Nucleosome assembly protein-1 is a linker histone chaperone in Xenopus eggs Keishi Shintomi*, Mari Iwabuchi*, Hideaki Saeki†, Kiyoe Ura†, Takeo Kishimoto*, and Keita Ohsumi*‡ *Laboratory of Cell and Developmental Biology, Graduate School of Bioscience and Biotechnology, Tokyo Institute of Technology, 4259 Nagatsuta, Midori-ku, Yokohama 226-8501, Japan; and †Division of Gene Therapy Science, Osaka University Graduate School of Medicine, 2-2 Yamada-oka, Suita, Osaka 565-0870, Japan Edited by Mark T. Groudine, Fred Hutchinson Cancer Research Center, Seattle, WA, and approved April 27, 2005 (received for review January 31, 2005)

In eukaryotic cells, genomic DNA is primarily packaged into nucleosomes through sequential ordered binding of the core and linker histone proteins. The acidic proteins termed histone chaperones are known to bind to core histones to neutralize their positive charges, thereby facilitating their proper deposition onto DNA to assemble the core of nucleosomes. For linker histones, however, little has been known about the regulatory mechanism for deposition of linker histones onto the linker DNA. Here we report that, in Xenopus eggs, the linker histone is associated with the Xenopus homologue of nucleosome assembly protein-1 (NAP1), which is known to be a chaperone for the core histones H2A and H2B in Drosophila and mammalian cells [Ito, T., Bulger, M., Kobayashi, R. & Kadonaga, J. T. (1996) Mol. Cell Biol. 16, 3112–3124; Chang, L., Loranger, S. S., Mizzen, C., Ernst, S. G., Allis, C. D. & Annunziato, A. T. (1997) Biochemistry 36, 469 – 480]. We show that NAP-1 acts as the chaperone for the linker histone in both sperm chromatin remodeling into nucleosomes and linker histone binding to nucleosome core dimers. In the presence of NAP-1, the linker histone is properly deposited onto linker DNA at physiological ionic strength, without formation of nonspecific aggregates. These results strongly suggest that NAP-1 functions as a chaperone for the linker histone in Xenopus eggs. Xenopus laevis 兩 chromatosome assembly 兩 cell-free system

I

n eukaryotic cells, genomic DNA is packaged primarily into the nucleosome, the basic repeating unit of chromatin (1, 2). Nucleosomes are assembled through sequential ordered binding of histone proteins, which are grouped into two types, the highly conserved core histones and the much more variable linker histones. Depending on the types of binding histones and the length of DNA protected by the histone binding from micrococcal nuclease digestion, two levels of the nucleosome structures are distinguished. The nucleosome core, the most fundamental structural unit of chromatin, comprises ⬇146 bp of DNA wrapped in 1.75 left-handed superhelical turns around two molecules of each core histone protein, H2A, H2B, H3, and H4. In addition to the core histones, metazoan nucleosomes also contain linker histones, such as histones H1 and H5, which bind to linker DNA, the DNA between nucleosome core particles, thereby protecting an additional ⬇20 bp of DNA. The chromatin particle that consists of an octamer of core histones, one molecule of the linker histone, and ⬇160–170 bp of DNA is called the chromatosome. In nucleosome assembly, histone binding to DNA is thought to occur in a three-step process: a tetramer or two dimers of histones H3 and H4 is first deposited onto DNA, then two dimers of histones H2A and H2B are deposited to yield the nucleosome core, and subsequently a single molecule of linker histone is deposited onto the core to form the chromatosome. For histones to be properly deposited onto DNA, regulatory factors that mediate the deposition are required, because, when mixed together at physiological pH and ionic strength, purified histones and DNA form an insoluble aggregate, but not nucleosomes. This aggregation is caused by the strong ionic interactions 8210 – 8215 兩 PNAS 兩 June 7, 2005 兩 vol. 102 兩 no. 23

