Identification of New p53 Acetylation Sites in COS-1 Cells

Abstract

The p53 tumor suppressor protein is a key regulator of cell cycle and death that is involved in many cell signaling pathways and is tightly regulated in mammalian cells. Post-translational modifications of p53 have been investigated previously mainly using antibodies. In this study, utilizing LC-MS/MS analysis, we have characterized p53 protein from COS-1 cells. Several already known post-translational modifications were observed, such as phosphorylation on serines 15, 33, 315, and 392 as well as acetylation on lysines 305, 370, 372, 373, 381, 382, and 386. Interestingly novel acetylation sites were identified at lysines 319 and 357. This study confirmed that p53 is a highly acetylated protein and revealed new acetylation sites that might aid the further understanding of p53 regulation.

p53 plays a key role in cellular homeostasis and is at the heart of a complex network of protective mechanisms safeguarding cellular integrity. Because of its central function in processes such as cell cycle regulation, apoptosis, DNA repair, cellular senescence, and apoptosis, the p53 pathway is crucial for effective tumor suppression, and mutations in p53 that compromise its function occur in more than 50% of cancers (1, 2). Interestingly p53 appears to be highly post-translationally modified, and although ubiquitination, neddylation, sumoylation, and methylation have been described, phosphorylation and acetylation are the most commonly reported modifications of p53 (3). Both phosphorylation and acetylation affect p53 stability and activity and are induced following various types of stress (4). For example, phosphorylation at Ser¹⁵, Ser²⁰, Thr¹⁸, and Ser³⁷ disrupts the interaction between p53 and its major negative regulator, MDM2 (3, 5), leading to an increase of p53 protein expression and activity. The acetylation sites are located mostly in the C-terminal end of p53 where the tetramerization and regulatory domains localize. Sites of acetylation have been reported at lysine residues 120, 164, 305, 320, 370, 372, 373, 381, 382, and 386 (6–14), and importantly, acetylation has recently been shown to be indispensable for p53 activation (14). In this context of high regulation of p53 through post-translational modifications, we aimed at identifying potential new p53 modifications by using mass spectrometry. p53 was obtained from the kidney fibroblast-like COS-1 cells that are known to produce a high amount of p53. In fact, in these cells, p53 is bound to SV40 large T antigen (15–17). This association sequesters the gene transactivation function of p53, rendering it inactive as a transcription factor. The sequestration leads to an accumulation of p53 as part of a complex with SV40 large T antigen (17). Utilizing CID analysis and high accuracy mass measurements, a number of different modifications, both known and novel, were deciphered. They encompass phosphorylation of serine residues 15, 33, 315, and 392 and acetylation of lysines 305, 319, 357, 370, 372, 373, 381, 382, and 386. The acetylation of p53 at Lys³¹⁹ and Lys³⁵⁷ is reported for the first time.

MATERIALS AND METHODS

Cell Culture—

COS-1 cells were obtained from the University of California San Francisco Cell Culture Facility and were cultured in Dulbecco's modified Eagle's medium supplemented with 10% heat-inactivated fetal calf serum, 100 units/ml penicillin/streptomycin and incubated in a 5% CO₂ atmosphere at 37 °C.

Purification of the p53 Protein—

p53 was immunoprecipitated from cell extracts representing the equivalent of 10⁸ cells. Cells were rinsed three times with ice-cold PBS before addition of lysis buffer (50 mm Tris, pH 7.8, 150 mm NaCl, 1 mm EDTA, 1% Nonidet P-40, 0.1% SDS, 5 mm sodium pyrophosphate, 10 mm β-glycerophosphate, 10 mm sodium butyrate, phosphatase inhibitors 1 and 2 (Sigma), protease inhibitors (Complete, Roche Applied Science)). Lysates were left rotating for 1 h at 4 °C. Insoluble material was pelleted at 13,000 × g for 30 min at 4 °C. Lysates were precleared for nonspecific binding to beads by incubation with normal serum Ig-bound agarose beads (Santa Cruz Biotechnology) for 1 h at 4 °C. After 5 min of centrifugation at 1500 rpm, the supernatants were transferred to fresh tubes for further analysis. Immunoprecipitations were performed by combining lysates and Fl-393 (Santa Cruz Biotechnology) p53-agarose-bound antibodies (ratio, 1 mg/5 μg) and allowed to rotate overnight at 4 °C. The beads were then washed three times with the lysis buffer, three times with lysis buffer containing 0.5 m NaCl, three times with PBS, and once with water. The proteins were then eluted by adding triethanolamine at pH 11.5 in H₂O. The triethanolamine eluate was immediately neutralized using 0.25 volume of 1 m Tris (pH 6.8). Resulting lysates were then loaded and fractionated on an SDS-polyacrylamide gel (10%, Novex, Invitrogen).

