Decoupling AMPK from fatty acid synthesis allows maintenance of fitness late in life
- Epigenetics Programme, Babraham Institute, United Kingdom
Figures
Figure 1 with 1 supplement
AMPK suppresses senescence through mitochondrial acetyl coenzyme A (Acetyl-CoA) import pathways.
(A) Experimental scheme for analysis of replicatively aged yeast. (B) Impact of SAK1 overexpression on Tom70-GFP (marker of senescence onset) and WGA (marker of replicative age). p-values calculated by two-way ANOVA, n = 7. (C) Separation of wild type and PGPD-SAK1 low senescence (high WGA low Tom70-GFP) and high senescence (low WGA high Tom70-GFP) populations by flow cytometry, and quantification of low senescence population at 48 hr. p-value calculated by t-test, n = 7. (D) Comparison of gene expression profiles of wild type and PGPD-SAK1 cells at log phase and after 48 hr ageing by mRNA-seq. Data is an average of 3 biological replicates (log) or 4 biological replicates (48 hr), significantly different genes (p < 0.01 by DEseq2) are shown in purple, genes both substantially (>3x indicated by dotted lines) and significantly different are highlighted in green. Substantially and significantly upregulated genes are annotated. (E) Simple schematic of the relationship between AMPK, mitochondria, and glycolysis pathways. Central glucose processing by glycolysis leading to fermentation is shown by blue arrows, AMPK activated pathways as purple arrows. Key enzymes in this study are shown in red. (F) Representative flow plots and quantification of the low senescence population at 48 hr in wild type and PGPD-SAK1 mutants lacking the carnitine shuttle (cat2Δ) and mutants lacking both the glyoxylate cycle (mls1Δ) and cat2Δ. p-values calculated by one-way ANOVA, n = 3–4.
Figure 1—figure supplement 1
Increased AMPK activity suppresses senescence in a subset of cells.
(A) mRNA expression level of constitutive overexpression of the upstream kinase Sak1 using a PGPD promoter fused to SAK1 at log phase, 24, and 48 hr. p-values calculated by one-way ANOVA, n = 3. (B) Impact of SAK1 overexpression on Rpl13a-mCherry in young and aged cells. p-values calculated by two-way ANOVA, n = 7. (C) Representative flow plots and quantification of the low senescence population at 48 hr in wild type and PGPD-SAK1 mutants lacking the glyoxylate cycle (mls1Δ) or succinate import (sfc1Δ). p-values calculated by one-way ANOVA, n = 4–6. (D) Representative flow plots and quantification of wild type and PGPD-SAK1 low senescence populations at 48 hr in mitochondrial mutants lacking a functional electron transport chain (cox9Δ) or pyruvate import (mpc1Δ). p-values calculated by one-way ANOVA, n = 4–5.
Figure 2 with 1 supplement
Combining AMPK activity and fatty acid synthesis to avoid senescence.
(A) Schematic highlighting the regulation of fatty acid synthesis by AMPK. Key proteins for this figure are highlighted in red. (B) Representative flow plots and quantification of the low senescence population in acc1S1157A, PGPD-SAK1 and A2A at 48 hr. p-values calculated by one-way ANOVA, n = 8–9. (C) Population medians for Tom70-GFP and WGA in acc1S1157A, PGPD-SAK1 and A2A at log phase, 24 and 48 hr. p-values calculated by two-way ANOVA, n = 7–9. (D) Representative flow plots of wild type and A2A at matched average replicative age (48 hr for wild type versus 40 hr for A2A), with quantification of the low senescence population and of median WGA (bud scars). p-values calculated by one-way ANOVA, n = 4–5. (E) Percentage cell viability in cultures aged for 40 and 48 hr. p-value calculated by t-test, n = 4. (F) Size of colonies formed in 24 hr on a YPD plate by log phase and age-matched wild type and A2A. Only cells that eventually formed colonies within 3 days were included to ensure all tested cells were viable. p-values calculated by two-way ANOVA, n = 14–39. (G) Size of colonies formed in 24 hr on a YPD plate by log phase and 48-hr-aged wild type, acc1S1157A, PGPD-SAK1 and A2A. Only cells that eventually formed colonies within 3 days were included to ensure all tested cells were viable. p-values calculated by two-way ANOVA, n = 15–61.
Figure 2—figure supplement 1
Combining AMPK activity and fatty acid synthesis to avoid senescence.
