FASEB J. Integrated DNA Technologies
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published as doi: 10.1096/fj.08-113464.
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Supplemental Data
Right arrow All Versions of this Article:
fj.08-113464v1
22/11/3853    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Quénet, D.
Right arrow Articles by Dantzer, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Quénet, D.
Right arrow Articles by Dantzer, F.
(The FASEB Journal. 2008;22:3853-3865.)
© 2008 FASEB

The histone subcode: poly(ADP-ribose) polymerase-1 (Parp-1) and Parp-2 control cell differentiation by regulating the transcriptional intermediary factor TIF1β and the heterochromatin protein HP1{alpha}

Delphine Quénet*, Véronique Gasser*, Laetitia Fouillen{ddagger}, Florence Cammas{dagger}, Sarah Sanglier-Cianferani{ddagger}, Régine Losson{dagger} and Françoise Dantzer*,1

* Département Intégrité du Génome, UMR7175, Ecole Supérieure de Biotechnologie de Strasbourg, Illkirch, France;

{dagger} Institut de Génétique et de Biologie Moléculaire et Cellulaire, Illkirch, France; and

{ddagger} Laboratoire de Spectrométrie de Masse Bio-organique, UMR7178, Ecole de Chimie, Polymères et Matériaux, Strasbourg, France

1Correspondence: Département IDG, UMR7175, ESBS, Bld. S. Brant, 67412 Illkirch, France. E-mail: fdantzer{at}esbs.u-strasbg.fr


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Recent advances reveal emerging unique functions of poly(ADP-ribose) polymerase-1 (Parp-1) and Parp-2 in heterochromatin integrity and cell differentiation. However, the chromatin-mediated molecular and cellular events involved remain elusive. Here we describe specific physical and functional interactions of Parp-1 and Parp-2 with the transcriptional intermediary factor (TIF1β) and the heterochromatin proteins (HP1) that affect endodermal differentiation. We show that Parp-2 binds to TIF1β with high affinity both directly and through HP1{alpha}. Both partners colocalize at pericentric heterochromatin in primitive endoderm-like cells. Parp-2 also binds to HP1β but not to HP1{gamma}. In contrast Parp-1 binds weakly to TIF1β and HP1β only. Both Parps selectively poly(ADP-ribosyl)ate HP1{alpha}. Using shRNA approaches, we provide evidence for distinct participation of both Parps in endodermal differentiation. Whereas Parp-2 and its activity are required for the relocation of TIF1β to heterochromatic foci during primitive endodermal differentiation, Parp-1 and its activity modulate TIF1β-HP1{alpha} association with consequences on parietal endodermal differentiation. Both Parps control TIF1β transcriptional activity. In addition, this work identifies both Parps as new modulators of the HP1-mediated subcode histone.—Quénet, D., Gasser, V., Fouillen, L., Cammas, F., Sanglier-Cianferani, S., Losson, R., Dantzer, F. The histone subcode: poly(ADP-ribose) polymerase-1 (Parp-1) and Parp-2 control cell differentiation by regulating the transcriptional intermediary factor TIF1β and the heterochromatin protein HP1{alpha}.


Key Words: epigenetics • pericentric heterochromatin • post-translational modification • protein interactions


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
DYNAMIC REORGANIZATION OF heterochromatic compartments and the epigenetic changes that are associated are currently recognized as important regulators of gene silencing that accompanies diverse cellular processes, including cell differentiation (1) . This finding is further determined by the compartmentalization of specific protein complexes containing histones and nonhistone proteins, chromatin-remodeling and chromatin modifying enzymes, and by the post-translational modifications of histones defined as the histone code (2 3 4) . However, how these activities are coregulated remains unknown.

A broad distribution of heterochromatin (HC) is observed at pericentromeric regions known to play an important role in gene silencing (4) . Its epigenetic nature is defined by methylation of cytosine-phosphate-guanine (CpG), hypoacetylation of histones, and methylation at lysine 9 of histone H3 (meH3K9), which is necessary for the enrichment in heterochromatin protein 1 (HP1) (5) . Three distinct mammalian HP1s have been characterized: HP1{alpha} and HP1β are primarily found within centromeric HC; whereas HP1{gamma} is enriched at euchromatic sites (6) . These proteins participate in chromatin packaging and have a well-established function in HC-mediated silencing. The structure of the HP1 proteins consists of an N-terminal chromodomain (CD) that binds meH3K9 and the histone fold motif of histone H3, a central hinge domain (hinge) that displays RNA/DNA binding activities, and a C-terminal chromoshadow domain (CSD) recognized as a protein-protein interaction domain (7 , 8) . It has been shown recently that HP1s are also targets of post-translational modifications similar to those described for histones, which define the existence of an HP1-mediated histone subcode (9) . Notably, HP1s interact with proteins involved in transcriptional regulation through a specific PxVxL motif called HP1 box, among them the transcription intermediary factor (TIF1β) (10 11 12) . TIF1β functions as a corepressor for the large family of Krüppel-associated box (KRAB)- domain-containing zinc-finger proteins and acts as a molecular scaffold to coordinate various activities that regulate chromatin structure and dynamics (13 14 15) . In addition, TIF1β exerts essential functions in early embryonic development (16) and spermatogenesis (17) .

The mouse embryonal carcinoma F9 cells represent a well-established model of endodermal differentiation that can be induced to differentiate into primitive-endoderm-like (PrE) cells when grown as a monolayer in the presence of retinoic acid (RA) and subsequently in parietal endoderm-like (PE) cells when grown in the presence of both RA and dibutyryl cAMP (18) . In this system, TIF1β-HP1s association plays an essential role in the relocation of TIF1β from euchromatin to HC and in the progression through differentiation by regulating the expression of endoderm-specific genes (19) .

Another regulatory mechanism that controls chromatin structure and integrity is the modification of histones and other nuclear proteins by poly(ADP-ribose) polymers catalyzed by poly(ADP-ribose) polymerases (Parps). Among the 17 members of the Parp family, Parp-1 and Parp-2 heterodimerize, share common binding partners and have been described as active players in the single-strand break/base excision repair process (20) . Parp-1- and Parp-2-deficient mice and cells are very sensitive to both ionizing radiation and alkylating agents, thus supporting a role of both Parps in the cellular response to DNA damage (21) . Moreover, Parp-1–/–Parp-2–/– embryos die at gastrulation, demonstrating the crucial role of poly(ADP-ribosyl)ation during embryonic development (21) . Several lines of evidence support the view that Parp-1 and Parp-2 play prominent roles in the maintenance of constitutive and facultative HC integrity, with, however, the emergence of specific functions for Parp-2. Both proteins localize to telomeres (22) , centromeres (23 , 24) , and rDNA (25) , where they interact and regulate specific partners. Parp-2–/– cells exhibit DNA damage-induced kinetochore defects; whereas the Parp-1+/– Parp-2–/– background displays specific female embryonic lethality associated with X chromosome instability (21) .

Interestingly, both Parp-1 and Parp-2 were also suggested to play critical roles in the progression through differentiation. In the developing mammalian central nervous system, Parp-1 serves roles in transcriptional events required for neuronal differentiation (26) . We have recently described the appearance of specific spontaneous defects in differentiation processes, including adipogenesis (27) , spermiogenesis (28) , and T-lymphocyte maturation (29) in the Parp-2–/– mice. However, the chromatin-mediated molecular mechanisms by which Parp-1 and Parp-2 may control differentiation have not yet been elucidated.

In this work we provide the first evidence for physical and functional selective interactions between Parp-2, Parp-1, TIF1β, and HP1{alpha} that have fundamental implications in HC structure and/or function governing endodermal differentiation. We show that Parp-2 physically binds to TIF1β with high affinity both directly and through HP1{alpha}. Both proteins relocate to pericentric HC throughout differentiation. We also identified a direct interaction of Parp-2 with HP1β. A weaker but significant direct interaction of Parp-1 with TIF1β and HP1β was also detected. Both Parps selectively poly(ADP-ribosylate) HP1{alpha}. Using shRNA approaches, we show that Parp-2-dependent poly(ADP-ribosyl)ation is required for primitive-endodermal differentiation possibly by targeting TIF1β to heterochromatic foci, whereas Parp-1 and its activity participate in the maintenance of TIF1β-HP1 association required for progression through parietal endodermal differentiation. Both Parps control TIF1β-mediated transcriptional activity. In addition, this work identifies Parp-1 and Parp-2 as new actors of the silencing subcode histone that underlies the histone code.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 REFERENCES
 
Plasmids and antibodies
Plasmids and antibodies used are detailed in Supplemental Information.

