A chromatin-associated protein from pea seeds preferentially binds histones H3 and H4 Josefa Castillo, A ´ ngel Zu ´ n ˜ iga*, Luis Franco and M. Isabel Rodrigo Department of Biochemistry and Molecular Biology, University of Valencia, Spain Pisum sativum p16 is a protein present in the chromatin of ungerminated embryonic axes. The purification of p16 and the isolation of a cDNA clone of psp54, the gene encoding its precursor have been recently reported [Castillo, J., Rodrigo, M. I., Ma ´ rquez, J. A., Zu´ n ˜ iga, A and Franco, L. (2000) Eur. J. Biochem. 267, 2156–2165]. In the present paper, we present data showing that p16 is a nuclear protein. First, after subcellular fractionation, p16 was clearly found in nuclei, although the protein is also present in other organelles. Immunocytochemical methods also confirm the above results. p16 seems to be firmly anchored to chromatin, as only extensive DNase I digestion of nuclei allows its release. Far Western and pull-down experiments demonstrate a strong in vitro interaction between p16 and histones, especi- ally H3 and H4, suggesting that p16 is tethered to chromatin through histones. Because the psp54 gene is specifically expressed during the late development of seed, the role of p16 might be related to the changes that occur in chromatin during the processes of seed maturation and germination. Keywords: chromatin; histones; nuclear proteins; histone acetylation; histone-binding proteins. The highly conserved nucleosome core particle is formed by 146 bp of DNA wrapped around a histone octamer. The structure of the histone octamer was resolved at 3.1 A ˚ resolution almost 10 years ago by the group of Moudri- anakis, who described it as a wedge-shaped, tripartite structure, formed by a tetramer of two copies each of histones H3 and H4 and two flanking H2A-H2B dimers [1]. An unexpected finding was the discovery of the histone fold, a common motif of tertiary structure, which generates the heterodimeric pair-wise association of histones via the handshake motif [1]. Six years later, the group of Richmond described the structure of reconstituted nucleosome cores at 2.8 A ˚ resolution [2]. The latter work confirmed that the overall structure of the isolated histone octamer is conserved in the whole core particle, and added some details to the known structure of histones and showed the exact path of DNA around the histones. It was recognized early on that the nucleosome structure represents a serious obstacle to the different dynamic nuclear processes such as transcription and, obviously, higher order organization of chromatin adds further impediments to the transcriptional machinery; the pack- aging of DNA in eukaryotic chromatin results in a high concentration of the nucleic acid, which may be as high as 0.1 mgÆmL )1 in some interphase nuclei [3]. Nevertheless, chromatin cannot be just considered as a static structure. A dynamic remodelling of chromatin continuously occurs at many loci and histones play a definite role in these changes, as they are the targets for many protein factors. The early view of histones as mere structural proteins changed about 10 years ago to envisage them as gene expression regulators, a role that is played via specific interactions with other proteins. Since the early genetic data on the involvement of histone N-terminal tails in silencing via the interaction with specific proteins [4–8], several lines of evidence have shown that histone-binding proteins typically act as silencers, corepressors or coactiva- tors in a way often modulated by histone post-translational modifications (reviewed in [9,10]). Apart from these typical functions, histones may bind other proteins that play diverse roles, such as chaperones in chromatin assembly (reviewed in [11,12]) or remodelling [13,14]. In many cases, however, the functional role of histone-binding proteins remains still unknown [15,16]. As mentioned above, the histone N-terminal tails are involved in protein binding. They are accessible both in the nucleosome [2] and in chromatin [17] and they are the site of post-translational modifications that can modulate protein binding [10]. Apart from the genetic evidence referred to above, biochemical data substantiated the actual existence of proteins able to bind histones via the N-terminal tails in an acetylation-dependent manner [18–20]. Nevertheless, non- histone proteins may bind histone domains other than the amino termini. The histone fold is not restricted to core histones and it may be involved in the dimerization of several proteins [21]. The use of novel methods of search for protein motifs [22,23] has allowed Sullivan et al.toexpandthe number of known proteins having the histone-fold motif [24]. Several of the proteins included in the generated database (http://genome.nhgri.nih.gov/histones/) may potentially bind histones through the histone fold. For instance, the centromeric protein CENP-A and macro-H2A, a rat nonhistone, H2A-related protein, have been proposed to Correspondence to L. Franco, Department of Biochemistry and Molecular Biology, University of Valencia, E-46100 Burjassot, Valencia, Spain. Fax: + 34 96 4635, Tel.: + 34 96 3864385, E-mail: luis.franco@uv.es Abbreviations: AUT, acetic acid/urea/Triton X-100; GST, glutathione S-transferase. *Present address: Hospital de La Ribera, E-46600 Alzira, Valencia, Spain. (Received 3 June 2002, revised 23 July 2002, accepted 1 August 2002) Eur. J. Biochem. 