Telomeric and rDNA Silencing in Saccharomyces cerevisiae Are Dependent on a Nuclear NAD⁺ Salvage Pathway

Corresponding author: Jeffrey S. Smith, Department of Biochemistry and Molecular Genetics, Jordan Hall, Box 800733, Charlottesville, VA 22908-0733., jss5y{at}virginia.edu (E-mail)

Communicating editor: L. PILLUS

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The Sir2 protein is an NAD⁺-dependent protein deacetylase that is required for silencing at the silent mating-type loci, telomeres, and the ribosomal DNA (rDNA). Mutations in the NAD⁺ salvage gene NPT1 weaken all three forms of silencing and also cause a reduction in the intracellular NAD⁺ level. We now show that mutation of a highly conserved histidine residue in Npt1p results in a silencing defect, indicating that Npt1p enzymatic activity is required for silencing. Deletion of another NAD⁺ salvage pathway gene called PNC1 caused a less severe silencing defect and did not significantly reduce the intracellular NAD⁺ concentration. However, silencing in the absence of PNC1 was completely dependent on the import of nicotinic acid from the growth medium. Deletion of a gene in the de novo NAD⁺ synthesis pathway BNA1 resulted in a significant rDNA silencing defect only on medium deficient in nicotinic acid, an NAD⁺ precursor. By immunofluorescence microscopy, Myc-tagged Bna1p was localized throughout the whole cell in an asynchronously growing population. In contrast, Myc-tagged Npt1p was highly concentrated in the nucleus in 40% of the cells, indicating that NAD⁺ salvage occurs in the nucleus in a significant fraction of cells. We propose a model in which two components of the NAD⁺ salvage pathway, Pnc1p and Npt1p, function together in recycling the nuclear nicotinamide generated by Sir2p deacetylase activity back into NAD⁺.

TRANSCRIPTIONAL silencing in the budding yeast Saccharomyces cerevisiae occurs within the silent mating-type loci (HML and HMR; for review see LOO and RINE 1995 ), telomeres (GOTTSCHLING et al. 1990 ), and the ribosomal DNA (rDNA) locus (BRYK et al. 1997 ; SMITH and BOEKE 1997 ). These loci are believed to form a heterochromatin-like structure that somehow prevents RNA polymerases and/or various recombination enzymes from gaining access to the associated DNA (GOTTSCHLING 1992 ; LOO and RINE 1994 ; FRITZE et al. 1997 ; SMITH and BOEKE 1997 ). Each form of silencing is fully dependent on the SIR2 gene, which encodes an NAD⁺-dependent histone/protein deacetylase with in vitro specificity for lysine 16 in the N-terminal tail of histone H4 and lysines 9 and 14 of histone H3 (IMAI et al. 2000 ; LANDRY et al. 2000B ; SMITH et al. 2000 ). Mutations in SIR2 that reduce histone deacetylation activity in vitro strongly weaken silencing in vivo (IMAI et al. 2000 ), suggesting that one of the major functions of Sir2p in silencing is histone deacetylation. This is supported by a study showing that SIR2 overexpression results in global histone hypoacetylation (BRAUNSTEIN et al. 1993 ). However, the in vivo target(s) of Sir2p-mediated deacetylation in yeast have not been identified.

Sir2p is the founding member of a highly conserved protein family with homologs identified in all three kingdoms of life (BRACHMANN et al. 1995 ; DERBYSHIRE et al. 1996 ). In S. cerevisiae there are five family members, including Sir2p. The others, Hst1p, Hst2p, Hst3p, and Hst4p, have been implicated in silencing, targeted gene repression (XIE et al. 1999 ), and maintenance of genomic stability (BRACHMANN et al. 1995 ). There are also at least seven homologs in human cells (BRACHMANN et al. 1995 ; AFSHAR and MURNANE 1999 ; FRYE 1999 ). It is improbable that all the Sir2-like proteins will be involved in regulating chromatin structure. For example, the CobB protein in Salmonella has been implicated in the synthesis of vitamin B12 (TSANG and ESCALANTE-SEMERENA 1998 ). However, for most homologs tested so far, including CobB, NAD⁺-dependent HDAC activity has been identified (IMAI et al. 2000 ; LANDRY et al. 2000B ; SMITH et al. 2000 ). It is therefore likely that nonhistone proteins are also in vivo targets for Sir2-like proteins, especially since an increasing number of nonhistone proteins are known to be acetylated (KOUZARIDES 2000 ; STERNER and BERGER 2000 ).

The deacetylase activity of Sir2p not only requires NAD⁺ for its activity, but also directly consumes NAD⁺ (LANDRY et al. 2000A ; TANNY and MOAZED 2001 ). For every acetyl group removed from a lysine residue, one molecule of NAD⁺ is hydrolyzed to form one molecule of nicotinamide (Nam; TANNY and MOAZED 2001 ). The acetyl group is transferred from the lysine residue to the ADP-ribose moiety of NAD⁺ to produce one molecule of the unusual compound, O-acetyl-ADP-ribose (TANNER et al. 2000 ; TANNY and MOAZED 2001 ). The reaction products are a mixture of 2' and 3' O-acetyl isomers (SAUVE et al. 2001 ). Sir2p and its family members therefore have the potential for consuming a large amount of NAD⁺.

The NAD⁺ dependence of Sir2p raises the interesting possibility that cellular processes regulated by Sir2p, such as silencing and aging, are influenced by changes in cellular NAD⁺ concentration. Indeed, mutations in the NPT1 gene reduce cellular NAD⁺ concentrations by approximately threefold and also cause a loss of rDNA and telomeric silencing (SMITH et al. 2000 ). NPT1 is also required for the extension of yeast cell life span when induced by caloric restriction (LIN et al. 2000 ). NAD⁺ is produced by at least two pathways in all organisms, a de novo pathway and one or more salvage pathways (PENFOUND and FOSTER 1996 ). Npt1p is highly related to the PncB nicotinic acid phosphoribosyltransferase (NAPRTase) of Salmonella typhimurium that is responsible for carrying out the last step of the Preiss-Handler NAD⁺ salvage pathway (LALO et al. 1993 ; RAJAVEL et al. 1998 ). In vitro, it catalyzes the formation of nicotinic acid mononucleotide (NaMN) from nicotinic acid (NA) and 5-phosphoribosyl-1-pyrophosphate (PRPP; RAJAVEL et al. 1998 ).

