Tài liệu Báo cáo khoa học: Poly(ADP-ribose) The most elaborate metabolite of NAD+ Alexander Burkle pptx

MINIREVIEW Poly(ADP-ribose) The most elaborate metabolite of NAD + Alexander Bu ¨ rkle Department of Biology, University of Konstanz, Germany Introduction The life cycle of poly(ADP-ribose) NAD + ⁄ NADH is among the most versatile biomole- cules, as it can be used not only as a coenzyme for a large number of oxidoreduction reactions, but in its oxidized version can also serve as substrate for several different of ADP-ribosyl transfer reactions, which are the overarching theme of this minireview series. The covalent transfer onto glutamic acid, aspartic acid or lysine residues of target proteins (‘acceptors’), followed by successive transfer reactions onto the protein– mono(ADP-ribosyl) adduct, and subsequently onto the emerging chain of several covalently linked ADP-ribo- syl residues is the basis of the formation of poly(ADP- ribose), which can be regarded the cell’s most elaborate metabolite of NAD + [1]. ADP-ribose chains may com- prise up to 200 ADP-ribose units, coupled via unique ribose (1¢¢fi2¢) ribose phosphate-phosphate linkages and display several branching points resulting from the formation of ribose (1¢¢¢fi2¢¢) ribose linkages (Fig. 1). Poly(ADP-ribosyl)ation occurs in multicellular organ- isms including plants and some lower unicellular eu- karyotes, but is absent in prokaryotes and yeast. Poly(ADP-ribosyl)ation is catalysed by the family of poly(ADP-ribose) polymerases (PARPs; Fig. 2), enco- ded in human cells by a set of 18 different genes [2]. Keywords PARP; tankyrase; poly(ADP-ribose); DNA damage; DNA repair; genomic instability; centrosome; centromere; telomeres; mitotic spindle Correspondence A. Bu ¨ rkle, Department of Biology, Box X911, University of Konstanz, D-78457 Konstanz, Germany Tel: +49 7531 884035 Fax: +49 7531 884033 E-mail: alexander.buerkle@uni-konstanz.de Website: http://gutenberg.biologie. uni-konstanz.de/ (Received 5 May 2005, accepted 14 July 2005) doi:10.1111/j.1742-4658.2005.04864.x One of the most drastic post-translational modification of proteins in eu- karyotic cells is poly(ADP-ribosyl)ation, catalysed by a family enzymes termed poly(ADP-ribose) polymerases (PARPs). In the human genome, 18 different genes have been identified that all encode PARP family members. Poly(ADP-ribose) metabolism plays a role in a wide range of biological structures and processes, including DNA repair and maintenance of genomic stability, transcriptional regulation, centromere function and mito- tic spindle formation, centrosomal function, structure and function of vault particles, telomere dynamics, trafficking of endosomal vesicles, apoptosis and necrosis. In this article, the most recent advances in this rapidly grow- ing field are summarized. Abbreviations ANK, ankyrin; BER, base excision repair; BRCA1, breast cancer 1 protein; DBD, DNA-binding domain; HPS, His-Pro-Ser-rich; IRAP, insulin- responsive amino peptidase; MVP, major vault protein; NuMa, nuclear mitotic apparatus protein; PARG, poly(ADP-ribose) glycohydrolase; PARP, poly(ADP-ribose) polymerase; RNP, ribonucleoprotein particle; Sir2, silent information regulator 2; TCDD, 2,3,7,8-tetrachlorodibenzo- p-dioxin; TRF, telomeric-repeat binding factor. 4576 FEBS Journal 272 (2005) 4576–4589 ª 2005 FEBS Apart from the covalent modification of acceptor proteins, poly(ADP-ribose) has also been shown to interact noncovalently, yet specifically, with a wide range of proteins [3]. This interaction is mediated through a sequence motif displaying a conserved pat- tern, and it is interesting to note that DNA damage checkpoint proteins such as p53 and p21 possess such poly(ADP–ribose) interaction domains. DNA methyl- transferase-1 (DNMT1) is also able to bind long and branched ADP-ribose polymers in a noncovalent fash- ion, which leads to inhibition of DNMT1 activity [4]. Taken together, poly(ADP-ribose) can also affect the function of proteins that are not direct modification targets. Enzymes involved in poly(ADP-ribose) metabolism PARP-1 The prototypic enzyme of the PARP family is poly (ADP-ribose) polymerase-1 (PARP-1; EC 2.4.2.30; Table 1). Its discovery was made four decades ago by Chambon, Weill and Mandel [5] and marked the birth of a most interesting and intriguing field originally occupied by biochemists, but later on also attracting radiobiologists, toxicologists, geneticists, molecular biologists, pharmacologists, cell biologists and experts from other biological disciplines. Nicotinamide NAD + P AR P AR n NAD + n Nicotinamide + n H + Rib O O PO O P O Rib O - O - N N NH 2 N N Acceptor protein CH 2 C O O - Rib N N NH 2 N N O O O - PO O O - P O Rib Rib N N NH 2 N N O O O - PO O O - P O Ri b Rib Rib N N NH 2 N N O O O - PO O O - P O Rib Rib N N NH 2 N N O O O - PO O O - P O Rib Rib N N NH 2 N N O O O - PO O O - P O Rib Rib N + H 2 NCO N H 2 NCO + H + Rib N N NH 2 N N O O O - PO O O - P O Rib + CH 2 C O O CH 2 C O O Acceptor protein Acceptor protein Fig. 1. Poly(ADP-ribose) structure. Poly(ADP-ribose) polymerases (PARPs) cleave the glycosidic bond of NAD + between nicotinamide and ribose followed by the covalent modification of mainly glu- tamate residues of acceptor proteins with an ADP-ribosyl unit. PARPs also catalyse an adduct elongation, giving rise to linear polymers with chain lengths of up to about 200 ADP-ribosyl units, characterized by their unique ribose (1¢¢fi2¢) ribose phos- phate-phosphate backbone. At least some of the PARP family members also catalyse a branching reaction by creating ribose (1¢¢¢fi2¢¢) ribose linkages. The sites of hydrolysis catalysed by poly(ADP-ribose) glycohydrolase (PARG), the major poly(ADP- ribose)-degrading enzyme, are indicated by arrows. A. Bu ¨ rkle Poly(ADP-ribosyl)ation FEBS Journal 272 (2005) 4576–4589 ª 2005 FEBS 4577 PARP-1 is a highly conserved and abundant nuclear protein, which is catalytically active as a dimer and is the major acceptor protein in intact cells, via the so-called automodification reaction. A range of other nuclear proteins can also serve as acceptor proteins (see below). PARP-1 displays a characteristic three- domain