269, 4641–4648 (2002) Ó FEBS 2002 doi:10.1046/j.1432-1033.2002.03164.x substitute for H3 and H2A, respectively, in some nucleo- somes [25]. In most cases, however, the histone fold of nonhistone proteins is involved in the dimerization-depend- ent acquisition of a tertiary structure suitable for further interacting either with DNA or with other proteins. For instance, it has been recently found that CHRAC-14 and CHRAC-16, two components of the Drosophila chromatin- remodelling complex CHRAC, are able to form heterodi- mers via their histone-fold domains. These dimers bind ISWI, the ATPase of the complex, under conditions of increased stringency where CHRAC-14 and CHRAC-16 alone are unable to interact with ISWI [26]. A similar situation seems to occur in HuCHRAC, the human homo- logue of the Drosophila chromatin-remodelling complex [27]. We have recently reported the purification of a protein, p16, abundant in the chromatin of ungerminated pea embryonic axes, and the isolation of a cDNA clone of its gene, psp54. The gene, which is expressed only during seed maturation or in adult tissues undergoing hydric stress, encodes a large polypeptide that is processed to yield p16. This protein seems to be associated to chromatin as it can be obtained together with octamer histones from formaldehyde cross-linked chromatin. Moreover, p16 is partially recov- ered from nuclei as heterodimers with H3 when the disulfide bridges occurring in vivo are preserved [28]. This means that at least some of the p16 molecules are close enough to H3 to allow them to interact in vivo with the histone, In the present paper, we report further experiments showing that p16 is a bona fide nuclear protein that interacts with histones, especially with H3 and H4. MATERIALS AND METHODS Materials Pea (P. sativum, cv. Lincoln) seeds were purchased locally. To obtain ungerminated embryo axes, seeds were imbibed in cold water as described previously [29] and the embryonic axes were excised from the cotyledons. Chicken erythrocyte histones were prepared as previously described [30] and they were kindly provided by E. Ballestar, CNIO, Madrid, Spain. Pea histones were prepared as reported elsewhere [31]. Chicken erythrocyte histones were acetylated in vitro with [ 14 C]acetyl-CoA in the presence of yeast recombinant Esa1p. AUT/PAGE (acetic acid/urea/Triton X-100/PAGE) analysis showed that H3 and H4 are almost fully acetylated, while H2A contains a mixture of acetylated and nonacet- ylated isoforms and H2B was not acetylated at all. The acetylation and AUT/PAGE analysis of histones was carried out by G. Lo ´ pez-Rodas, Dept. of Biochemistry and Molecular Biology, University of Valencia, Spain. Subcellular fractionation of pea embryonic axes and DNase I digestion of nuclei Pea embryonic axes were homogenized in buffer A [300 m M sucrose, 8 m M CaCl 2 ,8m M MgCl 2 ,10m M 2-mercapto- ethanol, 50 m M NaHSO 3 ,0.7m M phenylmethanesulfonyl fluoride, 20% (v/v) glycerol and 10 m M Mops pH 6.0] and nuclei were prepared from the extracts and purified by centrifugation through Percoll [29]. The supernatant obtained after sedimenting nuclei was further centrifuged at 10 000 g for 10 min, and the 10 000 g supernatant was centrifuged at 100 000 g for 1 h. To investigate the presence of p16 in the different subcellular fractions, the purified nuclear fraction and both the 10 000 g and 100 000 g sediments were acid-extracted with 0.25 M HCl and the soluble proteins were recovered [31]. The proteins present in the 100 000 g supernatant were recovered by precipitating with cold trichloroacetic acid (final concentration 25%, v/v). For DNase I digestion, Percoll-purified nuclei were washed twice by suspending and sedimenting (1100 g, 10 min) in digestion buffer (10 m M NaCl, 1 m M MgCl 2 , 5m M 2-mercaptoethanol, 0.1 m M phenylmethanesulfonyl fluoride, 0.25 M sucrose, 10 m M Tris/HCl, pH 7.4) and they were finally suspended in this buffer to give an attenuance at 260 nm of about 20 units per mL. DNase I digestion was then carried out as described elsewhere [31]. Electrophoresis of proteins and Western blots were carried out as described previously [28]. Electron microscopy Percoll-purified nuclei were fixed, infiltrated with Lowicryl K4M and polymerized in gelatin capsules. Thin sections (94 nm) were processed for colloidal gold cytochemistry as described previously [32]. The samples were treated with either p16 antiserum (diluted 1 : 5000), or with preimmune serum as a control, in immunostaining buffer (0.23 M NaCl, 0.1% bovine serum albumin, 20 m M Tris/HCl, pH 7.4), containing 1% fetal bovine serum. A goat anti-(rabbit IgG) Ig gold-conjugate (10 nm, Sigma) 10-fold diluted in