Enhancing lifespan of budding yeast by pharmacological lowering of amino acid pools

Myriocin increases CLS in prototrophic yeast cells by inducing quiescence

Although our previous studies demonstrated that Myr extends the CLS of yeast, all of the data relied on auxotrophic or nutrient-requiring yeast strains [39]. Conceivably, Myr might only extend CLS in auxotrophic strains because it impairs uptake of one or more of the required nutrients thereby producing a state of nutrient restriction. Furthermore, overall amino acid homeostasis is perturbed in auxotrophic cells and masks how wild-type (WT), prototrophic cells respond to nutrient changes and environmental threats [40]. Thus, for the studies described here we used BY4741 cells made prototrophic with a plasmid complementing all mutant genes as was done for analysis of amino acid pools in the yeast deletion strain collection [41].

Indeed, we find statistically significant enhancement of CLS with untreated prototrophic BY4741 cells dropping to 0.5 fraction viable by day 4±1 while Myr-treated cells don’t drop to this level until day 14±2 (Figure 1B, Area Under the Curve, AUC, 95% CI: 3.515-5.178 vs 14.77-18.05). Lifespan assays with triplicate biological replicas were done 5 times with similar results. These findings validate previous CLS analysis of Myr-treated cells and reveal that the extended lifespan in the presence of Myr is also observed in prototrophic yeast strains.

The SDC culture medium used in our current studies is weakly buffered with phosphate (pH 6) and some have maintained that lifespan enhancing strategies work in such conditions by protecting cells against acidification and accumulation of acetic acid [24]. We previously showed that Myr treatment enhanced lifespan of auxotrophic strains grown in SDC media buffered weakly or strongly, thus eliminating the possibility that drug treatment extended lifespan primarily by protecting cells against medium acidification [39]. We now find that Myr treatment increases lifespan in prototrophic BY4741 cells grown in strongly buffered SDC medium (Supplementary Figure 1). SDC culture medium has a relatively high concentration of amino acids, which increases lifespan by suppressing hyper-acidification of culture media [42].

A hallmark of long-lived yeast strains is the ability to enter a quiescent state referred to as Q and characterized by the ability to survive in the absence of nutrients and then enter the cell division cycle when nutrients become available. In contrast, strains with a normal lifespan enter a senescent or non-quiescent (NQ) state typically found when cells cease growing and enter stationary phase where they rapidly lose viability [43, 44]. Yeast Q cells represent a physiological state similar to G0 cells in higher eukaryotes as exemplified by stem and other long-lived cell types [45].

To determine if Myr treatment promotes entry into the Q state, we analyzed cells at day 3 of a CLS assay and found that drug treatment induces more than 90% of prototrophic BY4741 cells to enter a Q state while less than 10% of untreated cells enter a Q state (Figure 1C, 1D). These data are consistent with our published data showing that Myr treatment reduces activity of TORC1, a highly conserved feature of longevity and entry into quiescence [45], and the activity of one of its effector pathways, the Sch9 kinase (mammalian S6K) [39]. Cells lacking Sch9 activity (sch9Δ cells) are long-lived and enter the Q state [46] as do cells undergoing nutrient limitation and reduction in TORC1 activity [44, 45, 47, 48].

Free amino acid pools are smaller in Myr-treated cells

To understand how Myr reconfigures cellular physiology to enhance longevity, we optimized culture conditions to examine free intracellular amino acid pools (Materials and Methods). Under these conditions, 50-60% of cells bud after about 90 min of growth and complete their initial cell cycle after 2-3 h of growth (Supplementary Figure 2). The growth rate of Myr-treated cells slows at around 4±0.5 h of exposure (Figure 1E). Notably, Myr does not significantly reduce final cell density, since by day 1 of a CLS assay (72 h, Figure 1F) most cells have stopped growing with only a small, variable fraction of 0-20% slowly finishing their final cell cycle before entering quiescence on day 3 (144 h, Figure 1G).

To examine free intracellular amino acid pools, we first used a classical heat-extraction protocol followed by derivatization of amine groups with a fluorescent tag, compound separation by high-pressure liquid chromatography (HPLC) and quantification of fluorescent signals in eluted compounds. Stationary phase cells were inoculated into fresh culture medium and a sample was taken immediately and then at 1 h intervals. We find that the 18 amino acids detected by this procedure show a statistically significant difference in pool size with untreated cells having a larger pool than drug-treated cells (Figure 2A) (Cys not detectable and Ala not changed by Myr-treatment). In untreated cells, the pool of most amino acids begins to increase by the 1 h time point. In contrast, most pools remain relatively constant in Myr-treated cells. This effect of Myr was independent of pool size and was observed for amino acids normally maintained at intracellular concentrations that are relatively low (e.g., Met and Tyr) or high (e.g., Glu and Ser).

