Regulation of yeast chronological life span by TORC1 via adaptive mitochondrial ROS signaling

Abstract

Here we show that yeast strains with reduced target of rapamycin (TOR) signaling have greater overall mitochondrial electron transport chain activity during growth that is efficiently coupled to ATP production. This metabolic alteration increases mitochondrial membrane potential and reactive oxygen species (ROS) production, which we propose supplies an adaptive signal during growth that extends chronological life span (CLS). In strong support of this concept, uncoupling respiration during growth or increasing expression of mitochondrial manganese superoxide dismutase significantly curtails CLS extension in tor1Δ strains, and treatment of wild-type strains with either rapamycin (to inhibit TORC1) or menadione (to generate mitochondrial ROS) during growth is sufficient to extend CLS. Finally, extension of CLS by reduced TORC1/Sch9p-mitochondrial signaling occurs independently of Rim15p and is not a function of changes in media acidification/composition. Considering the conservation of TOR-pathway effects on life span, mitochondrial ROS signaling may be an important mechanism of longevity regulation in higher organisms.

Figure 1. Analysis of Mitochondrial Respiration Parameters in vivo and in vitro in tor1Δ and sch9Δ Strains

(A) Mitochondrial membrane potential of DBY2006 (wt) and tor1Δ strains during mid-log (12 hours after inoculation), late-log, post-diauxic (20 hours after inoculation) and early stationary phase (36 hours after inoculation) measured by DiOC6 staining. (B) Oxygen consumption of DBY2006 (wt) and tor1Δ in the presence (TET) or absence (mock) of 100 µM TET. (C) Oxygen consumption of wild-type (wt) and tor1Δ in the presence or absence of 10µM DNP. (D) ATP/O ratio of mitochondria isolated from wild-type (wt), tor1Δ and sch9Δ strains under NADH- or succinate-oxidizing conditions. Error bars represent the mean +/− SD with p-values from a Student’s unpaired t-test denoted as follows (*p < 0.05, **p < 0.01) and “ns” denoting no significance (p > 0.05) between the indicted comparisons. See also Figure S1 and S2.

Figure 2. Effects of Uncoupling Respiration on CLS of Wild-type and tor1Δ Strains

CLS assays (viability as a function of time after inoculation) of tor1Δ (A) or wild-type (B) strains treated with 10µM DNP (tor1Δ, DNP) or vehicle (tor1Δ, mock) for the first 24 hours post-inoculation (gray box). CLS assays of wild-type (C) or tor1Δ (D) strains treated with 10µM DNP from inoculation until termination of the aging experiment. Error bars represent the mean +/− SD.

Figure 3. Coupled Respiration Drives Mitochondria ROS Production in tor1Δ Yeast

(A) FACS analysis of DHE-stained wild-type and tor1Δ strains during early log (6 hours after inoculation), late log (16 hours after inoculation), post-diauxic (20 hours after inoculation) and stationary phase (36 hours after inoculation). (B) Microscopy of DHE fluorescence in wild-type (top) and tor1Δ (bottom) in early log (6 hours after inoculation). Arrows indicate punctate and tubular, mitochondria-like structures. Numbers represent the average percentage of cells displaying more than two foci of DHE fluorescence per cell plus the standard deviation of three biological replicates. Over 200 cells were counted for each replicate. (C) DHE fluorescence of wild-type and tor1Δ strains treated with 10µM DNP (+DNP) or vehicle for the first 20 hours of growth. FACS analysis was performed in late-log phase during post-diauxic growth (as indicated to the right). Error bars represent the mean +/− SD with p-values indicated as described in the legend to Figure 1. See also Figure S3.

Figure 4. Elevating Superoxide via Menadione or Rapamycin Treatment Only During Growth Extends CLS of Wild-type Yeast

CLS assays of wild-type (A) or tor1Δ (B) strains treated with 1µM menadione (MD) or vehicle (mock) for the first 24 hours of growth (grey box). (C) FACS analysis of stationary phase DHE staining of wild-type (wt) and tor1Δ strains that were treated with menadione (+MD) or vehicle for the first 24 hours of growth. (D) Stationary phase mitochondrial membrane potential measured by DiOC6 fluorescence of wild-type and tor1Δ strains treated as described in (C). (E) CLS assays of a wild-type strain treated with 200 nM rapamycin or vehicle (mock) for the first 24 hours of growth (gray box). (F) DHE fluorescence of a wild-type strain treated with rapamycin or vehicle (mock) for the first 24 hours of growth. Bars on the left indicate DHE staining during growth with and without rapamycin treatment; bars on the right indicate DHE staining in stationary phase following rapamycin treatment during growth. Error bars represent the mean +/− SD with p-values indicated as described in the legend to Figure 1. See also Figure S4.

Figure 5. Media Acidification Limits Chronological Life Span in Synthetic Dextrose Medium, But Extended CLS in tor1Δ and sch9Δ Strains is Due to Cell-intrinsic Effects

(A) CLS curves of wild-type (wt) and tor1Δ in un-neutralized (mock) and NaOH neutralized (neutralized) SD media. Media was neutralized at the beginning of stationary phase. (B) Oxygen consumption of neutralized (wt neutralized) or not (wt mock treated) wild-type cells 4 and 24 hours after neutralization. (C) Stationary phase cellular superoxide of wild-type (wt) and tor1Δ cells neutralized or untreated (mock). (D) Mitochondrial membrane potential of cells treated as described in (C). (E) CLS of wild-type and tor1Δ in original media (open symbols) or subjected to media swap (swap) in stationary phase. (F) Same as (E), except wild-type and sch9Δ strains were analyzed. Error bars represent the mean +/− SD with p-values indicated as described in the legend to Figure 1. See also Figure S5 and S6.

Figure 6. The Reduced Chronological Life Span of a rim15Δ Strain is Rescued by Deletion of SCH9 Via Effects on Mitochondrial Respiration and ROS

(A) Chronological life span, (B) mitochondrial oxygen consumption, (C) mitochondrial membrane potential, (D) MitoSox fluorescence, and (E) DHE fluorescence of DBY2006 (wt) and isogenic rim15Δ, sch9Δ and rim15Δ/sch9Δ strains. Mitochondrial parameters were measured in early stationary phase as described in Materials and Methods. Error bars represent the mean +/− SD with p-values indicated as described in the legend to Figure 1. See also Figure S7.

Figure 7. Adaptive Mitochondrial ROS Signaling and Activation of Rim15p-dependent Stress Responses Collaborate to Mediate CLS Extension by Reduced TORC1 Signaling

A speculative model of how reduced TORC1 signaling extends CLS by activating both adaptive mitochondrial ROS signaling (right arm) and Rim15-dependent stress-resistance pathways (left arm). These responses would cooperate in CLS extension by enhancing ROS detoxification and stress resistance and by an adaptive response to elevated cellular superoxide during growth that alters mitochondria function to decrease membrane potential and produce fewer ROS. This adaptive signal may activate some redox-sensitve factor that controls nuclear-encoded mitochondria and/or stress response genes (horizontal arrow), perhaps via epigenetic regulation, and results in altered respiration and enhanced stress responses in stationary phase. This model is meant to encapsulate those aspects of CLS extension by TOR inhibition involving ROS, mitochondria and oxidative-stress resistance. We acknowledge that reduced TOR signaling has other effects on cell physiology that are also important for CLS that are not pictured here.