immu- nostaining buffer containing 0.05% Tween 20 and 5% fetal bovine serum was used. After rinsing with immunostaining buffer, sections were further rinsed in distilled water and stained with uranyl acetate. The preparations were exam- ined and photographed in a transmission electron micro- scope Philips model CM-10. Pull-down assays To prepare the GST-p16 fusion protein, the cDNA encoding p16 was obtained from the psp54 (28) cDNA. The oligonucleotides used as primers were: 5¢-CCCCTCGA GATGTCTAGACAAAAAAAGAGTAG-3¢ and 5¢-CCC CTCGAGTCACACAACAGCACGAC-3¢.ThePCR product was excised at the XhoI site present in the primer termini and the resulting DNA was purified and cloned in phase at the XhoI site of plasmid pGEX4T-1 (Pharmacia Biotech). Escherichia coli BL21(DE3)pLysS cells (Invitro- gen) were transformed with the recombinant plasmid and with the vector alone. Proteins were expressed and immo- bilized on glutathione-Sepharose beads (Pharmacia Bio- tech), following the manufacturer’s indications. After isopropyl thio-b- D -galactoside induction, the cells were lysed and 600 lL aliquots of the soluble fraction were incubated with 20 lL of glutathione-Sepharose beads for 1 h at room temperature. The beads were then exhaustively rinsed with 140 m M NaCl, 2.7 m M KCl, 1.5 m M KH 2 PO 4 , 8.1 m M Na 2 HPO 4 , pH 7.3 and with buffer B [15 m M MgCl 2 , 150 m M NaCl, 15 m M EDTA, 10% glycerol, 0.3% Triton X-100, 0.02% NaN 3 ,1m M dithiothreitol, 0.2% protease inhibitor cocktail for bacterial cell extracts (Sigma), 25 m M Tris/HCl, pH 7.5]. Total core histones either from chicken erythrocytes or from pea (20 lg) or individual histone fractions (5 lg), 4642 J. Castillo et al. (Eur. J. Biochem. 269) Ó FEBS 2002 dissolved in 65 lL of buffer B were mixed with immobilized GSTorGST-p16.Themixturewasincubatedfor15minat room temperature in an orbital shaker. The beads were then sedimented (3000 g, 1 min) and the supernatant (unbound fraction) was saved. The beads were then successively washed with 65 lL of buffer B containing increasing amounts of NaCl (150, 250, 500, 700, 1000 and 2000 m M ). Proteins retained after the last washing were released by boiling the beads in gel sample buffer. Far Western blot analysis To obtain [ 35 S]methionine-labelled His 6 -p16, the cDNA encoding p16 was obtained as above and cloned into the XhoI site of plasmid pRSETA (Invitrogen). In vitro transcription-translation was carried out by using a TNT T7 Quick Coupled Transcription/Translation System (Pro- mega). Core histones from either chicken erythrocyte or pea were resolved by SDS/PAGE and blotted into nitrocellulose membranes in the presence of 192 m M glycine, 0.02% SDS, 25 m M Tris/HCl, pH 8.3. The membranes were incubated for 30 min at room temperature in binding buffer [75 m M KCl, 75 m M NaCl, 1 m M EDTA, 0.25 m M MgCl 2 ,0.5m M dithiothreitol, 0.05% Nonidet NP-40, 0.4% bovine serum albumin, 0.4% Ficoll 400, 0.4% poly(vinyl pyrrolidone), 20 m M Hepes, pH 7.7]. The membranes were afterwards incubated for 16 h at 4 °C in 2 mL of binding buffer containing the [ 35 S]methionine-labelled His 6 -p16 and they were washed three times (10 min each) with binding buffer. The binding of [ 35 S]methionine-labelled His 6 -p16 was monitored with a fluorescent image analyser FLA-3000 (Fujifilm). RESULTS P16 is present in nuclei The presence of p16 in pea embryonic axes nuclei has been checked by a variety of procedures, including immunologi- cal and biochemical methods. First, we performed a subcellular fractionation in which we prepared four frac- tions (see Materials and methods). Figure 1 shows the results of a representative experiment, where a protein with the mobility of p16 is clearly present in the nuclear fraction as well as in the 10 000 g pellet and, to a lesser extent, in the 100 000 g pellet, but it is absent from the soluble fraction (Fig. 1A). The Western blot of Fig. 1B allowed the unambiguous identification of that protein as p16. It has to be noted that the 1000 g pellet has been further purified by centrifugation through a Percoll gradient (see Materials and methods) and therefore it represents a bona fide nuclear fraction. Consequently, the high proportion of p16 in this fraction, where the most abundant proteins are obviously the histones, is an argument in favour of the actual presence of p16 in nuclei. The protein is also present in the 10 000 g fraction, which mainly contains mitochondria, protein bodies and other medium-sized organelles. The abundance of p16 in this fraction cannot be accounted for by the residual presence of nuclei or nuclear debris and it should be concluded that, apart from nuclei, p16 is also present in other organelles, but it is clear that p16 is not a cytosolic component (Fig. 1B). The partitioning of p16 between nuclei and other organelles will be further discussed below. Immunocytochemical analysis of Percoll-purified nuclei provides an additional proof as to the nuclear localization of p16. In the experiment of Fig. 2, a 1 : 5000 dilution of the antiserum was used to minimize any possible