between the negatively charged DNA backbone and the positively charged basic residues in the histones and can be prevented by the addition of negatively charged polymers that neutralize the positive charge of the histones. In the cell, the proper deposition of histones onto DNA within the nucleus is mediated by acidic proteins called histone chaperones, which bind and effectively neutralize highly charged histones, thereby preventing their nonspecific aggregation with DNA (for review, see ref. 3). For the core histones, several chaperones have been identified to date, with nucleoplasmin being the first on the list (ref. 4; for review, see refs. 5–7). Most chaperones for the core histones are shown to have a preference for either H3-H4 or H2A-H2B, and some of them are also implicated in various chromatin processes such as DNA repair, transcriptional silencing, and sperm chromatin decondensation (for review, see refs. 3, 6, and 7). For the linker histones, however, much less is known about the regulation of linker histone deposition onto the nucleosome core. Corresponding to the lack of knowledge on the mechanism of chromatosome formation, the precise location of the linker histone on the nucleosome core, i.e., the structure of the chromatosome, has not been determined (8, 9), although the detailed structure of the nucleosome core particle has been resolved (10). In this study, we examine linker histone-binding proteins in Xenopus eggs, of which chaperones for the core histones have been identified; the core histones H2A-H2B and H3-H4 stored in the egg cytoplasm are associated with histone chaperone nucleoplasmin and N1兾N2, respectively (11, 12). By using egg extracts, we found that B4 (⫽H1X or H1M), the linker histone of Xenopus eggs (13–15), is associated with nucleosome assembly protein-1 (NAP-1), first identified in mammalian cells as a protein that facilitates the assembly of nucleosome cores in vitro, and has been found in a wide variety of organisms, including yeast, soybean, Drosophila, and humans (16–19). Several lines of evidence also indicate that NAP-1 functions in vivo as a chaperone for H2A-H2B (19, 20). We also show that NAP-1 facilitates the proper deposition of the linker histone to the linker DNA, the result suggesting that NAP-1 functions as a linker histone chaperone in Xenopus eggs. Materials and Methods Isolation of B4-Binding Proteins. The soluble fraction of Xenopus egg extract was obtained by centrifugation (200,000 ⫻ g, 90 min at 4°C) of activated cytostatic factor-arrested extract, prepared as described (21). Column chromatography of the soluble fraction on a Superose 12 (HR 10兾30) and a Mono Q (HR 5兾5) column (Amersham Pharmacia) was performed as described

This paper was submitted directly (Track II) to the PNAS office. Abbreviations: NAP-1, nucleosome assembly protein-1; xNAP-1, Xenopus homologues of Nap-1; His-xNAP-1, histidine-tagged p60 xNAP-1; His-B4, histidine-tagged B4. Data deposition: The sequences reported in this paper have been deposited in the GenBank database (accession nos. DQ020266 –DQ020269). ‡To

whom correspondence should be addressed. E-mail: [email protected].

© 2005 by The National Academy of Sciences of the USA

www.pnas.org兾cgi兾doi兾10.1073兾pnas.0500822102

Cloning and Expression. Total RNA was isolated from unfertilized

Xenopus eggs. First-strand cDNA was synthesized from total RNA by superscript II reverse transcriptase and an oligo(dT) primer and then used as the template for PCR. 5⬘ and 3⬘ degenerate primers were designed to target two amino acid sequences of p60, AALQERL, and DWKKGKNV, respectively. The PCR amplification consisted of 30 cycles with 30 sec at 94°C, 1 min at 50°C, and 1 min at 72°C, followed by an extension step at 72°C for 10 min. Based on the sequences of the cDNA clones, 3⬘ and 5⬘ RACE was performed against second-strand cDNAs synthesized by using the Marathon cDNA Amplification Kit (Clontech). Two types of full-length cDNAs encoding p60 were isolated. The two cDNAs were cloned into pBS-RNT3, transcribed, and translated in vitro in the presence of [35S]methionine. cDNA of B4, the linker histone subtype specific to Xenopus early embryos (13), was isolated by PCR. Full-length cDNAs encoding p60 xNAP-1 and B4 were cloned into the pTrcHis plasmid vector (Invitrogen) and transformed into Escherichia coli BL21. His-6-tagged recombinant proteins were purified by using His䡠Bind Resin (Novagen) and used for immunization of rabbits and complex formation in vitro. Remodeling of Sperm Chromatin. Crude Xenopus egg extracts were

prepared as described (21). Demembranated Xenopus sperm prepared as described (23) was incubated with egg extract and histidine-tagged p60 xNAP-1 (His-xNAP-1) solutions at a final concentration of 5,000兾␮l. After 30-min incubation, sperm chromatin was sedimented onto a sucrose cushion by centrifugation and examined for linker histone binding. SDS兾PAGE of acidextracted proteins and micrococcal nuclease digestion assay were performed as described (23). Immunodepletion and Immunoblot Analysis. For immunodepletion