Cell culture and establishment of stable depleted cell lines
Wild-type and mutant F9 cells were grown in Dulbecco’s modified Eagle’s medium-4.5 g/L glucose (DMEM; Life Technologies, Inc., Gaithersburg, MD, USA) supplemented with 10% fetal calf serum (FBS; PanBiotech, Aidenbach, Germany) and 1% Gentamicin (Life Technologies, Inc.) at 37°C in 5% CO2. To induce PrE and PE differentiations, cells were treated, respectively, with 1 µM all-trans RA (Sigma, Lyon, France) alone or in combination with 250 µM dibutyryl cAMP (dbcAMP; Sigma) as described previously (30) . To establish stable scr, shParp-1, and shParp-2 F9 cell lines, 5 x 106 exponentially growing F9 cells were transfected with 5 µg of XmnI-linearized pSuper-scrParp-2, pSuper-shParp-1, or pSuper-shParp-2 vectors together with 250 ng of AflIII-linearized pGK-Hygro vector. Selection was started by adding 400 µg/ml Hygromycin B (Roche, Basel, Switzerland) to the growth medium 24 h posttransfections for over a period of 2 wk. Several drug-resistant colonies were isolated, expanded, and analyzed for the absolute levels of Parp-1, Parp-2, TIF1β, HP1{alpha}, and actin by Western blotting. To establish stable shParp-1;shParp-2 cell lines, 5 x 106 exponentially growing cells of a selected shParp-1 F9 clone were transfected with 5 µg of XmnI-linearized pSuper-shParp-2 together with 250 ng of AflIII-linearized pGK-Neo plasmid. Selection was started by adding 600 µg/ml neomycin to the growth medium 24 h posttransfections for over a period of 2 wk. Neomycin-resistant clones were isolated, expanded, and analyzed as mentioned above.

Immunoprecipitation, mass spectrometry, Western blot analysis, and glutathione S-transferase (GST) pull-down
For immunoprecipitation in testis cells, 30 testes were collected from 10-wk-old C57/Bl6 mice (Janvier, Le Genest Saint Isle, France) and homogenized by 20 Dounce (no. 2) strokes in lysis buffer [10 mM Tris-HCl, pH 8; 400 mM NaCl; 1% Nonidet P-40; 2 mM dithiothreitol (DTT); 0.5 mM Pefabloc, and protease inhibitor complex (PIC; Roche)]. Cell lysates were incubated on ice for 30 min, then digested with 300 U/ml of DNase I at 25°C for 30 min. After centrifugation at 15,000 rpm at 4°C for 20 min, cleared suspension was quantified by Bradford protein assay. Proteins (10 mg) were incubated overnight at 4°C with either purified anti-Parp-2 pAb or rabbit anti-mouse antibody as control, followed by 2 h incubation at 4°C with protein A sepharose (GE Healthcare, Little Chalfont, UK). The immunoprecipitates were washed twice with washing buffer (10 mM Tris-HCl, pH 8; 0,1% Nonidet P-40; 2 mM DTT; 0.5 mM Pefabloc; and PIC) containing 500 mM NaCl and twice with washing buffer containing 50 mM NaCl. Final pellets were resuspended in 50 µl of Laemmli buffer and subjected to 10% SDS-PAGE. Coprecipitated proteins were stained by SyproRuby (Molecular Probes, Eugene, OR, USA) according to the manufacturer’s instructions. Nano-liquid chromatography-mass spectrometry/mass spectrometry experiments were performed using a CapLC capillary liquid chromatography system (Waters, Milford, MA, USA) coupled to a hybrid quadrupole time-of-flight mass spectrometer (Q-TOF II, Waters) according to standard protocols (31) . For immunoprecipitation in F9 or F9-derived cells, 107 cells were lysed by 3 cycles of freezing and thawing in lysis buffer as above. Cleared lysates were quantified by Bradford protein assay. Following treatment with RNase I (1 mg/ml) for 30 min at room temperature, 200 µg of total proteins was incubated with purified anti-Parp-2 pAb, anti-TIF1β mAb, or the control antibody overnight at 4°C and immunoprecipitated using protein A sepharose for 2 h at 4°C. Beads were washed with (10 mM Tris HCl, pH 8; 50 to 500 mM NaCl; 0.1% Nonidet P-40; 2 mM DTT, and PIC), resuspended in Laemmli buffer, and analyzed by 10% SDS-PAGE and immunoblotting. Blots were probed with the appropriate specific antibodies followed by peroxidase-conjugated secondary antibodies, and developed using the ECL+ detection kit (Amersham, Little Chalfont, UK). When indicated, 100 nM of the Parp inhibitor Ku-0058948 (32) was added to the culture medium 2 h before lysis and maintained throughout the experiment. GST pull-down analysis was performed as described previously (22) .

Immunofluorescence
Immunofluorescence was performed as described previously (25) . Images were captured using a Leica microscope (Leica Microsystems, Heidelberg, Germany) and the capture software OpenLab (Improvision, Perkin Elmer, Inc., Coventry, UK).

In vitro binding assays
Escherichia coli expression and purification of GST, GST-TIF1β, and GST-HP1 fusion proteins were performed as described previously (10) . Equivalent amounts of purified proteins, quantified by Coomassie staining after SDS-PAGE, were incubated with 300 ng of either purified human Parp-1 or murine Parp-2 in binding buffer (20 mM Tris-HCl, pH 7.5; 300 mM NaCl; 0.5 mM Pefabloc; 0.1% Nonidet P-40; and PIC) for 2 h at 4°C. The beads were washed twice with washing buffer (10 mM Tris-HCl, pH 8; 2 mM DTT; 0.5% Nonidet P-40; 0.5 mM Pefabloc; and PIC) containing 500 mM NaCl and twice in washing buffer containing 50 mM NaCl. Beads were resuspended in Laemmli buffer and analyzed by Western-blotting.

Heteromodification and noncovalent binding of poly(ADP-ribose) on GST-fusion proteins
E. coli-expressed GST-fused proteins were purified as above and quantified by Coomassie staining on SDS-PAGE. Heteromodification of equivalent amounts of purified proteins by either human Parp-1 or mouse Parp-2 was performed as described previously (33) .

For noncovalent binding, 32P-labeled poly(ADP-ribose) was synthesized and purified according to Dantzer et al. (22) . Equivalent amounts of purified E. coli-expressed GST-fused proteins and 1 µg of purified recombinant XRCC1 were spotted directly onto nitrocellulose and incubated with 32P-labeled poly(ADP-ribose) as described previously (22) .

Quantitative RT-PCR
Quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) using total F9 RNA samples was performed by using the Quantitect SYBR Green PCR kit (Qiagen, Valencia, CA, USA) following the manufacturer’s instructions in combination with the Light Cycler Detection System. The PCR products were analyzed with the manufacturer’s software. The following primer sequences were used: Hprt-5', 5'-TGACACTGGCAAAACAATGCA-3'; Hprt-3', 5'-GGTCCTTTTCACCAGCAAGCT-3'; Mest-5', 5'-CTCCAAAAACTCTGGATACG-3'; Mest-3', 5'-GAAATTCAGAAGACGCTGGG-3'; HNF4–5', ACACGTCCCCATCTGAAGGTG; HNF4–3', CTTCCTTCTTCATGCCAGCCC.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 REFERENCES
 
Association of Parp-2 and Parp-1 with TIF1β and HP1{alpha} in mammalian cells
To identify interacting partners of Parp-2, testis cell extracts were immunoprecipitated with an anti-Parp-2 antibody or a control antibody, and captured proteins were analyzed by mass spectrometry (Fig. 1 A). Among the proteins identified, we isolated previously described partners of Parp-2, including Parp-1 and DNA polymerase β (33) and a partner of Parp-1 named macroH2A1.2 (34) , thus clearly supporting the validity of the approach used. Interestingly, we isolated 6 unique tryptic peptides from TIF1β. The functional similarities between Parp-2, Parp-1, and TIF1β, with regard to their accumulation on pericentric HC and their role in differentiation pathways, in addition to the previously described interaction of TIF1β with HP1{alpha} in mouse embryocarcinoma F9 cells, prompted us to investigate further whether these proteins could be physically associated in this model (Fig. 1B ). F9 cell extracts were immunoprecipitated with an anti-Parp-2 antibody or an irrelevant antibody, and coimmunoprecipitation of TIF1β and HP1{alpha} was assessed by Western blotting. We detected significant association of TIF1β and HP1{alpha} with Parp-2 in F9 cells (Fig. 1B , lane 2), whereas no association was detected using the control antibody (Fig. 1B , lane 1). In an analog experiment, when F9 cell extracts were immunoprecipitated with an anti-TIF1β antibody or a control antibody, significant fractions of HP1{alpha} and Parp-1 were found in the TIF1β immunoprecipitate (Fig. 1B , lane 6) but not in the control immunoprecipitate (Fig. 1B , lane 5). Taken together, these results describe an association of a subset of Parp-1 and Parp-2 with TIF1β and HP1{alpha} in mammalian cells.


