The N‐terminal and C‐terminal halves of histone H2A.Z independently function in nucleosome positioning and stability

Abstract Nucleosome positioning and stability affect gene regulation in eukaryotic chromatin. Histone H2A.Z is an evolutionally conserved histone variant that forms mobile and unstable nucleosomes in vivo and in vitro. In the present study, we reconstituted nucleosomes containing human H2A.Z.1 mutants, in which the N‐terminal or C‐terminal half of H2A.Z.1 was replaced by the corresponding canonical H2A region. We found that the N‐terminal portion of H2A.Z.1 is involved in flexible nucleosome positioning, whereas the C‐terminal portion leads to weak H2A.Z.1‐H2B association in the nucleosome. These results indicate that the N‐terminal and C‐terminal portions are independently responsible for the H2A.Z.1 nucleosome characteristics.


| 539
Genes to Cells SATO eT Al. 2017). The canonical histones are produced during the S-phase of the cell cycle, coupled with DNA replication, and form nucleosomes just after chromatin replication (Marzluff et al., 2008). In higher eukaryotes, the canonical histones are encoded by multiple genes (Marzluff et al., 2008). Histone variants are usually encoded by one or more genes as nonallelic isoforms, and their amino acid sequences are different from those of the canonical histones (Talbert & Henikoff, 2010. The production of the histone variants is independent of the cell cycle (Talbert & Henikoff, 2017). Histone variants are considered to play specific roles in the regulation of genomic processes, such as DNA replication, repair, recombination and transcription (Giaimo, Ferrante, Herchenröther, Hake, & Borggrefe, 2019;Talbert & Henikoff, 2017).
In the present study, we reconstituted nucleosomes containing human H2A.Z.1 or its swapping mutants, in which the N-terminal or C-terminal half of H2A.Z.1 was replaced by the corresponding canonical H2A sequence. We found that the N-terminal and C-terminal portions of H2A.Z.1 are independently responsible for the multiple positioning and unstable features of the H2A.Z.1 nucleosome, respectively.

| RESULTS
and determined the histone composition of each nucleosome. All three H2A.Z.1 nucleosomes (ZN1, ZN2 and ZN3), as well as the H2A nucleosome, contained stoichiometric amounts of each histone ( Figure 1c). These results indicated that H2A.Z.1 formed nucleosomes with multiple positions, which is consistent with the single-molecule probing analyses of the reconstituted H2A.Z nucleosome (Chen et al., 2019;Rudnizky et al., 2016).
To determine the nucleosome positioning on the 193 base-pair DNA, we performed a restriction enzyme (AluI) cleavage assay coupled with micrococcal nuclease (MNase) treatment. Purified ZN1, ZN2 and ZN3 nucleosomes were treated with MNase, which preferentially digests naked DNA, and the DNA fragments protected by the nucleosome formation were analyzed by AluI cleavage (Figure 1d). In the canonical H2A nucleosome, the nucleosome position was mapped at the center of the 193 base-pair DNA fragment ( Figure 1e, left panel, and f). On the other hand, the three H2A.Z.1 nucleosomes were differently mapped: at the DNA ends (ZN1), at about 10 base pairs away from the DNA end (ZN2) and at the same position as the canonical SATO eT Al. nucleosome (ZN3) (Figure 1e and f). These results indicated that H2A.Z.1 allows the nucleosome to be located at multiple positions. We then tested the stability of the nucleosomes containing H2A.Z.1 H2A(61-129) and H2A.Z.1 H2A(1-60) by a thermal stability assay (Taguchi, Horikoshi, Arimura, & Kurumizaka, 2014). In this assay, the free H2A-H2B dissociated from the nucleosome is detected by the binding of the fluorescent probe SYPRO Orange (Figure 3a). The nucleosomes were denatured by a biphasic process. The first phase corresponds to the H2A-H2B dissociation from the nucleosome, and the second phase represents the H3-H4 dissociation. The H2A.Z.1-H2B dimer is dissociated from the nucleosome at a lower temperature than the canonical H2A-H2B dimer . Consistently, in all three H2A.Z nucleosomes with different positions (ZN1, ZN2 and ZN3), the first phase was shifted toward a lower temperature, as compared to the H2A nucleosome ( Figure 3b). Interestingly, the H2A.Z.1 H2A(61-129) nucleosome did not show this lower temperature shift of the first phase and exhibited similar stability to that of the canonical H2A nucleosome (Figure 3c). In contrast, the H2A.Z.1 H2A(1-60) nucleosome was clearly more unstable than the canonical H2A nucleosome (Figure 3c). These results indicated that the C-terminal half of H2A.Z.1 is responsible for the weak H2A.Z.1-H2B association with the nucleosome.