(A) Growth curve of wild type and A2A cells in YPD media starting from log-phase pre-cultures. (B) Lifespan measured in YPD media for wild type and A2A cells based on %age of viable cells remaining at each time point; n = 5. (C) Population medians for Rpl13a-mCherry in wild type, acc1S1157A, PGPD-SAK1, and A2A at log phase, 24 and 48 hr. p-values calculated by two-way ANOVA, n = 7–10. (D) Population medians for Tom70-GFP and WGA (bud scars) in wild type, PGPD-SAK1, and A2A at log phase, 24 and 48 hr made from a compilation of experiments performed during this study. p-values calculated by two-way ANOVA, n = 27–38. (E) Manual bud scar counts of log phase and 48-hr-aged wild type and A2A cells. p-values calculated by one-way ANOVA, n = 14–20. (F) Southern blot analysis of extrachromosomal ribosomal DNA circles (ERCs) compared to genomic rDNA in wild type and A2A cells at log, 24, and 48 hr. (G) Analysis of median change in log2 mRNA abundance from log phase to given time points for all genes on ChrXIIr. Analysis by two-way ANOVA for aged time points only since log values are the reference point n = 2–4 biological replicates.
-
Figure 2—figure supplement 1—source data 1
Annotated pdf of full blot for Figure 2—figure supplement 1.
- https://cdn.elifesciences.org/articles/111611/elife-111611-fig2-figsupp1-data1-v1.zip
-
Figure 2—figure supplement 1—source data 2
Original tif format image file of blot in Figure 2—figure supplement 1.
- https://cdn.elifesciences.org/articles/111611/elife-111611-fig2-figsupp1-data2-v1.zip
Figure 3 with 1 supplement
Separate classes of A2A cells depend on respiration and fatty acid availability.
(A) Schematic of acetyl coenzyme A biosynthesis and the emergence of a negative feedback loop through AMPK regulation by Sip2. Key proteins for this figure are highlighted in red. (B) Western blot analysis of acetyl lysine on H3 in log phase wild type, acc1S1157A, PGPD-SAK1, and A2A cells. Membranes were probed with mouse anti-H3 and rabbit anti-pan-acetyl lysine using a two-colour detection system. p-values calculated by one-way ANOVA, n = 3–4. (C) Representative flow plots and quantification of low senescence population in sip2Δ and sip2Δ PGPD-SAK1 at 48 hr. p-values calculated by t-test, n = 5. (D) Representative flow plots and quantification of low senescence population in wild type and ald6Δ at 48 hr. Population medians are also shown for Tom70-GFP and WGA in wild type and ald6Δ at 48 hr. p-values calculated by t-test, n = 5–6. (E) Size of colonies formed in 24 hr on a YPD plate by log phase and 48-hr-aged wild type and ald6Δ. Only cells that eventually formed colonies within 3 days were included to ensure all tested cells were viable. p-values calculated by one-way ANOVA, n = 36–57. (F) Representative flow plots and quantification of low senescence population in A2A mutants lacking both carnitine acetyltransferase activity (cat2Δ) and the glyoxylate cycle (mls1Δ) at 48 hr. Arrows indicate two populations, p-value calculated by t-test, n = 4. (G) Size of colonies formed in 24 hr on a YPD plate by log phase and 40-hr-aged A2A and A2A cat2Δ mls1Δ. Only cells that eventually formed colonies within 3 days were included to ensure all tested cells were viable. Arrows indicate two populations, p-values calculated by two-way ANOVA, n = 29–55.
Figure 3—figure supplement 1
Genetic evidence that cytosolic acetyl coenzyme A (Acetyl-CoA) promotes senescence.
(A) Representative flow plots of the low senescence population at 48 hr in sip23R, sip23R PGPD-SAK1, and sip23R A2A. (B) Representative flow plots and quantification of low senescence population at 48 hr in A2A mutants lacking the electron transport chain (cox9Δ), p-value calculated by t-test, n = 3–4. (C) Representative flow plots and quantification of low senescence population in wild type and sip2Δ PGPD-SAK1 ald6Δ at 48 hr. Population medians are also shown for Tom70-GFP and WGA in wild type and ald6Δ at 48 hr. p-values calculated by t-test, n = 5–6. (D) Basal oxygen consumption rates for log and aged cells. p-values calculated by one-way ANOVA, n = 6–8. Note that due to unavoidable differences in cell number, the values for log and 48 hr samples are not comparable. (E) MA plots comparing change in normalised mRNA abundance between log and 48 hr to average mRNA abundance between log and 48 hr. Data represents an average of 3–4 biological replicate RNA-seq experiments. Line of best fit was calculated by linear regression, with the slope parameter quoted as a measure of gene expression dysregulation (Horkai et al., 2023). Transcripts from the rDNA intergenic spacer are highlighted in blue. (F) Comparison of gene expression profiles of PGPD-SAK1 and A2A cells at 48 hr ageing by mRNA-seq. Data is an average of 4 biological replicates, there are no significantly different genes (p < 0.05 by DEseq2). Threefold differences are indicated by dotted lines, only uncharacterised genes pass this threshold and are not significant.
