HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 30, 26921-26926, July 26, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/30/26921    most recent M200620200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Saxena, A. Articles by Choo, K. H. A. Search for Related Content PubMed PubMed Citation Articles by Saxena, A. Articles by Choo, K. H. A. Social Bookmarking                      What's this?

Centromere Proteins Cenpa, Cenpb, and Bub3 Interact with Poly(ADP-ribose) Polymerase-1 Protein and Are Poly(ADP-ribosyl)ated^*

Alka Saxena, Richard Saffery, Lee H. Wong, Paul Kalitsis, and K. H. Andy Choo^§

From the  Murdoch Childrens Research Institute, Royal Children's Hospital, Flemington Rd., Parkville 3052, Australia

Received for publication, January 22, 2002, and in revised form, April 12, 2002

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Poly(ADP-ribose) polymerase-1 (PARP-1) is activated by DNA strand breaks during cellular genotoxic stress response and catalyzes poly(ADP-ribosyl)ation of acceptor proteins. These acceptor proteins include those involved in modulation of chromatin structure, DNA synthesis, DNA repair, transcription, and cell cycle control. Thus, PARP-1 is believed to play a pivotal role in maintaining genome integrity through modulation of protein-protein and protein-DNA interactions. We previously described the association of PARP-1 with normal mammalian centromeres and human neocentromeres by affinity purification and immunofluorescence. Here we investigated the interaction of this protein with, and poly(ADP-ribosyl)ation of, three constitutive centromere proteins, Cenpa, Cenpb, and Cenpc, and a spindle checkpoint protein, Bub3. Immunoprecipitation and Western blot analyses demonstrate that Cenpa, Cenpb, and Bub3, but not Cenpc, interacted with PARP-1, and are poly(ADP-ribosyl)ated following induction of DNA damage. The results suggest a role of PARP-1 in centromere assembly/disassembly and checkpoint control. Demonstration of PARP-1-binding and poly(ADP-ribosyl)ation in three of the four proteins tested further suggests that many more centromere proteins may behave similarly and implicates PARP-1 as an important regulator of diverse centromere function.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Centromeres form the primary constriction of mammalian metaphase chromosomes. They are the sites of kinetochore formation and microtubule attachment and mediate the faithful division of sister chromatids during mitotic cell division (1). Electron microscopy reveals kinetochore as a trilaminar structure that can be delineated into three separate domains, the inner, middle, and outer zones (2-5). A host of centromere-specific proteins reside at one or another of these zones to provide essential functions such as structural organization, microtubule binding, motor movement, cytokinesis, and checkpoint control (1, 6-12). These proteins can be generally classified as constitutive or transient proteins. Constitutive proteins (CENP-A, CENP-B, CENP-C, CENP-G, and CENP-H) reside at the centromere at all stages of the cell cycle, whereas the transient proteins (e.g. CENP-E, CENP-F, INCENP, survivin, MCAK, ZWINT-1, ZW10, MAD-1, MAD-2, BUB1, BUBR1, and BUB3) associate with the centromere during specific stages of the cell cycle. Constitutive proteins CENP-A, CENP-C, and CENP-H are essential for correct kinetochore assembly and function as evidenced by a lethal phenotype in gene knockout and antibody/RNA inhibition studies in mouse, worm, and chicken DT40 cells (13-21), whereas CENP-B appears to be functionally redundant for centromere activity (22-24). Numerous gene knockout and inhibition studies have demonstrated that many of the transient proteins have essential roles in normal mitotic functions, such as INCENP, survivin, CENP-E, BUB3, and MAD2 (25-30).

Poly(ADP-ribose) polymerase-1 (PARP-1)¹ is a multifunctional enzyme that catalyzes the formation of poly(ADP-ribose) polymers on acceptor proteins involved in the maintenance of chromatin structure and DNA repair (31-37). The addition of poly(ADP-ribose) units makes the acceptor proteins more negatively charged, thus altering their structure, function, and binding properties to DNA (38-40). Poly(ADP-ribosyl)ation of acceptor proteins signals DNA lesion to downstream effectors involved in coordinating the recruitment of DNA repair complexes to the site of DNA damage (41-48). PARP-1 is also thought to play a role in the maintenance of genomic stability under genotoxic stress. Cells from PARP-1 knockout mice show increased chromosomal breakages, high genomic instability, and decreased ability to repair DNA damage (49-54). The continued presence of poly(ADP-ribosyl)ation in cells devoid of PARP-1 has been attributed to the activities of other members of the PARP family, of which five have now been identified (55-61).

