TY - JOUR
AU - Zebrowski, Jacek
AB - Abstract Yeast ageing has been gaining much attention in gerontology research, yet the process itself is still not entirely clear. One of the constraints related to the use of the Saccharomyces cerevisiae yeast in studies is the ambiguity of the results concerning ageing determinants for different genetic backgrounds. In this paper, we compare reproductive potentials and lifespans of seven widely used haploid laboratory strains differing in daughter cells production capabilities and highlight the importance of choosing an appropriate genotype for the studies on ageing. Moreover, we show here links between post-reproductive lifespan and lipid metabolism, as well as between reproductive potential, reproductive lifespan and phylogenetic relationship. Using FTIR spectroscopy that generated a biochemical fingerprint of cells, coupled with chemometrics, we found that the band of carbonyl (C = O) stretching vibration discriminates the strains according to post-reproductive lifespan. The results indicated that prolonged post-reproductive lifespan was associated with relatively lower amount of fatty acids esterified to phospholipids compared to a free acid pool, thus implying phospholipid metabolism for the post-reproductive lifespan of yeast. In addition, phylogenetic analysis showed a correlation between nucleotide similarity and the reproductive potential or reproductive lifespan, but not to the longevity expressed in time units. ageing, Saccharomyces cerevisiae, lifespan, longevity, FTIR spectroscopy, phylogenetics INTRODUCTION The unicellular Saccharomyces cerevisiae yeast has been used extensively as a simple model for studying fundamental phenomena occurring in eukaryotic cells. The main advantages of the organism include a small genome (12 Mbp), a short doubling time (80–90 min on rich medium), ease of genetic modifications and relatively high stability of haploid and diploid phases (Stansfield and Stark 2007). The yeast genome was completely sequenced in 1996 as the first among eukaryotic organisms (Goffeau et al.1996). A large number of genes have been found to encode similar proteins in yeast as well as in higher organisms such as fruit fly, worm or mouse, and similar metabolic pathways have been identified in these species (Longo, Mitteldorf and Skulachev 2005; Rego et al.2014). However, not every eukaryotic cellular activity can be modelled in yeast, and this constraint may concern some aspects of cellular ageing (Gershon and Gershon 2000). Achieving longevity and slowing down the ageing process have always been hot topics in the history of mankind. In general, ageing is defined by most evolutionary biologist as an age-dependent decline of physiological functions leading to a decrease in the survival rate and in the age-specific reproductive rate (Reznick et al.2004). Among various eukaryotic organisms used in the studies of ageing (e.g. yeast, worms, fruit flies, rats, mice or monkeys), the S. cerevisiae yeast often serves as a primary model for the understanding of genetic and biochemical determinants of the ageing phenomenon. Similar to many other organisms, yeast has two mechanisms of ageing: public and private (Martin, Austad and Johnson 1996). In respect of public mechanisms, highly conserved genes and metabolic pathways exist in a whole group of organisms, e.g. calorie restriction (Rona et al.2015), sirtuins (Tissenbaum and Guarente 2001; Burnett et al.2011; Kanfi et al.2012), S6 kinase (Fabrizio et al.2004; Pan et al.2007; Selman et al.2009). Private mechanisms, such as e.g. extrachromosomal rDNA circles (ERCs) formation, are linked only to yeast (Sinclair and Guarente 1997). In the case of S. cerevisiae, both the budding (reproductive) and non-budding (non-reproductive) state cells can be taken into account for ageing analysis. Several models of ageing have been proposed, among which the replicative and chronological models have been the most extensively explored ones. Originally, replicative lifespan was defined as the ability of a mother cell to produce daughters during its life before replicative ageing, and has been proposed as a model for dividing cells of higher eukaryotes, including mammals (Polymenis and Kennedy 2012). In turn, chronological lifespan was defined as the length of time during which a yeast cell can survive in the non-budding state. Chronological ageing (see Longo et al.2012; Piper 2012, for reviews) has been proposed as a model for post-mitotic cells of higher eukaryotes (Fabrizio and Longo 2003) Two other models of ageing, although much less widely applied, have been also used: hibernating lifespan, which is a special case of chronological lifespan (Postma, Lehrach and Ralser 2009), and colony lifespan (Vachova, Cap and Palkova 2012). In this paper, we focus on replicative ageing in S. cerevisiae, which has been extensively reviewed elsewhere (Lippuner, Julou and Barral 2014). Years of studies of yeast brought about several hypotheses and pointed out to a number of factors influencing replicative ageing. The most talked-about factors are associated with ERCs formation (Sinclair and Guarente 1997), oxidation macromolecules including proteins (Aguilaniu et al.2003; Erjavec et al.2007) and DNA damage (Sinclair, Mills and Guarente 1997). Cell organelles are also affected by changes associated with ageing. In general, the physiology of the entire cell undergoes significant changes that primarily affect the vacuole (Tang et al.2008), mitochondrion (Veatch et al.2009), nucleus (Dang et al.2009) and cytoplasm (Zhou et al.2011). Possible relationships between ageing and cell biochemistry or metabolic processes have not been extensively studied in yeast so far. Previously reported metabolome analysis of S. cerevisiae isogenic mutants differing in terms of longevity showed an association between lifespan and the mass spectroscopy (MS)-based fingerprint (Yoshida et al.2010). Very recent studies emphasised the role of lipid metabolism in chronological longevity (Handee et al.2016). In the light of these promising findings, we have searched here for possible associations between the selected longevity parameters and the metabolic fingerprint of cells generated by the Fourier-transform infrared (FTIR) spectroscopy. This high-throughput and powerful technique combined with chemometric tools has been successfully applied to comprehensive biochemical characterisation of yeast cells and establishing metabolic relationships (Adt et al.2010; Correa-Garcia et al.2014). We explored that approach for a set of laboratory strains differing in the capability of daughter cells production, which are widely used in laboratories around the world and considered as references in studies of ageing. In addition, the S. cerevisiae populations were analysed for the reproductive potential and longevity parameters expressed in units of time. We also determined possible phylogenetic links with the reproductive potential. MATERIALS AND METHODS Yeast strains All yeast strains used in this study are listed in Table 1. Table 1. Laboratory S. cerevisiae strains used in the study. Strain  Genotype  Source  BY4741  MATa his3Δ 1; leu2Δ 0; met15Δ 0; ura3Δ 0  EUROSCARF  BMA64–1A  MATa ura3–1; trp1Δ 2; leu2–3112; his3–11,15; ade2–1; can1–100  EUROSCARF  FY1679–01B  MATa ura3–52; LEU2; TRP1; HIS3; GAL2  EUROSCARF  FY1679–05A  MATa ura3–52; leu2Δ 1; trp1Δ 63; HIS3; GAL2  EUROSCARF  CEN.PK2–1C  MATa ura3–52; trp1–289; leu2–3112; his3Δ 1; MAL2–8C; SUC2  EUROSCARF  SY991  MATa ura3Δ 0; his3Δ 1; leu2Δ 0; trp1Δ 63; ade2Δ 0; lys2Δ 0; ADE8  EUROSCARF  SP-4  MATα leu1 arg4  Bilinski et al. (1985)  Strain  Genotype  Source  BY4741  MATa his3Δ 1; leu2Δ 0; met15Δ 0; ura3Δ 0  EUROSCARF  BMA64–1A  MATa ura3–1; trp1Δ 2; leu2–3112; his3–11,15; ade2–1; can1–100  EUROSCARF  FY1679–01B  MATa ura3–52; LEU2; TRP1; HIS3; GAL2  EUROSCARF  FY1679–05A  MATa ura3–52; leu2Δ 1; trp1Δ 63; HIS3; GAL2  EUROSCARF  CEN.PK2–1C  MATa ura3–52; trp1–289; leu2–3112; his3Δ 1; MAL2–8C; SUC2  EUROSCARF  SY991  MATa ura3Δ 0; his3Δ 1; leu2Δ 0; trp1Δ 63; ade2Δ 0; lys2Δ 0; ADE8  EUROSCARF  SP-4  MATα leu1 arg4  Bilinski et al. (1985)  View Large Table 1. Laboratory S. cerevisiae strains used in the study. Strain  Genotype  Source  BY4741  MATa his3Δ 1; leu2Δ 0; met15Δ 0; ura3Δ 0  EUROSCARF  BMA64–1A  MATa ura3–1; trp1Δ 2; leu2–3112; his3–11,15; ade2–1; can1–100  EUROSCARF  FY1679–01B  MATa ura3–52; LEU2; TRP1; HIS3; GAL2  EUROSCARF  FY1679–05A  MATa ura3–52; leu2Δ 1; trp1Δ 63; HIS3; GAL2  EUROSCARF  CEN.PK2–1C  MATa ura3–52; trp1–289; leu2–3112; his3Δ 1; MAL2–8C; SUC2  EUROSCARF  SY991  MATa ura3Δ 0; his3Δ 1; leu2Δ 0; trp1Δ 63; ade2Δ 0; lys2Δ 0; ADE8  EUROSCARF  SP-4  MATα leu1 arg4  Bilinski et al. (1985)  Strain  Genotype  Source  BY4741  MATa his3Δ 1; leu2Δ 0; met15Δ 0; ura3Δ 0  EUROSCARF  BMA64–1A  MATa ura3–1; trp1Δ 2; leu2–3112; his3–11,15; ade2–1; can1–100  EUROSCARF  FY1679–01B  MATa ura3–52; LEU2; TRP1; HIS3; GAL2  EUROSCARF  FY1679–05A  MATa ura3–52; leu2Δ 1; trp1Δ 63; HIS3; GAL2  EUROSCARF  CEN.PK2–1C  MATa ura3–52; trp1–289; leu2–3112; his3Δ 1; MAL2–8C; SUC2  EUROSCARF  SY991  MATa ura3Δ 0; his3Δ 1; leu2Δ 0; trp1Δ 63; ade2Δ 0; lys2Δ 0; ADE8  EUROSCARF  SP-4  MATα leu1 arg4  Bilinski et al. (1985)  View Large Media and growth conditions Yeast was grown in a standard rich medium YPD (1% yeast extract, 1% yeast Bacto-peptone, 2% glucose) on a rotary shaker at 150 rpm at the temperature of 28°C. Determination of reproductive potential Reproductive potential of individual mother yeast cells was defined as the number of mitotic cycles during the cell's life. After overnight growth, cells were arrayed on a YPD plate using a micromanipulator. Reproductive potential was determined microscopically by a routine procedure with the use of a micromanipulator. The number of buds formed by each mother cell signifies its reproductive potential. In each experiment, 45 single cells were analysed. The results represent measurements for at least 90 cells analysed in two independent experiments. Determination of mean doubling time Mean doubling time was calculated individually for each of the analysed cells during routine determination of reproductive potential as describe previously (Molon and Zadrag-Tecza 2016b). The results represent measurements for at least 90 cells analysed in two independent experiments. Determination of total lifespan Total lifespan was defined as the length of life of a single mother cell expressed in units of time. Total lifespan was calculated as the sum of reproductive and post-reproductive lifespans (PRLS; Zadrag, Bartosz and Bilinski 2008). Reproductive lifespan was defined as the length of time between the first and the last budding, and PRLS as the length of time from the last budding until cell death. The lifespan of the Saccharomyces cerevisiae yeast was determined as previously described by Minois et al. (2005) with small modifications by Zadrag, Bartosz and Bilinski (2008). Ten microlitre aliquots of an overnight-grown culture of yeast were collected and transferred on YPD plates with solid medium containing Phloxine B (10 μg/ml). Phloxine B was used to stain S. cerevisiae dead cells. Dead yeast cells lose membrane integrity and Phloxine B entered cell space giving pink/red colouration of cytosol. In each experiment, 45 single cells were analysed. During manipulation, the plates were kept at 28°C for 15 h and at 4°C during the night. The results represent measurements for at least 90 cells analysed in two independent experiments. DNA extraction and amplification of ITS1-5,8S-ITS2 DNA was extracted from 5 ml yeast liquid cultures according to Yeast DNA Extraction Kit (Thermo Fisher Scientific, Wilmington, DE, USA) methods. Amplifications of the target sequences were carried out in a 40 μl volume containing 1–10 ng genomic DNA, 100 nM of each primer, 200 pM of each of the four dNTPs, 1 U Phusion DNA polymerase (Thermo Scientific, Waltham, MA, USA) and the appropriate buffer supplied by the manufacturer. The primers used to amplify the ITS1–5,8S-ITS2 sequences were ITS1: TCCGTAGGTGAACCTGCGG and ITS4: TCCTCCGCTTATTGATATGC (White et al.1990). Sequencing and data analysis The ITS sequences were amplified under the PCR conditions described above. The amplified products were purified and used for sequencing. The sequencing procedures were performed by Genomed S.A. CLC Genomics Workbench 7 was used for data analysis. unweighted pair group method with arithmetic mean (UPGMA) and neighbour joining (NJ) were used for phylogenetic tree analysis. FTIR spectroscopy For the FTIR spectroscopy analysis, yeast cell culture was collected at the logarithmic growth stage and subsequently centrifuged at 3000 × g for 5 min., washed out twice with Mili-Q water and dried at 45°C for 24 h. Immediately before the measurements, the cell material was homogenised into powder using an agate mortar and pestle. Mid-infrared spectra were acquired by the iZ10 FTIR module of Nicolet iN10 MX infrared microscope (Thermo Scientific, USA) equipped with a deuterated triglycine sulphate detector and KBr beam splitter. The measurements were performed in attenuated total reflectance (ATR) mode using ATR accessory (Smart Orbit, Thermo Scientific) equipped with a single-bounce diamond crystal. A pinch of cell powder was deposited onto the crystal and pressured against its surface with attached pressure clamp. Sixty four interferograms were collected and co-added before Fourier transformation within the wavelength range between 4000 and 400 cm−1 at the 4 cm−1 resolution. The ATR crystal was carefully cleaned before each measurement with ethanol to remove any residual traces of the previous sample. The spectra were ATR, air vapour and baseline corrected and normalised to the amide I peak height as the reference using the OMNIC (v. 9, Thermo Fischer Scientific Inc.) software. Five independent experimental series with two technical replications were averaged and subjected to unsupervised multivariate analysis. The principal component analysis (PCA) was performed to obtain a metabolic snapshot of particular yeast strains using the Unscrambler X (v.10.1, CAMO, Norway) software. Statistical analysis Data were examined for normality using the Shapiro-Wilk test and Normal Quantile-Quantile (Q-Q) plot (Logan 2011). If the null hypothesis of normal distribution was rejected, we used box-plots and Kruskal-Wallis test for non-parametric comparison of the strains. The multiple pairwise Dunn's test (Dunn 1964) was applied to estimate which strains differed