View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Biochimica et Biophysica Acta 1823 (2012) 2287–2296

Contents lists available at SciVerse ScienceDirect

Biochimica et Biophysica Acta

journal homepage: www.elsevier.com/locate/bbamcr

SIAH-mediated ubiquitination and degradation of acetyl-transferases regulate the p53 response and protein acetylation

Inna Grishina a, Katherina Debus a, Carmen García-Limones b, Constanze Schneider a, Amit Shresta a, Carlos García c, Marco A. Calzado b, M. Lienhard Schmitz a,⁎

a Department of Biochemistry, Medical Faculty, Friedrichstrasse 24, Justus Liebig University, Member of the German Center for Lung Research, 35392 Giessen, Germany b Instituto Maimónides de Investigación Biomédica de Córdoba (IMIBIC), Universidad de Córdoba, 14004 Córdoba, Spain c Centro de Biología Molecular Severo Ochoa, CSIC-Universidad Autónoma de Madrid, 28049 Madrid, Spain

article info abstract

Article history: Posttranslational modification of proteins by lysine acetylation regulates many biological processes ranging Received 2 August 2012 from signal transduction to chromatin compaction. Here we identify the acetyl-transferases CBP/p300, Received in revised form 26 September 2012 Tip60 and PCAF as new substrates for the E3 ligases SIAH1 and SIAH2. While CBP/p300 can undergo Accepted 30 September 2012 ubiquitin/proteasome-dependent degradation by SIAH1 and SIAH2, the two other acetyl-transferases are ex- Available online 6 October 2012 clusively degraded by SIAH2. Accordingly, SIAH-deficient cells show enhanced protein acetylation, thus re-

Keywords: vealing SIAH proteins as indirect regulators of the cellular acetylation status. Functional experiments show SIAH that Tip60/PCAF-mediated acetylation of the tumor suppressor p53 is antagonized by the p53 target Ubiquitin E3 ligase SIAH2 which mediates ubiquitin/proteasome-mediated degradation of both acetyl-transferases and conse- p53 quently diminishes p53 acetylation and transcriptional activity. The p53 kinase HIPK2 mediates hierarchical Protein acetylation phosphorylation of SIAH2 at 5 sites, which further boosts its activity as a ubiquitin E3 ligase for several sub- HIPK2 strates and therefore dampens the late p53 response. © 2012 Elsevier B.V. All rights reserved.

1. Introduction Transcription of the human SIAH1 and SIAH2 encoding can be induced by p53 [9–11]. On the other hand, SIAH proteins can modu- The attachment of K48-branched ubiquitin chains to target pro- late the p53 response by ubiquitin/proteasome-dependent degrada- teins leads to their proteasomal degradation and thus controls their tion of p53-regulating enzymes such as HBX or the serine/threonine turnover rate. Ubiquitination is an ATP-dependent process that re- kinase HIPK2 [12,13]. quires a complex machinery consisting of E1 ubiquitin-activating, Regulation of SIAH1 and SIAH2 protein function is achieved by E2 ubiquitin-conjugating and E3 ubiquitin-ligating enzymes [1]. several mechanisms including inducible expression, differential for- Ubiquitin E3 ligases fall into several groups including the largest mation of homodimeric or heterodimeric complexes and also by group of RING (Really Interesting New Gene) finger and RING phosphorylation. Phosphorylation of murine Siah2 at Thr24 and finger-related E3s [2]. The RING family of ubiquitin E3 ligases also Ser29 increases its ability to degrade its substrate PHD3 [14]. Further- comprises the SIAH (seven in absentia) family of E3 ligases. The more, phosphorylation of SIAH1 at Ser19 [11] or phosphorylation of encodes SIAH1 and the highly homologous SIAH2 SIAH2 at positions 26, 28 and 68 and further unknown sites [12] di- protein. In mice a duplication of Siah1 created the Siah1a and Siah1b minishes the association with HIPK2. alleles [3]. SIAH1 and SIAH2 proteins have a largely divergent The evolutionary conserved kinase HIPK2 has the ability to phos- N-terminal part, but are highly conserved in the RING domain and phorylate p53 at Ser46, thus facilitating the CBP (CREB binding C-terminal substrate-binding domain. The SIAH proteins interact protein)-mediated acetylation of p53 at Lys382 which in turn pro- with a large list of cytosolic and nuclear proteins including transcrip- motes p53-dependent [15]. Lys382 is only one of tional repressors such as tramtrack and nuclear receptor co-repressor nine different acetylation sites that have been described for the (NCoR), the coactivator BOB/OBF-1, the histone methyltransferase tumor suppressor p53. Acetylation of p53 is a versatile mechanism SUV39H1 and the transcription elongation factor ELL2 [4–7]. The to activate its function, as acetylated p53 prevents degradative long list of SIAH-regulated signaling proteins also implies SIAH pro- ubiquitination at the same lysines, inhibits the formation of repres- teins in a variety of biological processes such as the regulation of sive HDM2/HDMX complexes and allows the recruitment of cofactors the hypoxic response, Ras-mediated signaling and cell death [8]. to direct promoter-specific activation of p53 transcriptional activity [16]. The six lysine residues in the C-terminal regulatory domains of ⁎ Corresponding author. Tel.: +49 641 9947570; fax: +49 641 9947589. p53 are acetylated by the lysine (K)-acetyl-transferases (KATs) CBP E-mail address: [email protected] (M.L. Schmitz). and the highly homologous p300 protein. However, p53 is also