To verify free pool data, we quantified amino acids by using gas chromatography-mass spectrometry (GC-MS) as this technology is well suited for examining amino acids [49]. We focused on the 5 h time point because by this time the pool size difference between drug-treated and untreated cells is maximal for most amino acids that can be analyzed by this technique (Asn, Cys, Gln and Pro could not be accurately measured). Samples were analyzed using full-scan mode of the mass spectrometer and ion peaks were normalized to norvaline, added as an internal standard. We present data normalized to the internal procedural standard in order to again emphasize large differences in the size of some amino acid pools (Figure 2B). In both assay procedures, the pools in drug-treated cells are lower than in untreated cells with the exception of His, which is higher at the 5 h time point (Figure 2B) in the GC-MS assay. This difference may result from more effective extraction with the GC-MS procedure. It is also noteworthy that Ala was significantly lower in Myr-treated cells in the GC-MS assay by the 5 h time point (Figure 2B) while there was no statistically significant difference over the 0-6 h assay as measured by AUC.

Lastly, we examined amino acid pools in auxotrophic cells (required uracil, tryptophan, leucine and histidine) grown in strongly buffered SDC medium to eliminate the possibility that weakly buffered medium was somehow necessary for pool lowering in Myr-treated cells. In this culture medium, 14 pools were significantly lowered by Myr treatment (Supplementary Table 1).

Taken together, our results indicate that Myr lowers amino acid pools and extends longevity independent of media acidification.

Myriocin lowers the rate of amino acid uptake

Since Myr inhibits the production of sphingolipids and thus causes a dramatic change in lipid content at the PM, we hypothesized that the effects of Myr treatment on free pools of amino acids may relate to the function or trafficking of amino acid transporters. Specifically, Myr could reduce amino acid uptake by directly interfering with transport activity, promoting endocytic clearance of amino acid transporters, or impairing their delivery to the PM. To understand the effect of Myr on amino acid transport, we first examined amino acid uptake into cells by using a mixture of ¹⁴C-labeled amino acids, consisting primarily of glutamate (16%), branched-chain amino acids (29%), alanine (13%) and aspartate (12.5%) with smaller fractions of the other amino acids. Uptake of radiolabel by cells was measured at 4 h when the pool size of most amino acids is significantly lower in Myr-treated cells (Figure 2). To avoid starving cells for specific amino acids in these assays, we quickly harvested cells directly from SDC medium and suspended them in the assay solution containing only the amino acid whose uptake was measured. The concentration of amino acids in assays using the ¹⁴C-mixture was equal to 11% of the concentration in SDC culture medium to give concentrations slightly below the K_m for most AATs [50]. The uptake rate of the ¹⁴C-amino acid mixture was 52% lower in Myr-treated cells than in untreated cells (Figure 3A), indicating that Myr treatment reduces uptake of several amino acids.

The general amino acid permease, Gap1, transports all 20 amino acids and could be responsible for reduced amino acid uptake [51]. To test this hypothesis, we measured uptake of the ¹⁴C-amino acid mixture in gap1Δ cells and found the rate to be identical to that in BY4741 cells without Myr treatment (Figure 3A). However, Myr treatment reduced uptake slightly more in gap1Δ than in WT cells (Figure 3A, compare BY4741 to gap1∆). Although the magnitude of difference was slight, the larger reduction of amino acid uptake in Myr-treated gap1Δ cells suggests that Gap1 activity contributes to the intracellular pool of amino acids even in the presence of Myr.

Uptake of individual amino acids was examined to identify drug effects on specific AATs. Met uptake was examined because Myr treatment significantly reduces its pool size (Figure 2) and it plays a central role in metabolism and longevity [33, 52–55]. Generally, AATs like the high affinity Met transporter, Mup1, are not highly expressed when substrate is present in culture medium [56, 57]. Since the SDC culture medium used for most experiments in this study contains a relatively high dose of Met (537 μmol/L, 80 mg/ml), Mup1 expression (and thereby activity) is likely to be low. However, when we compared ³H-Met uptake in WT and mup1Δ cells using assays containing 12 μmol/L Met, the substrate K_m for this transporter [58], Met uptake was relatively robust in WT cells but was reduced about 70% in mup1Δ cells (Figure 3B). Thus, in the presence of 12 μmol/L Met, Mup1 is the major Met transporter in BY4741 cells grown in SDC medium. Similar to WT cells, Myr-treatment reduces Met uptake in mup1Δ cells to about the same degree, further supporting the idea of other Met AATs are expressed in BY4741 cells and their activity is also reduced by drug treatment.

