The histone deacetylases Rpd3 and Hst1 antagonistically regulate de novo NAD+ metabolism in the budding yeast Saccharomyces cerevisiae

Abstract

NAD⁺ is a cellular redox cofactor involved in many essential processes. The regulation of NAD⁺ metabolism and the signaling networks reciprocally interacting with NAD⁺-producing metabolic pathways are not yet fully understood. The NAD⁺-dependent histone deacetylase (HDAC) Hst1 has been shown to inhibit de novo NAD⁺ synthesis by repressing biosynthesis of nicotinic acid (BNA) gene expression. Here, we alternatively identify HDAC Rpd3 as a positive regulator of de novo NAD⁺ metabolism in the budding yeast Saccharomyces cerevisiae. We reveal that deletion of RPD3 causes marked decreases in the production of de novo pathway metabolites, in direct contrast to deletion of HST1. We determined the BNA expression profiles of rpd3Δ and hst1Δ cells to be similarly opposed, suggesting the two HDACs may regulate the BNA genes in an antagonistic fashion. Our chromatin immunoprecipitation analysis revealed that Rpd3 and Hst1 mutually influence each other's binding distribution at the BNA2 promoter. We demonstrate Hst1 to be the main deacetylase active at the BNA2 promoter, with hst1Δ cells displaying increased acetylation of the N-terminal tail lysine residues of histone H4, H4K5, and H4K12. Conversely, we show that deletion of RPD3 reduces the acetylation of these residues in an Hst1-dependent manner. This suggests that Rpd3 may function to oppose spreading of Hst1-dependent heterochromatin and represents a unique form of antagonism between HDACs in regulating gene expression. Moreover, we found that Rpd3 and Hst1 also coregulate additional targets involved in other branches of NAD⁺ metabolism. These findings help elucidate the complex interconnections involved in effecting the regulation of NAD⁺ metabolism.

Keywords: NAD(+) biosynthesis; cell metabolism; gene regulation; histone deacetylase; metabolic regulation; yeast genetics; yeast metabolism.

Figure 1

Cells lacking RPD3 are deficient for de novo QA production.A, model of the NAD⁺ biosynthetic pathways in Saccharomyces cerevisiae. De novo NAD⁺ metabolism begins with TRP, which is converted into NaMN by the Bna enzymes (Bna2, Bna7, Bna4, Bna5, Bna1, and Bna6) (left). NaMN is also produced by salvage of NA and NAM, which is further connected with salvage of NR (right). NR is metabolized to NMN by Nrk1, which is then converted to NAD⁺ by Nma1, Nma2, and Pof1. Abbreviations of NAD⁺ intermediates are shown in bold and italicized. 3-HA, 3-hydroxyanthranilic acid; 3-HK, 3-hydroxykynurenine; ACMS, 2-amino-3-carboximuconate-6-semialdehyde; KA, kynurenic acid; KYN, kynurenine; NA, nicotinic acid; NaAD, deamido-NAD⁺; NAM, nicotinamide; NaMN, nicotinic acid mononucleotide; NFK, N-formylkynurenine; NMN, nicotinamide mononucleotide; NR, nicotinamide riboside; QA, quinolinic acid; TRP, l-tryptophan. Abbreviations of protein names are shown in ovals. Aro9/Aro8 and Bna3, kynurenine aminotransferase; Bna1, 3-hydroxyanthranilate 3,4-dioxygenase; Bna2, tryptophan 2,3-dioxygenase; Bna4, kynurenine 3-monooxygenase; Bna5, kynureninase; Bna6, quinolinic acid phosphoribosyl transferase; Bna7, kynurenine formamidase; Nma1/2, NaMN/NMN adenylyltransferase (NMNAT); Npt1, nicotinic acid phosphoribosyl transferase; Pof1, NMN adenylyltransferase (NMNAT); Pnc1, nicotinamidase; Qns1, glutamine-dependent NAD⁺ synthetase; Sir2 family, NAD⁺-dependent protein deacetylases; Urh1, Pnp1, and Meu1, nucleosidases; Nrk1, NR kinase; Isn1 and Sdt1, nucleotidases; Pho8 and Pho5, phosphatases. Tna1, NA, and QA transporter; Nrt1, NR transporter. B, illustration of the QA cross-feeding assay used to determine relative levels of QA release in strains of interest. Spots of haploid single-deletion feeder cells were applied to a lawn of QA-dependent recipient cells (bna4Δnrk1Δnpt1Δ) and allowed to grow for 2 to 3 days at 30 °C. The density of recipient cell growth around the feeder cell spots correlates with the amount of QA released by the feeder cells. C, deletion of RPD3 (rpd3Δ) decreases QA cross-feeding activities, whereas deletion of HST1 (hst1Δ) as well as HST1 and RPD3 together (hst1Δrpd3Δ) increases QA cross-feeding activities. Feeder cell spots along with recipient cells were grown on SC plates at 30 °C for 2 days. For clarity, inverse image is also shown (right). D, extracellular QA levels determined in the growth media. Deletion of RPD3 decreases QA release, whereas deletion of HST1 as well as HST1 and RPD3 together increases QA release. E, intracellular QA levels determined in the cell lysates. Deletion of RPD3 decreases intracellular stores of QA. For D and E, the graphs are based on data of two independent experiments. Error bars represent data from two biological replicates per strain each with three technical replicates (total n = 6 per strain). The p values are calculated using Student’s t test (∗p < 0.05). SC, synthetic complete.