0167-4889/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.bbamcr.2012.09.011 2288 I. Grishina et al. / Biochimica et Biophysica Acta 1823 (2012) 2287–2296 acetylated by further KATs, as Lys320 in the tetramerization domain occurrence of nonspecific binding. After rotating for 4 h at 4 °C, the is acetylated by the PCAF (P300/CBP-associated factor) and Lys120 beads were washed 5 times in NP40 lysis buffer. The precipitated pro- is modified by Tip60 (Tat-interactive protein 60 kD) [17,18]. teins were eluted by boiling in 2× SDS sample buffer for several mi- Here we describe the identification of the KATs CBP/p300, Tip60 nutes and were then further analyzed by Western blotting. Analysis and PCAF as further substrates for the SIAH proteins. While CBP/ of protein acetylation was done upon addition of 1 μM of the HDAC p300 was degraded by both SIAH variants, Tip60 and PCAF were inhibitor trichostatin A (TSA) to lysis and washing buffers. Denaturing only degraded by SIAH2. Cells lacking the Siah1a/Siah2 genes have lysis was done either by directly adding 1× SDS sample buffer or by more acetylated proteins and functional assays revealed that an im- lysis of cells in guanidinium-HCl buffer, followed by purification of portant role of SIAH2 for dampening the late p53 response. Full ubiquitinated proteins under denaturing conditions on Ni-NTA induction of early p53 activity is ensured by HIPK2 mediated phos- sepharose as described [24]. phorylation and Tip60/PCAF-dependent acetylation of p53. The late p53 activity is restrained by HIPK2-mediated phosphorylation of 2.5. Reverse transcription (RT)-PCR SIAH2 which in turn augments its ability to degrade both KATs and thus leads to impaired p53 acetylation and transcriptional activity. While one portion of the cells was analyzed by Western blot anal- ysis for protein expression, the remaining cells were lysed and total 2. Materials and methods RNA was extracted using the RNAeasy kit (Qiagen). One microgram of RNA was used as a template for the production of cDNAs from 2.1. Plasmids Oligo (dT)20 primers using the Superscript first strand synthesis sys- tem (Invitrogen). The various cDNAs encoding CBP, Tip60 and PCAF The plasmids encoding Flag-SIAH2, Flag-SIAH2 RM, HA-SIAH1, were detected with specific primers in a PCR reaction between 28 HA-SIAH1 RM, His-Ubiquitin, Flag-p53 [12], p53-luciferase [19] as and 32 cycles; the PCR products were detected after electrophoresis well as HA-HIPK2, HA-p300 and Flag-CBP [20] have been described. on ethidium bromide-stained agarose gels. Expression vectors for His-tagged p53 [21] (Dr. Jin Won Cho, Seoul), HA-tagged Tip60 [22] (Dr. Saadi Khochbin, Grenoble) and HA-PCAF 2.6. Mass spectrometry [23] (Melanie Ott, Heidelberg) were kind gifts. The point mutated versions of Flag-SIAH2 were generated using the QuikChange® HEK293T cells transfected to express Flag-SIAH2 alone or combi- site-directed mutagenesis kit (Stratagene). nation with HA-HIPK2 were treated 12 h before lysis with 10 μMof MG132. Cells were harvested and collected by centrifugation, 2.2. Antibodies and reagents followed by lysis in IP buffer (50 mM Hepes pH 7.5, 50 mM NaCl, 1% (v/v) Triton X-100, 2 mM EDTA, 10 mM sodium fluoride, 0.5 mM Mouse monoclonal antibodies against Flag (M2) and Tubulin sodium orthovanadate, 10 μg/ml leupeptine, 10 μg/ml aprotinin, (Tub2.1) were purchased from Sigma. Antibodies recognizing HA 10 μg/ml aprotinin, 10 μg/ml leupeptin and 1 mM PMSF). The tagged (3F10) (Roche Applied Science), HA (Y-11), Actin (C-2) and CBP SIAH2 proteins were immunoprecipitated with anti-Flag M2 affinity (A-22) (Santa Cruz Biotechnology Inc.), ubiquitin (P4D1) and gel (Sigma Aldrich) and the precipitated proteins were eluted by boil- acetyl-lysine (#9441) (Cell Signaling Technology) were from the ing in 1.5× SDS sample buffer for 5 min. After separation of proteins indicated suppliers. Rabbit polyclonal antibodies recognizing phos- by SDS–PAGE, the protein bands were stained, followed by excision phorylated SIAH2 proteins were generated by PolyPeptide Group of the SIAH2 bands and further analysis by RP-LC–MS/MS. (Strasbourg); the anti-phospho-SIAH2 (S28) antibodies were previous- ly described [12]. Secondary HRP-coupled antibodies were obtained 2.7. Protein structure prediction from Dianova and MG132 was purchased from Sigma. Secondary Cy-3-conjugated antibody (115-167-003) was purchased from Dianova Modeling of the SIAH2 protein structure was performed by I-TASSER and FITC-conjugated antibody (sc-2359) was from Santa Cruz. [25,26] using the human sequence NM_005067.5. The modeled structures were classified on the basis of C-score, TM-score 2.3. Cell culture, transient transfections and luciferase assays and root mean square deviation (RMSD), standard parameters used to specify the closeness of the predicted model to the native structure. Sec- HEK293T, H1299, U2OS, control MEFs and Siah1a−/−/Siah2−/− ondary structure prediction of SIAH2 showed a C-score of −2.45, Exp. MEFs were grown in DMEM supplemented with 10% (v/v) FCS, TM-score of 0.43±0.14 and Exp. RMSD of 12.2±4.4. Validation of the 2mM L-glutamine and 1% (v/v) penicillin/streptomycin. HEK293T modeled SIAH2 structure was performed by superimposing it with the and H1299 cells were transfected with Rotifect as described [24].In crystal structure of SIAH1a (2AN6A). The study of the structural similar- all transfections total DNA amounts were kept equal by adding ity between proteins was performed using the structure alignment pro- empty vector. Cells were lysed in 200 μl NP-40 buffer (50 mM Tris– gram TM-align [27]. SIAH2 model has an RMSD to target of 1.91 and HCl (pH 8), 5 mM EDTA, 150 mM NaCl, 0.5% NP-40, 0.5 mM PMSF, TM-score of 0.547. The validated structure was used to predict the 10 μg/ml aprotinin, 10 μg/ml leupeptin) and cleared by centrifuga- localization of the different HIPK2-phosphorylated residues. Figures tion. Ten microliters of the supernatant was mixed with 10 μl of lucif- were produced using PyMOL (Version 1.3, Schrödinger, LLC, San erase buffer and bioluminescence was immediately measured for 10 s Diego, CA.). in a luminometer (Berthold DuoLumat LB 9501). 2.8. Indirect immunofluorescence staining 2.4. Cell lysis, co-immunoprecipitation and Western blotting U2OS cells were seeded on cover slips and transfected to express Native lysis was done upon addition of NP-40 buffer to the cell SIAH2 RM together with the various KATs as indicated. Cells were pellets and incubation for 20 min on ice. Lysates were cleared by cen- washed twice with PBS and fixed for 1 min with cold methanol/ trifugation and proteins were either further analyzed by SDS-PAGE aceton (1:1). After air-drying, the cells were rehydrated with PBS and Western blotting or by co-immunoprecipitation. Lysates were in- and blocking was done with 10% (v/v) goat serum (Sigma) in PBS cubated with 1 μg of antibody together with 25 μl of the protein A/G for 60 min at room temperature. The primary and secondary anti- PLUS agarose (Santa Cruz) that was blocked for 1 h at room temper- bodies were incubated in 1% (v/v) goat serum overnight at 4 °C or ature in phosphate-buffered saline (PBS)/1% (w/v) BSA to avoid the 2 h at room temperature, respectively. After washing the cells several I. Grishina et al. / Biochimica et Biophysica Acta 1823 (2012) 2287–2296 2289 times with PBS, nuclei were stained with Hoechst 33342 (Invitrogen) were analyzed for protein and mRNA levels of CBP (Fig. 1A). Western and cells were mounted with Kaiser's glycerol gelatine (Merck). Cells blotting showed efficient degradation of the CBP protein upon expres- were analyzed for intracellular localization of proteins with a Nikon sion of increasing amounts of SIAH1, while the inactive SIAH1 point mu- eclipse TE2000-E microscope. tant did not change CBP protein levels. SIAH1 expression remained without consequences for the CBP mRNA levels, as revealed by RT-PCR. 3. Results Some SIAH substrates such as HIPK2 can be degraded by both SIAH isoforms; therefore it was then interesting to test whether also SIAH2 3.1. SIAH1 and SIAH2 degrade CBP and its related protein p300 has the ability to cause CBP degradation. Coexpression of SIAH2 also trig- gered CBP degradation, while a SIAH2 RING domain point mutant failed As SIAH proteins have the ability to degrade HIPK2 [11,12,28],we to affect CBP levels (Fig. 1B). Similar to SIAH1, also expression of SIAH2 were interested to test whether SIAH proteins also destabilize HIPK2- remained without any impact on CBP mRNA levels. To investigate interacting proteins such as CBP which constitutively binds and acety- whether SIAH-mediated CBP degradation is dependent on the lates HIPK2 [20]. To address this question, 293T cells were transfected proteasome, CBP and both SIAH forms were coexpressed in the absence to express CBP along with increasing amounts of SIAH1 or a mutant or in the presence of the proteasome inhibitor MG132. Subsequent im- thereof which is mutated in a catalytically important contained munoblotting showed that SIAH-mediated CBP degradation was largely in the RING domain (SIAH1 RM). Two days later cells were harvested and blocked in the presence of MG132 (Fig. 1C), suggesting that this