structure (Fig. 2), which can be further broken down into modules A–F [6]. The N-terminal 42 kDa DNA-binding domain (DBD) also comprises the pro- tein’s nuclear localization signal and is adjacent to a central 16 kDa automodification domain. The 55 kDa catalytic domain, which includes the active site, is located at the C-terminus. The DBD of PARP-1 binds to single- or double- strand breaks with high affinity via two zinc fingers but has also been reported to be involved in protein– protein interactions. The first zinc finger is essential for PARP-1 activation by DNA strand breaks, whereas the second is essential for PARP-1 activation by DNA single-strand breaks but not double-strand breaks. In the absence of DNA breaks, PARP-1 displays a very low basal enzyme activity. The molecular mechanism linking DNA binding to catalytic activation is unknown. Automodification of PARP-1 was reported to be mostly induced by single-strand breaks, whereas histone H1 seems to be modified preferentially when PARP-1 binds to double-strand breaks [7]. Following up earlier reports [8–10], several groups have provided recent evidence for alternative activation mechanisms for PARP-1, independent of DNA strand breakage. PARP-1 was reported to recognise distortions in the Table 1. Human enzymes involved in poly(ADP-ribose) formation or degradation and their genes. Official gene symbol Name [aliases] Chromosomal location Gene ID Protein size PARP1 Poly(ADP-ribose) polymerase family, member 1 [PARP-1, ADPRT, ADPRT1, PARP, PPOL, pADPRT-1] 1q41-q42 142 1014 aa (113 kDa) PARP2 Poly(ADP-ribose) polymerase family, member 2 [PARP-2, NAD + poly(ADP-ribose) polymerase-2] 14q11.2-q12 10038 583 aa (66 kDa) PARP3 Poly(ADP-ribose) polymerase family, member 3 [PARP-3, NAD + poly(ADP-ribose) polymerase-3] 3p22.2-p21.1 10039 532 aa (60 kDa); splice variant 539 aa (60.8 kDa) PARP4 Poly(ADP-ribose) polymerase family, member 4 [PARP4, ADPRTL1, PARPL, PH5P, VAULT3, VPARP, p193] 13q11 143 1724 aa (192.8 kDa) TNKS Tankyrase, TRF1-interacting ankyrin-related ADP-ribose polymerase [PARP-5a, PARP5A, PARPL, TIN1, TINF1, TNKS1] 8p23.1 8658 1327 aa (142 kDa) TNKS2 Tankyrase-2, TRF1-interacting ankyrin-related ADP-ribose polymerase 2 [PARP-5b, PARP-5c, PARP5B, PARP5C, TANK2, TNKL] 10q23.3 80351 1166 aa (126.9 kDa) TIPARP TCDD-inducible poly(ADP-ribose) polymerase [PARP-7] 3q25.31 25976 657 aa (75 kDa) PARP10 Poly (ADP-ribose) polymerase family, member 10 [PARP-10] 8q24.3 84875 1025 aa (150 kDa) PARG Poly(ADP-ribose) glycohydrolase [PARG] 10q11.23 8505 Three splice variants 976 aa (111 kDa); 893 aa (102 kDa); 866 aa (99 kDa) PARP-1 PARP-2 PARP-3 PARP-4 (VPARP) Tankyrase (PARP-5a) Tankyrase-2 (PARP-5b) N BRCT NLSZFI ZFII Module A B C E DF DNA-binding domain (DBD) NAD + -binding domain Automodification domain Active site DBD NLS BRCT SA 24 ankyrin HPS SA 24 ankyrin VIT TM VWA SH3 MVP-BD C Fig. 2. Structural organization of some PARP protein family mem- bers. BRCT, BRCA1 C-terminus; DBD, DNA-binding domain; HPS, His-Pro-Ser-rich domain; MVP-BD major vault binding domain; NLS, nuclear localization signal; SAM, sterile a-module; SH3, src homo- logy region; TM, transmembrane domain; VIT, vault protein inter- alpha-trypsin domain; VWA, von Willebrand factor type A domain; ZF; zinc finger. Poly(ADP-ribosyl)ation A. Bu ¨ rkle 4578 FEBS Journal 272 (2005) 4576–4589 ª 2005 FEBS DNA helical backbone and to bind to three- and four- way junctions as well as to stably unpaired regions in double-stranded DNA [11]. Such PARP-1 interactions with non-B DNA structures led to catalytic activation of the enzyme in the absence of free DNA ends. DNA hairpins, cruciforms and stably unpaired regions are all effective coactivators of PARP-1 automodification and poly(ADP-ribosyl)ation of histone H1. These data suggest a link between PARP-1 binding to non-B DNA structures in the genome and its function in the dynamics of local modulation of chromatin structure in normal cellular physiology (see below). Binding to and activation by unbroken DNA in the context of chromatin has also been reported to play a crucial role in the NAD + -dependent modulation of chromatin structure and transcriptional regulation, mediated by PARP-1 [12]. Along the same lines it was suggested that fast and transient decondensation of chromatin structure by poly(ADP-ribosyl)ation occurring in Aply- sia neurones in the absence of DNA strand breaks enables the transcriptional regulation needed to form long-term memory in this organism [13]. The automodification domain of PARP-1 is rich in glutamic acid residues, consistent with the fact that poly(ADP-ribosy)lation occurs on such residues. This domain also comprises a breast cancer 1 protein (BRCA1) C-terminus (BRCT) motif that is present in many DNA damage repair and cell-cycle checkpoint proteins. The C-terminal 55 kDa catalytic domain contains the residues essential for NAD + -binding, ADP-ribosyl transfer and branching reactions. The crystal structure of the C-terminal catalytic fragment revealed a striking homology with bacterial toxins that act as mono(ADP- ribosyl) transferases [14]. While PARP-1 is constitutively expressed, its charac- teristic ability of being activated by DNA strand breaks makes poly(ADP-ribosyl)ation an immediate and drastic cellular response to DNA damage as induced by ionizing radiation, alkylating agents and oxidants. In the absence of DNA single and double strand breaks, poly(ADP-ribosyl)ation seems to be a very rare event in live cells, but it can increase over 100-fold upon DNA damage [15]. Under these condi- tions about 90% of poly(ADP-ribose) is synthesized by PARP-1 [16]. Among many identified interaction partners of PARP-1 are also other members of the PARP-family, such as PARP-2 [17–19] and PARP-3 [20]. As men- tioned above, the most prominent target protein (acceptor) of this poly(ADP-ribosyl)ation reaction is PARP-1 itself but many other acceptor proteins have been described, including p53 [21], both subunits of NF-jB [22], histones, DNA-topoisomerases and the catalytic subunit of DNA-dependent protein kinase (DNA-PK cs ). Due to the high negative charge of the polymer, this modification significantly