We have previously shown that PARP-1 accumulates at active mammalian centromeres (including neocentromeres) on metaphase chromosomes (62). As a first step toward deciphering the possible functions of PARP-1 at the centromere, we investigated the association of PARP-1 with several centromere-specific proteins. By co-immunoprecipitation, we have identified CENP-A, CENP-B, and the spindle checkpoint protein, BUB3, but not CENP-C, as binding partners of PARP-1 and have demonstrated poly(ADP-ribosyl)ation of CENP-A, CENP-B, and BUB3 upon induction of DNA damage in cells.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cell Cultures and Preparation of Nuclear Extracts-- All cell lines, including wild-type mouse embryonic stem (ES) cells W9.5, ES-GFP, and ES-Cenpa:GFP, were grown in ES medium supplemented with 10% fetal calf serum (Trace Biosciences), leukemia inhibitory factor (ESGRO^TM, Chemicon International), and -mercaptoethanol in the presence of 5% CO₂ at 37 °C. The ES-GFP cell line was generated by electroporation of ES cells with plasmid pEGFP-C1 (Clonetech), which contained a neomycin-resistant gene and a gene expressing green fluorescent protein (GFP) using a Gene Pulsar transfection apparatus (Bio-Rad). Cells were allowed to grow for 24 h before the addition of 250 µg/ml G418 (Invitrogen) for selection of clones. ES-Cenpa:GFP was a heterozygous ES cell line containing one copy of the mouse Cenpa-GFP fusion protein. This cell line was generated using a targeted knock-in strategy, where GFP was placed at the C terminus of Cenpa.²

For the preparation of nuclear protein extracts, 10⁷ cells were harvested and lysed in ice-cold lysis buffer as previously described (24).

Irradiation of ES Cells-- Exponentially dividing ES cells were harvested and subjected to 10 or 30 Gy of  irradiation at room temperature in an irradiator using cobalt 60 as a source at a rate of 30 Gy/1.2 min over a period of 0.4-1.2 min.

Antisera-- Antisera used in this study included mouse monoclonal anti-PARP-1 antibody C2-10 (Trevigen, Gaithersburg, MD), mouse monoclonal anti-poly(ADP-ribose) antibody (PAR) (Trevigen, Gaithersburg, MD), goat polyclonal anti-GFP antibody (Rockland, Gilbertsville, PA), mouse monoclonal Cenpb 2D-7 (2), rabbit polyclonal anti-Cenpc antiserum (63), and rabbit polyclonal anti-Bub3 antiserum (64). Horseradish peroxidase-conjugated secondary antibodies were purchased from Jackson Immunoresearch Laboratories (West Grove, PA).

Immunoprecipitation-- Nuclear protein extracts (500 µl) were incubated with 5 µg of antisera overnight at 4 °C with gentle agitation. The extract/antibody mixtures were then incubated with 125 µl of 50% protein G-Sepharose resuspended in incubation buffer (50 mM NaCl, 20 mM Tris, 5 mM EDTA, 0.1 mM phenylmethylsulfonyl fluoride) containing a mixture of protease inhibitors (Complete^TM; Roche Molecular Biochemicals) for 3 h at room temperature with constant agitation. The Sepharose beads were then pelleted by centrifugation and washed in immunoprecipitation wash buffer (50 mM Tris, 10 mM EDTA, and 150 mM NaCl), prior to elution with incubation buffer containing 1% SDS.