of B4 and the p60 xNAP-1, rabbit Abs for B4 and p60 were conjugated to Protein G-Sepharose beads (Sigma–Aldrich) by incubating the beads with an equal volume of each antiserum for 1 h at 4°C, and crude extracts were incubated twice with 50% extract volume of each antibody beads for 30 min each on ice, with occasional agitation. For mock depletion, Protein G beads conjugated with rabbit IgG were used. For immunoblot analysis, extracts were mixed with SDS sample buffer (10% glycerol兾2% SDS兾5% 2-mercaptoethanol兾0.0025% bromophenol blue兾60 mM Tris䡠HCl, pH 6.8), boiled for 2 min, run on SDS polyacrylamide gels, and transferred to nitrocellulose membranes. After blocking with 10% skimmed milk, membranes were incubated with primary antibodies for 2 h at room temperature or for 12 h at 4°C. After incubation with alkaline phosphatase-conjugated secondary antibodies, membranes were processed for visualizing signals by the BCIP兾NBL phosphatase substrate system (Kirkegaard & Perry Laboratories). Incorporation of Linker Histones into Dinucleosomes. Radiorabeled dinucleosomes were reconstituted and purified as described (24). 5⬘-end-radiolabeled 5S rDNA was mixed with purified core histones from HeLa cells to reconstitute chromatin by salt Shintomi et al.

dialysis. The reconstituted mixtures were separated on a sucrose gradient depending on the number of histone octamers bound to the dinucleosome DNA fragments (24). Dinucleosomes (5 nM) were incubated with various concentrations of histidine-tagged B4 (His-B4), with or without being complexed with His-xNAP-1, in 5 ␮l of binding buffer (50 mM KCl兾0.4 mM EDTA兾0.7 mM MgCl2兾10 mM Hepes䡠KOH, pH 7.4) at room temperature for 30 min. Samples were loaded onto a 0.7% agarose 0.5 ⫻ TBE gel. After electrophoresis, the gel was dried and autoradiographed. In some experiments, recombinant B4 from which the tags for purification have been removed was prepared as described (25), and dinucleosomes were incubated with B4 or His-xNAP-1兾B4 complexes in buffer (50 mM KCl兾0.57 mM EDTA兾0.5% glycerol兾1.0 mg兾ml BSA兾10 mM Tris䡠HCl, pH 8.0) at room temperature for 30 min. Each dinucleosome sample (55 fmol) was digested with 0.075, 0.15, 0.30, or 0.60 units of nucleococcal nuclease and analyzed as described (26). Results and Discussion NAP-1 Is Associated with the Linker Histone B4 in Xenopus Eggs. In

Xenopus eggs, the core histones H2A-H2B and H3-H4 stored in the cytoplasm are associated with histone chaperones nucleoplasmin and N1兾N2, respectively (11, 12). To investigate whether B4, the linker histone of Xenopus eggs (13–15), is found complexed to other proteins, egg extract was fractionated by gel filtration (Superose 12), and fractions were probed with Abs to B4. The result showed that B4 contained in egg extract was eluted considerably faster from the column than purified B4 in physiological buffer, whereas it was eluted at the same position as purified B4 in high-salt buffer containing 0.5 M KCl (Fig. 1A). In addition, native PAGE兾immunoblot analysis showed that B4 of egg extract migrated into the separating gel, whereas purified B4 did not because of its strong positive charge (data not shown). These results suggested that, in egg extract, B4 is associated with acidic components that neutralize its positive charge. Beads coupled to anti-B4 Ab (23) were used to immunoprecipitate gel-filtration fractions that contained B4, then coprecipitated components were eluted from the beads by 0.5 M KCl buffer. After desalting, the eluate was applied to an anion exchange column (Mono Q) and eluted by a liner gradient of 0.1–0.6 M KCl. Separation by SDS兾PAGE revealed that the major acidic components were two proteins of 56 and 60 kDa (Fig. 1B). To identify the p56 and p60 proteins, both were electrophoretically purified and subjected to peptide mapping and amino acid sequence analysis. The peptide mapping of p56 and p60 revealed that the two proteins were almost identical to each other (data not shown), suggesting they were isoforms of a single entity. On the basis of the determined partial amino acid sequences, we obtained two independent cDNAs encoding the p56-p60 protein. DNA sequencing of the cDNAs revealed that both encoded Xenopus homologues of NAP-1 (xNAP-1) (Fig. 1C) (27). Although the ORFs of the two xNAP-1 cDNAs encode predicted polypeptides of 44,387 and 45,514 Da, their reticulocyte translated proteins exhibited mobility at 56 and 60 kDa on SDS兾PAGE, respectively (data not shown). Examination of the primary amino acid sequences of xNAP-1 revealed that the p56 and p60 xNAP-1 differ only at their C termini, with an additional 11 amino acids of p60 replacing the terminal valine residue of p56 (Fig. 1C). The two isoforms are most likely produced by alternative splicing, because quite similar splice variants are known for human NAP-2, a closely related protein of NAP-1, which binds linker histones as well as core histones in vitro (28). The deduced amino acid sequences also indicate that the identity between p60 xNAP-1 and human NAP-1 (hNAP-1) is as high as 88%; the identity between p60 xNAP-1 and hNAP-2 is 52%. Similar to NAP-1 from other organisms, xNAP-1 contains highly acidic segments, which are thought to interact with histones, as PNAS 兩 June 7, 2005 兩 vol. 102 兩 no. 23 兩 8211