Figure 1
View larger version (16K):
[in this window]
[in a new window]

  		Figure 1. Association of Parp-1 and Parp-2 with TIF1β and HP1{alpha} in mammalian cells. A) Identification of Parp-2-associated TIF1β by mass spectrometry. Testis cell extracts were immunoprecipitated using a control antibody (lane 1) or an anti-Parp-2 antibody (lane 2). Interacting proteins were resolved by SDS-PAGE, stained with Sypro-Red, and analyzed by mass spectrometry. Numbers in brackets refer to the number of peptides matched. Six different peptides (sequences shown) of TIF1β were found localized along the sequence. B) Top panel: coimmunoprecipitation of TIF1β and HP1{alpha} with Parp-2 in F9 cells. F9 cell extracts were immunoprecipitated with a control antibody (lane 1) or an anti-Parp-2 antibody (lane 2) and analyzed by Western blotting using successively anti-TIF1β, anti-HP1{alpha}, and anti-Parp-2 antibodies. Input corresponds to 1/20 of the amount of cell extract used for immunoprecipitation (lane 3). Bottom panel: coimmunoprecipitation of Parp-1 and HP1{alpha} with TIF1β in F9 cells. F9 cell extracts were immunoprecipitated with an anti-TIF1β antibody (lane 6) or a control antibody (lane 5) and analyzed by Western blot. For detection of TIF1β and HP1{alpha}, 1/8 of the immunoprecipitate was probed using successively anti-HP1{alpha} and anti-TIF1β antibodies. For detection of Parp-1, 7/8 of the immunoprecipitate was probed using an anti-Parp-1 antibody. Lane 4 (input) corresponds to 1/25 the amount of cell extract used for immunoprecipitation.

Parp-2 and Parp-1 interact differentially with HP1 isoforms and TIF1β in vitro
To characterize the complex further, binding assays between Parp-1, Parp-2, TIF1β, and HP1{alpha} were performed in vitro using purified recombinant proteins (Fig. 2 ). To carefully address the specificity of the interaction with members of the HP1 family, we also tested HP1β and HP1{gamma}. GST-HP1 fusion proteins were expressed in E. coli, purified on glutathione S-sepharose beads, and batched with purified recombinant Parp-1 or Parp-2. After GST-pull down followed by stringent washes, copurification of Parps were analyzed by Western blotting. As shown in Fig. 2A (top panel), we detected binding of Parp-2 to GST-HP1{alpha} (lane 4) and GST-HP1β (lane 6) but not to GST-HP1{gamma} (lane 10) nor GST (lanes 2, 8). In contrast, Parp-1 bound weakly but reproducibly to GST-HP1β only (lane 5).


Figure 2
View larger version (18K):
[in this window]
[in a new window]

  		Figure 2. Differential interaction of Parp-2 and Parp-1 with HP1 isoforms and TIF1β. A) In vitro interaction of Parp-1 and Parp-2 with HP1 isotypes. Top panel: purified recombinant Parp-1 (lanes 1, 3, 5, 7, 9) or Parp-2 (lanes 2, 4, 6, 8, 10) was incubated in a batch assay with purified GST (lanes 1, 2 and 7, 8), GST-HP1{alpha} (lanes 3, 4), GST-HP1β (lanes 5, 6) or GST-HP1{gamma} (lanes 9, 10). Bound Parps were analyzed by GST pull-down and Western blotting using successively anti-Parp-1, anti-Parp-2, and anti-GST antibodies. Bottom panel: purified recombinant Parp-2 was incubated in a batch assay with purified GST-fused CD (GST-CD1–66, lane 1), GST-fused hinge domain (GST-hinge67–119, lane 2) or GST-fused CSD (GST-CSD119–191, lane 3) of HP1{alpha}. Bound Parp-2 proteins were analyzed by GST pull-down and Western blotting using successively anti-Parp-2 and anti-GST antibodies. B) The HP1 box is required for the association of TIF1β with Parp-2 in F9 cells. Top panel: schematic representation of TIF1β indicating the TIF1βHP1box mutant used. Bottom panel: lysates from Cos1 cells expressing GST-Parp-2 (lanes 1, 3, 4) or GST (lane 2) together with either Flag-TIF1β (lanes 1–3) or Flag-TIF1βHP1box (lane 4) were analyzed by GST pull-down and Western blotting using anti-TIF1β and anti-GST antibodies, respectively. In lane 1, Ku-0058948 was added throughout the experiment. C) Parp-2 directly interacts with TIF1β. Top panel: purified recombinant Parp-1 (lanes 1, 3) or Parp-2 (lanes 2, 4) was incubated in a batch assay with purified GST or GST-TIF1β. Bound Parps were analyzed by GST pull-down and Western blotting using successively anti-Parp-1, anti-Parp-2, and anti-GST antibodies. Bottom panel: wild-type F9 (lanes 1, 2) and mutant TIF1βHP1box/– (lane 3) cell extracts were immunoprecipitated with a control antibody (lane 1) or an anti-Parp-2 antibody (lanes 2, 3) and analyzed by Western blotting using successively anti-TIF1β and anti-Parp-2 antibodies.

To identify the region of HP1{alpha} to which Parp-2 binds, purified GST fusion proteins expressing various HP1{alpha} deletion domains—GST-CD, GST-hinge, and GST-CSD—were tested for interaction with purified Parp-2. As shown in Fig. 2A (bottom panel), Parp-2 interacts with the CSD (lane 3) and hinge domain (lane 2) of HP1{alpha}, whereas no interaction was detected with the CD (lane 1). Thus Parp-2 but not Parp-1 can directly interact with HP1{alpha} in vitro.

Because TIF1β also interacts directly with HP1{alpha} (11) , we next addressed whether the association of TIF1β with Parp-2 detected in F9 cells (Fig. 1B ) requires HP1{alpha}. Flag-tagged proteins expressing either wild-type TIF1β (Flag-TIF1β) or TIF1β with a mutation in the HP1 box that disrupts its interaction with HP1{alpha} (Flag-TIF1βHP1box) were coexpressed in Cos1 cells together with GST-Parp-2 or GST alone. To test the role of poly(ADP-ribosyl)ation, the same assay was performed in the presence of the Parp inhibitor Ku-0058948. Coprecipitating proteins were analyzed by GST pull-down experiments and Western blotting. As shown in Fig. 2B , we observed a significantly weaker copurification of Flag-TIF1βHP1box with GST-Parp-2 (lane 4) when compared to Flag-TIF1β (lane 3). Thus, the HP1 box mutation of TIF1β that was previously described to disrupt the interaction between TIF1β and HP1 proteins (19) also impairs the interaction between TIF1β and Parp-2, which suggests that TIF1β-Parp-2 association involves—at least partly—HP1 proteins. No copurification with GST was detected (Fig. 2B , lane 2). The addition of Ku-0058948 also significantly impaired the binding of Flag-TIF1β to GST-Parp-2, thus indicating a role of poly(ADP-ribose) in the association of both partners (Fig. 2B , lanes 1, 3).

We next compared the direct binding efficiency of Parp-1 and Parp-2 to TIF1β (Fig. 2C ). Under similar conditions of binding assays as above, we identified an efficient binding of Parp-2 to TIF1β and a weaker but reproducible interaction with Parp-1 (Fig. 2C , top panel). To verify whether Parp-2 and TIF1β also associate independently of HP1 in vivo, nuclear extracts from TIF1βHP1box/– cells expressing the mutated TIF1βHP1box were immunoprecipitated with an anti-Parp-2 antibody, and the immunoprecipitates were probed for the presence of TIF1β (Fig. 2C , bottom panel). Coimmunoprecipitation of TIF1β with Parp-2 was detected both in TIF1βHP1box/– cells (Fig. 2C , lane 3) and in the parental F9 cells (Fig. 2C , lane 2), but not in control immunoprecipitate (Fig. 2C , lane 1).

Finally, we found that TIF1β interacts with either the central E or the catalytic F domains of Parp-2 but not the N-terminal DNA-binding domain (data not shown).

Taken together, these results describe an heterochromatic protein network characterized by 1) a selective efficient binding of Parp-2 to HP1{alpha} and HP1β but not HP1{gamma}, 2) an association of Parp-2 with TIF1β both directly and through HP1, and 3) a weaker but reproducible direct binding of Parp-1 to HP1β and TIF1β.

