IDENTIFICATION OF MUTATIONS THAT EXTEND THE FISSION YEAST
SCHIZOSACCHAROMYCES POMBE CHRONOLOGICAL LIFESPAN BY A
NOVEL PARALLEL SELECTION APPROACH
by
BO-RUEI CHEN
Submitted in partial fulfillment of the requirements
For the degree of Doctor of Philosophy
Dissertation Adviser: Kurt W. Runge, Ph.D.
Department of Genetics
CASE WESTERN RESERVE UNIVERSITY
January, 2011
CASE WESTERN RESERVE UNIVERSITY
SCHOOL OF GRADUATE STUDIES
We hereby approve the thesis/dissertation of
Bo-Ruei Chen a
candidate for the Ph.D. degree *.
(signed) Mark Adams, Ph.D. . (chair of the committee)
Kurt Runge, Ph.D. .
Steven Sanders, Ph.D. .
Peter Harte, Ph.D. .
Jo Ann Wise, Ph.D. .
(date) s 10/18/2010 _
*We also certify that written approval has been obtained for any proprietary material contained therein.
Tables of Contents
List of Tables v
List of Figures vi
Acknowledgements ix
List of Abbreviations x
Abstract 1
Chapter 1: Background, significance and specific aims 3 Aging, lifespan and longevity: a general introduction 4 Aging: an evolutionary perspective 6 The free radical theory of aging 7 Caloric restriction, stress management and lifespan 11 Mitochondrial activity, metabolism and aging 14 Down-regulation of the target of rapamycin (TOR) signaling and longevity 17 The insulin/insulin-like growth factor (IGF)-1 signaling and longevity 22 Sir proteins (sirtuins) and aging: more than just silencing transcription 26 Advantages of using yeast as a model to study evolutionarily conserved mechanisms of aging 33 The fission yeast Schizosaccharomyces pombe as an emerging model to study aging and longevity 35 Significance and specific aims of this study 37
Chapter 2: A new Schizosaccharomyces pombe chronological lifespan assay reveals that caloric restriction promotes efficient cell cycle exit and extends longevity 40 Abstract 41 Introduction 42 Materials and Methods 46 S. pombe strains 46 Chronological aging assays 48 Analysis of CLS assay data 49 FACS analysis (fixed cells) 49
i FACS analysis (live cells) 50 Stress sensitivity assays 50 Results 52 Chronological lifespan assay design 52 Cells in SD medium show the evolutionarily conserved lifespan shortening in response to over nutrition while cells in EMM medium do not 52 Caloric restriction extends lifespan in the SD medium-based assay 57 Longer CLS correlates with exhausting free glucose in the medium 63 Long-lived calorically restricted cells show increased stress resistance 64 The AKT orthologs sck1+ and sck2+ differentially affect lifespan under 66 normal and over nutrition conditions Deletions of the non-essential TOR kinase tor1+ or its potential 69 substrate gad8+ extends S. pombe chronological lifespan Discussion 71
Chapter 3: Construction and characterization of a barcode-tagged insertion mutant library in the fission yeast Schizosaccharomyces pombe 75 Abstract 76 Introduction 77 Materials and Methods 83 Strains and media 83 Construction of the bacterial barcode-tagged insertion DNA vector library 84 Construction of the fission yeast barcode-tagged insertion mutant library 88 Determination of the size of S. pombe barcode-tagged insertion mutant 89 library and bacterial barcode library Phenotypic assays to assess mutation diversity 90 Identification of insertion sites by thermal asymmetric interlaced (TAIL)-PCR 91 Results 93 Design of a linear barcode-tagged DNA vector for insertion mutagenesis in the fission yeast S. pombe 93 Construction of the bacterial barcode-tagged insertion DNA library 97
ii Construction of the fission yeast barcode-tagged insertion mutant library 97 The insertion mutant library contains multiple diverse mutations 100 Identification of the insertion sites by TAIL-PCR 102 Discussion 106
Chapter 4: A novel parallel selection approach identified a cyclin/CDK complex, Clg1p/Pef1p, whose inactivation leads to lifespan extension in the fission yeast Schizosaccharomyces pombe 111 Abstract 112 Introduction 114 Materials and Methods 119 Strains and media 119 Chronological aging assays 121 A parallel selection for long-lived mutations using PCR-mediated barcode sequencing 121 Identification of insertion sites by thermal asymmetric interlaced 126 (TAIL)-PCR and splinkerette PCR Construction of fission yeast mutant strains for lifespan analysis 127 Expression plasmids construction 130 Yeast two hybrid assay 131 Protein extraction, immunoprecipitation and Western blotting 131 Analysis of CLS assay data 133 Results 134 An unbiased parallel selection for long-lived barcode-tagged insertion mutants using a novel sequencing strategy 134 CLS of clg1-, spncrna.142- and 28S rRNA mutants 141 Establishment of a working model for Clg1p-dependent lifespan regulation 144 Clg1p physically interacts with the CDK Pef1p 145 Clg1p and the CDK Pef1p control CLS through the same pathway 147 Identification of other Pcl-like cyclins in S. pombe 149 Deletion of psl1+ or pas1+ does not extend CLS 152 Identification of potential S. pombe Rim15p orthologs 152
iii Cek1p physically interacts with Pef1p 156 A proposed model of chronological lifespan control in S. pombe by 157 Clg1p, Pef1p and Cek1p Discussion 159
Chapter 5: General discussion and future directions 165 Using yeast as a model to study aging: advantages and limitations 166 The fission yeast Schizosaccharomyces pombe as an emerging yeast aging 169 model Current approaches for large-scale genetic screen for genes determining 170 yeast lifespan Developing a S. pombe CLS assay and a novel genetic screen for long-lived 174 fission yeast mutants Limitations of the current barcoded insertion mutant library and probable 177 alternatives A proof-of-principle parallel genetic screen identified the Clg1p-Pef1p- 179 Cek1p lifespan-regulating pathway Cell cycle, senescence and aging 183 Future directions 185
Bibliography 188
iv List of Tables
Chapter 1 Table 1.1. Non-genetic interventions that extend model organism lifespan 5 Table 1.2. Evolutionarily conserved genes mentioned in this chapter whose orthologs have been implicated in the regulation of aging and longevity 5
Chapter 2 Table 2.1. Oligonucleotides used in this study 47 Table 2.2. Cells grown in caloric restriction (0.1% glucose) medium exhaust their glucose supply as they reach maximum density 64
Chapter 3 Table 3.1. The supplements and their amounts in the YC - uracil mixture used in the selection of insertion mutants 83 Table 3.2. Oligonucleotides used in this study 85 Table 3.3. Characterization of the diversity of mutations in the barcode- tagged insertion mutants by 4 phenotypic analyses. 102 Table 3.4. Insertion sites, detailed structures of the integrated insertion vector and adjacent chromosome sequences determined by TAIL-PCR and sequencing 105
Chapter 4 Table 4.1. Fission yeast strains used in this study 120 Table 4.2. Oligonucleotides used in this study 123 Table 4.3. Barcode sequencing of surviving mutants after selection for 14 days in stationary phase and insertion mutations identified from the overrepresented mutants 138 Table 4.4. Barcode sequencing of pool#1 and #2 insertion mutants in the initial culture (day0) for long-lived mutant selection 138 Table 4.5. Identification of S. cerevisiae orthologs of S. pombe Spbc20f10.10p by BLASTP search 150 Table 4.6. Comparison of S. cerevisiae Rim15p and its probable S. pombe orthologs Cek1p and Ppk18p identified by BLASTP search 154
v List of Figures
Chapter 1 Figure 1.1. The effects of oxidative stress on aging 8 Figure 1.2. Mitochondrial metabolism and aging 9 Figure 1.3. TORC1 signaling positively regulates cell growth and negatively 19 regulates longevity Figure 1.4. The insulin/IGF-1 signaling pathway negatively regulates stress resistance and longevity 23 Figure 1.5. Transcriptional silencing regulation by the budding yeast Sir proteins 27 Figure 1.6. Loss of Sir2p results in extrachromosomal rDNA circles (ERCs) formation and aging in budding yeast 28 Figure 1.7. A simplified schematic of the effects of SIR2/sirtuin on lifespan and health 30
Chapter 2 Figure 2.1. Over nutrition shortens S. pombe chronological lifespan in SD medium but lengthens lifespan in EMM medium 54 Figure 2.2. CLS curves in SMM-based medium are multiphasic and show an unusual response to caloric restriction 54 Figure 2.3. Cell density remains constant during the CLS assays in the SD and EMM media 55 Figure 2.4. EMM, but not SD medium, results in the accumulation of a large fraction of cells with sub-1N DNA content 56 Figure 2.5. Cell viability determined by propidium iodide staining and 58 FACS analysis parallels the decline in CFU/ml Figure 2.6. Caloric restriction extends lifespan in S. pombe 59 Figure 2.7. Long-lived S. pombe cells grown in 0.1% glucose maintain a constant DNA FACS profile while aging 62 Figure 2.8. Long-lived calorically restricted S. pombe show increased stress resistance 65 Figure 2.9. Deletion of the AKT kinase gene sck2+ extends lifespan under normal conditions while deletion of the gene for the paralogous kinase sck1+ does not 67 Figure 2.10. Deletion of either AKT kinase sck1+ or sck2+ extends lifespan in over nutrition conditions 68 Figure 2.11. Deletion of tor1+ or gad8+ kinase extends lifespan 70
vi Chapter 3 Figure 3.1. The deletion and tagging strategy used in S. cerevisiae and S. pombe gene deletion mutant collections 80 Figure 3.2. Amplification of insertion vector-genomic DNA junction by TAIL-PCR 92 Figure 3.3. The schematic of the insertion vector used to construct the barcode-tagged insertion mutant library 96 Figure 3.4. Discrimination of different products of linear insertion DNA with an ura4+ marker after transformation into fission yeast 98 Figure 3.5. A selection strategy for cells containing a single copy and multiple copies of ura4+ genes based on the hypothetical metabolic outcome of altered Ura4p levels and low concentrations of 5-FOA on cell survival. 99 Figure 3.6. The flowchart of fission yeast barcode-tagged insertion mutant library construction 100 Figure 3.7. Schematics of insertion mutations identified in this study 104
Chapter 4 Figure 4.1. Hypothetical survival curves of long-lived mutants with different median and maximum lifespans 135 Figure 4.2. Using the S. pombe barcode-tagged insertion mutant library and a novel barcode sequencing strategy to identify lifespan- extending mutations 136 Figure 4.3. Insertion sites determined in two of the isolated mutants form the day 14 culture 140 Figure 4.4. Characterization of the 28S ribosomal RNA (rRNA) insertion mutant by splinkerette PCR 141 Figure 4.5. The chronological lifespan of the three mutants isolated from the parallel selection for long-lived mutants 143 Figure 4.6. A working model of lifespan regulation by the Clg1p/Pef1p cyclin/CDK complex and the S. pombe Rim15p ortholog 145 Figure 4.7. The two-hybrid analysis between Gal4 DNA binding domain (DBD)-Pef1p and Gal4p activation domain (AD)-Clg1p 146 Figure 4.8. Co-immunoprecipitation (IP) identifies a physical interaction between Pef1p and Clg1p in fission yeast cells 147 Figure 4.9. Deletion of clg1+ or pef1+ extends S. pombe chronological lifespan 149 Figure 4.10. A co-IP assay indicates that Pef1p interacts with the cyclin homolog Spbc20f10.10p 151
vii Figure 4.11. Deletion of cyclin homologs psl1+ or pas1+ does not increase chronological lifespan 152 Figure 4.12. Deletion of cek1+ in clg1 background abolishes the lifespan- extending effect of clg1 155 Figure 4.13. Deletion of ppk18+ shortens chronological lifespan in both wild type and clg1-deleted strains 156 Figure 4.14. Pef1p interacts with Cek1p in fission yeast cells 157 Figure 4.15 A proposed model of CLS regulation by Clg1p, Pef1p, and 158 Cek1p
Chapter 5 Figure 5.1. A model of Clg1p-Pef1p-Cek1p-dependent CLS regulation 181
viii Acknowledgements
The completion of this dissertation would not have been possible without all
the support and care that I have received in the past few years. I would first like
to thank my mentor Dr. Kurt Runge for not only teaching me to do research with
a good standard and discipline, but also making sure that I learn to present my
work properly and confidently through so many practice talks before each of my
seminars or meeting presentations as well as countless back-and-forth revisions of
progress reports and meeting abstracts.
The next, I want to thank my committee members Drs. Mark Adams, Peter
Harte, Jo Ann Wise, and Steven Sanders for their insightful advice and challenges
as well as encouragements at our meetings. I also thank Dr. Phil Morgan for the
help he provided before leaving for Seattle.
I would also like express my appreciation to all the current and past
members of the Runge lab for their assistance, support, and friendship. Special
thanks are given to Anna Yakubenko, Becky Shtofman, and Ron Hector for the
nice experience of working together, great food and other fun stuff that we shared.
Finally, here comes the family. I would like to thank my parents Shi-Jiun
Chen and Min-Shu Wang for their unconditional, infinite care and love as well as
supporting every decision that I have made. I am also very grateful to my sister
Yu-Chin Chen and brother Yu-Hao Chen for sharing their life with me even though we are thousands of miles apart all the time. You all have made these years so much more wonderful. Thank you!
ix List of Abbreviations (listed in alphabetical order) Abbreviation Full description 2-DOG 2-deoxy-glucose 4E-BP eIF4E binding protein 5-FOA 5-fluoroorotic acid AMPK AMP-activated protein kinase CDK cyclin-dependent kinase CFU colony forming unit CLS chronological lifespan CR caloric restriction DR dietary restriction E-MAP epistatic miniarray profile EMM Edinburgh Minimum Medium ERC extrachromosomal rDNA circle ETC electron transport chain IGF insulin-like growth factor IGF1-R IGF-1 receptor InR insulin receptor MMEJ microhomology-mediated end joining NAD+ (oxidized) nicotinamide adenine dinucleotide NMEJ non-homologous end joining PKA protein kinase A (cAMP-dependent protein kinase) rDNA ribosomal DNA RLS replicative lifespan ROS reactive oxygen species rpl ribosomal protein large subunit (60S subunit) S6K ribosomal protein S6 kinase SGA synthetic genetic array SIR silent information regulator SLAM synthetic lethality analysis on microarray SOD super oxide dismutase TAIL-PCR Thermal asymmetric interlaced-PCR TOR target of rapamycin TORC TOR complex uORF upstream open reading frame UTR untranslated region
x Identification of Mutations that Extend the Fission Yeast
Schizosaccharomyces pombe Chronological Lifespan by a Novel
Parallel Selection Approach
Abstract
By
BO-RUEI CHEN
A chronological lifespan (CLS) assay was developed to study eukaryotic
cell aging using the fission yeast Schizosaccharomyces pombe. This assay
measures how long cells remain viable in a non-dividing state, allows a
continuous decline in viability without detectable re-growth until all cells in the
culture are dead, and recapitulates the evolutionarily conserved features of lifespan shortening by over nutrition and lifespan extension as well as increased stress resistance by caloric restriction.