A B Flag-CBP Flag-CBP HA-SIAH1 Flag-SIAH2 HA-SIAH1 RM Flag-SIAH2 RM

CBP Flag CBP Flag Flag SIAH2 HA SIAH1 Actin Actin Actin Actin

M M

RT-PCR CBP RT-PCR CBP

CDFlag-CBP HA-SIAH1 Flag-SIAH2 MG132 Ubi*CBP Flag CBP CBP

HA SIAH1 Ni-NTA CBP

Flag SIAH2 Ubiquitin Ubiquitin Actin Actin

Flag-CBP His-Ubiquitin E HA-p300 HA-SIAH1 HA-SIAH1 Flag-SIAH2 Flag-SIAH2 HA SIAH1 HA p300 Flag SIAH2 HA SIAH1 Input CBP CBP Flag SIAH2

Actin Actin Actin Actin

Fig. 1. SIAH-mediated ubiquitination and degradation of CBP and p300. (A) 293T cells transfected to express Flag-CBP along with HA-tagged versions of SIAH1 or a SIAH1 version where Cys 44 in the RING domain was mutated to Ser (SIAH1 RM). Two days later cells were harvested and two thirds of the cells were used for protein extraction and analysis of protein expression by immunoblotting (upper). The position of a non-specific band is shown by an asterisk. RNA was extracted from the remaining cells, followed by cDNA syn- thesis and detection of CBP by RT-PCR with specific primers. The PCR products were detected after electrophoresis on an agarose gel and ethidium bromide-staining, the position of a DNA marker (M) is indicated. (B) The experiment was done as in (A) with the exception that SIAH2 and a RING mutant thereof were used instead of SIAH1. (C) Cells were transfected to express Flag tagged CBP together with SIAH1 or SIAH2. The next day, MG132 (5 μM) or solvent control was added and further 20 h later the cells were lysed and tested for protein expression with the indicated antibodies. (D) Flag-CBP was coexpressed with His-tagged ubiquitin and SIAH1 or SIAH2 as shown and cells were further grown in the presence of 5 μM MG132. One fraction was lysed and used as an input control (lower) while the remaining fraction was lysed under denaturing conditions in guanidinium hydrochloride buffer, followed by purification of ubiquitinated proteins on Ni-NTA columns. Following extended electrophoresis on a 7% SDS gel and Western blotting, CBP and its ubiquitinated forms were revealed by CBP antibodies. (E) 293T cells were transfected to expresses HA-p300 along or together with increasing amounts of SIAH1 or SIAH2. Two days later cells were harvested and protein expression was analyzed by immunoblotting with the indicated antibodies. 2290 I. Grishina et al. / Biochimica et Biophysica Acta 1823 (2012) 2287–2296

ABHA-Tip60 HA-PCAF HA-SIAH1 HA-SIAH1 HA-SIAH1 RM HA-SIAH1 RM Flag-SIAH2 Flag-SIAH2 Flag-SIAH2 RM Flag-SIAH2 RM

HA Tip60 HA PCAF

HA SIAH1 HA SIAH1

Flag SIAH2 Flag SIAH2

Actin Actin Actin Actin M M

RT-PCR Tip60 RT-PCR PCAF

C HA-PCAF D HA-Tip60 Flag-SIAH2 MG132 Ni-NTA HA Ubi*PCAF

PCAF HA Tip60 HA-PCAF Flag-SIAH2 His-Ubiquitin Flag SIAH2 MG132

HA PCAF Actin Actin

Flag SIAH2 E

Input Ni-NTA HA Ubi*Tip60 Ubiquitin Ubiquitin

HA-Tip60 Flag-SIAH2 Actin Actin His-Ubiquitin MG132

HA Tip60

Flag SIAH2

Input Ubiquitin Ubiquitin

Actin Actin

Fig. 2. Tip60 and PCAF are degraded by SIAH2-mediated ubiquitination and proteasomal degradation. (A) Cells were transfected to express HA-Tip60 together with wildtype SIAH1, SIAH2 or mutants thereof which are mutated in their RING domains. Cells were analyzed either for protein expression (upper) or for Tip60 mRNA levels by RT-PCR and agarose gel electrophoresis as shown. (B) The experimental setting resembles that of (A) with the exception that a PCAF was expressed instead of Tip60. (C) Tagged versions of PCAF or Tip60 were expressed either alone or together with SIAH2 and further grown in the absence or presence of MG132 as shown. Cell lysates were tested by immunoblotting for expression of the KATs or SIAH2 as shown. (D) Cells transfected to express HA-PCAF along with SIAH2 and His-tagged ubiquitin in the presence or absence of MG132 as shown. While one fraction of the cells was used for the input controls, the remaining cells were lysed and His-tagged proteins were enriched under denaturing conditions. Ubiquitination of PCAF was revealed with an anti-HA antibody. (E) The experiment was done as in (D) with the exception that the ubiquitination of Tip60 was investigated. degradation process proceeds via the proteasome. The occurrence of revealed by immunoblotting with specific antibodies (Fig. 1D). These ex- SIAH-mediated CBP ubiquitination was then tested upon coexpression periments showed ubiquitination of CBP by both SIAH isoforms and a of CBP and the SIAH isoforms together with hexahistidine-tagged higher ubiquitinating activity of SIAH2. To test whether SIAH-proteins ubiquitin. After enrichment of ubiquitinated proteins under denaturing can also trigger degradation of the CBP-related protein p300, 293T cells conditions on Ni-NTA columns, the ubiquitinated forms of CBP were were transfected to express p300 alone or with increasing amount of I. Grishina et al. / Biochimica et Biophysica Acta 1823 (2012) 2287–2296 2291