Selective poly(ADP-ribosyl)ation and noncovalent binding of poly(ADP-ribose) to HP1{alpha}
To gain further insights into the functional interactions governing this protein network, we evaluated the ability of either Parp-1 or Parp-2 to poly(ADP-ribosyl)ate HP1 isotypes and TIF1β (Fig. 3 ). Purified GST-HP1 and GST-TIF1β fusion proteins or GST alone were incubated with either Parp-1 or Parp-2 or no protein, in the presence of {alpha}-32PNAD+ and DNase-I-treated calf thymus DNA. Autoradiography revealed that both Parp-1 and Parp-2 were able to poly(ADP-ribosyl)ate selectively HP1{alpha} but not HP1β, HP1{gamma}, TIF1β, or GST alone (Fig. 3A ). To identify the domain of HP1{alpha} poly(ADP-ribosyl)ated, we used the same approach with the purified E. coli-expressed deletion domains of HP1{alpha} and found that HP1{alpha} is poly(ADP-ribosyl)ated on its hinge domain by both Parp-1 and Parp-2 (Fig. 3B ).


Figure 3
View larger version (14K):
[in this window]
[in a new window]

  		Figure 3. The hinge domain of HP1{alpha} is poly(ADP-ribosyl)ated and binds poly(ADP-ribose). A, B) The hinge domain of HP1{alpha} is poly(ADP-ribosyl)ated by Parp-1 and Parp-2. A) Purified GST, GST-HP1{alpha}, GST-HP1β, GST-HP1{gamma}, and GST-TIF1β were incubated with either Parp-1 or Parp-2 in activity buffer containing [{alpha}-32P]NAD+ and DNase I-activated DNA. Right panel: autoradiography. Left panel: fusion proteins were analyzed by Western blotting with an anti-GST antibody. B) Top panel: schematic representation of HP1. Bottom panel: similar experiment was performed as in A, using purified GST, GST-CD1–66, GST-CSD119–191, or GST-hinge67–119 domains of HP1{alpha}. Right panel: autoradiography. Left panel: fusion proteins analyzed by Western blotting using an anti-GST antibody. C, D) Poly(ADP-ribose) binds to the hinge domain of HP1{alpha}. C) Purified GST, GST-HP1{alpha}, GST-HP1β, GST-HP1{gamma}, and GST-TIF1β were spotted on nitrocellulose and incubated with [{alpha}-32P]poly(ADP-ribose). Top panel: autoradiography. Bottom panel: the amount of protein loaded was controlled by SDS-PAGE and Coomassie staining. D) Similar experiment was performed as in C, using purified GST, GST-CD1–66, GST-CSD119–191, or GST-hinge67–119 domains of HP1{alpha}. XRCC1 (1 µg) was used as positive control (22) . Top panel: autoradiography. Bottom panel: the amount of protein loaded was controlled by SDS-PAGE and Coomassie staining.

To test whether HP1 isoforms or TIF1β could directly bind to poly(ADP-ribose), similar amounts of purified GST, GST-HP1, and GST-TIF1β fusion proteins were spotted onto nitrocellulose and incubated with radioactive poly(ADP-ribose) (Fig. 3C ). Detection of a radioactive signal was observed only for GST-HP1{alpha}, thus showing that HP1{alpha} binds tightly and stably to the poly(ADP-ribose). The same conditions were used to identify the domain of HP1{alpha} that could bind to poly(ADP-ribose) and showed that PAR binds specifically to the hinge domain (Fig. 3D ).

Interaction and colocalization of Parp-2 with TIF1β onto pericentric HC in PrE cells
The above results and the recent advances describing an essential role of TIF1β-HP1 interaction during endodermal differentiation (19) prompted us to follow the association of Parps and TIF1β in this process. We first analyzed the expression of Parp-1, Parp-2, TIF1β, and HP1{alpha} in nuclear extracts of undifferentiated F9 or differentiated PrE and PE cell lines. Whereas the expression of Parp-1 and TIF1β decreased throughout differentiation of F9 to PrE and PE cells, the expression of Parp-2 and HP1{alpha} remained constant (Fig. 4 A).


Figure 4
View larger version (31K):
[in this window]
[in a new window]

  		Figure 4. Parp-2 interacts and colocalizes with TIF1β in PrE cells. A) Expression of Parp-1, Parp-2, TIF1β, and HP1{alpha} throughout differentiation. F9 cells (lane 1) were induced to differentiate into PrE (lane 2) or PE cells (lane 3). Equivalent amounts of total protein extracts were separated by SDS-PAGE and analyzed by Western blotting with the appropriate antibodies. Lane 4: purified recombinant Parp-1 (50 ng). Lane 5: purified recombinant Parp-2 (50 ng). B) Left panel: colocalization of Parp-2 and TIF1β to centromeric HC in PrE cells. F9 cells were induced to differentiate into PrE and PE cells and stained red with Alexa 568-labeled mouse anti-TIF1β antibody (a, d, g) and green with Alexa 488-labeled rabbit anti-Parp-2 antibody (b, e, h). DNA was couterstained with 4',6'-diamidino-2-phenylindole (dapi) (c, f, i). Right panel: subcellular localization of Parp-1 throughout differentiation. F9 cells were induced to diffentiate into PrE and PE cells and stained red with Alexa 568-labeled mouse anti-TIF1β antibody (j, m, p) and green with Alexa 488-labeled rabbit anti-Parp-1 antibody (k, n, q). DNA was couterstained with dapi (l, o, r). C) F9 cells were induced to differentiate, and equivalent amounts of total protein cell lysates from F9, PrE, and PE cells were immunoprecipitated with a control antibody (lane 1) or an anti-Parp-2 antibody (lanes 2–4), and analyzed by Western blotting using successively anti-TIF1β and anti-Parp-2 antibodies. Input corresponds to 1/50 of the amount of total cell extract used for immunoprecipitation.

We next monitored the colocalization of Parp-1 and Parp-2 with TIF1β throughout the differentiation of F9 cells by indirect immunofluorescence (Fig. 4B ). As described previously, Parp-2 (Fig. 4Bb ) and Parp-1 (Fig. 4Bk ) displayed a nuclear punctate distribution in F9 cells, with, however, a stronger accumulation in nucleoli (25) , whereas TIF1β (Fig. 4Ba, j ) showed homogeneous nuclear staining and was excluded from nucleoli. Interestingly, the differentiation into PrE induced a dynamic targeting of Parp-2 (Fig. 4Be ) and TIF1β (Fig. 4Bd, m ) but not Parp-1 (Fig. 4Bn ) onto pericentric HC foci where both proteins colocalize. On further differentiation into PE cells, Parp-1 (Fig. 4Bq ), Parp-2 (Fig. 4Bh ), and TIF1β (Fig. 4Bg, p ) exhibited a pattern similar to undifferentiated cells, although with less accumulation of Parps in the nucleoli.

To support this result further, we examined the association of Parp-2 and TIF1β throughout differentiation. Nuclear extracts of F9, PrE, and PE cells were immunoprecipitated with an anti-Parp-2 antibody or an irrelevant antibody, and the immunoprecipitates were probed for the presence of TIF1β (Fig. 4C ). The amount of TIF1β coimmunoprecipitated was correlated with the expression of each partner throughout differentiation. Indeed, TIF1β was clearly detected in Parp-2 immunoprecipitates of F9 and PrE cell extracts (Fig. 4C , lanes 2, 3) containing significant amounts of proteins, but not readily in PE cell extracts, due to limited amounts of proteins expressed (Fig. 4C , lane 4). No TIF1β was detected in the control immunoprecipitate (Fig. 4C , lane 1). Together, these results indicate that the interaction of TIF1β with Parp-2 is maintained in the differentiated PrE cells.

Altogether, these data indicate that Parp-2 associates with TIF1β within regions of pericentric HC throughout differentiation of F9 to PrE cells in addition to the protein complex formed in the euchromatin compartment of F9 stem cells (Fig. 1) .

Parp-2 is required for the differentiation into PrE cells, whereas Parp-1 is required for terminal differentiation into PE and visceral-endoderm-like (VE) cells
To investigate whether the absence of either Parp-2, Parp-1, or both lead to a defect in the differentiation of F9 cells to endoderm-like cells, we used the shRNA approach to generate stable clones depleted in either Parp-1 (shParp-1), Parp-2 (shParp-2), or both proteins (shParp-1;shParp-2). A Western blot analysis of the selected clones is shown in Fig. 5 A. When compared to the expression of a housekeeping protein β-actin, the extent of Parps depletion was estimated to be more than 99% in the selected clones. Noticeably, the depletion of either Parp-1, Parp-2, or both had no effect on the level of TIF1β expression.