CELL BIOLOGY

(22), by using buffer (2 mM MgCl2兾20 mM Hepes䡠KOH, pH 7.4) containing various concentrations of KCl. To electrophoretically purify the p56 and p60 proteins, the Mono Q fractions containing both proteins were treated with SDS兾PAGE sample buffer, electrophoresed on a 12.5% polyacrylamide gel, and transferred to nitrocellulose membranes. After staining with Coomassie blue, the protein bands were excised. Peptide mapping and amino acid analysis of the proteins were carried out by using a commercial service (Advanced Protein Research Organization, Tokushima, Japan). Under our experimental conditions, ⬇0.6 ␮g of each of the p56 and p60 proteins was electrophoretically purified from 1 ml of egg extract.

Fig. 1. Identification of the B4-binding acidic protein in Xenopus eggs. (A) The B4 protein purified from Xenopus eggs and the soluble fraction of Xenopus egg extract were gel-filtrated on a Superose 12 column in physiological buffer (0.1 M) or high-salt buffer (0.5 M), and fractions were immunoblotted with Ab specific for B4. The position of elution for marker proteins is indicated at the top. (B) The gel-filtration fraction (0.1 M) containing B4 was treated with anti-B4 Ab beads, then B4-binding components were eluted from the beads by high-salt buffer, loaded onto a Mono Q column, and eluted by a linear gradient of 0.1– 0.6 M KCl. (C) Alignment of the predicted peptide sequences of the p60 and p56 B4-binding proteins and human NAP-1 (hNAP-1) with identical residues shaded. The identity between p60 and hNAP-1 is 88%. Acidic segments are underlined, and putative nuclear localization and nuclear export signals are indicated by shaded and open bars, respectively.

well as putative nuclear localization and nuclear export motifs (Fig. 1C). Involvement of NAP-1 in Nucleosome Assembly of Sperm Chromatin in Egg Extracts. The association of xNAP-1 with B4 in egg extract

suggests that xNAP-1 is involved in the regulation of B4 binding to linker DNA in Xenopus eggs, functioning as a chaperone. In addition, xNAP-1 may have a role in the postfertilization remodeling of nonnucleosomal sperm chromatin into nucleosomes, because another chaperone nucleoplasmin also plays an essential role in the removal of sperm-specific basic proteins from sperm chromatin, the first step of the remodeling (14, 29). Nucloplasmin is possibly involved in the regulation of linker histone binding, because it can remove linker histones from somatic nuclei incubated in egg extracts (30). To determine whether xNAP-1 is involved in nucleosome assembly in egg extract, and particularly assembly of sperm nuclei, we examined the remodeling of sperm chromatin in the absence of xNAP-1. To deplete xNAP-1 from egg extract, we prepared Abs against recombinant His-xNAP-1 (Fig. 2A, lane 1). The Abs specifically reacted to p56 and p60 of egg extract on immunoblot and precipitated both proteins from egg extract (Fig. 2 B and C). B4 was also coimmunoprecipitated (Fig. 2C). To confirm the specificity of B4 binding to xNAP-1, we examined whether the 8212 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0500822102

anti-B4 Abs coprecipitate other acidic proteins, particularly nucleoplasmin and the SET protein, which are homologous to NAP-1 in function and structure, respectively (3, 31, 32). The result clearly showed that xNAP-1 but neither nucleoplasmin nor the SET protein, was contained in the precipitate, demonstrating that B4 selectively binds to xNAP-1 (Fig. 2D). For reconstitution experiments, we prepared recombinant His-B4 (Fig. 2 A, lane 2) and mixed it with His-xNAP-1 in vitro to form complexes. On native PAGE analysis, most of the His-xNAP-1 migrated slower than the 140-kDa marker (lactase dehydrogenase), suggesting that His-xNAP-1 forms oligomers under physiological conditions, because the minor band migrating slightly slower than the 67-kDa marker (albumin) is thought to represent its monomeric form (Fig. 2E, lane 1). His-B4 itself did not migrate into the gel because of its strong positive charge (Fig. 2E, lane 5). However, when His-B4 was mixed with His-xNAP1, the mobility of HisxNAP-1 was decreased in proportion to the increase of the ratio of added His-B4. When His-xNAP-1 was mixed with His-B4 at a molar ratio of 2:1, it exhibited relatively constant mobility at 232 kDa (catalase) (Fig. 2E, lane 3), suggesting that xNAP-1 and B4 form a complex by stoichiometric binding. The in vitro binding of His-B4 to truncated versions of His-xNAP-1 revealed that the C-terminal acidic segment of His-xNAP-1 is required for complex formation with B4 (Fig. 5, which is published as Shintomi et al.