From previous studies it was unclear whether the silencing and aging phenotypes associated with npt1 mutations were directly caused by the reduction in NAD⁺ concentration or something more complex (LIN et al. 2000 ; SMITH et al. 2000 ). In this study, we therefore set out to more closely examine the role of Npt1p in silencing and to also test the relative contributions of the de novo and NAD⁺ salvage pathways in regulating silencing. In the process we discovered that the de novo NAD⁺ synthesis pathway was not essential for silencing, but that the salvage pathway in general was important. We also found that Npt1p is often highly concentrated within the nucleus, whereas the de novo NAD⁺ synthesis protein, Bna1p, always appears evenly distributed throughout the cell, suggesting that the NAD⁺ salvage pathway has an important nuclear function. Surprisingly, deletion of another salvage gene, PNC1, significantly weakened silencing without reducing the overall intracellular NAD⁺ concentration. However, silencing in the absence of PNC1 was completely dependent on nicotinic acid import from the growth medium. On the basis of these findings we present a model in which Pnc1p and Npt1p cooperate in the conversion of the nuclear nicotinamide generated by Sir2p back into NAD⁺.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Media and yeast strains:
Yeast media were as previously described (ROSE et al. 1990 ; COST and BOEKE 1996 ; SMITH and BOEKE 1997 ). All yeast growth was at 30°. SC medium contained 3.25 µM nicotinic acid (DIFCO 1998 ). When indicated, SC medium was supplemented with an additional 10 µM nicotinic acid. SC medium contained a limited concentration of adenine to allow development of a red colony color in the telomeric silencing assays.

NPT1, BNA1, and TNA1 were deleted from several strains using one-step PCR-mediated gene replacement (BAUDIN et al. 1993 ; LORENZ et al. 1995 ; SMITH et al. 1998 ). pRS400 (kanMX4) or pRS403 (HIS3) were used as templates for the PCR (BRACHMANN et al. 1998 ). DNA fragments for deleting QPT1 and PNC1 were produced by PCR amplification of kanMX4 from genomic DNA isolated from strains that were already qpt1::kanMX4 or pnc1::kanMX4 (WINZELER et al. 1999 ). Three hundred base pairs of flanking DNA were included on either side of kanMX4. The heterozygous diploid JS664 was constructed by mating the bna1::kanMX4 strain JS663 to the npt1-1 mutant JS646, which contains a Tn3::LacZ:: LEU2 transposon insertion in the C terminus of NPT1 (SMITH et al. 2000 ). The resulting diploid was sporulated and tetrads were dissected on YPD medium. Yeast strains and their genotypes are listed in Table 1.

Plasmid construction:
Genomic clones of NPT1 and BNA1 were constructed by amplifying the genes from a highly purified S. cerevisiae genomic DNA preparation (BACHMAN et al. 2001), using Vent DNA polymerase (New England Biolabs, Beverly, MA). To facilitate cloning, XhoI sites were incorporated into each end of the PCR fragment. The resulting PCR fragments were ligated into the XhoI site of pRS414 (CHRISTIANSON et al. 1992 ) to produce pJSS77 (NPT1) and pJOE22 (BNA1). The cloned NPT1 gene consisted of -400 to +250 bp surrounding the open reading frame (ORF). The cloned BNA1 gene consisted of -400 to +253 bp surrounding the ORF. To fuse the 5 x Myc tag in frame with NPT1 or BNA1, we used the Stratagene (La Jolla, CA) QuikChange site-directed mutagenesis method to initially introduce an NcoI site immediately downstream of the start codon, thus changing the second codon. For NPT1, serine 2 was changed to an alanine. For BNA1, phenylalanine 2 was changed to valine. The 5 x Myc tag was isolated as an NcoI-NcoI fragment from pAS90 (kindly provided by Doug Koshland, via Bob Skibbens) and then ligated into the engineered N-terminal NcoI sites of NPT1 and BNA1. The Myc-Npt1 vector was pJOE4, and the Myc-Bna1 vector was pJOE29. The 2µ plasmid versions were pJOE3 (2µ Myc-Npt1p) and pJOE28 (2µ Myc-Bna1p). The QuikChange method was also used to change His232 of Npt1p in pJSS77 to an asparagine residue in pJSS86. pTW1 was constructed by removing the NPT1 C terminus (including His232) from pJOE4 by excising a NotI-NsiI fragment and replacing it with the equivalent NotI-NsiI fragment from pJSS86. Plasmid descriptions are in Table 2.

Multiple alignments:
The entire open reading frames of Npt1-like proteins from Drosophila melanogaster (19%), Caenorhabditis elegans (15%), Arabidopsis thaliana (19%), Homo sapiens (18%), Escherichia coli (25%), S. typhimurium (25%), Borrelia burgdorferi (17%), S. cerevisiae, and Archaeoglobus fulgidus (9%) were aligned using ClustalW software and presented using Boxshade. The numbers in parentheses are percentage identity to the S. cerevisiae Npt1 protein. To identify the gene encoding for the human version of Npt1p, a BLAST search was carried out between the Drosophila protein sequence and the human expressed sequence tag (EST) database. Multiple ESTs were identified. The GenBank accession numbers were AI871064, AW081501, AA612667, AA069101, AA100897, and AI925574. A putative human ORF (with some gaps) was created and then compared to the working draft of the human genome using BLAST (LANDER et al. 2001 ). The gene is present on chromosome 8 at nucleotide positions 146382485 to 146385926. The gap-free protein sequence provided is from I.M.A.G.E. clone 3957135.

Silencing assays:
Two different rDNA silencing assays were used (SMITH et al. 1998 ). To assay for silencing of mURA3 in the rDNA, strains were grown overnight on YPD or SC medium as patches. These patches were then replica plated to SC, SC-Ura, or modified lead acetate (MLA) plates. The rDNA silencing color assay for expression levels of the MET15 gene was performed by streaking freshly grown cells onto Pb²⁺-containing MLA medium (COST and BOEKE 1996 ) and allowing the plates to grow for 5 days at 30° before photos were taken under a Leica stereoscopic microscope.

Two different telomeric silencing (TPE) assays were used. To measure silencing of URA3 adjacent to the left telomere of chromosome VII, cells were patched on YPD or selective SC media and grown for 1 day. The cells were then scraped off the plates and resuspended in sterile water. The cell mixtures were normalized to an A₆₀₀ of 1.0 and then serially diluted in fivefold increments. Five microliters of each dilution was spotted onto SC medium, SC medium lacking uracil (SC-Ura), or SC with 1% of 5-fluoroorotic acid (5-FOA) added. Plates were incubated at 30° for 2 days before photography. Loss of silencing activates expression of URA3, preventing growth on 5-FOA. The colony color TPE assay is based on the ADE2 reporter integrated next to the right telomere of chromosome V. Reporter strains are plated onto SC medium, which contains a limiting concentration of adenine. The cells turn red if they silence ADE2, but turn white if they do not silence ADE2. Switching between on and off states results in the characteristic sectored colonies (GOTTSCHLING et al. 1990 ).

NAD⁺ concentration measurements:
NAD⁺ measurements were performed as previously described (SMITH et al. 2000 ). Briefly, 500-ml yeast cultures were grown to an A₆₀₀ of 1.0 and then harvested by centrifugation. Cell pellets were extracted for 30 min with 5 ml of ice-cold 1 M formic acid (saturated with butanol). A total of 1.25 ml of 100% trichloroacetic acid (TCA) was added and incubated on ice for 15 min. The mixture was centrifuged at 4000 x g for 20 min, and the acid soluble supernatant (containing the NAD⁺) was saved. The pellet was washed with 2.5 ml of 20% TCA and pelleted again. The supernatants were combined and used for the NAD⁺ measurement. Acid extract (150 µl) was added to a reaction buffer (1 ml final volume) containing 300 mM Tris-HCl, pH 9.7, 200 mM lysine-HCl, 0.2% ethanol, 150 µg/ml alcohol dehydrogenase (Sigma, St. Louis). Reactions were incubated at 30° for 20 min. The absorbance was then measured at 340 nm and compared to a standard curve.