Western Blot Analysis-- Protein samples were subjected to PAGE on 10% gel and were transferred to Hybond C (Amersham Biosciences) by standard Western blotting. Blots were incubated with the relevant primary antisera at dilutions recommended by the manufacturers. After washing in phosphate-buffered saline containing 0.1% Tween 20, blots were incubated with relevant horseradish peroxidase-conjugated secondary antisera and developed using a chemiluminescence detection kit (Amersham Biosciences).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Cenpa Interacts with PARP-1 and Is Poly(ADP-ribosyl)ated-- To study the physical interaction of PARP-1 with centromere protein Cenpa, anti-PARP-1 antibody was used to immunoprecipitate the Cenpa:GFP fusion protein from nuclear cell extract prepared from the ES-Cenpa:GFP cell line (see "Materials and Methods"), followed by immunoblotting with anti-GFP antiserum. In a separate study,² the Cenpa:GFP fusion protein has been shown to localize specifically to mouse centromeres and provide normal centromere function in the ES-Cenpa:GFP heterozygous cell line. A parental ES cell line W9.5 and an ES-GFP cell line overexpressing GFP were included as controls in the immunoprecipitation assays. A band of 43 kDa corresponding to the Cenpa:GFP fusion protein was observed when the ES-Cenpa:GFP nuclear extract, but not extracts from the W9.5 or ES-GFP cells, was used for immunoprecipitation (Fig. 1A). To confirm the specificity of this interaction, anti-GFP antibody was used to immunoprecipitate nuclear extracts from the same cell lines, followed by immunoblotting with anti-PARP-1 antibody. Two bands of ~113 and 89 kDa, corresponding to the full-length and apoptotic fragments of PARP-1, were detected in the ES-Cenpa:GFP extract alone (Fig. 1B). These results therefore indicate that PARP-1 interacts with Cenpa.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   Western blot analysis of PARP-1 binding and poly(ADP-ribosyl)ation of Cenpa. A, nuclear extracts from ES-Cenpa:GFP fusion cell line, ES:GFP cell line, and control parental ES cell line W9.5 were immunoprecipitated using anti-PARP-1 antibody, Western blotted, and probed with a goat anti-GFP antibody. The solid arrow indicates the 43-kDa band corresponding to the Cenpa:GFP fusion product, seen only in the Cenpa:GFP cell line. The 27-kDa immunoglobulin light chain (open arrow) is seen in all three lanes. B, immunoprecipitation of W9.5 ES cell line, ES:GFP cell line, and ES-Cenpa:GFP cell line using anti-GFP antibody and probed with anti-PARP-1 antibody. Bands corresponding to full-length PARP-1 (large arrow) and its apoptotic fragment (small arrow) are seen only in the ES-Cenpa:GFP cell extract. C, immunoprecipitation of the Cenpa:GFP cell line irradiated with 10-Gy irradiation (left-hand lane) or no irradiation (0 Gy) using anti-PAR antibody and probed with anti-GFP antibody. A faint band corresponding to Cenpa:GFP (arrow) is detectable following irradiation but absent in nonirradiated cells. The open arrow indicates light chain of IgG. D, immunoprecipitation of cell lines treated with 10-Gy irradiation for the W9.5 cell line, ES-GFP cell line, and ES-Cenpa:GFP cell line using anti-PAR antibody or the ES-Cenpa:GFP cell line using anti-GFP antibody (right-hand lane). Immunoblotting was performed using anti-GFP antibody. The 43-kDa Cenpa:GFP protein band (solid arrow) is strongly present as expected in the anti-GFP immunoprecipitate and faintly in the anti-PAR immunoprecipitate of the ES-Cenpa:GFP cell line and is undetectable in both the control ES and ES:GFP cell lines. The open arrow indicates IgG light chain. E, the ES-Cenpa:GFP cell line was either not irradiated and immunoprecipitated using anti-PAR antibody or was irradiated at 30 Gy, followed by immunoprecipitation using anti-PAR or anti-GFP antibody and probed with anti-GFP antibody. The Cenpa:GFP band (solid arrow) is seen in both of the irradiated samples. The open arrow indicates IgG light chain.

Sequence analysis of Cenpa revealed the presence of a consensus motif for poly(ADP-ribose) or PAR-binding domain at the C terminus of the protein, a region also showing high homology to histone H3 (65) (Fig. 2). We directly investigated the poly(ADP-ribosyl)ation status of Cenpa using anti-PAR antibody in immunoprecipitation experiments. The results indicated no detectable levels of poly(ADP-ribosyl)ation on Cenpa in normal cells (Fig. 1, C and E; 0 Gy).

