![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
REVIEW
1 Department of Molecular Biology and Genetics, Cornell University, Ithaca, New York 14853, USA; 2 Department of Pharmacology, Weill Medical College of Cornell University, New York, New York 10021, USA
 |
Abstract |
|---|
 Top Abstract Structural and functional...The chemical biology of...Catabolism of PAR by...The molecular and cellular...The physiology and...Nuclear NAD+ metabolism and...How can PARP-1 and...AcknowledgmentsReferences |
|---|
[Keywords: NAD+; PARP-1; poly(ADP-ribosyl)ation]
|
Structural and functional domains of PARP-1, the founding member of the PARP family |
|---|
|
TopAbstractStructural and functional...The chemical biology of...Catabolism of PAR by...The molecular and cellular...The physiology and...Nuclear NAD+ metabolism and...How can PARP-1 and...AcknowledgmentsReferences |
|---|
). PARPs catalyze the polymerization of ADP-ribose units from donor NAD+ molecules on target proteins, resulting in the attachment of linear or branched polymers (Fig. 1). This enzymatic activity has been detected in organisms ranging from plants to mammals, but is apparently absent from yeast (Rolli et al. 2000; Amé et al. 2004). PARP family members share a conserved catalytic domain that contains the "PARP signature" motif, a highly conserved sequence (100% conserved in PARP-1 among vertebrates) that forms the active site (Rolli et al. 2000). Some PARP family members identified solely on homology, however, have not yet been shown to possess intrinsic PARP enzymatic activity (Amé et al. 2004). In addition to a catalytic domain, PARP family members typically contain one or more additional motifs or domains, including zinc fingers, "BRCA1 C-terminus-like" (BRCT) motifs, ankyrin repeats, macro domains, and WWE domains, each conferring unique properties on the particular PARP protein that contains them (Amé et al. 2004).
PARP-1 has a highly conserved structural and functional organization including (1) an N-terminal double zinc finger DNA-binding domain (DBD), (2) a nuclear localization signal, (3) a central automodification domain, and (4) a C-terminal catalytic domain (Fig. 2A; D'Amours et al. 1999; Rolli et al. 2000; Kraus and Lis 2003). PARP-1 binds to a variety of DNA structures, including single- and double-strand breaks, crossovers, cruciforms, and supercoils, as well as some specific double-stranded sequences (Rolli et al. 2000). PARP-1's basal enzymatic activity is very low, but is stimulated dramatically in the presence of a variety of allosteric activators, including damaged DNA, some undamaged DNA structures, nucleosomes, and a variety of protein-binding partners (D'Amours et al. 1999; Oei and Shi 2001; Kun et al. 2002, 2004; Kim et al. 2004). The targets of PARP-1's enzymatic activity include PARP-1 itself, which is the primary target in vivo, core histones, the linker histone H1, and a variety of transcription-related factors that interact with PARP-1 (Ogata et al. 1981; Huletsky et al. 1989; D'Amours et al. 1999; Kraus and Lis 2003). The automodification domain of PARP-1 contains several glutamate residues that are likely targets for automodification and a BRCT motif that functions in protein-protein interactions (D'Amours et al. 1999; Rolli et al. 2000). Collectively, the domains and activities of PARP-1 suggest important roles in a variety of nuclear functions. The activities and functions of the other PARP family members have not been studied to the same extent as PARP-1, although a clearer picture for some of the PARP family members has been emerging, as noted below and reviewed in more detail elsewhere (Smith 2001; Amé et al. 2004).
|
The chemical biology of PAR |
|---|
|
TopAbstractStructural and functional...The chemical biology of...Catabolism of PAR by...The molecular and cellular...The physiology and...Nuclear NAD+ metabolism and...How can PARP-1 and...AcknowledgmentsReferences |
|---|
) more than 40 years ago as a nicotinamide mononucleotide (NMN)-induced incorporation of 14C-adenine-labeled ATP into an acid-insoluble fraction in a nuclear extract. PAR is a branched polymer of repeating ADP-ribose units, which are linked via glycosidic ribose-ribose 1'' 
2' bonds (Fig. 3). Due to its chemical composition, PAR has been referred to as the "third type of nucleic acid" (D'Amours et al. 1999), although its biological role is considerably less well understood than RNA and DNA. PAR is heterogeneous with respect to length (as many as 200 ADP-ribose units in vitro) and extent of branching (approximately one branch per 20-50 ADP-ribose units) (D'Amours et al. 1999). The significance of this heterogeneity in PAR function is unknown, but it could play a role in determining specific functional outcomes in vivo.
|
; Bürkle et al. 2000; Kraus and Lis 2003). Each residue in PAR contains an adenine moiety capable of base stacking and hydrogen bonding, as well as two phosphate groups that carry negative charges (Amé et al. 2000). PAR may form definitive structures through intramolecular interactions (Minaga and Kun 1983a,b), and these structures have the potential for noncovalent attractive (or repulsive) interactions with other molecules (Mathis and Althaus 1987; Wesierska-Gadek and Sauermann 1988; Panzeter et al. 1992). Thus, PAR may alter protein activity by functioning as a site-specific covalent modification, a protein-binding matrix, or a steric block. For example, inhibition of PARP-1's DNA-binding activity by autoPARylation may be the result of steric effects of PAR on the biochemical properties of PARP-1, perhaps masking the PARP-1 DNA-binding domain, or through charge repulsion between PAR and DNA (D'Amours et al. 1999).