Figure 5
View larger version (30K):
[in this window]
[in a new window]

  		Figure 5. Parp-2-depleted F9 cells are impaired in the differentiation into PrE cells, whereas Parp-1-depleted F9 cells do not differentiate into PE cells. A) Analysis of the extent of depletion of Parps proteins. Western blot analysis for the expression of Parp-2, Parp-1, TIF1β, and β-actin in control (scr) (lanes 1, 3, 5), Parp-2-depleted (shParp-2) (lane 2), Parp-1-depleted (shParp-1) (lane 4), and Parp-1;Parp-2-depleted (shParp-1;shParp-2) (lane 6) F9 cells. B) Effect of Parp-1-, Parp-2-, and Parp-1;Parp-2-depletion on F9 cell differentiation. Scr, shParp-2, shParp-1, and shParp-1;shParp-2 F9 cells were induced to differentiate into PrE, and PE cells as described in Materials and Methods in the absence or in the presence of Ku-0058948 (100 nM) and photographed under a phase contrast microscope.

We induced each depleted F9 cell line to differentiate and followed differentiation by the morphological features characteristic of PrE and PE cells (30) . To evaluate the requirement of Parp-1 or Parp-2 catalytic activity, cells were grown in the absence or in the presence of Ku-0058948 throughout differentiation. As shown in Fig. 5B , when grown in the presence of the Parp inhibitor, control scr-F9 cells displayed a weak reduced capacity to differentiate to both PrE (Fig. 5Be ) and PE (Fig. 5Bf ), which suggests a potential involvement of poly(ADP-ribosyl)ation in both stages. In contrast, the knockdown of Parp-2 clearly impaired differentiation into PrE (Fig. 5Bh ), even though a population of remaining cells was still able to progress to PE (Fig. 5Bi ). The addition of Ku-0058948 had no major additional incidence on the differentiation of shParp-2 to PrE (Fig. 5Bk ) but significantly disrupted further differentiation to PE (Fig. 5Bl ), thus suggesting a role of Parp-1 catalytic activity in the second stage of differentiation. In line with this observation, Parp-1-depleted cells differentiated into PrE (Fig. 5Bn ) but did not differentiate further into PE (Fig. 5Bo ). The addition of Ku-0058948 significantly impaired the potential of shParp-1 to differentiate to PrE (Fig. 5Bq ), in agreement with an essential role of Parp-2-dependent poly(ADP-ribosyl)ation in primitive endodermal differentiation substantially increased in the absence of Parp-1. As expected, the depletion of both Parp-1 and Parp-2 completely inhibited the differentiation to both PrE and PE (Fig. 5Bt, u ).

Taken together, these data firmly assign an essential role of Parp-2 and its activity in the differentiation of F9 to PrE, whereas Parp-1 and the associated activity are crucial for terminal differentiation to PE. Expression of Troma-1 in these cells confirmed their differentiated status (data not shown). We also studied the requirement of either Parp-1, Parp-2, or both in an other model of terminal differentiation into VE cells and found an essential role of Parp-1 but not Parp-2 in this process (Supplemental Fig. 1).

Impaired relocation of TIF1β to pericentric HC in Parp-2-depleted PrE cells but not in Parp-1-depleted PrE cells
Differentiation to PrE is accompanied by relocation of TIF1β from euchromatin to HC (30) . To investigate the role of Parp-1 and -2 and poly(ADP-ribosyl)ation in this process, we compared the dynamic relocation of TIF1β onto pericentric HC during PrE differentiation in each stable depleted cell line and in the absence or in the presence of Ku-0058948 (Fig. 6 ). In nontreated cells, TIF1β staining was homogeneously distributed within the nucleus of scr, shParp-1, shParp-2, or shParp-1;shParp-2 F9 cells (data not shown). After 4 days of RA treatment, normal targeting of TIF1β to pericentric HC was observed in an average of 33% of shParp-1-PrE compared to 31% in the control scr-PrE, thus indicating that the absence of Parp-1 has no effect on TIF1β relocation. However, when grown in the presence of Ku-0058948, the percentage of control scr-PrE displaying focal staining of TIF1β significantly decreased to 12%, reflecting the involvement of poly(ADP-ribosyl) ation in TIF1β HC targeting. In similar conditions the differentiation to PrE was only slightly affected (Fig. 5Be ). Similarly, the knockdown of Parp-2 clearly impaired TIF1β redistribution to the same extent as for Ku-0058948-treated scr-PrE, thus suggesting an essential role of Parp-2 and its catalytic activity in this process. Accordingly, the addition of Ku-0058948 had no major additional incidence on TIF1β relocation in shParp-2-PrE but decreased the percentage of shParp-1-PrE with TIF1β foci.


Figure 6
View larger version (29K):
[in this window]
[in a new window]

  		Figure 6. Targeting TIF1β to pericentric HC in PrE cells requires Parp-2 but not Parp-1. A) Scr, shParp-1, and shParp-2 F9 cells were induced to differentiate in PrE in the absence or in the presence of Ku-0058948 and processed for immunofluorescence using an anti-TIF1β antibody (green). DNA was counterstained with dapi (blue). A representative cell according to the histogram in B is shown. B) Histogram showing percentage of cells displaying TIF1β heterochromatic foci. Cells in >20 randomly selected immunofluorescence fields were scored for the presence or absence of TIF1β nuclear foci. An average of 500 cells was scored per cell line and condition. Results are averages from at least 3 independent experiments.

Taken together, these results reveal an essential specific role of Parp-2, but not Parp-1, and its catalytic activity in the TIF1β targeting to heterochromatic foci during PrE differentiation. Under similar conditions, a wild-type-like accumulation of both HP1{alpha} and trimethyl-H3K9 to HC was detected in all cases (Supplemental Fig. 2).

Parp-1 but not Parp-2 controls the interaction between TIF1β and HP1{alpha}
The TIF1β-HP1 interaction is indispensable for PE differentiation (19) . The absence of PE differentiation in shParp-1, shParp-1;shParp-2, or shParp-2 cells treated with Ku-0058948 suggests a role of Parp-1 and its activity in the association of TIF1β with HP1. To test this hypothesis, we examined the association of both partners by coimmunoprecipitation in early differentiating PrE, shParp-1-PrE, and shParp-2-PrE in a window of time in which TIF1β-HP1 association is required for terminal differentiation (Fig. 7 A, left panel). Interestingly, the absence of Parp-1 but not Parp-2 caused a weak but reproducible decrease in the coimmunoprecipitation of HP1{alpha} with TIF1β compared to the association detected in the control F9 (Fig. 7A ; compare lanes 2, 3 with 1) thus revealing a partial but essential and specific role of Parp-1 compared to Parp-2 in the maintenance of TIF1β-HP1{alpha} interaction. To examine further the role of poly(ADP-ribosyl)ation in this association, similar coimmunoprecipitation experiments were performed in the presence of the Parp inhibitor (Fig. 7A , right panel). Inhibition of Parp activity also impaired TIF1β-HP1{alpha} coimmunoprecipitation (Fig. 7A ; compare lanes 6, 7). Taken together, these results are in favor of a role of Parp-1 but not Parp-2 and poly(ADP-ribosyl)ation in the association of TIF1β with HP1{alpha} that controls the progression through terminal differentiation.


Figure 7
View larger version (15K):
[in this window]
[in a new window]

  		Figure 7. Parp-1 but not Parp-2 is involved in the maintenance of TIF1β-HP1{alpha} interaction, whereas both proteins control TIF1β-dependent transcriptional activity. A) Left panel: F9 (lane 1), shParp-1 (lane 2), and shParp-2 (lane 3) cells were first grown for 24 h in the presence of RA to induce differentiation. Equivalent amounts of total protein cell lysates from early differentiating cells were then immunoprecipitated using an anti-TIF1β antibody and analyzed by Western blotting using successively anti-HP1{alpha} and anti-TIF1β antibodies. Input corresponds to 1/6 of the amount of total cell extract used for immunoprecipitation. Right panel: F9 cells were induced to differentiate for 24 h in the absence (lanes 4, 6) or in the presence of Ku-0058948 (lanes 5, 7) and treated for immunoprecipitation and Western-blot analysis as in A. B) Quantitative RT-PCR analysis of RNA from scr, shParp-1, shParp-2, or shParp-1;Parp-2 early differentiating cells, for the endoderm-specific transcript HNF4 (left panel) and the mesoderm-specific transcript Mest (right panel) normalized against HPRT control. Histogram displays the ratio of specific gene expression in RA-treated vs. untreated cells.

