The histone variant H2A.Z in gene regulation
The histone variant H2A.Z is involved in several processes such as transcriptional control, DNA repair, regulation of centromeric heterochromatin and, not surprisingly, is implicated in diseases such as cancer. Here, we review the recent developments on H2A.Z focusing on its role in transcriptional activation and repression. H2A.Z, as a replication-independent histone, has been studied in several model organisms and inducible mammalian model systems. Its loading machinery and several modifying enzymes have been recently identified, and some of the long-standing discrepancies in transcriptional activation and/or repression are about to be resolved. The buffering functions of H2A.Z, as supported by genome-wide localization and analyzed in several dynamic systems, are an excellent example of transcriptional control. Posttranslational modifications such as acetylation and ubiquitination of H2A.Z, as well as its specific binding partners, are in our view central players in the control of gene expression. Understanding the key-mechanisms in either turnover or stabilization of H2A.Z-containing nucleosomes as well as defining the H2A.Z interactome will pave the way for therapeutic applications in the future.
KeywordsH2A.Z H2Av Histone variant p400 Domino Tip60 CRISPR/Cas9
aryl hydrocarbon receptor
acute myeloid leukemia
androgen receptor-binding element
domain bromo-adjacent homology domain
bait protein–protein interaction-sequencing
bromodomain and PHD finger-containing transcription factor
bromodomain-containing protein 2
Egl-27 and MTA1 homology 2 domain
estrogen receptor alpha
estrogen receptor-binding element
facilitates chromatin transcription
forkhead box protein A111
forkhead box protein A2
fluorescence recovery after photobleaching
gain of function
general transcription factor
H3 lysine 4 monomethylation
H3 lysine 4 trimethylation
H3 lysine 14 acetylation
H3 lysine 27 acetylation
H3 lysine 122 succinylation
inverse fluorescence recovery after photobleaching
interferon-stimulated gene factor complex 3
loss of function
melanoma antigen-encoding gene
mouse embryonic stem cells
immunoprecipitation of H2A.Z-containing mononucleosomes obtained via MNase digestion of chromatin
mammalian target of rapamycin
nucleosome remodeling and deacetylase
nucleosome remodeling factor
open reading frames
Plasmodium falciparum GCN5
retinoic acid receptor γ
RNA polymerase II
- SANT domain
Swi3 Ada2 N-Cor and TFIIIB domain
something about silencing
SNF2-related CREBBP activator protein
small ubiquitin-like modifier
transcription factor binding site
trefoil factor 1
transcription start site
ubiquitin-specific protease 10
The chromatin structure represents the major modulator of all DNA-based processes such as gene transcription, DNA replication and repair. Chromatin is essentially composed of DNA and histone proteins that together form its basic unit, known as nucleosome. Within the core nucleosome, approximately 146 base pairs (bp) of DNA are wrapped in a left-handed superhelical turn around a protein structure composed of two copies each of the histones H3, H4, H2A and H2B whose crystal structure was solved more than 20 years ago . While histones H3 and H4 form a tetrameric structure known as nucleosome core that is positioned in the inner region of the nucleosome, the histones H2A and H2B are rather located on the nucleosomal surface. In addition, the linker histone H1 can contact the entry and exit sites of the nucleosomal DNA resulting in a more compact structure . Of note, histones are characterized by the presence of a characteristic histone fold domain from which unstructured N- and C-terminal tails protrude . In the process of nucleosome assembly, ATP-dependent remodelers assemble and arrange histone octamers in a highly dynamic fashion . Posttranslational modifications (PTMs) are placed predominantly on flexible histone tails but also within the histone fold domains [4, 5, 6]. Importantly, replication-independent (hereafter referred to as non-canonical) histone variants can substitute replication-dependent (hereafter referred to as canonical) histones and are specifically positioned within the genome.
While canonical histones are expressed exclusively during the replication phase of the cell cycle, histone variants are expressed throughout the cell cycle. Canonical histones are encoded by multi-copy genes that lack introns and present a stem loop structure at the 3′-end of their mRNAs. In contrast, genes encoding histone variants are biallelic, sometimes characterized by introns and poly-adenylated at the 3′-end of their mRNAs. As consequence, some histone variants-encoding genes are subjected to alternative splicing. Apart from histone H4, all histone protein families (H2A, H2B, H3 and H1) are characterized by specialized histone variants and among them the most studied family is the H2A, which comprises several members including macroH2A, H2A.X, H2A.Bbd and H2A.Z . As an aside, an H4 variant has been identified in the urochordata Oikopleura dioica  and in Trypanosoma brucei , suggesting the possibility that H4 variants may be expressed also in other organisms.
The histone variant H2A.Z has been intensively studied over the last three decades elucidating not only the enzymatic activities required for its chromatin deposition but also the interlinked posttranslational regulatory mechanisms as well as its dynamics in response to signaling pathways. The focus of this review is to summarize and discuss the current knowledge on the histone variant H2A.Z. In particular, we will emphasize the mechanisms of its chromatin deposition and removal, its posttranslational regulation and its interaction partners. Further, we will also review the latest developments concerning H2A.Z’s deregulation or mutations in diseases and how newest technologies can be used to manipulate histone variant levels.
Historical perspective and overview
The histone variant H2A.Z was originally identified in 1980 in mouse L1210 cells . Few years later, studies in Tetrahymena thermophila observed the presence of H2A.Z in the transcriptionally active macronucleus but not in the transcriptionally inactive micronucleus . Later, the Drosophila melanogaster homolog H2Av, was identified  and shown to be essential .
Subsequently, the mammalian H2A.Z gene was cloned in 1990  and similarly to Drosophila, it was found to be essential since the mouse knockout displays an early-lethal embryonic phenotype . Surprisingly, earlier studies revealed that H2A.Z depletion is not lethal in Saccharomyces cerevisiae but “only” leads to reduced growth, a phenotype that can be efficiently rescued via reintroduction of the H2A.Z-encoding gene from Tetrahymena , marking the evolutionary conservation of H2A.Z.
