Analysis of O-acetyl-ADP-ribose as a target for Nudix ADP-ribose hydrolases.

The Sir2 family of NAD(+)-dependent histone/protein deacetylases has been implicated in a wide range of biological activities, including gene silencing, life span extension, and chromosomal stability. Recent evidence has indicated that these proteins produce a novel metabolite O-acetyl-ADP-ribose (OAADPr) during deacetylation. Cellular studies have demonstrated that this metabolite exhibits biological effects when microinjected in living cells. However, the molecular targets of OAADPr remain to be identified. Here we have analyzed the ADP-ribose-specific Nudix family of hydrolases as potential in vivo metabolizing enzymes of OAADPr. In vitro, we found that the ADP-ribose hydrolases (yeast YSA1, mouse NudT5, and human NUDT9) cleaved OAADPr to the products AMP and acetylated ribose 5'-phosphate. Steady-state kinetic analyses revealed that YSA1 and NudT5 hydrolyzed OAADPr with similar kinetic constants to those obtained with ADP-ribose as substrate. In dramatic contrast, human NUDT9 was 500-fold less efficient (k(cat)/K(m) values) at hydrolyzing OAADPr compared with ADP-ribose. The inability of OAADPr to inhibit the reaction of NUDT9 with ADP-ribose suggests that NUDT9 binds OAADPr with low affinity, likely due to steric considerations of the additional acetylated-ribose moiety. We next explored whether Nudix hydrolytic activities against OAADPr could be observed in cell extracts from yeast and human. Using a detailed analysis of the products generated during the consumption of OAADPr in extracts, we identified two robust enzymatic activities that were not consistent with the known Nudix hydrolases. Instead, we identified cytoplasmic esterase activities that hydrolyze OAADPr to acetate and ADP-ribose, whereas a distinct activity residing in the nucleus is consistent with an OAADPr-specific acetyltransferase. These findings establish for the first time that select members of the ADP-ribose hydrolases are potential targets of OAADPr metabolism. However, the predominate endogenous activities observed from diverse cell extracts represent novel enzymes.

The Silent Information Regulator 2 (Sir2) family is a highly conserved group of genes present in the genomes of organisms ranging from archaebacteria to eukaryotes (1). The encoded Sir2 proteins are involved in diverse processes ranging from regulation of gene silencing to DNA repair and lifespan extension. The best characterized of these is the product of the Saccharomyces cerevisiae Sir2 gene, which is involved in silencing at telomeres (2)(3)(4), at the mating type loci (2,5), and in the ribosomal DNA (6 -9). Besides its role in silencing, Sir2 has been implicated in other cellular processes, including the repair of chromosomal double-strand breaks through non-homologous end-joining (10), cell cycle progression, and chromosome stability (11). The molecular function of Sir2, which results in the above silencing and cellular processes, has only recently been explored (reviewed in Refs. [12][13][14][15]. Initial examination of Sir2 molecular function reported a weak ADP-ribosyltransferase activity (16,17). Further analysis revealed Sir2 possesses an intrinsic NAD ϩ -dependent protein and histone deacetylation activity (18 -23). More recently it has been shown that Sir2 couples deacetylation to the hydrolysis of a high energy bond in NAD ϩ and transfers the acetyl group from its protein substrate to ADPr to generate a novel compound, Oacetyl-ADP-ribose (OAADPr) 1 (22)(23)(24)(25). Evidence that OAADPr production is evolutionarily conserved among Sir2 homologs, and that this metabolite may perform a critical function for this class of enzymes has recently been explored (26). Furthermore, a quantitative microinjection assay into starfish oocytes and blastomeres demonstrated that OAADPr can cause a delay/ block in oocyte maturation and embryo cell division in blastomeres, proposing that OAADPr may posses a unique bioactivity that may be linked to reported physiological effects/ functions of Sir2 (26). Thus OAADPr could act as a second messenger or metabolite in which other enzymes/proteins may utilize OAADPr to elicit the observed cellular response. Therefore, to clearly understand the biological functions of OAADPr, the proteins and enzymes that utilize/bind OAADPr should be identified and characterized.