unspecific reaction. No detectable accumulation of gold grains was observed when using preimmune serum, so the presence of immunogold in nuclei treated with p16 antiserum reflects the presence of the protein. Most of the grains are visible in perinuclear regions. It is known that digestion of nuclear DNA by DNase I results in the release of chromatin-bound proteins. Histones, which are tightly bound to DNA, are usually released only after extensive DNA digestion. We have therefore used digestion of pea embryo nuclei to analyse the tightness of p16 binding. The results of a representative experiment are given in Fig. 3. Several proteins were released simply by washing the nuclei with the low-ionic strength digestion medium, and their presence in the supernatant increased by incubating nuclei in the absence of added nuclease (Fig. 3A, lanes T). Therefore, these proteins could hardly be consid- ered as chromatin components, and they are probably components of the nucleoplasm. Some other proteins become soluble only after DNA digestion. Apart from some histones, four major polypeptides with apparent M r of 45 000, 29 000, 21 000 and 16 000, appeared in the supernatant after more or less prolonged digestion (Fig. 3A). The first polypeptide to be released, with M r 29 000, corresponds to the high mobility group protein 1P, previously identified in our laboratory [33]. The other polypeptides began to be released from chromatin only after more prolonged digestion. The Western blot in Fig. 3B shows that the released protein with M r 16 000 is p16. It appears in the soluble fraction only after 45 min of Fig. 1. Presence of p16 in different subcellular fractions. A purified nuclear fraction (lanes 1 and 5), the sediments of 10 000 g (lanes 2 and 6) and 100 000 g (lanes 3 and 7) centrifugation and a soluble fraction (lanes 4 and 8) were prepared as described under Materials and methods, and their proteins resolved by SDS/PAGE. The gels were either stained with Coomassie Blue (A) or Western-blotted and probed with 1 : 500 diluted p16 antiserum (B). The migration of size markers is shown on the right side of (A). Ó FEBS 2002 H3 and H4-binding chromatin protein from pea seeds (Eur. J. Biochem. 269) 4643 digestion, when 12–15% of DNA has been rendered acid soluble, and its release from nuclei was not complete even after 60 min (compare lanes S and P in Fig. 3A). It seems clear that a large destabilization of the nucleo- some core occurred before p16 was released, as both H2A and H2B, in addition to the linker H1 histone, were detectable in the soluble fraction prior to the appearance of p16 in the supernatants. The identity of these histones was also checked by their distinctive mobility in AUT/PAGE (data not shown). This fact seems to indicate that p16 is tightly anchored to DNA, either directly bound or tethered through a histone. We have previously found that treatment of nuclei from pea embryonic axes with formaldehyde results in the cross-linking of p16 to core histones [28], which suggests that the latter possibility, i.e. the existence of interactions between p16 and histones, is the primary cause for the occurrence of tightly bound p16 in nuclei. P16 binds histones in a specific manner To explore the above possibility, we first analysed p16– histone interactions by Far Western blotting. The results from a typical experiment are given in Fig. 4, which shows that p16 binds chicken erythrocyte and pea histones in vitro. Moreover, there is a preferential interaction with H3 and H4 and, to a lesser extent, with H2B. No interaction with H2A is detected. The possibility of artifacts due to the denatur- ation-renaturation of the electrophoresed proteins in Far Western blotting limits the validity of the above data. To corroborate them, we carried out a pull-down assay with an immobilized fusion protein GST-p16. The experi- ments were performed with histones from chicken erythro- cytes and from pea. Figure 5 shows that GST alone is unable to retain histones as all the input material appears in the unbound fraction. In constrast, the fusion protein effectively retains chicken erythrocyte histones. Most of H2A and H2B is present in both the 500 and 700 m M NaCl eluates, but H3 and H4 seem to be retained to a larger extent (Fig. 5A). These results suggest a somewhat preferential binding of H3 and H4, in agreement with those of the Far Western assays (Fig. 4). From a qualitative point of view, the experiment with pea histones (Fig. 5C) gave similar Fig. 3. Release of p16 upon DNase I digestion. Nuclei from ungermi- nated pea embryonic axes were digested for the indicated time periods (in minutes) and sedimented. Proteins recovered from the supernatant (lanes S) after 25% (w/v) trichloroacetic acid precipitation and the acid-soluble proteins from the residual nuclear pellet (lanes P) were analysed by SDS/PAGE. (A) shows a Coomassie Blue stained gel. For control, nuclei were sham-digested without added DNase I (lanes T). The migration of histones, previously identified high mobility group proteins (1P and 2P) and p16 is