A IP: Flag

PCAF HA Tip60

Flag SIAH2

Flag-SIAH2 RM HA-Tip60 HA-PCAF

PCAF HA Tip60 Input Flag SIAH2

Tubulin Tubulin

B Flag HA Merge Hoechst

Flag-SIAH2 RM & HA-CBP

SIAH2 CBP

Flag-SIAH2 RM & HA-p300

SIAH2 p300

Flag-SIAH2 RM & HA-PCAF

SIAH2 PCAF

Flag-SIAH2 RM & HA-Tip60

SIAH2 Tip60

Fig. 3. SIAH2 binds to PCAF and Tip60. (A) 293T cells were transfected to express PCAF or Tip60 either alone or together with Flag-SIAH2 RM. Cell lysates were prepared after 36 h and one fraction was used for the input control, while another fraction was used for immunoprecipitation of SIAH2 with anti-Flag antibodies. Coprecipitating PCAF and Tip60 proteins were detected with anti-HA antibodies; no signals were detected when the immunoprecipitation was done with isotype-matched control antibodies (data not shown). (B) U2OS cells express- ing the indicated HA-tagged KATs and Flag-SIAH2-RM were analyzed by indirect immunofluorescence for the intracellular localization of proteins. Overlapping localization is shown in yellow, DNA staining with Hoechst was used to visualize the nuclei.

SIAH1 and SIAH2. Two days later cells were harvested and analyzed for 3.2. Tip60 and PCAF are exclusively degraded by SIAH2 p300 protein levels by immunoblotting (Fig. 1E). Similar to the CBP pro- tein p300 was efficiently degraded upon expression of increasing As HIPK2 can also associate with Tip60 [29] we tested a potential amounts of SIAH1 and also SIAH2 proteins. influence of SIAH1 and SIAH2 on Tip60 protein levels. Coexpression 2292 I. Grishina et al. / Biochimica et Biophysica Acta 1823 (2012) 2287–2296

ABIP: Flag His-p53 Flag-p53 HA-Tip60 HA-Tip60 HA-PCAF HA-PCAF Flag-SIAH2

Ac-Lys Ac-p53 Ac-Lys Ac-p53 Ni-NTA Flag p53 p53 p53

PCAF PCAF HA HA Tip60 Tip60 Input Input Flag SIAH2 Flag p53 Actin Actin Actin Actin

C D IP: Ac-Lys

p53-Luc 250 130 95 72

55

Ac-Lys Fold Induction 36

28

Flag-p53 HA-Tip60 17 HA-PCAF Flag-SIAH2 SIAH1a-/- wt MEFs Flag-SIAH2 RM SIAH2-/-

Fig. 4. SIAH controls acetylation of p53 and further proteins. (A) Cells expressing Flag-p53 and increasing amounts of PCAF or Tip60 were lysed in a buffer containing the HDAC inhibitor TSA (1 μM), followed by immunoprecipitation of p53 and analysis of p53 acetylation using an antibody recognizing acetylated lysines. (B) Cells expressing His-tagged p53 along with Tip60, PCAF and SIAH2 were lysed under denaturing conditions in order to block HDAC activity, followed by purification of His-p53 on Ni-NTA columns and analysis of p53 expression and its acetylation with specific antibodies. (C) H1299 cells were transfected to express p53 along with expression vectors encoding Tip60 or PCAF, SIAH2 and a p53-dependent luciferase reporter gene. After 36 h, cell lysates were analyzed for luciferase activity. Gene expression by the empty vector was arbitrarily set as 1; error bars reflect standard deviations from two independent experiments performed in triplicates. (D) The indicated knockout MEFs lacking Siah1a and Siah2 genes and wildtype controls were treat- ed for 12 h with 1 μM TSA to enhance protein acetylation, followed by preparation of cell lysates, immunoprecipitation of acetylated proteins and their detection by immunoblot- ting using pan-acetyl-lysine-specific antibodies. The positions of molecular weight marker proteins are shown, differentially acetylated proteins are highlighted by arrows. of SIAH2 strongly diminished Tip60 protein levels, while SIAH1 and PCAF in samples of immunoprecipitated SIAH2 RM (Fig. 3A), thus re- remained without any influence. As SIAH2 did not affect Tip60 mRNA vealing mutual interaction between this ubiquitin E3 ligase and its sub- levels (Fig. 2A), these data show that Tip60 can be diminished selectively strate proteins. Subsequently we compared the intracellular localization by SIAH2. As HIPK2 contributes to PCAF-mediated p53 acetylation [30] of the acetyltransferases with that of the SIAH2 protein by indirect immu- we extended the search for new SIAH substrates also to PCAF. Strikingly, nofluorescence. As the expression of endogenous SIAH2 is extremely low, also the levels of the PCAF protein were strongly diminished in the pres- cells were transfected to express the HA-tagged acetyltransferases to- ence of SIAH2 (Fig. 2B), but similar to Tip60 also PCAF was selectively gether with Flag-tagged SIAH2-RM in order to avoid degradation of degraded by SIAH2. The SIAH2-mediated degradation of these two interacting proteins. While all four KATs were found exclusively in the KATs was largely abrogated in the presence of MG132 (Fig. 2C), showing nucleus, SIAH2 showed a mainly perinuclear distribution in the cytosol that degradation of PCAF and Tip60 proceeds via the proteasome. Ac- and only a minor fraction localized to the nucleus (Fig. 3B). These data cordingly, SIAH2 expression promoted ubiquitination of PCAF (Fig. 2D) suggest that only a small fraction of the KATs colocalizes and interacts and Tip60 (Fig. 2E), as revealed by coexpression of His-tagged ubiquitin with SIAH2, thus ensuring that not all proteins are degraded. and Ni-NTA pull-down experiments. Many ubiquitin E3 ligases such as SIAH2 can (at least transiently) 3.3. SIAH2-mediated control of KATs affects the acetylation status of p53 interact with their substrate proteins [12]. It was therefore relevant to and further proteins investigate the occurrence of SIAH2 binding to its substrate proteins Tip60 and PCAF. To address this question, Tip60 and PCAF were expressed The p53 tumor suppressor has multiple important functions in the either alone or together with the SIAH2 RING domain mutant protein cell, raising the need to control and specify its individual functions by SIAH2 RM, which is unable to degrade its interaction partners. additional mechanisms such as acetylation. To compare the abilities Co-immunoprecipitation experiments showed the occurrence of Tip60 of Tip60 and PCAF to acetylate p53, cells were transfected to express I. Grishina et al. / Biochimica et Biophysica Acta 1823 (2012) 2287–2296 2293

