Figures and data
Strain design
(A) Pleckstrin homology (PH) domain of SLM1 gene fused with NTT1 and inserted in the vector (pAG306GPD-ccdB-chr1). The constructed (PH-NTT1) plasmid, digested with NotI, was integrated into the yeast genome. (B) In a separate construct, yeGFP sequence was fused to the 3’ end of PH-NTT1 sequence and expressed in yeast and visualized by confocal microscopy to assess cellular localization. (C) Overall yeast strain design with NTT1 overexpression mediating ATP import from extracellular environment.
Measurement of relative ATP levels
(A) Yeast cells, expressing fluorescent QUEEN ATP reporters were imaged using fluorescence microscopy. Relative ATP levels were measured in cells grown in media containing (B) 2% glucose or (C) 20 mM 2DG and are presented as violin plots. Sample sizes for each strain are between 265 and 315 cells. Under 2DG condition, ATP treatment did not restore ATP levels in wild-type cells, while NTT1 cells increased intracellular ATP levels (D) LC-MS/MS measurement of intracellular ATP changes in Wt and NTT1 expressing cells. A non-parametric Kruskal-Wallis test was used for comparison of groups for the QUEEN measurements. For the LC-MS/MS data pairwise t-test was used to compare the means of treated versus non-treated samples of each strain and determine if a statistically significant difference exists between them. Statistical significance between groups and raw QUEEN ratio as well as LC- MS/MS data are available in Supplementary File 1.
Transcriptomic analysis of Wt and NTT1 cells with and without ATP treatment.
(A) Heatmap showing hierarchical clustering of gene expression profiles from Wt and NTT1 cells cultured with ATP (red) and without ATP (green). Expression values are Z-score normalized, with red indicating higher expression and blue indicating lower expression. Three independent cultures for each group were used to isolate total RNA and perform RNA-seq analysis and each cell in the columns shows the mean value from three biological replicates. (B, C) Venn diagrams displaying the number of genes uniquely or commonly (B) upregulated or (C) downregulated in Wt and NTT1 cells treated with ATP compared to untreated controls. (D, E) Enrichment analysis showing significantly enriched gene ontology (GO) terms associated with genes (D) upregulated and (E) downregulated in NTT1 cells compared to their corresponding untreated controls. Circle size reflects the number of genes in each GO term, and color indicates significance level as represented by -log10 (FDR). The normalized counts and the complete list of DEGs with p values can be found in Supplementary File 2.
ATP treatment represses Hog1 activation through a mitochondria-dependent mechanism in NTT1 but not in Wt cells.
(A and B) The effect of 5 mM ATP treatment on Hog1 activation was assessed by Western blot analysis using two independent antibodies against phospho-Hog1 in two independent biological replicates (see Method). Pgk1 was used as an internal loading control. Uncropped membranes are provided in Fig. S8. (C) Densitometric quantification of Hog1 phosphorylation. Quantification of the immunoblots in Figure 4A, B. Hog1P signal is shown as background-subtracted integrated density (left) and normalized to total Hog1 (center) or to Pgk1 (right) in the same lane, each expressed relative to untreated wild-type cells. Open circles and triangles show the individual replicate values and red bars represent the rho0 isolates of each cell types.
Quantifications of ATP levels and replicative lifespans of the Wt strain and NTT1 strain under ATP treatment conditions.
Heatmaps show intracellular ATP abundance in (A) WT and (B) NTT1 cells, as measured by the inverse of QUEEN fluorescence at the 488nm excitation (see Method), throughout entire lifespans of single cells. Each row represents a color-coded time trace of a single yeast cell. Left - no ATP supplementation; Right - 10 µM ATP supplementation. Color bar represents the relative ATP abundance (e.g., red color indicates higher ATP) (C) Average of single-cell time traces of ATP dynamics aligned by percentage of lifespan. Error bars are standard errors. (D) Replicative lifespans of Wt and NTT1 cells, with or without ATP supplementation. Wt: w/ ATP vs w/o ATP (p = 1.28E-12); NTT1-expressing cells: w/ ATP vs w/o ATP (p = 4.03E-18). P-values were calculated with Mann-Whitney U test. The raw fluorescence measurements and lifespan data and number of cells analyzed in each experiments can be found in Supplementary File 3.
Replicative lifespan of different aging modes and mode distribution under ATP treatment conditions.
(A) Without ATP, the mode-2 lifespans between the Wt and NTT1 cells showed no significant difference (p = 0.74). On the other hand, NTT1 cells showed significantly shorter mode-1 lifespan in comparison to the Wt control (p = 2.39E-4). (B) With ATP, the NTT1 strain showed significantly increased mode-1 lifespan (p = 1.24E-16), but not mode-2 lifespan (p = 0.15), whereas the Wt strain showed significantly increased mode-1 lifespan (p = 3.17E-7) and also mode-2 lifespan (p = 5.98E-6), which was similar to mode-1 lifespan (p = 0.4). The mode-1 lifespans between the 2 strains were similar (p = 0.849). Figures indicate mean replicative lifespan (mRLS) values and number of individual mothers tested in each assay (n) were also included (C) Both strains show increased percentage of mode-1 cells under the effects of ATP. The p-values were calculated using Mann-Whitney U test. The lifespan data can be found in Supplementary File 3.