Met can be transported by six additional AATs [56], with all of them having a higher K_m for Met than Mup1 [50]. By increasing the Met concentration in the uptake assay from 12 to 68 μmol/L, we assessed if activity of other Met AATs is lowered by Myr treatment. With 68 μmol/L Met and no drug treatment, the rate of uptake in mup1Δ cells is about 40% lower than in BY4741 cells (Figure 3C), supporting the idea that other Met AATs contribute to uptake in conditions of greater Met availability. Met uptake in drug-treated mup1Δ cells is reduced by about 55% compared to untreated cells (Figure 3C), further supporting the hypothesis of additional drug-impaired Met AATs being expressed in our experimental conditions. These results indicate that Myr is antagonizing Met uptake by Mup1 as well as other Met transporters.

Several nutrient transporters including Mup1, Can1 (arginine transporter), Lyp1 (lysine transporter), Tat2 (tryptophan and tyrosine transporter) and Fur4 (uracil transporter) are partitioned into a PM domain called the MicroCompartment of Can1 (MCC). MCCs were first associated with endocytosis of PM proteins [59, 60] and now have additional functions [57]. Some of these functions involve sphingolipids and the target of rapamycin complex 2 (TORC2), which is known to play roles in controlling de novo sphingolipid biosynthesis [61–64]. Because of these associations and the low pool of Arg in Myr-treated cells (Figure 2), we examined Arg uptake and found that Myr treatment inhibited uptake of ¹⁴C-Arg by about 50% (Figure 3D). Only Can1 and Gap1 are known to transport Arg. To determine if one or both of these AATs are present and responding to drug treatment, we examined Arg uptake in untreated gap1Δ cells. Uptake is not statistically different in untreated gap1Δ and BY4741 cells (Figure 3D), indicating that Can1 is the major Arg transporter under our assay conditions. Myr treatment of gap1Δ cells reduces ¹⁴C-Arg uptake by about 65%. In drug-treated cells there is also a statistically significant lower rate of Arg uptake in gap1Δ cells compared to BY4741 cells (Figure 3D). From these data we conclude that reduced uptake in both BY4741 and gap1Δ cells is due to effects of Myr on Can1 uptake activity.

Lastly, since Myr lowers the free pool of glutamate (Glu) (Figure 2), we examined its uptake, which is only known to occur by the activity of Dip5 and Gap1 [65]. Dip5 is a dicarboxylic acid transporter that also transports aspartate, Gln, Asn, Ser, Ala, and Gly [50, 65], all of which exhibit decreased levels in cells treated with Myr. Notably, Dip5 is not present in MCC domains [57] and thus, provides insights into whether the effects of Myr are restricted to MCC-localized AATs. Uptake of ¹⁴C-Glu in BY4741 cells is lowered about 50% by Myr-treatment (Figure 3E). To examine the effect of Myr treatment on Dip5-mediated Glu transport, we measured ¹⁴C-Glu uptake in gap1∆ cells and found a 65% decrease in uptake rate (Figure 3E). These data indicate that a majority of Glu is transported by Dip5 and its activity is lowered by Myr treatment. Importantly, these data indicate that the effects of Myr are not limited to AATs associated with MCCs at the PM.

Myriocin reduces Mup1 transport activity in the plasma membrane

Trafficking of Mup1 from its site of synthesis at the endoplasmic reticulum through the secretory pathway to the PM and eventual endocytosis by a ubiquitin-mediated process have been studied extensively [57, 66]. These previous studies provide valuable insights into the regulation of Mup1 trafficking and transport activity in the context of changing amino acid or nitrogen availability. We set out to understand how Myr reduces Mup1 activity and to determine if this affect is caused by (i) reduced delivery of Mup1 to the PM via the secretory pathway, (ii) increased removal of Mup1 from the PM by endocytosis, or (iii) direct inhibition of Mup1 transport activity at the PM.