Flag-SIAH2 WT S16A T26A S28AS68A T119A HA-HIPK2

Flag P -SIAH2 SIAH2

P -SIAH2 Thr26 P -SIAH2 Thr26

P -SIAH2 Ser28 P -SIAH2 Ser28

P -SIAH2 Ser68 P -SIAH2 Ser68

P -SIAH2 Thr119 P -SIAH2 Thr119 *

HA HIPK2

Actin Actin

Fig. 5. Functional characterization of HIPK2-mediated SIAH2 phosphorylation sites. The indicated expression plasmids for Flag-SIAH2 and its phosphorylation site mutants were expressed alone or together with HA-HIPK2 in 293T cells. Cell lysates were analyzed by immunoblotting with the indicated phospho-specific antibodies for the occurrence of site-specific SIAH2 phosphorylations. The position of a molecular weight marker is given in kDa; the asterisk highlights a non-specific band. p53 along with increasing amounts of Tip60 and PCAF, respectively. However, these phosphorylation sites were incompletely mapped, Following immunoprecipitation of p53, Western blotting with an an- raising the need to identify more SIAH2 phosphorylation sites using tibody recognizing acetylated lysines showed that PCAF and Tip60 an unbiased approach. Towards this goal SIAH2 was expressed either acetylate p53 to a comparable extent (Fig. 4A). In a next step it was alone or together with HIPK2, followed by immunoprecipitation of then interesting to test whether PCAF or Tip60-induced p53 acetyla- SIAH2 and denaturing SDS-PAGE. Bands representing SIAH2 were ex- tion can be regulated by SIAH2. Coexpression of this E3 ligase largely cised and phosphorylation sites were identified by mass spectrometry abrogated PCAF- and Tip60-mediated p53 acetylation (Fig. 4B), thus (supplementary Fig. 1). This analysis confirmed the published SIAH2 identifying SIAH2 as an additional protein that can influence the modifications at residues 26, 28 and 68, but also allowed the identifi- p53 acetylation status, albeit in an indirect manner. To determine cation of Ser16 and Thr119 as new HIPK2-dependent phosphoryla- the impact of SIAH2 on p53 acetylation-dependent transcription, tion sites (supplementary Fig. 2). HIPK2 phosphorylation sites are luciferase assays were performed (Fig. 4C). Expression vectors for typically flanked by a , which is also the case for four (Thr26, p53, both KATs and active/inactive SIAH2 were transfected into Ser28, Ser68 and Thr119) out of the five phosphosites identified p53-deficient H1299 cells along with a p53-dependent luciferase here. To confirm HIPK2-mediated phosphorylation at these residues gene. While induction of gene expression by p53 was further aug- by an independent experimental approach, specific SIAH2 antibodies mented by Tip60 or PCAF, coexpression of SIAH2 turned gene expres- recognizing phosphorylated Ser16, Thr26, Ser68 and Thr119 were sion down to the levels reached upon expression of p53 alone. This raised. While the phospho-Ser28 antibody was previously described effect was strictly dependent on the integrity of the SIAH2 RING do- [12], only the quality of the newly produced antibodies recognizing main, showing that the ubiquitinating activity of SIAH2 is required phosphorylated Thr26, Ser68 and Thr119 was sufficient to use them to dampen p53 acetylation and thus its transcriptional activity. Col- for further experiments. To test the occurrence of SIAH2 phosphoryla- lectively these experiments show that SIAH E3 ligases control the sta- tion, the E3 ligase was expressed either alone or together with HIPK2 bility of several KATs, raising the possibility that they might have an and modification of SIAH2 was determined with the different impact on the general acetylation status of the cell. To address this modification-specific antibodies. These experiments showed HIPK2- question we compared the acetylation status of MEFs lacking the mediated phosphorylation of SIAH2 at all 4 different sites (Fig. 5). Siah2 and Siah1a genes to that of control wildtype MEFs. Cells were Mutation of the phosphorylated amino acids to alanine precluded rec- treated for 12 h with the broad-spectrum HDAC inhibitor trichostatin ognition by the respective phospho-specific antibodies, thus ensuring A(TSA),followedbyenrichmentofacetylatedproteinsbyimmunopre- their specificity. The absent reactivity of the SIAH2 phospho-Thr26 anti- cipitation with acetyl-specific antibodies and immunoblotting. Detec- body towards the SIAH2 S28A mutant is most probably due to tion of acetylated proteins with pan-acetyl-lysine-specificantibodies overlapping epitopes, but interestingly some other phosphorylations oc- showed the occurrence of several additional bands representing acety- curred hierarchically (Fig. 5). SIAH2 phosphorylation at Ser68 was con- lated proteins in the Siah1a−/−/Siah2−/− cells (Fig. 4D), thus revealing stitutive and occurred also in the absence of overexpressed HIPK2. that SIAH proteins can also indirectly control the acetylome. Phosphorylation at Ser68 and Thr26 was absent after mutation of Thr119 to Ala, suggesting that phosphorylation of Thr119 is required 3.4. HIPK2-mediated SIAH2 phosphorylation augments its activity as a for Ser68 modification. Given the hierarchical occurrence of HIPK2- ubiquitin E3 ligase mediated phosphorylations on SIAH2 it was then interesting to visualize the relative localizations of the phosphorylation sites. As the structure of The activity of SIAH2 proteins can be regulated by various mecha- SIAH2 is not known we generated a three-dimensional model using the nisms including phosphorylation by the p53 kinase HIPK2 [12]. iterative threading assembly refinement (I-TASSER) procedure [25,26]. 2294 I. Grishina et al. / Biochimica et Biophysica Acta 1823 (2012) 2287–2296 A

SIAH2 SIAH1a SIAH2

SER-28 B SER-68 THR-26 SER-28 SER-16 THR-26

THR-119 THR-119

C THR-119

SER-28 THR-26

Fig. 6. Structure prediction for SIAH2. (A) Superposition of the crystal structure of SIAH1a (red) (residues 92–282, PDB entry 2AN6A) and the computationally predicted SIAH2 model (green). The study of the structural similarity between proteins was performed using the structure alignment program TM-align. SIAH2 model has an RMSD to target of 0.79 and TM-score of 0.585. The figure was produced using PyMOL (Version 1.3, Schrödinger, LLC, San Diego, CA.). The left part shows SIAH1a and SIAH2 as ribbon diagrams; the right part shows the ribbon diagram of SIAH2 with a superimposed space-filling model. (B) Space-filling models of SIAH2 are shown from different angles; the phosphorylated amino acids are highlighted in red. (C) Close-up of the patch containing phosphorylated Thr26, Ser28 and Thr119.