Figure 2

Determination of NAD⁺salvage pathway intermediates and NAD⁺levels in cells lacking RPD3 and HST1.A, hst1Δ and hst1Δrpd3Δ cells show increased NA–NAM cross-feeding activities. Feeder cells spots along with NA–NAM-dependent recipient cells (bna6Δnkr1Δnrt1Δ) were grown on NA-free SC plate at 30 °C for 3 days. B, quantification of NA–NAM production by measuring the extracellular (released) and intracellular (stored) levels of NA–NAM. rpd3Δ cells show decreased release and intracellular storage of NA–NAM, whereas hst1Δ and hst1Δrpd3Δ cells show increased release and storage of NA–NAM. C, rpd3Δ cells show increased NR cross-feeding activities, whereas hst1Δ and hst1Δrpd3Δ cells show decreased activities. Feeder cell spots along with NR-dependent recipient cells (npt1Δbna6Δpho5Δ) were grown at 30 °C on YPD plate for 3 days. D, rpd3Δ and hst1Δ cells show increased intracellular storage of NR. The hst1Δrpd3Δ double mutant shows a further increase compared with the single mutants. Only the rpd3Δ mutant shows a significant increase in NR release. E, rpd3Δ and hst1Δrpd3Δ cells exhibit significantly reduced NAD⁺ levels in standard SC medium. F, rpd3Δ cells display reduced NAD⁺ levels in NA-free SC medium, whereas hst1Δ and hst1Δrpd3Δ display increased NAD⁺ levels. For B and D–F, graphs are representative of the trend observed across three independent experiments. For B and D, error bars represent data from three technical replicates for each strain in an experiment. For E and F, error bars represent data from two biological replicates, each with two technical replicates for each strain in an experiment. The p values are calculated using Student’s t test (∗p < 0.05; ns, not significant). NA, nicotinic acid; NAM, nicotinamide; NR, nicotinamide riboside; SC, synthetic complete; YPD, yeast extract/peptone/dextrose.

Figure 3

Rpd3 and Hst1 regulate homeostasis of de novo intermediates.A, mass spectrometry analysis of TRP levels in rpd3Δ, hst1Δ, and hst1Δrpd3Δ cells. Deletion of RPD3 leads to accumulation of TRP, whereas hst1Δ and hst1Δrpd3Δ cells show reduced TRP levels. B, rpd3Δ cells exhibit defective KYN production. C, rpd3Δ cells show reduced 3-HK levels, whereas hst1Δ and hst1Δrpd3Δ cells show increased 3-HK levels. D, rpd3Δ cells produce reduced levels of 3-HA, whereas hst1Δ and hst1Δrpd3Δ cells produce greater levels of 3-HA. E, deletion of HST1 and especially deletions of RPD3 and HST1 together increase NA levels. All values for each metabolite are normalized to levels in WT cells. Error bars represent data from three technical replicates. The p values are calculated using Student’s t test (∗p < 0.05; ns, not significant). 3-HA, 3-hydroxyanthranilic acid; 3-HK, 3-hydroxykynurenine; KYN, kynurenine; NA, nicotinic acid; TRP, l-tryptophan.

Figure 4

Rpd3 positively regulates de novo NAD⁺metabolism.A, gene expression quantitative PCR (qPCR) analysis of BNA mRNA in WT, rpd3Δ, hst1Δ, and hst1Δrpd3Δ cells. Values shown are relative expression levels normalized to TAF10 as a control. Deletion of RPD3 decreases expression of all BNA genes shown. BNA expression in hst1Δrpd3Δ cells is generally increased relative to WT cells and slightly less than levels in hst1Δ cells. B, comparisons of Bna protein expression in HDAC mutants. HA-tagged Bna1, Bna2, and Bna5 proteins were generated in WT, rpd3Δ, hst1Δ, and hst1Δrpd3Δ cells. Protein expression was determined by Western blot analysis. Arrows make the positions of molecular weight markers. C, overexpression of BNA2 (BNA2-oe) slightly increases the levels of QA release. D, BNA2-oe increases QA release in rpd3Δ, whereas overexpression of both BNA2 and BNA6 in rpd3Δ clears accumulated QA. E, BNA2-oe alone or BNA2-oe and BNA6-oe together is insufficient to raise NAD⁺ levels in rpd3Δ cells grown in SC. F, restoration of de novo pathway activity is necessary to rescue NAD⁺ levels in rpd3Δ cells. NAD⁺ levels in rpd3Δ cells are increased to WT levels when supplemented with QA (at 10 μM) and with BNA6-oe. For A, C, E, and F, the graphs are representative of the trend observed across three independent experiments. For A and C, error bars represent data from three technical replicates for each strain in an experiment. For E and F, error bars represent data from two biological replicates each with two technical replicates for each strain in an experiment. The p values are calculated using Student’s t test (∗p < 0.05; ns, not significant). BNA, biosynthesis of nicotinic acid; HDAC, histone deacetylase; QA, quinolinic acid.