Genome-wide studies in the evolutionarily distant yeast Saccharomyces
cerevisiae have uncovered several longevity pathways that are conserved in other
eukaryotes. The current approaches rely on analysis of individual ORF deletion mutants from a large, pre-defined library or microarray analysis of the barcodes associated with these deletions. The creation of this barcoded, ORF deletion
strain set was a major effort that is difficult to extend to other model systems.
Therefore, we constructed a barcode-tagged random insertion mutant library in S.
pombe that allowed for the parallel selection of mutants with extended
chronological lifespan. Mutants selected from this library can be analyzed by
1 routine molecular biology techniques and without prior knowledge of the barcode
sequences.
In a proof-of-principle experiment, ~3,600 barcoded insertion mutants were
selected in a single culture, in which the mutants with the longest lifespans (and
their associated barcodes) increased in proportion as the cells with shorter
lifespans died. By amplifying, oligomerizing and sequencing the barcodes from
the viable cells late in the lifespan, we identified a lifespan-extending insertion
mutation in the cyclin gene clg1+/mug80+. Complete deletion of the clg1+ ORF
also extended lifespan. Clg1p, Pas1p and Psl1p are three cyclins that associate
with the CDK Pef1p, but only loss of Clg1p extended lifespan. Clg1p/Pef1p is
homologous to the Pho80p/Pho85p complex in S. cerevisiae, which promotes
entry into quiescence through the Rim15p kinase. We found that the increased
longevity of clg1 cells requires the Cek1p kinase, a homolog of Rim15p. Thus,
long-lived mutants affecting evolutionarily conserved pathways can be directly
selected from a pool of random S. pombe mutants. As Pef1p and Pho85p are similar to human CDK5, this longevity pathway may function in metazoans as well.
2 Chapter 1
Background, significance and specific aims
3 Aging, lifespan and longevity: a general introduction
Aging is a life-long decline in biological functions due to accumulation of molecular and cellular damage imposed by the collective actions of intrinsic and extrinsic factors on living organisms, and has a terminal consequence of mortality. Although the lifespans of humans are measured chronologically from birth to death, they can be considered as the combination of two types of cellular lifespans: the replicative lifespans (RLS) of continuously/frequently dividing cells and the chronological lifespans (CLS) of post-mitotic, terminally differentiated cells. RLS is the number of mitotic divisions a cell (e.g. fibroblasts and hemopoietic cells) can undergo before senescence, and CLS is the length of time a non-dividing cells (e.g. neurons and skeletal muscles) can remain viable. To discover the underlying mechanisms of organismal aging, it is important to know how these two types of cellular lifespans are regulated.
While studies related to human aging and aging-related diseases are the major focus of gerontology, the complexity of human bodies and the long human lifespan make research on the biology of human aging quite challenging. Model organisms such as yeast, worms and flies have therefore been extensively utilized because of their short lifespans, their powerful molecular genetics and the conserved molecular and genetic pathways they share with humans and other mammals [81, 194, 269].
Work in model organisms has identified several conserved genetic and environmental interventions that control aging [81, 269, 276] (Table 1.1 and 1.2) .
These systems have also facilitated studies of potential therapeutic compounds
4 Table 1.1. Non-genetic interventions that extend model organism lifespan Budding Known/major affected Treatment Flies Worms Mammals yeast genes/pathways Caloric TOR, insulin/IGF-1, sirtuins, PKA [81] restriction v v v v Resveratrol v v v v* Sirtuins [14, 134, 376] Rapamycin v v v TOR [20, 119, 278] Caffeine v TOR [368] Spermidine v v v Autophagy [64] * The lifespan extending effect was observed in an obese mouse model.
Table 1.2. Evolutionarily conserved genes mentioned in this chapter whose orthologs have been implicated in the regulation of aging and longevity Orthologs in model organisms* Budding Fission Genes product yeast yeast Flies Worms Mice sir2+, hst2+, Sirtuin SIR2, HST1-4 dSir2 sir-2.1 SIRT1-7 hst4+ ROS scavenging activity Superoxide sod1+, Sod, Sod2, Sod3, SOD1, SOD2 sod-1-5 SOD1-3 dismutase SPAC1486.01 Sodq, CG30128 catalase CTT1, CTA1 ctt1+ dCat ctl-1, 2, 3 CAT TOR signaling pathway TOR kinase TOR1 tor1+ dTor let-363 mTOR GTPase activating NA tsc1+, tsc2+ dTsc1, dTsc2 NA TSC1, TSC2 protein (GAP) S6 kinase SCH9 ND dS6K rsks-1 S6Ka-c Insulin/IGF-1 signaling pathway Insulin/IGF-1 NA NA dInR daf-2 InR, IGF1-R receptor Insulin/IGF-1 NA NA Chico ist-1 IRS1-4 receptor substrate PI3 kinase ND ND Dp110 age-1 PI3K AKT SCH9 sck1+, sck2+ dAkt akt-1, akt-2 AKT1-3 Forkhead ND ND dFOXO daf-16 FOXO1, 3, 4, 6 transcription factor NA: orthologous genes are not annotated in the genome. ND: sequence homologs exist; orthologous genes have not been determined. * Information in this table was obtained from NCBI (http://www.ncbi.nlm.nih.gov), SGD (http:// www.yeastgenome.org) , S. pombe GeneDB (http://old.genedb.org/genedb/pombe/) as of July 2010 and review articles [188, 249]
5 such as resveratrol, rapamycin and spermidine, which have clinical value in the treatment of aging-related diseases [150, 152] (Table 1.1).
Despite the fast-accumulating discovery of aging- and longevity-related
genes, questions regarding which processes are universal and which processes are
specific to model systems remain. To aid in the identification of evolutionarily
conserved core mechanisms that determine lifespan, this thesis describes a series
of experiments that develop the fission yeast Schizosaccharomyces pombe as a
novel aging model system. S. pombe has a sequenced, well-annotated genome
and well-established genetics. Phylogenetic studies have shown that S. pombe is
quite distinct from other model systems and shares more molecular and cellular
similarity with metazoans than the more popular yeast system, Saccharomyces
cerevisiae [124, 326]. Thus, S. pombe aging studies may identify lifespan
regulators conserved in metazoans, but not in S. cerevisiae. This chapter
reviews some of the well-studied conserved aging and longevity regulators, and
the significance and specific aims of this thesis.
Aging: an evolutionary perspective
Early hypotheses suggested that aging, like other biological processes (e.g.
development), is a genetically programmed event in organisms designed to
promote senescence and death and aiming to eliminate old and unfit individuals
considered as harmful threats to a strong and sound population [180, 181].
However, it was soon argued that in many species, individuals living in the wild
rarely survive long enough to allow such “aging genes” to manifest their effects
6 and therefore be preserved during evolution [180, 181]. The current view of
aging favors that a life-history trade-off strongly affects longevity by balancing
the amount of metabolic resources allocated to growth, somatic maintenance and
reproduction in different stages of lifespans to ensure continuous survival and
propagation of the population [180, 181]. According this idea, natural selection preserves a combination of genetic programs that provide an optimal chance for a species to grow, reproduce and survive in a specific environment/habitat.
Therefore, aging and longevity are controlled by genes modulating proliferation, development, metabolism and other biological activities required for maintaining life and proper interaction with the environment, rather than “aging-facilitating genes” whose activities specifically manifest in old individuals.
The free radical theory of aging
The free radical theory of aging, first proposed by Denham Harman in the mid-1950s [118], argues that one of the main causes of aging is the production of free radicals and other types of reactive oxygen species (ROS) during normal metabolic processes, and the subsequent oxidative damages to DNA, protein, lipids and other cellular components caused by these active molecules [25, 332]
(Figure 1.1).
When transported along the electron transport chain (ETC) embedded in the inner mitochondrial membrane, electrons leaking from ETC components oxidize
¯ O2 to form the highly reactive product superoxide anion (O2 ), which can be
¯ converted to hydroxyl radical (OH ), hydrogen peroxide (H2O2) and other types
7
Figure 1.1. The effects of oxidative stress on aging. Oxidative stress (e.g. reactive oxygen species (ROS)) generated from mitochondrial and other metabolisms or extrinsic sources (e.g. UV, ionizing radiation) can cause damage to DNA, proteins, lipids and other cellular components. Accumulation of these damage results in dysfunction of tissues and organs during organismal aging which eventually cause death. Dietary/ caloric restriction or low levels of stress can attenuate the deleterious effects of oxidative damage through reducing the levels of ROS and repairing damaged molecules.
of ROS [313, 364] (Figure 1.2). Because of its proximity to and its role in the
generation of ROS, mitochondrial DNA is one of the major targets of ROS and
increasing mitochondrial DNA damages have been observed with increasing age
across species [183, 227, 304, 387]. Functionally compromised mitochondria can even be a bigger source of ROS, which augments cellular oxidative damage and results in accelerated aging and shortened lifespan [27, 140, 349, 389, 393].
In addition to DNA, oxidative damage on proteins can cause oxidative cleavage of the polypeptide backbone, generation of protein-protein cross links and, most frequently, protein carbonylation [335]. Similar to mitochondrial DNA lesions,
8 levels of oxidized protein have also been shown to increase with age and in pre- mature aging diseases [240, 261, 385]. Several studies have also demonstrated that oxidative damage to lipid and RNA not only has functional consequences at molecular levels but also are the causes of some pathological disorders [110, 254,
255, 266, 345].
Figure 1.2. Mitochondrial metabolism and aging. In oxidative phosphorylation, electrons are transferred on the electron transport chain (ETC) to create proton gradients cross the inner mitochondrial membrane, which is utilized in ATP production
by F1F0-ATP synthase. In normal metabolisms, electrons may leak from the ETC and
oxidize O2 to generate superoxide anion and other types of ROS, which gradually promote aging. While severe mitochondrial dysfunction facilitates ROS production and rapid aging, specific mutations in some ETC components that partially reduce mitochondrial efficiency have been shown to increase lifespan in yeast and worms. The underlying mechanisms of lifespan extension by these specific mutations remain to be determined. CR also prolongs lifespan through mitochondria by mechanisms such as increased respiration rate, energy production efficiency and stress resistance.
9 Since ROS appears to be an inevitable byproduct of normal metabolism, several protective mechanisms have evolved to shield cells from their harmful effects. Enzymatic ROS scavengers such as Cu, Zn- and Mn-superoxide dismutase (SOD) can process superoxide anions to H2O2, which is then converted to H2O by catalase and glutathione peroxidase. Other scavengers such as glutathione reductase and thioredoxin reductase help maintain the reduced state of small molecular antioxidants and protein thiols [386]. Consistent with their molecular functions, it has shown that increased expression of Cu, Zn-SOD and catalase in flies not only decreased protein oxidation, but also resulted in lifespan extension [263]. Similarly, budding yeast cells overexpressing Sod1p or Sod2p had longer lifespan and SOD2 was also required for the lifespan extending phenotype of deletion of the protein kinase coding gene SCH9 [73]. Transgenic mice overexpressing human catalase in mitochondrial have also been reported to exhibit long lifespan [309]. On the other hand, down-regulation of SOD activity by gene deletion or RNAi have been shown to augment oxidative damages and reduce lifespan in yeast, flies and mice [65, 179, 189, 195, 209, 275, 370]. The strong correlation between stress resistance and longevity has led to successful identification of long-lived mutant yeast, worms and files [6, 77, 244].
It has to be noted that despite the generally accepted strong correlation between increased resistance to oxidative stress and longevity, the relationship between oxidative stress and aging has quite some complexity. For example, overexpression of catalase and Cu, Zn-SOD in mice was found to have no effect on lifespan, even though cultured embryonic fibroblasts from Cu, Zn-SOD
10 transgenic animals showed increased resistance to superoxide-generating
compound paraquat [271, 272]. The same was also observed in Mn-SOD and
other ROS scavenging enzyme transgenic mice [142, 272]. In C. elegans,
deletion of most of sod isoforms reduced resistance to oxidative stress but not
lifespan [60], and deletion of sod-2 (Mn-SOD) even promoted longevity [355].
These results indicate that whether increased levels of antioxidants extend
lifespan appears to be context-dependent and the length of lifespan is not solely
dependent on resistance to oxidative stress. In addition, ROS and other types of
oxidants in metazoans also function as activators of a variety of signaling
molecules and pathways that regulate many biological processes, including
proliferation, stress response and apoptosis [78]. Therefore, reduction in oxidant
levels not only can alleviate cellular oxidative damage but also may affect activity
of these signaling pathways. Reaching a balance between these two aspects are
important for optimal survival for metazoans.
Caloric restriction, stress management and lifespan
Caloric restriction (CR), also termed dietary restriction (DR), is a nutritional
intervention in which reduction of calorie or dietary intake without malnutrition extends organismal lifespan. Although the relation of longevity and food limitation had first been documented in spider in late 19th century [148], the
earliest experimental demonstration of a delay of aging-associated phenotypes
and lifespan extension upon CR was reported by McCay et al. in 1935 [224].
Since then, the effects of CR have been repeated in many other model organisms,
11 including yeast, flies, worms and fish [111]. Similar studies have been extended
to primates and strongly suggest that rhesus monkeys under a CR diet (~30%
reduced food intake than normal diet) exhibit reduced or delayed incidence of
age-related functional decline and diseases, and have longer lifespan than control
monkeys on a normal diet [3, 46]. Although it is not known whether CR extends
human longevity, individuals under CR diet show beneficial metabolic, hormonal and functional changes, as well as reduced risk to aging-related diseases, such diabetes, cardiovascular diseases and cancer [80].
