Identification of a modulator of the actin cytoskeleton, mitochondria, nutrient metabolism and lifespan in yeast

Abstract

In yeast, actin cables are F-actin bundles that are essential for cell division through their function as tracks for cargo movement from mother to daughter cell. Actin cables also affect yeast lifespan by promoting transport and inheritance of higher-functioning mitochondria to daughter cells. Here, we report that actin cable stability declines with age. Our genome-wide screen for genes that affect actin cable stability identified the open reading frame YKL075C. Deletion of YKL075C results in increases in actin cable stability and abundance, mitochondrial fitness, and replicative lifespan. Transcriptome analysis revealed a role for YKL075C in regulating branched-chain amino acid (BCAA) metabolism. Consistent with this, modulation of BCAA metabolism or decreasing leucine levels promotes actin cable stability and function in mitochondrial quality control. Our studies support a role for actin stability in yeast lifespan, and demonstrate that this process is controlled by BCAA and a previously uncharacterized ORF YKL075C, which we refer to as actin, aging and nutrient modulator protein 1 (AAN1).

Fig. 1. Actin stability declines with age.

a Representative SIM images of F-actin and bud scars of young and old wild-type (WT) cells stained with Alexa488-phalloidin and Alexa594-wheat germ agglutinin (WGA594), respectively. Yellow arrow identifies a depolarized actin cable. b Quantification of Alexa488-phalloidin corrected mean signal intensity in mother cells of young and old cells isolated from mCAD. Combined n of cells/strain from 3 trials: 248 (young) and 152 (old). The p value (<1.00e⁻¹⁵) was determined by a two-tailed non-parametric Mann–Whitney test. c Representative SIM images of Alexa488-phalloidin-stained actin and Alexa594-WGA-stained bud scars in young and old WT cells treated with vehicle (DMSO) or 1 µM Lat-A for 10 mins. d Time-course of actin cable loss during treatment with 1 µM Lat-A. Greater than 126 cells/timepoint/condition/trial were analyzed for n of 3 trials. Error bars: SEM. e Growth curves of deletion strains that exhibit reduced sensitivity to Lat-A (10 µM) in our genome-wide screen. f Actin cable abundance in WT cells and the 13 hits from the screen. Actin was visualized by rhodamine-phalloidin staining and wide-field fluorescence microscopy. Error bars: SEM. Scale bars, 2 µm.

Fig. 2. Identification of a gene (YKL075C) that affects actin cable stability, mitochondria, and lifespan.

a Growth curves of WT and ykl075c∆ cells treated with vehicle (DMSO) or 2 µM Lat-A for 16 h. Data from a representative trial is shown (n = 3 trials). b Representative images of Alexa488-phalloidin stained F-actin of WT and ykl075c∆ cells. c Actin cable abundance in mother cells of mid-log phase WT and ykl075c∆ cells. n = 44 cells (WT) and 26 cells (ykl075c∆). Data from a representative trial is shown (n = 3 trials). p value (1.00e⁻⁰⁶) was determined by a two-tailed non-parametric Mann-Whitney test. d Representative images of Alexa488-phalloidin stained F-actin of WT and ykl075c∆ cells treated with vehicle (DMSO) (top panels) or 2 µM Lat-A for 20 min (bottom panels). e Quantitation of actin cables in WT and ykl075c∆ cells during treatment with 2 µM Lat-A. Greater than 256 cells/strain/timepoint were analyzed for n = 5 trials. Error bars: SEM. p value (8.8e⁻⁰³) was determined by simple linear regression analysis. f RLS of WT and ykl075c∆ cells determined by manual micromanipulation. n of cells/strain: 35 (WT) and 36 (ykl075c∆). p value (1.23e⁻⁰⁶) was determined by the two-sided Mantel-Cox test. Data from a representative trial from n of 3 trials. g Representative images of mitochondrial redox state (reduced:oxidized ratio) visualized using mdito-roGFP1 in WT and ykl075c∆ cells. Higher numbers/warmer colors indicate more reducing mitochondria. h Quantification of mito-roGFP1 redox ratios in WT and ykl075c∆ cells. Combined n of cells/strain from three trials: 284 (WT) and 260 (ykl075c∆). p value (< 1.00e⁻¹⁵) was determined by a two-tailed non-parametric Mann–Whitney test. Scale bars, 2 µm.

Fig. 3. Ykl075cp protein localization and links to branched-chain amino acid (BCAA) metabolism.