Figure 5

Analysis of Rpd3 and Hst1 binding to the BNA2 promoter.A, design of the chromatin immunoprecipitation (ChIP) studies. HA-tagged Hst1 was generated in both WT and rpd3Δ cells. HA-tagged Rpd3 was generated in both WT and hst1Δ cells. Expression was confirmed by Western blot analysis (top). BNA2 promoter regions for ChIP studies are shown as BNA2 #1, #2, #3, #4, and #5 (bottom). B, confirming the QA cross-feeding phenotypes of HA-tagged strains. Both Rpd3-HA WT and Hst1-HA WT cells show WT levels of QA release. Deletion of HST1 in Rpd3-HA WT cells increases levels of QA release. Deletion of RPD3 in Hst1-HA WT cells decreases levels of QA release. C, ChIP analysis of Rpd3 binding to the BNA2 promoter. The pattern of Rpd3 binding is altered in hst1Δ cells. Binding activity of Rpd3 is most significant near BNA2 #2 in WT cells, which shifts to BNA2 #5 in hst1Δ cells. Relative IP levels were normalized to TAF10. D, Hst1-binding activity is the highest near the transcription start site (BNA2 #5) in an ascending pattern. Hst1-binding activity is decreased when Rpd3 is absent. For C and D, the graphs are representative of the trend observed across three independent experiments. Error bars represent data from three technical replicates for each strain in an experiment. The p values are calculated using Student’s t test (∗p < 0.05; ns, not significant). E, model of Rpd3 and Hst1 binding to the BNA2 promoter. BNA, biosynthesis of nicotinic acid; QA, quinolinic acid.

Figure 6

Rpd3 and Hst1 have opposing effects on histone H4 acetylation status at the BNA2 promoter.A, relative abundance of acetylated H4K5 (H4K5-Ac) at sites 3 and 5, depicted in Figure 5, A, of the BNA2 promoter (left). Deletion of RPD3 slightly decreases the amount of H4K5-Ac, whereas deletion of HST1 as well as deletions of RPD3 and HST1 together increase the level of H4K5-Ac, suggesting that Hst1 is the main deacetylase for this residue (right). B, relative abundance of H4K12-Ac at sites #3 and #5 of the BNA2 promoter (left). rpd3Δ cells show reduced acetylation of H4K12, whereas hst1Δ and hst1Δrpd3Δ cells show increased acetylation of H4K12, suggesting that Hst1 is primarily responsible for the deacetylation of this residue (right). C, relative abundance of H4K8-Ac at sites #3 and #5 of the BNA2 promoter (left). rpd3Δ and hst1rpd3Δ cells show increased acetylation of H4K8, whereas deletion of HST1 alone does not have a significant influence on H4K8-Ac levels, suggesting that Rpd3 is the main deacetylase for this residue. Values are relative to levels of H4 protein-bound DNA in each strain, and all values are normalized to those of WT cells. The graphs are representative of the trend observed across three independent experiments. Error bars represent data from three technical replicates for each strain in an experiment. The p values are calculated using Student’s t test (∗p < 0.05; nd, not detected; ns, not significant). D, model of BNA2 expression and putative chromatin structure produced by the effects of Rpd3 and Hst1. BNA, biosynthesis of nicotinic acid.

Figure 7

Rpd3 and Hst1 regulate different downstream target genes in NA–NAM and NR salvage pathways.A, relative expression analysis of the genes of the NA–NAM (left) and NR (right) salvage pathways in WT, rpd3Δ, hst1Δ, and hst1Δrpd3Δ cells by quantitative PCR (qPCR). B, relative expression analysis of the genes of the NR salvage pathway in WT, rpd3Δ, hst1Δ, and hst1Δrpd3Δ cells by qPCR. All values shown are relative expression levels normalized to TAF10 as a control. The graphs are representative of the trend observed across three independent experiments. Error bars represent data from three technical replicates for each strain in an experiment. The p values are calculated using Student’s t test (∗p < 0.05; ns, not significant). NA, nicotinic acid; NAM, nicotinamide; NR, nicotinamide riboside.