alters the phys- ical and biochemical properties of the modified pro- teins, such as their DNA-binding affinity, and it is likely that such alteration will have a regulatory func- tion concerning the interaction with other proteins [23]. Three different PARP-1 knockout mouse models (Parp1 – ⁄ – ) have been created independently [24–26], lacking the PARP-1 protein. Parp1 – ⁄ – mice are viable and fertile, but show hypersensitivity to alkylation treatment or ionizing radiation and loss of genomic stability. In stark contrast, however, they display pro- tection against various pathophysiological phenomena, such as lipopolysaccharide-induced septic shock or streptozotocin-induced diabetes [27]. PARP-2 Apart from PARP-1, PARP-2 is the only other PARP isoform known to be activated by DNA strand breaks [28] (Table 1; Fig. 2). This enzyme displays automodifi- cation properties similar to PARP-1. Its DBD, how- ever, is different from that of PARP-1, consisting of only 64 amino acids and lacking any obvious DNA- binding motif. The crystal structure of the catalytic fragment of murine PARP-2 has recently been solved, thus providing a basis for the development of isoform- specific inhibitors by rational drug design [29]. Despite major structural differences between PARP-1 and PARP-2, including size and the absence of zinc fingers or the BRCT motif, they are both targeted to the nuc- leus, and they bind to and become activated by DNase I-treated DNA. Both PARP-2 and PARP-1 can homo- and heterodimerise, and both are involved in the base excision repair (BER) pathway where they form a complex with X-ray cross-complementing factor 1 (XRCC1) [18]. Parp1 – ⁄ – ⁄ Parp2 – ⁄ – double mutant mice are nonviable and die at the onset of gastrulation, highlighting that the expression of both PARP-1 and PARP-2 and ⁄ or DNA strand break-dependent poly(ADP-ribosyl)ation is essential during early embryo- genesis. PARP-3 The protein domain structure of PARP-3 is very similar to PARP-2, featuring a small DNA-binding domain consisting of only 54 amino acids and compri- sing a centrosome-targeting motif [20]. Overexpression of PARP-3 or its N-terminal domain in HeLa cells A. Bu ¨ rkle Poly(ADP-ribosyl)ation FEBS Journal 272 (2005) 4576–4589 ª 2005 FEBS 4579 interfered with the G 1 ⁄ S cell cycle transition. PARP-3 catalyses the synthesis of poly(ADP-ribose) in vitro and in purified centrosome preparations, and forms a stable complex with PARP-1, in agreement with other reports of PARP-1 localization at the centrosome [30,31]. PARP-4 Vault particles are cytoplasmic ribonucleoprotein parti- cles (RNPs) composed of several small untranslated RNA molecules and three proteins of 100, 193 and 240 kDa. With a total mass of 13 MDa, vaults are the largest RNP complexes found in the cytoplasm of mammalian cells. The 193 kDa vault protein was iden- tified as a novel PARP [32,33] and is now termed PARP-4 (Table 1, Fig. 2). PARP-4 poly(ADP-ribo- syl)ates the p100 subunit (major vault protein; MVP) within the vault particle and to a lesser extent itself. The N-terminal region of PARP-4 contains a BRCT domain similar to the automodification domain of PARP-1, suggestive of a related function. Tankyrase This protein was initially identified through its inter- action with the telomeric-repeat binding factor 1 (TRF1), which is another negative regulator of telo- mere length [34]. The N-terminus of tankyrase contains a so-called His-Pro-Ser-rich (HPS) domain consisting of stretches of consecutive histidine, proline and serine residues, followed by 24 ankyrin (ANK) repeats, which is a structural feature only found in tankyrase and tankyrase-2 (see below) within the known members of the PARP family. Adjacent to the ANK domain is another protein interaction motif, the sterile alpha- module. The C-terminus of tankyrase displays homo- logy to the PARP-1 catalytic region. Consistent with the absence of any DNA-binding domain, tankyrase activity does not depend on the presence of DNA strand breaks but seems to be regulated by the phos- phorylation state of the protein [35]. About 10% of cellular tankyrase protein is recruited to telomeres through binding of its ANK domain to TRF1. Thus the binding of TRF1 to telomeric DNA controls telo- mere length in cis by inhibiting the action of telo- merase at the ends of individual telomeres [36]. Tankyrase-2 This enzyme was originally described as a tumour anti- gen that elicited antibody responses in certain tumour patients (Table 1, Fig. 2) [37,38]. Later, tankyrase-2 was reported to interact with several other proteins such as TRF1 [39], insulin-responsive amino peptidase (IRAP) [40], or Grb14, an SH2 domain-containing adaptor protein that binds to the insulin and fibroblast growth factor receptors [41]. Despite being encoded by a separate gene, tankyrase-2 displays a domain struc- ture that is strikingly similar to tankyrase except for the N-terminal HPS domain, which is missing in tankyrase-2 (Fig. 2) [39,41]. Tankyrase and tankyrase- 2 also show a significant functional overlap: both pro- teins possess PARP activity and poly(ADP-ribosyl)ate some of their interaction partners (IRAP, TAB182 and TRF1, but not TRF2) as well as themselves, whereas tankyrase-2 displays preferential automodification activity. Strikingly, overexpression of tankyrase-2, but not of tankyrase, caused rapid poly(ADP-ribo- syl)ation-dependent cell death [39]. Both tankyrases can self-associate via the sterile alpha-module domain to form high-molecular-mass complexes, indicative of a function as master scaffolding molecules in organ- izing protein complexes [42]. PARG While there are 18 genes currently known or assumed to encode different PARP isoforms [2] there is only a single gene known to encode an enzyme catalysing the hydrolysis of ADP-ribose polymers to free ADP- ribose. This is the gene encoding poly(ADP-ribose) glycohydrolase (PARG; EC 3.2.1.143; Table 1) [43–45]. The human PARG gene shares a 470 bp bidirectional promoter with the gene encoding the translocase of the inner mitochondrial membrane 23 (TIM23). Promoter activity is several fold higher for TIM23 than for PARG [46]. Three splice variants