A phylogenetically conserved NAD+-dependent protein deacetylase activity in the Sir2 protein family

Abstract

The yeast Sir2 protein, required for transcriptional silencing, has an NAD(+)-dependent histone deacetylase (HDA) activity. Yeast extracts contain a NAD(+)-dependent HDA activity that is eliminated in a yeast strain from which SIR2 and its four homologs have been deleted. This HDA activity is also displayed by purified yeast Sir2p and homologous Archaeal, eubacterial, and human proteins, and depends completely on NAD(+) in all species tested. The yeast NPT1 gene, encoding an important NAD(+) synthesis enzyme, is required for rDNA and telomeric silencing and contributes to silencing of the HM loci. Null mutants in this gene have significantly reduced intracellular NAD(+) concentrations and have phenotypes similar to sir2 null mutants. Surprisingly, yeast from which all five SIR2 homologs have been deleted have relatively normal bulk histone acetylation levels. The evolutionary conservation of this regulated activity suggests that the Sir2 protein family represents a set of effector proteins in an evolutionarily conserved signal transduction pathway that monitors cellular energy and redox states.

Figure 1

NAD⁺-dependent HDA activity in yeast extracts and purified proteins. Time courses of the reactions done in parallel are shown (A–D). (A) Wild-type (strain YCB180; squares) and quintuple mutant (strain YCB496; sir2 hst1 hst2 hst3 hst4; diamonds) yeast extracts were assayed for deacetylase activity by using a [³H]acetate-labeled chicken histone substrate. Purified Sir2ps were also assayed: (B) GST-Sir2p protein from S. cerevisiae, 30°C; (C) Sir2AF from A. fulgidus, 55°C; (D) GST-Sir2A protein from human, 30°C. GST protein purified in parallel with the GST-Sir2 and GST-Sir2A was completely inactive in several experiments. By using 2 μg GST protein in a 4-h reaction, 7 and 8 cpm were released; with 4, 8, and 16 μg protein 8, 14, and 17 cpm were released, respectively; the activities were all NAD⁺ dependent. With CobB, 33% of the label was released by 1 μg protein after 4 h at 37°C in the presence of NAD⁺ and 0.9% was released in the absence of NAD⁺, by using a 10-fold scaled-down reaction, indicating that CobB is approximately as active as Sir2AF. E and F show the effect of NAD⁺ titrations on the Sir2AF and human Sir2A proteins, respectively. By fitting these data to a hyperbola, we estimate that the NAD⁺ concentrations yielding half-maximal activity are 55 ± 12 μM for Sir2AF and 23 ± 8 μM for human Sir2A.

Figure 2

NAD⁺-dependent histone deacetylase activity in extracts prepared from individual mutants, assayed for 4 h at 30°C. We assayed extracts prepared from the indicated mutants in the presence or absence of 200 μM NAD⁺. The amount of NAD⁺-dependent activity is plotted. Experiments were performed at least in triplicate. Strains used were YCB180 (wild type), YCB178 (sir2), YCB232 (hst1), YCB172 (hst2), YCB394 (hst3), YCB526 (hst4), and YCB234 (hst1 hst2).

Figure 3

NPT1 is required for rDNA silencing and rDNA recombinational stability. (A) Silencing of mURA3 reporter. Strains JS311 (WT), M217 (npt1–1), M269 (npt1–2), and JS566 (sir2Δ) were spotted onto SC medium lacking uracil to measure silencing or nonselective SC medium as a growth control. (B) Silencing of MET15 reporter. Strains are as in A, except the sir2Δ strain is JS218.

Figure 4

Effects of npt1 mutation on telomeric and HMR silencing. (A) Silencing of a telomeric URA3 was analyzed in strains YCB647 (WT), YCB652 (sir2Δ), and JS641 (npt1Δ). Each strain was grown on 5-fluoroorotic acid (Foa) or SC medium. Growth on Foa indicates silencing. The npt1 mutant shows a ≈3,000-fold reduction in the frequency of Foa^R colonies relative to wild-type cells. In the sir2 mutant, the frequency was reduced >15,000-fold relative to wild type. (B) Silencing of a TRP1 reporter gene integrated at HMRΔA was analyzed in YL559 (WT), MC119 (sir2Δ), JS643 (npt1Δ), and JS644 (npt1Δ). Strains were spotted on SC-Trp and SC medium. Growth on the SC-Trp medium indicates loss of silencing.

Figure 5

An NAD⁺ salvage pathway important for transcriptional silencing in yeast. The diagram indicates the enzymatic steps known to occur in synthesis of NAD⁺. For the part of the pathway conserved in Salmonella, the names of the relevant genes proposed or known to carry out each step are given in lowercase letters inside the circle, and the relevant yeast ORFs are written in uppercase. NaMN, nicotinic acid mononucleotide; NaAD, desamido NAD⁺; Nam, nicotinamide; Na, nicotinic acid.

Figure 6

Histone acetylation in vivo in wild-type and mutant strains. Histones were extracted from yeast spheroplasts labeled with [³H]acetic acid. The acid extracts were then separated electrophoretically on acid-urea-Triton gels. (Left) Protein samples stained for total protein content by Coomassie brilliant blue staining. (Right) Fluorograph of the gel, after a 6-month exposure at −70°C. Strains tested were YCB496, 497 (mutant), and 180 (wild type).