A class of enzymes that are capable of hydrolyzing compounds of ADPr and possibly OAADPr include the Nudix Family of hydrolases. This family of enzymes catalyzes the hydrolysis of a nucleoside diphosphate linked to another moiety x, hence the acronym "Nudix" (27,reviewed in Ref. 28). They are found in archaea, eubacteria, animal, plant, and fungi and are characterized by the highly conserved array of amino acids GX 5 EX 7 REUXEEXGU, where U represents a bulky, hydrophobic amino acid, usually Ile, Leu, or Val and X represents any amino acid (27). Based on predictive (29) and structural studies (30), the Nudix box designates a unique loop-helix-loop motif, involved in the binding of the substrate (31) and in catalysis (32,33). The nucleoside diphosphate linkage is the common feature of the wide range of substrates for the family, which include NADH, dinucleoside polyphosphates, nucleotide sugars, and (deoxy)ribonucleoside triphosphates. These substrates are thought to be either potentially deleterious compounds, cell signaling molecules, regulators, or metabolic intermediates whose concentrations require modulation during fluctuations of the cellular environment (34). Hence, proposed physiological functions of these enzymes are to eliminate potentially toxic nucleotide metabolites from the cell (35,36) and to regulate the concentrations of nucleotide cofactors and signaling molecules for optimal cell growth and survival (37).
The purpose of the current study is to investigate the Nudix hydrolase family of enzymes as viable targets of OAADPr metabolism. To explore this idea, we examined the ability of Nudix enzymes, specifically the ADP-ribose (ADPr) hydrolases, to catalyze the hydrolysis of OAADPr. We found that select members of the ADP-ribose hydrolases are capable of efficient hydrolysis of OAADPr, suggesting that certain members may be crucial in regulating in vivo levels of OAADPr. To examine this possibility, we analyzed endogenous enzymatic activities toward OAADPr from a variety of cellular sources. We show that human HeLa, mouse 3T3 fibroblast, and yeast cell extracts harbor robust enzymatic activities toward OAADPr. However, we demonstrate that the products generated from these cellular activities are not consistent with known Nudix hydrolase activities. This finding raises important biological questions concerning the physiological functions of OAADPr metabolizing enzymes detected in cellular extracts as well as the observed diverse catalytic efficiency of ADPr hydrolases on OAADPr.
Preparation of Nuclear Extracts-Cell monolayers were washed twice with ice-cold phosphate-buffered saline, pH 7.4, at 4°C and removed from the culture dish by scraping. Cells were pelleted by centrifugation using a clinical centrifuge for 10 min at 4°C. The cell pellet was resuspended in phosphate-buffered saline, pH 7.4, and the resulting suspension was transferred to microcentrifuge tubes. Cells were repelleted in a microcentrifuge by spinning at 14,000 rpm (16,000 ϫ g) for 20 s at 4°C, then lysed by the addition of ice-cold hypotonic solution (buffer A) consisting of 10 mM HEPES, pH 8, 1.5 mM MgCl 2 , 10 mM KCl, 0.5 mM DTT, 0.5 mM PMSF, 0.5% Nonidet P-40, 1 mg/ml leupeptin, and 1 mg/ml aprotinin and incubating the suspension on ice for 5 min. The suspension was recentrifuged, and the supernatant containing the cytoplasmic fraction was collected and stored and Ϫ80°C. The pellet representing the nuclei were lysed in an ice-cold solution (buffer C), consisting of 20 mM HEPES, pH 8, 1.5 mM MgCl 2 , 420 mM NaCl, 0.2 mM EDTA, 1 mM DTT, 0.5 mM PMSF, 1 mg/ml leupeptin, and 1 mg/ml aprotinin. Cellular debris was removed by centrifugation, and the supernatant containing the DNA binding proteins was combined with an equal volume of buffer D (20 mM HEPES, pH 8, 100 mM KCl, 0.2 mM EDTA, 20% glycerol, 1 mM DTT, 0.5 mM PMSF, 1 mg/ml leupeptin, and 1 mg/ml aprotinin). Extracts were stored at Ϫ80°C until use.
Preparation of Yeast Extracts-Yeast cells were cultured to mid-log phase in YPD medium (1% (w/v) yeast extract, 2% peptone, 2% dextrose) with vigorous shaking at 30°C. Cells were harvested by centrifugation for 5 min at 3,000 rpm (1,600 ϫ g). The net weight of the yeast cells was determined. This weight was said to be equal to the packed cell volume (in milliliters), and for all subsequent steps was considered as 1 volume. Cells were resuspended in 4 volumes if ice-cold water and immediately centrifuged for 5 min at 3,500 rpm (2,170 ϫ g) and 4°C. The cell pellet was resuspended in 3 volumes of glass bead disruption buffer (20 mM Tris-Cl, pH 7.9, 10 mM MgCl 2 , 1 mM EDTA, 5% glycerol, 1 mM DTT, 0.3 M ammonium sulfate, 1 mM PMSF), and a cell paste was created. Four volumes of chilled acid-washed beads (Sigma) were added to the cell paste, the suspension was vortexed at maximum speed for 30 s, and the tubes were placed on ice for 4 min. Vortexing was repeated three times. The glass beads were allowed to settle, and the supernatant was decanted. To remove insoluble protein, the supernatant was centrifuged for 60 min at 12,000 rpm (15,800 ϫ g). The supernatant, which represents the crude extract, was collected and stored at Ϫ80°C.