Identification of p53 Post-translational Modifications by Mass Spectrometry—

After being separated by SDS-PAGE and Coomassie Blue-stained (Simply Blue, Invitrogen), the p53 protein band was excised and processed for digestion with trypsin (Promega) or Asp-N (Roche Applied Science) as described previously (18) with minor modifications. In brief, 100 ng of trypsin or Asp-N was used for each gel band, and digestions were carried out at 37 °C. Peptides were extracted from gel pieces with 100 μl of 50% acetonitrile, 2% formic acid twice; vacuum centrifuged to dryness; and resuspended in 10 μl of 0.1% formic acid in water prior to analysis of 1 μl. The peptides were fractionated by reversed phase high pressure liquid chromatography using an Eksigent one-dimensional NanoLC system (Dublin, CA). The column was a 100-μm-inner diameter × 150-mm self-packed column with Ultro120 (Peeke Scientific, Redwood City, CA) C₁₈ resin. The flow rate was 350 nl/min; solvents were water, 0.1% formic acid (A) and acetonitrile, 0.1% formic acid (B); and the gradient was 5–30% B over 40 min. The HPLC system was coupled to a QSTAR Pulsar (of quadrupole/quadrupole orthogonal acceleration TOF geometry; MDS Sciex, Toronto, Canada) or LTQ-FT (linear ion trap, FT ICR; Thermo Fisher Scientific, San Jose, CA) mass spectrometer operating in positive ion mode. The peak list-generating software was Analyst Mascot.dll v1.6b19 (Applied Biosystems) for QSTAR data or Mascot Distiller v2.1 for LTQ-FT data. The search engine used was Protein Prospector v4.25.4 (in-house version) (19). The database searched was Swiss-Prot.2007.04.19 (264,492 entries). The database search parameters were as follows. The enzyme specificity considered was either trypsin or Asp-N; the number of missed cleavages permitted was 3. Variable modifications considered included oxidation of methionine; N-terminal protein acetylation; pyroglutamate formation from N-terminal glutamine residues; modification of cysteine by carbamidomethylation or propionamide; phosphorylation on Ser, Thr, or Tyr; GlcNAcylation on Ser, Thr monomethylation of Lys; dimethylation of Lys, trimethylation of Lys; acetylation of Lys; methylation of Arg; dimethylation of Arg; methylation of His; methylation of Glu; and ubiquitination of Lys. The mass tolerance for precursor ions was set at 100 ppm, and the mass tolerance for fragment ions was set at 0.8 Da. For the LTQ-FT instrument the mass tolerance for precursor ions was set at 20 ppm, and the mass tolerance for fragment ions was set at 0.6 Da. The threshold for peptide score and maximum E-value for accepting individual MS/MS spectra were 15 and 0.01, respectively. All modification site assignments were determined by manual spectrum interpretation.

RESULTS

Sequencing of p53 Protein in COS-1 Cells—

p53 was immunoprecipitated from COS-1 cells. The resulting pellet was resolved by SDS-PAGE (Fig. 1A). The bands of interest were then excised and digested with trypsin or Asp-N, and the resulting digests were subjected to LC-/MS/MS analyses. Fig. 1B shows the sequence coverage observed from tryptic and Asp-N digestions of p53, illustrating that complete sequence coverage was achieved from a total of 107 unique peptides identified. This emphasizes the advantages of using different proteolytic enzymes in protein characterization. The tryptic digest yielded 87% of the p53 sequence, whereas the Asp-N digest yielded 50% sequence coverage. The use of two different mass spectrometers in this case did not alter the sequence coverage observed in each digest.