To evaluate the effect of Myr treatment on Mup1 protein level and cellular location, we examined cells with the chromosomal MUP1 gene fused to the green fluorescent protein (GFP) to determine if drug treatment impaired trafficking of Mup1-GFP to the PM. Starting with log phase cells treated with Met to deplete Mup1-GFP from the PM (Figure 4A, top-left panel, +Met) cells were shifted to methionine-deficient SC medium to promote synthesis of Mup1-GFP in the absence or presence of a range of Myr concentrations. Trafficking of Mup1-GFP to the PM 1 h after the media shift was not impaired by any concentration of Myr (Figure 4A). In these assays Vph1-MARS served as a visual marker for the vacuolar membrane to track the internal pool of Mup1-GFP localized to the lumen of the vacuole. Consistent with these observations, we also found that yeast cells grown to stationary phase (which triggers Mup1 clearance from the PM) [67] and inoculated into methionine-deficient SC medium exhibited delivery of newly-synthesized Mup1-GFP to the PM within 1 h, which was unaffected by Myr treatment (Figure 4B). From these data we conclude that Myr does not impair trafficking of newly synthesized Mup1 to the PM via the secretory pathway.

The media-shift experiments presented in Figure 4 allow for sensitive detection of Mup1-GFP delivery to the PM, but because of the media and culturing protocol cells exhibit a high vacuolar GFP signal at the beginning of each experiment, making it difficult to determine if Myr induces the endocytosis of Mup1-GFP. To assess Mup1 endocytosis, we examined cells harboring chromosomal MUP1 tagged with ecliptic pHluorin (a GFP variant that loses fluorescence at low pH) which does not fluoresce in the vacuole and late endosomal compartments and thus can be used to quantify the abundance of Mup1 at the PM and early endosomal compartments [68–70]. Stationary phase cells inoculated into fresh, methionine-deficient SC medium exhibited a steady increase in Mup1-pHluourin signal over a 6 h culture time course (Figure 5A, 5B), consistent with results observed in Figure 4. However, inoculation into Myr-containing media resulted in an initial increase in Mup1-pHluorin at 3 h followed by a sharp decline in Mup1-pHLuorin by 6 h (Figure 5A, 5B). This Myr-induced decline in Mup1-pHlourin signal at later time points was also detected in flow cytometry analysis (n≥10,000 cells per time point). Specifically, following transfer of stationary phase cells into fresh methionine-deficient medium, we observed very a low Mup1-pHluorin signal that increased until plateauing at about 4 h (Figure 5C). This early phase of Mup1-pHluorin synthesis and delivery to the PM was unaffected by Myr treatment (Figure 5C), consistent with the results of Figure 4. After plateauing around the 4 h time point, the Mup1-pHluorin signal began to decrease slowly (Figure 5C), suggesting that Mup1-pHluorin PM abundance had reached a steady state balanced by endocytic clearance. Importantly, the Mup1-pHluorin signal decreased more rapidly in the presence of Myr, an effect that was dose-dependent (Figure 5C).

To test if the loss of Mup1-pHluorin signal following Myr treatment is due to ubiquitin-mediated endocytosis, we generated yeast cells with chromosomal MUP1 tagged with pHluorin-UL36 (catalytic active and catalytic dead). UL36 is a viral deubiquitylase enzyme (DUB) that can reverse localized ubiquitin conjugation activity, thus protecting the fusion protein from ubiquitin-mediated degradation or trafficking events [71]. Importantly, Myr-treatment induced the clearance of a catalytic dead Mup1-pHluorin-UL36 while the catalytic active variant was partially stabilized (Figure 5D). These results indicate that ubiquitin-mediated endocytosis is at least partially responsible for the loss of Mup1-pHluorin signal following Myr treatment, although this does not exclude the possibility that other clearance mechanisms may contribute to the response. Taken together, these observations indicate that Myr does not affect Mup1 biosynthesis or delivery to the PM but can promote its endocytic clearance from the PM following prolonged exposures.

We next sought to analyze Mup1 trafficking in cells grown in SDC medium containing methionine (537 μmol/L). As expected, the Mup1-pHluorin signal is very low in cells grown in SDC medium and potential effects of Myr treatment are not observable (Supplementary Figure 3). To further characterize the effects of Myr treatment we examined Mup1 protein abundance in cells expressing Mup1-3xFLAG (tagged at the chromosomal locus) by immunodetection of Mup1-3xFLAG in cell lysates (Figure 5E, 5F). There is no detectable Mup1-FLAG at time 0 when stationary phase cells are inoculated into fresh SDC medium (Figure 5E, left lanes, top panel). After growth for 2 h similar FLAG-signals were detected in Myr-treated or untreated cells (Figure 5E). However, while Mup1-FLAG steady state abundance continues to increase through 6 h in untreated cells, its abundance plateaus at 4 h in Myr-treated cells, consistent with increased endocytic clearance at later time points in this experiment (Figure 5E, 5F). Thus, even in culture medium containing a fairly high concentration Met, the effects of Myr treatment on Mup1 abundance are similar to those observed in methionine-deficient SC medium (compare data in Figure 5C, 5E, 5F).