Conversion Assay Using HeLa Cell Nuclear Extract-The presence of an enzymatic activity in HeLa cells that can metabolize OAADPr was quantitated by incubating O-[ 3 H]AADPr with HeLa cell nuclear extract and monitoring the loss of radioactivity in the O-[ 3 H]AADPr peak when resolved by reverse-phase HPLC. The standard incubation mixture (100 l) contained 1.35 mM O-[ 3 H]AADPr and 500 g of HeLa cell nuclear extract. After 0 and 60 min the reaction was terminated by removing 20-l aliquots of this reaction and added them to 480 l of 0.05% trifluoroacetic acid/H 2 O. All samples were injected onto a Beckman Biosys 510 HPLC system and a Vydac C18 (1.0 ϫ 25 mm) small pore preparative column (Vydac, Hesperia, CA). Products from the enzymatic conversion of OAADPr were separated using a gradient system comprising solvent A (0.05% trifluoroacetic acid/H 2 0) and solvent B (0.02% trifluoroacetic acid/acetonitrile) and using a constant flow rate of 4 ml/min. Upon injection of the sample, the HPLC was run isocratically in solvent A for 5.0 min followed by a linear gradient of 0 -8% B over a 20-min period. The A 260 absorbing product derived from OAADPr had a retention time of 6 min and was collected directly from the HPLC after passage through the detector, frozen at Ϫ80°C, lyophilized, and then resuspended in the appropriate buffer for product analysis and identification. In verification of the HeLa cell A 260 absorbing product derived from OAADPr, either 50 M ADPr or ADP were co-injected with the OAADPr and HeLa cell reaction mixture.
Conversion Assay Using OAADPr and Nudix Enzymes-The standard incubation mixture (50 l) contained 50 mM Tris-Cl, pH 7.2, 16 mM MgCl 2 , 1.5 mM OAADPr, 1 g of YSA1 enzyme. After 10 min at 37°C, the reaction was stopped by the addition of 20 mM EDTA buffer. Control reactions were incubated without any Nudix enzyme. All reactions were analyzed by HPLC as described above. To determine the elution position of AMP, 50 M of AMP standard was used.
Purification of OAADPr-The generation and purification of OAADPr was performed as described previously (26).
Colorimetric Phosphate Assay for Nudix Activity-Enzyme kinetics were measured by converting substrates, OAADPr or ADPr, to products AMP and acetylated ribose 5Ј-phosphate or AMP and ribose 5Ј-phosphate, respectively. Calf intestinal alkaline phosphatase (AP) liberated inorganic phosphate, which was then detected by the colorimetric procedures of Ames and Dubin (38). Briefly, the reaction mixture was quenched with 850 l of an ascorbic-molybdate solution. This solution, prepared fresh daily, contained 1 part of 10% ascorbic acid to 6 parts of 0.42% ammonium molybdate⅐4H 2 O in 1 N H 2 SO 4 . The mixture was then incubated for 20 min at 45°C, and absorbance was read at 820 nm. A standard curve was generated each day. Two mol of phosphate was liberated for each mole of substrate hydrolyzed. The reaction mixture (150 l) for Nudix kinetic experiments contained 50 mM Tris-Cl, pH 7.5, 7 mM MgCl 2 , 0.010 -2.0 mM substrate, and 5 units of AP. In the assays, YSA1 ranged from 9.8 to 19.6 ng, mNudT5 ranged from 100 to 300 ng, and NUDT9 ranged from 100 to 200 ng for ADPr assays or from 900 to 1800 ng for OAADPr assays. The reaction mixture was incubated at 37°C for 1-15 min and quenched with the ascorbic-molybdate solution. Linearity of rates (i.e. initial velocities) was verified by measuring activities at various enzyme concentrations and by monitoring the reaction time course for up to 15 min. Product formation was linear with respect to enzyme concentrations and time. Typical assays were run between 3 and 7 min, where 5-20% substrate was converted to product. Substrate saturation curves were fitted to the Michaelis-Menten equation, v 0 ϭ (k cat * [S])/(K m ϩ [S]), using non-linear least squares analysis (Kaleidagraph, Synergy Software, Reading, PA). Due to the inability to saturate, the k cat /K m for NUDT9 with OAADPr was determined from the slope of the line fitted by linear-least squares regression.