Fig. 1.

CID-based sequence coverage of p53. p53 was isolated by immunoprecipitation, and the resulting pellet was fractionated by SDS-PAGE (A). The protein band corresponding to p53 was cut from the gel and digested, and peptides were analyzed by LC-MS/MS on QSTAR Pulsar and LTQ-FT mass spectrometers. Combining the results of trypsin and Asp-N digestions, 100% sequence coverage of p53 was achieved (B). Sequences observed after trypsin digestion are in red, and sequences observed after Asp-N digestion are underlined.

Identification of Known p53 Post-translational Modifications—

The data acquired were searched allowing for a wide variety of modifications, and all modified spectra reported here were manually verified. As shown in Table I, several peptides were identified as acetylated and phosphorylated, resulting in a total (known plus unknown) of 13 different residues modified along the p53 protein. Phosphorylation was detected on two residues near the N terminus of the protein. One was located on serine 15, and the CID spectrum to support this assignment is shown in Fig. 2A. The phosphorylation site was identified because of the detection of unmodified y₇ and y₈ ions, whereas y₁₁ and y₁₂ were both shifted by the mass of a phosphate group. Of the three residues between these fragments, LSQ, serine 15 is the only residue that can be phosphorylated. The other phosphorylation identified in the N-terminal region was on serine 33. In the spectrum in Fig. 2B, the mass difference between y₇ and y₈ ions corresponds to a phosphorylated serine (167 Da) and identifies serine 33 as the site of modification. The other modifications identified in this work lie in the C-terminal part of the protein where the regulatory domain of p53 is located (see Fig. 4). Many of these modifications were found in close proximity to each other. A series of acetylations on lysines 370, 372, and 373 were identified. In the case of Lys³⁷⁰, Fig. 2C shows the mass of the y₂ ion is as expected, but the y₃ ion is shifted by 42 Da, i.e. the mass of an acetyl group. Three modified forms of the peptide containing lysines 372 and 373 were identified. Peptides acetylated only on each of the lysines as well as one acetylated on both lysines at the same time were observed. The spectrum presented in Fig. 2D corresponds to the form that is doubly acetylated. In this spectrum the b₂ ion is shifted by 42 Da, and the mass of b₃ reflects an additional acetylation. Another pair of acetylated amino acids was identified on lysines 381 and 382. As Fig. 2E shows, the mass shifts of b₂ and b₃ ions clearly identify the modified residues. Another phosphorylation was also deciphered on the penultimate residue of the protein, on serine 392. Fig. 2F shows the corresponding CID spectrum. All y ions observed are shifted by the mass of the phosphate group. Two other modifications were also deciphered: acetylation on lysine 305 and phosphorylation on serine 315. Although these two modifications occur in the same tryptic peptide, they were not observed together. Their corresponding CID spectra are presented in Fig. 3, G and H, respectively. The last modification was found at the C-terminal end of the protein: acetylation of Lys³⁸⁶ was detected. Interestingly this modification was observed only in combination with two other modifications: acetylation on lysine 382 and phosphorylation on serine 392. The corresponding CID spectrum (m/z 783.31(2+)) is shown in Fig. 2I. The y ion series from y₂ onward includes the phosphorylation on serine 392. The next mass shift is between the y₇ and y₈ ions, identifying lysine 386 acetylation. All b ions display the mass shift corresponding to lysine 382 acetylation.

Table I.