Figure 4 with 1 supplement
Fatty acids are particularly important in ageing cells that do not respond to AMPK activation.
(A) Schematic of potential mechanisms for the beneficial effect of ACC1 activity: Inhibition of mTOR mediated by malonyl coenzyme A, the direct product of Acc1, or efficient synthesis of fatty acids, compared to a null hypothesis that Acc1 acts to remove cytosolic acetyl coenzyme A (Acetyl-CoA), for example if these cells are respiratory-deficient. (B) Representative flow plots and quantification of low senescence population at 48 hr in sip2Δ PGPD-SAK1 acc1S1157A and sip2Δ PGPD-SAK1 supplemented with 2.5 µM cerulenin (condition 3), 10 µM cerulenin + 0.04% oleic acid + 0.04% Tween 80 (condition 4) or 0.04% oleic acid + 0.04% Tween 80 (condition 5). p-values calculated by one-way ANOVA, n = 2–3. (C) Representative flow plots and quantification of low senescence population at 48 hr in wild type, PGPD-SAK1, and A2A supplemented with 0.04% oleic acid + 0.04% Tween 80. p-values calculated by one-way ANOVA, n = 4. (D) Size of colonies formed in 24 hr on a YPD plate by log phase and 48-hr-aged wild type, acc1S1157A, PGPD-SAK1, and A2A supplemented with 0.04% oleic acid + 0.04% Tween 80. Only cells that eventually formed colonies within 3 days were included to ensure all tested cells were viable. p-values calculated by one-way ANOVA, n = 19–52.
Figure 4—figure supplement 1
Fatty acids in senescence.
(A) Population medians for Tom70-GFP and WGA at 48 hr in sip2Δ PGPD-SAK1 acc1S1157A and sip2Δ PGPD-SAK1 supplemented with cerulenin and/or 0.04% oleic acid + 0.04% Tween 80. p-values calculated by one-way ANOVA, n = 2–3. (B) Manual bud scar counts of 48-hr-aged sip2Δ PGPD-SAK1 cells with and without 0.04% oleic acid. p-values calculated by one-way ANOVA, n = 20. (C) Population medians for Tom70-GFP and WGA at 48 hr in wild type, PGPD-SAK1, and A2A with and without 0.04% oleic acid + 0.04% Tween 80. p-values calculated by one-way ANOVA, n = 4. (D) Size of colonies formed in 24 hr on a YPD plate by log phase and 48-hr-aged ald6Δ cells with 0.04% oleic acid + 0.04% Tween 80. Only cells that eventually formed colonies within 3 days were included to ensure all tested cells were viable. p-values calculated by one-way ANOVA, n = 28–53. (E) MA plot comparing change in normalised mRNA abundance between log and 48 hr to average mRNA abundance between log and 48 hr. Data represents an average of 3–4 biological replicate RNA-seq experiments. Line of best fit was calculated by linear regression, with the slope parameter quoted as a measure of gene expression dysregulation (Horkai et al., 2023). Transcripts from the rDNA intergenic spacer are highlighted in blue. (F) Comparison of gene expression profiles of wild type cells after 48 hr ageing with or without 0.04% oleic acid + 0.04% Tween 80 by mRNA-seq. Data is an average of 4 biological replicates (untreated) or 3 biological replicates (oleic acid), significantly different genes (p < 0.01 by DEseq2) are shown in purple, genes both substantially (>3x indicated by dotted lines) and significantly different are highlighted in green. (G) Analysis of median change in log2 mRNA abundance from log phase to 48 hr for all genes on ChrXIIr. Analysis by t-test for aged time points only since log values are the reference point n = 3–4 biological replicates.
Figure 5
Model of different classes of ageing mediated by acetyl coenzyme A (Acetyl-CoA) and lipid metabolism.
Schematic demonstrating the two wild-type senescence fates. (i) Adaptation to future scarcity by synthesising excess cytosolic Acetyl-CoA to store as much energy as possible in lipids against future scarcity, and (ii) Adaptation to rapid growth by maximising growth from the current food supply by synthesising the bare minimum of cytosolic Acetyl-CoA.
Additional files
-
Supplementary file 1
Strains used in this work.
- https://cdn.elifesciences.org/articles/111611/elife-111611-supp1-v1.docx
-
Supplementary file 2
Oligonucleotides used for making strains.
- https://cdn.elifesciences.org/articles/111611/elife-111611-supp2-v1.docx
-
MDAR checklist
- https://cdn.elifesciences.org/articles/111611/elife-111611-mdarchecklist1-v1.docx
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Decoupling AMPK from fatty acid synthesis allows maintenance of fitness late in life
eLife 15:RP111611.
https://doi.org/10.7554/eLife.111611.3