Both Parps control TIF1β transcriptional activity
Gene expression analysis in early differentiating PrE cells has revealed an essential role of TIF1β-HP1 association for induction of the endoderm-specific gene HNF4 (19) and repression of the mesoderm-specific transcript Mest (unpublished data). To investigate the role of Parp-1 and Parp-2 in TIF1β transcriptional activity, we analyzed the transcriptional level of both genes in control (scr), shParp-1, shParp-2, and shParp-1;Parp-2-PrE by qRT-PCR (Fig. 7B ). In agreement with previous data, we detected a 2.4-fold reduction of HNF4 (19) and a 1.72-fold induction of Mest in TIF1βHP1box/– cells (data not shown). Surprisingly, whereas the expression of HNF4 was equivalent in control and shParp-1 cells, we found a significant reduced expression in shParp-2 and shParp-1;Parp-2 cells in comparison with control cells (Fig. 7B , left panel; 8- and 3.2-fold, respectively). In contrast, Mest was up-regulated by 5.9-, 7-, and 3.9-fold in the absence of either Parp-1, Parp-2, or both Parps, respectively (Fig. 7B , right panel). Together, these results clearly indicate a role of both Parps in the expression of TIF1β-regulated specific genes in addition to the critical role of TIF1β-HP1 complex.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
REFERENCES
 
It was previously shown that TIF1β-HP1 interaction plays a critical role during F9 cell differentiation (19) . Here, we provide evidence for physical and functional selective interactions between Parp-1, Parp-2, TIF1β, and HP1 in mammalian cells that provide an additional level of regulation through at least two distinct mechanisms: whereas Parp-2 activity controls TIF1β targeting to pericentric HC, Parp-1 activity essentially acts on TIF1β-HP1 association. Both might contribute to TIF1β-dependent gene expression.

Differential association of Parp-2 and Parp-1 with TIF1β and HP1 isotypes
In a proteomic screen of Parp-2 interacting proteins in testis cell extracts, we identified the transcriptional intermediary factor 1β (TIF1β) together with Parp-1. Interestingly, both Parps and TIF1β display several functional similarities in support of a dynamic interplay between these partners: 1) both Parps and TIF1β relocate on pericentric HC, in metaphase cells for Parp-1 (23) and Parp-2 (24) and in differentiated cells for TIF1β (30) ; 2) Parp-1 and Parp-2 (21) and TIF1β (16) exert central cellular functions during embryonic development; and 3) both Parps function in various differentiation processes, including neurogenesis (26) , dendritic cell differentiation (35) , spermiogenesis (28) , adipogenesis (27) , and T-lymphocyte development (29) , whereas TIF1β in association with HP1 is essential for F9 cells to differentiate into endoderm-like cells (19) . In line with this latter observation, we next confirmed that TIF1β and HP1{alpha} can be coimmunoprecipitated with Parp-2 and Parp-1 in F9 cells, thus describing the existence of a protein network involving Parp-1, Parp-2, TIF1β, and HP1{alpha}. We further characterized this complex combining in vitro protein-protein interaction assays and GST pull-down experiments and show here that both Parp-2 and Parp-1 interact directly but selectively and with significantly different affinities with both HP1 isoforms and TIF1β. Whereas Parp-2 binds efficiently and with high affinity to HP1{alpha}, HP1β, and TIF1β, Parp-1 interacts only weakly but reproducibly with HP1β and TIF1β. The association of Parp-2 with TIF1β is mediated both by a direct interaction of both partners and an indirect interaction through HP1{alpha}. Indeed, the PxVxL motif of TIF1β essential for its interaction with HP1 (11) is also required for its association with Parp-2. Together, these data describe Parp-1 and Parp-2 as new components of the TIF1β-HP1-containing heterochromatic network but strongly support the hypothesis that both proteins possibly display distinct functional roles during endodermal differentiation. Accordingly, we observed a dynamic relocation of Parp-2 to centromeric regions during PrE differentiation, where it colocalizes with TIF1β. In addition, Parp-2-TIF1β association is maintained in PrE cells. In contrast, Parp-1 subcellular localization remains unaffected throughout differentiation.

We next evaluated the importance of poly(ADP-ribosyl)ation in the complex. Despite the high 80% sequence identity and structural similarities among all three HP1 isotypes, we identified a selective poly(ADP-ribosyl)ation of HP1{alpha} by both Parp-2 and Parp-1. This result suggests a specific direct role of Parp activity in the modulation of higher-order chromatin structures and molecular interactions involving HP1{alpha}. The mechanism by which Parp-2 and Parp-1 control HP1{alpha} function involves both a covalent heteromodification and a noncovalent binding of poly(ADP-ribose) to HP1{alpha}. By targeting the hinge domain of HP1{alpha}, both ways of regulation can help to adjust various hinge-specific functions of HP1{alpha}. It is well recognized that the RNA-binding activity of HP1{alpha} that resides within its central hinge domain contributes to its recognition of pericentric HC (36 , 37) . Thus, the effect of HP1{alpha} poly(ADP-ribosyl)ation would be basically to modulate its ability to bind HC, owing to electrostatic repulsion of the negatively charged ADP-ribose polymers present on HP1{alpha} from RNA. The normal accumulation of HP1{alpha} on pericentric HC observed in shParp-2 and shParp-1 cells could then be assigned to the compensating meH3K9-binding property of the CD of HP1{alpha}. Alternatively, the modification of the hinge domain could modulate the contact of HP1{alpha} with selective partners inside the Parp-1-Parp-2-TIF1β-HP1 complex that would reflect a dynamic equilibrium and important regulatory events occurring in this heterochromatic protein network. Accordingly, we found reduced coimmunoprecipitation of both Parp-2 and HP1{alpha} with TIF1β in the presence of the Parp inhibitor. Altogether it is tempting to propose Parp-1, Parp-2, and poly(ADP-ribose) as auxiliary factors that contribute to the dynamic nature of HP1 either 1) by facilitating its association/dissociation activity to chromatin, in addition to the previously described histone methyl transferases ACF1 or SUVAR39 (38) or 2) by modulating the protein interaction network at pericentric HC. As such, both Parps could regulate various HP1-dependent processes, including 1) transcriptional silencing of HC (39 , 40) ; 2) its participation in kinetochore formation and maintenance during chromosome segregation (41) ; or 3) its role in HC dynamics during DNA replication (5) .

Distinct functions of Parp-2 and Parp-1 in endodermal differentiation and TIF1β dependent transcriptional activity
Despite the established shared functions of Parp-1 and Parp-2 in cellular response to DNA damage and a similar contribution in the maintenance of HC integrity, both enzymes have distinct DNA and/or protein targets, which suggests that they might also play unique functions that have only started to be clarified. In the present study, we provide direct evidence that Parp-2 and Parp-1 display key specific functions throughout endodermal differentiation even though both proteins poly(ADP-ribosyl)ate HP1{alpha}.

The dynamic accumulation of Parp-2 to centromeric regions in PrE cells, where it colocalizes and interacts with TIF1β, combined with the observation that the depletion of Parp-2 or the inhibition of its activity significantly impairs the relocation of TIF1β to nuclear foci, provide compelling evidence that Parp-2 and the associated poly(ADP-ribose) synthesis participate in the targeting of TIF1β to pericentric HC. One major mechanism by which TIF1β is targeted to centromeric regions is through HP1 interaction (30) . This association is required for terminal differentiation of F9 to PE and VE cells (19) . The finding that TIF1β-HP1 association is not significantly disrupted in Parp-2-depleted cells argues for the existence of an additional Parp-2-regulated process involved in TIF1β compartmentalization during cellular differentiation. Accordingly, we show here that compromised TIF1β heterochromatic targeting is associated with impaired differentiation of shParp-2 into PrE cells, whereas terminal differentiation of the remaining cells is not affected. These observations indicate key functions of Parp-2 in initiating differentiation. It is conceivable that through TIF1β HC selective targeting, Parp-2 controls the expression of TIF1β-dependent yet-to-be-identified genes involved in early endodermal differentiation. In this respect, it is noteworthy that Parp-2 has recently been shown to associate with and to act as a transcriptional cofactor for PPAR{gamma} during adipocyte differentiation (27) and for TTF-1 during lung development (42) . In addition, we found a severe down-regulation of the endoderm-specific gene HNF4 in Parp-2-depleted and Parp-1-Parp-2-depleted cells similar to that previously observed in TIF1βHP1box/– cells but not in Parp-1-depleted cells. Therefore, Parp-2 participates in the regulation of TIF1β functions, including HC targeting and expression of endoderm-specific genes through a process that might be independent of TIF1β-HP1 interaction.

In contrast to Parp-2, we show that Parp-1 and the associated polymerizing activity are crucial for terminal differentiation of F9 cells into PE or VE cells even though Parp-1 does not interact with HP1{alpha} and only weakly binds to TIF1β. Furthermore, Parp-1 is at least partly required for stable HP1-TIF1β interaction. Given the essential role of TIF1β-HP1 interaction in terminal differentiation (19) , it seems likely that the abrogated differentiation of shParp-1 F9 cells to PE or VE cells could be a consequence of the unstable TIF1β-HP1 association observed in these cells. However, alternative roles of Parp-1 dependent-poly(ADP-ribosyl)ation of HP1{alpha} that might govern differentiation appear to be involved as a significant level of TIF1β-HP1 interaction remained reproducibly detected in shParp-1 F9 that is sufficient for the wild-type-like expression of HNF4.