of the human PARG have been described, giving rise to PARG isoforms tar- geted either to the nucleus or to the cytoplasm [47]. Overexpression studies revealed that the largest iso- form of PARG is targeted to the nucleus while the two smaller isoforms show mostly cytoplasmic localization. PARG is an enzyme that possesses both endoglyco- sidic and exoglycosidic activity and is the only protein known to catalyse the hydrolysis of ADP-ribose poly- mers to free ADP-ribose (Fig. 1) [43]. Its products are free poly(ADP-ribose) and monomeric ADP-ribose, the latter being a potent protein-glycating carbohy- drate capable of causing protein damage [48]. Interest- ingly, an ADPR pyrophosphatase has been described that converts ADPR to AMP and ribose 5-phosphate, thus decreasing the risk of nonenzymatic protein gly- cation [49]. As a consequence of the combined action of PARPs and PARG, poly(ADP-ribose) undergoes a dynamic turnover in live cells. Poly(ADP-ribosyl)ation A. Bu ¨ rkle 4580 FEBS Journal 272 (2005) 4576–4589 ª 2005 FEBS Biological functions of poly(ADP- ribosyl)ation DNA repair and maintenance of genomic stability Mechanistic aspects A plethora of studies have firmly established that DNA damage-induced poly(ADP-ribosyl)ation contributes to cellular recovery from cytotoxicity in proliferating cells inflicted with low or moderate levels of DNA damage by alkylation, oxidation or ionizing radiation. PARP-1 as well as the related protein PARP-2 are those members of the PARP family that are responsive to DNA damage and play important roles in DNA repair and mainten- ance of genomic integrity, thus behaving as ‘survival fac- tors’ [50,51]. Specifically, PARP-1 and PARP-2 have been shown to play a crucial role in the BER pathway. In mechanistic terms, no clear picture has yet emerged, despite intense research conducted by many groups. Attractive scenarios developed over the last couple of years are (a) the localized relaxation of chromatin at the site of DNA damage, mediated either by direct modifi- cation of histones or noncovalent interaction of histones with poly(ADP-ribose) present as automodification on PARP-1 or -2; (b) a damage signalling function [3]; or (c) recruitment of specific DNA repair proteins to the site of damage via noncovalent interaction with poly(ADP-ribose). These are but a few of the current hypotheses. Recent work has added some new interesting aspects. PARP-1 is a known interaction partner of the Werner syndrome protein, a protein involved in DNA repair, maintenance of genomic stability and the pre- vention of premature ageing. Recently PARP-1 was shown to regulate both the exonuclease and helicase activities of the Werner syndrome protein, suggesting a possible mechanism of action of PARP-1 [52]. Another scenario that has been put forward is the provision of ATP from pyrophosphorolytic cleavage of poly(ADP- ribose). Such ATP could be used for the DNA ligation step in BER [53]. Accordingly, the decision between the short-patch and the long-patch pathway in BER in living cells appears to be dependent on the availability of ATP. The long-patch pathway would be pre- ferred under conditions of energy shortage, as this pathway might lead to increased generation of ATP from poly(ADP-ribose) via increased provision of pyrophosphate from deoxynucleotide incorporation into DNA [54]. On the other hand, recent data clearly show that DNA-damage induced poly(ADP-ribosyl)ation has roles to play beyond classical BER. For instance, it accelerates the repair of oxidative base damage as well as of UV-induced pyrimidine dimers in a pathway dependent on Cockayne syndrome B protein [55]. Fur- thermore, PARP-1 plays a role in regulating double- strand break repair, independently of p53 [56]. In this context, a novel route for DNA double-strand breaks rejoining seems to involve PARP-1 and XRCC1 ⁄ DNA ligase III, i.e. proteins classically associated with BER [57]. A particular type of DNA damage is represented by stalled DNA topoisomerase I mole- cules. Poly(ADP-ribose) reactivates stalled DNA topo- isomerase I and induces resealing of the DNA strand breaks caused by topoisomerase stalling, which may be viewed as a direct repair activity of PARP-1 [58]. Another function of poly(ADP-ribosyl)ation in the absence of any exogenous DNA damage was highligh- ted recently. During spermiogenesis, spermatid differ- entiation is marked by dramatic changes in chromatin density and composition. The extreme condensation of the spermatid nucleus is characterized by a shift from histones to transition proteins and then to protamines as the major nuclear proteins. Recently poly(ADP- ribose) formation driven by endogenous DNA strand breaks was discovered in spermatids of steps 11–14 of spermiogenesis, i.e. those steps that immediately pre- cede the most pronounced phase of chromatin con- densation in spermiogenesis. Transient ADP-ribose polymer formation may therefore facilitate the process- ing of DNA strand breaks arising endogenously during the chromatin remodelling steps of sperm cell matur- ation [59]. It is interesting to note that there is apparently some specificity in the downstream consequences of PARP-1 deficiency in cells surviving a genotoxic attack, in that an increase of deletion mutations and insertions ⁄ rear- rangements was recorded in vivo after treatment with an alkylating agent [60]. The importance of a proficient poly(ADP-ribosyl)ation system for the maintenance of genomic stability and thus the prevention of cancer is also mirrored in the finding that the V762A genetic variant of PARP-1 [61,62], which is associated with diminished enzyme activity, contributes to prostate cancer susceptibility [63]. This is perfectly in line with the diminished PARP activity previously recorded in normal peripheral blood lymphocytes of laryngeal cancer patients [64]. Apart from experiments aiming at inhibition of cellu- lar poly(ADP-ribosyl)ation, experimental systems have also been established to raise cellular poly(ADP-ribose) levels above normal [65,66]. Supranormal levels of cellular poly(ADP-ribose) achieved by overexpression of PARP-1 in cultured cells proved to block genomic