supporting information on the PNAS web site), whereas this segment is dispensable for core histone binding (33). To induce the nucleosome assembly of sperm chromatin in the absence of xNAP-1, egg extract was immunodepleted of both xNAP-1 and B4 and reconstituted with either free His-B4 or His-B4 in complex with His-xNAP-1 to the original concentration of endogenous B4 (Fig. 2F). Demembranated Xenopus sperm were incubated in the extracts for 60 min, then isolated for examination of His-B4 binding to the linker DNA by both SDS兾PAGE of acid-extracted nucleosomal proteins and micrococcal nuclease digestion. PAGE analysis showed that, in extract either with or without His-xNAP-1, sperm chromatin proteins were exchanged with nucleosomal histones, including His-B4, with H2A replaced by H2AX (12), an H2A variant comigrating with H3 (Fig. 2G; see also figure 2 in ref. 23). However, quantitative analysis revealed that, in the absence of His-xNAP-1, a considerably greater amount of His-B4 (⬇1.4-fold) was bound to chromatin. It should be noted that, consistent with the function of NAP-1 as a histone chaperone, His-xNAP-1, which is acid-soluble, was not recovered in the chromatin proteins. Furthermore, the micrococcal nuclease digestion assay revealed a discernible difference in the structure between nucleosomes formed in the presence and absence of xNAP-1. As shown in previous studies (23, 34), extensive nuclease digestion of B4-free nucleosomes, i.e., nucleosome cores, produced a DNA Shintomi et al.

fragment of 146 bp (Fig. 2H, lane 2), which was protected by the core histones alone, whereas nuclease digestion of nucleosomes containing B4 produced a DNA fragment of ⬇170 bp (Fig. 2H, lane 1), which represents DNA protected by the core histones plus the linker histone. When nucleosomes formed in the extract reconstituted with the His-xNAP-1兾His-B4 complex were extensively digested with micrococcal nuclease, the major product was a DNA fragment of ⬇170 bp (Fig. 2H, lane 3), the same size as the digestion product of nucleosomes formed in intact egg extract. This result indicates that, in the presence of xNAP-1, His-B4 binds to the linker DNA in the same manner as endogenous B4. In contrast, when nucleosomes formed in extract reconstituted with free His-B4 were similarly analyzed, DNA fragments generated were more heterogeneous in size and longer than 170 bp, as seen in the tailing of the band (Fig. 2H, lane 4). This tailing was not due to incomplete nuclease digestion, because more extensive digestion made the band less prominent without changing the tailing pattern (data not shown) but reflected size variations of the protected DNA because of nonspecific higher-amount association of B4 to chromatin. Thus, in the absence of xNAP-1, B4 binding to nucleosomes is irregular and兾or in excess. Moreover, the defect in nucleosome assembly caused by the removal of xNAP-1 might be underrepresented, because in the crude egg extract, some histone chaperone-like components could partly substitute for xNAP-1 in the deposition of the linker histone onto DNA. PNAS 兩 June 7, 2005 兩 vol. 102 兩 no. 23 兩 8213

CELL BIOLOGY

Fig. 2. The remodeling of Xenopus sperm chromatin into nucleosomes in egg extract with and without xNAP-1. (A) Histidine-tagged recombinant proteins of p60 xNAP-1 (His-xNAP-1, lane 1) and B4 (His-B4, lane 2) were produced in E. coli and purified. (B) On immunoblotting of egg extract, Abs raised against His-xNAP-1 detected specifically p56 and p60 xNAP-1 (lane 1), and these bands were also recognized by Ab for yeast NAP-1 (lane 2). (C) Anti-His-xNAP-1 Abs precipitated p56 and p60 xNAP-1 from egg extract, along with B4 (lane 3). (D) Anti-B4 Abs coprecipitated xNAP-1 but neither nucleoplasmin (Npl) nor SET from egg extract, along with B4 (lane 3). Distinct bands of nucleoplasmin are due to the difference in the phosphorylation state. (E) His-xNAP-1 and His-B4 were mixed in physiological buffer at various ratios and electrophoresed on a native-polyacrylamide gel. (F) Egg extract was mock-depleted with control Ab beads (lane 1) or immunodepleted of both xNAP-1 and B4 (lanes 2– 4) with anti-xNAP-1 and B4 beads, then reconstituted with free His-B4 (lane 3) or the His-B4兾His-xNAP-1 complex (lane 4). Npl (nucleoplasmin) is used as a dilution control. (G) SDS兾PAGE analysis of acid extracted proteins from sperm chromatin incubated in egg extract where the xNAP-1兾B4 complex has been depleted (lane 1) or replaced by the His-xNAP-1兾His-B4 complex (lane 2) or His-B4 (lane 3). (H) Micrococcal nuclease digestion assay of sperm chromatin incubated in the egg extracts where the xNAP-1兾B4 complex has been mock-depleted (lane 1), immunodepleted (lane 2), or replaced by the His-xNAP-1兾His-B4 complex (lane 3) or His-B4 (lane 4).