The functions of PAR in vivo are likely to involve specific interactions with a variety of effector proteins. Proteomic approaches have led to the identification of a 20-amino-acid PAR-binding motif in a heterogeneous group of PAR-binding proteins, including core histones, p53, and XRCC-1, that maps to important functional domains within the proteins (Pleschke et al. 2000; Gagne et al. 2003). In addition, the "macro domain," a conserved 
190-amino-acid domain found in a wide variety of proteins, was recently shown to function as an ADP-ribose-binding module (Karras et al. 2005). This includes the macro domain of macroH2A, a histone variant thought to be involved in transcriptional repression, and the double macro domain of PARP-9, a PARP involved in leukemia (Karras et al. 2005). Some non-sequence-specific DNA-binding proteins, such as H1, may even have a greater affinity for PAR than DNA (Malanga et al. 1998). Understanding the functional consequences of PAR-protein interactions will be the key to understanding the biology of PAR itself. This is an area that will require further investigation in the future.
|
Catabolism of PAR by PARG |
|---|
|
TopAbstractStructural and functional...The chemical biology of...Catabolism of PAR by...The molecular and cellular...The physiology and...Nuclear NAD+ metabolism and...How can PARP-1 and...AcknowledgmentsReferences |
|---|
; Davidovic et al. 2001). PARG proteins, which have been identified in mammals, flies, worms, and plants, have a catalytic domain, nuclear localization signal, nuclear export signal, and a putative regulatory domain (Fig. 2C; Davidovic et al. 2001; Meyer-Ficca et al. 2005). Only the catalytic domain shows a high level of homology across species, and the arrangement of the other motifs and domains within the protein varies from species to species (Amé et al. 2000).
In mammals, a single PARG gene encodes multiple PARG proteins with two predominant isoforms: a long nuclear/cytoplasmic isoform (110 kDa) and a short cytoplasmic isoform (65 kDa), both of which possess catalytic activity (Davidovic et al. 2001; Meyer-Ficca et al. 2005). The results of a recent study suggest that multiple long PARG isoforms may be produced by alternative mRNA splicing: a low-abundance nuclear form (110 kDa, which contains a NLS encoded by exon 1) and two high-abundance cytoplasmic forms (102 kDa lacking exon 1; 99 kDa lacking exons 1 and 2) that exhibit a perinuclear distribution (Meyer-Ficca et al. 2004). The abundance of PARG in the cytoplasm seems at odds with the fact that many of the PARP enzymes are located in the nucleus, but may indicate that low levels of PARG are sufficient for the catabolism of nuclear PAR. Generating a more comprehensive and unified picture of the nature and localization of the different PARG isoforms will require additional studies in a wider variety of cell types.
In vivo, the steady-state levels of PAR are regulated by the opposing actions of PARPs and PARG. The degradation of PAR may begin immediately upon the initiation of PAR synthesis and can be completed within minutes after the cessation of PAR synthesis has occurred (D'Amours et al. 1999; Tulin and Spradling 2003). This suggests that PAR and PAR-metabolizing enzymes are highly regulated. Although PARP-1 is present at a five-fold to 20-fold molar excess over PARG in some cell types, a variety of regulatory mechanisms act to control the levels of PAR in the nucleus (Alvarez-Gonzalez et al. 1999; D'Amours et al. 1999; Davidovic et al. 2001). For example, PARP-1 has a low basal enzymatic activity, which is stimulated dramatically by PARP-1's binding partners, including various proteins and forms of DNA (Oei and Shi 2001; Kun et al. 2002; Kim et al. 2004). PARG, on the other hand, has a higher specific activity than PARP-1, and its enzymatic activity increases with increased PAR length (D'Amours et al. 1999). Furthermore, PARG activity in the nucleus may be modulated by nucleo-cytoplasmic shuttling of the protein (Bonicalzi et al. 2003; Ohashi et al. 2003), which could control the levels of nuclear PARG enzymatic activity through regulated subcellular distribution.
|
2-3/2-3), resulting in depletion of the 110-kDa PARG isoform, are viable, phenotypically normal, and show similar overall PAR metabolism as wild-type animals (Cortes et al. 2004). The Parg2-3/2-3 animals do, however, exhibit increased sensitivity to genotoxins and septic shock, manifested as increased lethality relative to wild-type animals in response to treatment with N-methyl-N-nitrosourea (MNU), 
-irradiation, and lipopolysaccharide (LPS) (Cortes et al. 2004). Mice homozygous for a targeted deletion of exon 4 (Parg4/4), resulting in a complete depletion of all PARG isoforms (i.e., Parg null), die at embryonic day 3.5 (E3.5) (Koh et al. 2004). Parg4/4 embryos and embryonic trophoblast stem cells accumulate high levels of PAR and undergo increased cell death by apoptosis. Likewise, Drosophila containing loss-of-function mutations in PARG exhibit increased lethality in the larval stages at normal developmental temperatures (Table 1; Hanai et al. 2004). Together, these studies demonstrate the importance of PAR catabolism for the maintenance of normal physiology.