These observations, coupled with the observed reduction in ³H-Met uptake (Figure 3), prompted us to correlate the kinetics of methionine transport with Mup1 trafficking over the course of a 7-hour Myr treatment period (as shown in Figure 5C). The rate of Met uptake is very low following inoculation from a stationary phase culture (Figure 5G), which was expected since Mup1 is not expressed in cells at stationary phase. However, Met uptake capacity increased rapidly over 2 h of growth and to the same degree in cells with and without Myr treatment (Figure 5G). After 2 h of growth, Met uptake capacity slowly decreased in untreated cells, while uptake capacity decreased to a much greater extent in Myr-treated cells. Importantly, the majority of this uptake activity can be attributed to Mup1, since uptake of ³H-Met was reduced by ~70% in mup1∆ cells (Figure 5G). Interestingly, Myr treatment reduced ³H-Met uptake even in mup1∆ cells (Figure 5G), consistent with data shown in Figure 3 indicating that other transporters also contribute to Met uptake in these experiments.

We hypothesized that decreased Met uptake during Myr treatment may be due to Myr-induced endocytic clearance of Mup1, however such clearance was only observed in later time points following Myr treatment. Alternatively, it is also possible that Myr decreases Met uptake by directly inhibiting Mup1 transport activity at the PM. To distinguish between these possibilities, we examined ³H-Met uptake activity of cells grown in SDC medium and observed, as in cells grown in SC medium, very low uptake at time 0 in cells with and without Myr treatment (Figure 5H). In untreated cells, ³H-Met uptake increases steadily during 4 h of growth in SDC, followed by a slight decrease in Met uptake observed at 6 h (Figure 5H). In contrast, Myr-treated cells exhibit an initial increase in Met uptake at 2 h of growth followed by a significant decline in Met uptake to levels observed at the zero-time point (Figure 5H) despite the observed increase in Mup1 abundance during that time frame (Figure 5A–5C, 5E, 5F). This disparity between increased Mup1 abundance and decreased Met uptake in Myr-treated cells indicates that Myr treatment reduces Mup1 transport activity prior to stimulating endocytic clearance from the PM.

Based on these data, we hypothesized that Myr treatment gradually reduces the concentration of sphingolipids in membranes, which reduces Mup1 transport activity. To test this hypothesis, we analyzed Met uptake in culture medium supplemented with phytosphingosine, a downstream product of serine palmitoyltransferase (Figure 1A, SPT). Such supplementation is an established way to bypass both pharmacological and genetic impairment of SPT activity [72, 73]. While phytosphingosine had no effect on ³H-Met uptake in cells that were not treated with Myr, it restored uptake in Myr-treated cells (Figure 5I). Taken together, these results indicate that Myr-treatment induces a loss of Mup1 transport activity followed by endocytic clearance from the plasma membrane.

Sensing of amino acid pool size by TORC1 is vital for increased lifespan

An evolutionarily conserved function of TORC1 is to adjust growth and proliferation to a rate commensurate with amino acid levels [38, 43, 74]. For this reason, amino acid abundance promotes TORC1 activity while amino acid scarcity reduces it. Accordingly, impairment of amino acid pool expansion by Myr treatment (Figure 2) is likely a key upstream regulatory mechanism for slowing the growth rate after 4 h of drug treatment (Figure 1E) and for reducing TORC1 activity as we noted previously [39]. Because reduced TORC1 activity is a prominent and conserved feature of longevity [25, 26, 38], we determined if lower amino acid pool size in drug-treated cells is sensed by TORC1 to reduce its activity and promote CLS.

The Gtr1/Gtr2 GTPase signaling pathway in yeast (RagA or RagB/RagC or RagD in mammals) senses amino acids and regulates TORC1 activity [75–77]. Hence, amino acids promote GTP loading onto Gtr1/2 which activates TORC1 [78]. Conversely, amino acid scarcity turns off Gtr1/2 by causing Iml1, a GTPase-activating protein (GAP) found in a complex with Npr2/Npr3 (SEACIT complex, GATOR1 in mammals), to hydrolyze GTP and inactivate Gtr1/2 which reduces TORC1 activity. We reasoned that deleting IML1 orNPR2 should prevent TORC1 from sensing the smaller amino acid pools in drug-treated cells because Gtr1/2 would remain active, promote TORC1 activity and impair lifespan enhancement. Reduced Myr concentrations were required for lifespan assays to partially mitigate the slower growth rate of iml1Δ and npr2Δ cells and reduced final cell density caused by drug treatment. With reduced Myr concentrations, growth rate and final cell density are slightly reduced compared to BY4741 cells (Figure 6A). We find that untreated iml1Δ and npr2Δ cells lose viability slightly faster than WT BY4741 cells (Figure 6B, 6C). Notably, Myr treatment does not increase the lifespan of iml1Δ and npr2Δ cells. These data support the hypothesis that the Gtr1/2 pathway is essential for sensing lowered amino acid pools, down-regulating TORC1 activity and increasing lifespan in Myr-treated cells.