Indirect immunofluorescence microscopy:
A total of 10 ml of exponentially growing yeast cells (A₆₀₀ of 0.5 to 1.0) was fixed for 90 min in 3.7% formaldehyde. Cells were recovered by centrifugation and washed three times in 1 ml SK buffer (1 M sorbitol, 50 mM KPO₄ pH 7.5) with a final resuspension in 900 µl SK buffer. Spheroplasts were prepared by adding 100 µl 10x Zymolyase cocktail (0.1% ß-mercaptoethanol, 0.34 mg/ml Zymolyase-20) and incubating cells at 37° for 25 min. Cells were recovered by microcentrifugation, washed twice with 1 ml SK buffer, and resuspended in 250 µl SK buffer. Fixed spheroplasts were dropped onto 0.1% polylysine-treated 10-well glass slides and incubated in a humid chamber for 10 min. Slides were washed with SK buffer and then immersed in methanol (-20°) and acetone (-20°). Blocking solution (1x PBS, 0.1% BSA, 0.1% NaN₃, 10 mM MgSO₄, 10 mM CaCl₂) was placed in each well and incubated at room temperature for 15 min. The blocking solution was removed and the primary antibody (Roche -c-myc, monoclonal 9E10) was added at a 1:1000 dilution and incubated in a humid chamber for 2 hr. SK buffer was used to wash each well eight times. Secondary antibody (Cy3-conjugated goat -mouse IgG) was then added at a dilution of 1:1000 and slides were incubated for another 2 hr in a humid chamber. SK buffer was used to wash each well eight times. Mounting medium was Vectashield w/DAPI (H-1200; Vector Laboratories, Burlingame, CA). The slides were viewed and photographs were taken on a Nikon Eclipse E600 using the 100x objective.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Two NAD⁺ synthesis pathways in yeast:
To more clearly understand the role of NAD⁺ in the regulation of silencing, we first wanted to test the validity of the current models for synthesizing NAD⁺ in yeast. Eukaryotic NAD⁺ synthesis is generally divided into two separate pathways, salvage and de novo (Fig 1A). The salvage pathway begins with the breakdown of NAD⁺ into Nam and ADP-ribose. In the case of Sir2p-mediated deacetylase activity, the ADP-ribose byproduct is actually a mixture of 2' and 3' O-acetyl-ADP-ribose (SAUVE et al. 2001 ). The nicotinamide is then deamidated by a nicotinamide deamidase to form NA and ammonia. The nicotinamide deamidase in E. coli is PncA (FROTHINGHAM et al. 1996 ). Through a BLAST search we identified YGL037C as the equivalent S. cerevisiae gene. This gene was recently assigned the name PNC1 by the Saccharomyces Genome Database. Npt1p then converts NA to NaMN (RAJAVEL et al. 1998 ). At this point the salvage pathway merges with the de novo pathway to convert NaMN to NAD⁺ through the sequential action of nicotinamide mononucleotide adenylyltransferases (YLR328W and YGR010W) and NAD⁺ synthetase (YHR074W; EMANUELLI et al. 1999 ). Nicotinic acid can also be imported into the cell through the action of the high affinity nicotinic acid plasma membrane permease encoded by TNA1 (YGR260W; LLORENTE and DUJON 2000 ) and then presumably converted to NaMN by Npt1p (Fig 1A).

View larger version (53K): In this window In a new window Download PPT slide   Figure 1. Genetic analysis of NAD+ synthesis in S. cerevisiae. (A) Schematic diagram outlining the de novo and salvage pathways for NAD+ synthesis. The de novo pathway starts with tryptophan (Trp) and proceeds through four steps to 3-hydroxyanthranilic acid (3HA). Bna1p converts 3HA to 2-amino-3-carboxymuconic semialdehyde (2AC), and Qpt1p converts quinolinic acid (Qa) to nicotinic acid mononucleotide (NaMN). In the salvage pathway, the Nam produced by Sir2p is converted to NaMN by Pnc1p and Npt1p. Npt1p can also utilize nicotinic acid (Na) imported into the cell by the nicotinic acid permease (Tna1p). PRPP, 5-phosphoribosyl-1-pyrophosphate; NaAD, deamido NAD. (B) Nicotinic acid auxotrophy of bna1 and qpt1 mutants. WT (BY4741), bna1 (SY8), qpt1 (SY15), pnc1 (SY10), and npt1 (SY16) strains were streaked for single colonies on minimal medium either missing nicotinic acid (–Na) or containing 3.25 µM nicotinic acid (+Na). Colonies were grown for 3 days. (C) Synthetic lethality between bna1 and npt1-1 mutations. Eleven tetrads of diploid JS664 were dissected into four individual spores (A, B, C, and D) on YPD medium. Phenotypes of the viable spores were consistent with synthetic lethality.

The de novo pathway in yeast begins with tryptophan, which through multiple enzymatic steps is converted to NaMN. In yeast, two of the de novo pathway genes are BNA1 (YJR025C) and QPT1 (YFR047C; KUCHARCZYK et al. 1998 ; IMAI et al. 2000 ). Mutants for bna1, qpt1, pnc1, and npt1 were all viable on rich YPD and SC media, indicating that they all have the capacity to make NAD⁺ if provided with the proper nutrients. Strains that are deleted for BNA1 are known to be nicotinic acid auxotrophs that are unable to grow on medium lacking nicotinic acid (KUCHARCZYK et al. 1998 ). Based on the proposed metabolic scheme in Fig 1A, only de novo pathway mutants should be nicotinic acid auxotrophs because they need nicotinic acid to produce NAD⁺ through the salvage pathway. As predicted, the bna1 and qpt1 mutants were nicotinic acid auxotrophs, but the npt1 and pnc1 mutants were not (Fig 1B). We tested deletions of two other predicted de novo pathway genes, kynurenine 3-hydroxylase (YBL098W) and kynureninase (YLR231C), and as expected, these were also nicotinic acid auxotrophs (data not shown).

The metabolic scheme in Fig 1A also predicted that combining a de novo pathway gene mutation with a salvage pathway gene mutation would be lethal. To test this hypothesis, we crossed a bna1::kanMX4 strain with a strain carrying the npt1-1 mutation (tagged with the Tn3::lacZ::LEU2 insertion) and dissected tetrads from the resulting diploid strain. We observed 25% spore inviability, as predicted for two mutations lethal in combination (Fig 1C). The genotype of 16/18 dead spores was inferred to be npt1-1 bna1::kanMX4. The double mutant spores arrested as four- to six-cell microcolonies. Moreover, out of 21 tetrads dissected, we never observed a single viable G418-resistant (kanMX-containing) Leu⁺ spore clone. Synthetic lethality was also observed when the bna1::kanMX4 allele was combined with an npt1::kanMX4 allele (data not shown). Consistent with our findings, npt1 qpt1 double mutants were previously shown to have a severe slow growth phenotype (LIN et al. 2000 ). It is currently unclear why the qpt1 npt1 mutant combination was not completely lethal, but there could potentially be an unknown reaction that can weakly carry out the conversion of quinolinic acid to NaMN. We conclude that NPT1 and BNA1 represent genes in separate pathways required for NAD⁺ synthesis, confirming our predicted view of NAD⁺ synthesis in yeast and providing a framework for the following silencing experiments.