|
|
The molecular and cellular biology of PAR and PARP-1 |
|---|
|
TopAbstractStructural and functional...The chemical biology of...Catabolism of PAR by...The molecular and cellular...The physiology and...Nuclear NAD+ metabolism and...How can PARP-1 and...AcknowledgmentsReferences |
|---|
DNA damage detection and repair
The covalent modification of proteins by PARylation is an immediate and dramatic biochemical response to DNA damage induced by oxidation, alkylation, and ionizing radiation. The binding of PARP-1 to damaged DNA, including single-strand breaks (SSBs) and double-strand breaks (DSBs), through its double zinc finger DNA-binding domain potently activates PARP-1 enzymatic activity (as much as 500-fold) (D'Amours et al. 1999). As such, PARP-1 can function as a DNA damage sensor. With low levels of DNA damage, PARP-1 acts as a survival factor involved in DNA damage detection and repair. In contrast, with high levels of DNA damage, PARP-1 promotes cell death (see below) (Bürkle 2001a). PARP-1 has been implicated in multiple DNA repair pathways, including the SSB, DSB, and base excision repair (BER) pathways (Bürkle 2001b; Masutani et al. 2003). As might be expected, PARP-1 interacts physically and functionally with various proteins involved in these DNA repair pathways, and may recruit the repair proteins to sites of DNA damage (e.g., XRCC-1 in BER, DNA-dependent protein kinase in DSB repair) (Masson et al. 1998; Ruscetti et al. 1998; Ariumi et al. 1999; Okano et al. 2003; Lan et al. 2004). PAR itself, as a covalent attachment of automodified PARP-1, may also act to recruit repair proteins to sites of DNA damage (Malanga and Althaus 2005). PARP-2, the only other PARP enzyme whose catalytic activity is known to be stimulated by damaged DNA, has also been implicated in BER through interactions with XRCC-1 and PARP-1 (Schreiber et al. 2002).
|
; Farmer et al. 2005). BRCA1 and BRCA2 are tumor-suppressor proteins important for DSB repair by homologous recombination, and mutation of the genes encoding these proteins causes predisposition to breast and ovarian cancers (Tutt and Ashworth 2002; Wooster and Weber 2003). Depletion or chemical inhibition of PARP-1, but not PARP-2, in BRCA1- or BRCA2-deficient cells reduces the clonogenic survival of the cells compared to PARP-1-depleted/inhibited, BRCA1- or BRCA2-sufficient cells (Bryant et al. 2005; Farmer et al. 2005). Furthermore, BRCA1- or BRCA2-deficient cells treated with a PARP-1 inhibitor show major mitotic chromosome aberrations and a loss of formation of RAD51 foci, suggesting defects in DSB repair involving RAD51-dependent homologous recombination (Bryant et al. 2005; Farmer et al. 2005). Collectively, these studies suggest that the persistent single-strand breaks formed upon PARP-1 inhibition, which promote the collapse of replication forks and would normally be repaired by homologous recombination, cannot be repaired effectively in the absence of functional BRCA1 or BRCA2. This results in the accumulation of chromosomal abnormalities, cell cycle arrest in G2/M, and apoptosis (Bryant et al. 2005; Farmer et al. 2005). Although these and other studies have demonstrated a role for PARP-1 in several DNA repair pathways, a clear picture of the exact mechanisms by which PARP-1-mediated PARylation modulates these pathways is missing.
Cell death pathways
In contrast to its role as a survival factor in the presence of low levels of DNA damage, PARP-1 acts to promote cell death in the presence of extensive DNA damage. As such, chemical inhibition or genetic deletion of PARP-1 can protect animals from several DNA-damage-dependent pathophysiological conditions leading to aberrant cell death, including (1) ischemia-reperfusion injury, (2) glutamate excitotoxicity in the central nervous system, (3) 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-induced Parkinsonism, (4) cardiac infarction, (5) inflammatory injury, and (6) streptozotocin (STZ)-induced diabetes (Szabo and Dawson 1998; Shall and de Murcia 2000). Although a role for PARP-1 in these conditions has been well established, the mechanisms by which PARP-1 activation leads to cell death are still under active debate in the literature. Several mechanisms have been proposed, including energy-failure-induced necrosis and apoptosis-inducing factor (AIF)-dependent apoptosis.
Necrosis is a cell death process in which a cell swells and ruptures as it dies, releasing intracellular components into the surrounding tissue, which promotes an inflammatory response (Edinger and Thompson 2004). Hypersynthesis of PAR by PARP-1 in response to extensive DNA damage can promote cell death through necrosis, which occurs as a result of the depletion of cellular NAD+ and ATP, and subsequent cellular energy failure (Decker and Muller 2002; Bouchard et al. 2003). In contrast, apoptosis is an ordered cell death process in which the cell is systematically dismantled within membrane-enclosed vesicles that are engulfed by phagocytes, preventing the release of intracellular components into the surrounding tissue (Edinger and Thompson 2004). Studies from Yu et al. (2002) show that PARP-1 can play a role in caspase-independent apoptotic cell death through AIF. AIF is a pro-apoptotic flavoprotein residing in the mitochondrial intermembrane space (like cytochrome c and other apoptotic modulators) that is among the most powerful triggers of apoptosis (Chiarugi and Moskowitz 2002). In the nucleus, AIF induces peripheral chromatin condensation and high-molecular-weight (50 kb) DNA fragmentation (Susin et al. 1999). The translocation of AIF into the nucleus is impaired in Parp-1-/- fibroblasts after treatment with N-methyl-N-nitro-N-nitrosoguanidine (MNNG), a DNA-alkylating agent that potently activates PARP-1 and promotes PARP-1-dependent cell death (Yu et al. 2002). Exactly how PARP-1 activation triggers the release of AIF from mitochondria is not clear, but depletion of NAD+ following PARP-1 activation, or perhaps a product of PAR catabolism, could serve as a potential signal.