Alignment and comparison with the published crystal structure of HIPK2. A comparison between the wildtype and phosphomimetic SIAH1a (PDB 2AN6A) [31] showed the structural conservation of SIAH2 proteins showed that the increased activity of SIAH2 5D also oc- the regions forming the substrate-binding domain (Fig. 6A). All 5 curred for HIPK2 (Fig. 7C), providing a possible mechanism how exces- phosphorylation sites are not found in the region known to be direct- sive SIAH2 phosphorylation is prevented. ly involved in direct substrate recognition [32], but 4 of them (Ser16, Thr26, Ser28 and Thr119) are found on one surface of SIAH2 4. Discussion (Fig. 6B). Three phosphorylated amino acids (Thr26, Ser28 and Thr119) localize in close vicinity and form a narrow patch that Here we identify SIAH proteins as ubiquitin E3 ligases targeting the upon phosphorylation would be strongly negatively charged. The KATs CBP/p300, PCAF and Tip60. The KATs neither contain the consen- priming phosphorylation site Thr119 is predicted to be at an exposed sus sequence RPVAxVxPxxR which is contained in some of the SIAH position and protrudes from the surface (Fig. 6C). To investigate the substrate proteins [33] nor display a significant degree of sequence ho- consequences of SIAH2 phosphorylation for its ability to degrade mologies. The KATs investigated here are typically contained in large substrate proteins, 293T cells were transfected to express Tip60 multisubunit complexes [34]. While Tip60 is one of the 18 subunits of (Fig. 7A) and PCAF (Fig. 7B)alongwithincreasingamountsof the NuA4 complex [35], PCAF associates with further subunits to form SIAH2 wildtype, and SIAH2 mutants where all 5 phosphorylation the SAGA or the ATAC complexes [36]. It is presently not clear whether sites have been either mutated to alanine (SIAH2 5A) or in a SIAH proteins degrade free KATs outside from these complexes or phosphorylation-mimicking fashion to aspartic acid (SIAH2 5D). whether SIAH can be transiently recruited as a non-stoichiometric com- Subsequent analysis of Tip60 and PCAF levels showed an augmented ponent to these complexes. In any case, SIAH-mediated degradation of a ability of the phosphomimetic SIAH2 5D mutant to degrade the KATs fraction of the KATs will counteract acetylation. It will be very interest- (Fig. 7A, B), thus revealing that HIPK2-mediated phosphorylation of ing to see whether also other components of these multi-protein com- SIAH2 stimulates its activity as an E3 ligase. It was then interesting to plexes can be targeted by SIAH proteins. Although SIAH proteins are see whether SIAH2 5D also has an increased ability to degrade its kinase primarily found in the cytoplasm, a fraction localizes to the nucleus to I. Grishina et al. / Biochimica et Biophysica Acta 1823 (2012) 2287–2296 2295

ABHA-Tip60 HA-PCAF Flag-SIAH2 Flag-SIAH2 Flag-SIAH2 5A Flag-SIAH2 5A Flag-SIAH2 5D Flag-SIAH2 5D

HA Tip60 HA PCAF

Flag SIAH2 Flag SIAH2

Actin Actin Actin Actin

C Flag-HIPK2 Flag-SIAH2 Flag-SIAH2 5D

Flag HIPK2

HA SIAH2

Actin Actin

D SIAH2 P P

biquitin U biquitin -- U biquitin -- U HIPK2 HIPK2 -- Tip60 PCAF Tip60 PCAF Proteasome

Ac P Ac SIAH2 p53 p53 p53 p53

induction phase termination phase

Fig. 7. Phosphorylation of SIAH2 augments its ability to degrade KATs. (A) Tip60 was expressed alone or together with increasing amounts of SIAH2 wildtype, SIAH2 5A or the phosphomimetic SIAH2 5D protein as shown. Two days later cells were lysed and the abundance of proteins was checked by immunoblotting as shown. (B) The experiment was done as in (A) with the difference that the various SIAH2 mutants were compared for their ability to degrade PCAF. (C) 293T cells were transfected to express HIPK2 alone or together with increasing amounts of SIAH2 wildtype or SIAH2 5D as shown. Equal amounts of proteins contained in cell lysates were tested for protein expression with the in- dicated antibodies. (D) Schematic summary of the data. Initial p53 activation is accompanied by HIPK2-mediated p53 phosphorylation and KAT-dependent p53 acetylation. Follow- ing expression of p53 target genes including SIAH2, this E3 ligase helps to degrade HIPK2 [19] and Tip60/PCAF in order to dampen the late p53 response. degrade its nuclear targets including the corepressor NCoR [37]. The re- of the enzymes mediating p53 deacetylation is Sirt1, which can be pressive NCoR complex silences the activity of unliganded nuclear re- stabilized by the ubiquitin-specific peptidase USP22, one of the 11 ceptors and is tightly associated with HDAC3 activity, thus leading to death-from-cancer signature genes [43]. USP22-mediated stabiliza- local histone deacetylation. Also HDAC3 can be degraded by SIAH2 tion of Sirt1 results in decreased levels of p53 acetylation and conse- [38], raising the fascinating scenario that SIAH proteins can govern the quently in suppression of p53-mediated functions. Another example general acetylation status by controlling the stabilities of acetylating comes from the tumor suppressor WTX, which frequently inactivated and deacetylating protein complexes. These findings also imply that in- in Wilms tumor. Recent evidence shows that WTX enhances binding creased expression of SIAH proteins which is seen in breast cancer [9] or of p53 to CBP, consequently increasing CBP-mediated p53 acetylation the increased nuclear localization of SIAH as it occurs in human hepato- and thus increasing p53 function [44]. cellular carcinomas [39] may well have a general impact on protein In conclusion our data suggest that full p53 activation is followed acetylation and chromatin compaction. Interestingly, the CBP/p300 pro- by the increased synthesis of SIAH proteins. Hierarchical phosphory- teins can be degraded by both SIAH forms, while Tip60 and PCAF are lation of SIAH2 by the p53 kinase HIPK2 further augments its function only sensitive to SIAH2. The molecular basis for this is not understood, as a ubiquitin E3 ligase and will then lead to degradation of the KATs but it seems plausible that differential dimerization and the formation and impaired p53 acetylation. As SIAH proteins can also degrade the of SIAH2 homodimers accounts for this specificity [40]. p53 kinase HIPK2 [12], also the critical p53 Ser46 phosphorylation As SIAH1 and SIAH2 are well-known p53 target genes [10,11,41], will be lost, as schematically depicted in Fig. 7D. These mechanisms the induced expression of these E3 ligases will have an impact on will contribute to dampen the late p53 response and thus help to the acetylation status of p53. After induction of the early p53 re- avoid prolonged and exaggerated p53 activity. sponse under adverse conditions such as genotoxic stress, the p53 protein can be acetylated which activates DNA-binding and transcrip- Funding tional activity of p53 [16]. In addition, acetylated p53 is refractory to ubiquitination, which results in the stabilization of this tumor sup- This work was supported by grants from the German Research pressor [42]. Given the relevance of p53 acetylation for its function, Foundation project SCHM 1417/7-1, TRR81 and the international the acetylation status is regulated by a variety of mechanisms. One graduate program PROMISE (IRTG1566) to M.L.S. and by MICINN 2296 I. Grishina et al. / Biochimica et Biophysica Acta 1823 (2012) 2287–2296

