Functional Interaction between Nucleosome Assembly Proteins and p300/CREB-Binding Protein Family Coactivators

Isolation of p300-interacting proteins.

For the isolation of p300-interacting proteins, the yeast strain CTY10.5 containing the LexA–β-galactosidase reporter vector pLex (HIS) and pLex-p300^611–2283 were as described previously (8, 47). Screening a 10.5-day-postcoitum mouse embryo random-primed cDNA library fused to the VP16 trans activation domain (55) yielded two positive clones containing NAP2 and one containing the NAP1 sequence. Full-length NAP2 clones were isolated through screening cDNA libraries prepared from F9 EC (18).

Immunoprecipitation.

For immunoprecipitation, U2OS cells were transfected with pG4-p300^611–2283 (30 μg), pG4 (30 μg), or pCMV-HA-NAP2 (30 μg) and after 48 h harvested in a solution containing 50 mM Tris-HCl (pH 7.4), 120 mM NaCl, 5 mM EDTA, 0.5% NP-40, 50 mM NaF, 1 mM dithiothreitol (DTT), 1 mM phenylmethylsulfonyl fluoride (PMSF), 0.2 mM sodium orthovanadate, leupeptin (0.5 μg/ml), trypsin inhibitor (0.5 μg/ml), aprotinin (0.5 μg/ml), and bestatin (40 μg/ml) and incubated on ice for 30 min. The cell extract was precleared by incubation with protein A-agarose for 1 h at 4°C, and the supernatant was incubated with a mouse anti-Gal4 monoclonal antibody (Santa Cruz) for a further 1 h at 4°C. Samples were incubated with protein A agarose for another 1 h, collected, and washed three times in extraction buffer. Immunocomplexes were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and thereafter immunoblotted with a mouse antihemagglutinin (anti-HA) monoclonal antibody (Boehringer Mannheim).

For immunoprecipitation of endogenous NAP2, extracts were prepared from A31 cells as described above. Immunoprecipitation was performed as described above with anti-NAP2 peptide antibody in the presence or absence of the competing homologous peptide (see below). After electrophoresis, immunoblotting with anti-NAP2, anti-p300 (NM11; Calbiochem), or anti-JMY (47) was performed.

Expression vectors.

The NAP2 1–123, 98–386, 1–279, 234–386, 98–386, Δ110–230, and 110–230 derivatives were prepared by PCR using the appropriate primers, and the resulting fragments were isolated and cloned into the pCDNA3 (Invitrogen) backbone at the BamHI/XhoI site. For the glutathione S-transferase (GST)–p300 constructs, the appropriate fragments were removed by a restriction enzyme digest from Gal4-p300 (37) and cloned directly into the pGEX-3KG vector (Pharmacia). pCBP HAT was as previously described (40), and expression vectors for p53 and E2F-1, together with pBax-luc, were as previously described (37). pG4-p300^19–567 was prepared by subcloning the relevant fragment from Gal4-p300 into pCDNA3.

NAP and p300 binding assays.

For the in vitro binding assay, the indicated NAP2 proteins were in vitro translated by using a TNT-coupled reticulocyte lysate system (Promega) in the presence of [³⁵S]methionine-cysteine (Amersham) and incubated with GST beads loaded with the fusion protein in TNN buffer (50 mm Tris [pH 8.0], 150 mM NaCl, 0.5% NP-40, 1 mM EDTA, and protease inhibitor cocktail [Calbiochem]). After 1 h of incubation at 4°C, the beads were washed three times with 1 ml of TNN buffer. Protein bound to the beads was eluted directly into SDS loading buffer. Flag epitope-tagged p300 protein (residues 1135 to 2414) and wild-type p300 protein (43) were expressed in Sf9 cells in the baculovirus system and bound to M2-anti-Flag antibody–agarose (Sigma). For the control, the M2 beads were treated with cell extract from uninfected Sf9 cells. NAP2 protein in pET28 (Novagen) was expressed and purified using nickel beads according to the manufacturer's instructions (Pharmacia). The eluted NAP2 protein was dialyzed against 50 mM Tris (pH 7.5)–100 mM KCl–20% glycerol–0.2 mM PMSF–0.1 mM DTT. NAP2 protein was incubated with either p300 coupled or control beads with or without histones at 4°C for 2 h in buffer containing 50 mM Tris (pH 7.5), 1.5 mM MgCl, 0.2 mM EDTA, 0.5% NP-40, 10% glycerol, protease inhibitor cocktail (Calbiochem), 0.1 mM DTT, 0.1 mM PMSF, and 0.5 mg of bovine serum albumin per ml. The protein complex on the beads was washed three times in TNN buffer and subjected to SDS-PAGE, followed by a Western blot with anti-NAP2 antiserum. Alternatively, in vitro-translated [³⁵S]methionine-labeled NAP2 and mutant derivatives were used in the binding reaction, and binding efficiency was assessed after SDS-PAGE.