indicated on the margin. The poly- peptides with M r 45000 and 21000 (see the text) are marked by dots on lane S-60. In the lane marked st, molecular size markers (M r 94 000, 67 000, 43 000, 30 000, 20 100 and 14 400 from top to bottom) were run. (B) shows a Western blot of the samples from the supernatant at 45 and 60 min revealed with the p16 antiserum. Fig. 2. Immunocytochemical detection of p16 in nuclei purified from pea embryonic axes. Percoll-purified nuclei were fixed, thin-sectioned, treated with preimmune serum (A) or 1 : 5000 diluted p16 antiserum (B) and with a goat anti-(rabbit IgG) Ig gold-conjugate as described under Materials and methods. Several areas with immunogold deposits were seen, especially in the perinuclear regions, in nuclei treated with p16 antiserum (B). One of these areas, marked B 1 ,is magnified and shown below. For comparison, another area from (A), marked A 1 , is also magnified to verify the absence of gold deposits. The bars in (A) and (B) represent 1 lm. 4644 J. Castillo et al. (Eur. J. Biochem. 269) Ó FEBS 2002 results, although the preferential interaction of p16 with H3 and H4 seems to be stronger than with chicken histones. To further explore this question, we carried out a similar experiment with purified, individual histones from chicken erythrocytes (Fig. 5B). The results show that chicken H2A and H2B are released from the immobilized fusion protein at comparatively low ionic strength (500 m M NaCl). In fact, no histones remain bound to the immobilized fusion protein after 2 M NaCl washing. In contrast, neither H3 nor H4 are substantially released with the saline washing (Fig. 5B) and most of them remained bound to the immobilized GST-p16. That the complex between p16 and either histone remains stable even in high salt indicates that the strong binding of p16 to H3 and H4 is not predominantly due to ionic interactions (note that p16 has an isoelectric point of about 10 [28]). In contrast to the Far Western experiments, interaction of p16 with H2A is detected in the pull-down assays. The already-mentioned variation in histone structure due to the denaturation-renaturation processes in Far Western blotting may account for these differences. Acetylation of the e-amino groups of lysyl residues in the N-terminal tails often modulates the interaction of histones with other protein factors [20]. To check whether this occurs in the p16-histone binding, we also carried out pull-down assays with chicken erythrocyte histones acetylated in vitro with recombinant yeast Esa1p. The yield of the acetylation reaction, as revealed by AUT/PAGE (not shown), was high, but some nonacetylated isoforms still remain. The pull- down experiments (Fig. 6) indicate that p16 displays a certain preference for nonacetylated histones. The compari- son of input and unbound lanes in Fig. 5A clearly shows that a large proportion of nonacetylated histones is retained on the immobilized fusion protein and most of the bound proteins, especially H2A and H2B, are released in the saline washing (predominantly at 500 and 700 m M NaCl). On the contrary, when acetylated histones were used (Fig. 6), the proportion of unbound histones is much higher and no proteins are detectable in the saline washing. These circum- stances are particularly clear in the autoradiogram (Fig. 6B) that, obviously, displays only the acetylated histones. In conclusion, the results given in this section reveal that p16 interacts with histone in vitro, with specificity towards H3 and H4, and that the nonacetylated isoforms are somewhat preferred. DISCUSSION Several of the results reported in this paper show that a proportion of p16 is localized in nuclei. Immunocytochem- ical evidence (Fig. 2) clearly shows that p16 is present in purified nuclei and the results of the subcellular fraction- ation (Fig. 1) point are consistent with this assertion. The latter experiments added an interesting finding, namely, that p16 is not uniquely located in the nucleus in pea embryonic axes but it is also present in other particulate fractions. When the cDNA of psp54, the gene encoding the p16 precursor, was sequenced, both nuclear localization signals Fig. 5. Pull-down assay to show the interaction in vitro between p16 and histones. A GST-p16 fusion protein was immobilized on glutathione- Sepharose beads. The gel was loaded with chicken erythrocyte core histones (A), purified chicken erythrocyte histone fractions (B), or pea core histones (C). The beads were washed successively with loading buffer to recover the nonretained fraction and with buffers of increasing salt concentration as indicated. After the last saline washing, the beads were recovered, boiled in SDS/PAGE loading buffer and run. These later results are shown only for the experiments in (B) and (C). The gels in (A) and (C) were stained with Coomassie Blue, while (B) shows a silver-stained gel. I, input fractions; U, unbound fractions. The concentration (m M ) of NaCl in the successive washing solutions is indicated by the numbers above the lanes. Fig. 4. Interactions between p16 and histones probed by Far Western blotting. Core histones from chicken erythrocytes (lanes c) and from pea (lanes p) were separated by SDS/PAGE. (A) shows a Coomassie- stained gel and (B) a similar gel blotted and probed with [ 35 S]methio- nine-labelled p16. The migration of size markers is indicated on the right. Ó FEBS 2002 H3 and H4-binding chromatin protein from pea seeds (Eur. J. Biochem. 