Modified peptides identified from tryptic and Asp-N digestions representing each post-translational modification detected

Sequences in bold represent novel modifications. p, phosphorylation; Ac, acetylation; Ox, oxidation.

m/z	z	Peptide sequence	Score	Expectation value	Amino acid modified
976.78	3	1Ac-M(Ox)EEPQSDPSIEPPLS(p)QETFSDLWK24	38.7	4.90e−07	Ser15
771.72	3	21DLWKLLPENNVLS(p)PLPSQAVD40	40.1	1.80e−09	Ser33
536.25	3	291GEPC(CAM)HELPPGSTK(Ac)R306	33.5	1.90e−04	Lys305
710.78	2	307ALPNNTSSS(p)PQPK319	46.3	5.50e−10	Ser315
755.91	2	307ALPNNTSSSPQPK(Ac)K320	38.1	5.10e−09	Lys319
756.73	3	343ELNEALELKDAQAGK(Ac)EPAGSR363	70.3	8.60e−09	Lys357
346.19	3	364AHSSHLK(Ac)SK372	33.6	5.80e−06	Lys370
531.77	2	371SK(Ac)K(Ac)GQSTSR379	45.3	2.90e−06	Lys372 and Lys373
340.85	3	371SK(Ac)KGQSTSR379	31.2	2.10e−05	Lys372
403.22	2	373K(Ac)GQSTSR379	37.8	4.40e−06	Lys373
350.51	3	380HK(Ac)K(Ac)FMFK386	33.2	1.60e−04	Lys381 and Lys382
371.68	2	382K(Ac)FMFK386	23.8	6.30e−04	Lys382
783.31	2	382K(Ac)FMFK(Ac)TEGPDS(p)D393	32.3	1.60E−07	Lys382, Lys386, and Ser392
762.31	2	382K(Ac)FMFKTEGPDS(p)D393	31.4	1.60e−07	Lys382 and Ser392
677.26	2	383FMFKTEGPDS(p)D393	48.5	4.50e−11	Ser392

Fig. 2.

Identification of p53 modifications reported in the literature. The CID spectra shown here represent post-translational modifications of p53 that have already been described previously. Data were analyzed using Protein Prospector for peptide identification. Only the detected y and b ions are labeled. p, phosphorylation; Ac, acetylation; Ox, oxidation.

Fig. 4.

Map of p53 post-translational modifications identified. P represents phosphorylations; Ac represents acetylations. The asterisks indicate modifications reported previously for endogenous p53 from COS-1 cells; # indicates novel modifications.

Fig. 3.

Identification of new p53 modifications. The CID spectra shown here represent post-translational modifications of p53 reported for the first time. Data were analyzed using Protein Prospector for peptide identification. Only the detected y and b ions are labeled. Ac, acetylation.

Identification of Novel p53 Post-translational Modifications—

The experiments also allowed new modifications to be uncovered. Acetylation on lysine 319 was identified. As shown in Fig. 3A, an unmodified y₁ followed by a modified y₂ ion allowed the identification of lysine 319 as the acetylated amino acid. Lysine 357 was also detected as acetylated. The CID spectrum in Fig. 3B shows an unmodified y₆ but a modified y₇ ion, allowing the residue of modification to be pinpointed. These two modified residues border the tetramerization domain of p53 (Fig. 4).

DISCUSSION

We have identified phosphorylation and acetylation sites on endogenous p53 from COS-1 cells. The phosphorylation sites detected were localized at the N- and C-terminal ends of the protein on serines 15, 33, 315, and 392. Our results confirmed some phosphorylation sites previously described by Tack and Wright (20) using Edman sequencing and radiolabeling. They identified phosphorylation on Ser⁹, Ser¹⁵, Ser²⁰, either Ser³³ or Ser³⁷, at least one of Ser⁹⁰ or Ser⁹⁹, Ser³¹⁵, and Ser³⁹². The new results presented here were able to pinpoint phosphorylation on Ser³³, resolving ambiguity for one of the sites in their study. The additional phosphorylations described by Tack and Wright (20) that were not detected in the current study might be related to the use by the authors of CV-1 cells, the parent cell line of the COS-1 cells used in our present study. They infected the CV-1 cells with SV40 wild-type virus. However, the COS-1 cell line is stable because it contains only a defective mutant of SV40 that codes only for wild-type T antigen. The new discoveries presented here are the acetylation sites in the C-terminal region of p53. Acetylation occurred on Lys residues in positions 305, 319, 357, 370, 372, 373, 381, 382, and 386. None of these sites have been described previously on endogenous p53 from COS-1 cells, although many have been reported in other species (4, 21, 22). Work by Borger and DeCaprio (25) described the acetylation of SV40 large T-bound p53 at lysine 373 by Western blot analysis. In their experiment, they used p53 co-immunoprecipitated with SV40 large T from whole-cell lysates obtained from human osteosarcoma U-2 OS cells stably expressing wild-type SV40 large T antigen.