instability, assessed as sister-chromatid exchange, A. Bu ¨ rkle Poly(ADP-ribosyl)ation FEBS Journal 272 (2005) 4576–4589 ª 2005 FEBS 4581 induced by DNA damage [65], thus yielding the mirror image of what experiments with PARP inhibition have shown. Viewed together, the data demonstrate that poly(ADP-ribose) acts as a negative regulator of DNA- damage induced genomic instability [67]. PARP inhibitors in cancer treatment Proficient DNA repair is pivotal to the survival and maintenance of genomic stability of cells and organ- isms, given the relentless attack by endogenous and exogenous DNA-damaging agents. In the setting of cancer therapy with cytotoxic agents, however, DNA repair in tumour cells will counteract the desirable cell killing effect of the treatment. Accordingly, pharmaco- logical PARP inhibitors, which interfere with DNA repair pathways, have long been considered a useful addition to current cancer chemotherapy ⁄ radiotherapy protocols, drawing on the cocytotoxic effect of PARP inhibition under conditions of DNA damage [68]. Very recently, the specific killing of BRCA2-defici- ent tumour cells by the sole administration of PARP inhibitors, i.e. in the absence of any exogenous DNA-damaging treatment, was demonstrated [69,70]. Apparently this phenomenon is related with the defici- ency of BRCA2-deficient cells to perform homologous recombination, a pathway most prominently used by cells lacking PARP-1 activity. These findings should allow targeting of the DNA repair defect in BRCA- mutant cells as a new therapeutic strategy for some forms of human cancer. Regulation of transcription It has long been postulated that poly(ADP-ribo- syl)ation could influence the regulation of gene expres- sion via regulation of chromatin remodelling [23,71]. Indeed, numerous physical and functional interactions of PARP-1 with transcription factors have been des- cribed [72]. PARP-1, for example, plays a pivotal role in NF-jB-dependent gene expression, which makes it an important cofactor in immune and inflammatory responses [73–75]. Furthermore, recent data reveal functions of PARP-1 in the CaM kinase IId-dependent neurogenic gene activation pathway [76], in the NAD + -dependent modulation of chromatin structure and transcription mediated by nucleosome binding of PARP-1 [12], and in the determination of specificity in a retinoid signalling pathway via direct modulation of Mediator [77]. The involvement of poly(ADP-ribo- syl)ation in long-term potentiation in Aplysia neurones, occurring in the absence of DNA strand breaks, has also been linked with transcriptional effects [13]. Implications for apoptosis The specific cleavage of PARP-1 by caspase-3 ⁄ -7 within the nuclear location signal of PARP-1 generates a 24 kDa and an 89 kDa fragment, and this phenom- enon has been used extensively as a biochemical mar- ker of apoptosis. Caspase-mediated PARP-1 cleavage is thought to cause a loss of stimulation of the cata- lytic PARP-1 activity in the presence of DNA strand breaks. Recently a Parp1 knock-in mouse model (PARP-1(KI ⁄ KI)) was reported, in which the caspase cleavage site of PARP-1 has been mutated so as to render the protein resistant to caspases during apopto- sis [78]. Perhaps surprisingly the mice developed nor- mally. They also proved highly resistant to endotoxic shock and ischaemia–reperfusion damage, which was associated with reduced inflammatory responses in the target tissues and cells, due to the reduced production of specific inflammatory mediators. Despite normal binding of NF-jB to DNA, NF-jB-mediated tran- scription activity was impaired in these knock-in mice, which explains the above phenotype and creates a new and unexpected link between PARP-1 cleavage typical of apoptosis and the regulation of NF-jB, a master switch in inflammation. Despite the above-mentioned PARP-1 cleavage, massive formation of poly(ADP-ribose) can be observed during the early stages of apoptosis indica- ting that PARP family proteins are involved in this process [79,80]. Furthermore it could be shown that PARP-1 activation is required for the translocation of apoptosis-inducing factor from the mitochondria to the nucleus and that apoptosis-inducing factor is neces- sary for PARP-1-dependent cell death [81]. NAD + depletion, necrotic cell death and pathological conditions Twenty years ago, a mechanism of cell death depend- ing on the overactivation of PARP-1, and on severe and irreversible depletion of its substrate NAD + , was proposed for the first time [82]. Subsequently this para- digm was confirmed experimentally in mammalian sys- tems and also in plants [83]. In recent years this mechanism has been demonstrated to play a major role in a wide variety of pathophysiological conditions, including ischaemia–reperfusion damage and a wide range of inflammatory conditions (reviewed in [84,85]). Mitochondrial dysfunction triggered by PARP-1 over- activation seems to play a critical role for the ensuing cell death [86]. Based on this mechanism, PARP-inhibi- tory compounds are currently being developed as novel therapeutics to treat such kind of diseases. Poly(ADP-ribosyl)ation A. Bu ¨ rkle 4582 FEBS Journal 272 (2005) 4576–4589 ª 2005 FEBS Another intriguing scenario of how the decline of NAD + and the rise of nicotinamide triggered by extensive PARP-1 activation might impact on cellular physiology is the possible down-regulation of the activ- ity of silent information regulator 2 (Sir2)-like mam- malian proteins (sirtuins), a class of NAD + -dependent deacetylases [87]. Sir2 activity depends on high concen- tration of NAD + and is inhibited by nicotinamide. Sirtuins have been implicated in mediating gene silen- cing, longevity of organisms and genome stability. It was proposed that poly(ADP-ribosyl)ation by PARP- 1, which is induced by DNA damage, could modulate protein deacetylation by Sir2 via the NAD + ⁄ nicotina- mide connection [87]. Relevance of poly(ADP-ribose) catabolism A PARG loss-of-function mutant was described in Drosophila melanogaster, lacking the conserved cata- lytic domain of PARG. This mutant exhibits lethality in the larval stages at the normal developmental tem- perature