The mechanisms underlying the choice of PARP-1-dependent cell death pathways (i.e., necrosis vs. apoptosis) in response to genotoxic stimuli have not been determined, but may be influenced by the type, strength, and duration of the stimuli, as well as the cell type (Virag 2005). One feature of apoptosis is its dependency on ATP for the ordered degradation of cellular structures and maintenance of membrane integrity (Edinger and Thompson 2004). Thus, in cells whose ATP pools have been depleted due to PARP-1 activation, cell death occurs by necrosis. In addition, recent studies indicate that cellular metabolic status is a key factor in determining how ATP levels are affected by PARP-1 activation (Zong et al. 2004; Ying et al. 2005). Actively proliferating cells, such as cancer cells, are dependent on glucose catabolism through aerobic glycolysis for ATP production. Nonproliferating cells, in contrast, can catabolize a variety of metabolic substrates, including amino acids and lipids, and maintain ATP levels through oxidative phosphorylation in the mitochondria. PARP-1 activation in the nucleus preferentially depletes the nuclear and cytosolic pools of NAD+, but not the mitochondrial pools, thereby inhibiting glycolysis, but not oxidative phosphorylation (Zong et al. 2004). Consequently, proliferating cells are more sensitive to PARP-1 activation, becoming depleted of ATP and dying by necrosis. In contrast, nonproliferating cells are resistant to ATP depletion and cell death under the same conditions (Zong et al. 2004). Similarly, in MNNG-treated astrocytes, the decision between cell death or survival is regulated by the availability of metabolic substrates. Supplying substrates that bypass cytosolic glycolysis for ATP production enhances cell survival (Ying et al. 2005). Thus, PARP-1 activity plays a central role in this form of "programmed necrosis" (Edinger and Thompson 2004), which integrates signals from cellular proliferation, metabolism, and DNA damage to determine cell fate.
Interestingly, PARP-1 is inactivated during the execution phase of apoptosis (Soldani and Scovassi 2002). PARP-1 is cleaved by caspase-3 and -7, signature proapoptotic proteases, into an 25-kDa N-terminal fragment containing the DBD, and an 85-kDa C-terminal fragment that retains basal, but not DNA-damage-activated, enzymatic activity (Kaufmann et al. 1993). This cleavage eliminates PARP-1 activation in response to DNA fragmentation during apoptosis, protecting the cells from ATP depletion and subsequent necrotic death. In addition, by preventing futile attempts at DNA repair, PARP-1 cleavage may help to commit cells to the apoptotic pathway (Soldani and Scovassi 2002). PARG is also cleaved during apoptosis by caspase-3, releasing a C-terminal fragment that retains full enzymatic activity (Affar et al. 2001). Although the consequences are not clearly understood, the cleavage of both PARP-1 and PARG by caspases during apoptosis suggests an important function for PAR metabolism in regulating apoptosis.
Modification of chromatin structure
PARP-1 and PAR have been shown to play a role in regulating chromatin structure in the presence or absence of DNA damage. Early biochemical studies suggested that PARP-1 could disrupt chromatin structure by PARylating histones and destabilizing nucleosomes (Poirier et al. 1982; Mathis and Althaus 1987; Huletsky et al. 1989). Histones H1 and H2B are the main histone targets for PARylation in vivo by PARP-1 and PARP-2, respectively (Poirier et al. 1982; Huletsky et al. 1989), although the other core histones are modified as well (D'Amours et al. 1999). Non-histone chromosomal proteins, including high mobility group proteins, are also PARylated in vivo (Tanuma and Johnson 1983). PARylation of histone proteins has been implicated in the decondensation of chromatin through destabilization of nucleosomes (Poirier et al. 1982; Mathis and Althaus 1987; Kraus and Lis 2003). Note, however, that although much emphasis has been placed on histone proteins as targets for PARylation, PARP-1 itself is the primary target for PARylation in vivo, with >90% of PAR found on PARP-1 (Ogata et al. 1981; Huletsky et al. 1989; D'Amours et al. 1999). Additional biochemical studies suggested that polyanionic PAR, either free or attached to proteins such as PARP-1, may provide an attractive matrix for histones released from destabilized nucleosomes (Mathis and Althaus 1987; Realini and Althaus 1992) or even strip basic proteins, such as histones, from DNA (Mathis and Althaus 1987; Wesierska-Gadek and Sauermann 1988; Panzeter et al. 1992).
Complementing these biochemical studies, Tulin and Spradling used Drosophila, which has only two PARP genes (PARP-1-like, which expresses three isoforms, and tankyrase-like) (Hanai et al. 1998; Miwa et al. 1999), as a model system to examine the role of PARP-1 in the regulation of chromatin structure in vivo (Tulin et al. 2002; Tulin and Spradling 2003). Inhibition of PARP-1 enzymatic activity by using 3-aminobenzamide (3-AB; a general PARP inhibitor) or disruption of all PARP-1 expression blocks the accumulation of PAR, decondensation ("puffing"), and transcription at loci containing highly inducible genes (e.g., heat-shock- and ecdysone-induced genes) (Tulin and Spradling 2003). The authors proposed that in response to an external stimulus (e.g., heat shock or ecdysone), PARP-1 becomes activated and modifies chromatin proteins, resulting in chromatin decondensation and transcriptional activation (Tulin and Spradling 2003). However, the definitive targets of PARylation at the decondensed loci were not determined. Interestingly, the same group also reported that PARP-1 can play a role in promoting the formation of more compact chromatin structures (Tulin et al. 2002). Genetic disruption of PARP-1 gene expression in Drosophila, which causes larval lethality (Table 1; Miwa et al. 1999; Tulin et al. 2002), promotes a dramatic increase in the micrococcal nuclease sensitivity (i.e., decondensation) of heterochromatic, but not euchromatic, regions (Tulin et al. 2002). PARP-1, therefore, has opposite effects on chromatin structure (i.e., condensation vs. decondensation) depending on the type of chromatin. The mechanistic differences between these two types of regulation have yet to be elucidated.
