Acetylation of MyoD by p300 Requires More Than Its Histone Acetyltransferase Domain*

MyoD, an essential transcription factor involved in muscle cell terminal differentiation, is regulated by acetylation, as are a number of other transcription fac-tors, but the histone acetyltransferase enzyme responsi-ble for this acetylation is a matter of controversy. In particular, contradictory findings have been reported concerning the ability of CBP/p300 to acetylate MyoD in vitro . Here we provide an explanation for this discrepancy: although full-length p300 does indeed acetylate MyoD, a fragment of p300 corresponding to its histone acetyltransferase domain does not. In addition to clearly demonstrating that p300 acetylates MyoD in vitro , these results underscore the necessity of using full-length histone acetyltransferase enzymes to draw valid conclusions from acetylation experiments. Acetylation has recently emerged as a central mode of regu-lation for proteins. Histone acetyltransferases (HATs), 1 which are able to acetylate histone and non-histone proteins, are involved in a variety of essential cellular processes such as muscle-cell terminal differentiation. Muscle-cell terminal differentiation involves several families of transcription factors, including myogenic basic helix-loop-helices (MyoD, Myf-5, myo-genin, and MRF-4) (1), and transcriptional co-regulators with histone acetyltransferase activity, the PCAF/GCN5 family (2–4) and the CBP/p300 family (5, PCAF (7) and CBP/p300

HATs are involved at different steps of the differentiation program; differentiation triggers the acetylation of histones on muscle-specific promoters (11). In addition, the myogenic basic helix-loop-helix protein MyoD is also acetylated in myogenic cells (12). MyoD acetylation increases its transcriptional activity by influencing its ability to bind DNA (13) and to interact with other proteins (11). In vitro, MyoD is acetylated by PCAF (13). Acetylation of MyoD by CBP/p300, on the other hand, has been somewhat controversial. We reported that MyoD is acetylated by CBP or p300 with an efficiency similar to that observed with PCAF (12); in contrast, others have concluded from similar experiments that MyoD is acetylated by PCAF and not by p300 in vitro (14). The latter study relied on a truncated ver-sion of the p300 protein (from amino acid 965 to amino acid 1810) that is often employed in this type of study. It is of note that the p300 965-1810 fragment has lost the main protein-protein interaction domains of p300. In particular, the regions of p300 previously shown to interact with MyoD, the CH3 domain (amino acids 1620 -1891) (15) and the N-terminal KIX domain (amino acids 379 -654) (16), are truncated or altogether deleted in p300 965-1810 . Thus a possible explanation for the discrepancy between these two series of experiments might be that a physical contact between the HAT and its substrate is required for acetylation and that the interaction domains are critical for the reaction to take place. Here, we present the results of a direct test of this hypothesis, which show that acetylation of MyoD by p300 is linked to the ability of the two proteins to physically interact.

EXPERIMENTAL PROCEDURES
HAT Assay and Protein Acetylation-Recombinant GST-PCAF and FLAG-tagged p300 and p300 965-1810 were purified from bacteria and insect cells as described previously (12). HAT activity was measured as described previously (17) using either nucleosomes purified from HeLa cells or a peptide corresponding to the first 24 amino acids of histone H3. Bacterially produced recombinant MyoD was produced and acetylated as described previously (12).
Co-immunoprecipitation-20 ng of recombinant Escherichia coli-produced MyoD protein was incubated with 50 -100 ng of baculovirusproduced FLAG-p300 wild type or FLAG-p300 965-1810 in 25 mM HEPES, pH 7.2, 150 mM potassium acetate, 2 mM EDTA, and 0.1% Nonidet P-40 followed by immunoprecipitation with anti-FLAG antibody (M-2, Sigma). Proteins were resolved on SDS-polyacrylamide gel electrophoresis, and the presence of MyoD in the complexes was revealed by Western blot with anti-MyoD antibody (C-20, Santa Cruz Biotechnology).

RESULTS AND DISCUSSION
To determine whether p300 integrity is required for MyoD acetylation, we directly compared the two forms of p300 for their ability to acetylate MyoD. The histone acetyltransferase activities of the two proteins, as well as of PCAF, were first standardized based on their ability to acetylate histones, as measured by incorporation of 14 C from radiolabeled acetyl-CoA. We used as substrates either a synthetic peptide corresponding to the first 24 amino acids of histone H3 (17) (Fig. 1A) or nucleosomes prepared from HeLa cells (Fig. 1B). Note that we purposely chose to use a concentration of the p300 965-1810 fragment that acetylates histone substrates with a significantly higher efficiency than the full-length p300 and PCAF proteins (about 3-fold higher). We next assayed the three HATs using MyoD as a substrate. In contrast to the results obtained with the histones, the p300 965-1810 fragment did not induce detectable incorporation of 14 C in MyoD (Fig. 1C). As published previously (12), PCAF and p300, when full-length, catalyzed the incorporation of 14 C into MyoD to similar levels.
These data demonstrate that MyoD acetylation by p300 requires more than just its HAT domain. The most likely explanation is that physical interaction between the enzyme and its * This work was supported by grants from the Association Française contre les Myopathies and by Grant QLG1-1999-00866 from the European 5th PCRDT. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ Supported by the Association pour la Recherche sur le Cancer. § To whom correspondence should be addressed: CNRS Unité Propre de Recherche 9079, Institut André Lwoff, 7 Rue Guy Moquet, 94800 Villejuif, France. E-mail: ahbellan@vjf.cnrs.fr. substrate is required for efficient acetylation of the substrate. Indeed, the results of a co-immunoprecipitation assay indicated that although full-length p300 strongly interacts with MyoD in vitro, the p300 965-1810 fragment does not (Fig. 2). These results show a correlation between MyoD acetylation and physical interaction with MyoD. They demonstrate without ambiguity that p300 is in fact able to acetylate MyoD in vitro, and most importantly, they show that, in order to arrive at valid conclusions, full-length histone acetyltransferases must be used in in vitro assays. In light of this finding, and given that the widely used p300 965-1810 fragment has lost the main protein-protein interaction domains of p300, some previously published data may need to be re-evaluated.
FIG. 1. Full-length p300, but not p300 965-1810 , acetylates MyoD in vitro. A, recombinant GST-PCAF or baculovirus-produced p300 or p300 965-1810 incubated with a peptide corresponding to the first 24 amino acids of histone H3 and C14 acetyl-CoA. The radioactivity was measured in a ␤ counter (mean of three determinations). B and C, nucleosomes purified from HeLa cells (B) or recombinant MyoD (C) acetylated in vitro as above. Proteins were analyzed by SDS-polyacrylamide gel electrophoresis. Coomassie stains and autoradiograms of the gels are shown as indicated.