Loss of vacuolar acidity results in iron-sulfur cluster defects and divergent homeostatic responses during aging in Saccharomyces cerevisiae

Abstract

The loss of vacuolar/lysosomal acidity is an early event during aging that has been linked to mitochondrial dysfunction. However, it is unclear how loss of vacuolar acidity results in age-related dysfunction. Through unbiased genetic screens, we determined that increased iron uptake can suppress the mitochondrial respiratory deficiency phenotype of yeast vma mutants, which have lost vacuolar acidity due to genetic disruption of the vacuolar ATPase proton pump. Yeast vma mutants exhibited nuclear localization of Aft1, which turns on the iron regulon in response to iron-sulfur cluster (ISC) deficiency. This led us to find that loss of vacuolar acidity with age in wild-type yeast causes ISC defects and a DNA damage response. Using microfluidics to investigate aging at the single-cell level, we observe grossly divergent trajectories of iron homeostasis within an isogenic and environmentally homogeneous population. One subpopulation of cells fails to mount the expected compensatory iron regulon gene expression program, and suffers progressively severe ISC deficiency with little to no activation of the iron regulon. In contrast, other cells show robust iron regulon activity with limited ISC deficiency, which allows extended passage and survival through a period of genomic instability during aging. These divergent trajectories suggest that iron regulation and ISC homeostasis represent a possible target for aging interventions.

Keywords: Aging; DNA damage; Geroscience; Iron-sulfur clusters; Lysosomal acidity; Vacuolar acidity; Yeast replicative lifespan.

Fig. 1

Enhanced iron uptake or additional iron allows respiratory growth following impairment of the V-ATPase. a Overexpression of FET4 from a high copy (2 μ) plasmid rescues growth of vma21Δ mutants under respiratory conditions. b A vma21Δ mutant strain containing a suppressing loss of function mutation in the ROX1 gene (rox1-A10T) rescues the vma21Δ respiratory deficiency phenotype. cFET4 is required for rox1Δ rescue of vma21Δ respiratory growth. d Supplemental iron (500 μM ferrous ammonium sulfate) rescues pleiotropic phenotypes of vma21Δ mutants. e Chemical inhibition of the V-ATPase with 3 μM bafilomycin A1 preferentially impairs growth under respiratory conditions and iron rescues this phenotype. f Replicative lifespan under fermentative growth conditions (YPD media) of wild type or vma21Δ mutants with or without 20 mM sodium ascorbate and/or 0.5 mM iron II sulfate. Where indicated, fermentative growth is on YPD media containing 2% glucose, and non-fermentative respiratory growth conditions is growth on YPG media containing 3% glycerol. Lifespan statistics are shown in Table S7

Fig. 2

Iron dyshomeostasis occurs following disruption of the V-ATPase. a Flow cytometry analysis of GFP levels under control of the Aft1 responsive FIT2 promoter of wild type and vma21Δ mutants grown under fermentative conditions (YPD media) with or without 1 mM iron. Values are the mean of three samples, each consisting of 20,000 cells per condition. Statistics are shown in Table S8. Error bars represent the standard deviation of the three sample means. b Fluorescent microscopy images of wild type or vma21Δ mutants in the presence or absence of 100 μg/ml bathophenanthrolinedisulfonic acid (BPS, iron chelator) or 1 mM iron ammonium sulfate. c Iron levels were measured in wild type and vma21Δ cells grown in YPD media. No statistical difference was found using Student’s t test. n = 3, error bars represent standard deviation. d Aconitase activity was measured for wild-type cells, vma21Δ, and aco1Δ mutants. Each group is statistically significantly different from each other group by ANOVA and Bonferroni’s multiple comparison test P < 0.001. n = 3, error bars represent standard deviation

Fig. 3

Age-associated decline in vacuolar acidity is predictive of replicative lifespan and is associated with iron regulon activity which increases with age. However, a large subset of cells displays little to no iron regulon activity during aging. a Vacuolar acidity trend during aging, as measured by fluorescence ratio of vacuole-localized Prc1-pHluorin2. Acidity is normalized per cell to young cell value (average before the second division). Pearson r = − 0.259, p = 1 × 10⁻¹⁴⁹, n = 9857 cell divisions, error bars are standard error of the mean (SEM). b Survival curves for the population partitioned into two groups by the rate of vacuolar acidity loss during early life (0–12 divisions). The rate of vacuolar acidity loss rate for each cell is calculated using the slope of the least-squares regression line through the acidity values during divisions 0–12. Population is split in half using the median rate of vacuolar acidity loss. Fast rate of decline in vacuolar acidity is associated with shorter lifespan, logrank p = 2.22 × 10⁻¹⁶, n = 289 cells per group. c Age associated trend of population average iron regulon activity (Fit2-mRuby2 fluorescence). Iron regulon activity increases with age, Pearson r = 0.13, p < 10⁻³⁸. d Scatter plot of iron regulon activity (Fit2-mRuby2 fluorescence) and vacuolar acidity for all cells at all ages. Iron regulon activity is correlated with lower vacuolar acidity even when controlled for age (all ages: Spearman ρ = − 0.21, p < 10⁻⁴; age 10: Spearman ρ = − 0.17, p = 0.0017; age 20: Spearman ρ = − 0.38, p < 10⁻⁴). e Single-cell lifetime trajectories of measured iron regulon activity. Color values are log (Fit2-mRuby2 fluorescence). Many cells show limited to no iron regulon activation