In model organisms, CR can be achieved by many different methods (e.g. different feeding methods and limitation of different nutrients) and activate different signaling pathways [59, 104]. For example, glucose limitation generally leads to inhibition of Ras/cAMP-dependent kinase (PKA) pathway, insulin signaling and activation of AMP-activated kinase (AMPK) and sirtuins.
On the other hand, amino acid restriction results in down-regulation of the target of rapamycin (TOR) signaling pathway [59]. By tuning the activities of these signaling pathways, specific genes can be induced or repressed to allow proper adaptation to different the nutrient-limited conditions to achieve maximal survival at organismal levels.
Among many underlying mechanisms of CR, increased stress resistance appears to be conserved in most organisms [332]. For example, increased resistance to oxidative stress elicited by paraquat treatment or dysfunctional mitochondria, and elevated thermal tolerance were detected in long-lived calorie
restricted C. elegans [161, 190]. In rodents and monkeys, CR can attenuate age-
12 associated increases in ROS production and oxidative lesions [217, 332, 388].
Similar phenomena have also been observed in unicellular organisms. For
example, glucose-restricted yeast cells were found to be more resistant to heat
shock and H2O2 than cells grown in standard growth medium with normal glucose
concentrations [39, 95, 392]. Recently, this property of CR has been applied to
improve cancer treatment by protecting normal cells, but not cancer cells, from
oxidative stress induced by chemotherapeutic agents [281].
Accumulating evidence indicates that the increased stress resistance is likely the result of elevated detoxification, damage repair activity and reduction of ROS
generation [228, 322, 332] (Figure 1.1). Consistent with this idea, microarray
analyses on gene expression across species in response to reduction in dietary
intake have constantly found increased expression of stress response genes [114].
In addition to reducing the levels of free radicals and other ROS, it is
equally important that cells have strategies to purge damaged macromolecules and
organelles. Autophagy is one of the mechanisms by which eukaryotic cells
degrade and recycle macromolecules and organelles (e.g. oxidized proteins and
impaired mitochondria) through vacuoles or lysosomes [306]. Consistently, CR
has been reported to induce autophagy in rodents [49], and autophagy is required
for lifespan extension induced by CR in C. elegans [145]. Unlike protein, lipid
or some organelles which can be discarded and easily replaced, DNA and the
entire genome do not have this option, and DNA mutation and genomic instability
have been shown to increase with age and correlate with cancer incidence [54].
In this regard, CR can promote DNA damage repair and genome stability through
13 increasing expression of genes involved damage repair, and in some cases the activity and fidelity of certain DNA polymerases [128].
The hormesis hypothesis has been proposed to explain the effects of CR on metabolism, stress response and lifespan, and argues that stressors which are harmful at high intensities or dosages can bring beneficial outcome when provided at low levels. As long as the stress levels are tolerable, these stressors act as extracellular stimuli which activate signaling pathways (e. g. induction of protective proteins that scavenge ROS, chaperones for proper protein folding and modification) that counter the deleterious effect of the same stress at higher levels so that organisms are better protected in the future [90, 322]. Consistent with this idea, severe starvation and overfeeding are widely known to be detrimental while CR extends lifespan [277]. Other mild stresses (e.g. heat, osmotic stress, low-dosage radiation) have also been reported to increase lifespan in worms and yeast [147, 153, 384].
Mitochondrial activity, metabolism and aging
Another major effect of CR that has been observed across species is the promotion of mitochondrial activity (Figure 1.2). Lin et al. first reported that
CR in budding yeast increased lifespan in part by promoting respiration [202] and mutations compromising mitochondrial functions (e.g. deletion of CYT1
(cytochrome c1) or LAT1 (a subunit of pyruvate dehydrogenase complex)) abolished CR-induced lifespan extension [62, 202]. Barros et al. later showed that CR in yeast not only enhanced mitochondrial respiratory activity but also
14 reduced ROS production [10]. In C. elegans, 2-deoxy-glucose (2-DOG, a non metabolizable sugar and an alternative CR regimen in worms) feeding also increased respiration and overall stress resistance, although increased ROS production was observed [312]. The elevated levels of ROS and increased stress resistance in 2-DOG-fed worms are consistent with the prediction of hormesis.
CR has also been shown to promote mitochondrial biogenesis and increase ATP production efficiency in cultured mammalian cells and mice, although one study showed an accompanying increased oxygen consumption and the other detected a decrease in respiration rate [211, 252]. These differences in respiration rates might result from the different CR methods used.
Mitochondria are not only the generators of most of the energy in cells, but also the biggest contributor of ROS through of the ETC [364] (Figure 1.2).
Mitochondrial dysfunction caused by mitochondrial DNA mutations and impairment of mitochondrial activity augments ROS levels and is the hallmark of aging and aging-related pathologies [109, 352]. Zuin et al. showed that in fission yeast, deletion of genes encoding mitochondrial ETC components reduced mitochondrial respiration, increased ROS production and shortened lifespan
[393]. De-regulation of mitochondrial protein expression by mutations in the mitochondrial RNA polymerase coding gene RPO41 also shortened budding yeast lifespan [27]. In C. elegans, a mutation in the mev-1 gene, which encodes a subunit of the enzyme succinate dehydrogenase cytochrome b in complex II of the
ETC, also resulted in high oxidative stress and short lifespan [140]. Similarly, targeted-deletion of frataxin, a protein involved in the biosynthesis of
15 intramitochondrial iron-sulfur clusters and reduced expression of which in humans causes a neurodegenerative disorder named Friedreich Ataxia, impaired mitochondrial function in mouse hepatocytes, increased tumor formation and shortened lifespan [349]. RNAi-mediated inhibition of the frataxin homologue frh-1in C. elegans impaired respiration and reduced lifespan as well [389].
Although many mutations affecting mitochondrial functions result in deleterious phenotypes and reduced lifespan, it was noticed that in C. elegans,
RNAi knockdown of some mitochondrial genes can produce the opposite effects on the length of lifespan in different experiments [193, 358, 360]. For example,
Ventura et al. reported that RNAi-mediated inactivation of the frataxin-coding gene frh-1 by feeding worms with bacteria carrying the RNAi construct increased lifespan [360]. However, a shortened lifespan was observed when the same gene
was knocked down by injecting the RNAi construct to worms in a study carried
out by Vázquez-Manrique et al. [358]. Similarly, orthologous mitochondrial
genes whose mutations in humans result in life-shortening diseases have been
found to increase lifespan in worms when inactivated by RNAi knockdown [15,
37, 99, 193, 360, 361].
To address these discrepancies, Rea et al. used a novel RNAi dilution
approach to test the consequences of inactivating six different genes encoding
mitochondrial proteins (i.e. atp-3, nuo-2, isp-1, cco-1, mev-1, and frh-1). They found that generally there was a narrow window of RNAi concentrations and reduction of gene expression that could extend the lifespan of these animals.
RNAi expressed at higher concentrations further reduced the expression levels of
16 these mitochondrial proteins and resulted in gradually shortening lifespan [282].
This similar phenomenon was also observed in budding yeast F1F0-ATPase - subunit ATP2 gene. The atp2 mutant had a shortened lifespan. However, cells expressing reduced levels of Atp2p (i.e. from a weak promoter) had a lifespan longer than the wild type cells (Runge, K.W., Yakubenko, A. and
Shtofman, R., in preparation).
The different consequences between complete (or severe) loss and moderate reduction of mitochondrial functions are consistent with the theory of mitochondrial threshold effects [292] (Figure 1.2). In this theory, as long as the impairment of mitochondrial function (e.g. mutations or reduced expression of mitochondrial proteins) is below a detrimental threshold, cells have the ability to offset and compensate the functional decline by altering metabolisms or other cellular activities (e.g. a shift from oxidative phosphorylation to glycolysis for
ATP production) [292]. However, if the damage is above the threshold, deleterious effects will manifest and result in reduction of fitness and lifespan, as observed in the above worm (high concentration of RNAi) and yeast (complete loss of Atp2p) experiments. The mechanism of lifespan extension by this mildly reduced expression of mitochondrial proteins is yet to be determined.
Down-regulation of the target of rapamycin (TOR) signaling and longevity
Several nutrient-sensing kinases and associated pathways have been implicated in the regulation of lifespan. Among them, the protein kinase target of rapamycin (TOR) has been shown to affect lifespan in yeast, worms, flies and
17 mice. Tor kinase was first identified in the budding yeast S. cerevisiae from a
genetic screen for mutants resistant to the immunosuppressant compound
rapamycin [125]. Tor kinase is the catalytic subunit of two structurally and
functionally distinct complexes, TOR complex 1 (TORC1) and TOR complex 2
(TORC2). TORC1 is the primary target of rapamycin, which inhibits TORC1 as
a complex with the FKBP12 protein [208]. TORC2 is generally insensitive to
rapamycin although it has been shown that prolonged treatment of rapamycin has
an inhibitory effect on TORC2 activity [141, 303].
TORC1 regulates many growth related processes, including transcription,
translation, cell growth and autophagy in response to changes in the levels of
nutrients and growth factors, whereas TORC2 is implicated in actin cytoskeleton
reorganization [7, 17]. Among the identified TORC1 substrates, translation
initiation factor 4E binding protein (4E-BP) and ribosomal protein S6 kinase
(S6K) are the two most extensively studied targets for their roles in regulating
translation, ribosome biogenesis and cell growth (Figure 1.3). Phosphorylation
of 4E-BP by TORC1 leads to its dissociation from eIF-4E and promotes
translation of 5’ capped mRNAs. S6K phosphorylation by TORC1 is important
for the synthesis of ribosomal proteins and translational elongation factors [212].
Inactivation of TORC1 activity has been shown to increase lifespan in many model organisms (Figure 1.3). Unlike most other eukaryotes where Tor kinase is encoded by a single gene, S. cerevisiae has two Tor kinases encoded by TOR1
and TOR2 genes with Tor1p and Tor2p as the catalytic kinases of TORC1 and
TORC2, respectively [208]. Deletion of TOR1 gene, genes involved in the TOR
18 Figure 1.3. TORC1 signaling positively regulates cell growth and negatively regulates longevity. Activation of TORC1 promotes S6K activity, inhibits 4E-BP and results in increased translation and ribosomal biogenesis, conditions favoring cell proliferation. On the other hand, inhibition of TORC1 by mutations, increased levels of the inhibitory Tsc complex or rapamycin treatment increases stress resistance, autophagy and mitochondrial biogenesis, which are thought to be the underlying mechanisms of lifespan extension by TROC1 inactivation. It should be noted that Tsc1 and Tsc2 orthologs are not present in all model systems (Table 1.2). This diagram is compiled from review articles [164] and [115].
pathway or rapamycin treatment conferred long lifespan [26, 160, 278].
Budding yeast cells grown in caffeine-containing medium also had reduced
TORC1 activity and extended lifespan [368]. In C. elegans, RNAi knockdown
of Tor kinase or TORC1-specific subunit raptor also increased lifespan [144,
359]. In flies, inhibition of TOR signaling by overexpression of a dominant negative allele of dTor, the negative regulators of TOR, dTsc1 and dTsc2, or rapamycin feeding demonstrated similar lifespan extending effects [20, 165].
19 Recently, Harrison et al. showed that mice fed with rapamycin lived longer than
the control cohort, further demonstrating the evolutionarily conservation of TOR
signaling-mediated lifespan regulation [119].
Protein translation and homeostasis have been proposed to be a key factor
that affects cellular and organismal aging [155, 156, 172] (Figure 1.3).
Consistent with the role of TOR in the synthesis of ribosomal components,
deletions of genes encoding ribosomal structural proteins in yeast and worms extended lifespan in the same way as inactivation of TOR signaling [50, 117, 160,
337]. Similarly, RNAi knockdown of several translation initiation factors and
S6K in C. elegans not only reduced overall translation rate but also increased lifespan and stress resistance [50, 117]. Inhibition of S6K activity in flies and mice has also been reported to promote longevity [165, 315]. In the case of
S6K-deficient mice, these animals showed reduced symptoms of age-related pathologies, including bone, immune and motor dysfunction, insulin insensitivity, and they were resistant to age- and diet-dependent obesity [315, 354].
While suppressing overall translation is a reasonable response to nutrient limitation, it has been proposed that the some beneficial effects of down- regulating TOR signaling and ribosomal abundance are the results of increased expression of specific genes whose activities are required for optimal survival in response to environmental changes (Figure 1.3) [337, 390]. A systemic analysis of yeast deletion mutants of genes encoding ribosomal 60S subunit proteins (rpl mutants, for ribosomal protein large subunit) found that the expression of the
nutrient-responsive transcription factor Gcn4p was elevated in many rpl mutants
20 and the long-lifespan of phenotype of these rpl mutants required the expression of
Gcn4p [337]. The translation of GCN4 mRNA is known to be controlled by four
small upstream open reading frames (uORFs) in the 5’ untranslated region (UTR)
of GCN4 mRNA [129]. Because the budding yeast genome has duplicated
copies of many ribosomal protein coding genes [94], these deletions likely
resulted in reduced abundance, but not completely loss of mature ribosomal 60S
subunits. Therefore, Steffen et al. proposed that the reduced abundance of the
60S subunit may help the ternary complex containing initiation factors and a 40S
subunit pass through the four inhibitory uORFs and reach the GCN4 ORF before
binding a 60S subunit and formation of an active ribosome. In this way, the
access of active ribosomes to GCN4 coding sequence becomes more efficient and
results in increased Gcn4p translation, up-regulation of its downstream targets and
long lifespan [337].
Alternatively, increased expression of specific proteins can be achieved by
up-regulation of general translation inhibitors. For example, induction of
translation inhibitor 4E-BP has been shown to increase, but not reduce the
translation of some mitochondrial genes and enhance mitochondrial activity in
Drosophila upon CR. Close examination of the 5’ UTRs of these up-regulated
genes revealed that they are in general shorter, have weaker secondary structures, and have lower GC content than other genes. The easily accessible 5’UTRs may provide advantages of loading eIF-4E and promoting translation of these mitochondrial genes compared to other genes in the presence of increasing levels of 4E-BP [390]. The long-lived budding yeast tor1 mutant has also been
21 reported to show increased mitochondrial protein expression and respiration [26], although it was not determined whether yeast 4E-BP (encoded by EAP1) is responsible for this altered translational activity. Nonetheless, these two results suggest a conserved genetic interaction between TOR signaling and mitochondria which regulates metabolism, growth and lifespan.