between each other. Time-related lifespan characteristics were analysed using a survival analysis (Clark et al.2003). The Benjamini-Hochberg procedure (Benjamini and Hochberg, 1995) was used to correct p-value for the multiple comparisons. A Spearman's rank correlation test was run to determine relationship between the examined data. The events in the survival analysis were defined as the last budding (reproductive lifespan) or the cell death (PRLS, total lifespan). In the case of PRLS, the survival time was counted from the last budding. The uncensored Kaplan-Meier method was used to determine the survival function. The log-rank test (Peto et al.1977) was applied with the same weight to all events to assess differences between the survival curves. A significance level of 0.05 was assumed for the statistical hypothesis testing and the pairwise group comparison. The statistical computing and majority of data presentation was performed using R programming (R Core Team 2013) and special R packages (Dinno 2016; Therneau 2017). RESULTS AND DISCUSSION Yeast ageing is a process influenced by a number of factors, such as genes mutation, organelles dysfunction or the effect of extracellular environment (Molon et al.2016; Molon and Zadrag-Tecza 2016a). So far, a number of mechanisms have been identified as responsible for changes in the reproductive potential and longevity of this unicellular organism. Historically, yeast age was presented in terms of its reproductive potential (previously known as replicative lifespan). However, reproductive potential only provides information on the fertility rather than the age of a strain. For the last few years, there has been a heated debate concerning the units in which yeast's lifespan should be expressed (Minois et al.2005; Molon, Zadrag-Tecza and Bilinski 2015) Nowadays many laboratories use different laboratory yeast strains as reference in yeast ageing studies. A comparison of a number of such strains in this study showed that they may have extremely different parameters of ageing, which may lead to divergent conclusions in the field of ageing. Taking into account fertility as the criterion of choice, the most desirable model strains should have a relatively long reproductive potential, or we should compare the results with multiple laboratory strains that do not have a relatively short reproductive potential (Bitterman, Medvedik and Sinclair 2003). In this paper, we first analysed the reproductive potential of selected laboratory strains (Fig. 1). Before computing the interferential statistics, we checked the data for normality. The sample points deviated from the straight line (Fig. 1a) for the majority of the data in the (Q-Q) plots, while the Shapiro-Wilks test showed p-value much less than the significance level (0.05), which indicated a departure of the data set from the normal distribution. To get insight into the data statistics and to enable strain comparison, we used box plots as a non-parametric graphing tool (Krzywinski and Altman 2014). Figure 1b presents the box plot for the reproductive potential of the strains with increasing order of median (solid line inside box). The box borders are the lower and upper hinges, corresponding to Q1 and Q3 quartiles, respectively, and the Tukey-style whiskers extend to the most extreme points (no more than 1.5 × IQR from the edge of the box). This plot indicates a considerable skewness, presence of heavy tails and outliers in the data as well as variability in the 95% confidence interval approximated by the notch size. The CEN.PK2–1C strain showed the lowest, while the SP-4 the highest values of the median. The strains clustered into four homogeneous groups (the same letters) that were derived from the post hoc multiple pairwise Dunn's test, following a significant (P < 0.05) Kruskal-Wallis test. The same grouping pattern can be observed if we take into account the overlap of the notches in Fig. 1b. The strain variability in the reproductive potential is also shown in Fig. 1c as a function of the percentage of reproductive cells versus the number of daughters. The log-rank test (P < 0.001) indicated that the strains differed not only in the medians but also in terms of the curves. Figure 1. View largeDownload slide Reproductive potential of various wild-type S. cerevisiae strains. The Q-Q plot (a) visualises the deviation of the data from the normal distribution. The box plot (median, interquartile range and 95% confidence interval) was used for these non-parametric data to characterise diversity of the strains in terms of the reproductive potential (b).The statistical differences (P < 0.05) between the strains ordered according to median were assessed by the multiple pairwise Dunn's test and are indicated by different letters. The outliers are marked by the shaded squares. The survival function plot (c) presents the relationship between the fraction of viable cells and the number of daughters. The log-rank test showed a significant (P < 0.001) difference between the curves. The results represent measurements for at least 90 cells analysed in two independent experiments. Figure 1. View largeDownload slide Reproductive potential of various wild-type S. cerevisiae strains. The Q-Q plot (a) visualises the deviation of the data from the normal distribution. The box plot (median, interquartile range and 95% confidence interval) was used for these non-parametric data to characterise diversity of the strains in terms of the reproductive potential (b).The statistical differences (P < 0.05) between the strains ordered according to median were assessed by the multiple pairwise Dunn's test and are indicated by different letters. The outliers are marked by the shaded squares. The survival function plot (c) presents the relationship between the fraction of viable cells and the number of daughters. The log-rank test showed a significant (P < 0.001) difference between the curves. The results represent measurements for at least 90 cells analysed in two independent experiments. As seen in Fig. 1b, the reproductive potential of SP-4 is extended relative to other analysed strains. In the case of SY991, the parameter value is only slightly lower compared to SP-4. In turn, the lowest value of the reproductive potential was observed for the CEN.PK2–1C strain. Kaeberlein et al. (2005) showed that BY4742 (haploid strain of the opposite MAT character compared to the BY4741 strain used in these studies) had the highest reproductive potential among the analysed strains. So far this has been the most commonly used genetic background in yeast studies, including studies of ageing. Unfortunately, that strain often showed different average reproductive potentials depending on the laboratory (Molon, Zadrag-Tecza and Bilinski 2015; Pernice, Vevea and Pon 2016). Here we demonstrate that there are other laboratory strains with reproductive potentials exceeding that of BY4741, e.g. SY991, SP-4, FY1679. Next, we verified whether the strains displaying the same genetic background but different auxotrophic markers vary in terms of ageing parameters. Choosing the FY1679 genetic background we did not find differences in the reproductive potential between FY1679–1B and FY1679–5A strains, which supports the previous data (Kaeberlein et al.2005). To assess the strains diversity in the lifespan, we used survival analysis and the uncensored Kaplan-Meier method. The survival curves with the 95% confidence intervals (coloured ribbons) for the reproductive lifespan (i), PRLS (ii) and total lifespan (iii) are given in Fig. 2. The log-rank test showed statistical differences (P < 0.001) between the strains for the each of the lifespan parameters. Figure 2a shows that reproductive lifespan differs considerably among the investigated strains. In general, the time period during which cells are capable of budding strongly depends on two major parameters: the number of cells produced by the mother and the doubling time. Gene mutations and some growth conditions may have an influence on both these parameters (Molon et al.2016; Molon and Zadrag-Tecza 2016a). The data presented in this paper indicate that reproductive lifespan varies in different genetic backgrounds. This is not surprising as the analysed strains have various reproductive potentials. Particularly, the SP-4 and SY991 strains, which have extended reproductive potential, have the longest reproductive lifespans, while the CEN.PK2–1C strain with the shortest reproductive potential has the shortest reproductive lifespan. Figure 2. View largeDownload slide Survival functions for reproductive lifespan (a), PRLS (b) and total lifespan (c) obtained using the uncensored Kaplan-Meier method. The log-rank test showed significant (P < 0.001) differences between the strains for the each of the lifespan parameters. The results represent measurements for at least 90 cells analysed in two independent experiments. Figure 2. View largeDownload slide Survival functions for reproductive lifespan (a), PRLS (b) and total lifespan (c) obtained using the uncensored Kaplan-Meier method. The log-rank test showed significant (P < 0.001) differences between the strains for the each of the lifespan parameters. The results represent measurements for at least 90 cells analysed in two independent experiments. We then calculated the doubling time (Fig. 3) during the routine lifespan analysis. The doubling time data, similarly to other lifespan parameters, deviated from normality (Shapiro-Wallis test, P < 0.001) and showed statistically significant differences (Kruskal-Wallis test, P < 0.001) between the strains. Figure 3 presents the box plot with increasing order of median. The post hoc multiple pairwise Dunn's test showed the strains clustered into three groups. The FY1679–5A, FY1679–1B and SP4 represented the group of the lowest doubling time, while CEN.PK2–1C was characterised by the highest value of the parameter. Figure 3. View largeDownload slide Doubling time of various wild-type S. cerevisiae strains. The box plots (median, interquartile range and 95% confidence interval) describe variability of the strains in the order of increasing median. Statistical (P < 0.05) differences between the strains assessed by the non-parametric pairwise Dunn's test are indicated by different letters. The outliers are marked by the shaded squares. The results represent measurements for at least 90 cells analysed in two independent experiments. Figure 3. View largeDownload slide Doubling time of various wild-type S. cerevisiae strains. The box plots (median, interquartile range and 95% confidence interval) describe variability of the strains in the order of increasing median. Statistical (P < 0.05) differences between the strains assessed by the non-parametric pairwise Dunn's test are indicated by different letters. The outliers are marked by the shaded squares. The results represent measurements for at least 90 cells analysed in two independent experiments. Reports on the lifespan of yeast during the reproductive and post-reproductive stages are scarce. Recent studies have shown that it is possible to verify reproductive lifespan of yeast not only through the standard procedure (Minois et al.2005) but also through automated tracking of single yeast cells (Jo et al.2015). So far there has been few reports showing the cell's length of life after the period of reproduction, although mother yeast cells usually do not die after the last budding. The data obtained in previous studies with yeast mutants suggested that the length of the cell's life after reproduction was inversely proportional to the reproductive potential (Zadrag-Tecza et al.2013). In this work, we showed, using Kaplan-Meier survival analysis, that PRLS was related to and probably specific to genetic backgrounds (Fig. 2b). This was clearly manifested in the case of the SP-4 (extremely increased reproductive potential) and BMA64-1A laboratory strains with very short time after the reproduction. In the latter case, cells achieved a large volume and exploded, which dramatically decreased their PRLS despite the relatively low reproductive potential (Molon and Zadrag-Tecza 2016b). The fact that the length of life after reproduction depends on the genetic background is also evidenced by the significant extension of both the reproductive potential and the PRLS of the FY1679–01B and SY991 strains (Figs 1c and 2b). Longevity of an organism is measured as a calendar lifespan expressed in units of time. In yeast, total lifespan is defined as the sum of the time period when yeast cells are capable of budding and the PRLS. In our study, we showed that FY1679–01B and SY991 are high-longevity strains while BMA64–1A and CEN.PK2–1C are the short-lived ones (Fig. 2c). These data indicated that the lifetime of cells depended to a large degree on the time after reproduction. Our observations suggest thus that genetic background may have a significant impact not only on the reproductive potential but also longevity. In addition, different auxotrophic markers may have an impact on longevity, mainly by regulation of PRLS (Fig. 2b and c), but this phenomenon requires further experiments. It seems that selection of a laboratory strain with gene disruption or overexpression can generate pseudo-positive or pseudo-negative results (Dmello et al.1994; Kaeberlein et al.2005). Further, we attempted to establish phylogenetic matches between the analysed strains. To accomplish this purpose, we used a standard rDNA internal transcribed spacer region (ITS1–5,8S-ITS2) and the S288c laboratory strain as a reference. The ITS region is the most widely sequenced DNA region in fungi, particularly useful for molecular systematics at the species level as well as within species. Regions of fungal ribosomal DNA (rDNA) are highly variable sequences of great importance in phylogenetic analysis (Martin and Rygiewicz 2005). The phylogenetic analysis performed in our study with the use of UPGMA and NJ showed a relationship between the reproductive potential and reproductive lifespan, and genetic similarity for the analysed strains. The phylograms consisted of two main branches. The strains with an increased reproductive potential (above ∼30 buddings) and extended reproductive lifespan, including SP-4, FY1679–05A, FY1679–01B, SY991, centred around one of the branches, while the remaining strains gathered on the other (Fig. 4a and b). It seems probable that closely related strains may have similar reproductive potentials and our results support