Replicative Lifespan analyses of rho0 Isolates.
Replicative lifespan (RLS) phenotypes of (A) parental controls and (B) their rho0 isolates were measured using a microfluidic chip under untreated and 10 mM ATP-treated (w/ATP) conditions. Elimination of mtDNA in rho0isolates alleviates the toxic effect of ATP treatment on NTT1 cells. Figures indicate mean replicative lifespan (mRLS) values and number of individual mothers tested in each assay (n) were also included. Raw lifespan data and statistical significance are provided in Supplementary File 3.
Proposed Models for the Effects of ATP Treatment in Wt and NTT1 Cells.
The type- 1 effect involves extracellular ATP sensing, analogous to mammalian purinergic signaling, which activates G-protein signaling and regulates the stress-associated MAPK cascade to modulate mitochondrial bioenergetic function and downstream transcriptional responses (left panel). In contrast, the type-2 effect is specific to NTT1 cells and is driven by increased intracellular ATP abundance resulting from Ntt1-mediated ATP import from the medium. In this model, elevated intracellular ATP inhibits catabolic processes, reprograms mitochondrial bioenergetic function, and, in parallel, modulates some of the same pathways associated with the type-1 ATP-sensing response (right panel). In the type-2 effect, inhibition of Hog1 activity depends on the respiratory status of mitochondria. Both type-1 and type-2 effects increase replicative lifespan (RLS), with the type-2 effect producing a greater lifespan extension.
Differential expression in NTT1-overexpressing cells.
(A) Volcano plot of log2 fold change vs −log10 nominal p-value for all expressed genes. Blue and red points denote genes passing p < 0.05 with negative or positive fold changes, respectively. Genes with the largest effect sizes or strongest statistical support are labelled. (B) Numbers of significantly down- and up-regulated genes (p < 0.05); the asymmetry (268 down vs 92 up) is significant by binomial test against an unbiased null. (C) Functional enrichment of NTT1 vs Wt differential expression. Categories with BH-FDR < 0.05 are shown, ranked by significance. Red, predominantly up- regulated; blue, predominantly down-regulated; circle size scales with fold enrichment. Numbers in parentheses are differentially expressed genes / total in category. Curated gene sets cover central metabolic, signaling and biogenesis pathways in S. cerevisiae.
Purine specificity of the NTT1 response.
(A) Log2 fold changes of de novo purine (ADE), BAS1-co-regulated (HIS4) and one-carbon/folate (MTD1, SHM2, GCV2, ADE3) genes that reached p < 0.05. Asterisks indicate p < 0.05. All thirteen captured genes are down-regulated. None of the de novo pyrimidine (URA) or ribonucleotide reductase (RNR) genes reached p < 0.05; the parallel pathways are transcriptionally unaffected. (B) The LC-MS/MS measurement of purine metabolites in Wt versus NTT1 cells. The raw LC-MS/MS data along with statistical values are available in Supplementary File 1.
Extracellular ATP treatment activates Gpa2–cAMP–PKA pathway signaling.
Log2 fold change (versus untreated) for Gpa2–cAMP–PKA pathway genes in Wt (solid) and NTT1 cells (hatched) after ATP treatment. Red, upregulated; blue, downregulated (padj < 0.05); n.d., not significant. Upregulation of negative regulators alongside repression of positive components and PKA targets indicates active PKA signaling in both strains.
Gene Ontology and KEGG Pathway Analyses for Wt Cells.
Gene ontology (GO) enrichment analysis for (A) upregulated and (B) downregulated genes, and KEGG pathway analysis for (C) upregulated and (D) downregulated genes in Wt cells after ATP treatment. GO categories include BP (Biological Process), CC (Cellular Component), and MF (Molecular Function).
Alteration of MAP Kinase Pathway in Wt Cells.
Genes uniquely upregulated (green) or downregulated (red) are shown for each component of the MAPK pathway in Wt cells under ATP treatment: (A) Cell Wall Stress, (B) High Osmolarity, (C) Pheromone, and (D) Starvation.
Gene Ontology and KEGG Pathway Analyses for NTT1 Cells.
Gene ontology (GO) enrichment analysis for (A) upregulated and (B) downregulated genes, and KEGG pathway analysis for (C) upregulated and (D) downregulated genes in NTT1 cells after ATP treatment compared to untreated controls. GO categories include BP (Biological Process), CC (Cellular Component), and MF (Molecular Function).
Snf1-Mediated Regulatory Gene Network in NTT1 Cells.
Genes regulated by Snf1 in NTT1 cells under ATP treatment are depicted in red (downregulated), green (upregulated), and black (no expression changes). The types of interactions are represented by arrows and lines.