The NAD⁺ salvage pathway is important for rDNA and telomeric silencing:
Mutations in NPT1 were already known to cause a loss of silencing at the rDNA and telomeres (SMITH et al. 2000 ). However, we did not know whether this was a general defect of the NAD⁺ salvage pathway or an npt1-specific phenotype. We also wanted to know whether the de novo NAD⁺ synthesis pathway had any role in silencing. We therefore deleted NPT1, PNC1, BNA1, or QPT1 from the rDNA silencing reporter strain JS306, which contained a MET15 colony color reporter integrated into the rDNA nontranscribed spacer (SMITH et al. 1999 ). MET15 is repressed by the rDNA, resulting in a tan colony color on rich medium containing Pb²⁺ (SMITH and BOEKE 1997 ). As expected (SMITH et al. 2000 ), deletion of NPT1 caused a loss of silencing phenotype as indicated by a white colony color background and multiple dark brown sectors (Fig 2A). The pnc1 mutant had a similar loss of silencing phenotype, although it was less severe in that very few sectors were observed (Fig 2A). The entire NAD⁺ salvage pathway is therefore important for rDNA silencing, not just NPT1. In contrast, the bna1 and qpt1 mutants produced a colony color similar to the wild-type (WT) strain, indicating that the de novo NAD⁺ synthesis pathway is not required for rDNA silencing (Fig 2A). Since npt1 mutants also dramatically weaken telomeric silencing, we tested whether the pnc1, bna1, or qpt1 deletions affected TPE. As shown in Fig 2B, the pnc1 deletion derepressed a telomeric URA3 reporter gene by 25-fold compared to WT, which was not as dramatic as the npt1 defect, but was similar to the intermediate defect observed with rDNA silencing. Deletions of BNA1 or QPT1 had no effect on TPE (Fig 2B).

View larger version (63K): In this window In a new window Download PPT slide   Figure 2. Silencing phenotypes of de novo and salvage pathway mutants. WT (JS306), npt1 (JS673), pnc1 (JS804), bna1 (JS663), and qpt1 (JS805) strains were tested for rDNA silencing, telomeric silencing, and cellular NAD+ concentration. (A) rDNA silencing phenotypes using a colony color assay. Lighter colony colors indicate a loss of rDNA silencing. (B) Telomeric silencing phenotypes using a telomeric URA3 reporter gene. Loss of silencing is indicated by less growth on 5-FOA. (C) Relative intracellular NAD+ concentrations. Values on the y-axis are the mean absorbance at 340 nm. Error bars indicate standard deviation from three independent experiments.

The above results were surprising because elimination of the de novo NAD⁺ synthesis pathway was expected to reduce intracellular NAD⁺ levels and thus weaken silencing. We therefore measured NAD⁺ in all four mutants grown in YPD medium. As expected, the npt1 mutant lowered the intracellular NAD⁺ concentration by approximately threefold compared to the WT strain (Fig 2C). NAD⁺ levels in the bna1 and qpt1 mutants were unchanged compared to the WT parent (Fig 2C), which was consistent with the lack of a silencing defect. Given its defects in rDNA and telomeric silencing, we expected the pnc1 mutant to have a reduced intracellular NAD⁺ concentration. Surprisingly, we found that the mutant actually had a near-normal NAD⁺ concentration (Fig 2C). The phenotypic differences between npt1 and pnc1 mutants are addressed further below.

Regulation of rDNA silencing by nicotinic acid concentration:
As described above, de novo pathway mutants were unable to grow in medium lacking nicotinic acid. The activity of Npt1p, which converts nicotinic acid to NaMN, allows these mutants to make NAD⁺ from an exogenous source of nicotinic acid imported by Tna1p. We hypothesized that if nicotinic acid levels became limiting in a bna1 mutant, Npt1p activity would become impaired due to reduced substrate concentration, possibly resulting in an rDNA silencing defect. To test this possibility, we first pregrew patches of WT (JS306) and bna1 (JS663) strains on either YPD, SC, or SC with excess nicotinic acid added (SC + NA). These patches were then replica plated to SC, SC-Uracil, or Pb²⁺ (MLA) media (Fig 3A). Compared to WT, silencing of the mURA3 and MET15 rDNA silencing reporters in the bna1 mutant was reduced when the strains were pregrown on SC medium (Fig 3A; more Ura⁺ growth and lighter-colored patch, respectively). As expected, no difference was observed when the strains were pregrown on YPD medium prior to replica plating. This result suggested that there was a limiting nutrient in SC medium that was required for rDNA silencing. We surmised that this nutrient was nicotinic acid, which is present at 3.25 µM in normal SC medium. When pregrown on SC medium supplemented with an additional 10 µM nicotinic acid (13.25 µM final concentration), silencing was restored to the replica plated bna1 mutant (Fig 3A; note reduced Ura⁺ growth and a darker-colored patch).

View larger version (61K): In this window In a new window Download PPT slide   Figure 3. rDNA silencing in a bna1 mutant is altered by variations in nicotinic acid concentration. (A) Silencing in JS306 (WT) and JS663 (bna1) strains was measured qualitatively using a standard yeast patch assay. The media on which the strains were pregrown is indicated on the left side of the panel. After replica plating to the appropriate indicator medium (listed at the top of the panel), the patches were incubated for 1 or 2 days as noted. More Ura+ growth and a lighter colony color on MLA medium indicate weaker rDNA silencing by the bna1 mutant. (B) Relative intracellular NAD+ levels. NAD+ concentration was measured from strains JS306 (WT), JS663 (bna1), and JS673 (npt1). Each growth condition is indicated by a different shaded bar. The values are the average absorbances (340 nm). Error bars are the standard deviation from three experiments. SC, synthetic complete (3.25 µM nicotinic acid); SC + NA, synthetic complete supplemented with an additional 10 µM nicotinic acid.

We next tested whether the silencing defect of a bna1 mutant on SC medium correlated with a reduction in NAD⁺ concentration. Interestingly, the wild-type cells (JS306) grown on SC medium contained 33% less NAD⁺ compared to YPD (Fig 3B). There was also an additional 25% reduction in NAD⁺ concentration in the bna1 mutant (JS663) compared to WT when grown on SC (Fig 3B), consistent with the loss of silencing observed in Fig 3A. When SC was supplemented with excess nicotinic acid, the NAD⁺ concentration in the WT and bna1 strains increased to a level comparable to growth on YPD (Fig 3B). For the npt1 mutant (JS673), the NAD⁺ concentration was low for all growth conditions. These results indicate that in the absence of the de novo NAD⁺ synthesis pathway, rDNA silencing is highly sensitive to changes in environmental nicotinic acid concentrations.