earlier observations (Kaeberlein et al.2005). Unfortunately, it is difficult to merge strain relatedness and longevity expressed in units of time. Figure 4. View largeDownload slide UPGMA (a) and NJ (b) tree for the S. cerevisiae laboratory strains obtained from a distance matrix based on pairwise alignments of the ITS (ITS1–5.8S-ITS2) loci. Figure 4. View largeDownload slide UPGMA (a) and NJ (b) tree for the S. cerevisiae laboratory strains obtained from a distance matrix based on pairwise alignments of the ITS (ITS1–5.8S-ITS2) loci. Finally, we focused on searching possible biochemical determinants of ageing using mid-infrared spectroscopy combined with chemometrics. Earlier reported metabolic analysis of Saccharomyces cerevisiae isogenic mutants differing in terms of replicative longevity showed an association between lifespan and the MS-based fingerprint (Yoshida et al.2010). Taking into account these promising findings, we attempted to establish possible relationships between selected parameters characterising the strain longevity (reproductive potential, reproductive lifespan, PRLS, total lifespan) and the strain metabolic fingerprints generated by FTIR spectroscopy. This method provides a comprehensive snapshot of cellular biochemistry, including all macromolecules, and thus gives insight into abundance and mutual proportions of major chemical compounds constituting a cell. The average mid-infrared spectrum of the analysed strains is shown in Fig. 5. The most prominent peaks at 2926, 1642 and 1539 cm−1 were attributable to (C-H) asymmetric stretching vibrations of CH2 groups in lipids, and the amide I + II vibrations characterising proteins, respectively (Silverstein, Webster and Kiemle 2005). The low-frequency bands at 1405, 1237 and 1052 cm−1 corresponded to the spectral fingerprint region and involved overlapped absorbance due to proteins, nucleic acids, lipids and carbohydrates. The PCA was used as the unsupervised multivariate method to identify hidden structures in the spectral dataset and to visualise potential grouping and separating patterns of the data. This approach tremendously reduced the multidimensional space formed by the spectra into a few orthogonal (uncorrelated) variables while preserving crucial information on the date variance. Figure 5. View largeDownload slide Average FTIR spectrum of whole cell material for S. cerevisiae strains acquired in the wavenumber range from 4000 to 900 cm−1. The most prominent absorbance bands are indicated and correspond to (C-H) asymmetric stretching vibrations of CH2 groups in lipids (2926 cm−1), amide I (1642 cm−1) and amide II (1539 cm−1) bands of proteins and to the fingerprint region between 1450 and 900 cm−1. The carbonyl (C = O) stretching vibration region (1760–1700 cm−1) is magnified and given in the inset. Figure 5. View largeDownload slide Average FTIR spectrum of whole cell material for S. cerevisiae strains acquired in the wavenumber range from 4000 to 900 cm−1. The most prominent absorbance bands are indicated and correspond to (C-H) asymmetric stretching vibrations of CH2 groups in lipids (2926 cm−1), amide I (1642 cm−1) and amide II (1539 cm−1) bands of proteins and to the fingerprint region between 1450 and 900 cm−1. The carbonyl (C = O) stretching vibration region (1760–1700 cm−1) is magnified and given in the inset. We applied this analysis both to whole spectrum and selected spectral regions. No clear relationship was established between the strain longevity parameters, including the reproductive potential, reproductive lifespan, PRLS and total lifespan, and the metabolic mid-infrared spectral features corresponding to the whole mid-infrared spectrum, the amide I and amide II bands, and the spectral fingerprint region between 1450 and 900 cm−1 (data not shown). The only meaningful spectral information related to the longevity was observed for the region between 1760 and 1710 cm−1 wavenumbers. This band is attributable to the carbonyl (C = O) group stretching vibration but the precise peak location and assignation depends on the molecular environment of the group (Stuart 2005). Absorbance is generally the most prominent in the whole spectrum for pure chemical compounds containing the C = O group in their molecular structure. However, in highly heterogeneous yeast cell material, the signal, mostly from lipids, was overwhelmed by vibration from other more abundant cell compounds. As a consequence, the FTIR measurements generated relatively low-intensity spectral signature, difficult to be observed directly (Fig. 5 inset). We attempted to reveal the hidden information in this spectral region by the use of spectroscopy in combination with chemometrics, particularly employing the PCA. The lowest three principal components (PC1–PC3) were found to account for above 95% of the total spectral variability in this region. The PC1 scores did not correlate with any of the longevity parameters (data not shown). This may suggest that the majority of metabolic information detected in that spectral range was related to strain variability in terms of structural and physiological traits that are not related directly to longevity but reflect fundamental variability of the genetic background. The most interesting association was observed between the PC2 scores (explaining 7% of total variability) and PRLS (PRLS) (Fig. 6a). The strains with low values of PRLS had generally highly negative PC2 scores. Moreover, the PC2 coordinates increased along with increment of PRLS. Only the FY1679–01B strain that showed the highest observed PRLS was exceptionally located in the middle of the PC2 score scale. Application of the Spearman's rank correlation coefficient (s) confirmed a monotonic relationship between the PC2 scores and the lifespan parameter expressed both as the mean (s = 0.97, P = 0.001 and s = 0.88, P = 0.002) and the expected event sum (E) predicted using log-rank test and the Kaplan-Meier procedure (s = 0.97, P = 0.001 and s = 0.88, P = 0.002) for the strain set excluding FY1679–01B. Figure 6. View largeDownload slide (a) Two-dimensional scatter plot of scores for principal components PC2 vs PC3 estimated on the basis of mid-infrared spectra in the region of carbonyl (C = O) stretching vibrations for S. cerevisiae strains. Each point is a projection of the average FTIR spectral data for a given strain in the range of 1760–1710 cm −1 on the PC2. The PC2 explained 7% of total variance of the data set. Values at the top describe the PRLS of strains and correspond to the PC2 scores indicated by the arrows. (b) Loadings of the PC2 for FTIR spectral data. Extremes at 1716 and 1736 cm−1 indicate wavenumbers contributing the most prominently to the variance and clustering pattern of the PC2 scores given in (a). (c) Correlation plot for the mean and for the total number of expected events (E) calculated using log-rank test and survival analysis vs PC2 scores. The shaded area corresponds to 95% confidence intervals. The Spearman's rank correlation coefficient and the P-value are indicated. The strain FY1679–01B (marked with the dark blue ring) was considered an outlier and was not included into the statistical calculations. Figure 6. View largeDownload slide (a) Two-dimensional scatter plot of scores