Mammalian two-hybrid assays.

For the mammalian two-hybrid assay, 0.5 μg of pG4-p300^611–2283, -p300^611–1257, -p300^1302–1572, -p300^1572–2283, or -p300^1572–1906 (37) was transfected with 0.5 μg of pVP16 or pVP16-NAP2 into U2OS cells. The Gal4 reporters pG5-luc and pG4-AdML-luc have been described previously (37).

Transfection.

Transfection of SAOS2 and U2OS cells was carried out as previously described (47, 49). Immunoblotting of transfected cell extracts was performed as previously described using anti-p53 (DO1; Santa Cruz), anti-p21 (C19; Santa Cruz), and anti-NAP2 peptide antiserum.

Anti-NAP2 peptide antibody.

The anti-NAP2 peptide antibody was raised in rabbits against the peptide taken from the C-terminal region of NAP2 containing residues GDEEGEDEDDDDDDADVNPKK.

Core histone octamer preparation.

Core histone octamers were prepared from chicken blood as described previously (22). The composition was checked by SDS-PAGE and verified to contain H2A, H2B, H3, and H4.

Histone binding assays.

Glutathione beads containing Drosophila GST histone tails (17) derived from H2A, H3, H2B, and H4 (2 μg of each) and in vitro-translated NAP were incubated in the binding buffer (10 mM Tris-HCl [pH 8], 1 mM EDTA, 150 mM NaCl, 0.1% Triton X-100, 0.5 mM PMSF, 1 mM DTT) for 1 h at 4°C. Beads were collected by centrifugation, washed twice with the same buffer, and resuspended in SDS sample buffer. [³⁵S]methionine-radiolabeled NAP was resolved on a 7.5% gel.

Sucrose gradient analysis.

NAP2 and p300 (residues 1135 to 2414) proteins were purified as described above, and core histones were obtained from Sigma. For the in vitro analysis, protein complexes were formed in 50 mM Tris [pH 7.5], 1.5 mM MgCl₂, 0.4 mM EDTA, 10% glycerol, and 0.25 mg of bovine serum albumin per ml at 4°C for 30 min. A 20 to 50% sucrose gradient was sedimented at 47,000 rpm for 17 h in an SW55 rotor (Beckman). Fractions were collected from the bottom of the gradient. Typically, about 25 fractions of about 200 μl each were collected and analyzed by immunoblotting with either the anti-NAP2 peptide antibody, anti-p300 (NM11; Calbiochem), or anti-histone H2A-H2B (Santa Cruz), as described above. Each gradient analysis was repeated several times. Standard molecular mass proteins (Amersham Pharmacia Biotech) included in the gradient analysis were aldolase (158 kDa), catalase (232 kDa), and ferritin (440 kDa).

NAP proteins bind to p300.

To investigate the mechanisms through which p300/CBP coactivators regulate transcription, we screened for proteins capable of interacting with p300 in the yeast two-hybrid assay. Using LexA-p300^611–2283 as the bait, a 10.5-day-postcoitum mouse embryo activation domain-tagged library was screened from which we isolated a previously identified protein, known as NAP1, together with another less well characterized and more recently identified member of the family, referred to as NAP2 (23, 45). A comparison of NAP1 and NAP2 indicated that both proteins are of similar size (391 to 386 amino acid residues, respectively) and in certain regions possess striking similarity (Fig. 1a). The NAP family of proteins also includes TAF1, which is closely related to both NAP1 and NAP2 (Fig. 1a) and exists as two alternatively spliced variants, known as α and β (42). In the following study, we present data derived mostly from assays performed with NAP2, although it should be noted that unless otherwise stated, the properties of all three NAP proteins (NAP1, NAP2, and TAF1) were similar in the study reported here.