SIR2 overexpression can restore rDNA silencing to an npt1 mutant:
Sir2p is a limiting component of rDNA silencing, and its overexpression improves silencing (FRITZE et al. 1997 ; SMITH et al. 1998 ). We therefore hypothesized that SIR2 overexpression would not strengthen rDNA silencing in an npt1 mutant due to the suboptimal intracellular NAD⁺ concentration. WT and npt1 strains containing a high-copy SIR2 vector were tested for increased rDNA silencing strength using a patch-replica-plating assay (Fig 4A). In this assay, increased silencing is measured by reduced Ura⁺ growth. As expected from previous studies (SMITH et al. 1998 ), the SIR2 high-copy vector strengthened silencing in the WT strain (Fig 4A, day 4). Surprisingly, high-copy SIR2 almost completely restored rDNA silencing to the npt1 strain (Fig 4A). SIR2 overexpression also restored rDNA silencing to a bna1 mutant under limiting nicotinic acid conditions (data not shown). How this happens is currently unclear, but it suggests that Sir2p could also have an additional NAD⁺-independent function in rDNA silencing, perhaps a structural role. On the other hand, overexpression of SIR2 could lower the threshold concentration of NAD⁺ required to effect silencing. In contrast, NPT1 was required for the strengthening of telomeric silencing caused by SIR3 overexpression (J. SMITH, unpublished data), suggesting that TPE is more sensitive than rDNA silencing to changes in NAD⁺ concentration.

View larger version (66K): In this window In a new window Download PPT slide   Figure 4. SIR2 overexpression restores rDNA silencing to an npt1 mutant. (A) The high-copy vector pRS425 or the high-copy SIR2 plasmid pSIR 2µ were transformed into JS306 (WT) and JS673 (npt1::kanMX4). Two independent transformants were patched onto SC-Leu media, grown for 1 day, and then replica plated to SC-Leu and SC-Leu-Ura media. Photos of the patches were taken after the indicated number of days. (B) Relative NAD+ levels in sir2 and SIR2 overexpressing strains grown in SC-Leu medium. The strains were JS740 (SIR2+), JS860 (sir2), and JS742 (SIR2 2µ). The values reported are the mean absorbances (340 nm). Error bars are the standard deviations.

Since Sir2p consumes NAD⁺ as part of its deacetylation reaction, it was possible that changes in SIR2 dosage (such as in Fig 4A) could alter the intracellular NAD⁺ concentration. To test this possibility, we measured NAD⁺ levels in strains with varying SIR2 copy number (Fig 4B). The NAD⁺ concentration in strains that were deleted for SIR2 or contained a high-copy SIR2 plasmid was actually similar to the concentration from a normal SIR2⁺ strain (Fig 4B). Therefore, mutations in SIR2 do not significantly affect the overall intracellular NAD⁺ concentration. This lack of a change could be partially due to compensatory changes in activity of the de novo and/or salvage NAD⁺ synthesis pathways.

Npt1p enzymatic activity is required for silencing:
NAPRTase activities have been described from bacteria, yeast, and humans. However, genes encoding these proteins have been described only for S. typhimurium, S. cerevisiae, and Mycobacterium tuberculosis (RAJAVEL et al. 1998 ). The PncB NAPRTase of Salmonella is known to utilize an unusual ATP-mediated energy-coupling mechanism to achieve a highly specific and efficient reaction. ATP increases the V_max by >10-fold and decreases K_m values by 200-fold (RAJAVEL et al. 1998 ). A phosphate group from ATP is covalently transferred to His219 during each catalytic cycle. It has been proposed that this is a regulatory mechanism to increase flux toward NAD⁺ when cellular ATP levels are high or to render NaMN formation irreversible. A H219N mutation eliminates this energy coupling, but the basal kinetic properties of the enzyme in the absence of ATP are unchanged (RAJAVEL et al. 1998 ). Purified yeast Npt1p also displays similar ATP-modulated kinetics (RAJAVEL et al. 1998 ). In S. cerevisiae, the equivalent histidine residue is predicted to be His232 (RAJAVEL et al. 1998 ).

We wanted to know whether a high level of enzymatic activity by yeast Npt1p was required for silencing. The catalytic residues of NAPRTases have not been identified, but activity is severely impaired by knocking out the energy-coupling mechanism. BLAST searches for NPT1-like genes were conducted on the available public databases to determine whether the phosphorylated histidine residue of PncB was conserved across all species. NPT1-related genes were identified from all kingdoms. A multiple alignment of several proteins, including the putative human version, is shown in Fig 5. S. cerevisiae Npt1p was most closely related to the E. coli and Salmonella PncB proteins (25% amino acid identity). For all family members, the conservation extended throughout the entire protein length. Importantly, one absolutely conserved block of homology included His232 of S. cerevisiae, suggesting that the energy-coupling mechanism is universal in NaPRTases.

View larger version (116K): In this window In a new window Download PPT slide   Figure 5. Multiple alignment of proteins related to S. cerevisiae Npt1p. The species and amino acid numbering are indicated to the left. Black boxes represent amino acid identity in >50% of the species. Shaded boxes represent amino acid similarity in >50% of the species. The conserved histidine involved in energy coupling is indicated by a solid circle (His232 for S. cerevisiae).

To determine whether Npt1 enzymatic activity was critical for silencing in S. cerevisiae, we mutated the equivalent Histidine 232 residue in Npt1p to an asparagine. This mutant H232N npt1 gene contained on a CEN plasmid was introduced into an npt1 TPE reporter strain. Silencing of a telomeric ADE2 gene in this reporter strain results in a red colony color. We previously showed that deletion of NPT1 caused this strain to have a white (derepressed) colony color (SMITH et al. 2000 ). Wild-type NPT1 restored silencing but the H232N mutant (npt1-3) did not (Fig 6A). Western blotting using Myc-tagged versions of these NPT1 alleles verified that the H232N (npt1-3) mutation did not affect steady-state protein levels (Fig 6B), compared to WT Myc-tagged Npt1p. These results indicate that a high level of Npt1p enzymatic activity is required for efficient silencing.

View larger version (64K): In this window In a new window Download PPT slide   Figure 6. Npt1p catalytic activity is required for silencing. (A) An H232N mutant of Npt1p (npt1-3) is defective for telomeric silencing. An npt1 mutant was transformed with CEN plasmids containing WT, NPT1, or npt1-3. (B) Western blot analysis of steady-state Npt1 protein levels. Myc-tagged Npt1 protein was detected from strain JJSY2 (lane 1). A Myc-tagged version of Npt1-3p was detected from five isolates of TWY1 (lanes 2–6). Parallel Millipore-P membranes were probed with -Myc antibody or -tubulin antibody.

Npt1p is highly concentrated in the nucleus:
Sir2p and its homologs are histone/protein deacetylases that consume one NAD⁺ molecule for every lysine residue that they deacetylate (LANDRY et al. 2000A ; TANNY and MOAZED 2001 ), suggesting a large demand for NAD⁺ in the nucleus. Furthermore, this NAD⁺ hydrolysis generates nicotinamide as one of the by-products (LANDRY et al. 2000A ; TANNY and MOAZED 2001 ). We therefore wanted to determine whether the NAD⁺ salvage pathway, which converts nicotinamide to NaMN, was compartmentalized in the nucleus. To test this possibility we analyzed yeast cells expressing 5 x Myc-tagged Npt1p or 5 x Myc-tagged Bna1p by indirect immunofluorescence microscopy using a Myc-specific primary antibody.

