Andrus Gerontology Center, Department of Biological Sciences, Department of Molecular and Computational Biology, University of Southern California, Los Angeles
Wei, M., Madia, F., Longo, V. D. Studying Age-dependent Genomic Instability using the S. cerevisiae Chronological Lifespan Model. J. Vis. Exp. (55), e3030, doi:10.3791/3030 (2011).
Studies using the Saccharomyces cerevisiae aging model have uncovered life span regulatory pathways that are partially conserved in higher eukaryotes1-2. The simplicity and power of the yeast aging model can also be explored to study DNA damage and genome maintenance as well as their contributions to diseases during aging. Here, we describe a system to study age-dependent DNA mutations, including base substitutions, frame-shift mutations, gross chromosomal rearrangements, and homologous/homeologous recombination, as well as nuclear DNA repair activity by combining the yeast chronological life span with simple DNA damage and mutation assays. The methods described here should facilitate the identification of genes/pathways that regulate genomic instability and the mechanisms that underlie age-dependent DNA mutations and cancer in mammals.
Two lifespan models have being used to study aging in S. cerevisiae: RLS and CLS. The replicative (budding) life span (RLS) is based on the observation that yeast mother cells undergo a finite number of divisions3-6. We will focus on the chronological lifespan (CLS), a model based on the chronological survival of non-dividing yeast in culture or on plates7-11.Wild type yeast grows exponentially for 10-12 hours in synthetic dextrose complete (SDC) medium. When glucose level reduces, yeast switches from fermentation to respiration, a process termed diauxic shift. Cells continue to divide slowly during the post-diauxic phase for approximately 48 hours before entering G0 arrest. The metabolic rate of the cells remains high till day 5-6 (day 0 is the inoculation day). Cell viability over time can be accessed by the percentage of the chronologically aging cells, sampled every two days, which can exit G0 arrest and form colonies on rich YPED plates.
1. Yeast chronological lifespan (CLS) in liquid culture
Variations of in situ viability assay include:
2. in situ viability assay
In the liquid culture, a small fraction of the surviving cells may re-entered the cell cycle and grow utilizing remaining nutrients or those released form lysed dead cells in the medium, a phenotype termed regrowth/gasping13. We developed a viability-on-a-plate system that utilizes the auxotrophy of the DBY746 strain (trp1)to circumvent the regrowth/gasping problem and also allow the testing of the effect of constant exposure or deprivation of various external nutrients or stimuli on yeast CLS11. This in situ viability assay also mimics the replicative lifespan model such that cells are constantly exposed to abundant nutrients for the duration of the lifespan analysis.
Variations of in situ viability assay include:
3. DNA damage and mutation frequency during chronological aging
Canavanine resistance (Canr) and can1 sequencing
Spontaneous mutation frequency can be evaluated by measuring the frequency of canavanine resistance (Canr) in chronologically aging cultures. Mutations in the CAN1 (YEL063) gene, which encodes the plasma arginine permease, render cells resistant to the arginine analogue L-canavanine.Canr colonies collected at different time points can also be saved for extraction of genomic DNA and subsequent sequencing of the CAN1 gene, which can provide mutation spectrum data (with Mutation Surveyor, SoftGenetics).
Base substitutions (Trp+ reversions)
Strains with trp1-289 contain an amber mutation (C403T) in the TRP1 coding sequence. Measurement of frequency of trp1-289 to Trp+ reversion15 allows the estimation of base substitution rate during yeast chronological aging.
The Lys- strain EH150 (MATa, lys2ΔBglII, trp1-Δ, his3-Δ200, ura3-52, ade2-1o) harbors a lys2ΔBglII mutation that was constructed by inserting 4 nucleotides to create a BglII restriction enzyme site in the LYS2 gene. The resulting +4 shift in the open reading frame results in auxotrophy for lysine that can be reversed by small insertion/deletion mutations16-17.
Gross chromosomal rearrangements (GCRs)
To detect gross chromosomal rearrangements (GCRs), we generated a mutant strain, in which HXT13 (YEL069), encoding a highly redundant hexose transporter, was disrupted by a URA3 cassette18. HXT13 is located 7.5 kb telomeric to CAN1 on chromosome V. Mutations in both CAN1 and URA3 genes render cells resistance to L-canavanine and 5-fluoroorotic acid (5FOA), respectively. Considering the low frequency of point mutations that occur in both genes, analysis of the Canr 5FOAr frequency provides an estimation of GCRs that result in the loss of both genes.
Homologous and homeologous recombination
To monitor the level of homologous (100%) and homeologous recombination (91%) during chronological aging, we generated mutants in which linearized plasmids (HIS3::intron-IR-URA3) carrying either 100% homologous inverted repeats (IRs) (pSR406) or 91% homeologous IRs (pSR407) at the HIS3 locus19. Recombination between the IRs allows the expression of functional His3 protein.
Complementary to the in situ viability assay, age-dependent Trp+ reversion, Lys+ frame-shift mutation, recombination, or Canr can also be studied in cells aged on plate.