for principal components PC2 vs PC3 estimated on the basis of mid-infrared spectra in the region of carbonyl (C = O) stretching vibrations for S. cerevisiae strains. Each point is a projection of the average FTIR spectral data for a given strain in the range of 1760–1710 cm −1 on the PC2. The PC2 explained 7% of total variance of the data set. Values at the top describe the PRLS of strains and correspond to the PC2 scores indicated by the arrows. (b) Loadings of the PC2 for FTIR spectral data. Extremes at 1716 and 1736 cm−1 indicate wavenumbers contributing the most prominently to the variance and clustering pattern of the PC2 scores given in (a). (c) Correlation plot for the mean and for the total number of expected events (E) calculated using log-rank test and survival analysis vs PC2 scores. The shaded area corresponds to 95% confidence intervals. The Spearman's rank correlation coefficient and the P-value are indicated. The strain FY1679–01B (marked with the dark blue ring) was considered an outlier and was not included into the statistical calculations. These results support previous findings (Yoshida et al.2010) suggesting that biochemical status of cells may be associated to some extent with yeast longevity. Subsequent analysis of loading factors for PC2 allowed us to establish which spectral features contributed most markedly to this variance. The loadings (Fig. 6b) showed a positive peak at 1716 cm−1 and a negative peak (local minimum) at 1736 cm−1. The former can be assigned tentatively to the carbonyl group (CO) of aldehydes or free fatty acids (FFA), while the latter, shifted to higher frequencies, is mainly related to ester bonds (Silverstein, Webster and Kiemle 2005). The most intense absorbance in the region of ester bonds was located at about 1747 cm−1 (inset in Fig. 5), which can be ascribed to triacylglycerols (TAGs). TAG usually shows maximum absorbance above 1740 cm−1 but its precise location depends on the molecular environment and molecular interactions. FTIR spectromicroscopic mapping of a cell demonstrated that TAG constituting of lipid stores showed a band centred at 1744 cm−1 (Gazi et al.2007). In turn, phospholipid standards of the most abundant membrane compounds in yeast (Vanderrest et al.1995) including phosphatidylcholine, phosphoethanolamine and phosphoinsositol (Cagnasso et al.2010; Derenne, Vandersleyen and Goormaghtigh 2014) exhibited carbonyl bands shifted below 1740 cm−1. However, a possible small contribution of the esterified sterols to the band should not be excluded (Acuna-Johnson and Oehlschlager 1989). The PC2 score plot predicted a negative association of the peak tentatively corresponding to FAs esterified to phospholipids and a positive relationship of the peak ascribed to FFAs to PRLS. Thus, the results indicate a reverse relationship across the examined strains, between the abundance of the compounds containing the C = O functional group in a non-esterified form and those esterified to phospholipids. Consequently, the proportion of both lipid forms discriminated the laboratory genotypes in a way associated with post-reproductive longevity. There was, however, no spectral indication from PCA that the pool of TAG ranked the laboratory strains according to the lifespan characteristics. In earlier studies, a link between metabolic profile and ageing was explored using a set of longevity mutants and replicative lifespan model (Yoshida et al.2010). The analysis of small molecule profile determined by means of MS indicated that amino acids and nucleotide derivatives are the major discriminating biochemical compounds related to lifespan extension. Among other metabolites, those related to tricarboxylic acid cycle were positively correlated while glycolytic intermediates were negatively associated with lifespan. In turn, our studies focused on the overall biochemical status of the cell that gave insight on the profile of the major cell constituents, and thus pertained to quite different composition features. We did not find among the examined laboratory strains any link between the reproductive potential or reproductive lifespan and the biochemical fingerprint of the cells. In the understanding of biochemical fundamentals of ageing in yeast, an essential role is ascribed to lipid metabolism and lipid homeostasis (Petschnigg et al.2009). Cellular prolongevity function of TAG has been demonstrated very recently on the basis of a survey of laboratory and wild isolates differing in chronological lifespan as well as through genetic manipulation of TAG biosynthesis and catabolism (Handee et al.2016). The TAG is accumulated in the cell in form of lipid particles as a storage of surplus chemical energy and a source of building blocks for membrane biogenesis. Increased level of TAG has been proved to coincide with enhanced longevity, in a way independent from energy release or stress response factors (Handee et al.2016). Although TAG is dispensable for cell survival in optimal growth conditions (Fei et al.2008), the cells deprived of molecular mechanisms to incorporate FAs into glycerides generally show a delayed growth and reduced cell viability (Zhang et al.2003; Garbarino et al.2009; Connerth et al.2010). The function of TAGs related to the longevity promotion seems to be associated with the control of FAs channelling into phospholipid synthesis rather than being a source of stored energy. TAG shares common precursors with phospholipids including diacylglycerol and phosphatidic acid and thus may serve as a buffer for excess FAs, particularly cytotoxic unsaturated species, preventing massive proliferation of phospholipidic membranes that ultimately may lead to cell death (Petschnigg et al.2009). The lipid-mediated cell death may be also related to activation of apoptotic-like response through targeting different cellular pathways (Zhang et al.2003; Low et al.2005; Garbarino et al.2009). The role of TAG was particularly prominent for chronological lifespan (Handee et al.2016), when TAGs are biosynthesised to a high level. For reproductive lifespan, the relationship appeared less obvious since only the TAG synthesising capability (but not TAG hydrolysis) was needed to attain full replicative potential (Handee et al.2016). Our results support those concepts, highlighting significance of lipid metabolism and homeostasis in ageing. We showed here for the first time that lipid homeostasis might be related to post-reproductive longevity. Previously reported observations on lipid metabolism and homeostasis involvement in lifespan extension concerned different longevity models, or in fact reproductive capabilities. FTIR spectroscopy combined with chemometrics indicated that maintenance of phospholipids at low content in a cell relative to a pool of FFAs might have a prolongevity effect for the post-reproductive phase. Minimisation of phospholipids rather than maximisation of TAG appeared critical in our study. This does not exclude, however, the role of TAG typically ascribed in the control of the FFAs canalling in to phospholipid biosynthesis. The mid-infrared spectral band in the C = O stretching region showed maximum at the wavenumbers specific for the FA esters of glycerol and thus provided higher absorbance signal compared to that from the FAs esterified to phospholipids, shifted to lower frequencies. However, it did not rank the strains according to the longevity parameters, possibly due to the interfering effects of background genetic variability. Phospholipids are relatively a much more stable compound during yeast cell growth than TAG (Taylor and Parks 1979), and thus may be regarded as another convenient marker in studies on ageing. FUNDING This research was supported by the University of Rzeszow's task grant no. WBR/KBiBK/DS/1/2016 and partially by grant no. DEC-2013/09/B/NZ3/01352 from the Polish National Science Centre. Conflict of interest. None declared. REFERENCES Acuna-Johnson P, Oehlschlager AC. Identification of sterols and biologically significant steroids by UV and IR spectroscopy. In: Parish EJ (ed). Analysis of Sterols and Other Biologically Significant Steroids . New York: Academic Press, 1989, 267– 84. Adt I, Kohler A, Gognies S et al.  . FTIR spectroscopic discrimination of Saccharomyces cerevisiae and Saccharomyces bayanus strains. Can J Microbiol  2010; 56: 793– 801. Google Scholar CrossRef Search ADS PubMed  Aguilaniu H, Gustafsson L, Rigoulet M et al.  . Asymmetric inheritance of oxidatively damaged proteins during cytokinesis. Science  2003; 299: 1751– 3. Google Scholar CrossRef Search ADS PubMed  Benjamini Y, Hochberg Y. Controlling the false discovery rate - a practical and powerful approach to multiple testing. J Roy Stat Soc B-Met  1995; 57 : 289– 300. Bilinski T, Krawiec Z, Liczmanski A et al.  . Is hydroxyl radical generated by the Fenton reaction in vivo? Biochem Bioph Res Co  1985; 130: 533– 9. Google Scholar CrossRef Search ADS   Bitterman KJ, Medvedik O, Sinclair DA. Longevity regulation in Saccharomyces cerevisiae: linking metabolism, genome stability, and heterochromatin. Microbiol Mol Biol R  2003; 67: 376. Google Scholar CrossRef Search ADS   Burnett C, Valentini S, Cabreiro F et al.  . Absence of effects of Sir2 overexpression on lifespan in C. elegans and Drosophila. Nature  2011; 477: 482– 5. Google Scholar CrossRef Search ADS PubMed  Cagnasso M, Boero V, Franchini MA et al.  . ATR-FTIR studies of phospholipid vesicle interactions with alpha-FeOOH and alpha-Fe2O3 surfaces. Colloid Surface B  2010; 76: 456– 67. Google Scholar CrossRef Search ADS   Clark TG, Bradburn MJ, Love SB et al.  . Survival analysis part I: Basic concepts and first analyses. Br J Cancer  2003; 89 : 232– 8. Google Scholar CrossRef Search ADS PubMed  Connerth M, Czabany T, Wagner A et al.  . Oleate inhibits steryl ester synthesis and causes liposensitivity in Yeast. J Biol Chem  2010; 285: 26832– 41. Google Scholar CrossRef Search ADS PubMed  Correa-Garcia S, Bermudez-Moretti M, Travo A et al.  . FTIR spectroscopic metabolome analysis of lyophilized and fresh Saccharomyces cerevisiae yeast cells. Anal Methods  2014; 6: 1855– 61. Google Scholar CrossRef Search ADS   Dang WW, Steffen KK, Perry R et al.  . Histone H4 lysine 16 acetylation regulates cellular lifespan. Nature  2009; 459: 802– 7. Google Scholar CrossRef Search ADS PubMed  Derenne A, Vandersleyen O, Goormaghtigh E. Lipid quantification method using FTIR spectroscopy applied on cancer cell extracts. Biochim Biophys Acta  2014; 1841: 1200– 9. Google Scholar CrossRef Search ADS PubMed  Dinno A. dunn.test: Dunn's Test of Multiple Comparisons Using Rank Sums. R Package Version 1.3.2  2016 https://CRAN.R-project.org/web/packages/dunn.test (19 May 2017, date last accessed). Dmello NP, Childress AM, Franklin DS et al.  . Cloning and characterization of lag1, a longevity-assurance gene in yeast. J Biol Chem  1994; 269: 15451– 9. Google Scholar PubMed  Dunn OJ. Multiple comparisons using rank sums. Technometrics  1964; 6: 241– 52. Google Scholar CrossRef Search ADS   Erjavec N, Larsson L, Grantham J et al.  . Accelerated aging and failure to segregate damaged proteins in Sir2 mutants can be suppressed by overproducing the protein aggregation-remodeling factor Hsp104p. Gene Dev  2007; 21: 2410– 21. Google Scholar CrossRef Search ADS PubMed  Fabrizio P, Longo VD. The chronological life span of Saccharomyces cerevisiae. Aging Cell  2003; 2: 73– 81. Google Scholar CrossRef Search ADS PubMed  Fabrizio P, Pletcher SD, Minois N et al.  . Chronological aging-independent replicative life span regulation by Msn2/Msn4 and Sod2 in Saccharomyces cerevisiae. FEBS Lett  2004; 557: 136– 42. Google Scholar CrossRef Search ADS PubMed  Fei WH, Alfaro G, Muthusamy BP et al.  . Genome-wide analysis of sterol-lipid storage and trafficking in Saccharomyces cerevisiae. Eukaryot Cell  2008; 7: 401– 14. Google Scholar CrossRef Search ADS PubMed  Garbarino J, Padamsee M, Wilcox L et al.  . Sterol and diacylglycerol acyltransferase deficiency triggers fatty acid-mediated cell death. J Biol Chem  2009; 284: 30994– 1005. Google Scholar CrossRef Search ADS PubMed  Gazi E, Dwyer J, Lockyer NP et al.  . Biomolecular profiling of metastatic prostate cancer cells in bone marrow tissue using FTIR microspectroscopy: a pilot study. Anal Bioanal Chem  2007; 387: 1621– 31. Google Scholar CrossRef Search ADS PubMed  Gershon H, Gershon D. The budding yeast, Saccharomyces cerevisiae, as a model for aging research: a critical review. Mech Ageing Dev  2000; 120: 1– 22. Google Scholar CrossRef Search ADS PubMed  Goffeau A, Barrell BG, Bussey H et al.  . Life with 6000 genes. Science  1996; 274: 546 . Google Scholar CrossRef Search ADS PubMed  Handee W, Li XB, Hall KW et al.  . An energy-independent pro-longevity function of triacylglycerol in yeast. PLos Genet  2016; 12: 1– 25. Google Scholar CrossRef Search ADS   Jo MC, Liu W, Gu L et al.  . High-throughput analysis of yeast replicative aging using a microfluidic system. P Natl Acad Sci USA  2015; 112: 9364– 9. Google Scholar CrossRef Search ADS   Kaeberlein M, Kirkland KT, Fields S et al.  . Genes determining yeast replicative life span in a long-lived genetic background. Mech Ageing Dev  2005; 126: 491– 504. Google Scholar CrossRef Search ADS PubMed  Kanfi Y, Naiman S, Amir G et al.  . The sirtuin SIRT6 regulates lifespan in male mice. Nature  2012; 483: 218– 21. Google Scholar CrossRef Search ADS PubMed  Krzywinski M, Altman N. Visualizing samples with box plots. Nat Methods  2014; 11 : 119– 20. Google Scholar CrossRef Search ADS PubMed  Lippuner AD, Julou T, Barral Y. Budding yeast as a model organism to study the effects of age. FEMS Microbiol Rev  2014; 38: 300– 25. Google Scholar CrossRef Search ADS PubMed  Logan M. Biostatistical Design and Analysis Using R: A Practical Guide . Chichester, UK: John Wiley & Sons, 2011. Longo VD, Mitteldorf J, Skulachev VP. Opinion - Programmed and altruistic ageing. Nat Rev Genet  2005; 6: 866– 72. Google Scholar CrossRef Search ADS PubMed  Longo VD, Shadel GS, Kaeberlein M et al.  . Replicative and chronological aging in Saccharomyces cerevisiae. Cell Metab  2012; 16: 18– 31. Google Scholar CrossRef Search ADS PubMed  Low CP, Liew LP, Pervaiz S et al.  . Apoptosis and lipoapoptosis in Schizosaccharomyces the fission yeast pombe. FEMS Yeast Res  2005; 5: 1199– 206. Google Scholar CrossRef Search ADS PubMed  Martin GM, Austad SN, Johnson TE. Genetic analysis of ageing: Role of oxidative damage and environmental stresses. Nat Genet  1996; 13: 25– 34. Google Scholar CrossRef Search ADS PubMed  Martin KJ, Rygiewicz PT. Fungal-specific PCR primers developed for analysis of the ITS region of environmental DNA extracts. BMC Microbiol  2005; 5: 1– 11. Google Scholar CrossRef Search ADS PubMed  Minois N, Frajnt M, Wilson C et al.  . Advances in measuring lifespan in the yeast Saccharomyces cerevisiae. P Natl Acad Sci USA  2005; 102: 402– 6. Google Scholar CrossRef Search ADS   Molon M, Szajwaj M, Tchorzewski M et al.  . The rate of metabolism as a factor determining longevity of the Saccharomyces cerevisiae yeast. Age  2016; 38. Molon M, Zadrag-Tecza R. Effect of temperature on replicative aging of the budding yeast Saccharomyces cerevisiae. Biogerontology  2016a; 17: 347– 57. Google Scholar CrossRef Search ADS   Molon M, Zadrag-Tecza R. The links between hypertrophy, reproductive potential and longevity in the Saccharomyces cerevisiae yeast. Acta Biochim Pol  2016b; 63: 329– 34. Google Scholar CrossRef Search ADS   Molon M, Zadrag-Tecza R, Bilinski T. The longevity in the yeast Saccharomyces cerevisiae: A comparison of two approaches for assessment the lifespan. Biochem Bioph Res Co  2015; 460: 651– 6. Google Scholar CrossRef Search ADS   Pan KZ, Palter JE, Rogers AN et al.  . Inhibition of mRNA translation extends lifespan in Caenorhabditis elegans. Aging Cell  2007; 6: 111– 9. Google Scholar CrossRef Search ADS PubMed  Pernice WM, Vevea JD, Pon LA. A role for Mfb1p in region-specific anchorage of high-functioning mitochondria and lifespan in Saccharomyces cerevisiae. Nat Commun  2016; 7: 1– 12. Google Scholar CrossRef Search ADS   Peto R, Pike MC, Armitage P et al.  . Design and analysis of randomized clinical-trials requiring prolonged observation of each patient .2. Analysis and examples. Br J Cancer  1977; 35 : 1– 39 Google Scholar CrossRef Search ADS PubMed  Petschnigg J, Wolinski H, Kolb D et al.  . Good fat, essential cellular requirements for Triacylglycerol synthesis to maintain membrane homeostasis in yeast. J Biol Chem  2009; 284: 30981– 93. Google Scholar CrossRef Search ADS PubMed  Piper PW. Maximising the yeast chronological lifespan. Aging Res Yeast  2012; 57: 145– 59. Polymenis M, Kennedy BK. Chronological and replicative lifespan in yeast: Do they meet in the middle? Cell Cycle  2012; 11: 3531– 2. Google Scholar CrossRef Search ADS PubMed  Postma L, Lehrach H, Ralser M. Surviving in the cold: yeast mutants with extended hibernating lifespan are oxidant sensitive. Aging-US  2009; 1: 957– 60. Google Scholar CrossRef Search ADS   R Core Team R. A Language and Enviromental for Statistical Computing . R Foundation for Statistical Computing 2013. http://www.R-project.org/ (19 May 2017, date last accessed). Rego A, Trindade D, Chaves SR et al.  . The yeast model system as a tool towards the understanding of apoptosis regulation by sphingolipids. FEMS Yeast Res  2014; 14: 160– 78. Google Scholar CrossRef Search ADS PubMed  Reznick DN, Bryant MJ, Roff D et al.  . Effect of extrinsic mortality on the evolution of senescence in guppies. Nature  2004; 431: 1095– 9. Google Scholar CrossRef Search ADS PubMed  Rona G, Herdeiro R, Mathias CJ et al.  . CTT1 overexpression increases life span of calorie-restricted Saccharomyces cerevisiae deficient in Sod1. Biogerontology  2015; 16: 343– 51. Google Scholar CrossRef Search ADS PubMed  Selman C, Tullet JMA, Wieser D et al.  . Ribosomal protein S6 kinase 1 signaling regulates mammalian life span. Science  2009; 326: 140– 4. Google Scholar CrossRef Search ADS PubMed  Sinclair DA, Guarente L. Extrachromosomal rDNA circles—a cause of aging in yeast. Cell  1997; 91: 1033– 42. Google Scholar CrossRef Search ADS PubMed  Sinclair DA, Mills K, Guarente L. Accelerated aging and nucleolar fragmentation yeast sgs1 mutants. Science  1997; 277: 1313– 6. Google Scholar CrossRef Search ADS PubMed  Silverstein R, Webster F, Kiemle D. Spectrometric Identification of Organic Compounds . New York: John Wiley & Sons, 2005. Stansfield I, Stark M. Yeast Gene Analysis . London: Academic Press, 2007. Stuart B. Infrared Spectroscopy: Fundamentals and Applications . Chichester, UK: John Wiley & Sons, 2005. Tang FS, Watkins JW, Bermudez M et al.  . A life-span extending form of autophagy employs the vacuole-vacuole fusion machinery. Autophagy  2008; 4: 874– 86. Google Scholar CrossRef Search ADS PubMed  Taylor FR, Parks LW. Triaglycerol metabolism in Saccharomyces cerevisiae. Relation to phospholipid synthesis. Biochim Biophys Acta  1979; 575 : 204– 14. Google Scholar CrossRef Search ADS PubMed  Therneau T. A Package for Survival Analysis in S. version 2.40-1 . https://CRAN.R-project.org/web/packages/survival (19 May 2017, date last accessed). Tissenbaum HA, Guarente L. Increased dosage of a sir-2 gene extends lifespan in Caenorhabditis elegans. Nature  2001; 410: 227– 30. Google Scholar CrossRef Search ADS PubMed  Vachova L, Cap M, Palkova Z. Yeast colonies: a model for studies of aging, environmental adaptation, and longevity. Oxid Med Cell Longev  2012, DOI https://doi.org/10.1155/2012/601836. Vanderrest ME, Kamminga AH, Nakano A et al.  . The plasma-membrane of saccharomyces-cerevisiae - structure, function, and biogenesis. Microbiol Rev  1995; 59: 304– 22. Google Scholar PubMed  Veatch JR, McMurray MA, Nelson ZW et al.  . Mitochondrial dysfunction leads to nuclear genome instability via an iron-sulfur cluster defect. Cell  2009; 137: 1247– 58. Google Scholar CrossRef Search ADS PubMed  White TJ, Bruns T, Lee S et al.  . Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics. In: Innis MA, Gelfand DH, Sninsky JJet al.   (eds). PCR Protocols: A Guide to Methods and Applications . New York: Academic Press, 1990. Yoshida R, Tamura T, Takaoka C et al.  . Metabolomics-based systematic prediction of yeast lifespan and its application for semi-rational screening of ageing-related mutants. Aging Cell  2010; 9: 616– 25. Google Scholar CrossRef Search ADS PubMed  Zadrag R, Bartosz G, Bilinski T. Is the yeast a relevant model for aging of multicellular organisms? An insight from the total lifespan of Saccharomyces cerevisiae. Curr Aging Sci  2008; 1: 159– 65. Google Scholar CrossRef Search ADS PubMed  Zadrag-Tecza R, Molon M, Mamczur J et al.  . Dependence of the yeast Saccharomyces cerevisiae post-reproductive lifespan on the reproductive potential. Acta Biochim Pol  2013; 60: 111– 5. Google Scholar PubMed  Zhang Q, Chieu HK, Low CP et al.  . Schizosaccharomyces pombe cells deficient in triacylglycerols synthesis undergo apoptosis upon entry into the stationary phase. J Biol Chem  2003; 278: 47145– 55. Google Scholar CrossRef Search ADS PubMed  Zhou CK, Slaughter BD, Unruh JR et al.  . Motility and segregation of Hsp104-associated protein aggregates in budding Yeast. Cell  2011; 147: 1186– 96. Google Scholar CrossRef Search ADS PubMed  © FEMS 2017. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
TI - Phylogenetic relationship and Fourier-transform infrared spectroscopy-derived lipid determinants of lifespan parameters in the Saccharomyces cerevisiae yeast
JF - FEMS Yeast Research
DO - 10.1093/femsyr/fox031
DA - 2017-06-01
UR - https://www.deepdyve.com/lp/oxford-university-press/phylogenetic-relationship-and-fourier-transform-infrared-spectroscopy-Ak5A4SC9KI
SP - fox031
VL - 17
IS - 4
DP - DeepDyve
ER -