When the Myc-Npt1 or Myc-Bna1 constructs were integrated into the genome, the tagged proteins were undetectable by microscopy, but detectable by Western blotting (data not shown). However, when expressed from a CEN or 2µ vector, Myc-Npt1p was detectable and indeed concentrated in the nucleus in 40% of the cells in which Myc-Npt1p was visible (Fig 7). In cells that did not have a nuclear concentration, Myc-Npt1p displayed a whole-cell localization pattern (Fig 7). There was no obvious indication of any subnuclear localization. In contrast, Myc-Bna1p expressed from a CEN or 2µ vector had a whole-cell localization pattern in close to 100% of the cells (Fig 7). Strong localization of Bna1p in the nucleus was never observed. For both Npt1p and Bna1p, the CEN and 2µ vectors produced similar localization patterns, but the staining intensity was much lower from the CEN vectors because of their lower copy number (data not shown). For clarity, the 2µ expression data are presented in Fig 7. These results strongly suggest that a large proportion of NAD⁺ salvage in the cell takes place in the nucleus, which is consistent with the role it plays in modulating transcriptional silencing. The dual localization of Npt1p also suggests that Npt1p may shuttle between the nucleus and cytoplasm in a cell-cycle-regulated fashion.

View larger version (20K): In this window In a new window Download PPT slide   Figure 7. Differential cellular localization of Npt1p and Bna1p. 5 x Myc-tagged Npt1p or 5 x Myc-tagged Bna1p were expressed from their own promoters off 2µ vectors in npt1 or bna1 strain backgrounds, respectively. Localization was observed using indirect immunofluorescence microscopy. Cy3 signal indicates the Myc-tagged protein (red) and DAPI staining indicates the location of the nucleus (green). In the merged image, yellow indicates co-localization. The Bna1p- specific strain was JJSy54 and the Npt1p-specific strain was JJSy51.

Effects of nicotinic acid import on silencing:
As described above, Npt1p is often concentrated in the nucleus where it must convert nicotinic acid to NaMN. Nicotinic acid can be produced by Pnc1p-mediated deamidation of nicotinamide or Tna1p (the nicotinic acid plasma membrane permease) can import it into the cell. As shown in Fig 2, the NAD⁺ concentration in a pnc1 mutant was normal, but there was a partial silencing defect. We hypothesized that silencing was only partially defective in the pnc1 mutant because Npt1p could still utilize the environmentally imported nicotinic acid to produce NaMN (see Fig 1A). To test this hypothesis, we eliminated nicotinic acid import by deleting TNA1 and then assayed for silencing of the telomeric URA3 reporter gene. Deletion of TNA1 caused a very slight silencing defect indicated by smaller colonies on FOA, but not a reduced colony number compared to the WT strain (Fig 8). However, silencing was almost completely eliminated when the pnc1 and tna1 mutations were combined (Fig 8; very little growth on the FOA plate). Interestingly, the silencing defect of the pnc1 tna1 mutant was even more dramatic than the defect caused by the npt1 mutation. Therefore, the import of nicotinic acid from the growth medium is required for telomeric silencing only when the NAD⁺ salvage pathway is disrupted by deletion of PNC1.

View larger version (35K): In this window In a new window Download PPT slide   Figure 8. The effect of nicotinic acid import on telomeric silencing. Loss of silencing is indicated by less colony growth on 5-FOA medium compared to growth on SC (complete) medium. Fivefold serial dilutions of cells were plated onto each plate. The strains were WT (YCB647), npt1 (JS641), pnc1 (JS807), tna1 (JS813), and pnc1 tna1 (JS861).

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Sir2p as an NAD⁺ consumer in yeast:
NAD⁺ turnover is rapid in both bacteria and eukaryotic cells, with a half-life ranging from 25 to 90 min (MANSER et al. 1980 ; PENFOUND and FOSTER 1996 ). However, the mechanisms responsible for this high rate of NAD⁺ consumption are not well understood. For example, although E. coli DNA ligase consumes NAD⁺, substitution of the ligase gene with an ATP-dependent form has no effect on NAD⁺ flux (PARK et al. 1989 ). In mammalian cells, poly(ADP) ribosylation reactions catalyzed by poly(ADP) ribose polymerase (PARP) can account for a large amount of NAD⁺ consumption in the nucleus when cells undergo DNA damage (BERGER 1985 ). However, neither PARP activity nor a PARP-related gene has been identified in yeast. It is now known that Sir2p and its family members are NAD⁺-dependent protein deacetylases that consume NAD⁺ as part of their reaction mechanism (TANNER et al. 2000 ; TANNY and MOAZED 2001 ). Sir2p also has a weak NAD⁺-dependent phosphoryl transferase activity that could potentially contribute to NAD⁺ consumption (TANNY et al. 1999 ). In this study we have shown that deletion or overexpression of SIR2 does not measurably change the intracellular NAD⁺ concentration. There are four other Sir2 family members in yeast, and Hst2p actually contributes most of the NAD⁺-dependent deacetylase activity from whole-cell extracts (SMITH et al. 2000 ). It may take changes in some or all of these genes to observe a change in NAD⁺ concentration. It is still very possible that the Sir2p family members in other species also significantly contribute to NAD⁺ turnover.

Nuclear localization of Npt1p:
It is unclear how and from where Sir2p acquires its NAD⁺ substrate/cofactor. Presumably, the histone deacetylase activity of Sir2p acts in the nucleus in vivo, although this has not yet been directly demonstrated. Since Sir2p generates one nicotinamide molecule for each acetyl group removed (LANDRY et al. 2000A ), it would be very efficient for the cell to recycle this nicotinamide by-product back into NAD⁺ near the sites of silencing. This would ensure a constant supply of NAD⁺ for Sir2p in the nucleus. Indeed, we have shown that a Myc-tagged Npt1 protein is concentrated in the nucleus in a large percentage of cells.

Finding Npt1p in the nucleus raises the possibility that NAD⁺ synthesis and NAD⁺ salvage may be compartmentalized. We observed that a Myc-tagged Bna1 protein, which is part of the de novo NAD⁺ synthesis pathway, was dispersed evenly throughout the cell. This suggests that a large proportion of the de novo synthesis pathway operates outside the nucleus, while in many cells the salvage pathway operates primarily inside the nucleus. After NaMN synthesis, the two pathways converge with the action of nicotinamide mononucleotide (NMN) adenylyltransferase, which catalyzes the essential formation of NAD⁺ or deamido-NAD. Interestingly, this enzymatic activity is nuclear in vertebrate cells (HOGEBOOM and SCHNEIDER 1952 ). Furthermore, the yeast, chicken, and human enzymes have been shown to associate with the nuclear matrix and/or chromatin (CANTAROW and STOLLAR 1977 ; BALDUCCI et al. 1992 ; MAGNI et al. 1997 ). The yeast enzyme is encoded by the YLR328W open reading frame (Fig 1A; EMANUELLI et al. 1999 ). Yeast also encodes a highly related, but uncharacterized gene called YGR010W, with 72% identity to YLR328W (EMANUELLI et al. 1999 ). Since ylr328w mutants are viable (WINZELER et al. 1999 ), the Ygr010w protein likely provides a redundant function. This class of enzyme was named as an NMN adenylyltransferase, but the bacterial and mammalian forms actually prefer NaMN as the substrate to produce deamido-NAD (MAGNI et al. 1997 ), the step immediately downstream of the Npt1-catalyzed reaction. Some older studies using enucleated human cells have reported that NAD⁺ synthesis occurs almost exclusively in the nucleus (RECHSTEINER and CATANZARITE 1974 ).