A yeast pathway for sensing glutamine and leucine upstream of TORC1 includes Pib2, which harbors a PI3P-binding FYVE-domain, and Vps34 which synthesizes this lipid signal [79–81]. We determined if the Pib2 pathway plays a role in Myr-enhanced lifespan by using pib2Δ cells which are not as sensitive to Myr treatment as iml1Δ and npr2Δ cells. In a lifespan assay, untreated pib2Δ cells die slightly faster than WT BY4741 just like iml1Δ and npr2Δ cells (Figure 6D). However, there is a small difference because pib2Δ cells tend to respond to Myr since the 1.309 uM (525 ng/ml) dose slightly increases survival by a small but statistically significant amount relative to pib2Δ cells without drug treatment (AUC: 95% CI, 1.672-2.247 vs 2.980-4.104). These results are consistent with data showing that the Vps34-Pib2 and Gtr1/2 pathways cooperate to control TORC1 activity [82, 83].

Strains, culture conditions, lifespan assays and statistical significance

Yeast strains are described in the Table 1. BY4741 cells and mutant strains from the BY4741 gene knockout library were made prototrophic by transformation with the pHLUM plasmid (Addgene, Watertown, Massachusetts) [40]. SDC culture medium and procedures for CLS assays followed published procedures [117] except for the following changes: culture medium was warmed to 30° C in 125 ml or 1 L flasks. The order of additions was: 95% EtOH, Myr (to give a final concentration of 0.3% EtOH) and then cells to give an initial A600nm units/ml of 0.15 or as indicated in the text. After each addition, cultures were mixed for 10-15 sec. Cultures were incubated in a gyratory shaker (200 rpm) at 30° C for 72 h (day 1 of CLS assay) and viability was quantified by diluting cells in sterile water, plating on YPD plates and counting colonies after 3 days incubation at 30° C. Colony values were plotted as ‘Fraction Viable’ with day 1 set to a value of 1, and subsequent days as a fraction of the day 1 colony count. Myriocin (Cayman Chemicals, #63150) stock solutions of 2.49or 0.996 mM (0.4 mg/ml) in 95% EtOH were treated before use to three cycles of 1 min at 65° C followed by 1 min vortexing. Myr stocks were stored at -20° C. Phytosphingosine (MilliporeSigma, P2795), dissolved in 95% EtOH to give a 20 mM stock, was added to cultures after addition of EtOH and before addition of Myr. The combination of phytosphingosine and Myr slows growth, therefore, we used 1.246 μM (500 ng/ml) of Myr for the uptake assays shown in Figure 5I. In experiments conducted using SC media, cells were maintained and cultured as previously described [69]. A two-tailed Student’s t-test (*= ρ<0.05; **= ρ<0.01; ***= ρ<0.001) and the Area Under the Curve (AUC, Prism 8, Graphpad) were used to evaluate statistical significance of CLS and other data representing at least 3 biological replicates. Results were verified by at least one repeat experiment.

To understand how Myr reconfigures cellular physiology to enhance survival requires time course assays and larger numbers of cells than are available from typical 25 ml cultures used for CLS assays. In addition, a higher starting density than 0.005 A600nm units of cells is needed to access early changes induced by Myr treatment (Figure 1B). By varying the starting culture density and the concentration of Myr, we found conditions in which Myr stimulated CLS to approximately the same extent in 200 ml cultures (1 L flask) started at an A600nm of 0.15 (Figure 1G), as in 25 ml cultures started at an A600nm of 0.005 (Compare Figure 1B, 1G). One limitation of lifespan assays performed on larger cultures and with prototrophic cells is a tendency for mutant cells that feed off of dead cells, to multiply and produce an increase in viability as noted by others [24]. This is what occurred with two of the three drug-treated cultures around day 15 of the experiment shown in Figure 1G and is why no further data points are presented. The amino acid pool and uptake assays and Mup1 protein trafficking data presented here are performed long before re-growth begins and are not affected by it.