Fig. 3. The remodeling of Xenopus sperm chromatin in the in vitro reconstitution system. (A and B) Demembranated Xenopus sperm (lane 1) were incubated with the nucleoplasmin兾H2A兾H2B complex (lane 2) or this complex plus either the His-xNAP-1兾His-B4 complex (lane 3) or His-B4 (lane 4). Protein composition and basic structure were examined by SDS兾PAGE of acid extracted proteins (A) and micrococcal nuclease digestion (B), respectively.

Requirement of NAP-1 for Sperm Chromatin Remodeling into Nucleosomes in the Reconstitution System. To clarify the role of xNAP-1

in nucleosome assembly from sperm chromatin, we used a reconstitution system for the remodeling of sperm chromatin in vitro. It has been established that the chromatin of Xenopus sperm, which contains histones H3 and H4 along with spermspecific basic proteins (Fig. 3A, lane 1) can be remodeled into nucleosome cores by incubation solely with nucleoplasmin in complex with histones H2A and H2B (29, 35) (see also Fig. 3A, lane 2). We supplemented the in vitro remodeling system with free His-B4 or the His-xNAP1兾His-B4 complex and examined the remodeling of sperm chromatin in this supplemented reconstituted system. When sperm chromatin was incubated with the nucleoplasmin兾H2A兾H2B along with His-xNAP-1兾His-B4 complexes, sperm chromatin proteins were remodeled to nucleosomal histones including His-B4 but not His-xNAP-1 (Fig. 3A, lane 3). Micrococcal nuclease digestion of the nucleosomes produced a 170-bp DNA fragment (Fig. 3B, lane 3), indicating that His-B4 was appropriately deposited onto the linker DNA. In contrast, when sperm chromatin was incubated with the nucleoplasmin兾 H2A-H2B complex along with free His-B4, considerable amounts of sperm-specific basic proteins remained on the chromatin (Fig. 3A, lane 4). The result of micrococcal nuclease digestion assay was consistent with a failure in sperm chromatin remodeling to nucleosomes, because the 170-bp DNA fragment was not a prominent component of digestion products (Fig. 3B, lane 4). Thus, in the in vitro reconstitution system, the formation of a complex between B4 and xNAP-1 is an essential requirement for successful remodeling of sperm chromatin into nucleosomes containing the appropriate deposition of linker histones. Facilitation by NAP-1 of Proper Deposition of B4 onto Dinucleosomes.

To investigate directly a role of xNAP-1 in the deposition of B4 onto the nucleosomal linker DNA, we used synthetic 5S dinucleosomes that contain two nucleosome cores spaced by intact linker DNA; the physiological binding of one or two molecules of linker histones to dinucleosomes results in a slight decrease in mobility on nucleoprotein agarose gel electrophoresis and can be distinguished from nonspecific histone binding (26). Dinucleosomes, whose DNA had been radiolabeled, were incubated with 8214 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0500822102

Fig. 4. Facilitation by xNAP-1 of the physiological deposition of B4 onto dinucleosomes. (A) Various concentrations of free B4 (lanes 1– 6), the HisxNAP-1兾His-B4 complex (lanes 7–12), was added to dinucleosomes, and the binding of the linker histones was examined by the mobility shift of dinucleosomes on agarose gel electrophoresis. The positions of free DNA, dinucleosomes, and dinucleosomes incorporating one (1) and two (2) molecules of the linker histone are indicated. (B) Micrococcal nuclease (MNase) digestion of B4-incorporated dinucleosomes. Dinucleosomes, which had incorporated none (Dinucleosome) or two molar equivalents of B4 in the absence (Di ⫹ B4) or presence of His-xNAP-1 (Di ⫹ B4兾NAP-1) were digested with increasing amounts of micrococcal nuclease. The positions of DNA fragments of dinucleosomes (Di), chromatosome (CH), and nucleosome core (NC) are indicated. M, MspI-deigested pBR322 size marker.