If yeast NMN adenylyltransferase is located exclusively in the nucleus, then this suggests that one of the de novo NAD⁺ synthesis enzymes makes an NAD⁺ precursor molecule that can easily diffuse or be transported into the nucleus. Alternatively, one of the de novo synthesis enzymes could shuttle between the nucleus and the cytoplasm. It will be interesting to determine if and how such a cytoplasmic/nuclear transition occurs. Nuclear localization of Npt1p in 40% of asynchronously growing cells also raises the possibility that this process is cell cycle regulated. High levels of NAD⁺ salvage in the nucleus may be required at certain phases of the cell cycle to help Sir2p establish and/or maintain silencing. At times when Npt1p is not concentrated in the nucleus, silent chromatin could become more susceptible to chromatin remodeling. For example, telomeric chromatin becomes accessible to transcriptional activators during the G₂/M phases of the cell cycle (APARICIO and GOTTSCHLING 1994 ), and Sir2p is released from the rDNA (nucleolus) at the end of mitosis (STRAIGHT et al. 1999 ). Perhaps nuclear localization of Npt1p is also lost during this time. Future work will address this possibility.

Nicotinic acid and rDNA silencing:
Why does rDNA silencing in a bna1 mutant become sensitive to nicotinic acid concentration in the growth media? In a bna1 mutant, NAD⁺ synthesis is completely dependent on the salvage pathway and an exogenous source of nicotinic acid. The cell internalizes nicotinic acid through action of the nicotinic acid transporter Tna1p (LLORENTE and DUJON 2000 ). The limiting concentration of nicotinic acid in the medium could drop the intracellular concentration below a critical value needed for efficient catalysis by Npt1p. For Salmonella PncB, the K_m for nicotinic acid is 1.5 µM (VINITSKY and GRUBMEYER 1993 ). The concentration of nicotinic acid in SC medium is 3.25 µM, which is approaching the K_m. Subsequent inhibition of Npt1p activity would reduce the cellular NAD⁺ concentration, as we observed in Fig 3B, and result in weakened silencing. Alternatively, changes in the nicotinic acid concentration could also influence the expression of genes involved in rDNA silencing. Deletion of SIR2 is known to elevate the expression of Bna1p (BERNSTEIN et al. 2000 ), and the expression of the nicotinic acid transporter Tna1p is downregulated by increasing nicotinic acid concentrations (LLORENTE and DUJON 2000 ). It is possible that other silencing-related genes could be affected.

What is the function of the NAD⁺ salvage pathway in silencing?
Npt1p and Pnc1p are both part of the NAD⁺ salvage pathway. However, pnc1 and npt1 mutants do not have identical silencing phenotypes or the same effects on intracellular NAD⁺ concentration. As shown in Fig 2, the pnc1 mutant has partial rDNA and telomeric silencing defects, but has a normal intracellular NAD⁺ concentration. The npt1 mutant has a more severe silencing defect at the rDNA and telomeres that correlates well with the reduced intracellular NAD⁺ concentration. This difference could be caused by differential usage of nicotinic acid pools (see below).

The role of nicotinamide deamidase (Pnc1p for S. cerevisiae) in the NAD⁺ salvage pathway is to produce nicotinic acid. As a result, some nicotinic acid is derived intracellularly from NAD⁺ hydrolysis and the rest is imported from the growth medium (environmental). Since Npt1p is responsible for converting both sources of nicotinic acid to NaMN, loss of Npt1p function causes a large reduction in NAD⁺ concentration and severe silencing defects. In a pnc1 mutant, nicotinic acid is no longer produced intracellularly, but Npt1p can still convert the imported (environmental) nicotinic acid to NaMN. The result is no significant net change in the overall intracellular NAD⁺ concentration. However, there is still a modest silencing defect. One possibility to explain this phenomenon is that the nicotinic acid generated by Sir2p and Pnc1p in the nucleus has a greater probability of conversion to Sir2p-utilized NAD⁺ than does imported nicotinic acid. In support of this hypothesis, deletion of TNA1 causes only a slight derepression of telomeric silencing (Fig 8), even though nicotinic acid import is eliminated. An alternative model is that Pnc1p and Npt1p cooperate to convert the nicotinamide specifically produced by Sir2p to NaMN. Nicotinamide is a noncompetitive inhibitor of the NAD⁺- dependent histone deacetylase activity (LANDRY et al. 2000A ). Therefore conversion of Sir2p-produced nicotinamide to nicotinic acid and NaMN could help facilitate the deacetylation activity of Sir2p by reducing product inhibition. Future work will be aimed at testing these models, which are not mutually exclusive. In both cases, Npt1p would be expected to be concentrated in the nucleus when silencing is being established and/or maintained.

The relationship between NAD⁺, silencing, and aging:
The Guarente lab has shown that yeast cell longevity is dependent on the SIR2 gene (KAEBERLEIN et al. 1999 ). Part of this effect is due to the role of Sir2p in regulating rDNA recombination and silencing. Since Sir2p is an NAD⁺-dependent histone deacetylase, changes in NAD⁺ concentration could have profound effects on aging, similar to the effects on silencing. Indeed, as yeast cells age, the intracellular ATP content increases 2.6-fold and the NAD⁺ content increases 20% (ASHRAFI et al. 2000 ). It has been proposed that the increased NAD⁺ concentration may be a response designed to prevent genomic instability (ASHRAFI et al. 2000 ).

NAD⁺ has major roles in metabolic reactions and the maintenance of the cellular redox state. High metabolic activity corresponds to increased glycolysis (as an example) and therefore more NAD⁺ is converted to NADH. It has been proposed that this would divert NAD⁺ away from Sir2p, which uses NAD⁺ for silencing (GUARENTE 2000 ). Sir2p has also been proposed to be a molecular sensor of NAD⁺ levels that translates this information into the appropriate chromatin silencing (GUARENTE 2000 ; SMITH et al. 2000 ). Npt1p is also involved in this process. The extension of yeast life span by caloric restriction requires NPT1 (LIN et al. 2000 ). In contrast, QPT1 of the de novo NAD⁺ synthesis pathway was not required for the life-span extension (LIN et al. 2000 ). This is consistent with our data showing that qpt1 mutants do not affect rDNA or telomeric silencing. It will therefore be interesting to determine whether PNC1 and/or TNA1 also functions in extension of life span by caloric restriction.