Recently, this H2A.Z isoform scenario has become more complex due to the identification of an alternatively spliced and primate-specific isoform of H2A.Z.2 (hereafter referred to as H2A.Z.2.1 ), known as H2A.Z.2.2 (Fig. 1) that is expressed in a wide range of tissues with maximum transcript expression in human brain tissues . H2A.Z.2.1 and H2A.Z.2.2 differ within their C-terminal region, and H2A.Z.2.2-containing nucleosomes are less stable compared to H2A.Z.2.1-containing ones due to reduced binding to neighboring histones within one octamer . Previously, Adam and colleagues have shown that the C-terminal region of the yeast H2A.Z protein interacts with RNA polymerase II (RNAPII), promoting its recruitment at promoters . Given that the C-terminus of H2A.Z.2.2 significantly differs from the one of H2A.Z.1 and H2A.Z.2.1, it will be interesting to test whether also H2A.Z.2.2 is able to interact with RNAPII and to evaluate whether the different histone variants, present within a different genomic localization, may differentially influence RNAPII recruitment and finally transcription.
H2A.Z has been linked to diverse biological processes such as memory [29, 30, 31, 32] and epithelial-to-mesenchymal transition (EMT, ). At molecular level, it has been implicated in heterochromatin regulation [34, 35, 36, 37, 38], anti-silencing function at boundaries in yeast [39, 40, 41, 42], DNA repair [25, 43, 44, 45, 46, 47, 48] and transcriptional regulation [21, 23, 28, 29, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88]. How H2A.Z regulates such a wide spectrum of different processes is not fully understood, and it is even more surprising that H2A.Z regulates both transcriptional repression and activation.
Interestingly, the Drosophila H2Av variant, encoded by the His2Av gene, fulfills the functions of both mammalian H2A.Z and H2A.X variants. Similar to mammalian H2A.Z, H2Av regulates heterochromatin formation  and gene regulation  as also marked by its enrichment in euchromatic regions . On the other hand, the mammalian histone variant H2A.X is a pivotal factor for DNA damage responses. H2A.X is phosphorylated on its unique serine 139 (called γ-H2A.X) upon DSBs and helps recruiting the DNA repair machinery . Similarly, upon DSBs, Drosophila H2Av is specifically phosphorylated on a serine residue conserved in mammalian H2A.X . Subsequently, phosphoH2Av is acetylated by the histone acetyltransferase (HAT) dTip60, leading to its exchange with an unmodified H2Av at DSB sites . In line with that, loss-of-function (LoF) of Tip60 leads to accumulation of phosphoH2Av .
In this review, we will discuss the recent literature elucidating the contrasting facets of H2A.Z in gene regulation.
Mechanisms of loading and removal of the histone variant H2A.Z
Composition of the SWR1, NuA4, Ep400/Tip60, SRCAP and Domino complexes
FLJ11730 (Meaf6; hEaf6)
Swr1 is evolutionary conserved: The Drosophila homolog is known as Domino, while in mammals there are two homologs called SRCAP (SNF2-related CREBBP activator protein) and Ep400, which are both able to catalyze the incorporation of H2A.Z within chromatin [49, 112]. Biochemical purifications of the human Ep400-containing complex surprisingly unveiled that it is composed of not only homologous subunits of the SWR1 complex but also contains subunits that are exclusively found within the yeast NuA4 complex (Table 1, [109, 113, 114, 115, 116]). The same is true for the Drosophila complex (Table 1, ). This suggests that the Drosophila/human complex, known as p400/Tip60 complex, represents a physical merge of the yeast SWR1 and NuA4 complexes. This hypothesis is further supported by the observation that human Ep400 represents a fusion of yeast Swr1 and Eaf1, subunits of SWR1 and NuA4 complexes, respectively . In contrast, biochemical purification of the human SRCAP-containing complex showed that it does not contain any histone acetyltransferase activity (Table 1, [112, 116]).
While the SWR1, p400/Tip60 and SRCAP complexes load H2A.Z within chromatin, there are also mechanisms to evict H2A.Z. ANP32E was recently shown to remove H2A.Z from nucleosomes in human cells during DNA damage [130, 131]. Its depletion leads to increased H2A.Z occupancy, and it co-localizes genome-wide with H2A.Z [130, 131, 132].
Together, the identification of the protein machineries placing H2A.Z is an important step forward for the better understanding of the dual role of H2A.Z in gene regulation: based on the loading machinery involved in the locus-specific deposition of H2A.Z (SRCAP or p400/Tip60), different PTMs of H2A.Z can be deposited leading to the recruitment of different H2A.Z interactors that finally result in a different transcriptional output (repression or activation).
Posttranslational regulation of H2A.Z determines the transcriptional output
Histone variants, like all the canonical histones, can be dynamically decorated by various PTMs including acetylation, methylation, phosphorylation, SUMOylation and ubiquitination. Genetic data indicate that H2A.Z can serve as a buffer to quench phenotypic noise via modulating transcriptional efficiencies . Looking at gene expression, H2A.Z depletion can either lead to the upregulation of genes, for example ∆Np63α and Notch signaling targets [29, 49, 57, 72, 134], or to downregulation such as estrogen signaling [28, 50, 54]. Thus, H2A.Z is able to modulate, by dampening, either transcriptional repression or activation. In our view, PTMs of H2A.Z play a major role in this transcriptional buffering function.