269) 4645 and a leader peptide sequence, were detected [28]. This may account for the fact that p16 is partitioned between the nuclei and some other subcellular organelles. The temporal relationship between this differential targeting and the processing of p54, the precursor 54 kDa peptide, is not known, but there are putative bipartite nuclear localization signals both in p16 and in the N-terminal region of the precursor polypeptide [28], so it is theoretically possible that p16 enters the nucleus either in the form of a precursor or after maturation. We do not yet know the nature of the non- nuclear organelles containing p16. As p16 and p54 share homology with seed storage proteins, these extranuclear organelles may be protein bodies. Preliminary evidence (J. Castillo & M. I. Rodrigo, unpublished results) seems to support this hypothesis. The presence of p16 in nuclei is the result of a strong interaction with chromatin components. The data of Fig. 3 distinctly show that, in clear contrast with the behaviour of other nuclear proteins, p16 is released only after a somewhat extensive digestion of DNA. We have carried out digestions of pea embryo chromatin with micrococcal nuclease; the nucleosomes were separated in a nucleoprotein gel and a second dimension in SDS/PAGE was run to analyse the protein complement. In these experiments (M. I. Rodrigo, J. Castillo & L. Franco, unpublished results), p16 appears as a component of a subset of nucleosomes. These results support the idea that p16 is a component of chromatin. In a previous paper, we showed that most, if not all, of nuclear p16 can be recovered bound to octameric histones after formaldehyde cross-linking of chromatin. Moreover, part of p16 was found to be close enough to H3 to became bound to the histone through a disulfide bridge when reducing agents were avoided during p16 extraction [28]. This latter result does not provide evidence that p16 is recruited to chromatin via disulfide bonds but both results suggested that p16 lies in close vicinity of core particles and that it may result bound to chromatin through some interaction with histones. The data presented in this paper confirm this assumption and both the Far Western blotting experiments and the pull-down assays (Figs 4 and 5) show that p16 interacts in vitro with histones, particularly with H3 and H4. In this context it is noteworthy that the experiments of Fig. 3 show that the release of histones H2A and H2B is substantially easier than that of p16. This fact indicates that the presence of p16 in nuclei is due to a strong interaction that requires a large disorganization of chromatin to be broken down. When the preference of p16 to interact with H3 and H4, which occupy a central position in the histone octamer [1] and in the core particle [2], is considered, it is reasonable to think that p16 is tethered to chromatin in vivo due to its ability to bind H3-H4. The physical basis for the histone-binding ability of p16 is not known, although the results reported here suggest that electrostatic forces are not fundamental in this process. We have analysed the sequence of p16 in search of histone folds [22,23], but this motif seems not to be present in the protein. On the other hand, acetylation of histones influences to some extent the p16–histone interactions, so the N-terminal tails of histones are probably involved in binding, although the structured domains of the histones also might partici- pate. In the literature there are several examples of proteins that bind sequences of the histones corresponding to their structured domains and yet do not seem to possess the histone fold. Among them are Saccharomyces Spt6p [15], and the human proteins p46, a component of the histone acetyltransferase B complex, and the highly related p48, present in the chromatin assembly factor CAF-1 [34]. It seems obvious that the mechanisms of p16–histone interac- tion would deserve further analysis. Finally, we wish to discuss on the possible function of p16. It is likely, in view of our previous results [28], that the role of p16 is related to the hydric stress accompanying seed desiccation. The gene encoding p16, psp54, is expressed at a high rate during seed formation and, as a result, p16 accumulates to amount to about 8% of the histones. It is possible that the cells dispose of the excess p16 by storing it in the protein bodies. We have to note that many cases of proteins partitioned between nuclei and other subcellular compartments have been described. These include the well- known high mobility group B nonhistone proteins [35] and there are also other examples in plants [36]. It seems evident that the