has been observed in tumors, where their elevated mRNA levels are associated with lower patient survival 227778798081. Cells overexpressing CenH3CENP-A fail to arrest in the presence of DNA-damaging agents, suggesting that CenH3CENP-A overexpression may contribute to resistance against current chemotherapeutic drugs 82. Remarkably, when over-expressed, CenH3CENP-A will localize at sites of high histone turnover, forming heterotypic nucleosomes with the H3.3 variant 82. This ectopic localization of CenH3CENP-A is not dependent on HJURP, but rather on DAXX, a histone chaperone known to primarily bind H3.3–H4 82. These data illustrate the redundancy among chaperones and the malleability of histone binding. Nonetheless, the tight functional connection between CenH3CENP-A and HJURP highlights a unique specific link between this histone variant and chaperone. All together, these observations suggest that alongside the co-evolution of specific histone variant–chaperone pairs, flexibility can be accommodated in the system, particularly in disease conditions. H3.3 and DAXX: water-mediated specificity Compared to CenH3, the histone variant H3.3 is fairly widespread throughout the genome, marking euchromatin, pericentromeres, and telomeres 5683. It differs from canonical H3.1 and H3.2 by only five and four residues, respectively (Fig 2A). In humans, H3.3 is encoded by two genes with identical sequence and is expressed in a replication-independent manner 13. Knockout experiments result in phenotypes ranging from lethality in mice to infertility in flies 8485. Interestingly, replacing the endogenous H3.3 gene with H3.2 in flies can rescue the infertility phenotype, suggesting that histone supply outside of S-phase may be more relevant than sequence specificity in this context 86. Nonetheless, engineering single-point mutations on H3.1 to make it resemble H3.3 leads to a replication-independent incorporation of H3.1 in fly cells 13. This highlights that the variant residues are important for histone dynamics, which are likely mediated by histone chaperones. H3.3 was the first histone variant to be shown to be directly involved in carcinogenesis. Two-point mutations in the N-terminal tail, Lys27 to Met and Gly34 to Val or Arg, have been found in patient samples 8788. These mutations affect post-translational modifications of Lys27 and Lys36, key residues targeted by the mammalian epigenetic program. H3.3 chaperones are also mutated in a wide range of cancers 8990, and their mutation can correlate with H3.3 mutations in the same tumors 88. To date, two distinct H3.3 chaperones have been identified with separate functional niches. The first is HIRA, which controls H3.3 deposition at transcriptionally active loci, while the second is DAXX, which functions at pericentromeric and telomeric regions 56 (Table 1). DAXX was discovered as H3.3 incorporation into chromatin that was not completely abolished in HIRA knockout cells 5691. DAXX directly interacts with H3.3, although its chromatin-related functions are linked to ATRX, a member of the SNF2 ATP-dependent chromatin remodeling family 5692. Structures of the central domain of DAXX bound to H3.3–H4 have shed light on the mode of histone binding and H3.3 specificity 9394 (Fig 3C). DAXX envelopes H3.3–H4 and makes extensive contacts with both H3.3 and H4. DAXX forms a coiled coil with α2 helix of H4, somewhat similar to the one formed by HJURP or Scm3 with CenH3, suggesting a conserved mode of interaction. Like many other chaperones, DAXX interferes with H3.3–H4 tetramer formation and DNA binding as required for nucleosome formation 9394. Alongside structural studies, pull-down experiments using whole cell extracts found that a single DAXX residue, Glu225, was key in determining H3.3 specificity 9394. In cells, a Glu225Ala DAXX mutation is sufficient to induce a gain-of-binding effect for H3.2 in cells. As seen in the structure, Glu225 engages in a network of water-mediated hydrogen bonds to the histone pair 93. This network also involves H3.3 residue Gly90, which is a Met in H3.1 and H3.2 (Fig 3D). Pull-down experiments using single swap mutants between H3.1/2 and H3.3 residues show that Gly90 is the primary determinant for DAXX binding. A Gly90Met H3.3 mutant can no longer bind DAXX, while a Met90Gly H3.1 mutant has gained the ability to bind DAXX 9394. Despite these pull-down results, the Gly90Met H3.3 mutant can be crystallized in complex with DAXX, and the DAXX Glu225Ala mutant can be crystallized in complex with H3.3–H4 93. These structures have a different number of water molecules at the interface, suggesting a role for water-mediated interactions in DAXX recognition of H3–H4 variants. DAXX once again highlights the lack of tight structural stringency in chaperone recognition of histone variants, reinforcing the need for combined in vivo and in vitro studies. HIRA is a heterotrimer composed of HIRA, UBN1, and CABIN1. It is functionally connected to H3.3 dynamics as its depletion in cell culture affects H3.3 incorporation at promoters and gene bodies 56. Chromatin immunoprecipitation followed by high-throughput DNA sequencing shows that HIRA co-localizes with H3.3, primarily at transcriptional start sites of highly expressed genes 95. This link between HIRA and active transcription is further supported by its co-localization with both initiating and elongating RNA polymerase 83. Moreover, HIRA functions at sites of UV damage, where it is required for transcriptional restart 96. A short peptide of the UBN1 subunit has been identified as the primary interface of the HIRA complex to H3.3–H4 dimers 97. This peptide is specific for the H3.3 variant as it interacts with Gly90 97. The crystal structure of this peptide with H3.3–H4 bound by ASF1 confirms that these chaperones interact with structurally distinct re