Fig. 4

Iron-sulfur clusters become deficient during aging. Activation of iron regulon is inversely correlated with iron-sulfur cluster deficiency. Cells with robust iron regulon activity during aging generally follow a trajectory of limited iron-sulfur cluster deficiency. n = 209 total cells. a Aging trend of population mean iron-sulfur cluster insufficiency (Rps2 nuclear enrichment). Iron-sulfur cluster insufficiency increases during aging. Pearson r = 0.52, p < 10⁻⁴. Error bars are SEM. b Scatter plot of iron-sulfur cluster insufficiency (nuclear enrichment of Rps2-GFP) and iron regulon activation (Fit2-mRuby2 fluorescence) in aged cells (all cells and ages > 12 divisions). Spearman ρ = − 0.26, p < 10⁻⁴, n = 189 (cells that lived 12 generations or more). c Aging trend of iron-sulfur cluster deficiency for iron regulon competent and incompetent cells during aging. Iron regulon-active cells are defined as having a maximum Fit2 fluorescence level that is 3× the median baseline. Iron regulon-inactive cells are not more iron-starved during early life, but have less severe iron-sulfur cluster deficiency from middle to old age (p < 0.05 for ages 15–30). n_{active iron regulon} = 59, n_{inactive iron regulon} = 150. Error bars are SEM. d Single-cell trajectories of iron-sulfur cluster deficiency and iron regulon activity during aging. Cells display divergent iron metabolism trajectories during aging. Red: iron regulon active with limited iron-sulfur cluster deficiency. Blue: no iron regulon activation. n = 209. e Example aged cells in adjacent traps from each iron metabolism trajectory. Top cell in trap: iron-sulfur cluster deficiency (Rps2-GFP nuclear retention visible as bright green dot) with little iron regulon activation. Bottom cell in trap: robust iron regulon activation (Fit2-mRuby2) with no visible Rps2-GFP nuclear retention

Fig. 5

Activation of the DNA damage response increases during aging, is correlated with the loss of vacuolar acidity, and is predictive of a shorter replicative lifespan. a Aging trend of DNA damage response activation as measured by presence of Rad52-GFP foci. As cells age, the fraction of cells that activate the DNA damage response increases, Pearson r = 0.27, p < 10⁻⁴, n = 266 cells. Error bars are SEM. b Scatter plot of age of first DNA damage response activation (Rad52-GFP foci) and rate of vacuolar acidity loss in early life for cells which activate the DNA damage response at any point during life (ages 0–12 divisions). n = 244 cells that developed foci. Faster loss of vacuolar acidity during early life (ages 1–12 divisions) correlates with earlier appearance of DNA damage response. Pearson r = − 0.23, p = 0.0006. c Comparison of vacuolar acidity for cells with and without activated DNA damage response (Rad52-GFP foci) at various ages. n = 266. Presence of activated DNA damage response (Rad52-GFP foci) is associated with lower vacuolar acidity when controlled for age. All comparisons ranksum p < 0.05 except for ages 15: p = 0.089, 18: p = 0.45, and 20: p = 0.20. Error bars are SEM

Fig. 6

Iron regulon-active cells survive longer after the first activation of the DNA damage response during aging and undergo more lifetime divisions during which the DNA damage response (Rad52-GFP foci formation) is activated. Total lifespan of iron regulon-active cells is highly correlated to the number of lifetime divisions during which the DNA damage response is activated. Within the iron regulon-inactive group, longer-lived cells do not undergo any more divisions with DNA damage than shorter lived cells. This suggests that within the iron regulon competent subpopulation, the longest-lived cells are those with the most robust ability to survive age-associated DNA damage. Within the iron regulon incompetent subpopulation, the longest-lived cells are those with minimal age-associated DNA damage. a Single-cell trajectories of DNA damage response activation during aging. Top: cells that activate the iron regulon during aging (maximum lifetime Fit2-mCherry fluorescence > 3× median baseline level). Bottom: cells that fail to activate the iron regulon during aging. Empty spaces indicate that data was not collected for a single cell at that age, e.g., due to temporary cell segmentation or tracking error or initial age of capture after first division. b Survival curves comparing remaining lifespan after first observed activation of the DNA damage response (Rad52-GFP foci formation) for iron regulon competent and incompetent cells. Cells that activate the iron regulon during aging have increased survival after the first activation of the DNA damage response logrank p = 0.018, n_{active iron regulon} = 76, n_{inactive iron regulon} = 129. Error bars are SEM. c Comparison of DNA damage response activation between iron regulon active and iron regulon-inactive cells. During aging, cells that activate the iron regulon undergo more divisions during which the DNA damage response is activated (Student’s t p < 10⁻⁴). d Scatter plots showing correlation between total lifespan and number of divisions during which the DNA damage response is activated. Longer lifespan of the iron regulon-active cells is well-correlated (Pearson r = 0.41) with a higher number of divisions with an activated DNA damage response. Longer lifespan in iron regulon-inactive cells is not correlated (Pearson r = 0.08) with more divisions with DNA damage. Correlation coefficients are significantly different by Fisher’s Z-transform p = 0.01