Autophagy is a regulated process that promotes degradation of proteins, organelles and other cellular components in lysosomes (vacuoles in yeast) and is induced in part through inhibition of TOR signaling, subsequent activation of
Atg1 kinase and autophagosome formation [57, 149]. Recent work has linked autophagy to lifespan determination [60] (Figure 1.3). For example, lifespan extension by RNAi knockdown of TORC1-specific subunit raptor and rapamycin treatment in worms and yeast, respectively, is abolished by inactivation of autophagy [2, 116]. In addition, autophagy is activated by CR, rapamycin, resveratrol or spermidine, and these treatments extend lifespan in yeast, worms, flies and improve survival of cultured mammalian cells [2, 64, 236, 238].
The insulin/insulin-like growth factor (IGF)-1 signaling
The insulin/insulin-like growth factor-1 (IGF-1) pathway is one of the most conserved signaling pathways that regulate nutrient and energy homeostasis at organismal levels. Upon ligand binding, the insulin/IGF-1 receptor activates its downstream signals through phosphoinositide 3-kinase (PI3 kinase) and
AKT/PKB kinase, eventually resulting in phosphorylation and inhibition of the forkhead transcription factor FOXO [300, 346] (Figure 1.4). Inhibition of this
22 pathway promotes FOXO nuclear translocation and activates, or inhibits, FOXO- dependent transcription [29, 246, 346].
Figure 1.4. The insulin/IGF-1 signaling pathway negatively regulates stress resistance and longevity. Upon ligand binding and activation, the insulin/IGF-1 receptor in worms, flies and mammals activates its downstream effectors, including PI3 kinase and AKT kinase. Activated AKT kinase phosphorylates the forkhead transcription factor FOXO, whose activity is important for the expression of genes required for stress response and survival. Consequently, inhibition of insulin/IGF-1 signaling by elimination of the components in this pathway leads to activation of FOXO and increasing stress resistance and longevity. Downstream components of insulin/IGF-1 pathway have also been identified in the budding yeast S. cerevisiae, including the AKT kinase Sch9p. Depletion of Sch9p has also been reported to extend yeast lifespan and stress resistance. The pathways shown in this figure are compiled from review articles [249] and [269].
23 The effects of perturbations of insulin/ IGF-1 signaling on organismal
lifespan and the underlying mechanisms have been best studied in C. elegans and
D. melanogaster and shown to be conserved in mice (Figure 1.4). In C. elegans,
mutations in the genes coding the insulin/IGF-1 receptor daf-2, the downstream
PI3 kinase age-1 or RNAi knockdown of the AKT kinase coding gene akt-1
promote lifespan extension that depends on the transcription factor DAF-16 [87,
113, 174, 178, 197]. Mutations that inactivate components in this pathway (i.e.
InR (insulin/IGF-1 receptor), chico (insulin receptor substrate-like protein)) or
overexpression of dFOXO were also found to increase Drosophila lifespan [43,
138, 347, 353]. It is also worth noting that genes homologous to the components of the insulin/ IGF-1 pathway may also regulate yeast lifespan (Figure 1.4). The budding yeast Sch9p has high homology to the AKT kinase and elimination of
Sch9p not only affected cell growth but also extended yeast lifespan [77, 160].
The association of longevity with reduced insulin/IGF-1 signaling was also observed in mammals. In flies and worms, there is only a single receptor that mediates insulin/IGF-1 signaling. However, in mice and other mammals, the corresponding pathway is control by the insulin receptor and the IGF-1 receptor,
which are activated in response to changes in the levels of blood sugar and growth
hormones, respectively (Figure 1.4) [29]. Mice with inactivating mutations that affect IGF-1 signaling (e.g. mutations in Pou1f1, Prop1, growth hormone receptor and growth hormone receptor binding protein) or insulin signaling (e.g. mutations in insulin receptor or insulin receptor substrate 1) have extended lifespan and delayed onset of age-related phenotypes [11, 22, 81, 314]. In dogs, the body size
24 is known to inversely correlate with the length of lifespan [105]. Recently, a specific IGF-1 gene haplotype was found to associate with small breed dogs and most of these relatively long-lived small dogs have lower levels of serum IGF-1 levels [63, 173, 342].
Although there is no direct evidence implicating that inactivation of the insulin/IGF-1 pathway delays aging in primates, accumulating evidence in population genetics suggests that the longevity-promoting effects observed in model organisms may also be conserved in human aging as natural mutations and
DNA variants in the components of this pathway are found associating with many centenarian populations around the world [173, 175].
To understand the molecular mechanism of lifespan extension by inactivation of insulin/IGF1 signaling, two independent microarray studies in C. elegans have identified several candidate effectors of this pathway whose expression is affected by both DAF-2 and DAF-16 [192, 246]. Among those induced in daf-2 RNAi but repressed in daf-2 RNAi daf-16 RNAi animals are genes encoding stress response and antimicrobial proteins, suggesting that elevated resistance to oxidative damage and the lethal effect from the accumulated bacteria and other pathogens in old worms are important in determining C. elegans lifespan. Consistent with this expression profile, daf-2 mutants have been shown to be more resistant to oxidative damage than wild type animals [58] and RNAi knockdown of genes in these two groups (stress response and antimicrobial) in a daf-2 RNAi background partially reversed the longevity phenotype [246].
25 Sir proteins (sirtuins) and aging: more than just silencing transcription
Sirtuins are evolutionarily conserved proteins that can be found in
organisms from bacteria to mammals [356]. Sirtuins have one or both of the
following activities: NAD+-dependent deacetylase or ADP-ribosyltransferases and
the NAD+ dependence of sirtuins has their activities closely regulated by the
metabolic states of cells [112, 356]. While the substrate of the budding yeast
sirtuin Sir2p appears exclusively to be histones, studies of mammalian sirtuins
have identified histones and several non-histone proteins (e.g. p53, FOXO, NFκB,
glutamate dehydrogenase) as their substrates [112]. Different sirtuins also have
a wide distribution of subcellular localizations (i.e. nuclei, cytoplasm and
mitochondria) and are proposed to act as a central regulator to integrate
intracellular and extracellular signals to regulate a variety of cellular and
physiological processes (e.g. apoptosis, fat metabolism, insulin secretion, DNA damage repair) [112].
The molecular mechanisms of sirtuins in lifespan regulation were first investigated in the budding yeast S. cerevisiae. The budding yeast Sir (silent
information regulator) proteins are encoded by four genes (SIR1-4) and were
initially characterized as proteins involved in transcriptional silencing (Figure
1.5). All four Sir proteins have been shown to regulate transcription silencing at
the mating type loci, loss of which results in sterility in haploid cells [284]. At
telomeres, Sir2p, Sir3p and Sir4p regulate transcriptional activity in telomere-
proximal region which renders positional effect variegation (PEV) of gene
expression [98]. The sirtuin protein Sir2p also acts at the ribosomal DNA
26 (rDNA) loci where Sir2p regulates the transcription of rRNA and represses recombination of the rDNA repeats [31, 96, 330].
Figure 1.5. Transcriptional silencing regulation by the budding yeast Sir proteins. In budding yeast, Sir2p, Sir3p and Sir4p are required for the maintenance of transcriptional silencing at telomeres and mating type loci. Silencing at mating type loci also involves Sir1p although it has been shown the Sir1p only plays a minor role. Sir2p has an additional function at the repeated ribosomal DNA (rDNA) loci, where Sir2p represses recombination of rDNA repeats and maintains transcriptional silencing.
The role of Sir proteins in lifespan regulation in budding yeast was reported in the early 1990s when a hypermorphic sir4 allele was shown to increase RLS by
~30% in a semi-dominant manner accompanied with loss of silencing at telomeres and the mating type loci [170]. Later experiments revealed that the lifespan-extending sir4-42 mutation also caused re-distribution of Sir proteins to the nucleolus, where the ribosomal DNA (rDNA) repeats are enriched, and implied that Sir complex might regulate yeast lifespan through its activity at rDNA loci [171]. This hypothesis was confirmed when Sir2p was shown to be required for suppressing recombination at rDNA loci and formation of extrachromosomal rDNA circles (ERCs), which had been demonstrated as a
27 senescing factor in the budding yeast [158, 323]. Consistently, overexpression of Sir2p extended lifespan while deletion of SIR2 shortened lifespan and promoted ERC accumulation [158] (Figure 1.6).
Figure 1.6. Loss of Sir2p results in extrachromosomal rDNA circles (ERCs) formation and aging in budding yeast. In young and SIR2 wild type cells, Sir2p distributed at the repeated rDNA arrays represses recombination of rDNA repeats. In old cells or sir2 mutants, the lack of Sir2p promotes rDNA recombination and formation of ERCs, which preferentially accumulate in the mother cells. The origin of DNA replication in the rDNA repeats allows ERC replication and quickly increases its copy number in cells, resulting in rapid aging possibly through competing essential cellular binding proteins.
28 Sir2p was proposed as a mediator of CR-induced lifespan extension in yeast
as Lin et al. showed that CR promoted mitochondrial respiration and concomitant
increase in the levels of the oxidized form of nicotinamide adenine dinucleotide
(NAD+) and the ratio of NAD+/NADH [199, 200, 202]. While NADH inhibits
Sir2p, NAD+ is required for the histone deacetylase activity of Sir2p [200].
Therefore, CR, through increasing NAD+ and reducing NADH, was proposed to
activate Sir2p to suppress ERC formation and extend lifespan. It has to be noted
that there has been debates that the requirement of Sir2p in CR-induced longevity
appears to be strain- and the CR regimen- dependent [154, 157, 159, 201].
Although the Sir2p-dependent ERC formation appears to be a yeast-specific
senescing factor, extensive studies conducted in yeast and other model organisms
have identified additional mechanisms which are likely conserved in other species
(Figure 1.7). The Nyström group provided data showing that during each cell
division, oxidatively damaged proteins were asymmetrically distributed in mother
cells and this biased inheritance of oxidized proteins depended on functional actin
cytoskeleton and Sir2p specifically, but not other Sir proteins [1, 67]. The lower levels of oxidative damage inherited by daughter cells correlated with higher cytosolic catalase Ctt1p activity after cytokinesis in these new born cells and deletion of CTT1 abolished the unequal distribution of oxidized proteins [67].
These results suggest that Sir2p and other sirtuins may regulate aging through controlling the distributions of cellular damage and oxidative stress scavengers.
Another Sir2p-dependent mechanism that may also be conserved in other organisms is related to the effects of its deacetylase activity on gene expression
29
Figure 1.7. A simplified schematic of the effects of Sir2p/sirtuins on lifespan and health. Activation of Sir2p/sirtuins by caloric restriction, resveratrol, NAD+ and other stimuli have been shown to extend lifespan in model organisms such as yeast and worms, and alleviate symptoms of aging-related diseases and disorders, including diabetes, obesity and cancer. Molecular genetic and cell biological studies have identified several probable mechanisms underlying the beneficial effects of up-regulation of Sir2p/sirtuins, including asymmetric distribution of cellular damage, improved genomic stability, stress resistance, cell survival, insulin sensitivity and mitochondrial activity. In the budding yeast S. cerevisiae, Sir2p also suppresses senescing factor ERC formation.
and chromatin structure [51]. Dang et al. showed that late in the lifespan, the abundance of Sir2p decreased, histone H4 lysine 16 quickly became hyper- acetylated at telomeres and overall histone H3 and H4 protein levels were gradually reduced, all of which could be reversed by deletion of the H4K16 acetyltransferase coding gene SAS2. Epistasis tests further proved that over- expression of Sir2p and deletion of SAS2 extended lifespan through the same pathway [51]. Interestingly, mammalian SIRT6 was shown to possess histone
30 H3 lysine 9 deacetylase activity that may regulate mammalian senescence through telomere structure maintenance and NFB-dependent gene expression [168, 231].
Sirtuin also regulates the lifespan of invertebrates. In C. elegans, lifespan
can be extended by overexpression of SIR-2.1 [16, 350]. Increased levels of
SIR2.1 protein promoted the expression of sod-3, a target gene of the forkhead
transcription factor DAF-16, and stress resistance, both of which depended on 14-
3-3 proteins. Consistently, elimination of DAF-16 or 14-3-3 proteins abolished
lifespan extension by SIR-2.1 overexpression [16]. In C. elegans, SIR-2.1 is not
required for lifespan extension by reducing insulin/IGF-1 signaling or CR through
complete food removal [16, 161]. It has also been reported that ubiquitous or
neuron-specific overexpression of Drosophila dSir2 increased lifespan, although
this result remains controversial in the field [286]. Unlike C. elegans SIR-2.1,
dSir2 is required for CR-induced lifespan extension in Drosophila and CR
increased the levels of dSir2 mRNA [286, 287]. Consistent with the above
genetic studies, sirtuin activating compound resveratrol has been shown to extend
lifespan in worms and flies in sir-2.1- and dSir2-dependent manner [376].
Mammals have 7 sirtuin members that involve in a variety of cellular and
physiological functions, including rRNA transcription, DNA repair, cell survival,
cell cycle, energy metabolism, mitochondrial biogenesis, insulin/glucose
homeostasis [111, 112]. Currently, there is no evidence indicating that
overexpression of sirtuins extends mammalian lifespan as it does in yeast and
invertebrate model organisms. However, up-regulation of sirtuin expression or
activity has been shown to alleviate the symptoms of aging-related diseases
31 (Figure 1.7). For example, a common symptom of type II diabetes is insulin
resistance. Ectopic expression of SIRT1, the closest ortholog of yeast Sir2p, in
cultured cells increased insulin sensitivity under insulin-resistant conditions [341],
and SIRT1 transgenic mice showed increased hepatic insulin sensitivity, glucose tolerance and reduced susceptibility to diabetes [8]. Administration of
resveratrol and other small molecule SIRT1 activators also improved the health
and survival of mice on high-fat diet. The treated mice also exhibited enhanced
insulin sensitivity, mitochondrial function and lower plasma glucose levels [14,
187, 233]. In addition, increased SIRT1 expression and activation of SIRT1 by
resveratrol has been reported to suppress intestinal tumorgenesis and breast cancer
cell proliferation in mice [79, 367]. Regardless their effect on the overall length
of lifespan, these results indicate that sirtuins in mammals are important in
maintaining physiological health and potential targets for delaying and alleviating
late-onset diseases.