4. Translesion synthesis (TLS)
The age-dependent mutation frequency increase may involve increased macromolecule damage, diminished cellular protection/repair, and increased erroneous DNA repair (such as Polζ-dependent translesion synthesis, TLS) during aging. For example, the long-lived sch9Δ mutants exhibit elevated expression of SOD2, and decreased expression of the error-prone DNA repair enzyme Rev120. Here we describe an assay combining damaged DNA template and whole nuclear extract to evaluate the TLS in vitro.
5. Representative Results
We typically used the CFU counts on day 3 as 100 percent survival in the standard liquid CLS analysis. The percentage survival in the subsequent days can be fitted to calculate the mean (50% survival) and maximum (10% survival) life spans12. Life span results obtained from the in situ viability assay are generally consistent with those using the liquid CLS assay, but with reduced the mean life span, which is due in part to the fact that cells are constantly exposed to nutrients. Presence of glucose inhibits the activity of cellular protection as showed by the reduced transactivation of stress response transcription factor such as Msn2/4 and Gis112.
Age-dependent mutation frequency varies greatly depending on the strain background, genetic manipulation, culture conditions, and age. Table 2 shows the typical results obtained in the wild-type strain (DBY746).
|Nitrogen base (-AS/-AA)||1.8|
Table 1. Synthetic complete glucose medium, SDC (adjust to pH 6.0 with NaOH). 4-fold excess of histidine, leucine, tryptophan, and uracil are included to compensate the auxotrophy of the DBY746 strain.
|Day 3 (mean ± SEM)||Up to (during aging)|
|Canr||1.76 ± 0.12 x10-6||6-8 x10-6|
|Trp+ reversion||6.60 ± 1.70 x10-8||1.60 x10-6|
|Lys+ frame-shift mutation||3.00 ± 0.78 x10-8||0.70 x10-6|
|GCRs||0.62 ± 0.10 x10-8||0.30 x10-6|
|Homologous recombination||5.55 ± 2.74 x10-6||27.00 x10-6|
|Homeologous recombination||0.12 ± 0.04 x10-6||0.48 x10-6|
Table 2.Typical mutation frequencies of wild type cells (DBY746) during chronological aging.
Figure 1. Result of translesion synthesis (TLS)20. Nuclear extracts from 3 day-old stationary phase wild-type (DBY746) and sch9Δ mutant cells were incubated with undamaged or abasic site-containing DNA templates for 30 min at 30°C. TLS products are indicated by solid (with undamaged template) or dotted (with damaged template) lines. There was no translesion synthesis observed in nuclear extract from sch9Δ mutants. Free primers are indicated by the open arrow.
Liquid aging cultures sometime exhibit regrowth/gasping phenotype13, which complicates the CLS analysis. Regrowth typically occurs when more than 90-99% of the population has lost viability. This phenotype is often associated with increased oxidative stress and/or decreased protection in the cell. For example, the frequency of this phenotype more than doubles in cells lacking cytosolic superoxide dismutase and greatly reduces in long-lived mutants (e.g., sch9Δ or ras2Δ) or mutants overexpressing superoxide dismutases22. In practice, we define regrowth as an increase in viability or stabilization in viability for 3 consecutive samplings in the high mortality phase during chronological aging.
Age-dependent frequencies of different DNA mutations vary greatly depending on the strain background, genetic manipulation, culture conditions, as well as regrowth/gasping and extremely low survival. Pre-experiments should be carried out at time points such as the mean and maximum survival with various plating density for mutation assays to determine mutation frequency range before performingthe full scale life span analysis. The wild type strain should always be included in a life span or mutation frequency study in parallel to any treatment or genetic mutant, such that inter-experiment-variations can be accounted for. Multiple biological replicas should be included in the study; and, both liquid and in situ viability/mutation assays should be carried out to corroborate the results.
Instead of focusing on one specific type of DNA mutation, profiling of age-dependent genomic instability using multiple assays in combination, may shed light on specific DNA damage and DNA damage repair system(s) that contribute to age-dependant genomic instability. For example, a significant increase in gross chromosomal rearrangements (GCRs), compared to those of other DNA mutations, is observed in wild type yeast suggesting an elevationof double strand break and/or an impairment in non-homologous end joining (NHEJ) during yeast chronological aging (Table 2). The can1 mutation spectrum (sequencing of the CAN1 gene) obtained in the wild-type aging yeast suggested an increase in oxidative damage during chronological aging and error-prone DNA repair; whereas, in long-lived sch9Δ mutants, less oxidative DNA damage and much reduced error-prone translesion synthesis were observed20.
Methods described here can be further expended to study age-dependant genomic instability. For example, quiescent and non-quiescent cells can be isolated from yeast stationary-phase cultures using the density gradient method described by Allen et al.23. Combined with the mutation assays described here, we have reported that a large portion of age-dependent mutations arises from quiescent cells, rather than the dividing, damaged or apoptotic cells20,24.
No conflicts of interest declared.
We thank S. Jinks Robertson and E. Heidenreich for providing plasmids and yeast strains; P. Pham and M.F. Goodman for help withthe TLS assay. This work was supported, in part, by the American Federation for Aging Research grant and by NIH AG20642, AG025135.