functional role of nuclear p16 involves a histone-mediated chromatin binding, which only occurs in vivo during seed dessication. It may be speculated that p16 is involved in protection of chromatin structure or even in the silencing of genes in preparation for, or during dormancy. A second possibility for the role of nuclear p16 arises from the work of Galvez and de Lumen [37]. These authors have cloned a cotyledon-specific cDNA from soybean encoding a 2-S albumin. The primary polypeptide is processed to give lunasin, an acidic protein of 43 amino acids. The temporal pattern of the protein expression is similar to that of p16 and it also has histone-binding capacity, with a preference for the hypoacetylated isoforms. Interestingly, lunasin, when transfected into mammalian cells, causes an arrest of cell division and the authors suggest that the role of the protein may be related to the cessation of mitosis in the last stages of plant embryogenesis via chromatin binding [37]. Although no sequence homology is found between lunasin and p16, different genes may be employed along evolution for similar functions and plants offer many examples in this line. Fig. 6. Influence of the acetylation of histones on their interaction with p16. A pull-down experiment like that of Fig. 5 was carried out with chicken erythrocyte core histones acetylated in vitro with [ 14 C]acetyl- CoA in the presence of yeast recombinant Esa1p. (A) shows the Coomassie-stained gel and its autoradiogram is given in (B). All the symbols are as in Fig. 5. 4646 J. Castillo et al. (Eur. J. Biochem. 269) Ó FEBS 2002 It has been recently pointed out that although the fundamental mechanisms involved in chromatin-dependent gene regulation are common to all eukaryotes, the data obtained from plants have revealed some interesting pecu- liarities [38]. In this context, we should mention that the events accompanying seed dessication and germination are unique in the eukaryote kingdoms and probably many lessons could be learned by studying in detail their molecular bases. We are currently studying the role of p16, in the hope that the results may give information about the physiological function of a plant protein, but also to throw some light on the mechanisms that govern the structural changes of chromatin. ACKNOWLEDGEMENTS This work was supported by Grant PB97-1368 from the Ministry of Education and Culture, Spain and by Grant BMC2001-2868 from the Ministry of Science and Technology, Spain. J. C. is the recipient of a fellowship from the Conselleria de Cultura Educacio ´ iCiencia (Valencia, Spain). We are very indebted to G. Lo ´ pez-Rodas for the gift of acetylated histones, to E. Ballestar for purified histones and for his critical comments on the manuscript and to J. Renau for his advice in electron microscopy. REFERENCES 1. Arents, G., Burlingame, R.W., Wang, B.W., Love, W.E. & Moudrianakis, E.N. (1991) The nucleosomal core histone octamer at 3.1 A ˚ resolution: a tripartite protein assembly and a left-handed superhelix. Proc. Natl Acad. Sci. USA 88, 10148–10152. 2. Luger, K., Ma ¨ der, A.W., Richmond, R.K., Sargent, D.F. & Richmond, T.J. (1997) Crystal structure of the nucleosome core particle at 2.8 A ˚ resolution. Nature 389, 251–260. 3. Daban, J.R. (2000) Physical constraints in the condensation of eukaryotic chromosomes. Local concentration of DNA versus linear packing ratio in higher order chromatin structures. Biochemistry 39, 3861–3866. 4. Kayne,P.S.,Kim,U J.,Han,M.,Mullen,J.R.,Yoshizaki,F.& Grunstein, M. (1988) Extremely conserved histone H4 N terminus is dispensable for growth but essential for repressing the silent mating loci in yeast. Cell 55, 27–39. 5. Johnson, L.M., Kayne, P.S., Kahn, E.S. & Grunstein, M. (1990) Genetic evidence for an interaction between SIR3 and histone H4 repression of the silent mating loci in Saccharomyces cerevisiae. Proc. Natl Acad. Sci. USA 87, 6286–6290. 6. Megee, P.C., Morgan, B.A., Mittman, B.A. & Smith, M.M. (1990) Genetic analysis of histone H4: essential role of lysines subject to reversible acetylation. Science 247, 841–845. 7. Durrin, L.K., Mann, R.K., Kayne, P.S. & Grunstein, M. (1991) Yeast histone H4 N-terminal sequence is required for promoter activation in vivo. Cell 65, 1023–1031. 8. Durrin, L.K., Mann, R.K. & Grunstein, M. (1992) Nucleosome loss activates CUP1 and HIS3 promoters to fully induced levels in the yeast Saccharomyces cerevisiae. Mol. Cell. Biol. 12, 1621–1629. 9. Struhl, K. (1998) Histone acetylation and transcriptional regula- tory mechanisms. Genes Dev. 12, 599–606. 10. Strahl, B.D. & Allis, C.D. (2000) The language of covalent histone modifications. Nature 403, 41–45. 11. Adams,C.R.&Kamakaka,R.T.(1999)Chromatinassembly: biochemical identities and genetic redundancy. Curr. Opin. Genet. Dev. 9, 185–190. 12. Verreault, A. (2000) De novo nucleosome assembly: new pieces in an old puzzle. Genes Dev. 14, 1430–1438. 13. McQuibban, G.A., Commisso-Cappelli, C.N. & Lewis, P.N. (1998) Assembly, remodeling, and histone binding capabilities of yeast nucleosome assembly protein 1. J.Biol.Chem.273, 6582– 6590. 