View larger version (20K): [in this window] [in a new window]   Fig. 2.   Presence of potential poly(ADP-ribose) binding sites in mouse centromere proteins Cenpa, Cenpb, and Bub3. The consensus PAR binding motif is shown in boldface type and starts at the N terminus with one or more basic amino acid residues Lys or Arg, followed by interspersion of basic (b, underlined) and hydrophobic (h, asterisk) residues, where x denotes a nonconsensus amino acid (69). Sequences from Cenpa, Cenpb, and Bub3 show similar amino acid distribution patterns (in boldface type) to the consensus PAR binding motif, in that basic residues (Lys/Arg (K/R), underlined) at the N-terminal position are followed by downstream basic residues (Lys/Arg/His (K/R/H), underlined) interspersed with hydrophobic residues (ACGVILMFYW, marked with an asterisk). The numbers show the N-terminal amino acid positions of the peptides for Mus musculus Cenpa, Cenpb, and Bub3 (NCBI accession numbers O35216, P27790, and Q9WVA3, respectively).

We next investigated the poly(ADP ribosyl)ation status of Cenpa following the induction of DNA damage. W9.5, ES-GFP, and ES-Cenpa:GFP cells were irradiated with either 10 Gy (Fig. 1, C and D) or 30 Gy (Fig. 1E) of  radiation followed by immediate preparation of nuclear extracts. These were subjected to immunoprecipitation with anti-PAR antibody and immunoblotting with anti-GFP antibody. As shown in Fig. 1, C and D, a low but reproducibly detectable level of poly(ADP-ribosyl)ation of the Cenpa:GFP protein was observed following irradiation at 10 Gy. The level of poly(ADP-ribosyl)ation of the Cenpa:GFP protein increased noticeably following irradiation at 30 Gy (Fig. 1E). These results indicate that Cenpa is significantly poly(ADP-ribosyl)ated following high dose irradiation.

Cenpb Interacts with PARP-1 and Is Poly(ADP-ribosyl)ated-- Anti-PARP-1 and anti-Cenpb antibodies were used to immunoprecipitate nuclear extracts from W9.5 cells, followed by immunoblotting with anti-Cenpb antibody. A band of 80-kDa corresponding to Cenpb was detected in immunoprecipitates using both of the primary antibodies, suggesting an interaction between PARP-1 and Cenpb (Fig. 3A). This result was confirmed by performing the reverse experiment in which the anti-PARP-1 and anti-Cenpb immunoprecipitates were immunoblotted with anti-PARP-1 antibody. The results indicated the expected detection of the 113-kDa full-length PARP-1 and 89-kDa apoptotic fragment in both of the samples (Fig. 3B).

View larger version (27K): [in this window] [in a new window]   Fig. 3.   Western blot analysis of PARP-1 binding and poly(ADP-ribosyl)ation of Cenpb. A, Nuclear extracts from wild-type ES cell line W9.5 were immunoprecipitated with anti-PARP-1 or anti-Cenpb antibody and probed with anti-Cenpb antibody. The 80-kDa band (solid arrow) corresponding to Cenpb is seen in both immunoprecipitates. The open arrow shows IgG heavy chain. B, as A except probed with anti-PARP-1 antibody. Bands corresponding to the full-length (large arrow) and apoptotic (small arrow) fragment of PARP-1 are seen in both lanes. C, immunoprecipitates obtained using anti-PAR antibody on W9.5 cells irradiated with 10 Gy or nonirradiated (0 Gy) and probed with anti-Cenpb antibody. The Cenpb band (arrow) is seen in the irradiated cell line only. The large and small open arrows represent heavy and light chains of IgG, respectively. D, 10-Gy -irradiated W9.5 cells immunoprecipitated with anti-PAR or anti-Cenpb antibody and probed with anti-Cenpb antibody. The Cenpb band (arrow) is clearly seen in both lanes.

Sequence analysis also revealed a possible PAR consensus domain in Cenpb (Fig. 2). Direct detection of poly(ADP-ribose) residues in normal W9.5 cells gave a negative result indicating an absence of significant poly(ADP-ribosyl)ation of this protein under normal conditions (Fig. 3C, 0 Gy). When the cells were irradiated at 10 Gy before immunoprecipitation with either anti-PAR or anti-Cenpb antibody and immunoblotted with anti-Cenpb antibody, Cenpb was clearly detected in both of the immunoprecipitates (Fig. 3, C and D), indicating that Cenpb was poly(ADP-ribosyl)ated following 10 Gy of -radiation.