Product Determination for HeLa Cell Activity-The A 260 absorbing product of HeLa cell activity was collected from the HPLC and used as a substrate in subsequent reactions with AP and/or NUDT9. The standard reaction mixture (150 l) contained 50 mM Tris-Cl, pH 7.5, 7 mM MgCl 2 , 50 or 100 M substrate (concentration determined by extinction coefficient ⑀ 259 nm ϭ 15.4 mM Ϫ1 cm Ϫ1 ), 5 units of AP and either the presence or absence of 2.0 g of NUDT9. The reaction mixture was incubated at 37°C for 10 min, and free phosphate was detected by quenching with the ascorbic-molybdate solution (as described above).
Product Determination for Nudix Activity-The standard reaction mixture as described above was scaled up, quenched with trifluoroacetic acid (1% final concentration), and injected onto the HPLC as described above. Two products were collected: an A 260 absorbing species and a species containing the acetate that was detected by using O-[ 3 H]AADPr as the substrate. Each product, to a final concentration of 50 M, was introduced to separate reaction mixtures (150 l) containing 50 mM Tris-Cl, pH 7.5, 7 mM MgCl 2 , and 5 units of AP. The reaction mixture was incubated for 5 min at 37°C, and free phosphate was detected by quenching with the ascorbic-molybdate solution (as described above).
Matrix-assisted Laser Desorption Ionization Mass Spectrometry Analyses-MALDI mass spectrometry was performed at the Environmental Health Sciences Center, Oregon State University, as described previously (22).

Certain Members of the ADPr Hydrolase Family Efficiently
Hydrolyze OAADPr-We performed initial literature searches to identify enzymes capable of metabolizing the closest known metabolite of OAADPr, ADPr. The most promising finding was the Nudix hydrolase family of enzymes, which consists of a group of phosphoanhydrases present in organisms ranging in complexity from viruses to humans (34). The substrates of this enzyme family include nucleoside triphosphates, co-enzymes, sugar nucleotides, and dinucleoside polyphosphates (34). Interestingly, ADPr is a favored substrate for several members of the Nudix hydrolase family (Refs. 34, 39 -43; reviewed in Ref. 27). We have demonstrated that echinoderm oocytes (26), human ( Fig. 1) and S. cerevisiae cell extracts (data not shown) possess robust enzymatic activities that metabolize OAADPr, although the products of these reactions had not been elucidated. It was therefore logical to screen a number of Nudix ADPr hydrolases as potential enzymes responsible for the robust metabolism of OAADPr observed in these cellular extracts. To explore the idea that OAADPr may be a substrate for these hydrolases, we assayed the ADP-sugar pyrophosphatase, YSA1, from S. cerevisiae (YSA1) (40), the murine ADP-sugar diphosphatase mNudT5 (42), and the human ADPr diphosphatase NUDT9 (44). These enzymes were chosen because their preferred substrates are ADP-sugars with preference for ADPr (44,45). A common mechanism of virtually all members of the Nudix hydrolase superfamily is a nucleophilic attack by water on a pyrophosphate linkage in the substrate (43). This nucleophilic attack occurs on the alpha or beta phosphorus producing a reaction product corresponding to the nucleoside 5Ј-monophosphate (43). Because the products generated from ADPr hydrolysis are AMP and ribose 5Ј-phosphate (39,45,46), the reaction products predicted from the conversion of OAADPr would be AMP and acetylated ribose 5Ј-phosphate. To provide evidence for this prediction, YSA1 was incubated with either OAADPr or ADPr and the products were analyzed by reversephase HPLC and a colorimetric phosphatase assay.
