Regulation of Histone Acetylation and Transcription by Nuclear Protein pp32, a Subunit of the INHAT Complex

Histone acetylation by CBP/p300 and coactivators is considered to be a key mechanism of and A multiprotein cellular complex, INHAT, containing the Set/TAF-I β oncoprotein and pp32 strongly inhibits the HAT activity of p300/CBP and PCAF by histone masking. Here we report that the INHAT complex and its subunits have overlapping but distinct HAT-inhibitory and histone binding characteristics. We provide evidence suggesting that the histone binding and INHAT activity of pp32 can be regulated by its physical association with other INHAT subunits. In vivo colocalization and transfection studies show that pp32-INHAT domains are responsible for histone binding, HAT-inhibitory activity, and repression of transcription. We propose INHAT and its subunits may function by modulating histone acetyltransferases through a histone masking mechanism and may play important regulatory roles in the establishment and maintenance of the newly proposed “histone code” of chromatin. followed conjugated anti-mouse IgG (Jackson ImmunoResearch Laboratories). For nucleus detection, cells were incubated with DAPI 1 (Molecular Probes) stain and mounted with Gel/Mount. Slides were examined under oil immersion with a confocal microscope with BioRad 1024-ES using Confocal Assistant TM program and Nikon Eclipse E-600 with X60 objectives. Raw data images were processed further using Confocal Assistant TM program.

INTRODUCTION this domain in the function of these proteins. Besides their C-terminal acidic domains, little is known about any common or distinct characteristics of Set/TAF-Iβ and pp32 in HAT and transcriptional regulation. Additionally, it is not clear how the activity of the individual INHAT subunits contributes to the overall activity of the complex and whether the activity of the subunits can be dictated based on whether they are free or part of the complex.
In this report, we have analyzed the HAT inhibitory characteristics of INHAT subunits with emphasis on pp32. We show that the INHAT complex and each of its subunits show overlapping but distinct specificities in their HAT inhibitory activity by targeting different histone subunits. The colocalization studies show that while the full length pp32 colocalizes with distinct histone domains in the nucleus, a mutant pp32 defective in HAT inhibition and histone binding fails to colocalize with histones indicating that histone binding is a prerequisite for pp32 function in the regulation of histone acetylation and transcription. p300 and Flag-PCAF was carried out as described (22). The INHAT complex was purified from HeLa cells as described (7).
HAT assay -Histone and nucleosome acetyltransferase assays were performed as described previously (7).

In vitro immunoprecipitation and interaction assays-[ 35 S]methionine-labeled INHAT
subunits and pp32 deletion mutant proteins were generated by in vitro transcription/translation and used in histone binding and immunoprecipitation assays as described (7). For histone binding specificities of INHAT subunits, individual histones were used for each assay instead of total histones.
Transfection assays-CV-1 cells were seeded at 20, 000 cells/well of a 48 well dish as described (7)  Peptides-The pp32 peptide (residues 150-180) was synthesized commercially (Annovis, Aston, PA). Product purity was greater than 85 % and molecular weight of synthesized peptide was confirmed by mass spectrometer analysis.
Immunostaining-For endogenous pp32 detection, NIH-3T3 cells seeded in 35 mm 2 plate were fixed and incubated with anti-pp32 antibodies and immunocytostained with secondary antibody conjugated with FITC (Jackson ImmunoResearch Laboratories).
NIH3T3 cells were seeded overnight and transiently transfected with pEGFPC1-INHAT subunits and pEGFPC1-pp32-C2 as indicated. After 24 hours, cells were washed with PBS, and fixed in 50 % acetone/methanol. For histone staining, cells were incubated with anti-histones antibodies (Chemicon International), followed by incubation with Cy3 conjugated anti-mouse IgG (Jackson ImmunoResearch Laboratories). For nucleus detection, cells were incubated with DAPI 1 (Molecular Probes) stain and mounted with Gel/Mount. Slides were examined under oil immersion with a confocal microscope with BioRad 1024-ES using Confocal Assistant TM program and Nikon Eclipse E-600 with X60 objectives. Raw data images were processed further using Confocal Assistant TM program.