Cenpc neither Interacts with PARP-1 nor Becomes Poly(ADP-ribosyl)ated-- Nuclear extracts from W9.5 cells immunoprecipitated with either anti-PARP-1 or anti-Cenpc antibody were subjected to Western blotting with either anti-PARP-1 antibody (Fig. 4A) or anti-Cenpc antibody (Fig. 4B). Both the 113-kDa full-length PARP-1 and the 89-kDa apoptotic fragment were present when anti-PARP-1 antibody was used in the immunoprecipitation but were absent when anti-Cenpc was used instead (Fig. 4A). Similarly, in the reverse experiment, the 140-kDa Cenpc band was not detected in the anti-PARP-1 immunoprecipitate but was observed in the anti-Cenpc immunoprecipitate (Fig. 4B). Analysis of poly(ADP-ribosyl)ation status further indicated that Cenpc was not immunoprecipitated by anti-PAR antibody in nuclear extracts from both the nonirradiated ES cells (Fig. 4C; 0 Gy) or cells irradiated with either 10 Gy (Fig. 4, C and D) or 30 Gy (data not shown). These combined results suggest that Cenpc does not interact with PARP-1 and is not poly(ADP-ribosyl)ated at detectable levels even following major DNA damage. In support of the latter conclusion, sequence analysis of Cenpc failed to show the presence of a consensus PAR-binding domain.

View larger version (28K): [in this window] [in a new window]   Fig. 4.   Western blot analysis of PARP-1 binding and poly(ADP-ribosyl)ation of Cenpc. A, nuclear extracts from wild-type W9.5 ES cells were immunoprecipitated with anti-Cenpc or anti-PARP-1 antibody and probed with anti-PARP-1 antibody. Bands corresponding to the full-length (large solid arrow) and apoptotic (small solid arrow) fragment of PARP-1 are seen in anti-PARP-1 but not anti-Cenpc immunoprecipitate. B, W9.5 immunoprecipitated with anti-Cenpc or anti-PARP-1 antibody and probed with anti-Cenpc antibody. The band corresponding to Cenpc at 140 kDa (solid arrow) is present only in the Cenpc-immunoprecipitated sample. C, immunoprecipitates obtained using anti-PAR antibody on W9.5 cells irradiated with 10 Gy or nonirradiated (0 Gy) and probed with anti-Cenpc antibody. The band corresponding to Cenpc is undetectable in both lanes. D, 10-Gy -irradiated W9.5 cells immunoprecipitated with anti-Cenpc or anti-PAR antibody and probed with anti-Cenpc antibody. The Cenpc band (arrow) is seen in the Cenpc-immunoprecipitated sample but not in the PAR immunoprecipitate. The large and small open arrows show heavy and light chains of IgG, respectively.

Bub3 Interacts with PARP-1 and Is Poly(ADP-ribosyl)ated-- Bub3 is a mitotic checkpoint protein that has previously been shown to accumulate at the centromere transiently (66). We were interested in determining the association between PARP-1 and Bub3. When nuclear extracts from W9.5 cells were immunoprecipitated with anti-PARP-1 or anti-Bub3 antibody and probed with anti-Bub3 antibody, a band of 40 kDa corresponding to the Bub3 protein was detected in both of the immunoprecipitates (Fig. 5A). In the reverse experiment, where the same two immunoprecipitates were Western blotted and probed with anti-PARP-1 antibody, the bands corresponding to the full-length and apoptotic forms of PARP-1 were detected (Fig. 5B). These results therefore demonstrate a physical interaction between PARP-1 and Bub3.