various concentrations of free His-B4 or the His-xNAP-1兾 His-B4 complex at physiological ionic strength, and incorporation of B4 into these dinucleosomes was analyzed by gel electrophoresis. When low concentrations of free His-B4 (⬍50 nM) was added to dinucleosomes, their mobility was decreased to an extent that indicated the physiological binding of linker histones (Fig. 4A, lanes 2 and 3). However, by addition of higher concentrations of free His-B4 (⬎100 nM), the mobility of dinucleosomes was further reduced and, in addition, a tail of slower-migrating forms was evident, including forms unable to migrate from the slots. This indicated nonspecific aggregation of dinucleosomes, and the amount of aggregates increased in proportion to the concentration of His-B4 added (Fig. 4A, lanes 4–6). In contrast, when His-B4 in complex with His-xNAP-1 was Shintomi et al.

added to dinucleosomes, the physiological linker histone binding was induced (Fig. 4A, lanes 8–12), and nonspecific aggregates were not formed even with the highest concentration of His-B4 examined (Fig. 4A, lane 12). His-xNAP-1 alone did not affect the mobility of dinucleosomes (Fig. 4A, lane 13). Notably, dinucleosomes mixed with 400 nM His-B4 in complex with His-xNAP-1 (Fig. 4A, lane 12) exhibited a similar extent of mobility shift as those mixed with 200 nM free His-B4 (Fig. 4A, lane 5). Nevertheless, no aggregate was formed in the former, whereas considerable aggregation occurred in the latter. Moreover, the band of the former is sharper than that of the latter, suggesting more stable B4 binding to dinucleosomes in the presence of HisxNAP-1. The facilitation by xNAP-1 of stable B4 binding to dinucleosomes was confirmed by micrococcal nuclease digestion of reconstitutes. When dinucleosomes incubated with B4兾 xNAP-1 complexes were treated with micrococcal nuclease, a 170-bp DNA fragment indicative of proper linker histone binding was more stably produced (Fig. 4B). These results demonstrate that xNAP-1 restrains excess and兾or irregular binding of B4 to dinucleosomes, which causes nonspecific aggregation but does not affect physiological B4 binding. Thus, xNAP-1 functions as a histone chaperone that facilitates regulated and normal deposition of B4 onto the linker DNA.

some assembly is to neutralize the strong positive charge of histones by complex formation, thus preventing DNA aggregation that would otherwise occur. We have shown in Xenopus egg extract that xNAP-1 forms a complex with B4 and ensures the proper deposition of B4 onto nucleosomes, playing the role of chaperone for the linker histones. We therefore conclude that xNAP-1 is a linker histone chaperone in Xenopus eggs. NAP-1 binding to linker histones may be specific to early embryos, because B4 is the linker histone subtype specific to early embryonic cells (13–15), and NAP-1 has been shown to be a chaperone for H2A and H2B in other cells (19, 20). However, recent findings that histone H1 binding to chromatin is dynamic in mammalian culture cells (36, 37) suggest a role for a similar linker histone chaperone to effect rapid exchange of histone H1 in somatic chromatin. Although we have confirmed that xNAP-1 can function as a chaperone for somatic H1 in its deposition onto dinucleosomes in vitro (25), the true identity of the somatic linker histone chaperone remains to be determined. Nevertheless, our results demonstrate that NAP-1 can be a useful tool for in vitro reconstitution of the linker histone-bound nucleosomes, which has been difficult because of nonspecific aggregation.

Conclusion Extensive studies on the assembly of nucleosome cores in vitro indicate that the primary role of histone chaperones in nucleo-

We thank A. Philpott for critical reading of the manuscript, K. Nagata for the SET protein Ab, and K. Tachibana and E. Okumura for helpful discussions. We also thank T. Horiguchi and Y. Kitamura for their help in some experiments. This work was supported by grants from the Japanese Ministry of Education, Culture, Sports, Science and Technology (to K.O. and T.K).

