Acetic acid effects on aging in budding yeast

In a recent issue of Cell Cycle, Burtner et al. presented evidence that the accumulation of acetic acid in stationary phase budding yeast cultures is “the primary mechanism of chronological aging in yeast”.¹ Burtner et al. suggest that “how acetic acid-induced cell death could contribute to aging in higher organisms is not readily apparent.” We also recently investigated the effects of pH on chronological lifespan (CLS) of budding yeast. The results of our experiments and those of previous studies point to a mechanism of acetic acid toxicity in yeast related to the induction of growth signaling pathways and oxidative stress. These mechanisms are relevant to aging in all eukaryotes.

CLS is defined as the length of time budding yeast cells survive after undergoing a nutrient depletion-induced arrest of the cell cycle in stationary phase. To simplify discussion, we refer to this arrest as “growth arrest.” As in the experiments of Burtner et al. buffering medium to maintain a higher pH lengthened CLS compared to CLS in unbuffered medium (Fig. 1A). Buffering medium also inhibited the age-dependent accumulation of reactive oxygen species detected by the fluorescent probe dihydroethidium, which is mostly sensitive to superoxide anions (Fig. 1B). Thus, the accumulation of acetic acid in stationary phase induces oxidative stress, a factor previously implicated in chronological aging of yeast and aging in other organisms as well.

Figure 1.

Effects of buffering pH in yeast chronological aging experiments. (A) Viability of BY4741 wild type cells cultured in synthetic complete (SC) medium⁴ for indicated times in the absence (“unbuffered”) or presence (“buffered”) of 50 mM citrate-phosphate buffer (pH 6.0). The osmolarity of unbuffered medium was adjusted with sorbitol to maintain the same osmolarity as buffered medium. Viability was measured as colony-forming units. Data represent averages and standard deviations of three biological replicas. Unbuffered cultures achieved a final average pH of 3.19, and buffered cultures achieved a final average pH of 5.26. Similar results were obtained in a second genetic background (DBY746; not shown). (B) Reactive oxygen species in aliquots of 25,000 cells from buffered or unbuffered cultures were measured by flow cytometry at the indicated times of medium depletion using dihydroethidium (DHE). Y axis indicates number of cells producing DHE fluorescence signals in individual channels of the flow cytometer; X axis indicates channels detecting progressively higher levels of DHE fluorescence and is scaled exponentially. (C) The fraction of visibly budded cells (“% visible buds”) was determined microscopically at indicated times in aliquots of 500 or more cells from each time point. Data represent averages and standard deviations in three biological replicas. (D) Viability of cells with or without visible buds was assessed by staining aliquots of cells collected at indicated times with propidium iodide (PI) followed by microscopic examination with appropriate filters. Total cells counted in each sample exceeded 2,000 cells. Data are representative of results in 4 biological replicas. Additional PI staining data are presented in Figure S1.

Buffering medium also increased the frequency with which cells arrest growth in G₁ when they enter stationary phase, as indicated by a reduced number of stationary phase cells with microscopically detectable buds (Fig. 1C). Counting budded cells underestimates the number of stationary phase cells that fail to arrest in G₁, because cells with small buds are difficult to detect. It also does not account for a potential reduction in the number of budded compared to unbudded cells associated with apoptotic cell destruction, which occurs more frequently in stationary phase populations enriched for cells with buds.² In fact, staining cells with the DNA stain propidium iodide (PI) to detect dead cells that have lost membrane integrity—which includes cells at late stages of apoptosis—revealed a minimum twofold increase in the rate at which cells with visible buds compared to those without visible buds were dying in unbuffered stationary phase cultures (Fig. 1D). Since cells with small buds are difficult to detect, some of the PI-stained cells without visible buds were also likely to be cycling when they died. In contrast, in buffered stationary phase cultures, a significant difference was not detected in PI staining of cells with or without visible buds (Fig. S1). These findings indicate that the accumulation of acetic acid in stationary phase cultures inhibits growth arrest of cells in G₁ and is preferentially toxic to cells that fail to undergo a G₁ arrest.

The data of Burtner et al. show that osmotic stress or mutational inactivation of conserved growth signaling pathways (deletion of genes encoding the AKT homologue Sch9 or the RAS homologue Ras2) confer resistance to acetic acid toxicity in stationary phase cultures via undefined mechanisms. They also show that caloric restriction extends CLS by reducing the accumulation of acetic acid in stationary phase cultures. Deletion of SCH9 induces the superoxide dismutase Sod2 in stationary phase cells.³ Since oxidative stress is increased by acetic acid in wild type cells (Fig. 1B), induction of Sod2 in sch9Δ cells could explain some of the resistance of this strain to acetic acid toxicity. However, similar to the effects of buffered medium on stationary phase G₁ arrest described above, CLS extension by deletion of SCH9, RAS2, increased osmolarity or by caloric restriction are all accompanied by more frequent stationary phase growth arrest in G₁.⁴ Together with the findings reported here, these and other findings described in this earlier study point to DNA replication stress—i.e., inefficient DNA replication—arising downstream of growth signaling that inhibits G₁ arrest in stationary phase as an additional component of the lifespan-shortening effects of acetic acid in chronological aging experiments. This conclusion is supported by the recent discovery of a distinct population of “non-quiescent” stationary phase cells that, in addition to its enrichment for budded and apoptotic cells, exhibits elevated expression of genes encoding proteins that respond to replication stress.²

A role for acetic acid in the induction of growth signaling pathways and replication stress in yeast is not surprising. In budding yeast and many other fungi, intracellular acidification—which occurs when acetic acid is introduced into medium⁵—activates the same highly conserved Ras2 and cAMP-dependent signaling pathways that respond to glucose.⁶ Thus, despite the absence of glucose (which has been completely consumed), nutrient-depleted stationary phase cells are continuously subjected to acetic acid-induced growth signals that promote entry into S phase, but in the absence of nutrients or the regulatory mechanisms required for the synthesis of dNTPs and efficient DNA replication. This is a recipe for replication stress.