Rationale for drug doses

Myriocin has been used in a wide variety of studies and in organisms ranging from yeasts, worms and flies to mammals [14, 92, 99]. In contrast to some studies which use an acute drug dose to induce on-off types of physiological responses, our studies use a moderate drug dose or chronic treatment to induce relatively small, hormesis-like stress and related physiological responses. To improve reproducibility across different types of assays and strain backgrounds, the activity of drug stocks was quantified from the mass yield [118] measured at A600nm after 24 h of growth in SDC containing. A concentration of drug that reduced the A600nm by 20-25% was used for assays (typically 1.245-1.745 μM (500-700 ng/ml)).

Metabolite analyses

A protocol for measuring free amino acid pools was modified to enable time-course assays [78]. Flasks (1 L) containing 200 ml of SDC were warmed to 30° C and EtOH +/- Myr was added as described for CLS assays. Starter cultures of prototrophic BY4741 cells, grown for 24 h in SDC at 30° C, were diluted into the 200 ml cultures to an A600nm of 0.15 units/ml. Flasks were incubated at 30° C on a gyratory shaker at 200 rpm. Every hour for 6 h, 5 A600 units of cells were collected on membrane filters (0.45 μm, HAWP02500, MF-Millipore, MilliporeSigma, Burlington Massachusetts). Filters were immediately rinsed twice with 5 ml of ice-cold Nanopure water, quickly transferred to 0.5 ml cold water, vortexed and incubated 15 min in a boiling water bath. Cells were chilled on ice for 5 min and centrifuged 1 min at 13,000 x g. Fifty microliter of supernatant was saved for protein assay (BCA assay, Pierce) while the rest was filtered (Pall Nanosep® OD003C33 centrifugal device with 3K Omega membrane, Pall Corp.) to remove molecules larger than 3,000 Daltons. Filtrates were transferred to screw-capped tubes, frozen in a dry ice/ethanol bath and stored at -80° C. Amino acid analyses were done at the Proteomics and Mass Spectrometry Facility at the Donald Danforth Plant Science Center (St. Louis, Missouri, USA). Data are expressed as pmoles/ug of protein.

For amino acid analyses by GC-MS, cells were grown as in the heat extraction method, but they were collected on nylon membrane filters (Whatman, 0.45 μm, 25 mm dia., WHA7404002) followed by washing once with 5 ml ice-cold Nanopure water. Filters were immediately immersed into 2 ml of ice-cold acetonitrile:water (1:1, v/v, containing 20 μM L-norvaline as an internal standard) in a 5 ml capped, conical centrifuge tube and vortexed 10 sec before removing the filter. Samples were shaken up and down sixty times and incubated on ice for at least 20 min before freezing in liquid nitrogen and storing overnight at -80° C. Tubes thawed on ice were vortexed 3 min at room temperature followed by centrifugation at 4,500 x g for 5 min at 4° C (Eppendorf #5010R). The supernatant fluid was poured into a chilled 2 ml screw-capped tube, centrifuged again and equal amounts were transferred to microfuge tubes and dried at 10^-3 mBar using a Labconco CentriVap system (ThermoFisher Scientific). Dried samples were either immediately derivatized and analyzed by GC-MS or stored at -80° C. Cells pellets were washed once with 1 ml cold methanol (-20° C) and air-dried. Cell pellets were suspended in 250 μl of 2% sodium dodecyl sulfate and 5 mM Tris-HCl (pH 7.4) overnight at 4° C. Following addition of 250 μl of water, the samples were stored at -20° C. Thawed cell pellets were heated 5 min at 95° C before measuring protein concentration by using a micro-BCA assay (Pierce, ThermoFisher Scientific) and a NanoDrop One spectrophometer (ThermoFisher Scientific).

Sample derivatization

Extracted samples were derivatized by the addition of 239 mM (20 mg/ml) methoxyamine hydrochloride in pyridine and incubation for 1.5 h at 30° C. Sequential addition of N-methyl-N-trimethylsilyl trifluoroacetamide followed with an incubation time of 30 min at 37° C with thorough mixing between addition of solvents. The mixture was then transferred to a V-shaped amber glass chromatography vial and analyzed by GC-MS.