View larger version (13K): [in this window] [in a new window]   Fig. 5.   Western blot analysis of PARP-1 binding and poly(ADP-ribosyl)ation of Bub3. A, nuclear extracts from wild-type ES cell line W9.5 were immunoprecipitated with anti-PARP-1 or anti-Bub3 antibody and probed with anti-Bub3 antibody. The 40-kDa band (solid arrow) corresponding to Bub3 is seen in both immunoprecipitates. The open arrow shows IgG light chain. B, as in A, except probed with anti-PARP-1 antibody. Bands corresponding to the full-length (large arrow) and apoptotic (small arrow) fragment of PARP-1 are seen in both lanes. C, 10-Gy -irradiated W9.5 cells, non--irradiated W9.5 cells (0 Gy) immunoprecipitated with anti-PAR antibody and probed with anti-Bub3 antibody. The band corresponding to Bub3 is present only in the irradiated cells. The light chain of IgG (open arrow) is seen in both lanes.

As with Cenpa and Cenpb, sequence analysis of Bub3 showed the presence of hydrophobic amino acids interspersed with basic amino acids, with an accumulation of basic residues at the N-terminal side of this motif, suggesting the possible presence of a PAR consensus domain (Fig. 2). Use of anti-PAR antibody in immunoprecipitation and Western blot analysis similarly detected no measurable levels of poly(ADP-ribosyl)ation in nonirradiated W9.5 cells (Fig. 5C, 0 Gy). Following 10 Gy of irradiation and upon immunoprecipitation with anti-PAR antibody and immunoblotting with anti-Bub3 antibody, a protein band of 40 kDa corresponding to Bub3 was detected (Fig. 5C), indicating that the Bub3 protein is poly(ADP-ribosyl)ated.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The post-translational modification of centromere-associated proteins is now widely accepted as a mechanism for regulating kinetochore assembly and centromere activity. Phosphorylation of several proteins, including CENP-A (67), CENP-E (68), 3F3/2 epitope (69), and topoisomerase IIA (70), has been reported, and protein kinases, including BUBR1, MPS1, aurora, and polo-like kinases, associate directly with the mammalian centromere structure (71-74). Farnesylation (post-translational modification of proteins by the addition of isoprenoids) of CENP-E and CENP-F has recently been reported and has been shown to be essential for the correct attachment of centromeres to the mitotic spindle (75). Ubiquitination of centromere-associated proteins also plays a pivotal role in the progression from metaphase to anaphase (76).

Poly(ADP-ribosyl)ation of proteins by PARP-1 is a conserved post-translational modification implicated in DNA repair, apoptosis, regulation of transcription, anti-recombination, scaffold attachment, and modulation of chromatin structure (33, 45, 47, 77-81). We have previously demonstrated the association of PARP-1 with mammalian centromeric DNA and active centromeres, providing circumstantial evidence for PARP-1 function at the centromere (62). In the present study, we show that PARP-1 interacts directly with Cenpa, Cenpb, and Bub3 under normal conditions. We further demonstrate that whereas no significant poly(ADP-ribosyl)ation of these proteins is measurable in unchallenged cells, all three proteins undergo poly(ADP-ribosyl)ation upon DNA damage induced by -irradiation. In contrast, Cenpc does not appear to either interact with PARP-1 or undergo poly(ADP-ribosyl)ation with or without induction of DNA damage. Given that 3 of 4 proteins tested proved to be substrates for PARP-1 binding and poly(ADP-ribosyl)ation and that in excess of 40 proteins have now been shown to associate with the mammalian centromere (1, 7-10, 63, 64, 82), it is likely that other similarly behaving proteins have yet to be identified. This together with the fact that poly(ADP-ribosyl)ation of proteins generally changes activities or functions of these proteins suggests that PARP-1 is likely to be involved in modulating many different aspects of centromere behavior, including kinetochore assembly/disassembly (e.g. by Cenpa and Cenpb) and spindle checkpoint control (e.g. by Bub3).

CENP-A replaces histone H3 in at least a subset of centromeric nucleosomes (83, 84). It is present at all active centromeres and neocentromeres (64, 85) and plays an essential role in the targeting of Cenpc to the centromere (19, 86). CENP-A is thus critical for the correct functioning of all centromeres. Although an interaction between PARP-1 and Cenpa is seen in the absence of DNA damage, a significant level of poly(ADP-ribosyl)ation of Cenpa is detected only after high doses of irradiation. This suggests that the site of poly(ADP-ribosyl)ation on Cenpa may be inaccessible or remain dormant in centromeric nucleosomes in the absence of an extreme assault. We detected a consensus poly(ADP-ribosyl)ation site within the histone fold domain of Cenpa, which is a conserved region between Cenpa and histone H3. In noncentromeric nucleosomes, this region on histone H3 is located deep within the final nucleosomal structure (87). If a similar positioning occurs in the centromeric nucleosomes with Cenpa, this may prevent access to poly(ADP-ribosyl)ation.

