Independent regulation of age-associated fat accumulation and longevity

The study addresses how age-associated lipid droplet (LD) accumulation is regulated and whether it impacts longevity. As organisms age, metabolic rates decline and neutral lipids accumulate in LDs. Prior work linked age-related metabolic decline to reduced NAD+ levels, prompting interventions to boost NAD+. However, such interventions do not consistently elevate NAD+ or extend lifespan. Alternative mechanisms modulating fat accumulation during aging have been reported in mammals and flies, but whether lipid accumulation itself reduces lifespan remains unclear. Here, the authors examine LD dynamics during replicative aging in Saccharomyces cerevisiae, test whether manipulating the NAD+ biosynthesis (kynurenine) pathway via BNA2 overexpression alters LD accumulation and lifespan, and dissect the metabolic routes (shikimate and aromatic amino acid biosynthesis, SA) upstream of BNA2 that might redirect glycolytic flux away from lipid synthesis. The purpose is to determine if LD accumulation and longevity are causally linked or independently regulated and to uncover metabolic strategies to modulate LDs in aging cells.

Background work indicates that metabolic decline with age often coincides with reduced NAD+ and increased fat deposition. Interventions supplying NAD+ precursors (niacin, NMN, NR) can improve aspects of metabolism and reduce fat but do not always increase NAD+ or extend lifespan. Additional mechanisms affecting age-related fat include DNA-PK activity promoting diet-induced weight gain in aging mice and HDAC6 overexpression suppressing ectopic LD accumulation in aging Drosophila muscles. Yeast LD biology and glycerolipid regulation are influenced by numerous genes, but the connection between NAD+ biosynthesis and LD regulation in aging yeast had not been established prior to this work.

• Model and aging: Saccharomyces cerevisiae replicative aging was assessed (number of divisions per mother cell). Aging mother cells were enriched using magnetic-bead labeling and MACS columns at defined time points corresponding to median ages 0 (young), 8, 16, and 23 divisions.
• LD quantification: LDs were visualized by fluorescence microscopy using BODIPY 493/503 staining and Erg6p-mCherry as a LD marker; quantified by flow cytometry (BD FACSCanto II). Neutral lipid versus phospholipid ratios were measured by GC-MS of FAMEs after lipid extraction and SPE fractionation.
• Genetic perturbations: A limited overexpression screen (280 high-copy genomic fragments) identified BNA2 as a suppressor of age-associated LD accumulation. Overexpression constructs (GPD promoter-driven BNA2 integrated at chromosome I) and homozygous gene deletions were generated. Deletions targeted core BNA pathway genes (BNA1, BNA2, BNA5, BNA6, BNA7) and the branch gene BNA3; ARO1 (rate-limiting in the shikimate pathway) was also deleted to probe upstream pathway dependency.
• Lifespan assays: Replicative lifespan (RLS) was measured by (i) traditional microdissection and (ii) a high-throughput modification of the Mother Enrichment Program (MEP) with β-estradiol induction, fluorescent labeling of mother cells, and bud scar counting across multiple biological replicates.
• NAD+/NADH measurement: Cellular NAD+ and NADH were extracted from young and aged cells and quantified via a colorimetric assay (Sigma MAK037), normalized per cell.
• Metabolomics: Global LC-MS metabolomics (C18 negative ion and HILIC positive ion modes; Thermo Q-Exactive Plus) quantified relative metabolite levels in middle-aged control vs BNA2-OE cells. Targeted LC-QTOF metabolomics (Agilent 6520 Q-TOF) quantified specific SA-BNA pathway intermediates (e.g., shikimate, chorismate, anthranilate, kynurenine, 3-HK, kynurenic acid, NAD+), normalized to cell counts. Standards were used for identification; data analyzed with Maven and Agilent MassHunter; statistics via GraphPad Prism.
• Cold stress regimen: Middle-aged cells (median age 16) were subjected to intermittent cold exposure (4 °C) vs maintained at 30 °C, and RLS was determined to assess the relationship between LD abundance and stress survival.
• Statistics: Two-way ANOVA for flow cytometry data; paired two-tailed t-tests for metabolite comparisons; log-rank tests for lifespan curves. Replicates and p-values are reported per figure panels.