Together, this work highlights key distinct functions of Parp-1 and Parp-2 in endodermal differentiation, although redundant activities cannot be excluded, as indicated by the similar up-regulation of the mesoderm-specific Mest gene in the absence of either Parp-1, Parp-2, or both proteins.

Toward a role of Parp-2 and Parp-1 in the histone subcode
It is well established that the dynamic regulation of chromatin structure and function is accomplished by a tuned combination of histone modifications, defined as the "histone code" (3) . In recent studies, poly(ADP-ribosyl)ation is turning out to be another of the many global epigenetic modulators involved in chromatin dynamics during physiological processes (28) .

More recently, a histone subcode hypothesis has been reported, which predicts that the posttranslational modification of HP1 provides a second regulatory layer of the chromatin code involved in transcriptional repression (9) . HP1 proteins can be extensively modified by phosphorylation, acetylation, methylation, sumoylation, and ubiquitination, similar to histones. Here we identified a selective poly(ADP-ribosyl)ation of HP1{alpha} by both Parp-1 and Parp-2, thus describing PAR as an emerging new modification of subcode proteins. In addition, this property points out Parp-1 and Parp-2 as new regulators of the histone subcode. The next challenge will be to identify the site-specific residue poly(ADP-ribosyl)ated in HP1{alpha}, with the aim of further deciphering the tight regulatory network that regulates these highly similar proteins.

In conclusion, the work described here identifies another step toward the role of TIF1β-HP1 association in endodermal differentiation, defined by distinct physical interactions of Parp-1 and Parp-2 with TIF1β and HP1 isotypes and a selective poly(ADP-ribosyl)ation of HP1{alpha}. A model describing Parp-1 and Parp-2 as essential players in the histone subcode can be proposed in which the selective poly(ADP-ribosyl)ation of HP1{alpha} is a major regulatory event required for TIF1β relocation on heterochromatic foci and TIF1β-HP1{alpha}-mediated transcriptional activity throughout differentiation (Fig. 8 ). Identifying the mechanism by which Parp activity is spontaneously induced during this process, as shown in Supplemental Fig. 3, remains an exciting challenge. In addition, these findings reinforce previous evidence that PAR plays fundamental roles in pericentric HC structure and integrity and open the way toward forthcoming fascinating issues aimed at understanding its contribution in HP1-mediated centromere function and cell division.


Figure 8
View larger version (17K):
[in this window]
[in a new window]

  		Figure 8. Model describing Parp-1 and Parp-2 as new components of the heterochromatic protein network at pericentric HC in PrE cells. By poly(ADP-ribosyl)ating HP1{alpha}, Parp-2 and Parp-1 can be defined as new actors of the HP1-mediated histone subcode required for TIF1β relocation on heterochromatic foci and TIF1β-HP1{alpha} association throughout endodermal differentiation, respectively. Both ways contribute to the expression of the endoderm-specific gene HNF4 and the mesoderm-specific gene Mest. In addition, the formation of this complex in undifferentiated F9 stem cells suggests additional important roles played by ADP-ribose modifications in controlling chromatin structure and activity during cell proliferation and/or division. Some of these effects might be mediated by HP1{alpha} and/or TIF1β.


   ACKNOWLEDGMENTS
 
We thank V. Schreiber for critical reading of the manuscript and A. van Dorsselear for access to the proteomic platform. This work was supported by Association pour la Recherche Contre le Cancer, Ligue Nationale Contre le Cancer et Comité Régional, Centre National de la Recherche Scientifique, and Agence Nationale de la Recherche.

Received for publication May 13, 2008. Accepted for publication June 26, 2008.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 