We addressed whether NAP family proteins can interact with p300 in mammalian cells, using both two-hybrid cell-based assays and immunoprecipitation, which we performed on both exogenous and endogenous proteins. In a two-hybrid assay performed with U2OS cells with the bait G4-p300^611–2283, clear and specific stimulation of reporter gene activity occurred when VP16 activation domain-tagged NAP1, NAP2, or TAF1β was coexpressed (Fig. 1b), supporting the idea that each of these three NAP family proteins can interact with p300. The stimulation in activity was specific for G4-p300^611–2283, since VP16-NAP had little effect on the Gal4 DNA binding domain (Fig. 1b). Further evidence for an interaction between NAP and p300 was provided by the introduction of expression vectors for G4-p300^611–2283 and HA-NAP2 into U2OS cells, followed by immunoprecipitation with anti-Gal4 and immunoblotting with anti-HA, where NAP2 was found to be specifically coimmunoprecipitated with p300 (Fig. 1c). Importantly, similar observations were made when the immunoprecipitation was performed on endogenous proteins from nontransfected murine A31 cells, where p300 was found to coimmunoprecipitate with endogenous NAP2. Thus, by using an anti-NAP2 peptide antibody that recognized murine NAP2, we found that when the immunoprecipitation was performed with anti-NAP2 in the presence or absence of a competing NAP2 peptide, p300 was detected only in the presence of the control peptide (Fig. 1d, panel ii). In summary, therefore, these results provide considerable support for the idea that p300 can physically interact with members of the NAP family of proteins and further that this interaction occurs under physiological conditions.

Binding domains in p300 and NAP2.

To resolve which region in p300 is responsible for the interaction with NAP2, we used the mammalian two-hybrid assay together with a panel of G4-p300 hybrids which were studied in the presence of coexpressed VP16-NAP2. An interaction was apparent between VP16-NAP2 and the C-terminal region of p300, encompassed within residues 1572 to 2283, which was further resolved to the region between amino acid residue 1572 and 1906 (Fig. 2a). These results were further confirmed through a biochemical binding assay in which different regions of p300 were expressed as GST fusion proteins and incubated with in vitro-translated NAP2. In agreement with the two-hybrid results, we found that GST-p300^1572–1906 bound efficiently to NAP2, whereas the overlapping fusion proteins GST-p300^1302–1737 and GST-p300^1818–2080 failed to do so (Fig. 2b and d). Moreover, using in vitro-translated luciferase as a negative control, we could not detect any binding between luciferase and p300 (Fig. 2c). Overall, these results indicate that p300 and NAP2 are likely to directly interact in mammalian cells, and the results identify a C-terminal region in p300, encompassing the CH3 region, as a likely binding domain that can interact with NAP2.

Using similar approaches, we investigated the region in NAP2 that is required for the interaction with p300 and present data derived from a biochemical binding assay, although similar conclusions were drawn from a mammalian two-hybrid analysis (data not shown). A comparison of the activity of different in vitro-translated NAP2 mutant derivatives in an assay that assessed binding to GST-p300^1572–1906 indicated that there are two broad regions that can bind to p300. Although the N-terminal region of NAP2 from residue 1 to 123 possessed marginal binding activity, an internal domain from residue 110 to 230 and the C-terminal region from 234 to 386 both bound with greater efficiency to GST-p300^1572–1906 (Fig. 3a, b, and c). These results therefore suggest that NAP2 contains at least two autonomous and separable interaction domains for p300 (Fig. 3g).

We also tested whether NAP proteins could form homo- or heteromeric protein complexes with other members of the NAP family. By assessing biochemical binding activity, we found that His-NAP2 could bind to in vitro-translated NAP2 (Fig. 3d, e, and f); similar results were obtained for NAP1 and TAF1, which formed complexes with NAP1, TAF1, and NAP2 (data not shown). An analysis of the panel of NAP2 mutant derivatives indicated that there were at least two distinct domains capable of allowing complex formation, one encompassed within the N-terminal region up to amino acid residue 123 and the other in the C-terminal region from residue 234 to 386; an internal region from residue 110 to 230 failed to bind to NAP2 (Fig. 3e). The data derived from these binding studies indicate that NAP family proteins can form both homo- and heteromeric protein complexes and that at least two binding domains are involved in facilitating these interactions (Fig. 3g).

NAP proteins augment p300-dependent transcription.