Posttranslation modifications (PTMs) identified in human, mouse and/or rat histone variant H2A.Z
Ubiquitination of H2A.Z (H2A.Zub) occurs on different lysine residues as summarized in Fig. 1 and Table 2. However, the function of only few ubiquitinated lysine residues has been described. Sarcinella and colleagues observed ubiquitination of H2A.Z on K120 and K121 and linked these modifications to X-chromosome inactivation (XCI) . K120, K121 and K125 monoubiquitination (K120ub1, K121ub1 and K125ub1, respectively) is mediated by RING1B [58, 145]. Active H2A.Z deubiquitination, mediated by USP10 (Ubiquitin-Specific Protease 10), is required to induce gene expression . Furthermore, RNF168 ubiquitinates H2A.Z, but the exact target lysine is still unknown . Surprisingly, Ku and colleagues observed in mouse embryonic stem cells (mESCs) that a fraction of H2A.Zub1 is also acetylated on its N-terminal tail: This population is more acetylated and contains a differential acetylation profile compared to the non-ubiquitinated H2A.Z . It still needs to be investigated whether such a dually modified H2A.Z is an exclusive feature of mESCs or does also occur in other cell types.
The small ubiquitin-like modifier (SUMO) is another member of the ubiquitin peptide family (Fig. 1 and Table 2). SUMOylation of H2A.Z (H2A.Zsu) in yeast has been linked to DNA repair, as it is required for the recruitment of DSBs to the nuclear periphery . Similarly, in HeLa cells, H2A.Z.2su by the SUMO E3 ligase PIAS4 is involved in DNA repair , but the exact site modified by PIAS4 has not yet been identified.
In the last years, also H2A.Z methylation was identified which, based on methylation state and the specific lysine residue to be modified, can have different transcriptional outputs (Fig. 1 and Table 2). Monomethylation of lysine 7 of H2A.Z (H2A.ZK7me1), mediated by SETD6, is associated with gene repression in mESCs , while dimethylation of lysine 101 (H2A.ZK101me2) is linked to gene induction in human cells .
Together, like the PTMs of canonical histones, there is a complicated network of activating and repressing marks also for H2A.Z. However, there are few valuable marks that will, in our view, pave the way for unraveling the molecular mechanisms of H2A.Z in gene regulation.
The H2A.Z interactome
Assuming that placement of H2A.Z and PTMs of H2A.Z are read and interpreted, it is important to first define the “H2A.Z interactome”. The working hypothesis is interacting factors will give decisive insights about the molecular mechanisms conducting either gene activation or repression.
Two of the many found H2AZ interactors (Table 3) were biochemically verified and functionally characterized. The first one is BRD2 that was identified by affinity purification of MNase-digested chromatin as an H2A.Z binder on chromatin level . Further, BRD2 was proposed to be a decisive downstream mediator that couples H2A.Z to AR-induced gene activation . It binds H2A.Z-containing nucleosomes via its bromodomains promoted by H4 hyperacetylation and prefers, mediated by a so far unknown mechanism, binding to the H2A.Z.1 over the H2A.Z.2 isoform [21, 144]. Strikingly, H2A.Z.2 was shown to promote and/or maintain BRD2, E2F1 and histone acetylation levels in malignant melanoma . H2A.Z.2 recruits BRD2 and E2Fs, along with HAT activity, to promoters of E2F target genes in melanoma cells, facilitating expression of cell cycle genes and, ultimately, promoting cell proliferation. The other, recently identified protein is PWWP2A that was shown to tightly bind H2A.Z via a multivalent binding mode . PWWP2A’s direct binding to H2A.Z is predominantly mediated by a C-terminal section of its internal protein region of no known homology or structure. Real-time-lapse microscopy imaging showed halt of PWWP2A-depleted cells in mitosis for up to 10 h. A similar effect has been observed in H2A.Z double knockout vertebrate cells . Hence, PWWP2A might be the mediator of the H2A.Z-dependent cell cycle progression phenotype. Interestingly, PWWP2A, as well as H2A.Z, interacts with an MTA1-specific subcomplex of the NuRD complex that was named “M1HR” . This subcomplex consists exclusively of MTA1, RBBP4/7 and HDAC2 and excludes CHD, GATAD2 and MBD proteins. Depletion of PWWP2A increased acetylation of histones in a subset of H2A.Z-containing enhancers bound by PWWP2A where it presumably regulates histone acetylation levels via M1HR recruitment.
Furthermore, H2A.Z was shown to interact with components of complexes involved in a multitude of biological processes, for example DNA damage repair (e.g., MSH2 and MSH6 of the mismatch repair complex, as well as PIR51, RAD23B and XPC), gene activation (e.g., MLL/KMT complex, PHF2, BRD8, MEAF6, ING3), gene repression (e.g., TIP27/JAZF1, BAHD1, BCORL1, MIER1 and CDYL), various transcription factors (e.g., DIDO1, MYPOP, ZFX/Y), chromatin remodeling (e.g., SMCA1 of the nucleosome remodeling factor (NuRF) complex) and proteins whose function(s) remain yet elusive like the RAI1 complex [154, 155], ZNF512B, MAGEA10, PHF20L1 and ZNF768.
Besides the further need to validate all mentioned putative H2A.Z interaction partners in independent biochemical and functional assays, it is tempting to speculate that these many interactors are one important reason why H2A.Z bears transcriptional activating as well as repressing features. At the same time, it shows that although a lot about H2A.Z’s interactome was resolved, its role in recruiting transcription-regulating complexes to their destinations on chromatin still remains a puzzle.