Van Holde, K. E. (1989) Chromatin (Springer, New York). Wolffe, A. (1998) Chromatin (Academic, London), 3rd Ed. Philpott, A., Krude, T. & Laskey, R. A. (2000) Semin. Cell Dev. Biol. 11, 7–14. Laskey, R. A., Honda, B. M., Mills, A. D. & Finch, J. T. (1978) Nature 275, 416–420. Tyler, J. K. (2002) Eur. J. Biochem. 269, 2268–2274. Haushalter, K. A. & Kadonaga, J. T. (2003) Nat. Rev. Mol. Cell. Biol. 4, 613–620. Loyola, A. & Almouzni, G. (2004) Biochim. Biophys. Acta 1677, 3–11. Travers, A. (1999) Trends Biochem. Sci. 24, 4–7. Bharath, M. M. S., Chandra, N. R. & Rao, M. R. S. (2003) Nucleic Acids Res. 31, 4264–4274. Luger, K., Mader, A. W., Richmond, R. K., Sargent, D. F. & Richmond, T. J. (1997) Nature 389, 351–360. Dilworth, S. M., Black, S. J. & Laskey, R. A. (1987) Cell 51, 1009–1018. Kleinschmidt, J. A., Seiter, A. & Zentgraf, H. (1990) EMBO J. 9, 1309–1318. Smith, R. C., Dworkin-Rastl, E. & Dworkin, M. B. (1988) Genes Dev. 2, 1284–1295. Ohsumi, K. & Katagiri, C. (1991) Dev. Biol. 147, 110–120. Dworkin-Rastl, E., Kandolf, H. & Smith, R. C. (1994) Dev. Biol. 161, 425–439. Ishimi, Y., Hirosumi, J., Sato, W., Sugawara, K., Yokota, S., Hanaoka, S & Yamada, M. (1984) Eur. J. Biochem. 142, 431–439. Ishimi, Y. & Kikuchi, A. (1991) J. Biol. Chem. 266, 7025–7029. Yoon, H. W., Kim, M. C., Lee, S. Y., Hwang, I., Bahk, J. D., Hong. J. C., Ishimi, Y. & Cho, M. G. (1995) Mol. Gen. Genet. 249, 465–473. Ito, T., Bulger, M., Kobayashi, R. & Kadonaga, J. T. (1996) Mol. Cell. Biol. 16, 3112–3124. Chang, L., Loranger, S. S., Mizzen, C., Ernst, S. G., Allis, C. D. & Annunziato, A. T. (1997) Biochemistry 36, 469–480.

21. Iwabuchi, M., Ohsumi, K., Yamamoto, T. M., Sawada, W. & Kishimoto, T. (2000) EMBO J. 19, 4513–4523. 22. Ohsumi, K. & Katagiri, C. (1991) Dev. Biol. 148, 295–305. 23. Ohsumi, K., Katagiri, C. & Kishimoto, T. (1993) Science 262, 2033–2035. 24. Ura, K., Araki, M., Saeki, H., Masutani, C., Ito, T., Iwai, S., Mizukoshi, T., Kaneda, Y. & Hanaoka, F. (2001) EMBO J. 20, 2004–2014. 25. Saeki, H., Ohsumi, K., Aihara, H., Ito, T., Hirose, S., Ura, K. & Kaneda, Y. (2005) Proc. Natl. Acad. Sci. USA 102, 5697–5702. 26. Ura, K., Nightingale, K. & Wolffe, A. P. (1996) EMBO J. 15, 4959–4969. 27. Steer, W. M., Abu-Daya, A., Brickwood, S. J., Mumford, K. L., Jordanaires, N., Mitchell, J., Robinson, C., Thorne, A. W. & Guille, M. J. (2003) Mech. Dev. 120, 1045–1057. 28. Rodriguez, P., Munroe, D., Prawitt, D., Chu, L. L., Bric, E., Kim, J., Reid, L. H., Davies, C., Nakagama, H., Leobbert, R., et al. (1997) Genomics 44, 253–265. 29. Philpott, A. & Leno, G. H. (1992) Cell 69, 759–767. 30. Dimitrov, S. & Wolffe, A. (1996) EMBO J. 15, 5897–5906. 31. Von Lindern, M., Vas Baal, S., Wiegant, J., Raap, A., Hagemeijeer, A. & Grosveld, G. (1992) Mol. Cell. Biol. 12, 3346–3335. 32. Kellogg, D. R., Kikuchi, A., Fujita-Nakata, T., Turck, C. W. & Murray, A. W. (1995) J. Cell Biol. 130, 661–673. 33. Fujii-Nakata, T., Ishimi, Y., Okuda, A. & Kikuchi, A. (1992) J. Biol. Chem. 29, 20980–20986. 34. Dimitrov, S., Dasso, M. & Wolffe, A. P. (1994) J. Cell Biol. 126, 591–601. 35. Katagiri, C. & Ohsumi, K. (1994) Int. J. Dev. Biol. 38, 209–216. 36. Lever, M. A., Th’ng, J. P. H., Sun, X. & Hendzel, M. J. (2000) Nature 408, 873–876. 37. Misteli, T., Gunjan, A., Hock. R., Bustin, M. & Brown, D. T. (2000) Nature 408, 877–881.

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

CELL BIOLOGY

20.

Shintomi et al.

PNAS 兩 June 7, 2005 兩 vol. 102 兩 no. 23 兩 8215