[SAF2010-17122] and Consejería de Salud (Junta de Andalucía) [19] M.A. Calzado, L. de la Vega, E. Munoz, M.L. Schmitz, Autoregulatory control of the p53 response by Siah-1L-mediated HIPK2 degradation, Biol. Chem. 390 (2009) [PI-0650-2010] grants to M.A.C. 1079–1083. [20] L. de la Vega, I. Grishina, R. Moreno, M. Kruger, T. Braun, M.L. Schmitz, A redox- regulated SUMO/acetylation switch of HIPK2 controls the survival threshold to Acknowledgments oxidative stress, Mol. Cell 46 (2012) 472–483. [21] W.H. Yang, J.E. Kim, H.W. Nam, J.W. Ju, H.S. Kim, Y.S. Kim, J.W. Cho, Modification We are grateful to Daniela Stock and Stephanie Steitz for technical of p53 with O-linked N-acetylglucosamine regulates p53 activity and stability, – assistance. We thank Dr. David Bowtell for SIAH knockout MEFs and Nat. Cell Biol. 8 (2006) 1074 1083. [22] E. Col, C. Caron, C. Chable-Bessia, G. Legube, S. Gazzeri, Y. Komatsu, M. Yoshida, M. Dres. Jin Won Cho (Seoul), Saadi Khochbin (Grenoble) and Melanie Benkirane, D. Trouche, S. Khochbin, HIV-1 Tat targets Tip60 to impair the apopto- Ott (Heidelberg) for sending plasmids. tic cell response to genotoxic stresses, EMBO J. 24 (2005) 2634–2645. [23] A. Dorr, V. Kiermer, A. Pedal, H.R. Rackwitz, P. Henklein, U. Schubert, M.M. Zhou, E. Verdin, M. Ott, Transcriptional synergy between Tat and PCAF is dependent on Appendix A. Supplementary data the binding of acetylated Tat to the PCAF bromodomain, EMBO J. 21 (2002) 2715–2723. [24] F. Renner, V.V. Saul, A. Pagenstecher, T. Wittwer, M.L. Schmitz, Inducible SUMO Supplementary data to this article can be found online at http:// modification of TANK alleviates its repression of TLR7 signalling, EMBO Rep. 12 dx.doi.org/10.1016/j.bbamcr.2012.09.011. (2011) 129–135. [25] A. Roy, A. Kucukural, Y. Zhang, I-TASSER: a unified platform for automated pro- tein structure and function prediction, Nat. Protoc. 5 (2010) 725–738. References [26] Y. Zhang, I-TASSER server for protein 3D structure prediction, BMC Bioinformatics 9 (2008) 40. [1] D. Vucic, V.M. Dixit, I.E. Wertz, Ubiquitylation in : a post-translational [27] Y. Zhang, J. Skolnick, TM-align: a protein structure alignment algorithm based on modification at the edge of life and death, Nat. Rev. Mol. Cell Biol. 12 (2011) the TM-score, Nucleic Acids Res. 33 (2005) 2302–2309. 439–452. [28] S.Y. Kim, D.W. Choi, E.A. Kim, C.Y. Choi, Stabilization of HIPK2 by escape from [2] S. Lipkowitz, A.M. Weissman, RINGs of good and evil: RING finger ubiquitin proteasomal degradation mediated by the E3 Siah1, Cancer Lett. ligases at the crossroads of tumour suppression and oncogenesis, Nat. Rev. Cancer 279 (2009) 177–184. 11 (2011) 629–643. [29] Q. Li, S. Lin, X. Wang, G. Lian, Z. Lu, H. Guo, K. Ruan, Y. Wang, Z. Ye, J. Han, et al., [3] N.G. Della, P.V. Senior, D.D. Bowtell, Isolation and characterisation of murine ho- Axin determines cell fate by controlling the p53 activation threshold after DNA mologues of the Drosophila seven in absentia gene (sina), Development 117 damage, Nat. Cell Biol. 11 (2009) 1128–1134. (1993) 1333–1343. [30] V. Di Stefano, S. Soddu, A. Sacchi, G. D'Orazi, HIPK2 contributes to PCAF-mediated [4] J. Boehm, Y. He, A. Greiner, L. Staudt, T. Wirth, Regulation of BOB.1/OBF.1 stability p53 acetylation and selective transactivation of p21Waf1 after nonapoptotic DNA by SIAH, EMBO J. 20 (2001) 4153–4162. damage, 24 (2005) 5431–5442. [5] J. Frasor, J.M. Danes, C.C. Funk, B.S. Katzenellenbogen, Estrogen down-regulation [31] G. Polekhina, C.M. House, N. Traficante, J.P. Mackay, F. Relaix, D.A. Sassoon, M.W. of the corepressor N-CoR: mechanism and implications for estrogen derepression Parker, D.D. Bowtell, Siah ubiquitin ligase is structurally related to TRAF and mod- of N-CoR-regulated genes, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 13153–13157. ulates TNF-alpha signaling, Nat. Struct. Biol. 9 (2002) 68–75. [6] M. Liu, J. Hsu, C. Chan, Z. Li, Q. Zhou, The ubiquitin ligase Siah1 controls ELL2 sta- [32] C.M. House, N.C. Hancock, A. Moller, B.A. Cromer, V. Fedorov, D.D. Bowtell, M.W. bility and formation of super elongation complexes to modulate gene transcrip- Parker, G. Polekhina, Elucidation of the substrate binding site of Siah ubiquitin li- tion, Mol. Cell 46 (2012) 325–334. gase, Structure 14 (2006) 695–701. [7] R. Tiedt, B.A. Bartholdy, G. Matthias, J.W. Newell, P. Matthias, The RING finger pro- [33] C.M. House, I.J. Frew, H.L. Huang, G. Wiche, N. Traficante, E. Nice, B. Catimel, D.D. tein Siah-1 regulates the level of the transcriptional coactivator OBF-1, EMBO J. 20 Bowtell, A binding motif for Siah ubiquitin