H2A.Z at enhancers and promoters
In the past, genomic localization of H2A.Z has been mostly reported at the TSS of genes, but more recently it is becoming increasingly clear that H2A.Z is also found at enhancers. In yeast, H2A.Z is strongly enriched at the TSS of both active and inactive genes . Its occupancy at the TSS negatively correlates with gene expression: H2A.Z occupancy is more pronounced at poorly expressed genes compared to induced genes [61, 62, 63, 64]. In contrast, genome-wide studies in human CD4+ T cells observed H2A.Z enrichment mainly at the TSS of active genes [157, 158]. Subsequently, this scenario was further refined with the observation that H2A.Z occupancy at TSS correlates with the level of transcriptional output: While Ku and colleagues observed a negative correlation , other studies observed a positive correlation between gene expression and H2A.Z occupancy [60, 67, 84]. Furthermore, usage of dynamic systems showed that gene induction is associated with reduced H2A.Z occupancy at TSS [54, 68, 70, 71, 72] as well as at enhancers [49, 54, 57, 69, 71, 72]. Similarly, in Drosophila as well as in plants, H2A.Z occupies the promoter in absence of gene expression but it decreases upon gene induction [65, 66]. Notably, H2A.Z occupancy strongly correlates with H3K4 methylation states [58, 74, 157, 159, 160, 161], further marking its involvement in gene poising and activation. The inverse correlation between H2A.Z occupancy and transcription is also reflected in RNAPII occupancy, : H2A.Z is actively excluded from coding regions by the RNAPII-associated remodelers FACT (facilitates chromatin transcription) and spt6 . Deletion of spt16, a gene encoding a FACT subunit, or of spt6, leads to H2A.Z accumulation at coding regions, a phenotype associated with increased cryptic transcription . This is in line with the observation that H2A.Z-containing nucleosomes are not enriched with H3K36me3, a histone mark associated with transcriptional elongation [143, 159] and provide a mechanistic explanation to the increased H2A.Z occupancy observed at coding regions upon reduced transcription [67, 76, 77]. Additionally, H2A.Z knockdown leads to reduced RNAPII recruitment at TSSs in Saccharomyces cerevisiae and human cells [28, 76] and reduced TBP (TATA-binding protein) occupancy in Saccharomyces cerevisiae . However, it plays a positive function in preventing RNAPII stalling, as its depletion increases this phenomenon [78, 79], further marking the strong relationship between H2A.Z and RNAPII. Increased cryptic transcription observed at coding regions upon H2A.Z accumulation in yeast  would suggest the involvement of H2A.Z in promoter usage, but it must be noted that, at least in human cells, H2A.Z is strongly enriched at facultative heterochromatin without leading to cryptic transcription . A further indication that H2A.Z may be involved in promoter usage is represented by the observation that dispersed core promoters (promoters in which the TSS spreads over hundreds of nucleotides) show a stronger H2A.Z enrichment compared to focused core promoters (promoters in which the TSS occurs in a narrow genomic window of few nucleotides ). This dispersion in promoter usage may be the consequence of a different stability of H2A.Z-containing nucleosomes , an aspect that will be further discussed in the next section. To note, not only H2A.Z enrichment but also the proximity of H2A.Z-containing nucleosomes to the TSSs influences gene expression .
Histone variant H2A.Z at enhancers
Effects of H2A.Z depletion
Reduced induction upon treatment
Reduced induction upon treatment
Lack of induction upon treatment
Increased induction upon treatment
H2A.Z and the nucleosome-free regions (NFRs)
The TSS of active genes was previously known as a nucleosome-free region (NFR) or nucleosome-depleted region (NDR). Interestingly, this has been challenged by Jin and Felsenfeld in 2007 . In this study, the authors observed that nucleosomes containing both H2A.Z and H3.3 histone variants are highly unstable and found at regulatory regions such as promoters and enhancers . When nucleosomes are isolated at low salt concentrations, it is possible to observe occupancy of H3.3/H2A.Z double-containing nucleosomes at NFRs, an occupancy that is lost when high salt concentrations are used , further suggesting that NFRs might indeed be not nucleosome-free . Such unstable nucleosomes are also enriched at NFRs in Drosophila and yeast [170, 171]; however, the reason why the H3.3/H2A.Z double-containing nucleosomes are highly unstable remains unclear. While H2A.Z and the canonical H2A differ significantly in their L1 loop , cell-free studies observed that nucleosomes composed of both H2A and H2A.Z (defined as heterotypic, ) are more stable than H2A.Z only-containing nucleosomes (defined as homotypic), which are less stable than H2A homotypic nucleosomes [174, 175]. However, a highly unstable heterotypic nucleosome occupies the TSS in the G1 phase of the cell cycle [176, 177]. Furthermore, it should be noted that cell-free studies observed that H3.3 does not alter the stability of H2A.Z both homo- and heterotypic nucleosomes [175, 178]. Based on these data, one could think that neither the incorporation of H3.3 into an H2A.Z-containing nucleosome or the presence of an H2A.Z/H2A heterotypic nucleosome could be responsible for the nucleosome instability observed at NFRs; however, the different approaches used, in vitro (cells) versus cell-free assays, may lead to discrepancies and actually the cell-free approaches may lead to underestimate the nucleosome instability as consequence of the lack of PTMs and/or interactors that may contribute to the regulation of nucleosome stability in a physiological context. It is possible that such nucleosomes would be H2A.Z homotypic. However, at least in Drosophila, homotypic H2A.Z nucleosomes are not enriched at the TSS , excluding this possibility. In contrast, two more studies observed increased stability of the H2A.Z-containing nucleosomes compared to the H2A-containing nucleosomes in cell-free assays [180, 181]. It seems that this increased stability can be counteracted by histone acetylation, including H2A.Zac . In more detail, it appears that H2A.Zac is the key modification that destabilizes the nucleosome and that acetylation of other histone proteins alone is not sufficient to achieve this destabilization; even more, heterotypic nucleosomes are destabilized by H2A.Zac . While the previous studies focused on H2A.Z.1, another study found structural differences in the L1 loop when comparing this isoform with H2A.Z.2.1 . Furthermore, H2A.Z.2.2-containing nucleosomes seem to be less stable than the H2A.Z.2.1-containing ones . As consequence, at least in vertebrates (or in primates in the case of H2A.Z.2.2), the high instability of the nucleosomes located at NFRs can still be due to the occupancy of the different H2A.Z isoforms that can organize different homo- and heterotypic nucleosomes (also in combination with H3.3) that are regulated by different combinations of PTMs. However, this remains a hypothesis that needs to be tested to identify the mechanism(s) how the nucleosomes occupying the “NFR” become unstable.