CR has been reported to up-regulate the expression of SIRT1, SIRT2,
SIRT3 and SIRT6 [45, 163, 280, 318, 365], suggesting that some effects of CR in
mammals may be mediated by these sirtuins. Among the sirtuins, SIRT1 has
been shown to be required for several CR-induced phenotypes. For example, in
a mouse model of Alzheimer’s disease, it was proposed that CR-mediated
reduction of -amyloid content in the brain was through up-regulation of SIRT1
activity (i.e. SIRT1 expression and NAD+ levels), which in turn promoted -
secretase activity [280]. Cohen et al., using cultured mammalian cells,
demonstrated that SIRT1 was required for CR-dependent attenuation of apoptosis
32 induced by overexpression of apoptotic protein Bax [45]. A distinct phenotype
of mice under restricted diet, increased physical activity, was also absent in
SIRT1 knockout mice [40]. The function of other sirtuins in response to CR
remains to be determined.
Advantages of using yeast as a model to study evolutionarily conserved mechanisms of aging
As a unicellular eukaryotic organism, yeast has been proved to be a valuable model in understanding the fundamental molecular and cellular mechanisms of life that are evolutionarily conserved across species [84, 124]. One of the most well-known examples is the identification of the mechanisms and key regulators of cell cycle control by Lee Hartwell and Paul Nurse (along with the biochemical studies by Tim Hunt) in the 1970s, which provided not only the foundation to study the more complex cell division events in metazoans but also a solid ground toward the understanding of cancer formation and treatment [120, 137, 256].
Compared to other popular model organisms, yeast cells are easy to grow and maintain, both technically and economically. There are many powerful genetic and molecular biological tools readily available for the studies of functions of yeast genes as well as identification of orthologs in other organisms
[28, 83]. These advantages have been extended since the genome of the budding yeast S. cerevisiae was completely sequenced [94] and increasingly well annotated in the following years. The sequenced genome of the budding yeast led to the generation of a collection of complete ORF deletion mutants [91, 375],
33 which help the development of many extended approaches that facilitate genome-
wide studies of individual gene functions and genetic interactions (e.g. synthetic
genetic arrays (SGA) [351], synthetic lethality analysis on microarray (SLAM)
[262, 267] and epistatic miniarray profile (E-MAP) [310, 311]).
Aging research field also benefits from yeast genetics and biology. The
relative ease of micromanipulation and the distinct morphologies of the mother
(parent) and daughter (new born) cells enabled early experiments in the budding
yeast S. cerevisiae to show that a given yeast cell could only undergo a limited
number of mitotic divisions before senescence in 1950s [12, 239], and these
observations were eventually extended to the research topic of replicative aging and replicative lifespan (RLS). Yeast aging can also be measured as chronological aging and chronological lifespan (CLS), which is the length of time that cells can remain viable in a non-dividing state, also called stationary phase, quiescence or G0 [74]. In metazoans, organismal aging is determined by both
replicative and chronological lifespan of different cell types and understanding
mechanisms that determine both kinds of lifespan is a prerequisite to prolonging
healthy lifespan and discovering cures for aging-associated diseases. Studying
yeast replicative aging can help elucidate the underlying mechanisms of
senescence of proliferating cell types in mammals (e.g. fibroblasts, hemopoietic
stem cells) which are crucial in repairing injured tissues and replenishing blood
cells. On the other hand, studying yeast chronological aging may reveal
fundamentals of the regulatory processes that help maintain the survival and
health of post-mitotic cells (e.g. neurons, skeletal muscles), whose decline has
34 been implicated in the phenotypes of the elderly, such as neurodegenerative
diseases and sarcopenia. To date, it has become evident that many fundamental
processes required to maintain cell viability, allow cell proliferation, and extend
lifespan are conserved from yeast to mammals, making yeast an important model
system to investigate the regulation of aging and longevity [81, 151].
The fission yeast Schizosaccharomyces pombe as an emerging model to study
aging and longevity
The fission yeast S. pombe was the second fungal species to have its genome
sequenced and annotated [377]. Like S. cerevisiae, S. pombe also possesses
well-established methods that allow easy genetic and molecular biological
analysis [85]. Phylogenic and molecular biology studies have suggested that S.
pombe appears more similar to the last common ancestor of metazoans and fungi
[124, 326] and has more similarities to mammals than S. cerevisiae in several
processes including the centromere structure, telomere function, RNA splicing,
the existence of RNAi-mediated gene silencing and the requirement of mitochondrial genome for survival in wild type cells [166, 167, 237, 320, 377,
380]. In addition, some of the genes that may control lifespan in metazoans and are absent in S. cerevisiae exist in S. pombe (Table 1.2). These facts argue that using S. pombe to study aging will not only recover lifespan-regulating genes and pathways that exist in S. cerevisiae, but also those conserved in mammals and fission yeast, absent in budding yeast.
Despite the difficulty in distinguishing S. pombe progenies produced by
35 symmetric fission, Barker and Walmsley used a pedigree analysis approach to
show that as in budding yeast, replicative aging accompanied with gradual loss of symmetric morphology between divided cells also occurred in fission yeast [9].
The Nyström lab, which previously showed that S. cerevisiae Sir2p plays a role in
the retention of oxidized proteins in the mother cells, applied this assay to S. pombe and obtained a similar result: fission yeast undergoes age-related asymmetric division and asymmetric portioning of damaged proteins, a process that depends on the cell polarity protein Tea1p, functional cytoskeleton and Sir2p
[66]. The large evolutionary distance between S. pombe and S. cerevisiae suggests a conserved role of sirtuins in regulating replicative lifespan through asymmetric damage partitioning [1, 66, 67].
Compared to the replicative aging assay that requires relatively difficult techniques, fission yeast chronological aging assay is easy to perform and research on CLS regulation has started to bloom in the past few years. Roux et al. were among the first to employ a chronological aging assay generally used in budding yeast to show that deletion of pka1+ or sck2+ (encoding the fission yeast
AKT kinase) extended CLS in part due to reduced accumulation of ROS and
increased oxidative stress resistance (pka1 mutant). The conserved role of
ROS scavenging proteins and mitochondria in lifespan determination in fission
yeast have also been examined as well [247, 393]. These early steps
demonstrate the promising role of S. pombe in studying aging and longevity. In
addition to these “conspicuously-titled” aging research work, extensive efforts
made by the Yanagida group on the regulation of entry, maintenance and exit of
36 quiescence (mostly achieved by acute nitrogen starvation) also provide valuable information that is anticipated to facilitate the understanding the biology of fission yeast chronological aging [235, 299, 319, 343, 381].
Significance and specific aims of this study
The success in the budding yeast aging research appears to affect (at least partially) the way to study fission yeast aging (e.g. [297]). However, it is necessary to validate that the assay used in this new aging model can properly recapitulate evolutionarily conserved features of aging and longevity. Also, the newly developed assay should allow its application to genome-wide analysis that facilitates isolating as many lifespan-extending mutations as possible in a cost- and labor-effective manner to advance comprehensive understanding of the genetics of aging.
Recently, a barcode-tagged S. pombe deletion mutant collection was released [176]. Despite its potential in promoting genome-wide studies, some concerns have arisen in the fission yeast research community. For example, a significant number of mutants have large incomplete deletions (i.e. over 600 mutants retain 20% of their coding sequences, and over 200 mutants retain more than 40% [176]), which may not produce loss-of-function phenotypes as expected in an “ORF deletion” mutant and therefore bring in confusion or impede the isolation of certain mutants in some experiments. Also, the lack of a commercialized and readily available microarray for the barcode tags prevents parallel analysis similar to those performed in the budding yeast (e.g. [91]).
37 This dissertation shows how we used S. pombe to study the evolutionarily conserved mechanisms that regulate aging and longevity. To achieve this goal, the following three aims were examined within this dissertation:
Aim #1 (Chapter 2): Development and characterization of a new S. pombe chronological lifespan assay. In this aim, a new S. pombe chronological lifespan assay that allows monitoring the lifespan of a yeast culture until the complete extinction of viable cells was established. This assay was able to recapitulate the evolutionarily conserved responses of the length of lifespan to nutrient levels.
Stress resistance upon caloric restriction and the effects of deletions of the AKT kinase coding genes (sck1+ and sck2+) and two other kinase coding genes in the
TOR signaling pathway on lifespan were also examined.
Aim #2 (Chapter 3): Construction and characterization of a barcode-tagged S. pombe insertion mutant library. To construct a mutant library for large-scale parallel selection for long-lived mutants, an insertion mutagenesis approach was used to create a collection of barcode-tagged S. pombe insertion mutants.
Randomly picked insertion mutants were examined by easily scored phenotypic selections to show the randomness of mutations. The insertion sites of several mutants were determined by thermal asymmetric interlaced (TAIL)-PCR. The insertion structures and their effects on the application of this library were discussed.
38 Aim #3 (Chapter 4): Identification of lifespan-extending mutations using a novel
parallel selection strategy and the barcode-tagged S. pombe insertion mutant
library. The barcode-tagged S. pombe insertion mutant library constructed in
aim #2 was used in a proof-of-principle experiment to demonstrate a parallel selection for mutants that had longer chronological lifespan. A novel barcode sequencing approach was used to score the mutants with the desired phenotypes.
Detailed analysis of one of the isolated mutants led to the discovery a conserved
signaling pathway that appears to regulate chronological lifespan in both budding
and fission yeast.
39
Chapter 2
A new Schizosaccharomyces pombe chronological lifespan assay
reveals that caloric restriction promotes efficient cell cycle exit
and extends longevity
Note: A modified version of this chapter was published as Chen BR and Runge
KW. 2009. A new Schizosaccharomyces pombe chronological lifespan assay reveals that caloric restriction promotes efficient cell cycle exit and extends longevity. Experimental Gerontology 44: 493-502.
40 Abstract
This chapter describes a new chronological lifespan (CLS) assay for the
fission yeast Schizosaccharomyces pombe. Yeast CLS assays monitor the loss
of cell viability in a culture over time, and this new assay shows a continuous
decline in viability without detectable re-growth until all cells in the culture are
dead. Thus, the survival curve is not altered by the generation of mutants or
epigenetic variants that can grow during the experiments, and one can monitor the
entire lifespan of a strain until the number of viable cells has decreased over 106- fold. This S. pombe CLS assay recapitulates the evolutionarily conserved features of lifespan shortening by over nutrition, lifespan extension by caloric restriction, increased stress resistance of calorically restricted cells and lifespan control by the AKT kinases and TOR signaling pathway. Both S. pombe AKT kinase orthologs regulate CLS: loss of sck1+ extended lifespan in the over
nutrition condition, loss of sck2+ extended lifespan under both normal and over
nutrition conditions, and loss of both genes showed that sck1+ and sck2+ control different longevity pathways. Deletions of the tor1+ and gad8+ genes, both encoding protein kinases involved in the TOR signaling pathway in S. pombe, also resulted in evolutionarily conserved lifespan-extending phenotype. The longest-lived S. pombe cells upon caloric restriction showed the most efficient cell cycle exit, demonstrating that caloric restriction links these two processes. This new S. pombe CLS assay will provide a valuable tool for aging research.
41 Introduction
The lifespans of multicellular organisms are most often measured chronologically, i.e. the length of time the entire organism is alive, and are dependent on both the replicative lifespans (RLS) and chronological lifespans
(CLS) of different cell types. RLS is the number of times a cell can divide before division ceases, and is applicable to stem cells such as those that
repopulate the human hematopoietic and immune systems [30, 391]. CLS is the
length of time a cell can survive without dividing, and is applicable to post-
mitotic cells such as neurons or skeletal muscle [207]. Thus, understanding the
molecular mechanisms that control RLS and CLS is important to understanding
human aging.
Reducing the amount of calories in the diet (caloric restriction) is an
intervention that affects both cellular RLS and CLS. Early caloric restriction
experiments demonstrated a delay in the onset of aging phenotypes and a
significant increase in lifespan in rodents [224], and these observations are
recapitulated in several non-mammalian model systems such as the budding yeast
Saccharomyces cerevisiae, the nematode Caenorhabditis elegans and the fruit fly
Drosophila melanogaster [18, 59, 109]. An additional property of these long-
lived organisms is resistance to stress: a higher proportion of long-lived yeast,
nematodes and flies can survive lethal stresses compared to their normal
counterparts [90, 170, 218, 219, 322]. Thus, lifespan extension by caloric
restriction and the increased stress resistance of long-lived cells are evolutionarily
conserved properties of aging in both unicellular and multicellular eukaryotes.
42 In addition to caloric restriction, altering the activities of nutrient sensing
signaling pathways also has broad effects on aging in many organisms [81, 269].
The protein kinase target of rapamycin (TOR) and insulin/insulin-like growth factor (IGF)-1 pathways are among the most well-studied nutrient sensing pathways that also determine the length of lifespan [269]. TOR exerts its function through two distinct complexes, TOR complex 1 (TORC1) and TOR complex 2 (TORC2) to regulate a variety of cellular activities, including translation, cell growth and organization of actin cytoskeleton [7, 17].
Inactivation of TOR, especially TORC1, has been shown to increase lifespan in model organisms such as budding yeast, worms, flies and mice [119, 336]. As to the insulin/IGF-1 signaling pathway, mutations in the insulin/IGF-1 receptors, their ligands or the downstream components, such PI3 kinase and AKT kinase, also promote longevity in flies, worms and mice [249]. The downstream components of insulin/IGF-1 pathway have also been found in yeast, such as the
AKT kinase. Deletion of the budding yeast SCH9 gene, which encodes the AKT kinase ortholog, also increases S. cerevisiae RLS and CLS [77, 160].
Several of the non-mammalian model systems used in aging research have the benefit of shorter lifespans and well-developed genetics compared to human populations. In the cases of nematode and fruit fly CLS and budding yeast RLS, most genetic studies follow the same approaches as survival studies in humans: scores to hundreds of individuals are monitored for survival over a given time period until all individuals are dead. This approach has been successful in identifying evolutionarily conserved genes that affect lifespan such as AKT [268],
43 TOR [336], insulin-related growth factors [198, 243], and sirtuins [108, 322].
The availability of defined mutant collections in model organisms has also been a major benefit in identifying genes whose mutations can increase longevity [113,
160, 193, 278].