ligase, Proc. Natl. Acad. Sci. U. S. A. 100 (2001) 4143–4152. (2003) 3101–3106. [8] C.M. House, A. Moller, D.D. Bowtell, Siah proteins: novel drug targets in the Ras [34] K. Anamika, A.R. Krebs, J. Thompson, O. Poch, D. Devys, L. Tora, Lessons from and hypoxia pathways, Cancer Res. 69 (2009) 8835–8838. genome-wide studies: an integrated definition of the coactivator function of his- [9] P. Chan, A. Moller, M.C. Liu, J.E. Sceneay, C.S. Wong, N. Waddell, K.T. Huang, A. tone acetyl transferases, Epigenetics Chromatin 3 (2010) 18. Dobrovic, E.K. Millar, S.A. O'Toole, et al., The expression of the ubiquitin ligase [35] V. Sapountzi, I.R. Logan, C.N. Robson, Cellular functions of TIP60, Int. J. Biochem. SIAH2 (seven in absentia homolog 2) is mediated through gene copy number in Cell Biol. 38 (2006) 1496–1509. breast cancer and is associated with a basal-like phenotype and p53 expression, [36] Z. Nagy, A. Riss, S. Fujiyama, A. Krebs, M. Orpinell, P. Jansen, A. Cohen, H.G. Breast Cancer Res. 13 (2011) R19. Stunnenberg, S. Kato, L. Tora, The metazoan ATAC and SAGA coactivator HAT [10] S. Matsuzawa, S. Takayama, B.A. Froesch, J.M. Zapata, J.C. Reed, p53-inducible complexes regulate different sets of inducible target genes, Cell. Mol. Life Sci. 67 human homologue of Drosophila seven in absentia (Siah) inhibits : (2010) 611–628. suppression by BAG-1, EMBO J. 17 (1998) 2736–2747. [37] J. Zhang, M.G. Guenther, R.W. Carthew, M.A. Lazar, Proteasomal regulation of [11] M. Winter, D. Sombroek, I. Dauth, J. Moehlenbrink, K. Scheuermann, J. Crone, T.G. nuclear receptor corepressor-mediated repression, Genes Dev. 12 (1998) Hofmann, Control of HIPK2 stability by ubiquitin ligase Siah-1 and checkpoint ki- 1775–1780. nases ATM and ATR, Nat. Cell Biol. 10 (2008) 812–824. [38] H.L. Zhao, N. Ueki, M.J. Hayman, The Ski protein negatively regulates Siah2- [12] M.A. Calzado, L. de la Vega, A. Moller, D.D. Bowtell, M.L. Schmitz, An inducible mediated HDAC3 degradation, Biochem. Biophys. Res. Commun. 399 (2010) autoregulatory loop between HIPK2 and Siah2 at the apex of the hypoxic re- 623–628. sponse, Nat. Cell Biol. 11 (2009) 85–91. [39] A. Brauckhoff, M. Malz, D. Tschaharganeh, N. Malek, A. Weber, M.O. Riener, C. Soll, [13] J. Zhao, C. Wang, J. Wang, X. Yang, N. Diao, Q. Li, W. Wang, L. Xian, Z. Fang, L. Yu, J. Samarin, M. Bissinger, J. Schmidt, et al., Nuclear expression of the ubiquitin li- E3 ubiquitin ligase Siah-1 facilitates poly-ubiquitylation and proteasomal degra- gase seven in absentia homolog (SIAH)-1 induces proliferation and migration of dation of the hepatitis B viral X protein, FEBS Lett. 585 (2011) 2943–2950. liver cancer cells, J. Hepatol. 55 (2011) 1049–1057. [14] A. Khurana, K. Nakayama, S. Williams, R.J. Davis, T. Mustelin, Z. Ronai, Regulation [40] A. Depaux, F. Regnier-Ricard, A. Germani, N. Varin-Blank, Dimerization of hSiah of the ring finger E3 ligase Siah2 by p38 MAPK, J. Biol. Chem. 281 (2006) proteins regulates their stability, Biochem. Biophys. Res. Commun. 348 (2006) 35316–35326. 857–863. [15] T.G. Hofmann, A. Moller, H. Sirma, H. Zentgraf, Y. Taya, W. Droge, H. Will, M.L. [41] J. Tanikawa, E. Ichikawa-Iwata, C. Kanei-Ishii, A. Nakai, S. Matsuzawa, J.C. Reed, S. Schmitz, Regulation of p53 activity by its interaction with homeodomain- Ishii, p53 suppresses the c-Myb-induced activation of heat shock transcription interacting protein kinase-2, Nat. Cell Biol. 4 (2002) 1–10. factor 3, J. Biol. Chem. 275 (2000) 15578–15585. [16] C.L. Brooks, W. Gu, The impact of acetylation and deacetylation on the p53 path- [42] M. Benkirane, C. Sardet, O. Coux, Lessons from interconnected ubiquitylation and acet- way, Protein Cell 2 (2011) 456–462. ylation of p53: think metastable networks, Biochem. Soc. Trans. 38 (2010) 98–103. [17] G. Legube, L.K. Linares, S. Tyteca, C. Caron, M. Scheffner, M. Chevillard-Briet, D. [43] Z. Lin, H. Yang, Q. Kong, J. Li, S.M. Lee, B. Gao, H. Dong, J. Wei, J. Song, D.D. Zhang, Trouche, Role of the histone acetyl transferase Tip60 in the p53 pathway, J. Biol. et al., USP22 Antagonizes p53 Transcriptional Activation by Deubiquitinating Chem. 279 (2004) 44825–44833. Sirt1 to Suppress Cell Apoptosis and Is Required for Mouse Embryonic Develop- [18] L. Liu, D.M. Scolnick, R.C. Trievel, H.B. Zhang, R. Marmorstein, T.D. Halazonetis, S.L. ment, Mol. Cell 46 (4) (2012) 484–494. Berger, p53 sites acetylated in vitro by PCAF and p300 are acetylated in vivo in re- [44] W.J. Kim, M.N. Rivera, E.J. Coffman, D.A. Haber, The WTX tumor suppressor en- sponse to DNA damage, Mol. Cell. Biol. 19 (1999) 1202–1209. hances p53 acetylation by CBP/p300, Mol. Cell 45 (2012) 587–597.