The role of H2A.Z in nucleosome positioning
H2A.Z and DNA methylation
Studies in Arabidopsis thaliana have shown that H2A.Z is excluded from sites enriched in DNA methylation and H2A.Z-occupied sites display low levels of DNA methylation . This anti-correlation between H2A.Z occupancy and DNA methylation is recapitulated in other organisms [88, 185, 186, 187, 188] and involves acetylation of H2A.Z , which is known to be the predominant mechanism in gene activation. Gene reactivation observed upon loss of DNA methylation, obtained via pharmacological inhibition or knockdown of DNA methyltransferases (DNMTs), is associated with a gain in H2A.Z occupancy [184, 189]. Similarly, increased H2A.Z occupancy, obtained via LoF of ANP32E that removes H2A.Z from chromatin , leads to reduced DNA methylation , while the opposite is observed upon gain-of-function (GoF) of ANP32E, LoF of the H2A.Z loading machinery or depletion of H2A.Z itself [88, 184, 188]. Notably, this mechanism may involve also another histone variant: macroH2A. In fact, it reversely correlates with H2A.Z occupancy but positively correlates with DNA methylation and gene silencing . However, in contrast to the study of Yang and colleagues , Barzily-Rokni and colleagues do not observe gain in H2A.Z occupancy upon pharmacological inhibition of DNMTs alone or combined with macroH2A depletion even if gene expression is re-established . One possibility to explain this discrepancy is represented by the different concentration of 5-azacytidine used in these studies [189, 190].
Our current model for H2A.Z in gene regulation
The AR and ER systems represent good examples to explain the function of H2A.Z in gene transcription (Fig. 3). In the AR system, the PSA gene can be considered as the prototype of this pathway (Fig. 3a): In the absence of androgen (OFF state), H2A.Z is loaded by the SRCAP  and/or p400/Tip60  complexes. In this repressed configuration, H2A.Z is monoubiquitinated at both enhancers and promoters  potentially by RING1B [58, 145]. Upon androgen stimulation (ON), H2A.Z is deubiquitinated by USP10 and its occupancy decreases [51, 52]. Of note, H2A.Zac correlates with AR induction [59, 122] and similarly, the occupancy of the p400/Tip60 complex increases upon AR induction . The recruitment of the p400/Tip60 complex is mediated by its MRG15 subunit which recognizes H3K4 methylation states  while SRCAP has been shown to interact with AR . In the case of the estrogen signaling cascade, we focus on the case of the TFF1 locus (Fig. 3b): In the OFF state, forkhead box protein A1 (FoxA1) binds to a distal enhancer (FoxA1-binding site, FBS) of the TFF1 locus where it recruits the p400/Tip60 complex supporting H2A.Z loading . Lack of H2A.Z at the TFF1 promoter, leads to a poorly defined nucleosome occupancy in the repressed/poised state (OFF, ). Upon activation of the pathway, the p400/Tip60 complex is recruited at the TFF1 promoter by ERα which binds to its cognate sequences (ERE). At the TFF1 promoter, the p400/Tip60 complex loads H2A.Z leading to a better-defined nucleosome positioning. At the same time, H2A.Z occupancy decreases at the FoxA1-bound distal enhancer .
From the above, some general rules for H2A.Z in gene regulation can be postulated: At genes that are poised/repressed (OFF), repressive marks of H2A.Z are found and as consequence its LoF leads to upregulation. At genes that are active, activating PTMs of H2A.Z, such as H2A.Zac, are found and as consequence H2A.Z LoF leads to downregulation.
In a repressed (OFF) or poised state, the H2A.Z deposition machinery is recruited by TFs and/or histone modifications to chromatin. This recruitment can be transient but still allows an exchange of H2A with H2A.Z. In the OFF state, H2A.Z is deacetylated by the deacetylation machinery and ubiquitinated on its C-terminus by RING1B. Upon gene activation (ON), additional TFs and/or histone modifications lead to the recruitment of the loading/acetylation/deubiquitination machinery. This triggers H2A.Z acetylation and deubiquitination, finally leading to transcriptional activation.
Still a number of open questions remain there. For example, the specificity of the p400/Tip60 and SRCAP complexes awaits to be determined in higher eukaryotes. Furthermore, there may be several p400-containing complexes that may act at different stages of gene transcription, the one without the acetyltransferase function and the other with Tip60 present. It is possible that these two complexes act in a stepwise fashion, one after the other. It is also possible that the different complexes are preferentially found at either promoters or enhancer elements (as described above). In addition, the H2A.Z removing factor ANP32E is also found in complex with Ep400 and Tip60  suggesting that deposition, acetylation and removal of H2A.Z could be coupled in a stepwise manner. Chromatin-IP (ChIP) experiments with the relevant antibodies in dynamically regulated systems will answer such questions.
H2A.Z in diseases
The histone variant H2A.Z, its PTMs and interacting proteins have been linked to several diseases, most notably cancer. H2A.Z expression is upregulated in metastatic melanoma , breast [191, 192], prostate [52, 59, 193], colorectal , liver [195, 196], bladder  and lung  cancer. Similarly, H2A.Z protein levels are elevated during cardiac hypertrophy  but decreased in diseased vascular tissues .
In prostate cancer, H2A.Zac has a pro-oncogenic role. It promotes activation of oncogenes and repression of tumor suppressor genes . Regarding genomic occupancy, H2A.Zac increases at the TSS of oncogenes, while it decreases at the TSS of tumor suppressor genes [59, 60] and its genomic redistribution in prostate cancer leads to the activation of AR-associated neo-enhancers . It must be marked that in metastatic melanoma both H2A.Z isoforms, H2A.Z.1 and H2A.Z.2, are upregulated; however, only the depletion of the H2A.Z.2 but not H2A.Z.1 leads to reduced proliferation . In contrast to that, H2AFZ but not H2AFV is overexpressed in liver cancer and its knockdown results in reduced proliferation and inhibits the cancer cells’ metastatic potential . These data support the notion that the different H2A.Z isoforms, which differ in only three amino acids, have distinct roles in the development of different tumor types.