| Possible mechanism for the multiple positioning and instability of the H2A.Z.1 nucleosome
In the present study, we found that the N-terminal and C-terminal halves of H2A.Z.1 independently affect the nucleosome positioning and stability, respectively. In the nucleosome, the H2A-H2B dimers have multiple contact sites with DNA (Koyama & Kurumizaka, 2018). In fact, several H2A.Z-specific amino acid residues are located near the DNA and on the histone-histone interfaces (Figure 4a, colored blue and red). Among them, the H2A.Z.1 N-terminal half contacts the DNA backbone around the super helical locations (SHLs) 3.5 and 4.5 of the nucleosome (Figure 4a). In the H2A nucleosome structure, the H2A Lys15 and Lys36 residues are located near the SHL4.5 and SHL3.5 sites, respectively. In H2A.Z.1, these basic H2A Lys15 and Lys36 residues are replaced by the Val17 and Ser38 residues, respectively ( Figure 4a). These basic to neutral amino acid substitutions may reduce the histone-DNA interaction and could play roles in inducing the multiple positioning of the H2A.Z.1 nucleosome. It should be noted that the H2A.Z.1 Ser38 residue is substituted by Thr38 in H2A.Z.2 and is responsible for the enhanced mobility of H2A.Z.1 in living cells . Although the H2A.Z.1 Val17 and Ser38 residues are not highly conserved in the H2A.Z proteins of other species, neutral amino acid residues are conserved at these positions. These neutral amino acid residues may be important for the multiple positioning character of the H2A.Z.1 nucleosome.
Hydrophobic interactions are a major factor in the formation of the stable histone octamer (Taguchi et al., 2014). In the C-terminal half of H2A, the Val114 and Leu115 residues form a hydrophobic cluster with the H3 Ile112 and Val117 residues (Figure 4b, right panel). In the H2A.Z.1 nucleosome, the H2A Val114 residue is substituted by the hydrophilic Ser116 residue (Figure 4b, left-right). Other hydrophobic residues, H2A Ile62, Ile87 and Leu97, form a hydrophobic cluster with the H2B Phe65 residue in the H2A nucleosome (Figure 4c, right panel). In the H2A.Z.1 nucleosome, the corresponding residues, Val65, Ile90

| 541
Genes to Cells SATO eT Al. and Ile100, also form a hydrophobic cluster with the H2B Phe65 residue, but the interaction may be weaker because the side chain of H2A.Z.1 Val65 residue is shorter than that of the corresponding H2A Ile62 residue (Figure 4c, left panel). These H2A.Z.1specific residues located in the C-terminal half may be important for the weakened H2A.Z.1-H2B association in the nucleosome. Interestingly, the H2A.Z.1 Val65 and Ile100 residues are conserved between yeasts and mammals. These conserved residues in the hydrophobic cluster may play an important structural role and contribute to the dynamics of the H2A.Z nucleosome.

| Preparation of histone octamers and nucleosomes
For nucleosome reconstitution, the 193 base-pair DNA, H2A-H2B and H3.1-H4 were mixed at a 1:3.5:1.5 molar ratio, in a solution containing 10 mM Tris-HCl (pH 7.5), 1 mM dithiothreitol, 1 mM EDTA and 2 M KCl. Nucleosomes were reconstituted by the salt-dialysis method in 10 mM Tris-HCl buffer (pH 7.5), containing 1 mM dithiothreitol, 1 mM EDTA and 0.25 M KCl Kujirai et al., 2018). The reconstituted nucleosomes were heated at 55°C for 2 hr and then subjected to a native PAGE analysis, as described below. The reconstituted nucleosomes were separately purified by nondenaturing polyacrylamide gel electrophoresis using a Prep Cell apparatus (Bio-Rad), in 20 mM Tris-HCl (pH 7.5) and 1 mM dithiothreitol Kujirai et al., 2018). The separately purified nucleosomes  were analyzed by SDS-PAGE and native PAGE analyses, as described below.

| Native PAGE analysis
Nucleosomes (100 ng of DNA in Figure 1b; 1.3 µg of DNA in Figure 2d) were analyzed by nondenaturing 6% polyacrylamide gel electrophoresis in 0.2x TBE (17.8 mM Tris, 0.4 mM EDTA and 17.8 mM boric acid) with ethidium bromide staining. The gel images were obtained using an LAS4000 image analyzer (GE Healthcare).

| AluI restriction enzyme digestion assay coupled with MNase treatment
The purified nucleosomes (400 ng of DNA) were treated with MNase (TAKARA, 0.6 units for the H2A nucleosome, 0.7 units for ZN1, 0.5 units for ZN2 and 0.5 units for ZN3) at 25°C for 5 min, in a 20 µL reaction containing 32.5 mM Tris-HCl (pH 8.0), 10 mM Tris-HCl (pH 7.5), 25 mM NaCl, 1.25 mM CaCl 2 and 1.5 mM dithiothreitol. The reactions were stopped by adding 5 µL of stop solution, containing 6.2 mg/ml Proteinase K (Roche), 3.3% SDS and 100 mM EDTA. The DNA fragments were extracted with a Miniprep DNA Purification Kit (Promega). After digestion with 10 units of the AluI (New England Biolabs) restriction enzyme, the DNA fragments were fractionated by nondenaturing 10% polyacrylamide gel electrophoresis, stained with SYBR Gold and detected with an Amersham Typhoon scanner (GE Healthcare). The migration profiles of the DNA fragments were analyzed with the Image Gauge software (GE Healthcare). We then estimated the lengths of the DNA fragments produced by AluI cleavage with the 20 base-pair DNA markers as references, using the Multi Gauge software (Fujifilm).

| Thermal stability assay
The thermal stability of the nucleosomes containing H2A, H2A.Z, H2A.Z H2A(61-129) and H2A.Z H2A(1-60) was tested, as described previously (Taguchi et al., 2014). Briefly, the purified nucleosome (12 pmol) was mixed with 50 × SYPRO Orange dye (Sigma-Aldrich) and incubated in a solution containing 20 mM Tris-HCl buffer (pH 7.5) and 1 mM dithiothreitol. The SYPRO Orange fluorescence emitted by its binding to denatured histones was detected with a StepOnePlus TM Real-Time PCR system (Applied Biosystems). The temperature gradient ranged from 25°C to 95°C, in steps of 1°C/min. The fluorescence intensity was normalized relative to the fluorescence signal at 95°C.