To investigate the functional consequence of NAP2 on the activity of p300, we studied the effect of NAP2 on two p300-dependent transcription factors, p53 and E2F-1 (4, 21, 37, 38, 54). The transcriptional activity of p53 was assayed on different p53-responsive promoters, and we present data derived from the bax promoter, which is a p53-responsive gene that encodes a protein involved with inducing apoptosis (41). As expected, the bax promoter was efficiently activated in the presence of p53, and a titratable increase in p53-dependent transcription was apparent as the level of NAP2 was increased, with a marginal but significant enhancement evident upon the coexpression of p300 (Fig. 4a). The stimulation of transcription by NAP2 and p300 was dependent upon the integrity of the N-terminal activation domain, since a p53 derivative containing an inactive trans activation domain, p53^22/23 (39), failed to respond to NAP2 and p300 (Fig. 4a). We also found that both NAP1 and TAF1β could similarly induce p53 activity (data not shown).

We studied the effect of NAP2 on another p300-responsive transcription factor, namely, E2F-1 (37, 54). The E2F-1 trans activation domain, which is located in the C-terminal region of E2F-1, is a Gal4 hybrid protein-activated transcription of a Gal4-responsive promoter (Fig. 4b). Coexpression of NAP2 increased the level of transcription of G4-E2F-1^380–437, and there was a further increase in activity upon the coexpression of p300 (Fig. 4b). Overall, these results indicate that NAP proteins can enhance the activity of different p300-dependent transcription factors.

The specificity of NAP2 for transcriptional activation was assessed by employing derivatives of p300/CBP expressed as hybrid proteins with the Gal4 DNA binding domain. These p300/CBP hybrids were chosen because they are known to be transcriptionally active and thus could be used to test the specificity of the effect of NAP2 on transcription. Two hybrids were studied, G4-p300^19–567, encompassing the CH1 (48) region, and G4-CBP HAT, which contains the HAT domain (residue 1099 to 1758) taken from CBP (40); both hybrids are transcriptionally active (40) (Fig. 4c). While G4-p300^19–567 and G4-CBP HAT were transcriptionally active (at a level similar to that observed for G4-E2F-1^380–437), there was little effect on their transcriptional activity upon coexpression of NAP2 (Fig. 4c), thus implying that NAP2 does not function as a general activator of transcription.

Although the previous results show that NAP2 can augment the transcriptional activity of p53 and E2F-1, the data were derived from experiments performed on transfected templates, and therefore the chromatin environment of the template may not necessarily reflect the situation encountered with endogenous cellular genes. To address this point, we assessed the effect of NAP and p300 upon endogenous genes by studying the activity of the p53-responsive gene that encodes p21^Waf1/Cip1 (15). We introduced p53 into SAOS2 cells (which are p53^−/− and Rb^−/−) together with NAP2 and p300, either alone or together, and monitored the levels of endogenous p21. As expected, we found that p53 could stimulate p21 expression as the level of p53 was increased (Fig. 4d). At a subsaturating amount of p53, we introduced NAP2 and p300. Whereas neither NAP2 nor p300 had a profound effect on the p53-dependent induction of p21, when both NAP2 and p300 were introduced together a greater level of p21 was apparent, usually exhibiting about a threefold induction (Fig. 4d). These results therefore further strengthen the idea that NAP2 and p300 can functionally interact in the induction of genes within a chromatin environment.

NAP binds to core histone tails.

In the next series of experiments, we explored the biochemical properties of NAP proteins and thereafter established a likely relevance of the association between p300 and NAP. Previous studies have indicated that NAP1 copurifies with the core histones H2A and H2B, an interaction that may facilitate the assembly of nucleosomes onto a DNA template (9, 11, 16). We extended these observations by determining if NAP2 showed binding activity for N-terminal tails and thereafter whether there was any specificity for the N-terminal tail region of core histones. In an assay using GST tail proteins derived from Drosophila H2A, H2B, H3, and H4 (17), we found that NAP2 could bind to isolated histone tails and reproducibly exhibited stronger binding activity for the H2A and H3 tail regions, although it was able to bind specifically to H2B and H4 (Fig. 5a). Similar results were observed on the tail binding specificity of NAP1 and TAF1β, which also exhibited preferential binding activity for the H2A and H3 tail regions (Fig. 5a). It is important to note that this binding specificity did not appear to reflect the conservation of protein sequence between the Drosophila and murine tail regions, which were all greater than 90% identical in sequence across the N-terminal 50 residues. Thus, these results suggest that NAP proteins can directly interact with the tail region of core histones and further imply that a level of specificity may exist that allows NAP proteins to bind more efficiently to certain core histone tails.

So far, the data indicate that NAP2 can interact with two types of protein, p300/CBP coactivators and core histones. It was therefore of interest to assess the likelihood that histones could form a ternary complex with NAP and p300. To gain evidence for this idea, we employed sedimentation analysis by centrifugation through a sucrose gradient with purified NAP2, p300, and core histones, which was performed either on isolated NAP2 or p300, on NAP2 and p300 together, or on all three proteins. By monitoring the distribution of NAP2, p300, or core histones, each analysis gave rise to a characteristic sedimentation profile, which reflected the properties of the protein complex in the gradient (Fig. 5b). Only when all three proteins were incubated together and thereafter analyzed on the gradient was there a significant and substantial shift in the sedimentation profile of NAP2 towards a protein complex with a greater mass (Fig. 5b), a result that is consistent with an interaction between all three types of proteins. Furthermore, the fractions containing the greater-mass NAP2 complex also contained p300 and core histones (Fig. 5b). The profiles of NAP2 alone and of NAP2 and p300 were similar (Fig. 5b), a result that may reflect the ability of NAP2 to form oligomers. Nevertheless, based on these results, the most likely explanation for the increased sedimentation properties observed with NAP2, p300, and histones is that they result from the formation of a ternary complex that involves all three proteins.

p300 efficiently binds to NAP in the presence of core histones.

Having gained evidence that a ternary complex between p300, NAP2, and core histones may occur, we investigated the effect that core histones had on the interaction between p300 and NAP2. To pursue this question, we identified binding conditions in which subsaturating levels of NAP2 bound specifically to p300^1135–2414 (Fig. 6a, compare lanes 2 and 4). We reasoned that under these conditions, histones may play a role in facilitating the interaction between NAP2 and p300^1135–2414. We therefore assessed the effect of core histones and found that the addition of core histones caused a notable and specific increase in the efficiency of the interaction between p300^1135–2414 and NAP2 (Fig. 6a, compare lanes 2 and 3), suggesting that the presence of core histones could act to enhance the interaction between p300 and NAP2. It should be noted that this effect was unlikely to be caused by the charged nature of histones, since polyanions, such as spermidine, failed to cause a similar effect (data not shown).

In an extension to this analysis, we assessed whether domains exist in NAP2 which can be influenced by the presence of core histones and therefore are likely to play a role in regulating the interaction of NAP2 with p300. Relative to the interaction between wild-type NAP2 and p300, we found in the absence of core histones that a C-terminal deletion in NAP2 up to residue 279 bound to p300 at about a 10-fold-greater efficiency than wild-type NAP2, a similar enhancement in binding efficiency being apparent for the internal deletion mutant Δ110–230 (Fig. 6b, compare lanes 1, 2, and 3). For both of these NAP2 mutants, we assessed the effect on the interaction with p300 of adding core histones, which we found enhanced the interaction to a far lesser extent than that observed for wild-type NAP2 binding to wild-type p300, which was, as previously noted, significantly enhanced by core histones (Fig. 6b, compare lanes 5, 6, and 7). The effects were specific to the interaction between NAP2 and p300, since similar effects were not observed on the binding of NAP2 to Flag-tagged control beads (Fig. 6b and c, compare lanes 5, 6, and 7). In contrast to results for the other two NAP2 mutants, the binding of the internal region of NAP2, 110 to 230, to p300 was enhanced by the presence of core histones (Fig. 6b), suggesting that this region of NAP2 possesses a domain that is influenced by the presence of core histones. Overall, these results suggest that the presence of core histones may influence the interaction between p300 and NAP.

NAP proteins functionally interact with p300/CBP coactivators.

The NAP proteins are an evolutionarily conserved family of chaperone-like proteins, to which the functions of both transcription and DNA replication-related mechanisms have been ascribed (1, 25, 26, 27, 42). For example, NAP proteins can act to facilitate the binding of sequence-specific transcription factors to a nucleosomal template (59) and further may play a role in nucleosome assembly by interacting with ACF (27). Other reports are consistent with NAP proteins taking part in the intracellular transport of histones (26), and a separate interaction with cyclin B has also been documented for yeast (30).

In this study, we have identified a hitherto unexpected role for NAP proteins in contributing to the activity of p300/CBP coactivators. The results strongly suggest that NAP proteins can physically associate with p300 and imply that this interaction allows NAP to augment p300 activity and thus p300-responsive transcription factors, such as p53 and E2F-1. Moreover, although NAP can form a complex with p300, we also found that NAP possesses the intrinsic capacity for binding to histone tails. Most interestingly, we acquired some evidence to suggest that NAP, p300, and core histones may interact to form a ternary complex and further that core histones may influence the interaction between NAP and p300. It seems likely, therefore, that certain aspects of p300 activity will entail a functional interaction with NAP proteins.