14. Ito, T., Ikehara, T., Nakagawa, T., Kraus, W.L. & Muramatsu, M. (2000) p300-mediated acetylation facilitates the transfer of histone H2A-H2B dimers from nucleosomes to a histone cha- perone. Genes Dev. 14, 1899–1907. 15. Bortvin, A. & Winston, F. (1996) Evidence that Spt6p controls chromatin structure by a direct interaction with histones. Science 272, 1473–1476. 16. Borowski, P., Kuhl, R., Laufs, R., Schulze zur Wiesch, J. & Heiland, M. (1999) Identification and characterization of a histone binding site of the non-structural protein 3 of hepatitis C virus. J. Clin. Virol. 13, 61–69. 17. Stemmer, C., Briand, J.P. & Muller, S. (1997) Mapping of linear histone regions exposed at the surface of the nucleosome in solu- tion. J. Mol. Biol. 273, 52–60. 18. Juan, L J., Utley, R.T., Adams, C.C., Vetesse-Dadey, M. & Workman, J.L. (1994) Differential repression of transcription factor binding by histone H1 is regulated by the core histone amino termini. EMBO J. 13, 6031–6040. 19. Vettese-Dadey, M., Walter, P., Chen, H., Juan, L J. & Workman, J.L. (1994) Role of the histone amino termini in facilitated binding of a transcription factor. GAL4-AH, to nucleosome cores. Mol. Cell Biol. 14, 970–981. 20. Vetesse-Dadey, M., Grant, P.A., Hebbes, T.R., Crane-Robinson, C., Allis, C.D. & Workman, J.L. (1996) Acetylation of histone H4 plays a primary role in enhancing transcription factor binding to nucleosomal DNA in vitro. EMBO J. 15, 2508–2518. 21. Arents, G. & Moudrianakis, E.N. (1995) The histone fold: a ubiquitous architectural motif utilized in DNA compaction and protein dimerization. Proc. Natl Acad. Sci. USA 92, 11170–11174. 22. Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Miller, W. & Lipman, D.J. (1997) Gapped BLAST and PSI- BLAST: a new generation of protein database search programs. Nucleic Acids Res. 25, 3389–3402. 23. Eddy, S.R. (1998) Profile hidden Markov models. Bioinformatics 14, 755–763. 24. Sullivan, S.A., Aravind, L., Makalowska, I., Baxevanis, A.D. & Landsman, D. (2000) The histone database: a comprehensive www resource for histones and histone fold-containing proteins. Nucleic Acids Res. 28, 320–322. 25. Wolffe, A.P. & Pruss, D. (1996) Deviant nucleosomes: the func- tional specialization of chromatin. Trends Genet. 12, 58–62. 26. Corona, D.F.V., Eberharter, A., Budde, A., Deuring, R., Ferrari, S., Varga-Weisz, P., Wilm, M., Tamkun, J. & Becker, P.B. (2000) Two histone fold proteins, CHRAC-14 and CHRAC-16, are developmentally regulated subunits of chromatin accessibility complex (CHRAC). EMBO J. 19, 3049–3059. 27.Poot,R.A.,Dellaire,G.,Hu ¨ lsmann, B.B., Grimaldi, M.A., Corona, D.F.V., Becker, P.B., Bickmore, W.A. & Varga-Weisz, P.D. (2000) HuCHRAC, a human ISWI chromatin remodelling complex contains hACF1 and two novel histone-fold proteins. EMBO J. 19, 3377–3387. 28. Castillo, J., Rodrigo, M.I., Ma ´ rquez, J.A., Zu´ n ˜ iga, A. & Franco, L. (2000) A pea nuclear protein that is induced by dehydration belongs to the vicilin superfamily. Eur. J. Biochem. 267, 2156– 2165. 29. Ull, M.A. & Franco, L. (1986) The nucleosomal repeat length of pea (Pisum sativum) chromatin changes during germination. Plant Mol. Biol. 7, 25–31. 30. Ballestar, E., Abad, C. & Franco, L. (1996) Core histones are glutaminyl substrates for tissue transglutaminase. J. Biol. Chem. 271, 18817–18824. 31. Rodrigo, M.I. & Franco, L. (1990) Histone variants from pea (Pisum sativum): their differential presence in fractions obtained by DNase I digestion of nuclei. Physiol. Plantarum 78, 602–608. Ó FEBS 2002 H3 and H4-binding chromatin protein from pea seeds (Eur. J. Biochem. 269) 4647 32. Bendayan, M., Nancy, A. & Kan, F.W.K. (1987) Effect of tissue processing on colloidal gold cytochemistry. J.Histochem. Cytochem. 35, 983–996. 33. Ull,M.A.,Herrero,M.E.&Franco,L.(1991)PutativeHMG non-histone chromosomal proteins from pea (Pisum sativum). Plant Sci. 75, 55–62. 34. Verreault, A., Kaufman, P.D., Kobayashi, R. & Stillman, B. (1998) Nucleosomal DNA regulates the core-histone-binding subunit of the human Hat1 acetyltransferase. Curr. Biol. 8, 96–108. 35. Bustin, M. & Neihart, N. (1979) Antibodies against chromosomal HMG proteins stain the cytoplasm of mammalian cells. Cell 16, 181–189. 36. Ashgar, R., Fenton, R.D., DeMason, D.A. & Close, T.J. (1994) Nuclear and cytoplasmic localization of maize embryo and aleurone dehydrin. Protoplasma 177, 87–94. 37. Galvez, A.F. & de Lumen, B.O. (1999) A soybean cDNA encoding a chromatin-binding peptide inhibits mitosis of mam- malian cells. Nature Biotechnol. 17, 495–500. 38. Lusser, A., Ko ¨ lle, D. & Loidl, P. (2001) Histone acetylation: lessons from the plant kingdom. Trends Plant Sci. 6, 59–65. 4648 J. Castillo et al. (Eur. J. Biochem. 269) Ó FEBS 2002 . A chromatin-associated protein from pea seeds preferentially binds histones H3 and H4 Josefa Castillo, A ´ ngel Zu ´ n ˜ iga*, Luis Franco and M. Isabel. showed that H3 and H4 are almost fully acetylated, while H 2A contains a mixture of acetylated and nonacet- ylated isoforms and H2B was not acetylated at all.