The discovery of PARP-1 binding and poly(ADP-ribosyl)ation of Cenpb is interesting given that a functional role of this protein at the centromere has not been clearly identified. CENP-B binds to a highly conserved 17-bp "CENP-B box" motif located in a subset of centromeric satellites and has been proposed to have a role in the higher order organization of these satellite repeats (88, 89). This protein is not detected on some functional centromeres (including neocentromeres), and its complete absence such as in Cenpb null mice does not disrupt normal centromere function (22-24, 64, 90). It is possible, however, that Cenpb may be replaced by a functional homologue (91) in its absence.

Following poly(ADP-ribosyl)ation, the negatively charged ADP-ribose polymers generally render acceptor proteins incapable of DNA binding (32, 45, 81, 92-95). A host of chromatin-associated proteins, including all of the normal histone subunits, topoisomerase I and II, high mobility group proteins, transcription factors, DNA synthesome complex proteins like DNA ligase and DNA polymerase  and , and others, are known to undergo this modification (45, 81). Such a modification of chromatin-associated proteins results in "opening up" of the chromatin to allow access to the DNA (95). The detection of poly(ADP-ribosyl)ation in two constitutive centromere proteins, Cenpa and Cenpb, following radiation insult suggests a possible primary and/or secondary mechanism for the decondensation of the highly compact chromatin structures of the centromere for DNA repair. It is possible that such a role is also maintained under normal physiological conditions to provide the cells with a mechanism for dealing with the small amounts of DNA damage caused by normal DNA processes or exposure to DNA-destabilizing agents such as free oxygen radicals and UV radiation (42, 81, 96-101).

Bub3 undergoes a transient association with centromeres prior to the commencement of anaphase and is involved in the mitotic spindle checkpoint function until all chromosomes are correctly aligned at the metaphase plate (102). We have identified Bub3 as a substrate for poly(ADP-ribosyl)ation. This modification may affect the interaction of this protein with BUB1 and BUBR1 (103, 104), thereby altering the functioning of the spindle checkpoint and anaphase progression. It is interesting to speculate that poly(ADP-ribosyl)ation of Bub3 might signal kinetochore damage to the spindle checkpoint. Alternatively, during DNA repair, when the kinetochore is disassembled, poly(ADP-ribosyl)ation may keep the mitotic checkpoint suppressed.

This study represents a first step toward delineating the role of PARP-1 in centromere function. The demonstration of poly(ADP-ribosyl)ation as a possible regulator of both constitutive kinetochore proteins and those involved in spindle checkpoint control adds another level of complexity to the mechanisms underlying centromere assembly and function. Further studies should determine the poly(ADP-ribosyl)ation status of other known centromere proteins. The absence of any drastic centromeric phenotype in PARP-1 null mice (51-54, 105-108) suggests some functional redundancy of PARP-1 at the centromere. It will be interesting to investigate the binding behavior of other members of the PARP family toward the centromere and its many proteins.

	ACKNOWLEDGEMENTS

We thank J. Dix for use of the  irradiator and P. Farlei and D. Magliano for advice on Western analysis.

	FOOTNOTES

^* This work was supported by a grant from National Health and Medical Research Council of Australia (to K. H. A. C.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

^§ Senior Principal Research Fellow of the National Health and Medical Research Council of Australia. To whom correspondence should be addressed. Tel.: 61-3-8341-6306; Fax: 61-3-9348-1391; E-mail: choo@ cryptic.rch.unimelb.edu.au.

Published, JBC Papers in Press, May 13, 2002, DOI 10.1074/jbc.M200620200

² P. Kalitsis, E. Earle, K. Fowler, and K. M. A. Choo, manuscript in preparation.

	ABBREVIATIONS

The abbreviations used are: PARP, poly(ADP-ribose) polymerase; PAR, poly(ADP-ribose); GFP, green fluorescent protein; ES, embryonic stem; Gy, gray(s).

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CiteULike

Complore