While the data discussed so far highlight the upregulation of H2A.Z in cancer, additional mechanisms of H2A.Z deregulation may involve aberrant expression of the machineries involved in H2A.Z modifications and/or chromatin deposition/removal. For example, the methyltransferase SMYD3, which is upregulated in several cancer types, promotes proliferation of breast cancer cells and tumorigenesis . This is achieved because SMYD3 supports H2A.Z methylation, which is required to activate the expression of the cyclin A1-encoding (CCNA1) gene . Similarly, Tip60 is downregulated in acute myeloid leukemia (AML) samples  and present with mono-allelic loss in lymphomas, head-and-neck and mammary carcinoma . In line with that, Tip60 has a tumor suppressor function in colon . However, whether decreased Tip60 expression has any impact on H2A.Z deposition or acetylation is unknown.
Finally, the enzymes involved in the H2A.Z deposition can also be useful therapeutic targets, for example knockdown of SRCAP reduces cell proliferation of prostate cancer cells .
The CRISPR/Cas9 system as a tool to deplete H2A.Z
In summary (Fig. 4), the CRISPR/Cas9 technology can be used to target the 5′UTR or 3′UTR leading to deregulation of the target by promoting mRNA destabilization or translational block. However, targeting the 3′UTR can also be used to upregulate gene expression by increasing mRNA stability or translation. In our view, it is tempting to speculate that other histone variants or even canonical histones could also be targeted (or tagged/mutated/replaced) in this way as well as, the loading and removal machineries involved in the chromatin regulation of histone variants.
Conclusions and perspective
We propose that the plastic behavior of H2A.Z at regulatory regions is ideally suited to buffer phenotypic noise by modulating transcriptional efficiency, both repressive and activating. Mechanistically, high levels of unmodified H2A.Z and ubiquitination of H2A.Z may serve as a roadblock for transcription; acetylation of H2A.Z and subsequent removal of H2A.Z may enhance transcription rate and/or help recruiting RNAPII or other activating complexes.
In the future, new tools such as highly specific antibodies against single-modified H2A.Z residues are needed to characterize the function of H2A.Z in development and in pathological settings. Since the current antibodies against H2A.Z (unmodified and pan-acetyl-H2A.Z) are exquisitely specific, there is a high chance that more specific reagents will have a major impact on chromatin and transcription research.
All authors wrote the manuscript. BDG and AH prepared the figures and tables. All authors read and approved the final manuscript.
The authors declare that they have no competing interests.
Consent for publication
All authors have approved the manuscript.
This work was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation)—Project number 109546710—TRR81 to T.B. and S.B.H, and the Heisenberg program (BO 1639/5-1) by the DFG (German Research Foundation) and the Excellence Cluster for Cardio Pulmonary System (ECCPS) to T.B and the Cardio-Pulmonary Institute (CPI) to S.B.H. Funding for open access charge was provided by the DFG collaborative research TRR81. B.D.G. is supported by a Research Grant of the University Medical Center Giessen and Marburg (UKGM).
Ethics approval and consent to participate
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 19.Eirin-Lopez JM, Gonzalez-Romero R, Dryhurst D, Ishibashi T, Ausio J. The evolutionary differentiation of two histone H2A.Z variants in chordates (H2A.Z-1 and H2A.Z-2) is mediated by a stepwise mutation process that affects three amino acid residues. BMC Evol Biol. 2009;9:31. https://doi.org/10.1186/1471-2148-9-31.CrossRefPubMedPubMedCentralGoogle Scholar
- 24.Dunn CJ, Sarkar P, Bailey ER, Farris S, Zhao M, Ward JM, et al. Histone hypervariants H2A.Z.1 and H2A.Z.2 play independent and context-specific roles in neuronal activity-induced transcription of Arc/Arg3.1 and other immediate early genes. Neuro. 2017. https://doi.org/10.1523/eneuro.0040-17.2017.CrossRefGoogle Scholar
- 27.Bonisch C, Schneider K, Punzeler S, Wiedemann SM, Bielmeier C, Bocola M, et al. H2A.Z.2.2 is an alternatively spliced histone H2A.Z variant that causes severe nucleosome destabilization. Nucleic Acids Res. 2012;40(13):5951–64. https://doi.org/10.1093/nar/gks267.CrossRefPubMedPubMedCentralGoogle Scholar
- 52.Dryhurst D, McMullen B, Fazli L, Rennie PS, Ausio J. Histone H2A.Z prepares the prostate specific antigen (PSA) gene for androgen receptor-mediated transcription and is upregulated in a model of prostate cancer progression. Cancer Lett. 2012;315(1):38–47. https://doi.org/10.1016/j.canlet.2011.10.003.CrossRefPubMedGoogle Scholar
- 58.Ku M, Jaffe JD, Koche RP, Rheinbay E, Endoh M, Koseki H, et al. H2A.Z landscapes and dual modifications in pluripotent and multipotent stem cells underlie complex genome regulatory functions. Genome Biol. 2012;13(10):R85. https://doi.org/10.1186/gb-2012-13-10-r85.CrossRefPubMedPubMedCentralGoogle Scholar