  1. Schneider, R., Grosschedl, R. (2007) Dynamics and interplay of nuclear architecture, genome organization, and gene expression. Genes Dev. 21,3027-3043[Abstract/Free Full Text]
  2. Narlikar, G. J., Fan, H. Y., Kingston, R. E. (2002) Cooperation between complexes that regulate chromatin structure and transcription. Cell 108,475-487[CrossRef][Medline]
  3. Jenuwein, T., Allis, C. D. (2001) Translating the histone code. Science 293,1074-1080[Abstract/Free Full Text]
  4. Grewal, S. I., Jia, S. (2007) Heterochromatin revisited. Nat. Rev. Genet. 8,35-46[CrossRef][Medline]
  5. Maison, C., Almouzni, G. (2004) HP1 and the dynamics of heterochromatin maintenance. Nat. Rev. Mol. Cell Biol. 5,296-304[CrossRef][Medline]
  6. Eissenberg, J. C., Elgin, S. C. (2000) The HP1 protein family: getting a grip on chromatin. Curr. Opin. Genet. Dev. 10,204-210[CrossRef][Medline]
  7. Lomberk, G., Wallrath, L., Urrutia, R. (2006) The heterochromatin protein 1 family. Genome Biol. 7,228[CrossRef][Medline]
  8. Singh, P. B., Georgatos, S. D. (2002) HP1: facts, open questions, and speculation. J. Struct. Biol. 140,10-16[CrossRef][Medline]
  9. Lomberk, G., Bensi, D., Fernandez-Zapico, M. E., Urrutia, R. (2006) Evidence for the existence of an HP1-mediated subcode within the histone code. Nat. Cell Biol. 8,407-415[CrossRef][Medline]
  10. Le Douarin, B., Nielsen, A. L., Garnier, J. M., Ichinose, H., Jeanmougin, F., Losson, R., Chambon, P. (1996) A possible involvement of TIF1 alpha and TIF1 beta in the epigenetic control of transcription by nuclear receptors. EMBO J. 15,6701-6715[Medline]
  11. Nielsen, A. L., Ortiz, J. A., You, J., Oulad-Abdelghani, M., Khechumian, R., Gansmuller, A., Chambon, P., Losson, R. (1999) Interaction with members of the heterochromatin protein 1 (HP1) family and histone deacetylation are differentially involved in transcriptional silencing by members of the TIF1 family. EMBO J. 18,6385-6395[CrossRef][Medline]
  12. Ryan, R. F., Schultz, D. C., Ayyanathan, K., Singh, P. B., Friedman, J. R., Fredericks, W. J., Rauscher, F. J., 3rd (1999) KAP-1 corepressor protein interacts and colocalizes with heterochromatic and euchromatic HP1 proteins: a potential role for Kruppel-associated box-zinc finger proteins in heterochromatin-mediated gene silencing. Mol. Cell Biol. 19,4366-4378[Abstract/Free Full Text]
  13. Sripathy, S. P., Stevens, J., Schultz, D. C. (2006) The KAP1 corepressor functions to coordinate the assembly of de novo HP1-demarcated microenvironments of heterochromatin required for KRAB zinc finger protein-mediated transcriptional repression. Mol. Cell Biol. 26,8623-8638[Abstract/Free Full Text]
  14. Schultz, D. C., Friedman, J. R., Rauscher, F. J., 3rd (2001) Targeting histone deacetylase complexes via KRAB-zinc finger proteins: the PHD and bromodomains of KAP-1 form a cooperative unit that recruits a novel isoform of the Mi-2alpha subunit of NuRD. Genes Dev. 15,428-443[Abstract/Free Full Text]
  15. Ayyanathan, K., Lechner, M. S., Bell, P., Maul, G. G., Schultz, D. C., Yamada, Y., Tanaka, K., Torigoe, K., Rauscher, F. J., 3rd (2003) Regulated recruitment of HP1 to a euchromatic gene induces mitotically heritable, epigenetic gene silencing: a mammalian cell culture model of gene variegation. Genes Dev. 17,1855-1869[Abstract/Free Full Text]
  16. Cammas, F., Mark, M., Dolle, P., Dierich, A., Chambon, P., Losson, R. (2000) Mice lacking the transcriptional corepressor TIF1beta are defective in early postimplantation development. Development 127,2955-2963[Abstract]
  17. Weber, P., Cammas, F., Gerard, C., Metzger, D., Chambon, P., Losson, R., Mark, M. (2002) Germ cell expression of the transcriptional co-repressor TIF1beta is required for the maintenance of spermatogenesis in the mouse. Development 129,2329-2337[Medline]
  18. Strickland, S., Mahdavi, V. (1978) The induction of differentiation in teratocarcinoma stem cells by retinoic acid. Cell 15,393-403[CrossRef][Medline]
  19. Cammas, F., Herzog, M., Lerouge, T., Chambon, P., Losson, R. (2004) Association of the transcriptional corepressor TIF1beta with heterochromatin protein 1 (HP1): an essential role for progression through differentiation. Genes Dev. 18,2147-2160[Abstract/Free Full Text]
  20. Schreiber, V., Dantzer, F., Ame, J. C., de Murcia, G. (2006) Poly(ADP-ribose): novel functions for an old molecule. Nat. Rev. Mol. Cell Biol. 7,517-528[CrossRef][Medline]
  21. Menissier de Murcia, J., Ricoul, M., Tartier, L., Niedergang, C., Huber, A., Dantzer, F., Schreiber, V., Ame, J. C., Dierich, A., LeMeur, M., Sabatier, L., Chambon, P., de Murcia, G. (2003) Functional interaction between PARP-1 and PARP-2 in chromosome stability and embryonic development in mouse. EMBO J. 22,2255-2263[CrossRef][Medline]
  22. Dantzer, F., Giraud-Panis, M. J., Jaco, I., Ame, J. C., Schultz, I., Blasco, M., Koering, C. E., Gilson, E., Menissier-de Murcia, J., de Murcia, G., Schreiber, V. (2004) Functional interaction between poly(ADP-Ribose) polymerase 2 (PARP-2) and TRF2: PARP activity negatively regulates TRF2. Mol. Cell. Biol. 24,1595-1607[Abstract/Free Full Text]
  23. Saxena, A., Saffery, R., Wong, L. H., Kalitsis, P., Choo, K. H. (2002) Centromere proteins Cenpa, Cenpb, and Bub3 interact with poly(ADP-ribose) polymerase-1 protein and are poly(ADP-ribosyl)ated. J. Biol. Chem. 277,26921-26926[Abstract/Free Full Text]
  24. Saxena, A., Wong, L. H., Kalitsis, P., Earle, E., Shaffer, L. G., Choo, K. H. (2002) Poly(ADP-ribose) polymerase 2 localizes to mammalian active centromeres and interacts with PARP-1, Cenpa, Cenpb and Bub3, but not Cenpc. Hum. Mol. Genet. 11,2319-2329[Abstract/Free Full Text]
  25. Meder, V. S., Boeglin, M., de Murcia, G., Schreiber, V. (2005) PARP-1 and PARP-2 interact with nucleophosmin/B23 and accumulate in transcriptionally active nucleoli. J. Cell Sci. 118,211-222[Abstract/Free Full Text]
  26. Ju, B. G., Solum, D., Song, E. J., Lee, K. J., Rose, D. W., Glass, C. K., Rosenfeld, M. G. (2004) Activating the PARP-1 sensor component of the groucho/ TLE1 corepressor complex mediates a CaMKinase IIdelta-dependent neurogenic gene activation pathway. Cell 119,815-829[CrossRef][Medline]
  27. Bai, P., Houten, S. M., Huber, A., Schreiber, V., Watanabe, M., Kiss, B., de Murcia, G., Auwerx, J., Menissier-de Murcia, J. (2007) PARP-2 controls adipocyte differentiation and adipose tissue function through the regulation of the activity of the RXR/PPARgamma heterodimer. J. Biol. Chem. 282,37738-37746[Abstract/Free Full Text]
  28. Dantzer, F., Mark, M., Quenet, D., Scherthan, H., Huber, A., Liebe, B., Monaco, L., Chicheportiche, A., Sassone-Corsi, P., de Murcia, G., Menissier-de Murcia, J. (2006) Poly(ADP-ribose) polymerase-2 contributes to the fidelity of male meiosis I and spermiogenesis. Proc. Natl. Acad. Sci. U. S. A. 103,14854-14859[Abstract/Free Full Text]
  29. Yelamos, J., Monreal, Y., Saenz, L., Aguado, E., Schreiber, V., Mota, R., Fuente, T., Minguela, A., Parrilla, P., de Murcia, G., Almarza, E., Aparicio, P., Menissier-de Murcia, J. (2006) PARP-2 deficiency affects the survival of CD4+CD8+ double-positive thymocytes. EMBO J. 25,4350-4360[CrossRef][Medline]
  30. Cammas, F., Oulad-Abdelghani, M., Vonesch, J. L., Huss-Garcia, Y., Chambon, P., Losson, R. (2002) Cell differentiation induces TIF1beta association with centromeric heterochromatin via an HP1 interaction. J. Cell Sci. 115,3439-3448[Abstract/Free Full Text]
  31. Miguet, L., Pacaud, K., Felden, C., Hugel, B., Martinez, M. C., Freyssinet, J. M., Herbrecht, R., Potier, N., van Dorsselaer, A., Mauvieux, L. (2006) Proteomic analysis of malignant lymphocyte membrane microparticles using double ionization coverage optimization. Proteomics 6,153-171[CrossRef][Medline]
  32. Farmer, H., McCabe, N., Lord, C. J., Tutt, A. N., Johnson, D. A., Richardson, T. B., Santarosa, M., Dillon, K. J., Hickson, I., Knights, C., Martin, N. M., Jackson, S. P., Smith, G. C., Ashworth, A. (2005) Targeting the DNA repair defect in BRCA mutant cells as a therapeutic strategy. Nature 434,917-921[CrossRef][Medline]
  33. Schreiber, V., Ame, J. C., Dolle, P., Schultz, I., Rinaldi, B., Fraulob, V., Menissier-de Murcia, J., de Murcia, G. (2002) Poly(ADP-ribose) polymerase-2 (PARP-2) is required for efficient base excision DNA repair in association with PARP-1 and XRCC1. J. Biol. Chem. 277,23028-23036[Abstract/Free Full Text]
  34. Nusinow, D. A., Hernandez-Munoz, I., Fazzio, T. G., Shah, G. M., Kraus, W. L., Panning, B. (2007) Poly(ADP-ribose) polymerase 1 is inhibited by a histone H2A variant, MacroH2A, and contributes to silencing of the inactive X chromosome. J. Biol. Chem. 282,12851-12859[Abstract/Free Full Text]
  35. Aldinucci, A., Gerlini, G., Fossati, S., Cipriani, G., Ballerini, C., Biagioli, T., Pimpinelli, N., Borgognoni, L., Massacesi, L., Moroni, F., Chiarugi, A. (2007) A key role for poly(ADP-ribose) polymerase-1 activity during human dendritic cell maturation. J. Immunol. 179,305-312[Abstract/Free Full Text]
  36. Maison, C., Bailly, D., Peters, A. H., Quivy, J. P., Roche, D., Taddei, A., Lachner, M., Jenuwein, T., Almouzni, G. (2002) Higher-order structure in pericentric heterochromatin involves a distinct pattern of histone modification and an RNA component. Nat. Genet. 30,329-334[CrossRef][Medline]
  37. Muchardt, C., Guilleme, M., Seeler, J. S., Trouche, D., Dejean, A., Yaniv, M. (2002) Coordinated methyl and RNA binding is required for heterochromatin localization of mammalian HP1alpha. EMBO Rep. 3,975-981[CrossRef][Medline]
  38. Eskeland, R., Eberharter, A., Imhof, A. (2007) HP1 binding to chromatin methylated at H3K9 is enhanced by auxiliary factors. Mol. Cell Biol. 27,453-465[Abstract/Free Full Text]
  39. Nielsen, S. J., Schneider, R., Bauer, U. M., Bannister, A. J., Morrison, A., O'Carroll, D., Firestein, R., Cleary, M., Jenuwein, T., Herrera, R. E., Kouzarides, T. (2001) Rb targets histone H3 methylation and HP1 to promoters. Nature 412,561-565[CrossRef][Medline]
  40. Richards, E. J., Elgin, S. C. (2002) Epigenetic codes for heterochromatin formation and silencing: rounding up the usual suspects. Cell 108,489-500[CrossRef][Medline]
  41. Obuse, C., Iwasaki, O., Kiyomitsu, T., Goshima, G., Toyoda, Y., Yanagida, M. (2004) A conserved Mis12 centromere complex is linked to heterochromatic HP1 and outer kinetochore protein Zwint-1. Nat. Cell Biol. 6,1135-1141[CrossRef][Medline]
  42. Maeda, Y., Hunter, T. C., Loudy, D. E., Dave, V., Schreiber, V., Whitsett, J. A. (2006) PARP-2 interacts with TTF-1 and regulates expression of surfactant protein-B. J. Biol. Chem. 281,9600-9606[Abstract/Free Full Text]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Supplemental Data
Right arrow All Versions of this Article:
fj.08-113464v1
22/11/3853    most recent
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Quénet, D.
Right arrow Articles by Dantzer, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Quénet, D.
Right arrow Articles by Dantzer, F.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS