A microarray-based genetic screen for yeast chronological aging factors

Abstract

Model organisms have played an important role in the elucidation of multiple genes and cellular processes that regulate aging. In this study we utilized the budding yeast, Saccharomyces cerevisiae, in a large-scale screen for genes that function in the regulation of chronological lifespan, which is defined by the number of days that non-dividing cells remain viable. A pooled collection of viable haploid gene deletion mutants, each tagged with unique identifying DNA "bar-code" sequences was chronologically aged in liquid culture. Viable mutants in the aging population were selected at several time points and then detected using a microarray DNA hybridization technique that quantifies abundance of the barcode tags. Multiple short- and long-lived mutants were identified using this approach. Among the confirmed short-lived mutants were those defective for autophagy, indicating a key requirement for the recycling of cellular organelles in longevity. Defects in autophagy also prevented lifespan extension induced by limitation of amino acids in the growth media. Among the confirmed long-lived mutants were those defective in the highly conserved de novo purine biosynthesis pathway (the ADE genes), which ultimately produces IMP and AMP. Blocking this pathway extended lifespan to the same degree as calorie (glucose) restriction. A recently discovered cell-extrinsic mechanism of chronological aging involving acetic acid secretion and toxicity was suppressed in a long-lived ade4Delta mutant and exacerbated by a short-lived atg16Delta autophagy mutant. The identification of multiple novel effectors of yeast chronological lifespan will greatly aid in the elucidation of mechanisms that cells and organisms utilize in slowing down the aging process.

Figure 1. Microarray-based screen for chronological longevity factors.

(A) Schematic representation of the UP and DOWN tags flanking KanMX. Universal primer sequences (U1/U2 and D1/D2) flank the UP and DOWN tags. (B) Experimental flow of the screen. Aliquots were removed from SC cultures of the pooled YKO population at days 1, 9, 21, and 33, and spread onto YPD plates to allow growth of survivors. A typical BY4741 CLS time course in 2% glucose (NR) is shown. The TAGs were PCR amplified from genomic DNA and fluorescently labeled with either Cy5 (day 1) or Cy3 (days 9, 21, and 33). The day 1 TAGs were co-hybridized with day 9, 21, or 33 TAGs onto the microarray to generate the abundance ratio for each mutant at that particular day (D9/D1, D21/D1 and D33/D1). (C) Box plot of the mutant abundance ratios within the aging population at days 9, 21, and 33 from the NR medium. (D) Box plot showing the general increase in mutant viability within the aging population for CR medium compared to NR medium.

Figure 2. Deletion mutants that shorten CLS.

(A) Various deletions of autophagy genes isolated from the screen as short-lived were retested individually for CLS in NR and CR media. (B) Semi-quantitative CLS assay comparing a fet3Δ mutant to WT in NR and CR media. (C) Quantitative CLS assay for the same fet3Δ and WT strains in NR and CR media. Colony forming units (CFU) are plotted over time. (D) CLS assay showing CR-mediated extension of lifespan in ftr1Δ and fit3Δ mutants.

Figure 3. Lifespan-extending mutants include those that block de novo and salvage biosynthesis of purines.

(A) Examples of CLS assays for various long-lived mutants isolated from the screen, including genes related to purine metabolism (ade3Δ, ade4Δ, and fcy2Δ). (B) Schematic diagram of the de novo purine biosynthesis pathway and its connections to one-carbon metabolism and purine import/salvage pathways. Partially adapted from . (C) Examples of additional deletion mutants from the de novo purine biosynthesis, purine salvage, and one-carbon metabolism pathways that were not originally isolated from the screen.

Figure 4. Epistasis analysis of the de novo purine biosynthesis pathway in CLS.

(A) The lifespan extending ade4Δ mutation was combined with lifespan shortening ade17Δ and atg16Δ mutations through genetic crosses, and the double mutants tested for CLS when grown in SC 2% glucose (NR) medium. An ade16Δ ade17Δ double mutant that completely blocks the AICAR to FAICAR step of the de novo pathway was also tested. (B) CLS assays with WT, ade4Δ, fcy2Δ, tor1Δ, and gln3Δ strains grown in standard SC medium that contains (30 mg/L adenine), or SC medium supplemented with 4 times more adenine (4x Ade; 120 mg/L).

Figure 5. Growth media effects on CLS.

(A) Semi-quantitative CLS assay showing that deleting ADE4 extends CLS to the same degree as CR extends lifespan of a WT strain. The ade4Δ mutation and CR are not additive for CLS extension. (B) Quantitative CLS assay showing similarities between lifespan extension caused by the ade4Δ mutation and the CR growth condition. CFU  =  colony forming units. (C) SC medium with generally lower concentrations of amino acids (CPMB media) extends CLS when compared to the richer SC medium used in the genetic screen (Hopkins media). Deleting ATG16 (blocking autophagy) or FET3 (iron metabolism) prevented the extension of CLS induced by CPMB medium.

Figure 6. Cell-extrinsic effects of the atg16Δ and ade4Δ mutants on CLS.

(A) Schematic diagram of a reciprocal media swap experiment. WT and mutant cell cultures were grown to day 5 in standard SC media containing 2% glucose (NR), at which point the cells were pelleted. The media was removed, filtered, and then exchanged such that the cell pellets received expired media derived from the mutant (E-mutΔ) or WT (E-WT) strains. The cultures were then allowed to age and the standard CLS assay continued. (B) CLS assay of the media swap experiment. WT, ade4Δ and atg16Δ strains without the swapped media are included as controls. (C) pH measurements of the SC growth media over time in NR and CR conditions. WT, ade4Δ, ade17Δ, and atg16Δ mutants were tested. A four-fold excess of adenine was added to the WT and ade4Δ strains under the NR condition where indicated.

Figure 7. Effects of elevated pH on CLS.

(A) Standard SC cultures (2% glucose) that were either not pH adjusted (started at the default pH of ∼4.0) or pre-adjusted to 6.0 were inoculated with the indicated strains. At day 2 (48 hours), the pH of a subset (open symbols) of the originally untreated cultures was adjusted to a pH of 6.0. For each culture, the pH was measured at the start of incubation (day 0), and then days 1, 2, 5, and 9. (B) CLS assays showing the effect of raising the SC medium pH to 6.0 at either at inoculation (D0), or after 2 days growth (D2).

Figure 8. CLS mutants affect acetic acid secretion and resistance.

(A) Acetic acid concentrations in the filtered media from cultures of WT, ade4Δ, and atg16Δ strains that were growing in 2% glucose (NR) or 0.5% glucose (CR) conditions. Measurements were taken from log phase cultures, or cultures grown for 2 and 5 days. (B) Acetic acid measurements from WT and atg16Δ strains grown in unbuffered SC, or SC with the pH adjusted to 6 at the time of inoculation. (C) Resistance of the three strains to a 200 minute exposure of 300 mM acetic acid was measured for cells that were grown for 2 or 5 days in SC media (NR or CR levels of glucose). Percent survival for the treated cells is normalized to the untreated cells, which would be 100%.