Importantly two acetylation sites identified in our study have never been described so far and are located on lysines 319 and 357. Although mutations at lysine 319 have occasionally been detected in non-small cell lung carcinomas (23) and hepatocellular carcinomas (24), lysines 319 and 357 have never been reported to be post-translational modification sites, and consequently, their potential roles in p53 regulation have not been explored. Our results provide novel and interesting observations as they were obtained without stressing the cells. Modifications on p53 are triggered by stress. Few residues are thought to be constitutively modified. Constitutive phosphorylations at Ser⁶, Ser³³, Ser³¹⁵, and Ser³⁹² (26) as well as at Ser³⁷⁶ or Ser³⁷⁸ (27) have been reported. Three of these modifications (Ser³³, Ser³¹⁵, and Ser³⁹²) were observed in our experiments. The role of these modifications has not been clearly established. It has been demonstrated that, in vitro, phosphorylation of Ser³⁹² stimulated formation of p53 tetramers, whereas phosphorylation of Ser³¹⁵ reversed it (28). p53 binds to large T antigen as a monomer (29). Does a balance between the two modifications lead to a tendency toward a p53 monomer formation in COS-1 cells? Ser¹⁵ phosphorylation along with Thr¹⁸, Ser²⁰, and Ser³⁷ has been shown to lead to a conformational change in p53 that prevents its interaction with MDM2, thus inhibiting p53 ubiquitination and degradation. In our case, only Ser¹⁵ was detected as phosphorylated. We can hypothesize that the binding of large T antigen to p53 by itself could stabilize p53. Hence it has been proposed that large T antigen enhances the stability of p53 partly by complexing with MDM2 (30, 31). In our experiment, we identified large T antigen in the p53 pulldown assay but not MDM2. p53 N-terminal phosphorylation can also promote the binding of the acetyltransferases CBP1/p300 to p53 (3). Kishi et al. (32) suggested that phosphorylation of Ser³³ may stimulate acetylation of p53. Upon stress, p53 is known to be specifically acetylated at lysine residues (in positions 164, 370, 372, 373, 381, and 382) by CBP/p300 (6, 13), at Lys³²⁰ by p300/CBP-associated factor (PCAF) (7), at Lys³⁰⁵ by p300 (10), and at Lys¹²⁰ by the p19ARF/oncogene pathway (33, 34). It was also demonstrated that Tip60, Ada3, and Pin1 participate in the control of p53 acetylation (8, 35, 36). The effects of acetylation on p53 function are not totally clear, but they have been implicated in modulating p53 transcriptional activity and stability (37). Like the acetylation of lysine residues in histones, acetylation of p53 has been linked to gene transcription regulation. Knights et al. (38) showed that distinct p53 acetylation “cassettes” differentially influence gene expression patterns and cell survival versus death. Acetylation sites are also important to keep p53 from ubiquitination and subsequent degradation by MDM2 as p53 is ubiquitinated and acetylated on similar sites at the C terminus (39, 40), suggesting that these modifications may compete for the same residues. Importantly recent work by Tang et al. (14) described p53 acetylation as an indispensable event that destabilizes the p53-MDM2 interaction enabling p53 activation. In our study, lysines 319 and 357, which were observed to be acetylated, border the tetramerization domain of p53 (Fig. 4). The recent crystal structure of large T antigen complexed with p53 revealed a hexameric complex of large T antigen binding six p53 monomers (29), and one could speculate that the new acetylation sites that we report here might impact p53 conformation and eventually participate in the formation and stabilization of the complex with large T antigen.

In conclusion, our study confirmed that p53 is a highly acetylated protein and unveiled new acetylation sites. Given the increasingly reported importance of acetylation for the regulation of p53, and although further functional exploration will be needed, the new acetylation sites identified here open new perspectives for the refinement of p53 mechanism of regulation.