NAP proteins augment p300 activity.

Although it has previously been shown that NAP proteins can augment the binding of sequence-specific transcription factors to a nucleosomal template (59), the data described here define a new level of control through which NAP can regulate transcription. In this respect, it is interesting that we gained evidence to support the idea that the ability of NAP to bind to p300 may be influenced by the presence of core histones (Fig. 6), thus providing a plausible mechanism that could account for certain aspects of transcriptional control through NAP. Perhaps such a mechanism is important in maintaining the association of coactivator complexes with areas of chromatin that are destined to become, or already have become, transcriptionally active. In addition, another relevant consideration relates to the ability of NAP to influence nucleosome assembly in conjunction with ACF (27). It is possible that such an activity may favor the recruitment and stable association of p300 complexes with newly replicated DNA, which could be an important step in regulating the transcriptional activity of newly replicated genes. A role for the NAP-p300 interaction in a DNA replication-related context cannot therefore be excluded.

Transcriptional control by NAP.

Although the results reported here generally support the NAP-p300 interaction in transcriptional activation, the stage or stages at which this interaction is required in the activation process have yet to be determined. However, it has become clear that chromatin can provide a substantial barrier to transcription, mostly through the inability of sequence-specific transcription factors to penetrate to DNA in a chromatin environment, and that a variety of ATP-dependent chromatin-remodeling activities can act to overcome this barrier (3, 10, 31). One potential model suggests that chromatin-remodeling activities create a fluid chromatin environment by regulating the rate of interconversion between different chromatin states and that activating transcription complexes lock chromatin in an active state, whereas repressing complexes fix chromatin in an inactive state. In the context of this model, it is likely that the interaction between NAP and p300 functions as an activating complex which, in collaboration with a remodeling activity, locks transcription into the on state. Thus, another role for the interaction between p300 and NAP may be relevant to a later stage of a multistep activation process, once chromatin-remodeling activities have been targeted and chromatin has been modified. In this respect, it is noteworthy that our results show that NAP proteins can bind to histone tail regions. While the copurification of NAP1 with H2A and H2B has been previously noted to occur (9, 11, 26), the data presented here suggest in addition that NAP proteins can bind to the tail region of core histones. Furthermore, within this tail binding activity we obtained evidence that specificity resides for certain core histones. It will be most interesting to establish the relevance of this tail binding activity to NAP function in vivo.

The results presented in this study raise a number of important questions. For example, while we find that the activities of p53 and E2F-1 are likely to be influenced by NAP and p300, we have yet to determine whether this process represents a general feature of p300-dependent transcriptional activation. Furthermore, the p300 HAT activity appears to be important in allowing p300 to activate some transcription factors (34, 35, 36, 40, 44), although the results presented here suggest that NAP does not directly influence the isolated HAT domain (Fig. 4d). It is possible, though it has yet to be proven, that other HATs in p300 complexes are responsible for augmenting transcription when NAP assembles with p300 or, alternatively, that NAP augments p300 activity through a process that does not involve HAT.

A model for NAP-dependent transcriptional activation through p300 coactivators.

The ability of NAP to interact with both p300 and core histones, combined with the fact that the presence of core histones appears to regulate the interaction between p300 and NAP, implies a plausible model that may account for the capacity of NAP to stimulate p300-dependent transcription (Fig. 7). We suggest that the stable association of p300 coactivators with transcriptional activation domains and the subsequent p300 binding to NAP may result in an overall stabilizing influence which acts to strengthen the association of coactivator complexes with chromatin. According to the results presented here, the interaction between p300 and NAP may be augmented by the histone, perhaps nucleosomal, environment of a chromatin template. We suggest that this is a useful mechanistic feature that is a desirable property of transcriptional coactivators, since it provides a process that is destined to aid transcriptional activation. Moreover, we would also anticipate, based on the results in previous reports (59), that the presence of NAP in close proximity to a transcription factor binding site positioned within a nucleosome will favor DNA binding activity, which we imagine is an equally important event that will contribute to enhanced transcriptional activity.

Overall, we suggest that NAP's ability to recruit and stabilize the interaction of p300 with chromatin, together with its capacity to facilitate stable transcription factor binding to a nucleosomal template, places it in a central position of control in regulating p300-dependent transcription.