- 60.Valdes-Mora F, Song JZ, Statham AL, Strbenac D, Robinson MD, Nair SS, et al. Acetylation of H2AZ is a key epigenetic modification associated with gene deregulation and epigenetic remodeling in cancer. Genome Res. 2012;22(2):307–21. https://doi.org/10.1101/gr.118919.110.CrossRefPubMedPubMedCentralGoogle Scholar
- 62.Li B, Pattenden SG, Lee D, Gutierrez J, Chen J, Seidel C, et al. Preferential occupancy of histone variant H2AZ at inactive promoters influences local histone modifications and chromatin remodeling. Proc Natl Acad Sci USA. 2005;102(51):18385–90. https://doi.org/10.1073/pnas.0507975102.CrossRefPubMedGoogle Scholar
- 63.Buchanan L, Durand-Dubief M, Roguev A, Sakalar C, Wilhelm B, Stralfors A, et al. The Schizosaccharomyces pombe JmjC-protein, Msc1, prevents H2A.Z localization in centromeric and subtelomeric chromatin domains. PLoS Genet. 2009;5(11):e1000726. https://doi.org/10.1371/journal.pgen.1000726.CrossRefPubMedPubMedCentralGoogle Scholar
- 64.Wan Y, Saleem RA, Ratushny AV, Roda O, Smith JJ, Lin CH, et al. Role of the histone variant H2A.Z/Htz1p in TBP recruitment, chromatin dynamics, and regulated expression of oleate-responsive genes. Mol Cell Biol. 2009;29(9):2346–58. https://doi.org/10.1128/mcb.01233-08.CrossRefPubMedPubMedCentralGoogle Scholar
- 67.Cui K, Zang C, Roh TY, Schones DE, Childs RW, Peng W, et al. Chromatin signatures in multipotent human hematopoietic stem cells indicate the fate of bivalent genes during differentiation. Cell Stem Cell. 2009;4(1):80–93. https://doi.org/10.1016/j.stem.2008.11.011.CrossRefPubMedPubMedCentralGoogle Scholar
- 76.Hardy S, Jacques PE, Gevry N, Forest A, Fortin ME, Laflamme L, et al. The euchromatic and heterochromatic landscapes are shaped by antagonizing effects of transcription on H2A.Z deposition. PLoS Genet. 2009;5(10):e1000687. https://doi.org/10.1371/journal.pgen.1000687.CrossRefPubMedPubMedCentralGoogle Scholar
- 80.Bargaje R, Alam MP, Patowary A, Sarkar M, Ali T, Gupta S, et al. Proximity of H2A.Z containing nucleosome to the transcription start site influences gene expression levels in the mammalian liver and brain. Nucleic Acids Res. 2012;40(18):8965–78. https://doi.org/10.1093/nar/gks665.CrossRefPubMedPubMedCentralGoogle Scholar
- 81.Guillemette B, Bataille AR, Gevry N, Adam M, Blanchette M, Robert F, et al. Variant histone H2A.Z is globally localized to the promoters of inactive yeast genes and regulates nucleosome positioning. PLoS Biol. 2005;3(12):e384. https://doi.org/10.1371/journal.pbio.0030384.CrossRefPubMedPubMedCentralGoogle Scholar
- 97.Kobor MS, Venkatasubrahmanyam S, Meneghini MD, Gin JW, Jennings JL, Link AJ, et al. A protein complex containing the conserved Swi2/Snf2-related ATPase Swr1p deposits histone variant H2AZ into euchromatin. PLoS Biol. 2004;2(5):E131. https://doi.org/10.1371/journal.pbio.0020131.CrossRefPubMedPubMedCentralGoogle Scholar
- 100.Altaf M, Auger A, Monnet-Saksouk J, Brodeur J, Piquet S, Cramet M, et al. NuA4-dependent acetylation of nucleosomal histones H4 and H2A directly stimulates incorporation of H2AZ by the SWR1 complex. J Biol Chem. 2010;285(21):15966–77. https://doi.org/10.1074/jbc.m110.117069.CrossRefPubMedPubMedCentralGoogle Scholar
- 107.Auger A, Galarneau L, Altaf M, Nourani A, Doyon Y, Utley RT, et al. Eaf1 is the platform for NuA4 molecular assembly that evolutionarily links chromatin acetylation to ATP-dependent exchange of histone H2A variants. Mol Cell Biol. 2008;28(7):2257–70. https://doi.org/10.1128/MCB.01755-07.CrossRefPubMedPubMedCentralGoogle Scholar
- 111.Osada S, Sutton A, Muster N, Brown CE, Yates JR 3rd, Sternglanz R, et al. The yeast SAS (something about silencing) protein complex contains a MYST-type putative acetyltransferase and functions with chromatin assembly factor ASF1. Genes Dev. 2001;15(23):3155–68. https://doi.org/10.1101/gad.907201.CrossRefPubMedPubMedCentralGoogle Scholar
- 116.Robert F, Hardy S, Nagy Z, Baldeyron C, Murr R, Dery U, et al. The transcriptional histone acetyltransferase cofactor TRRAP associates with the MRN repair complex and plays a role in DNA double-strand break repair. Mol Cell Biol. 2006;26(2):402–12. https://doi.org/10.1128/MCB.26.2.402-412.2006.CrossRefPubMedPubMedCentralGoogle Scholar
- 122.Ito S, Kayukawa N, Ueda T, Taniguchi H, Morioka Y, Hongo F, et al. MRGBP promotes AR-mediated transactivation of KLK3 and TMPRSS2 via acetylation of histone H2A.Z in prostate cancer cells. Biochim Biophys Acta Gene Regul Mech. 2018. https://doi.org/10.1016/j.bbagrm.2018.07.014.CrossRefGoogle Scholar
- 128.Wang AY, Schulze JM, Skordalakes E, Gin JW, Berger JM, Rine J, et al. Asf1-like structure of the conserved Yaf9 YEATS domain and role in H2A.Z deposition and acetylation. Proc Natl Acad Sci U S A. 2009;106(51):21573–8. https://doi.org/10.1073/pnas.0906539106.CrossRefPubMedPubMedCentralGoogle Scholar