• LDs increase with age: Yeast LDs accumulate markedly during replicative aging, increasing 7.2-fold from median age 0 to 23 by BODIPY quantification; biochemical NL:PL ratios corroborated increased neutral lipids.
• BNA2 overexpression (BNA2-OE) suppresses age-associated LD accumulation: Young BNA2-OE and control cells have similar LD levels, but old BNA2-OE cells show ~40% lower LDs versus controls (flow cytometry; multiple experiments).
• BNA2-OE extends lifespan without altering cellular NAD+: RLS increased by >15%. Microdissection: median 25 (max 55) in control vs median 34 (max 66) in BNA2-OE (p=0.0116). MEP: control median 25.2±0.3, max 41.6±0.6; BNA2-OE median 28.9±0.5, max 49.9±1.3 (p<0.0001). NAD+ and NADH per cell and NAD/NADH ratios showed minimal or no significant differences between BNA2-OE and control across ages.
• Core BNA genes are required for normal longevity but not for LD accumulation changes: Deletion of BNA1, BNA5, BNA6, or BNA7 shortened lifespan, while BNA3 deletion did not. Despite reduced lifespan, these deletions did not prevent normal age-associated LD accumulation. Thus, LD accumulation is not causally shortening lifespan.
• LD accumulation and lifespan are separable phenotypes: BNA2-OE reduced LD accumulation during aging even when downstream BNA genes were deleted (bna1Δ, bna3Δ, bna5Δ, bna6Δ, bna7Δ), indicating that NAD+ production per se is not necessary for LD suppression by BNA2-OE. For lifespan, BNA2-OE extended lifespan without BNA1, BNA3, or BNA5 but not without BNA6 or BNA7, linking specific downstream steps to longevity benefits.
• Dependence on upstream SA pathway: Deletion of ARO1 (shikimate pathway) blocked the LD suppression by BNA2-OE, implicating SA flux as essential for reducing LDs. Lifespan in aro1Δ was severely reduced, but BNA2-OE partially rescued lifespan above WT levels, though less than BNA2-OE alone.
• Metabolic rewiring by BNA2-OE: Global and targeted metabolomics showed BNA2-OE dramatically increased SA intermediates (shikimate, shikimate-3-P, chorismate ~16–60-fold) and moderately increased aromatic amino acids (tyrosine ~4-fold, phenylalanine ~1.9-fold, tryptophan ~1.47-fold). Early BNA pathway intermediates (kynurenine, 3-hydroxykynurenine) increased >170-fold. Branch metabolites kynurenic acid and xanthurenate increased ~350-fold and ~49-fold, respectively. Downstream metabolites (3-hydroxyanthranilate ~2.9-fold; quinolinate minimal change) and NAD+ showed minimal changes. In BNA2-OE bna6Δ, chorismate and anthranilate remained high; in BNA2-OE aro1Δ they were low/undetectable. SA metabolite levels inversely correlated with LD accumulation.
• LDs and stress resistance: Under cold exposure (4 °C), control cells showed slightly increased lifespan (median 27.5±1.3 vs 25.8±1.3 at 30 °C), BNA2-OE aro1Δ showed minimal change (25 vs 25.1), whereas BNA2-OE cells exhibited significantly decreased lifespan (median 24.8±1.3 vs 29.5±1.7 at 30 °C). As aged control and BNA2-OE aro1Δ cells have more LDs and BNA2-OE fewer, results suggest LD accumulation provides protection against cold stress.

The study demonstrates that age-associated LD accumulation in yeast is not inherently detrimental to lifespan and can be genetically and metabolically uncoupled from longevity. BNA2 overexpression reduces LD accumulation during aging by diverting glycolytic flux through the shikimate and aromatic amino acid (SA) pathway toward the kynurenine pathway, decreasing substrate availability for lipid synthesis likely downstream of pyruvate/acetyl-CoA. This metabolic rewiring depends on the upstream SA pathway (ARO1) for LD suppression but does not require completion of NAD+ biosynthesis, as LD reduction persists despite deletions of downstream BNA genes and occurs without increased cellular NAD+. Lifespan extension by BNA2-OE, however, requires specific downstream steps (BNA6/BNA7), indicating a separate mechanistic basis for longevity benefits possibly involving kynurenine branch metabolites (e.g., xanthurenate). The data further reveal a context-dependent role for LDs: while not shortening lifespan during normal aging, accumulated LDs enhance survival under cold stress, aligning with protective roles reported for LDs in other systems. Overall, these findings clarify the relationship between lipid storage and longevity, providing a model in which aging-associated shifts in glycolytic flux favor neutral lipid/LD formation, whereas BNA2-OE reorients flux to SA-BNA pathways to reduce LDs without compromising or necessarily enhancing NAD+ levels.

This work establishes that lipid droplet accumulation during normal yeast aging is not a driver of reduced lifespan and can be independently modulated from longevity. Overexpression of BNA2 reroutes metabolic flux upstream through the SA pathway, reducing LD accumulation while extending lifespan through downstream components of the kynurenine pathway, with minimal impact on NAD+ levels. The study provides a mechanistic model for metabolic rewiring of aging cells and suggests LDs confer stress resilience. Future research should (i) delineate the causal roles of specific kynurenine branch metabolites (e.g., kynurenic acid, xanthurenate) in lipid regulation and longevity, (ii) apply isotopic flux analyses to directly quantify substrate redistribution, (iii) expand to other stress paradigms and organisms to evaluate conservation, and (iv) map the broader regulatory network linking glycolysis, pentose phosphate, lipid synthesis, and SA-BNA pathways during aging.

• Model organism and aging paradigm: Findings are in budding yeast using replicative aging; generalizability to metazoans and to chronological aging requires validation.
• Overexpression-based perturbation: BNA2 overexpression may have non-physiological effects; dosage sensitivity and off-target consequences were not exhaustively tested.
• NAD+ measurement sensitivity: Colorimetric assays at select ages may miss compartment-specific or transient NAD+/NADH changes.
• Pathway resolution: While metabolomics indicates flux redirection, direct metabolic flux measurements (e.g., 13C tracing) were not performed.
• Stress scope: Protective effect of LDs was shown under cold stress; other stresses (oxidative, heat, ER stress) were not systematically assessed.
• Limited genetic screen: Initial screen was focused on known lipid/LD-related genes; broader unbiased screens could reveal additional regulators.