of 25 °C [88]. However, about a quarter of the mutant fly population progressed to the adult stage at 29 °C but then displayed progressive neurodegenera- tion with reduced locomotor activity and a shortened lifespan. This phenotype was accompanied by extensive accumulation of poly(ADP-ribose) in the central ner- vous system. These results suggest that undisturbed poly(ADP-ribose) metabolism is required for mainten- ance of the normal function of neuronal cells. Recently, mouse models with mutant PARG gene have also been created. The complete loss of PARG activity induced by disruption of exon 1 led to the fail- ure of cells to degrade poly(ADP-ribose) and caused increased sensitivity to cytotoxicity and early embry- onic lethality [89]. By contrast, combined disruption of exons 2 and 3 led to the selective depletion of the 110 kDa isoform of the enzyme. This intervention was compatible with survival and fertility of mice and led to increased sensitivity to genotoxic and endotoxic stress in vivo [90]. Viewed together a picture emerges that loss of PARG activity may produce biological effects that depend very much on the cellular compart- ment(s) affected by such loss. This conclusion is rather sobering, but not implausible in view of the multiple cellular sites where poly(ADP-ribosyl)ation has been detected recently (see below). Centromere function and mitotic spindle formation PARP-2 together with PARP-1 has been detected at centromeres [19,91] where they both interact with constitutive and transient centromeric proteins [17,92], indicating that poly(ADP-ribosyl)ation might act as a regulator of both constitutive kinetochore proteins and those involved in spindle checkpoint control. Whereas PARP-2 localization is discrete at the centromere, PARP-1 shows a broader centromeric and pericentro- meric distribution [17]. The absence of any drastic centromeric phenotype in Parp1 knockout mice is sug- gestive of some functional redundancy for PARP-1 at the centromere [19]. On the other hand, an increase in centromeric chromatid breaks observed in Parp2 knockout mice exposed to c-irradiation has been reported. Furthermore female-specific lethality associ- ated with X-chromosome instability has been observed in Parp1 + ⁄ – ⁄ Parp2 – ⁄ – mice [19]. These data are sug- gestive of diverse roles of PARP-2 and PARP-1 in modulating the structure and checkpoint functions of the mammalian centromere, in particular during radi- ation-induced DNA damage. In addition, poly(ADP-ribose) was recently identified as a new component of the spindle, in addition to the known major spindle components including micro- tubules, microtubule-associated proteins and motors consisting of proteins and DNA [93]. The presence of poly(ADP-ribose) is required for bipolar spindle assembly and function. Centrosomal function The regulation of centrosome function is crucial to the accurate transmission of chromosomes to the daughter cells in mitosis. Both PARP-1 and PARP-3 (Table 1; Fig. 2) have been identified at centrosomes where they form a stable complex [20] and poly(ADP-ribosyl)ate p53 [31]. p53, in turn, has also been shown to localize at centrosomes and to control centrosome duplication [94]. Thus both PARP-1 and PARP-3 seem to be involved in centrosome duplication by modulating p53 activity via poly(ADP-ribosyl)ation. In particular, PARP-3 localizes preferentially to the daughter centri- ole throughout the cell cycle [20]. An attractive hypo- thesis is that the presence of both PARP-1 and PARP-3 at the centrosome may link the DNA damage surveillance network to the mitotic fidelity checkpoint. Tankyrase (Table 1, Fig. 2) is another member of the PARP family that localizes to the centrosome in a cell cycle-dependent manner. During mitosis, tankyrase colocalizes with nuclear mitotic apparatus protein (NuMa) [95], with which it was shown to form a stable complex at the centrosome [40]. When NuMa returns to the nucleus after mitosis this colocalization termin- ates and tankyrase associates with GLUT4 vesicles that coalesce around centrosomes [35] and function in A. Bu ¨ rkle Poly(ADP-ribosyl)ation FEBS Journal 272 (2005) 4576–4589 ª 2005 FEBS 4583 insulin-dependent glucose utilization. Thus, spindle poles and Golgi apparatus alternately contain most of cellular tankyrase, whereas only a small fraction of tankyrase functions at telomeres (see below). Structure and function of vault particles As mentioned above, vault particles are large cytoplas- mic RNPs. Although the cellular function of vaults is unknown, their subcellular localization and distinct morphology point to a role in intracellular transport, particularly nucleo–cytoplasmic transport [96]. It was reported that vault particles may also be involved in intracellular detoxification, as all three vault proteins display increased expression in many multidrug resist- ant human cell lines examined [97]. PARP-4 is identi- cal with the 193 kDa vault protein (Table 1, Fig. 2) and poly(ADP-ribosyl)ates the p100 subunit (major vault protein; MVP) within the vault particle and to a lesser extent itself. Immunofluorescence and biochemi- cal data show that PARP-4 is not exclusively associ- ated with the vault particle but can also localize to the nucleolus, the nuclear spindle and to nuclear pores [32,96]. The enzyme has also been found in association with mammalian telomerase but is dispensable for telomerase function and vault structure in vivo [98]. Telomere dynamics Parp1 knockout mice were reported to have shorter telomeres than wild-type mice [99]. This observation is indicative of a role of PARP-1 in maintaining telomere length and is compatible with the positive correlation between cellular poly(ADP-ribosyl)ation capacity (lar- gely reflecting PARP-1 activity) and lifespan of mam- malian species [100]. Apparently, the differences in enzyme activity may be due, at least in part, to chan- ges the primary structure of PARP-1 that arose during evolution [101]. A functional role of PARP-2 in the maintenance of telomere integrity is supported by the colocalization of PARP-2 and telomeric-repeat binding factor 2 (TRF2), which is a negative regulator of