Reverse-phase HPLC of this reaction indicated that the disappearance of substrate ADPr or OAADPr was coincident with the appearance of AMP (Fig. 2, A and B, respectively). To confirm the elution position of AMP, 100 M of an authentic AMP standard was co-injected onto the HPLC, and the elution profile was monitored (Fig. 2C). As expected, UV absorption at 260 nm demonstrated that AMP co-eluted with the A 260 peak seen with YSA1 enzymatic turnover of OAADPr (Fig. 2B). To detect the cleavage of OAADPr, we generated radiolabeled O-[ 3 H]AADPr using the Sir2 homolog HST2 and the synthetic, acetylated histone H3 peptide, [ 3 H]AcH3, where the radioactivity resides on the methyl position of the acetyl group. Cleavage of OAADPr by the ADPr hydrolase YSA1 between the ␣ and ␤ phosphates would be expected to generate acetylated ribose 5Ј-phosphate as the other reaction product. Upon O-[ 3 H]AADPr conversion by YSA1, we found a single tritiumlabeled product eluting at fraction 6 in reverse-phase HPLC analysis (Fig. 2D). Next, we attempted to verify this tritiated product as acetylated ribose 5Ј-phosphate. To accomplish this, YSA1 was incubated with O-[ 3 H]AADPr, and the reaction products were collected after separation by HPLC. If the product is acetylated ribose 5Ј-phosphate then we would expect to retrieve equal moles of acetyl-ribose to phosphate. The acetate-containing species was quantitated by using the specific activity of O-[ 3 H]AADPr (1.6 cpm/pmol) in the initial reaction. Using the phosphatase assay, a fraction containing 55 M acetylated product yielded 44.8 Ϯ 0.6 M free phosphate, indicating that this enzymatic product harbors a molar equivalent of phosphoryl and acetyl groups and is consistent with acetylated ribose 5Ј-phosphate as the cleaved product. To confirm that the other product contained the second molar equivalent of phosphate, the A 260 species predicted to be AMP yielded 40.8 Ϯ 2.1 M free phosphate. Combined, these results indicate that each product of the ADPr hydrolase reaction contains approximately equivalent molar amounts of phosphate, indicating cleavage of OAADPr between the ␣ and ␤ phosphates, producing reaction products in accordance with AMP and acetylated ribose 5Ј-phosphate.
These findings reveal the capacity of the ADPr hydrolase YSA1 to metabolize OAADPr to the reaction products AMP and acetylated ribose 5Ј-phosphate. To further characterize OAADPr as a substrate for the ADPr hydrolase family of enzymes, we next determined the steady-state kinetic constants of YSA1, mNudt5, and NUDT9 for OAADPr. For these assays, enzyme was incubated with varying concentrations of substrate, and the initial velocities were measured using a colorimetric phosphate assay described under "Experimental Procedures." Saturation curves were constructed and kinetic constants calculated from fits using the Michaelis-Menten equation. The activities toward OAADPr were compared with ADPr (Table I). Our results with ADPr were similar to those reported previously (Table I) 3A) and with a k cat /K m value ϳ3.5-fold less than with ADPr as substrate. The mNudT5 hydrolyzes each substrate at a similar maximal rate (0.8 -0.9 s Ϫ1 ), and the ratio of k cat /K m values is nearly 1:1 (Table I) (Fig. 3B). These data demonstrate the capacity of YSA1 and mNudT5 to bind and hydrolyze OAADPr with similar kinetic parameters as those for ADPr. Strikingly, this is not the case with NUDT9, which exhibits a k cat /K m value that is 500-fold less than with ADPr (Fig. 3C). Because the saturation curve for OAADPr (Fig. 3C) was linear through as high as 800 M substrate, accurate V max or K m values could not be obtained. The k cat /K m value (448 Ϯ 24 M Ϫ1 s Ϫ1 ) for NUDT9 with OAADPr was determined from the slope of the line fitted by linear-least squares regression (Table I). If it is assumed that V max is unchanged or 2-fold slower with OAADPr as substrate compared with ADPr (a reasonable assumption based on the results with NudT5 and YSA1, see Table I), then the Michaelis-Menten equation can be used to fit the data and obtain an estimated range for K m . A range of 7.9 -16.4 mM was obtained for NUDT9 using OAADPr as substrate. Compared with ADPr, the K m value with OAADPr is 240-to 500-fold higher, suggesting that the low activity of NUDT9 against OAADPr stems from poor binding affinity, rather than inefficient chemical catalysis. To provide additional evidence that indeed NUDT9 exhibits low binding affinity for OAADPr, we analyzed the ability of OAADPr to inhibit the hydrolysis of ADPr by NUDT9. If poor binding affinity were the cause of the observed high K m values and low k cat /K m values, then we would predict that OAADPr would have little affect on the ADPr hydrolytic activity of NUDT9. After incubating the enzyme with 12-96 M ADPr and up to 400 M OAADPr, we saw no significant inhibition (within the error of the assay), consistent with a lack of significant binding by OAADPr. It is reasonable to conclude that the decreased affinity for OAADPr is a result of steric hindrance caused by the acetyl group on the ribose ring of substrate. Overall, these data suggest that certain ADPr hydrolase enzymes, in this instance mNudT5 and YSA1, possess catalytic activity specific for OAADPr with similar kinetic parameters to those obtained for ADPr. This demonstrates for the first time the identification of select ADPr hydrolase family members capable of the enzymatic turnover of OAADPr. Human Cell Extracts Harbor Two Non-Nudix OAADPr Metabolizing Activities-As previously indicated, we found that TABLE I Kinetic parameters for YSA1, mNudT5, and NUDT9 using ADPr or OAADPr as substrates The colorimetric procedure described under "Experimental Procedures" was used, and substrates were varied between 0.010 and 2.0 mM. The steady-state kinetic constants were calculated from non-linear regression analysis using the data presented on the Michaelis-Menten plots, as in Fig. 3 c Due to the inability to saturate, the k cat /K m for NUDT9 with OAADPr was determined from the slope of the line fitted by linear-least squares regression. An estimated range for the K m value was determined by fixing the V max at the same rate or 2-fold slower than that obtained with ADPr as substrate (a reasonable assumption based on the results with NudT5 and YSA1, see above) and fitting the data to the Michaelis-Menten equation.
d These data from Ref. 44. human and yeast cells possess robust enzymatic activities that metabolize OAADPr (Fig. 1). To explore the hypothesis that the metabolism of OAADPr observed in cell extracts resulted from a Nudix hydrolase activity, we set out to identify the products of this reaction. If the reaction products were indeed AMP and acetylated ribose 5Ј-phosphate, we could conclude that a Nudix-like enzyme may be responsible for the observed turnover. Initially we analyzed whether these enzyme activities could be localized to either the cytoplasm or nucleus. To accomplish this, HeLa nuclear and cytoplasmic pools were fractionated and assayed for OAADPr metabolizing enzymes. Both nuclear and cytoplasmic extracts were incubated with O-[ 3 H]AADPr for various times, and the loss of radioactivity in the reverse-phase HPLC peak corresponding to O-[ 3 H]AADPr was monitored (Fig. 4). Interestingly, we observed activity in both the nuclear and cytoplasmic fractions, however, the labeled acetyl-containing product of OAADPr turnover was different between the two cellular compartments (Fig. 4). Upon O-[ 3 H]AADPr conversion by the cytoplasmic compartment, the tritium-labeled product elutes at fraction 5 in the reverse-phase HPLC analysis (Fig. 4,  left panel). The nuclear OAADPr metabolizing activity generates a tritium-labeled product that elutes at fraction number 10 (Fig. 4, right panel). Following the adenine moiety, we observed, by HPLC and UV-detection at A 260 , that both the cytoplasmic and nuclear enzymes appear to generate the same reaction product eluting at fraction 6 ( Figs. 5 and 6). Also, we have demonstrated that nuclear and cytoplasmic extracts from 3T3 fibroblasts and S. cerevisiae yeast whole cell extracts are also capable of metabolizing OAADPr, producing analogous reaction products to those seen with HeLa cytoplasmic and nuclear extracts. 2 These data suggest the existence of different cellular enzymes in both the nucleus and cytoplasm of HeLa, 3T3 fibroblast cells, and yeast extracts that utilize OAADPr. Characterization of a Cytoplasmic OAADPr Metabolizing Activity-Next, we attempted to characterize the products of these reactions and establish if they were consistent with Nudix hydrolase activity. Analysis of the HeLa cell cytoplasmic enzymatic activity toward OAADPr revealed that the reaction products are not consistent with generating AMP and acetylated ribose 5Ј-phosphate. Several biochemical criteria indicate that one of the cytoplasmic enzymes is an esterase, hydrolyzing OAADPr to ADPr and acetate. First, we verified acetate as one of the reaction products. HeLa cell cytoplasmic extract was incubated with O-[ 3 H]AADPr, the products were analyzed by reverse-phase HPLC, and the fractions were analyzed by liquid scintillation counting. We detected radioactivity resulting from the acetyl group in fraction 5 (Fig. 5E). The HPLC elution position of the radiolabeled product corresponded exactly to the position of authentic sodium [ 3 H]acetate (data not shown). Furthermore, HPLC fractions containing this product were subjected to ethyl acetate organic extraction. The radiolabeled compound was extracted with ethyl acetate, whereas negligible radioactivity remained in the aqueous layer, consistent with the presence of radiolabeled acetate. Ethyl acetate organic extraction is widely employed as a common assay for histone deacetylases, which produce acetate as the primary reaction product (47)(48)(49). Together, these data suggest that acetate is one reaction product generated from the HeLa cytoplasmic conversion of O-[ 3 H]AADPr. Consistent with an esterase like activity, ADPr would be the other predicted reaction product generated from OAADPr hydrolysis. Consequently, the production of ADPr as the other reaction product was examined. Reverse-phase HPLC and UV detection of the reaction products revealed an A 260 absorbing species eluting at fraction 6, which was consistent with a species containing the adenine ring (Fig.  5B). To provide evidence that this peak was ADPr, 10 M of an authentic ADPr standard was co-injected with the cytoplasmic reaction. UV absorption at 260 nm showed ADPr had the same retention time as the reaction product (Fig. 5C). Control experiments in the absence of a cytoplasmic extract showed minimal hydrolysis of substrate OAADPr (Fig. 5A). To further confirm that this product is indeed ADPr, we analyzed this A 260 product using the ADPr hydrolase YSA1. Because the products generated from ADPr conversion by ADPr hydrolases are AMP and ribose 5Ј-phosphate (39,45,46), conversion of the product to AMP would provide evidence that indeed ADPr is the authentic product generated by the action of the cytoplasmic enzyme on OAADPr. HPLC analysis demonstrated that YSA1 readily converted the A 260 reaction product to AMP (Fig. 5D). Collectively, these results illustrate that OAADPr is hydrolyzed to acetate and ADPr by one or more esterases located in the cytoplasm.
Characterization of a Nuclear OAADPr Metabolizing Activity-As previously discussed, we observed that the nuclear OAADPr metabolizing activity generates a tritium-labeled product that elutes at fraction 10 ( Fig. 4, right panel). This elution position is far more hydrophobic than the elution profile for acetate alone, which elutes at fraction 5 (Fig. 5E), or for acetylated ribose 5Ј-phosphate, which elutes at fraction 6 (Fig.  2D). These findings suggest that the nuclear OAADPr metabolizing activity is not consistent with either an esterase-or Nudix-like activity. To shed light on the nature of this nuclear OAADPr metabolizing enzyme, the products generated from the nuclear enzymatic turnover of OAADPr were extensively studied. HeLa cell nuclear extract was incubated with O-[ 3 H]AADPr, and the products were analyzed by reversephase HPLC, the colorimetric phosphatase assay, and MALDI mass spectral analysis. When nuclear extract was incubated with OAADPr, reverse-phase HPLC demonstrated that the disappearance of substrate O-[ 3 H]AADPr was coincident with the appearance of an A 260 absorbing species eluting at fraction 6 ( Fig. 6A). To explore the possibility that this product was ADPr, 50 M authentic ADPr was co-injected with the nuclear extract conversion assay described in Fig. 6A, and reversephase HPLC was performed. ADPr co-eluted with the A 260 peak generated from the enzymatic turnover O-[ 3 H]AADPr (Fig. 6B). Because acetate was clearly not a reaction product, we examined the possibility that the enzyme was cleaving O-[ 3 H]AADPr between the ␤-phosphate and ribose 5Ј-OH moiety to yield ADP and acetylated ribose. To explore this, co- injections with ADP were performed. Interestingly, HPLC analysis revealed that ADP eluted at the same position as ADPr (Fig. 6C). Because ADP and ADPr elute together on the HPLC, colorimetric phosphate assays were used to distinguish between these two possibilities. Either alkaline phosphatase or alkaline phosphatase and a ADPr hydrolase (NUDT9) was incubated with the unknown A 260 absorbing species, and the amount of phosphate liberated in each case was quantitated. ADP would be hydrolyzed to adenosine by alkaline phospha-tase whether NUDT9 was present or absent (1 mol of ADP producing 2 mol of P i and 1 mol of adenosine), whereas ADPr would only yield free phosphate in the presence of a Nudixspecific pyrophosphatase activity to generate products AMP and ribose 5Ј-phosphate, which are then sensitive to complete hydrolysis by alkaline phosphatase. Only in the presence of both alkaline phosphatase and NUDT9 enzyme was significant phosphate released. Of the total phosphate detected, only 13% could be liberated by alkaline phosphatase only, whereas 87% of the total required NUDT9 and alkaline phosphatase for detection. This suggests that the major product represented by the HPLC fraction 6 is indeed ADPr and not ADP.
To provide further evidence that ADPr is the product generated from the nuclear enzymatic metabolism of OAADPr, HPLC fractions containing this product were subjected to MALDI mass spectral analysis for mass determination. The adduct yielded a mass (positive molecular ion in MALDI) of 560 m/z (Fig. 7), consistent with the enzymatic formation of ADPr. Together these findings provide strong evidence that ADPr is a reaction product generated from the HeLa nuclear conversion of OAADPr. This finding, combined with the observation that the tritium-labeled product is distinct from either acetate or acetylated phospho-5Ј-ribose, suggests that the nuclear enzymatic activity is an OAADPr-specific acetyltransferase and is not consistent with an esterase or traditional Nudix hydrolase activity. Currently the properties of this acetyl-containing product in fraction 10 and the identity of the enzyme responsible for the transferase activity are under investigation.

DISCUSSION
This study demonstrates in vitro that certain members of the ADPr hydrolase family are potential targets of OAADPr metabolism, whereas others display a clear preference for ADPr turnover. Moreover, we demonstrate that HeLa, 3T3 fibroblast, and S. cerevisiae cell extracts harbor robust enzymatic activities toward OAADPr. However, our data suggest that the products generated from these endogenous enzymes are not consistent with known Nudix hydrolase reactions. Instead, we show that OAADPr is utilized by one or more cytoplasmic enzymes that catalyze OAADPr hydrolysis to acetate and ADPr, and by one or more nuclear enzymes that catalyze the transfer of the acetyl moiety to an acceptor protein or small molecule. Although no clear endogenous Nudix-like activity was observed in our studies, these observations do not rule out the possibility that such activities do exist and are important physiologically. The two observed activities may be present at much higher levels than specific Nudix hydrolases and may have greater catalytic efficiencies toward OAADPr compared with Nudix hydrolases, thereby competing with endogenous Nudix hydrolases for OAADPr turnover. Consequently, the detection of endogenous Nudix activity may be masked by these robust cellular activities. The findings, that ADPr hydrolases display unique specificity for OAADPr and that there exists other cellular enzymes capable of hydrolyzing OAADPr, provide strong circumstantial evidence for the importance of OAADPr in signaling and other cellular processes.
Although the function(s) of OAADPr have not been established, it has been proposed to act as a second messenger, a regulator of other enzymatic processes, or as a substrate for other enzymes (22,23). Our data are fully consistent with these possibilities. First, our finding of an esterase-like activity (or activities) located in the cytoplasmic compartment of cells suggests this activity may play an integral role in regulating the levels of OAADPr by catalyzing OAADPr to acetate and ADPr. Thus, in this instance OAADPr could initiate a signaling cascade as an important second messenger. Our identification of an OAADPr-specific acetyltransferase activity in the nuclear compartment of cells indicates that OAADPr can function as a co-enzyme substrate for other enzymatic reactions. The function of these specific transferases as well as the identity of the acceptor protein or small molecule remains to be uncovered. Our finding that select members of the ADPr hydrolase family of enzymes (for example mNudT5 and YSA1) are capable of efficient hydrolysis of OAADPr, whereas NUDT9 is essentially inactive toward OAADPr (as reflected in the k cat /K m values, Table I), further supports the notion that OAADPr is a potential second messenger as well as an in vivo substrate for other enzymes. Lack of activity between OAADPr and NUDT9 raises a number of intriguing questions pertaining to OAADPr function. Interestingly, NUDT9 has a mitochondrial leader sequence that gives rise to a mature 34.2-kDa mitochondrial protein (45). Recently, it has been reported that mitochondria from mammalian cells contain intrinsic NAD ϩ -dependent deacetylase activity (50). This activity was shown to be inhibited by nicotinamide but not by Trichostatin A, making it consistent with an Sir2 type of deacetylase activity. Furthermore, this deacetylase activity was identified as the nuclearencoded human Sir2 homolog hSirT3 (50). This finding raises interesting questions for the biological roles of OAADPr in the mitochondria and elsewhere. SIRT3-dependent generation of OAADPr in the mitochondria would be resistant to hydrolysis by NUDT9, implying an important role in maintaining OAADPr levels within the mitochondria. It has recently been proposed that OAADPr may be an important signaling molecule that mediates Sir2-like enzyme function on metabolic networks (51). Moreover, it has been speculated that the conserved family of Sir2 proteins are involved in sensing cellular energy and redox states (20). Supporting a link between Sir2-like enzymes and metabolism, Lin et al. (52) have demonstrated that yeast lifespan extension under limiting glucose requires Sir2 and increased respiration. These observations, combined with the results from this study, suggest that the Sir2 family of enzymes may control a variety of metabolic or signaling pathways through the formation of OAADPr. Identification of the enzymes and the reactions they catalyze will be a critical step in uncovering the functions of this unique metabolite.