A recent study from our laboratory has helped to clarify the mechanisms by which PARP-1 can direct the reversible modulation of chromatin structure in an NAD+-dependent manner (Kim et al. 2004). Using a combination of biochemical, cell-based, and cytological approaches with human and Drosophila cells, we showed that PARP-1 incorporates into chromatin by virtue of specific nucleosome-binding properties and promotes the formation of compact, transcriptionally repressed chromatin structures. This is reminiscent of the activities of the linker histone H1. In the presence of NAD+, PARP-1 autoPARylates and dissociates from chromatin, resulting in the formation of decondensed, transcriptionally active chromatin structures. In contrast to previous models, this process occurs in the apparent absence of histone modification. Also, we showed that nucleosomes are very potent activators of PARP-1's enzymatic activity, indicating that nucleosome-bound PARP-1 is poised for activation even in the absence of DNA damage. Finally, although PARP-1 and H1 behave similarly in several in vitro assays, they compete for binding to nucleosomes in vitro and localize to distinct chromatin domains in vivo, suggesting distinct chromatin-dependent functions for these nucleosome-binding proteins. The interplay between PARP-1 and H1 at the level of the nucleosome may also involve PARylation of H1 by PARP-1, which would presumably promote the release of H1, although this has not yet been examined directly.
Collectively, these studies support the idea that PARP-1 can facilitate both the compaction and decondensation of chromatin depending on the physiological signals available. Furthermore, these studies demonstrate that chromatin decondensation and transcriptional activation require PARylation events. The mechanisms by which cellular signals, such as hormones and stress, can trigger downstream events to activate PARP-1 are not known. In addition, it is not yet clear what signal, perhaps a specific pattern of histone modification, directs PARP-1 binding to H1-depleted regions of the genome.
A variety of in vitro and in vivo experiments have also suggested a role for PARylation in determining and maintaining the methylation patterns of genomic DNA (Zardo et al. 2003). For example, inhibition of cellular PARP activity by 3-AB has been shown to increase the extent and alter the pattern of DNA methylation in the promoter CpG island of the Htf9 gene in L929 mouse fibroblasts (Zardo and Caiafa 1998). This may occur through decreased expression of the DNA methyltransferase DNMT1 (Zardo et al. 2002) or inhibition of DNMT1 methyltransferase activity by noncovalent binding of PAR (Reale et al. 2005). Given the important role of DNA methylation in organizing chromatin structures (Robertson 2002), PARylation-dependent alterations in the extent and patterns of DNA methylation could provide another means for PARP-1 and related enzymes to modulate chromatin structure.
Transcriptional regulation
Roles for PARP-1 and PARylation in the transcriptional regulation of specific genes have been demonstrated in several physiological contexts using a variety of experimental approaches, including in vitro transcription assays, cell-based reporter gene assays, RNAi, and gene deletion in vivo (D'Amours et al. 1999; Hassa and Hottiger 2002; Kraus and Lis 2003). From these studies, at least two different mechanisms have been proposed for the regulation of transcription by PARP-1: (1) modulating chromatin structure, as described above, and (2) acting as part of gene-specific enhancer/promoter-binding complexes (Kraus and Lis 2003). In each case, the targets of PARP-1 enzymatic activity differ. In its capacity as a component of enhancer/promoter-binding complexes, PARP-1 acts to stimulate transcription with some activators, while inhibiting transcription with others, depending on the cell type and promoter context (D'Amours et al. 1999; Hassa and Hottiger 2002; Kraus and Lis 2003). In some cases, PARP-1 enzymatic activity is not required for its transcriptional coregulator function (e.g., with NF-
B, HTLV Tax, B-Myb) (Meisterernst et al. 1997; Anderson et al. 2000; Cervellera and Sala 2000; Hassa et al. 2001). Although a growing body of work has demonstrated a role for PARP-1 as a transcriptional coregulator, mechanistic details of how PARP-1 might serve such a role have been lacking.
Two recent papers have shed new light on the mechanisms by which PARP-1 can modulate signal-regulated transcription in an activator- and promoter-specific manner. Pavri et al. (2005) used a combination of biochemical and cell-based assays to demonstrate a role for PARP-1 in ligand-dependent transcription by retinoic acid receptor 
(RAR)/retinoid X receptor (RXR) heterodimers. With chromatin templates and a purified in vitro transcription system, they found that PARP-1 is required for retinoic acid (RA)-induced transcription by RAR/RXR. This activity does not require the catalytic domain of PARP-1, and the addition of NAD+ leads to a loss of RA-dependence. In Parp-1-/- mouse embryo fibroblasts (MEFs), RA-induced expression of the RAR-regulated RAR
2 gene is lost, but can be restored by complementation with full-length or catalytic-domain-deleted PARP-1. Chromatin immunoprecipitation (ChIP) assays indicate that PARP-1, RAR, and inactive Mediator (a coactivator) are bound constitutively to the RAR2 promoter in both Parp-1+/+ and Parp-1-/- MEFs in the absence of RA. In Parp-1-/- MEFs, however, the cdk8 component of inactive Mediator fails to dissociate from the RAR2 promoter upon treatment with RA, blocking subsequent transcription of the gene. The authors propose a model in which PARP-1 supports RA- and RAR-dependent transcription of the RAR2 gene by promoting the conversion of Mediator from an inactive form (+cdk8) to an active form (-cdk8). The mechanisms underlying this switch (e.g., possible PARP-1-cdk8 interactions), however, have not yet been determined. In addition, the mechanism by which nuclear NAD+ is prevented from disrupting this regulatory system in vivo is unclear.
Ju et al. (2004) provide another perspective on how PARP-1 might function as a regulated "promoter-specific exchange factor," focusing on PARP-1-dependent dissociation of TLE (Transducin-like enhancer of Split) corepressor complexes from the proneural MASH1 promoter. MASH1 expression is repressed in proliferating neuronal stem cells (NSCs) by the inhibitory basic helix-loop-helix transcription factor HES1 (Hairy/Enhancer of Split-1), which recruits TLE corepressor complexes containing PARP-1. ChIP assays in NSCs indicate that treatment with PDGF, which initiates a differentiation program in the cells, promotes the dissociation of the TLE corepressor complex, but not HES1 or PARP-1, from the MASH1 promoter. This is followed by the recruitment of HAT-containing coactivator complexes and expression of the MASH1 gene. Interestingly, calcium/calmodulin-dependent protein kinase II
(CaMKII) is recruited to the MASH1 promoter and is required for PDGF-dependent release of the TLE1 corepressor complex from the promoter and the derepression of MASH1 expression. The authors propose a model in which CaMKII directs the switch from repression to activation at the MASH1 promoter by (1) stimulating PARP-1 enzymatic activity, leading to PARylation and dissociation of the TLE repressor complex; and (2) phosphorylating HES1, thereby converting it from a TLE-dependent repressor to an activator capable of recruiting coactivator complexes.
The studies from Pavri et al. (2005) and Ju et al. (2004), both of which demonstrate roles for PARP-1 as a promoter-specific exchange factor, add to the growing literature defining roles for PARP-1 in signal-regulated transcription in the absence of DNA damage. What is perplexing is the wide variety of mechanisms for PARP-1 transcriptional coregulatory activity, which seem to vary in an activator- and gene-specific manner. What remains to be seen is if any unifying principles will be apparent when more of the mechanistic details are uncovered.
Insulator function
Recent results from Yu et al. (2004) have implicated PARylation in the regulation of CTCF, a ubiquitous DNA-binding protein, at transcriptional insulators. Insulators are elements that organize the genome into discrete regulatory domains by limiting the actions of enhancers and silencers through a "positional blocking" mechanism (Bell et al. 2001). In their studies, Yu et al. (2004) examined the insulator at the imprinting control region (ICR) of the Igf2-H19 locus, an element that binds CTCF and limits the extent of expression in a parent origin-specific manner. PAR containing >10 ADP-ribose units was only detected on the maternally inherited (unmethylated) ICR. Furthermore, the presence of PAR at the ICR was dependent on functional CTCF-binding sites in the maternal allele, but not the paternal allele. ICR insulator function was shown to be sensitive to 3-AB, a general inhibitor of PARPs that rapidly reduced the levels of PAR, but not CTCF, at the ICR, and activated expression of the normally repressed maternal Igf2 allele. The generality of these results was demonstrated using a ChIP-chip approach with a microarray containing a library of CTCF-binding sites. Nearly 80% of the CTCF target sites on the array were immunoprecipitated by antibodies to PAR and CTCF, although CTCF was not directly shown to be the PARylated protein at these sites. In an insulator trap assay, the insulator function of most of the CTCF-binding sites in the library was sensitive to 3-AB. Collectively, these results support a role for PAR in the activity of insulators, possibly as a covalent attachment to CTCF. This conclusion is extended by further studies showing that CTCF immunoprecipitated from cells is PARylated and that purified CTCF can be PARylated in vitro by PARP-1. Additional studies will be required to determine if CTCF is, indeed, the target of PARylation at insulators and, if so, exactly how PARylation affects CTCF's role in insulator function.
Mitotic apparatus function
PARP-1 and at least five other PARP family members (i.e., PARP-2, PARP-3, VPARP, tankyrase 1, and tankyrase 2) are associated with various components of the mitotic apparatus, which is required for the accurate segregation of chromosomes during cell division (Smith 2001; Amé et al. 2004). For example, both PARP-1 and PARP-2 localize to mitotic centromeres, the chromosomal regions where kinetochores form to capture microtubules from the mitotic spindle (Earle et al. 2000; Saxena et al. 2002a,b). In addition, several PARP enzymes, including PARP-1, PARP-3, and tankyrases, are associated with centrosomes, the cellular microtubule organizing center that functions as the spindle pole during mitosis (Smith and de Lange 1999; Kaminker et al. 2001; Augustin et al. 2003; Kanai et al. 2003). Furthermore, PAR and PAR-metabolizing enzymes, including VPARP and PARG, also localize to the mitotic spindle (Kickhoefer et al. 1999; Chang et al. 2004). The colocalization of PARPs and PARG to the mitotic apparatus suggests that dynamic regulation of PAR metabolism may play a key role in the control of mitotic functions. In support of this, the cellular concentration of PAR increases dramatically during metaphase and anaphase-telophase of mitosis (Bakondi et al. 2002).
Recent studies by Chang et al. (2004) suggest that the increased production of PAR during mitosis plays an essential role in the assembly and structure of bipolar spindles. Antibody staining for PAR using isolated spindles assembled in cycled Xenopus egg extract and in somatic cells demonstrated that PAR colocalizes with spindle microtubules, and is enriched at spindle poles and kinetochores. The addition of PARG or PAR antibody results in a rapid breakdown of spindle structure. The results of Chang et al. (2004) demonstrate a requirement for PAR in the assembly and structure of bipolar spindles, although the exact role and targets of PAR action at spindles are not known. PAR may act as a signaling component to regulate spindle proteins through covalent and noncovalent modifications. Alternatively, PAR may serve as a structural component of the spindle to provide a matrix for proper spindle assembly and function (Chang et al. 2004). In sum, a growing body of evidence has implicated roles for PAR and PAR metabolizing enzymes in the function of the mitotic apparatus. Determination of their specific roles in mitotic processes will require additional functional assays.