Specificity of inhibition of histone acetylation by pp32 -Using baculovirus expressed
Flag-tagged p300, PCAF and pp32 purified as gluthathione-S-transferase (GST) fusion proteins from E.coli, we tested pp32's ability to inhibit histone and nucleosome acetylation. In agreement with previous results, the acetylation of core histones by p300 and PCAF was inhibited by addition of saturating concentrations of pp32 (Fig. 1, compare lanes 1 and 3 to lanes 2 and 4, respectively). Additionally, nucleosomal histone acetylation by p300 and PCAF was also inhibited by pp32 (data not shown). No proteolytic degradation of histones in the presence of pp32 was evident (Fig. 1, coomassie and data not shown). These observations demonstrate that pp32 inhibits histone acetylation and prompted us to finely map the HAT-inhibitory domain(s) of pp32.
Mapping of the pp32 INHAT domain-pp32 contains previously characterized structural domains, including a highly acidic C-terminal domain homologous to that of TAF-Is (amino acids 167-249), an α-helical N-terminal domain (amino acids 1-206), and a central domain containing the tumor suppressive activity spanning amino acids 150-174 (30,36). We generated and purified a series of N and C-terminal deletion mutants of pp32 as GST-fusion proteins and examined their HAT inhibitory activity in order to identify the HAT inhibitory domain(s) of pp32 ( Fig. 2A). Approximately the same amount of expressed mutant proteins was used for HAT inhibitory activity assay and purified GST protein alone has no effect on the assay system (data not shown). When compared to the full length protein (construct 1), deletion of large parts of the acidic C-terminal residues 181-249 of pp32 (pp32-C1, construct 2) overall had no effect on the INHAT activity.
However, further deletion of additional C-terminal sequences (residues 151-180) resulted in the loss of entire inhibitory activity (pp32-C2, construct 3). Interestingly, this domain overlaps with the tumor suppressor domain of pp32 (36). These results suggest that a major HAT inhibitory domain of pp32 resides between amino acids 151-180.
A series of pp32 N-terminal deletion mutant proteins were also generated and tested for their ability to inhibit histone acetylation. It is interesting that the deletion of the first 1-59 residues of pp32 (pp32-N1, construct 4), showed a slight increase in INHAT activity when compared to that of the full-length pp32. This suggests the possible existence of a negative auto-regulatory domain in the N-terminus of pp32 (residues 1-59). The overall role of this domain in pp32 activity in vivo remains to be determined. Consistent with the results of C-terminal deletion analysis, both pp32-N2 (construct 5) lacking residues 1-119 and pp32-N3 (construct 6) which lacks residues 1-149 of the N-terminus showed strong HAT inhibitory activity. Although deletion of residues 181-249 did not significantly affect the HAT inhibitory activity of pp32-C1 (construct 2), pp32-N5 which only Histone binding specificity of pp32 and INHAT-Histone and chromatin binding activity of both TAF-I proteins has been previously reported (37). Additionally we have shown that the INHAT complex associates with histones on chromatin in vivo and blocks them from serving as acetylase substrates implying a critical role for histone binding in the INHAT mediated inhibition of acetylation (7). To determine whether HAT-inhibitory properties of pp32 mutants correlate with their histone binding characteristics, we compared the histone binding properties of pp32 mutants.
In vitro immunoprecipitation experiments utilizing radiolabeled full length and mutant pp32 proteins with total histones were carried out to determine the histone binding properties of pp32. pp32-C1, which retained histone acetylation inhibitory activity, was efficiently immunoprecipitated with anti-histones antibodies ( correlation between histone binding and the HAT inhibitory activity of pp32. In addition, the N-terminal mutant proteins were also tested for their histone binding ability and as predicted from their HAT inhibitory activity, all N-terminal mutants tested bound to histones with comparable affinity (Fig. 3B, lanes 3, 6, and 9). Together these results suggest histone binding as a prerequisite for HAT inhibition by pp32.
We also tested the hypothesis that pp32 and other INHAT subunits might have distinct binding affinities towards different histone subunits. In order to address this question of histone binding specificity, we analyzed the binding properties of pp32 and the other INHAT subunits to individual histones in vitro. For that purpose radiolabeled full-length pp32 was incubated with each histone subunit separately and immunoprecipitated with anti-histones antibodies. The anti-histones antibodies bind specifically to an antigenic determinant which is present on all four histone subunits and pulled down approximately the same amount of isolated histone subunits (data not shown). Histones H2B and H3 showed stronger interaction with pp32 (Fig. 4, panel I). Analysis of histone subunit binding by Set/TAF-Iβ demonstrated that this subunit has the strongest affinity to histone H3 and histone H4 (Fig. 4, panel II). TAF-Iα, bound with the higher affinity to histones H4 and H2B (Fig. 4,  To determine whether histone binding properties of INHAT and its subunits correlate with their ability to inhibit individual histone acetylation, the inhibitory activity of INHAT and its subunits towards each histone subunit was titrated (Fig. 4B). In agreement with histone subunit binding properties of pp32 (Fig. 4A), inhibition of histone H2B acetylation by pp32 was slightly higher than that of the other subunits (Fig. 4B, panel pp32). The inhibition by purified Set/TAF-Iβ protein was very similar for all histone subunits although it displayed a lower binding affinity for histone H2B.
However, the INHAT complex showed slightly lower inhibitory activity towards histone H2B in agreement with its histone subunit binding properties. Together these results indicate that INHAT and its individual subunits display distinct but overlapping histone substrate preferences in vitro and that the inhibition of histone acetylation specificity of INHAT subunits may be regulated by their physical association with other subunits.

Colocalization of INHAT subunits with histones-The observation that pp32 and other
INHAT subunits associate with histones in vitro led us to investigate whether they would individually colocalize with histones in vivo. We have previously shown, using ChIP assays, that INHAT forms a complex with histones in vivo (7). Localization of endogenous pp32 was examined using anti-pp32 antibodies and a secondary antibody localization, they nontheless showed nuclear colocalization with histones at varying levels ( Fig. 5D and E).
Our in vitro binding studies demonstrated that pp32-C2 failed to inhibit histone acetylation and did not bind to histones. We utilized GFP-pp32-C2 in colocalization analysis to further demonstrate the interaction specificity between pp32 and histones in vivo. Like GFP-pp32, GFP-pp32-C2 localized in the nucleus (Fig. 5C GFP-pp32-C2).
However, in contrast to the GFP-pp32 which showed extensive colocalization with histones ( Fig. 5B, overlay)  In summary, our results describe overlapping but distinct HAT inhibitory and histone binding properties of INHAT and its subunits. The recently proposed "histone code" hypothesis envisions that the level and combination of epigenetic marking including acetylation, methylation, and phosphorylation on histones will play a fundamental role in chromatin-based processes including transcription (2,6). An extension of this hypothesis is that the epigenetic marking of histones including acetylation therefore must be regulated. Given that the "histone code" dictates the transition between transcriptionally active and silent chromatin states, we propose that INHAT and other yet to be identified parallel complexes will play a significant role in the establishment and maintenance of the "histone code" reflecting the transcriptional off or on states of chromatin domains. Coomassie staining of p300 HAT inhibition assay gel is shown.