- 137.Myers FA, Lefevre P, Mantouvalou E, Bruce K, Lacroix C, Bonifer C, et al. Developmental activation of the lysozyme gene in chicken macrophage cells is linked to core histone acetylation at its enhancer elements. Nucleic Acids Res. 2006;34(14):4025–35. https://doi.org/10.1093/nar/gkl543.CrossRefPubMedPubMedCentralGoogle Scholar
- 139.Mehta M, Braberg H, Wang S, Lozsa A, Shales M, Solache A, et al. Individual lysine acetylations on the N terminus of Saccharomyces cerevisiae H2A.Z are highly but not differentially regulated. J Biol Chem. 2010;285(51):39855–65. https://doi.org/10.1074/jbc.m110.185967.CrossRefPubMedPubMedCentralGoogle Scholar
- 140.Chittuluru JR, Chaban Y, Monnet-Saksouk J, Carrozza MJ, Sapountzi V, Selleck W, et al. Structure and nucleosome interaction of the yeast NuA4 and Piccolo-NuA4 histone acetyltransferase complexes. Nat Struct Mol Biol. 2011;18(11):1196–203. https://doi.org/10.1038/nsmb.2128.CrossRefPubMedPubMedCentralGoogle Scholar
- 160.Won KJ, Choi I, LeRoy G, Zee BM, Sidoli S, Gonzales-Cope M, et al. Proteogenomics analysis reveals specific genomic orientations of distal regulatory regions composed by non-canonical histone variants. Epigenetics Chromatin. 2015;8:13. https://doi.org/10.1186/s13072-015-0005-9.CrossRefPubMedPubMedCentralGoogle Scholar
- 183.Lantermann AB, Straub T, Stralfors A, Yuan GC, Ekwall K, Korber P. Schizosaccharomyces pombe genome-wide nucleosome mapping reveals positioning mechanisms distinct from those of Saccharomyces cerevisiae. Nat Struct Mol Biol. 2010;17(2):251–7. https://doi.org/10.1038/nsmb.1741.CrossRefPubMedGoogle Scholar
- 189.Yang X, Noushmehr H, Han H, Andreu-Vieyra C, Liang G, Jones PA. Gene reactivation by 5-aza-2’-deoxycytidine-induced demethylation requires SRCAP-mediated H2A.Z insertion to establish nucleosome depleted regions. PLoS Genet. 2012;8(3):e1002604. https://doi.org/10.1371/journal.pgen.1002604.CrossRefPubMedPubMedCentralGoogle Scholar
- 194.Dunican DS, McWilliam P, Tighe O, Parle-McDermott A, Croke DT. Gene expression differences between the microsatellite instability (MIN) and chromosomal instability (CIN) phenotypes in colorectal cancer revealed by high-density cDNA array hybridization. Oncogene. 2002;21(20):3253–7. https://doi.org/10.1038/sj.onc.1205431.CrossRefPubMedGoogle Scholar
- 196.Yang HD, Kim PJ, Eun JW, Shen Q, Kim HS, Shin WC, et al. Oncogenic potential of histone-variant H2A.Z.1 and its regulatory role in cell cycle and epithelial-mesenchymal transition in liver cancer. Oncotarget. 2016;7(10):11412–23. https://doi.org/10.18632/oncotarget.7194.CrossRefPubMedPubMedCentralGoogle Scholar
- 206.Chang JW, Zhang W, Yeh HS, Park M, Yao C, Shi Y, et al. An integrative model for alternative polyadenylation, IntMAP, delineates mTOR-modulated endoplasmic reticulum stress response. Nucleic Acids Res. 2018;46(12):5996–6008. https://doi.org/10.1093/nar/gky340.CrossRefPubMedPubMedCentralGoogle Scholar
- 208.Collin J, Mellough CB, Dorgau B, Przyborski S, Moreno-Gimeno I, Lako M. Using zinc finger nuclease technology to generate CRX-reporter human embryonic stem cells as a tool to identify and study the emergence of photoreceptors precursors during pluripotent stem cell differentiation. Stem Cells. 2016;34(2):311–21. https://doi.org/10.1002/stem.2240.CrossRefPubMedGoogle Scholar
- 216.Boskovic A, Bender A, Gall L, Ziegler-Birling C, Beaujean N, Torres-Padilla ME. Analysis of active chromatin modifications in early mammalian embryos reveals uncoupling of H2A.Z acetylation and H3K36 trimethylation from embryonic genome activation. Epigenetics. 2012;7(7):747–57. https://doi.org/10.4161/epi.20584.CrossRefPubMedGoogle Scholar
- 219.Weinert BT, Scholz C, Wagner SA, Iesmantavicius V, Su D, Daniel JA, et al. Lysine succinylation is a frequently occurring modification in prokaryotes and eukaryotes and extensively overlaps with acetylation. Cell Rep. 2013;4(4):842–51. https://doi.org/10.1016/j.celrep.2013.07.024.CrossRefPubMedGoogle Scholar
- 225.Kettenbach AN, Schweppe DK, Faherty BK, Pechenick D, Pletnev AA, Gerber SA. Quantitative phosphoproteomics identifies substrates and functional modules of Aurora and Polo-like kinase activities in mitotic cells. Sci Signal. 2011;4(179):5. https://doi.org/10.1126/scisignal.2001497.CrossRefGoogle Scholar
- 228.Mertins P, Yang F, Liu T, Mani DR, Petyuk VA, Gillette MA, et al. Ischemia in tumors induces early and sustained phosphorylation changes in stress kinase pathways but does not affect global protein levels. Mol Cell Proteomics. 2014;13(7):1690–704. https://doi.org/10.1074/mcp.M113.036392.CrossRefPubMedPubMedCentralGoogle Scholar
- 234.Udeshi ND, Svinkina T, Mertins P, Kuhn E, Mani DR, Qiao JW, et al. Refined preparation and use of anti-diglycine remnant (K-epsilon-GG) antibody enables routine quantification of 10,000 s of ubiquitination sites in single proteomics experiments. Mol Cell Proteomics. 2013;12(3):825–31. https://doi.org/10.1074/mcp.O112.027094.CrossRefPubMedGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.