telomere length. PARP-2 activity regulates the DNA-binding activity of TRF2 via poly(ADP-ribosyl)ation of the dimerization domain of TRF2 as well as via noncovalent binding of poly(ADP-ribose) to the myb domain of TRF2 [102]. This protein interaction may well be involved in modu- lating t-loop formation in response to DNA damage. Tankyrase is another negative regulator of telomere length [34], but surprisingly only about 10% of cellular tankyrase protein is recruited to telomeres, and its binding to telomeres is mediated through TRF1. Poly(ADP-ribosyl)ation of TRF1 by tankyrase inhibits binding of TRF1 to telomeric DNA and so contributes to telomere length regulation, by reversing the negative effect of TRF1 on telomere length [103]. The catalytic activity of tankyrase is crucial for this effect, because nuclear overexpression of tankyrase, but not of a PARP-deficient mutant, causes the lengthening of telomeres [104]. Additional proteins, however, are also involved in TRF1 regulation, such as TRF1-interacting nuclear protein 2 (TIN2), which was reported to pro- tect TRF1 from poly(ADP-ribosyl)ation by tankyrase via formation of a ternary complex with TRF1 and tankyrase, yet without affecting tankyrase automodifi- cation [105]. As mentioned above, the domain structure of tanky- rase-2 is very similar to that of tankyrase except for the N-terminal HPS domain, which is missing in tankyrase-2 (Fig. 2) [39,41]. Likewise the two enzymes share several interaction partners including IRAP, TAB182 and TRF1. Overexpression of either tanky- rase or tankyrase-2 in the nucleus released endogenous TRF1 from the telomere, suggesting that the function of the two enzymes may partially be redundant [103,104]. Taken together, at least four members of the PARP family have been implicated in telomere regulation. A full understanding of their distinct functions at telo- meres, their regulation and possible functional cooper- ation will require a substantial amount of additional research work. Trafficking of endosomal vesicles Despite the effect of tankyrase on telomere dynamics, it should be noted that most of this protein is found outside the cell nucleus. As mentioned above, it can either be detected at centrosomes, where it seems to interact with NuMa [40,95], or in association with nuclear pore complexes [103] or at Golgi-associated GLUT4 vesicles [35]. Tankyrase was also reported to interact with a tankyrase-binding protein of 182 kDa (TAB182), displaying a heterochromatin-like staining pattern in the nucleus and colocalizing with cortical actin in the cytoplasm [106]. Endocytotic vesicles in myocytes and adipocytes contain the glucose transpor- ter GLUT4 as well as IRAP. The reversible transloca- tion of GLUT4 between these GLUT4 vesicles in the Golgi and the plasma membrane allows insulin to regulate glucose utilization. Tankyrase appears to be an important insulin-signalling target, as the protein not only interacts with IRAP located to GLUT4 stor- age vesicles in the Golgi, but is also phosphorylated by mitogen-activated protein kinase (MAPK) upon insulin Poly(ADP-ribosyl)ation A. Bu ¨ rkle 4584 FEBS Journal 272 (2005) 4576–4589 ª 2005 FEBS stimulation. Tankyrase poly(ADP-ribosyl)ates IRAP, as well as itself, and this activity is enhanced by MAPK-mediated phosphorylation indicating that tankyrase may mediate the long-term regulation of GLUT4 vesicles by the MAPK cascade [35,107]. Emerging new PARP family members TIPARP (PARP7) 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) is a pro- totype substance from the class of dioxins and causes pleiotropic effects in mammalian species through modulating gene expression. A novel, TCDD-inducible member of the PARP family (TiPARP or PARP7 [2]) was recently identified and characterized (Table 1) [108,109]. TiPARP mRNA is expressed in a broad range of mouse tissues and can poly(ADP-ribosyl)ate histones. Genetic analyses revealed that induction depends on the aromatic hydrocarbon receptor (AhR) and on the aromatic hydrocarbon receptor nuclear translocator (Arnt). PARP10 Recently a novel member of the PARP family (PARP- 10) has been characterized at the biochemical level, which is a novel c-myc-interacting protein with poly(ADP-ribose) polymerase activity (Table 1) [110]. PARP-10 is a 150 kDa protein that interacts with Myc and possesses a domain with homology to RNA recog- nition motifs. PARP-10 can poly(ADP-ribosyl)ate itself and core histones, but neither Myc nor Max, a well- known c-myc interactor. PARP-10 is localized to the nuclear and cytoplasmic compartments under the con- trol of a nuclear export sequence, which also seems to be relevant for the inhibitory effect of PARP-10 con- cerning c-Myc- and E1A-mediated cotransformation of rat embryo fibroblasts. Conclusion and outlook The field of poly(ADP-ribosyl)ation is currently a very exciting one and it has widened in every respect. Whereas until a couple of years ago a single enzyme was looked at (PARP-1), we are now dealing with over a dozen. Whereas DNA strand breakage used to be considered the only trigger of poly(ADP-ribose) syn- thesis and consequently the only relevant cellular condition for poly(ADP-ribose) function, several additional activation conditions and cellular pheno- mena related with poly(ADP-ribosyl)ation have been described. As a consequence, the range of specialists interested in poly(ADP-ribosyl)ation has broadened enormously. Obtaining a comprehensive picture of the biological functions of poly(ADP-ribosyl)ation and the underlying molecular mechanisms is highly desirable and would further accelerate the transfer of basic sci- entific information on this subject to the medical appli- cation. Hopes that this may be achieved in the not too distant future are increasing in view of the growing interest of the scientific community in this field. References 1Bu ¨ rkle A (2004) Poly(ADP-ribosyl)ation. Landes Bio- science, Georgetown, TX, USA. 2 Ame ´ JC, Spenlehauer C & de Murcia G (2004) The PARP superfamily. Bioessays 26, 882–893. 3 Pleschke JM, Kleczkowska HE, Strohm M & Althaus FR (2000) Poly(ADP-ribose) binds to specific domains in DNA damage checkpoint proteins. J Biol Chem 275, 40974–40980. 