a Localization of Ykl075cp and the actin cytoskeleton in cells containing either native YKL075C or YKL075C tagged with the Myc epitope of 3 trials. Ykl075cp-13Myc visualized by immunofluorescence using anti-Myc antibodies and Alexa488-tagged secondary antibodies. The actin cytoskeleton was visualized by staining with Alexa568-phalloidin. Scale bar, 2 µm. b Top 15 gene ontology (GO) terms of ykl075c∆ cells. Two-side multiple hypothetic correction test with Bonferroni correction statistical analysis from Yeastract+ database. c Table of BCAA proteins with altered transcript levels in ykl075c∆ cells. d, e Validation of mRNA transcript levels of BCAA transaminases BAT1 and BAT2 by quantitative PCR. Decreased BAT1 and increased BAT2 gene expression in ykl075c∆ cells are significantly different from WT cells (unpaired two-tailed t-test; p values: 2.9e⁻⁰³ (WT vs. bat1∆) and 3.3e⁻⁰³ (WT vs. bat2∆)). Data is representative of three trials. f Quantification of free intracellular BCAA levels in WT, ykl075c∆, bat1∆, and ykl075c∆ bat1∆ cells using Cell BioLabs BCAA colorimetric ELISA kit. p values: 2.43e⁻⁰³ (WT vs. ykl075c∆), 9.00e⁻⁰⁴ (WT vs. bat1∆), 5.00e⁻⁰⁴ (WT vs. ykl075c∆ bat1∆), 1.50e⁻⁰¹ (ykl075c∆ vs. bat1∆), 4.16e⁻⁰² (ykl075c∆ vs. ykl075c∆ bat1∆), and 9.10e⁻⁰¹ (bat1∆ vs. ykl075c∆ bat1∆). g Serial dilutions of WT, aan1(ykl075c)∆, tor1∆, and aan1∆ tor1∆ cells recovering from either DMSO or 200 nM rapamycin treatment on YPD plates. h Growth rate (OD₆₀₀/h for 72 h) of WT, aan1(ykl075c)∆, tor1∆, and aan1∆ tor1∆ cells during recovery from DMSO or 200 nM rapamycin treatment (RR). p values of two biological repeats: 8.49e⁻⁰¹ (WT_DMSO vs. aan1∆_DMSO), 9.96e⁻⁰¹ (WT_DMSO vs tor1∆_DMSO), 8.06e⁻⁰¹ (aan1∆_DMSO vs tor1∆_DMSO), 5.21e⁻⁰¹ (WT_RR vs. aan1∆_RR), 1.90e⁻⁰³ (WT_RR vs tor1∆_RR), and 8.00e⁻⁰⁴ (aan1∆_RR vs tor1∆_RR). i Representative western blot of the phosphorylation of ribosomal protein-S6 (Rps6p) in mid-log phase WT, aan1(ykl075c)∆, and tor1∆ cells in the presence or absence of rapamycin (200 µM). TCE total protein loading control. j Quantification of Rps6p phosphorylation of panel i of untreated or rapamycin-treated WT, aan1(ykl075c)∆, and tor1∆ cells. Data are representative of three trials. p value: 3.50e⁻⁰² (non-parametric Kruskal-Wallis test). Error bars: SEM for panels d-f, h and j.

Fig. 4. AAN1(YKL075C) affects actin cables and mitochondria through its function in BCAA homeostasis.

a F-actin visualized using Alexa488-phalloidin in mid-log phase WT, aan1∆, bat1∆, and aan1∆ bat1∆ cells. b Quantitation of actin cable abundance in panel a. The center band of the box and whisker plot represents the median; the box indicates middle quartiles; whiskers extend to the 10th and 90th percentiles. Combined n of cells/strain from 3 trials: 274 (WT), 246 (aan1∆), 236 (bat1∆), and 258 (aan1∆ bat1∆). P values: 5.16e⁻⁰⁸ (WT vs. aan1∆), 9.50e⁻¹⁰ (WT vs. bat1∆), 5.40e⁻¹¹ (WT vs aan1∆ bat1∆), >0.999 (aan1∆ vs bat1∆), >0.999 (aan1∆ vs aan1∆ bat1∆), and >0.999 (bat1∆ vs. aan1∆ bat1∆). P values were generated from two-tailed non-parametric Kruskal–Wallis tests. c AlexaFluor488-phalloidin stained F-actin of mid-log phase WT, aan1∆, bat1∆ or aan1∆ bat1∆ cells grown in synthetic complete (SC) medium in the presence or absence of leucine (+/−Leu). d Box and whisker plot of actin cable abundance + /−Leu. The center band represents the median; the box indicates middle quartiles; whiskers extend to the 10^th and 90^th percentiles. Combined n of cells/strain/group: 242 (WT_SC), 216 (WT_-Leu), 331 (aan1∆_SC), and 313 (aan1∆_-Leu). p values were generated from two-tailed non-parametric Mann–Whitney test of 3 trials: 5.88e⁻⁰⁹ (WT_SC vs. WT_-Leu) and 2.08e⁻⁰¹ (aan1∆_SC vs. aan1∆_-Leu). e Alexa488- phalloidin stained F-actin of mid-log phase WT, aan1∆, bat1∆, and aan1∆ bat1∆ cells treated with 2 µM Lat-A for 20 min. f Actin cable loss as a function of the time of Lat-A treatment in WT, aan1∆, bat1∆, and aan1∆ bat1∆ cells. Error bars: SEM of three trials (n = 100 cells/strain/trial). p values: 1.30e⁻⁰³ (WT vs. aan1∆), 3.63e⁻⁰² (WT vs. bat1∆), 1.99e⁻⁰² (WT vs. aan1∆ bat1∆), 7.45e⁻⁰¹ (aan1∆ vs. bat1∆), 6.83e⁻⁰¹ (aan1∆ vs. aan1∆ bat1∆), and 9.87e⁻⁰¹ (bat1∆ vs. aan1∆ bat1∆). p values were determined by simple linear regression analysis. g Mitochondrial redox state visualized using mito-roGFP1. h Reduced:oxidized mito-roGFP1 ratios in WT, aan1∆, bat1∆, and aan1∆ bat1∆ cells. Combined n of cells/strain of three trials: 546 (WT), 591 (aan1∆), 581 (bat1∆). p values: <1.00e⁻¹⁵ (WT vs aan1∆), <1.00e⁻¹⁵ (WT vs. bat1∆), and >0.999 (aan1∆ vs. bat1∆). p values were determined using a two-tailed non-parametric Kruskal–Wallis test. Scale bars, 2 µm.