In the context of the well-documented stimulation of growth signaling pathways by intracellular acidification in yeast and evidence for the induction of replication stress in stationary phase cells, it's likely that the resistance to acetic acid toxicity in sch9Δ and ras2Δ cells reported by Burtner et al. is related in part to reduced growth signaling that promotes the more frequent growth arrest in G₁ we detected earlier in these strains.⁴ This would protect against acetic acid-induced replication stress. The more efficient stationary phase G₁ arrest induced by osmotic stress is expected to similarly protect against acetic acid-induced replication stress, perhaps through its induction of the cyclin-dependent kinase inhibitor Sic1. Sic1, which blocks entry into S phase, is required for a Hog1-dependent G₁ arrest induced by osmotic stress.⁷ Hog1 also confers resistance to acetic acid toxicity in cycling cells.⁵ Whether deletion of SCH9 or RAS2 activates Sic1 is not known; except for our study,⁴ efforts to understand the role of Sch9 and other growth signaling molecules in CLS have focused on inhibition of Sod2 and other oxidative stress responses. Based on studies of the mammalian homologues of Sch9 and Sic1 (AKT and p27), however, this is likely to be the case—AKT inhibits G₁ arrest in mammalian cells by inhibiting the activity of p27.⁸

Are the CLS-shortening effects of acetic acid in yeast relevant to aging in higher eukaryotes? Although acetic acid is not a specific physiological trigger of growth signaling pathways in more complex organisms, induction of mitogenic signaling by low pH has been reported in higher eukaryotes, where it contributes to a variety of pathological states related to aging, such as inflammation and cancer. This includes, for example, activation by low pH of AKT and cellular proliferation in Barrett's esophagus, a precursor to esophageal cancer for which chronic acid reflux is a major risk factor.⁹ Extracellular acidification—which is detected in many solid tumors—also has been linked to proliferation of medulloblastoma cells by triggering the formation of IP3, a potent activator of AKT, and activating the MEK/ERK mitogen signaling pathway.¹⁰ There are many other examples in the literature of the induction by low pH of RAS, cAMP and/or AKT-dependent mitogenic signaling pathways associated with cancer, chronic acidosis-related diseases or regulation of differentiation during embryogenesis.

Even more relevant is the fact that growth signaling through conserved insulin-like growth factor (IGF-1) pathways regulated by AKT, RAS and cAMP and other molecules is a well-established factor in aging in all eukaryotes. Growth signaling through these pathways also contributes to neurodegenerative and other age-related diseases, including cancer.¹¹ Similar to the Sch9-dependent inhibition of both Sod2 expression and G₁ arrest that occurs in response to growth signaling by acetic acid in stationary phase yeast cells, growth signaling through AKT in mammals suppresses MnSod and other oxidative stress responses,^12,13 in addition to inhibiting p27 and G₁ arrest.⁸ Also similar to the effects of acetic acid in yeast stationary phase cells, replication stress arises in differentiating mammalian cells that normally arrest growth in G₁, but inappropriately arrest growth in S phase downstream of sustained growth signaling by RAS, AKT and other oncogenes at early stages of cancer.¹⁴ A role for replication stress in aging of higher eukaryotes has not yet been explored. However, longevity-promoting NAD-dependent deacetylases counteract the effects of growth signaling through AKT and other proteins by inducing p27 and G₁ arrest, in addition to oxidative stress responses.^15,16 The induction of p27 and G₁ arrest by NAD-dependent deacetylases—which is rarely described in discussions of the longevity-promoting capabilities of these proteins—likely promotes longevity by protecting against replication stress. Also relevant here is the observation that caloric restriction—which prolongs the lifespan of most, if not all, organisms—induces a “metabolic checkpoint” that (similar to the more frequent stationary phase G₁ arrest induced by caloric restriction in nutrient-depleted yeast cells⁴) drives mouse cells into G₁. Abrogation of this checkpoint leads to growth arrest in S phase followed by apoptosis.¹⁷

In summary, Burtner et al. raise an important question about the relevance of acetic acid effects in the yeast chronological aging model to aging in higher eukaryotes. Results presented here and in related studies that were not cited by Burtner et al. indicate that in yeast, accumulation of acetic acid in stationary phase cultures stimulates highly conserved growth signaling pathways and increases oxidative stress and replication stress, all of which have been implicated in aging and/or age-related diseases in more complex organisms. Low pH also stimulates growth signaling pathways in mammals. Although the reduced production of acetic acid identified by Burtner et al. as a factor in the CLS-extending effects of caloric restriction in yeast may be specific for this organism, the underlying mechanism by which this protects against chronological aging is likely to be same as for caloric restriction in higher eukaryotes—that is, reduced growth signaling that inhibits both oxidative and replication stress. The remarkable parallels between regulation of chronological aging in yeast and of aging in more complex organisms suggest that conserved growth signaling pathways impact aging in all eukaryotes via dual effects on oxidative and replication stress. The yeast chronological aging model will likely continue to provide insights that will guide us toward a better understanding of these and other aspects of aging and age-related diseases.