GC-MS quantitation

An Agilent 7800B GC coupled to a 5977B MS detector was used for this study. GC-MS protocols were similar to those described previously [119], except a modified temperature gradient was used for GC with initial temperature at 60° C, held for 1 min, rising at 10° C/min to 325° C and held for 10 min. The electron ionization energy was set to 70 eV. Full scan mode (m/z, 50–800) was used for metabolomics analysis. The identification of metabolites was performed using the Automated Mass Spectral Deconvolution and Identification System (AMDIS) software matched to the FiehnLib metabolomics library (available through Agilent, Santa Clara, CA, USA) with retention time and fragmentation pattern matching [120–123]. Relative quantitation was performed using the software package DExSI (Data Extraction for Stable Isotope-labelled metabolites) (https://github.com/DExSI/). Relative abundance was corrected for recovery using the L-norvaline standard and adjusted to mg/ml protein in cell pellets.

Amino acid uptake assay

Uptake assays were based on published procedures [124]. Cells grown for 24 h at 30° C in SDC or SC media were diluted into 25 mls of fresh culture medium as described for lifespan assays. After incubation for the times indicated in Figure legends, 3 A600nm units of cells were collected on a membrane filter, 0.45 μm dia. (Millipore-Sigma cat. #HAW02500) and washed twice with 5 ml of assay solution (SDC medium lacking amino acids and NH₄SO₄, no pH adjustment) warmed to 30° C. All remaining steps were done at 30° C. Cells were removed by placing the filter into a 50 ml conical, plastic tube containing 1.235 ml of assay solution and vortexed 15 sec before removing the filter. Tubes were shaken at 15 second intervals for 2 min to acidify the medium and then 12.5 ul (1-1.5 x 10⁶ dpm) of a mix of radioisotope and non-radioactive substrate were diluted while vortexing into the assay solution. Three min later, 0.5 ml of this solution was filtered and washed twice with 5 ml cold water. Filters were transferred to 4 ml of Liquid Gold scintillation fluid and radioactivity was quantified in a liquid scintillation counter. Specific activities were calculated as described [124]. Specifically, a standard curve of mg protein/A600nm units verses time of cell growth (0, 2, 4 and 6 h) was made for myriocin-treated and untreated and used to adjust small differences in protein concentration/A600nm unit over time and the two treatments.

Microscopy and protein trafficking assays

Mup1 localization and trafficking were analyzed by fluorescence microscopy and flow cytometry as previously described [69, 70] with minor modifications. Briefly, JTY240 (endogenously expressing Mup1-GFP and Vph1-MARS) and JMY1811 (endogenously expressing Mup1-pHluorin) cells were grown in Met-deficient SC medium and treated with 20 μg/ml methionine to stimulate Mup1 trafficking from the plasma membrane into the vacuole. Mup1-GFP localization in JTY240 cells was visualized over time using a DeltaVision Elite Imaging system (Olympus IX-71 inverted microscope; Olympus 100× oil objective (1.4 NA); DV Elite sCMOS camera, GE Healthcare). Mup1-pHluorin trafficking in JMY1811 cells was analyzed over time using the BD Accuri^TM C6 Plus Flow Cytometer by measuring the relative intensity (FITC channel) of pHluorin signal of 10,000 cells every sampling time point.

Immunoblotting

JMY78 cells (with chromosomal Mup1-FLAG) were grown in SDC medium with or without Myr treatment. Total protein, from 5 A600nm equivalents of cells harvested from each time point, were precipitated by addition of 1 ml of ice-cold 10% trichloroacetic acid in TE buffer (50 mM Tris pH 8.0, 1 mM EDTA) and left on ice for 1 h. Precipitates were washed twice with ice-cold acetone, disrupted by sonication, dried by aspiration and vacuum centrifugation, resuspended in lysis buffer (50 mM Tris pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% SDS), and vortexed vigorously with acid-washed glass beads (0.5 mm dia.). Proteins were diluted with equal volume of 2X sample buffer (150 mM Tris pH 6.8, 6 M urea, 6% SDS, 10% β-mercaptoethanol, 20% glycerol), separated by electrophoresis in 12% Bis-Tris polyacrylamide gels, and transferred onto a polyvinylidene difluoride membrane. The membrane was blocked with 3% bovine serum albumin in TBS-T buffer (50 mM Tris pH 7.5, 150 mM NaCl, 0.1% Tween-20) and incubated with anti-FLAG (Sigma, St. Louis, MO, cat# AB_262044, 1:1000 dilution) and anti-G6PDH (Sigma, St. Louis, MO, cat# AB_258454, 1:10000 dilution) primary antibodies followed with IRDye 680LT goat anti-mouse (LI-COR Biosciences, cat# 926-68020, 1:10000 dilution) and 800CW goat anti-rabbit (LI-COR Biosciences, cat# 926-32211, 1:10000 dilution) secondary antibodies. Fluorescence was visualized using the Odyssey CLx imaging system and quantified using Image Studio Lite (LI-COR Biosciences).