	ACKNOWLEDGMENTS

We thank Doug Koshland for the pAS90 plasmid; Patrick Grant, Charles Grubmeyer, and Rolf Sternglanz for helpful discussions; Forrest Spencer for use of the Leica stereoscopic microscope; and Mike Christman and Dan Burke for use of their fluorescent microscopes. We also thank Steve Buck and Patrick Grant for critical comments on the manuscript. This work was supported in part by National Institutes of Health grants GM62385 to J.D.B. and GM61692 to J.S.S.

Manuscript received May 2, 2001; Accepted for publication December 14, 2001.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AFSHAR, G. and J. P. MURNANE, 1999  Characterization of a human gene with sequence homology to Saccharomyces cerevisiae SIR2.. Gene 234:161-168[Medline].

APARICIO, O. M. and D. E. GOTTSCHLING, 1994  Overcoming telomeric silencing—A trans-activator competes to establish gene expression in a cell cycle-dependent way. Genes Dev. 8:1133-1146[Abstract/Free Full Text].

ASHRAFI, K., S. S. LIN, J. K. MANCHESTER, and J. I. GORDON, 2000  Sip2p and its partner Snf1p kinase affect aging in S. cerevisiae. Genes Dev. 14:1872-1885[Abstract/Free Full Text].

BACHMAN, N., M. BIERY, J. D. BOEKE, and N. L. CRAIG, 2002  Tn7-mediated mutagenesis of Saccharomyces cerevisiae genomic DNA in vitro. Methods Enzymol. in press.

BALDUCCI, E., M. EMANUELLI, G. MAGNI, N. RAFFAELLI, and S. RUGGIERI et al., 1992  Nuclear matrix-associated NMN adenylyltransferase activity in human placenta. Biochem. Biophys. Res. Commun. 189:1275-1279[Medline].

BAUDIN, A., O. OZIER-KALOGEROPOULOS, A. DENOUEL, F. LACROUTE, and C. CULLIN, 1993  A simple and efficient method for direct gene deletion in Saccharomyces cerevisiae.. Nucleic Acids Res. 21:3329-3330[Free Full Text].

BERGER, N. A., 1985  Poly(ADP-ribose) in the cellular response to DNA damage. Radiat. Res. 101:4-15[Medline].

BERNSTEIN, B. E., J. K. TONG, and S. L. SCHREIBER, 2000  Genomewide studies of histone deacetylase function in yeast. Proc. Natl. Acad. Sci. USA 97:13708-13713[Abstract/Free Full Text].

BRACHMANN, C. B., J. M. SHERMAN, S. E. DEVINE, E. E. CAMERON, and L. PILLUS et al., 1995  The SIR2 gene family, conserved from bacteria to humans, functions in silencing, cell cycle progression, and chromosome stability. Genes Dev. 9:2888-2902[Abstract/Free Full Text].

BRACHMANN, C. B., A. DAVIES, G. J. COST, E. CAPUTO, and J. LI et al., 1998  Designer deletion strains derived from Saccharomyces cerevisiae S288C: a useful set of strains and plasmids for PCR-mediated gene disruption and other applications. Yeast 14:115-132[Medline].

BRAUNSTEIN, M., A. B. ROSE, S. G. HOLMES, C. D. ALLIS, and J. R. BROACH, 1993  Transcriptional silencing in yeast is associated with reduced nucleosome acetylation. Genes Dev. 7:592-604[Abstract/Free Full Text].

BRYK, M., M. BANERJEE, M. MURPHY, K. E. KNUDSEN, and D. J. GARFINKEL et al., 1997  Transcriptional silencing of Ty1 elements in the RDN1 locus of yeast. Genes Dev. 11:255-269[Abstract/Free Full Text].

CANTAROW, W. and B. D. STOLLAR, 1977  Nicotinamide mononucleotide adenylyltransferase, a nonhistone chromatin protein. Arch. Biochem. Biophys. 180:26-34[Medline].

CHRISTIANSON, T. W., R. S. SIKORSKI, M. DANTE, J. H. SHERO, and P. HIETER, 1992  Multifunctional yeast high-copy number shuttle vectors. Gene 110:119-122[Medline].

COST, G. J. and J. D. BOEKE, 1996  A useful colony colour phenotype associated with the yeast selectable/counterselectable marker MET15.. Yeast 12:939-941[Medline].

DERBYSHIRE, M. K., K. G. WEINSTOCK, and J. N. STRATHERN, 1996  HST1, a new member of the SIR2 family of genes. Yeast 12:631-640[Medline].

DIFCO, 1998 Difco Manual. Becton Dickinson and Company, Sparks, MD.

EMANUELLI, M., F. CARNEVALI, M. LORENZI, N. RAFFAELLI, and A. AMICI et al., 1999  Identification and characterization of YLR328W, the Saccharomyces cerevisiae structural gene encoding NMN adenylyltransferase. Expression and characterization of the recombinant enzyme. FEBS Lett. 455:13-17[Medline].

FRITZE, C. E., K. VERSCHUEREN, R. STRICH, and R. E. ESPOSITO, 1997  Direct evidence for SIR2 modulation of chromatin structure in yeast rDNA. EMBO J. 16:6495-6509[Medline].

FROTHINGHAM, R., W. A. MEEKER-O'CONNELL, E. A. TALBOT, J. W. GEORGE, and K. N. KREUZER, 1996  Identification, cloning, and expression of the Escherichia coli pyrazinamidase and nicotinamidase gene, pncA. Antimicrob. Agents Chemother. 40:1426-1431[Abstract].

FRYE, R. A., 1999  Characterization of five human cDNAs with homology to the yeast SIR2 gene: Sir2-like proteins (Sirtuins) metabolize NAD and may have protein ADP-ribosyltransferase activity. Biochem. Biophys. Res. Commun. 260:273-279[Medline].

GOTTSCHLING, D. E., 1992  Telomere-proximal DNA in Saccharomyces cerevisiae is refractory to methyltransferase activity in vivo. Proc. Natl. Acad. Sci. USA 89:4062-4065[Abstract/Free Full Text].

GOTTSCHLING, D. E., O. M. APARICIO, B. L. BILLINGTON, and V. A. ZAKIAN, 1990  Position effect at S. cerevisiae telomeres: reversible repression of Pol II transcription. Cell 63:751-762[Medline].

GUARENTE, L., 2000  Sir2 links chromatin silencing, metabolism, and aging. Genes Dev. 14:1021-1026[Free Full Text].

HOGEBOOM, G. and W. SCHNEIDER, 1952  The synthesis of diphosphopyridine nucleotide by liver cell nuclei. J. Biol. Chem. 197:611-620[Free Full Text].

IMAI, S.-I., C. M. ARMSTRONG, M. KAEBERLEIN, and L. GUARENTE, 2000  Transcriptional silencing and longevity protein Sir2 is an NAD-dependent histone deacetylase. Nature 403:795-800[Medline].

KAEBERLEIN, M., M. MCVEY, and L. GUARENTE, 1999  The SIR2/3/4 complex and SIR2 alone pro