The budding yeast S. cerevisiae CLS assays are distinct from those lifespan assays described above in that one assays a far greater number of individuals, i.e. typically >109 cells. In most experiments, cell viability is analyzed as the percent of viable cells remaining after the culture has reached its maximal density.
Viability is usually followed until it has declined to 0.1% to 1% of its original value (e.g. [74, 77, 213]), which allows a simple comparison of yeast CLS assay results with other lifespan assays. Analysis beyond 3 logs can be a problem because large numbers of individuals (106 cells) remain alive and rare variants can arise and re-grow as other cells die [70]. Because the CLS assay measures aging as the decline in the number of viable cells, the generation of a new mutant population of viable, growing cells confounds analysis. The challenge in using this range to identify mutants that alter lifespan has required innovative approaches such as prescreening mutants for stress resistance [77], or analyzing very small cultures of a defined set of mutants (i.e. the gene deletion strain set)
[278].
The utility of the S. cerevisiae system prompted us to examine the feasibility of using the fission yeast Schizosaccharomyces pombe in CLS assays. S. pombe appears to be more similar to the last common ancestor of humans and fungi [124,
326], has more similarities to humans than S. cerevisiae in several processes
44 including RNA splicing, DNA repair and telomere function, and the presence of an RNAi system [237, 377], and has powerful molecular genetics and simple culturing conditions similar to S. cerevisiae. This chapter describes a CLS assay in S. pombe that follows the entire curve of the lifespan until all cells have died and recapitulates features of aging conserved throughout eukaryotes including lifespan extension by caloric restriction, lifespan shortening by over nutrition and lifespan regulation by TOR and AKT kinases. Analysis of the deletion mutants of tor1+ and its downstream target gad8+ indicates that S. pombe TOR signaling regulates lifespan in an evolutionarily conserved manner. There are two AKT paralogs in S. pombe, and analysis of the full lifespan curve of mutants lacking these paralogs showed that each AKT ortholog regulates lifespan under different nutritional conditions. Analysis of cells under normal and calorically-restricted conditions revealed a link between longevity and efficient cell cycle exit. These results demonstrate the utility of this new S. pombe CLS assay in studying the biology of aging.
45 Materials and Methods
S. pombe strains
The auxotrophic wild type strain KRP1 (h- ade6-M216 ura4-D18 leu1-32
his7-366, originally designated as CHP429 from C. Hoffman [4]) was used in all
lifespan analyses and to construct gene deletion strains. The gene deletion
strains are KRP19 (KRP1 leu1::LEU2), KRP14 (KRP1 sck1::LEU2), and
KRP20 (KRP1 sck2::LEU2), which have the entire open reading frame (ORF) of the deleted gene replaced with the plasmid pRS305 bearing the S. cerevisiae
LEU2 gene (as described in [366] except that a single CspC I site replaces the two
Sap I sites). KRP37 (KRP1 tor1∆) and KRP41 (KRP1 gad8∆), which also have the entire ORF deleted, were made by transforming linear ura4+ DNA flanked by
~500-800 bp of 5’ and 3’ untranslated regions (UTRs) of tor1+ or gad8+. The
linear ura4+ DNA with flanking 5’ and 3’ UTRs of tor1+ and gad8+ were
amplified from genomic DNA of S. pombe mutant strains FY13663 (tor1∆) and
FY13657 (gad8∆) provided by National BioResearch Project in Japan. Correct deletions were confirmed by PCR using primers external to the sequences for targeting the integration and the internal primers for pRS305, T7-R1 Extend and
T3 Extend, and for ura4+, BarcodePCR(888r) and InversePCR2 (Table 2.1).
The double deletion strain KRP21 (KRP1 sck1∆::LEU2 sck2∆::LEU2) was constructed from KRP14 and KRP20 by mating and tetrad dissection using standard genetic techniques [237]. Leu+ spores from tetrads with 2 Leu+ and
2Leu- spores were tested for the h- mating type and verified for the existence of
both deleted alleles by PCR.
46 Table 2.1. Oligonucleotides used in this study Oligonucleotide name Sequence sck1 deletion (KRP14) sck1UP1b* ACCGCGGCCGTTTCAGCAAAAGCAATGATCAATCAAG sck1UP2b* CATTCCAACTGCAGTGGCATAGTAAAATACGAATTAGTGTTCG sck1DWN1b* AGAGCTCGAGTAAGCTTGCTCCAATCAAAAAATTTG sck1DWN2b* GCCACTGCAGTTGGAAGATGACTACAAAATAGAACGTAAC sck2 deletion (KRP20) sck2UP1b* ACCGCGGCCGTTTATAATCACTCAAAAAACCAATTTATG sck2UP2b* CATTGAAGCGCAACTGCAGTGGCAGGATTATATCCAACTGAGGTG sck2DWN1b* AGAGCTCGAGTACGTTTGGTCTCCACGAATG sck2DWN2b* CCACTGCAGTTGCGCTTCAATGATTAAGCGCTG integration of pRS305 into leu1-32 (KRP19) leu1UP1b* ACCGCGGCCGGTTGTTGAAGAAGTTTTGTTGTGA leu1UP2b* TGTACCAACTGCAGTGGTGAACATCCATGATTTCATCACAA leu1DWN1b* AGAGCTCGAGGGGTTTAATGTAGAATAAATTCATATG leu1DWN2b* ACCACTGCAGTTGGTACACAGCGACAACTCGGTCATAA
external confirmation primers for KRP14, KRP21 sck1 confirm1 CGCCCCTCTGATATTATAACC sck1 confirm2 GAAAGCAACAAGTATTTGAGTTGGG
external confirmation primers for KRP20, KRP21 sck2 confirm1b CTTCTACTAGGGGTTGTTTCTTTGCACA sck2 confirm2b GAGATAGAATAGCCTTAAGCAAGATGAC
external confirmation primers for KRP19 leu1 confirm1 AGCGCTGATAATTGAGGCTACTAT leu1 confirm2 AGAATCGGGTTGTCATCGTTACAAT
internal confirmation primers for KRP14, KRP19, KRP20, KRP21 T7-R1 Extend GTAATACGACTCACTATAGGGCGAATTGG T3 Extend GCTCGGAATTAACCCTCACTAAAGGGA
confirmation primer for KRP14, KRP21 LEU2-EcoRV-AS CGGAGGCTTCATCGGAG
tor1 deletion (KRP37) Tor1 deletion 5' confirm ACAGTATGATAACCGCACGCGTTAGTGG Tor1 deletion 3' confirm CAGCCGTTACACAACCATAGACACACC gad8 deletion (KRP41) Gad8 deletion 5' confirm ACACGCAAGGTCTCGTAGTCATGCTTGGT Gad8 deletion 3' confirm CCAGGTGCTATTGTATTAATGTCGTCGG external confirmation primers for KRP37 Tor1 5' confirm-2 AAGATCGTTGGTAGAGTACTTCCAACCAAC Tor1 3' confirm-2 AAGGACTTGGGTGTAGCAAGTGAATTATGG external confirmation primers for KRP41 Gad8 5' confirm-2 AAAGTACCGTCGTGCAGCCTGGTTTATT Gad8 3' confirm-2 ACTTAGACGCTAATAAAGTCGAACAGGTGT
47 Table 2.1. Oligonucleotides used in this study (continued) Oligonucleotide name Sequence internal confirmation primers for KRP37, KRP41 BarcodePCR(888r) CACGACATGTGCAGAGATGCCGACGAAGCA InversePCR2 ACATGCTCCTACAACATTACCACAATCT * The ORF deletion mutations constructed according to the method of Wang et al. [366] used primers containing UP in their names to amplify the region upstream of the ORF and primers containing DWN in their name to amplify the downstream region. The primers ending in 2b overlap with one another to form a single PCR product with a central CspC I site.
Chronological aging assays
Two independent isolates of each mutant were assayed to determine the
CLS. For each CLS assay, cells were revived from frozen storage by streaking
onto rich medium plates (yeast extract + supplements or YES [237]) and grown at
30°C for 3 days to form single colonies. Cells for lifespan analysis were
inoculated from single colonies into medium (at an initial cell density of 5 104 cells/ml in 30 ml of medium in a 125 ml flask), and grown at 30°C rotating at 220 rpm. The medium for CLS assays was synthetic dextrose (SD) medium with 3% glucose with 150 mg/l of adenine, leucine, histidine and uracil [291] or Edinburgh
Minimal Medium (EMM) with 2% glucose with 225 mg/l of the same supplements [237] or synthetic minimal medium (SMM) with 3% glucose [291].
For over nutrition and caloric restriction experiments, SD or EMM medium with adenine, leucine, histidine and uracil or SMM medium were made with varying amounts of glucose (indicated in the figure legends). Cultures were grown to saturation, usually 2 days after starting the cultures, and this time point was designated as day 0 in all aging experiments. Starting from day 0, aliquots of cultures were taken, serially diluted in sterile milliQ water and multiple dilutions were plated on YES plates in duplicate and grown at 30oC for 4 days. Colonies
48 were then counted and used to calculate the number of colony forming units per
ml of culture (CFU/ml). Pilot experiments showed that incubating these plates for a total of 7 days did not significantly change the final CFU/ml values. For each experiment, the CFU/ml value was monitored until it reached < 10/ml.
Each experiment included both isolates of each mutant strain assayed and was done at least twice.
Analysis of CLS assay data
The CFU/ml values from each time point from 2 or more independent cultures were averaged and plotted on a log scale, with error bars corresponding to the range (for 2 cultures) or standard deviation (for > 3 cultures). The different growth conditions produced cultures whose cell densities on day 0 differed by ~10 fold. In order to allow a direct graphical comparison of these results, the log10 of each CFU/ml was obtained and the average, range or standard
deviation of these log10 values were then determined. These values were then
normalized to the day 0 log10(CFU/ml) value and plotted as Normalized
log10(CFU/ml) with values ranging from 0 to 1. Statistical comparisons were
performed using the Wilcoxon signed rank test in Prism 4 (Graphpad Software).
FACS analysis (fixed cells)
Cells for FACS analysis were taken from aging cultures (0.2 to 0.5 ml),
fixed in 70% ethanol, and stored at 4oC until all samples were collected. For
analysis, cells were washed once with 0.5 ml of 50 mM sodium citrate buffer pH
7.4 and resuspended in 0.5 ml of the same buffer for sonication at low power for 5
49 seconds to separate aggregated cells. Cells were treated with a final concentration of 250 g/ml RNase A in 50mM sodium citrate buffer pH 7.4 at
37oC overnight. After a second sonication, cells were stained with 4 g/ml
propidium iodide at room temperature for 30 minutes. Stained cells were diluted
to ~106 cells/ml and analyzed on Becton-Dickinson FACScan flow cytometer
using the FL2 channel. Data analysis was performed using FlowJo version 7
software (Tree Star, Inc).
FACS analysis (live cells)
Cells from a 0.15 ml aliquot of cultures grown in SD or EMM were spun
down, washed in 0.5 ml of 1X PBS, resuspended in 1X PBS with 4 µg/ml of
propidium iodide, and then incubated at room temperature for at least 45 minutes
on a rotator in the dark. Stained cells were spun down, resuspended in 1X PBS
and diluted to ~106 cells/ml in 1X PBS, and the fluorescence intensity of at least
104 cells was measured with a Becton-Dickinson FACScan flow cytometer using
the FL2 channel. The percentage of positive- and negative-staining cells was
determined using FlowJo version 7.
Stress sensitivity assays
For stress sensitivity assays, 5 106 cells were harvested and centrifuged at
2,000 x g. The supernatant was removed and cells were then resuspended in 0.5
ml of the appropriate solution. For the heat shock assay, cells were resuspended
in sterile milliQ water pre-warmed to 55°C, incubated at the same temperature for
50 5-25 minutes and then placed on ice for 2 minutes before plating for viability
(described below). The heat shock control cells were resuspended in 55°C pre- warmed sterile milliQ water and immediately chilled on ice for 2 minutes. For
H2O2 treatment, cells were resuspended in 0.5 ml of H2O2 solution (0, 100 or 300 mM) and incubated at 30°C for 90 minutes. Treated cells were washed once with 0.5 ml of sterile milliQ water and resuspended in 0.5 ml of sterile milliQ water. To monitor cell viability, stressed cells were serially diluted from 1 to
10,000-fold and 5 l of each dilution was spotted on YES plates and incubated at
30°C for 4 days.
51 Results
Chronological lifespan assay design
The CLS assay described here is based on the environmental niche of saprophytic fungi, which usually exist in a non-dividing state and then must rapidly grow once food becomes available in order to compete with other prokaryotic and eukaryotic microorganisms [86, 97, 102, 374]. A principal evolutionary advantage of S. pombe is its ability to ferment glucose in an aerobic environment, allowing a growing population of S. pombe to consume a key carbon source required by competing organisms. Thus, S. pombe has evolved under selective pressure to survive in a non-dividing state and resume growth, and the cellular machinery that regulates CLS has been adapted for this function.
Consequently, an assay that follows survival in a stationary phase culture and resumption of growth when food is reintroduced should allow one to examine evolutionarily conserved signaling pathways that control S. pombe lifespan. The new CLS assay was therefore based on the ability of cells to survive in stationary phase and then form colonies when plated on fresh medium.
Cells in SD medium show the evolutionarily conserved lifespan shortening in response to over nutrition while cells in EMM medium do not
Three types of media commonly used for S. pombe growth were tested for use in CLS assays: EMM + 2% glucose [237], SMM + 3% glucose [291], and SD
+ 3% glucose [184, 248, 253, 291, 339, 379]. Analysis of duplicate cultures of
EMM and SD media produced survival curves that showed a continuous decline in viability, and viability could be followed over at least six orders of magnitude
52 (Figure 2.1A, B). In contrast, cells in SMM medium showed a multiphasic
survival curve and responded to changes in nutritional conditions in an unusual
manner (Figure 2.2). Thus the subsequent work focused on EMM and SD media.
An evolutionarily conserved feature of lifespan regulation is that increasing
nutrient levels, or over nutrition, shortens lifespan [277, 322]. To establish whether the EMM or SD media were appropriate as a standard condition for a
CLS assay, the concentration of glucose, a major regulator of yeast metabolism
[132] and the primary carbon source in the medium, was varied. S. pombe can
grow in medium with glucose concentrations as high as 8% and as low as 0.1%
[133, 235]. We initially examined duplicate cultures containing the standard
glucose concentration (2% for EMM and 3% for SD) or 5% glucose to determine
if over nutrition shortened lifespan.
Cells grown in SD + 5% glucose medium showed a shorter lifespan than
those grown in the standard SD + 3% glucose medium (Figure 2.1A), showing
that this condition recapitulated the same response to over nutrition seen in other
species [229, 277, 322]. In contrast, cells grown in EMM + 5% glucose medium
had a longer lifespan than cells grown in the standard EMM + 2% glucose
condition (Figure 2.1B), indicating that over nutrition in EMM medium had the
opposite effect. As EMM and SD are made from nearly identical chemicals but
in different proportions [237, 317], the prolonged CLS in EMM with increased
calories may be related to the response to the levels of one or more of the
nutrients in the medium. Thus, only cells grown in SD medium had the
characteristics of a condition appropriate for CLS assays.
53 Figure 2.1. Over nutrition shortens S. pombe chronological lifespan in SD medium but lengthens lifespan in EMM medium. Wild type KRP1 cells were seeded at 5 104 cells/ml in a 30 ml culture and grown to stationary phase at 30°C (day 0, which is 48 hours after the culture was started). The cultures were maintained at 30°C and samples were taken at the intervals shown and plated on rich medium to assay the number of cells per ml that could form colonies (colony forming units or CFU per ml). Duplicate assays were performed and error bars show the ranges of the values (some error bars are too small to be visible). Cells were grown under standard (3% glucose for SD medium and 2% for EMM medium) or over nutrition conditions (5% glucose) in either SD medium (A) or EMM medium (B).
Figure 2.2. CLS curves in SMM-based medium are multiphasic and show an unusual response to caloric restriction. KRP1 cells were grown in SMM medium with 3% or 0.1% glucose. In both cases, the survival curve did not show a uniform decline but contained plateaus, which can be associated with low levels of regrowth. While the calorically restricted medium (0.1% glucose) did extend lifespan, it only did so in the latter portion of the lifespan in contrast to the uniform lifespan extension seen in other species [277, 322, 332]. The combination of the unusual CLS curves and response to caloric restriction indicates that SMM is an inappropriate medium for the assay of evolutionarily conserved features of aging.
54 In addition to cell viability, cell number and cellular DNA content were also monitored in the standard SD and EMM media. Cell number remained constant
throughout the entire lifespan, indicating that reduction in viability (CFU/ml curves) was not accompanied by a decrease in cell density (cells/ml curves)
(Figure 2.3A, B).
Figure 2.3. Cell density remains constant during the CLS assays in the SD and EMM media. (A) Cell density and CFU/ml for CLS assays in SD medium are shown. Duplicate assays were performed and the error bars show the ranges of the values. Cell density was determined by counting cells on a hemocytometer in duplicate for each time point, while the CFU/ml was determined as described in Materials and Methods. (B) Cell density and CFU/ml for duplicate CLS assays in EMM medium, determined as in panel A.
In contrast to cell number, cellular DNA content decreased as cultures aged in distinct ways in the two media. Cells grown in EMM medium showed both
1N and 2N DNA content at the start of the experiment (Figure 2.4A), consistent
55 with the previous reports that S. pombe can enter stationary phase from either G1
(i.e. 1N) or G2 (i.e. 2N) phase [47, 235, 319, 338]. However, by day 2, this culture showed substantial amounts of particles with sub-1N DNA content that dramatically increased as the assay progressed (Figure 2.4A, B). In contrast, cells in SD medium had a 2N DNA content at the beginning which slowly converted to 1N, indicating that cells were metabolically active despite the inability to re-grow when plated on fresh medium (Figure 2.4A). Thus, the change in the bulk population of cells in the culture occurred even as the fraction of viable cells declined (Figures 2.1 and 2.4).
Figure 2.4. EMM, but not SD medium, results in the accumulation of a large fraction of cells with sub-1N DNA content. (A) FACS analysis of DNA content in S. pombe cells from CLS experiments in EMM + 2% glucose or SD + 3% glucose. Cells were fixed with 70% ethanol, stored at 4°C until all cells could be treated with propidium iodide and analyzed together. The medium used for each series of FACS analyses is indicated along with the position of haploid S. pombe cells with 1N and 2N DNA content. All cells in the same series come from the same culture of a single CLS experiment. (B) The FACS analyses of the day 11 cultures in panel A plotted at scales that show the entire histogram and the relative proportions of the sub-1N, 1N and 2N cells.
56 This gradual diminution of 2N to 1N DNA content is distinct from normal
cell cycle progression where 2N cells divide to produce 1N cells, producing a
bimodal distribution of cells rather than cells with intermediate DNA content
[207]. These data suggest that cells in the CLS assay with SD medium were
slowly degrading their DNA. However, very few cells with less than 1N DNA
content were observed in the SD culture by day 11 (Figure 2.4B). A similar
reduction in DNA content has also been observed in CLS assays of several S. cerevisiae lab strains [372], showing that this response can be induced in evolutionarily distant fungi.
Cell viability was also monitored by the ability of cells to exclude the
membrane impermeable DNA stain propidium iodide. The fraction of viable,
unstained cells declined in parallel with the ability of cells to form colonies in
both the SD and EMM cultures (Figure 2.3, day 0 to day 6 and Figure 2.5),
confirming that the ability to form colonies is an appropriate assay for S. pombe
cell viability. Thus, aging in EMM and SD media was similar in that cell
number remained constant while cellular DNA content and cell viability showed
changes that paralleled the survival curve; however, only cells grown in SD
medium showed the evolutionarily conserved response to lifespan shortening due
to over nutrition. Consequently, SD medium was used in all subsequent assays.
Caloric restriction extends lifespan in the SD medium-based assay
To test whether S. pombe grown in SD medium exhibits the evolutionarily
conserved feature of lifespan extension by caloric restriction, it was first
57
Figure 2.5. Cell viability determined by propidium iodide staining and FACS analysis parallels the decline in CFU/ml. The fraction of viable cells in SD and EMM media in the CLS assays were determined from the same cultures in Figure 2.3A and B. Cell viability was measured by labeling cells with propidium iodide and analyzing them by FACS. The fraction of cells that did not stain with propidium iodide on day 0 was normalized to 1.0 in both cases and normalized values are shown. Cells grown for 5 days or more in EMM medium accumulate a large population of sub-1N cells that show greatly reduced staining with propidium iodide (Figure 2.4). This population that lacks nuclear DNA is unlikely to be viable, but scores the same as viable cells that do not stain with propidium iodide. Therefore, the fraction of viable cells is only shown out to day 6 when the FACS difference between cells that do and do not stain with propidium iodide was distinguishable.
necessary to determine a standard growth condition that did not shorten lifespan due to over nutrition nor lengthen lifespan due to caloric restriction. Therefore, to determine if 3% glucose was an appropriate condition, the lifespans of duplicate cultures with 3% and 4% glucose were compared. Both cultures gave similar, overlapping survival curves (Figure 2.6A). Thus, 3% to 4% glucose constitutes a range of glucose concentrations that does not shorten lifespan due to over nutrition nor lengthen lifespan due to caloric restriction, and SD + 3% glucose is a valid standard condition.
To directly test the effect of caloric restriction, the glucose concentrations in the media for CLS assays were reduced from 3% to 0.1%. The progressive
58 Figure 2.6. Caloric restriction extends lifespan in S. pombe. (A) SD + 3% glucose is an appropriate standard condition for CLS assays. Chronological lifespans were measured in SD + 3% or 4% glucose as described in the Materials and Methods. All lifespans were performed at the same time. The lifespan of the SD + 5% glucose culture from Figure 2.1A is included for comparison. The lifespans in 3% and 4% glucose were not significantly different (p = 0.28), while the 5% lifespan was significantly shorter than 3% glucose (p = 0.0005). Error bars represent the ranges of duplicate experiments. (B) CLS assays were performed and analyzed as in panel A with the glucose concentrations shown, except that the 0.3% and 0.1% cultures were sampled every two days. All lifespans were performed in duplicate and run concurrently. The 3% glucose cultures in panel B are independent from those in panel A. All lifespans for cultures with 2% glucose or less were significantly longer than the 3% glucose culture (p < 0.008). (C
and D) The normalized log10(CFU/ml) from the CLS experiments in panels A and B are plotted, respectively. The normalization of the log10 values starts all lifespans at the same point, and allows cultures that grow to different maximum densities (e.g. the 3% and 0.1% glucose cultures) to be directly compared. (E) Median lifespans from cultures with different glucose concentrations as defined as the point where the normalized log10(CFU/ml) equals 0.5. Median CLS values were calculated by linear interpolation between the values immediately above and below 0.5. The slightly different values for the two independent 3% glucose experiments performed on different days (in panels A and B) are shown.
59 decrease in glucose concentrations from 3% to 0.3% caused a progressive
increase in lifespan, while the 0.3% and 0.1% glucose cultures had remarkably similar, overlapping CLS curves (Figure 2.6B). This observation of reaching a maximum lifespan as calories decrease is strikingly similar to data from calorically restricted mice, where lifespan steadily increased as calories were decreased from ad lib to 85 kcal/wk to 50 kcal/wk, and overlapping survival curves were produced at 50 kcal/wk and 40 kcal/wk [373]. Thus, the S. pombe
CLS assay in SD medium showed the evolutionarily conserved trait of lifespan extension by caloric restriction.
The survival curves in many lifespan assays typically compare the median lifespans, i.e. the time when 50% of the population remains alive [241]; however, most studies analyze a much smaller population and follow viability over a smaller range (a decline of 100 - 1000 fold). As our assay follows the full lifespan of hundreds of millions of individuals until all cells are dead to give a survival curve that spans seven orders of magnitude, the point when 50% of the population remains alive would not summarize the changes in viability that occur in a large portion of the survival curve. An additional consequence of the large range of viable cells is that the sampling error for CFU/ml is high early in the lifespan (+106) and low at the end of the lifespan (+10). Therefore, the survival
curves of these different glucose cultures presented in Figure 2.6A and B required
a different statistical method for quantitative comparison. A common way to
compare the means or medians of these types of data is to take the logarithm of
each value and then compare these transformed data [21, 333].
60 To establish a median lifespan measurement to compare different survival
curves, the log10(CFU/ml) of aged cultures at each time point was normalized to
the log10(CFU/ml) at day 0 of the corresponding cultures so that all lifespans start
at 1.0 and then decline as cells in the culture die (Fig. 2.6C, D). The lifespan
medians were then calculated as the point where normalized log10(CFU/ml)
equals 0.5. This metric confirms that the median lifespans of the 3% glucose
cultures done at different times (Figure 2.6A, B) were similar to each other (8.1
and 6.7 days), the 5% culture had a shorter lifespan (3.4 days), the 3% and 4%
cultures had very similar values (8.1 and 8.7 days, respectively), as did the 0.3%
and 0.1% glucose cultures (22.8 and 23.1 days, respectively) (Figure 2.6E). This
normalization also revealed the differences between the 0.1% and 0.3% glucose
cultures early in the lifespans (Figure 2.6D), since the 0.1% glucose culture did
not reach as high a cell density and maintained the same number of viable cells
over a longer period. These curves were compared with a Wilcoxon signed rank
test, which showed that the 3% and 4% curves were indistinguishable and the 3%
and 5% curves were clearly different (Figure 2.6). Thus, transformation to
logarithmic values provided a more revealing presentation of these data, and
defining the median CLS as the point where the normalized log10(CFU/ml) equals
0.5 provided an accurate summary of the different lifespan curves.
To investigate how caloric restriction might extend S. pombe lifespan, we examined DNA content by FACS for the 3% - 0.1% glucose cultures (Figure 2.7).
Similar to the results from the 3% glucose cultures, cells in the 2% and 1% glucose cultures had a 2N DNA content at the beginning of the CLS assay that
61 Figure 2.7. Long-lived S. pombe cells grown in 0.1% glucose maintain a constant DNA FACS profile while aging. Aliquots from one of the two duplicate cultures from Figure 2.6 were collected and processed for FACS. All cell numbers are plotted on the same scale. For the purposes of comparison, day 0 of the assay is defined as 48 hr after the culture was started. In the case of cells grown in 0.1% glucose medium, the culture did not reach maximum cell density until day 1, i.e. 72 hrs after the culture was started. The 0.1% glucose culture maintained the same FACS profile until day 21 (data not shown).
gradually decreased to 1N DNA content (Figure 2.7) as the ability to form colonies declined by several logs (Figure 2.6B). However, the cultures with a longer CLS showed a slower shift in DNA content (Figures. 2.6 and 2.7). This difference can be seen, for example, by comparing days 0 and 7 of the lifespans for the 1%, 2% and 3% glucose cultures. The longer-lived 1% glucose culture retained a unimodal FACS profile on both days with a DNA content on day 7 midway between 1N and 2N. In contrast, the shorter-lived 3% and 2% cultures
62 showed a larger shift and proportion of 1N cells on day 7 than on day 0, and the
FACS profiles on day 7 were more heterogeneous. This correlation of higher
viabilities when the cells showed less of a shift over time was repeated in the
long-lived 0.3% and 0.1% glucose cultures. For example, a comparison on day
7 and day 1 of the 0.1% glucose culture revealed FACS profiles that were virtually the same. These results indicate that cells in the 0.3% and 0.1% glucose cultures had exited the cell cycle, and this exit was associated with longer
CLS. Exit in the 0.3% and 0.1% cultures was from both G1 and G2, as the
FACS profiles showed a bimodal pattern of 1N and 2N cells. This pattern is
similar to FACS profiles observed for other starvation conditions [235] and was
distinct from the 3%, 2% and 1% cultures (Figure 2.7).
Sub-1N cells were observed over time in the 0.3% glucose culture (Figure
2.7). Cells with sub-1N DNA content were detected by day 3 and the proportion of sub-1N cells increased until a maximum at day 12. These sub-1N cells were unique to the 0.3% glucose culture. Sub-1N cells were not observed in the 0.1% glucose culture in the experiment shown in Figure 2.6 as well as 2 additional independent experiments (data now shown), indicating that these cells showed fewer changes during the CLS assay. The combination of prolonged viability in
early lifespan and the lack of changes in the FACS profile indicated that 0.1%
glucose was the appropriate caloric restriction condition.