4 Reale A, Matteis GD, Galleazzi G, Zampieri M & Caiafa P (2005) Modulation of DNMT1 activity by ADP-ribose polymers. Oncogene 24, 13–19. 5 Chambon P, Weill JD & Mandel P (1963) Nicotina- mide mononucleotide activation of new DNA-depen- dent polyadenylic acid synthesizing nuclear enzyme. Biochem Biophys Res Commun 11, 39–43. 6 de Murcia G & Me ´ nissier de Murcia J (1994) Poly(ADP-ribose) polymerase: a molecular nick- sensor. Trends Biochem Sci 19, 172–176. 7 Kun E, Kirsten E & Ordahl CP (2002) Coenzymatic activity of randomly broken or intact double-stranded DNAs in auto and histone H1 trans-poly(ADP-ribosy- lation), catalyzed by poly(ADP-ribose) polymerase (PARP I). J Biol Chem 277, 39066–39069. 8 Gradwohl G, Mazen A & de Murcia G (1987) Poly(ADP-ribose) polymerase forms loops with DNA. Biochem Biophys Res Commun 148, 913–919. 9 Sastry SS & Kun E (1990) The interaction of adenosine diphosphoribosyl transferase (ADPRT) with a cruciform DNA. Biochem Biophys Res Commun 167, 842–847. 10 Oei SL, Herzog H, Hirsch-Kauffmann M, Schneider R, Auer B & Schweiger M (1994) Transcriptional regu- lation and autoregulation of the human gene for ADP- ribosyltransferase. Mol Cell Biochem 138, 99–104. 11 Lonskaya I, Potaman VN, Shlyakhtenko LS, Oussat- cheva EA, Lyubchenko YL & Soldatenkov VA (2005) Regulation of poly(ADP-ribose) polymerase-1 by DNA structure-specific binding. J Biol Chem 280, 17076– 17083. 12 Kim MY, Mauro S, Gevry N, Lis JT & Kraus WL (2004) NAD + -dependent modulation of chromatin structure and transcription by nucleosome binding properties of PARP-1. Cell 119, 803–814. A. Bu ¨ rkle Poly(ADP-ribosyl)ation FEBS Journal 272 (2005) 4576–4589 ª 2005 FEBS 4585 [...]... (ADP-ribose) glycohydrolase in the control of poly(ADP-ribose) metabolism Exp Cell Res 268, 7–13 Shimokawa T, Masutani M, Nozaki T, Nakagama H, Araki S, Aoki Y & Sugimura T (1998) The human poly(ADP-ribose) glycohydrolase maps to chromosome 10q11.23–21.1 by fluorescence in situ hybridization Hum Cell 11, 243–246 ´ Ame JC, Apiou F, Jacobson EL & Jacobson MK (1999) Assignment of the poly(ADP-ribose) glycohydrolase... Transient poly(ADPribosyl) ation of nuclear proteins and role of poly(ADPribose) polymerase in the early stages of apoptosis J Biol Chem 273, 13703–13712 81 Yu SW, Wang H, Poitras MF, Coombs C, Bowers WJ, Federoff HJ et al (2002) Mediation of poly(ADPribose) polymerase-1-dependent cell death by apoptosis-inducing factor Science 297, 259–263 82 Berger NA (1985) Poly(ADP-ribose) in the cellular response to DNA... Functions of poly(ADP-ribose) polymerase in controlling telomere length and chromosomal stability Nat Genet 23, 76–80 Grube K & Burkle A (1992) Poly(ADP-ribose) poly¨ merase activity in mononuclear leukocytes of 13 mammalian species correlates with species-specific life span Proc Natl Acad Sci USA 89, 11759–11763 Beneke S, Alvarez-Gonzalez R & Burkle A (2000) ¨ Comparative characterisation of poly(ADP-ribose). .. Ramroth H, Burkle ¨ A, Bartsch H, Dietz A & Becher H (2002) Reduced 4587 Poly(ADP-ribosyl)ation 65 66 67 68 69 70 71 72 73 74 75 76 77 poly(ADP-ribosyl)ation in lymphocytes of laryngeal cancer patients: results of a case-control study Int J Cancer 98, 780–784 Meyer R, Muller M, Beneke S, Kupper JH & Burkle ¨ ¨ ¨ A (2000) Negative regulation of alkylation-induced sister-chromatid exchange by poly(ADP-ribose). .. Brabeck C, Pfeiffer R, Leake A, Beneke S, Meyer R & Burkle A (2003) 1-selegiline potentiates the cellular ¨ poly(ADP-ribosyl)ation response to ionizing radiation J Pharmacol Exp Ther 306, 973–979 Burkle A (2001) Poly(APD-ribosyl)ation, a DNA dam¨ age-driven protein modification and regulator of genomic instability Cancer Lett 163, 1–5 Beneke S, Diefenbach J & Burkle A (2004) Poly(ADP¨ ribosyl)ation inhibitors:... Lis JT (2003) PARP goes transcription Cell 113, 677–683 Hassa PO & Hottiger MO (2002) The functional role of poly(ADP-ribose) polymerase 1 as novel coactivator of NF-kappaB in inflammatory disorders Cell Mol Life Sci 59, 1534–1553 Hassa PO, Buerki C, Lombardi C, Imhof R & Hottiger MO (2003) Transcriptional coactivation of nuclear factor-kappaB-dependent gene expression by p300 is regulated by poly(ADP)-ribose... (2000) Poly(ADP-ribose) polymerase-1: what have we learned from the deficient mouse model? Mutat Res 460, 1–15 ´ Ame JC, Rolli V, Schreiber V, Niedergang C, Apiou F, Decker P et al (1999) PARP-2, A novel mammalian DNA damage-dependent poly(ADP-ribose) polymerase J Biol Chem 274, 17860–17868 ´ Oliver AW, Ame JC, Roe SM, Good V, de Murcia G & Pearl LH (2004) Crystal structure of the catalytic fragment of. .. Brouwer D & Cornelissen M (2005) Poly(ADP-ribose) polymerase in plants affects energy homeostasis, cell death and stress tolerance Plant J 41, 95–106 84 Virag L & Szabo C (2002) The therapeutic potential of poly(ADP-ribose) polymerase inhibitors Pharmacol Rev 54, 375–429 85 Meli E, Pangallo M, Baronti R, Chiarugi A, Cozzi A, Pellegrini-Giampietro DE & Moroni F (2003) Poly(ADP-ribose) polymerase as a... Visochek L, Katzoff A, Levitan D, Susswein AJ, Klein R, Valbrun M & Schwartz JH (2004) Long-term memory requires polyADP-ribosylation Science 304, 1820–1822 14 Ruf A, Mennissier de Murcia J, de Murcia G & Schulz GE (1996) Structure of the catalytic fragment of poly(ADP-ribose) polymerase from chicken Proc Natl Acad Sci USA 93, 7481–7485 15 Juarez-Salinas H, Sims JL & Jacobson MK (1979) Poly(ADP-ribose). .. F, Blanche H, Verasdonck P, Le Gall I, Schachter F, Burkle A & Muiras ML (2000) New polymorph¨ isms in the human poly(ADP-ribose) polymerase-1 coding sequence: lack of association with longevity or with increased cellular poly(ADP-ribosyl)ation capacity J Mol Med 78, 431–440 63 Lockett KL, Hall MC, Xu J, Zheng SL, Berwick M, Chuang SC et al (2004) The ADPRT V762A genetic variant contributes to prostate . MINIREVIEW Poly(ADP-ribose) The most elaborate metabolite of NAD + Alexander Bu ¨ rkle Department of Biology, University of Konstanz, Germany Introduction The. chain of several covalently linked ADP-ribo- syl residues is the basis of the formation of poly(ADP- ribose), which can be regarded the cell’s most elaborate metabolite