|
The physiology and pathophysiology of PAR and PARP-1 |
|---|
|
TopAbstractStructural and functional...The chemical biology of...Catabolism of PAR by...The molecular and cellular...The physiology and...Nuclear NAD+ metabolism and...How can PARP-1 and...AcknowledgmentsReferences |
|---|
Genome maintenance
Both PARP-1 and PARP-2 are important for the maintenance of genome stability, and depletion of either one alone leads to a loss of genetic stability. For example, Parp-1-/- mice or Parp-1-/- embryonic fibroblasts exhibit defective SSB repair and increased homologous recombination, sister-chromatid exchange, and micronuclei formation (Table 1; de Murcia et al. 1997; Wang et al. 1997; D'Amours et al. 1999; Bürkle 2001b). In addition, PARP-1 deficiency causes increased deletion mutations and insertions/rearrangements after treatment with alkylating agents (Shibata et al. 2005). Likewise, Parp-2-/- mice or Parp-2-/- embryonic fibroblasts exhibit increased sensitivity to ionizing radiation, increased radiation-induced chromosomal breaks, increased sister-chromatid exchange, G2/M cell cycle block, and a high level of mitotic chromosomal aberrations (Menissier de Murcia et al. 2003). Although neither PARP-1 nor PARP-2 is individually required for survival in the absence of genotoxic insults, and Parp-1-/- and Parp-2-/- mice are viable and fertile (Wang et al. 1995; de Murcia et al. 1997; Masutani et al. 1999; Menissier de Murcia et al. 2003), Parp-1-/- Parp-2-/- mice exhibit embryonic lethality prior to E8.0 (Menissier de Murcia et al. 2003). Interestingly, Parp-1+/- Parp-2-/- mice show female-specific embryonic lethality at E9.5 associated with X-chromosome instability in those females, but not in males (Table 1; Menissier de Murcia et al. 2003). Collectively, the available data indicate that PARP-1 and PARP-2 possess both overlapping and nonredundant functions that are required for the maintenance of genomic stability.
Carcinogenesis
Carcinogenesis is a multistep process involving aberrations in a variety of cellular processes, including genome maintenance, cell cycle control, proliferation, differentiation, and cell death. PARP-1 and PARylation have been implicated in all of these processes, suggesting possible connections between PARP-1 function and carcinogenesis (Masutani et al. 2003). Although Parp-1-/- mice do not exhibit a propensity for the development of early onset tumors, they do show increased tumor formation in several chemically induced and transgenic cancer models (Table 1; Masutani et al. 2005). For example, Parp-1-/- mice treated with certain alkylating agents [e.g., N-nitrosobis(2-hydroxypropyl)amine and azoxymethane] have a higher incidence of colon and liver cancers than Parp-1+/+ mice (Tsutsumi et al. 2001; Nozaki et al. 2003). The incidence of tumors in response to some other DNA-damaging agents, however, is not different between Parp-1-/- and Parp-1+/+ mice (Masutani et al. 2005), suggesting that the nature of the DNA damage dictates the specific contribution of PARP-1 to cancer prevention.
Other mouse models have suggested important functional interactions between PARP-1 and other genome maintenance factors in preventing carcinogenesis. For example, Parp-1 knockout in severe combined immuno-deficiency (SCID) mice, which harbor a mutant gene encoding the DNA-PK catalytic subunit, promotes a dramatic increase in the incidence of T-cell lymphoma (Table 1; Morrison et al. 1997). As noted above, PARP-1 is likely to play a role in the recruitment and activation of DNA-PK complexes at sites of DSB repair (Ruscetti et al. 1998; Ariumi et al. 1999). Simultaneous depletion of PARP-1 and p53, a tumor-suppressor gene required for cell cycle checkpoints and apoptotic cell death following DNA damage (Hofseth et al. 2004), has produced seemingly conflicting results about the role of PARP-1 in carcinogenesis in two different mouse models. Tong et al. (2001) showed that PARP-1 deficiency from biallelic deletion of exon 2 in the Parp-1 gene accelerates spontaneous tumor formation in p53-/- mice relative to PARP-1-sufficient p53-/- mice. In contrast, Conde et al. (2001) showed that PARP-1 deficiency from biallelic deletion of exon 4 in the Parp-1 gene increases the latency for tumor formation in p53-/- mice relative to PARP-1-sufficient p53-/- mice (Table 1). The reason for the differences in these studies is not clear, but may relate to the specific exons targeted for disruption in the PARP-1 gene or other genetic factors. Both studies do agree, however, that Parp-1-/- p53-/- cells/animals show more genomic instability than Parp-1+/+ p53-/- cells/animals, pointing to functional cooperation between PARP-1 and p53 in genome maintenance, the cornerstone of cancer prevention. Collectively, the available data indicate that normal PARP-1 function is required to prevent cancer formation in response to DNA damage. Furthermore, PARP-1 is likely to cooperate with PARP-2 and other genome maintenance factors to maintain genomic integrity and prevent spontaneous tumor formation.
Aging
The accumulation of macromolecular damage, especially damage to genomic DNA, underlies the aging process. Given their roles in cellular responses to various types of genomic insults, PARP-1 and other PARPs have been implicated in aging and longevity (Beneke and Burkle 2004; Bürkle et al. 2005). The initial connection came from studies showing that the cellular capacity of blood mononuclear cells from various mammalian species to synthesize PAR is positively correlated with the life span of the species (Grube and Burkle 1992). Recent studies have suggested important physical and functional interactions between PARP-1 and WRN, a member of the RecQ helicase family that forms part of the DNA replication complex, in DNA damage responses and in the prevention of chromosome defects and cancer (Lebel et al. 2003; von Kobbe et al. 2003). Interestingly, the gene encoding WRN is mutated in Werner Syndrome, a rare genetic disorder that induces premature aging in humans (Lee et al. 2005). In addition to their roles in counteracting DNA damage and improper chromosome segregation, as noted above, PARP-1 and other PARPs (e.g., tankyrases and PARP-2) play a role in maintaining telomere length and structure (Cook et al. 2002; Dantzer et al. 2004; Dynek and Smith 2004; O'Connor et al. 2004). Telomere maintenance is important for the prevention of cellular senescence and replicative crisis in replicating cells. Although much of the current evidence suggesting a role for PARPs in aging is correlative, the potential connections are intriguing and warrant further examination.
Inflammatory responses
PARP-1 has been implicated in pathophysiological inflammatory responses. As such, Parp-1-/- mice show resistance in various models of inflammation, including streptozotocin-induced diabetes and LPS-induced septic shock (Oliver et al. 1999; Mabley et al. 2001; Hassa and Hottiger 2002), whereas Parg2-3/2-3 mice (i.e., selective depletion of the 110-kDa isoform) show increased susceptibility (Table 1; Cortes et al. 2004). PARP-1's role in promoting pathophysiological inflammatory responses appears to be twofold. First, in response to genotoxins, PARP-1 induces necrotic cell death, which releases intracellular components into the surrounding tissue, causing tissue damage and promoting inflammation (see above). Second, PARP-1 functions as a coactivator of transcription factors that regulate immune and inflammatory response genes (e.g., NF-B and AP-1) (Oliver et al. 1999; Hassa and Hottiger 2002; Andreone et al. 2003). In fact, NF-B- and AP-1-regulated pro-inflammatory genes are down-regulated in Parp-1-/- mice or Parp-1-/- immune cells (Zingarelli et al. 1998; Shall and de Murcia 2000; Hassa and Hottiger 2002) and up-regulated in immune cells treated with a PARG inhibitor (Rapizzi et al. 2004). Collectively, the available data indicate that PARP-1 plays an important role in pathophysiological inflammatory responses, associated primarily with the dysregulation of NF-B function. The mechanisms by which PARP-1 functions as a coactivator of NF-B are not clear, but are likely to involve direct interactions with NF-B and may not require PARP-1's enzymatic activity (Oliver et al. 1999; Hassa and Hottiger 2002).
Neuronal function
PARP-1 has been implicated in neuronal pathology and, more recently, in normal neuronal function as well. Most of the work in this area has focused on PARP-1's role in inducing necrotic cell death upon neuronal injury, including excitotoxicity (i.e., glutamate-mediated neuronal death), ischemia/reperfusion, and traumatic injury (Szabo and Dawson 1998; Ha and Snyder 2000; Cole and Perez-Polo 2004). Neuronal injury generates significant amounts of oxygen- and nitrogen-derived free radicals (e.g., superoxide, hydroxyl radical, nitric oxide), which are particularly toxic to neuronal cells (Cole and Perez-Polo 2004). These reactive oxygen species set off a cascade of DNA damage, PARP-1 activation, and necrosis in neuronal cells. Chemical inhibition or genetic depletion of PARP-1 in mammalian systems blocks these neurotoxic effects (Zhang et al. 1994; Ha and Snyder 2000; Mandir et al. 2000; Cole and Perez-Polo 2004). In contrast, loss-of-function mutation of PARG in Drosophila promotes progressive neurodegeneration and shortened life span in surviving adult animals (Table 1; Hanai et al. 2004). These results suggest that regulated PAR metabolism is important in maintaining normal neuronal functions.
Recent studies have also suggested a role for PARP-1 in learning and memory (Satchell et al. 2003; Cohen-Armon et al. 2004). Cohen-Armon et al. (2004) used the marine mollusk Aplysia as a model system to examine the role of a PARP-1 homolog in long-term memory formation. Interestingly, they found that PARP-1 enzymatic activity is increased in Aplysia neuronal cells in two different learning paradigms: response to an aversive stimulus and feeding-related learning. Treatment with the PARP-1 inhibitor 3-AB during training blocked long-term memory, whereas treatment with serotonin, which facilitates long-term memory formation, increases PARylation of histone H1. The authors speculate that PARylation of nuclear proteins that regulate the expression of genes involved in long-term memory formation might mediate the effects they observed. More extensive studies, however, are needed to determine how PARP-1 is activated and the specific role it plays during long-term memory formation.
|
Nuclear NAD+ metabolism and the regulation of PARP-1 activity |
|---|
|
TopAbstractStructural and functional...The chemical biology of...Catabolism of PAR by...The molecular and cellular...The physiology and...Nuclear NAD+ metabolism and...How can PARP-1 and...AcknowledgmentsReferences |
|---|
). Furthermore, in biochemical assays, the availability of NAD+ regulates PARP-1's effects on chromatin structure and transcription (Kim et al. 2004). How NAD+ functions as a nuclear signal to regulate PARP-1 activity in vivo, however, is unclear. Studies examining the biosynthesis of NAD+ have begun to provide some clues.
|
). The salvage pathways begin with either nicotinamide or nicotinic acid, collectively referred to as niacin or vitamin B3. The de novo pathway leads from tryptophan to quinolinate, which connects to the nicotinic acid salvage pathway through nicotinic acid mononucleotide (NaMN). In most mammalian tissues, nicotinamide, a product of NAD+ hydrolysis by PARP-1 and SIRT1, is likely the most relevant NAD+ precursor for the regulation of PARP-1 functions (Rongvaux et al. 2003). Nicotinamide is first converted to NMN by nicotinamide phosphoribosyltransferase (NAMPT). The production of NAD+ from NMN and ATP is then catalyzed by a family of NMN adenylyltransferases (NMNATs) (Magni et al. 2004). The enzymes possessing NMNAT activity also have NaMN adenylyltransferase (NaMNAT) activity and are, therefore, required for all NAD+ biosynthetic pathways. Interestingly