Longer CLS correlates with exhausting free glucose in the medium
To validate that the 0.1% glucose cells are in fact restricted for glucose, we
determined the free glucose concentration in the medium before and after the cells
63 reached their maximum density (day –1 and day 1, respectively) (Table 2.2). At day -1, the glucose concentrations in media from the 0.1% and 3% glucose
cultures were the same as that in the initial medium before cells were added.
However, by day 1 of the CLS, the amount of glucose remaining in the 0.1%
glucose culture was about 1000-fold less and at the lower limit of detection while
the amount remaining in the 3% culture was almost the same as the starting
concentration in the calorically restricted culture (0.09%). In addition, the 0.1%
cultures did not reach as high a cell density as the 3% glucose cultures (Table
2.2), indicating that glucose was a limiting nutrient for cell growth. Thus, the
longer CLS correlated with completely exhausting the free glucose in the medium
as cells reached their maximum density.
Table 2.2. Cells grown in caloric restriction (0.1% glucose) medium exhaust their glucose supply as they reach maximum density. Glucose Concentration Cell Density (cells/ml) Average (%) Std. Dev. (%) Average Std. Dev. Day -1 3.14 0.0718 2.88 106 1.06 106 SD + 3% Day 0 0.468 0.236 4.34 107 3.49 106 Glucose Day 1 0.0854 0.0679 4.50 107 3.17 106
Day -1 0.0976 0.0080 3.58 105 2.41 105 SD + 0.1% Day 0 0.0117 0.0111 4.34 106 1.32 106 Glucose Day 1 0.00007 0.00002 5.64 106 3.10 106 Aliquots (0.5 ml) of aged cultures harvested at different time points were filter-sterilized, stored at 4oC, and the glucose concentrations were determined using a Glucose (HK) Assay Kit (Sigma) according to the manufacturer’s instructions. Assays were performed on three separate cultures for each condition and the average and standard deviations (Std. Dev.) are shown.
Long-lived calorically restricted cells show increased stress resistance
To determine if the long-lived calorically restricted S. pombe share the
evolutionarily conserved feature of increased resistance to environmental stress
64 [322, 332], the ability of cells grown in 3% or 0.1% glucose cultures to survive exposure to an oxidizing agent (H2O2) or heat stress (55°C) was tested. The calorically restricted cells showed no loss of viability when exposed to 300 mM
H2O2, while the viability of normal cells decreased by ~1000-fold (Figure 2.8A).
The calorically restricted cells also showed an increased resistance to heat stress compared to normal cells, where survival of calorically restricted cells was ~20 fold higher than normal cells (Figure 2.8B). Consequently, calorically restricted
S. pombe showed increased stress resistance compared to S. pombe grown under normal conditions, as is commonly seen in other calorically restricted organisms.
Figure 2.8. Long-lived calorically restricted S. pombe show increased stress resistance. (A) Resistance to an oxidizing agent. Aliquots of cells from the SD + 3% glucose culture or the calorically restricted SD + 0.1% glucose culture were taken when cells had reached maximum density, washed with sterile water, and 5 x 106 cells were resuspended in varying concentrations of H2O2 and incubated at 30°C for 90 min. After washing the cells in water, 10-fold serial dilutions were made and 5 µl aliquots of each suspension were spotted onto rich medium where all cells can grow. (B) Resistance to heat stress. Cells grown in normal or calorically restricted condition as in A were heat shocked at 55°C for different lengths of time, placed on ice for 2 min and then diluted and spotted on plates. Fold differences in resistance for the assays in panels A and B were determined by counting the number of single colonies in the most dilute spots in the treated and untreated samples.
65 The AKT orthologs sck1+ and sck2+ differentially affect lifespan under normal
and over nutrition conditions
We tested whether the two AKT orthologs of S. pombe, sck1+ and sck2+,
played a role in lifespan control by determining the CLS of the strains lacking one
or both genes. When assayed under normal conditions (i.e. 3% glucose), the
strain bearing a deletion of the sck2+ gene, sck2∆, showed a significant increase in
lifespan (Figures. 2.9B, E), while the strain bearing a deletion of the sck1+ gene,
sck1∆, had a lifespan indistinguishable from wild type (Figure 2.9A, D). The
sck1∆ sck2∆ double mutant had the same lifespan as the sck2∆ single mutant
(Figures 2.9B, C and E, F), confirming that the sck1+ AKT paralog did not affect
lifespan under normal conditions even when its paralog sck2+ was not available to
potentially substitute for sck1+ function.
The products of paralogous genes often substitute for one another under
different conditions (e.g. [185]), suggesting that the sck1+ protein might control
lifespan under a different nutritional condition. The previous measurements of
glucose concentrations in the normal medium indicated that cells start growth in
3% glucose and glucose concentration gradually declines (Table 2.2). These
considerations suggested that sck2+ functioned at 3% and/or lower glucose concentrations to control lifespan. We therefore hypothesized that sck1+ might be adapted for a function at a higher glucose concentration, and assayed CLS in the 5% glucose over nutrition condition for the single and double AKT mutants.
Both the sck1∆ and sck2∆ mutations caused a detectable lifespan extension in the over nutrition condition. The sck1∆ strain had a longer CLS than wild
66 Figure 2.9. Deletion of the AKT kinase gene sck2+ extends lifespan under normal conditions while deletion of the gene for the paralogous kinase sck1+ does not. (A-C) Chronological lifespans of sck1∆ and sck2∆ single and double mutants grown in normal SD medium (3% glucose), with CFU/ml plotted on a log scale. The KRP14 sck1∆ strain shows a lifespan the same as KRP19 wild type (wt) cells (p = 0.46) while the KRP20 sck2∆ and KRP21 sck1∆ sck2∆ strains show extended lifespans (p < 0.001 for sck2∆ and sck1∆ sck2∆). The sck2∆ and sck1∆ sck2∆ lifespans are statistically indistinguishable from each other (p > 0.05). (D-F) The data from panels A-C plotted as normalized log10(CFU/ml) to show the fraction of viable cells over the course of the lifespan. The median lifespans determined from these data are: KRP19 wild type 8.3 days; KRP14 sck1∆ 8.7 days; KRP20 sck2∆ 10.6 days; KRP21 sck1∆ sck2∆ 10.5 days.
type cells that was most noticeable in the latter 40% of the lifespan (Figures
2.10A, D). The sck2∆ strain also extended lifespan over wild type cells (Figures
2.10 B, E). The double mutant strain had a longer lifespan than the wild type and either single mutant strain (Figures 2.9C, F), confirming that each AKT
67 ortholog made a contribution to the control of longevity in S. pombe. Thus, the
two AKT orthologs of S. pombe control lifespan under different nutritional
conditions.
Figure 2.10. Deletion of either AKT kinase sck1+ or sck2+ extends lifespan in over nutrition conditions. (A-C) Chronological lifespans of sck1∆ and sck2∆ single and double mutants grown in over nutrition SD medium (5% glucose), with CFU/ml plotted on a log scale. (D-F) The data from panels A-C plotted as normalized log10(CFU/ml) to show the fraction of viable cells over the course of the lifespan. The median lifespans determined from these data are: KRP19 wild type 3.6 days; KRP14 sck1∆ 4.4 days; KRP20 sck2∆ 7.1 days; KRP21 sck1∆ sck2∆ 9.4 days. KRP14 sck1∆ had a longer lifespan than the wild type KRP19 strain after day 5, which made the two curves distinguishable (p = 0.0215). Both the KRP20 sck2∆ and KRP21 sck1∆ sck2∆ strains show extended lifespans compared to wild type (p < 0.001 and p < 0.05 for sck2∆ and sck1∆ sck2∆, respectively). The lifespan curves of the sck2∆ strain and sck1∆ sck2∆ strains were significantly different (p < 0.05).
68 Deletions of the non-essential TOR kinase tor1+ or its potential substrate gad8+ extends S. pombe chronological lifespan
Inactivation of TOR signaling has been reported to extend lifespan in worms, flies and budding yeast. To test whether TOR activity in fission yeast also regulates lifespan, the CLS of deletion mutants of tor1+, encoding the non-
essential Tor kinase, and gad8+, a serine/threonine kinase and potentialTor1p
substrate [222], were determined. The tor1∆ and gad8∆ single deletion mutants had a slow growth phenotype. On solid medium, they formed smaller colonies compared to the wild type strains when grown for the same period of time (data not shown). In the liquid cultures for CLS determination, tor1∆ and gad8∆ mutants required additional 2 to 3 days to reach saturation (Figure 2.11A, B).
To directly compare the lifespans of the wild type strain, tor1∆ and gad8∆ mutants, the lifespan curves of these strains were re-graphed to adjust the time points at which the tor1∆ and gad8∆ cultures reached saturation to “Day 0”
(Figure 2.11C-E). Both tor1∆ and gad8∆ mutants had longer lifespans than the wild type control, consistent with the findings in other model organisms and demonstrating an additional evolutionally characteristic preserved in our aging assay.
69 Figure 2.11. Deletion of tor1+ or gad8+ kinase extends lifespan. (A and B) The KRP37 tor1∆ and KRP41 gad8∆ mutants exhibited slow growth phenotype. When grown in SD + 3% glucose medium, KRP37 tor1∆ and KRP41 gad8∆ cultures required 2 to 3 more days to reach saturation. (C and D) To directly compare the lifespans of the wild type, tor1∆ and gad8∆ cells, time points day2 and day3 of tor1∆ and gad8∆ cultures in panels A and B, respectively, were designated as “day0” in C and D. (E and F) The data from panels C and D plotted as normalized log10(CFU/ml) to show the fraction of viable cells over the course of the lifespan. The median lifespans determined from these data are: KRP34 wild type 8.6 days; KRP37 tor1∆ 14.9 days; KRP41 gad8∆ 13.3 days. Both mutants had longer CLS than the wild type KRP 34 (p=0.0002 for both mutants).
70 Discussion
The new S. pombe CLS assay described in this chapter recapitulates
features of lifespan control that are conserved throughout eukaryotes.
Conditions for over nutrition, normal nutrition and under nutrition were
established by varying the concentration of glucose and, like other eukaryotes
[229, 277, 322, 332], over nutrition shortened lifespan while caloric restriction
increased lifespan and stress resistance (Figures 2.6 and 2.8). An important
property of this assay is that after cells reached their maximum density, cell
viability showed a progressive decline with no plateaus, indicating that regrowth
of the cells was minimized and the measurement of viable cells was not
complicated by a subpopulation of growing cells. These data contrast with S.
cerevisiae where cell lysis and regrowth are observed in some CLS assays [70,
74], apparently due to mutagenic or epigenetic adaptation. The lack of regrowth
in the new S. pombe CLS assay allows an analysis of the full survival curve, i.e.
over a >106-fold range, until all cells in the culture are dead. This large range
contrasts with almost all previous assays in S. pombe and S. cerevisiae that follow a smaller (~102-103) portion of the survival curve (e.g. [75, 77, 214, 247, 258,
297, 371, 372, 393]).
The ability to follow the entire survival curve in the new CLS assay
allowed us to elucidate the distinct properties of the two S. pombe AKT kinase
family members, showing that sck1+ functions under an over nutrition condition
while sck2+ functions under both normal and over nutrition conditions (Figures
2.9 and 2.10). Roux et al. and Ohtsuka et al. have also shown that the sck2∆
71 mutation extends lifespan, and Roux et al. also examined the sck1∆ mutation but
found no effect [258, 297]. However, the sck1∆ mutant was only analyzed using
their standard condition and monitoring survival over 3 logs. Our results show
that the lifespan extending effect of the sck1∆ mutation observed in over nutrition
conditions was most evident at later points in the lifespan (Figures 2.10A, D).
This result illustrates the advantage of following lifespan over an extended range.
We noted that one S. cerevisiae CLS assay has recently been described, using
small cultures, which can follow cell viability over a ~106-fold range, and was
used to show that the lifespan extending effects of caloric restriction were
independent of sirtuins. Many of the effects were only detectable after viability
had dropped by >1000-fold [329], reinforcing the importance of analyzing the full
lifespan.
Because the increasing roles of nutrient-sensing kinases in longevity
regulation are being revealed, this assay was also to examine the effect of TOR
signaling on S. pombe CLS. As in budding yeast, there are two genes encoding
the catalytic cores of TOR complexes, tor1+ and tor2+. Previous work in S.
cerevisiae has established that the TOR1 and TOR2 genes function in nutrient sensing and the decision whether cells should begin a new cell cycle [290].
Biochemical analyses show that two Tor complexes exist, TORC1 and TORC2,
which can have different Tor proteins and different associated subunits [7].
Both S. pombe and S. cerevisiae are similar in that the tor2∆ mutants are inviable,
and the tor1∆ mutants have a longer lifespan ([160, 278] and Figure 2.11).
The current model for the Tor control of longevity is thought to occur
72 through the TORC1 complex [336]. Our S. pombe result was therefore
surprising in light of recent data indicating that S. pombe Tor1p is predominantly
in TORC2 complex [121, 208, 223]. Thus, our data suggest a role for TORC2-
dependent control of CLS or the presence of an undetected TORC1 fraction with
Tor1p. The long lifespan phenotype of the gad8∆ mutant appears to support the
first idea as phosphorylation of Gad8p was recently shown to depend on not only
Tor1p, but also a conserved TORC2-specific subunit, Sin1p [139]. Interestingly,
recent work by Soukas et al. showed that C. elegans fed with RNAi targeting
TORC2-specific subunit Rictor (encoded by rict-1 gene) exhibited an extended
lifespan on nutrient-rich diet [334]. Thus, data from S. pombe and C. elegans
suggest that both the TORC1 and TORC2 can regulate organismal lifespan.
The calorically restricted (0.1% glucose) S. pombe showed no change in
DNA content over the course of the CLS assay (Figure 2.7) and survived much
longer (Figure 2.6B, D). In contrast, S. pombe grown in media with higher
glucose levels showed a gradual reduction in DNA content as the fraction of cells
that were viable rapidly declined (i.e. the 1%, 2% and 3% glucose cultures in
Figures 2.6 and 2.7). Thus, the S. pombe in media with glucose concentrations >
1% were still metabolically active. These data indicate that caloric restriction induced a quiescent state, resulting in cell cycle exit to G0 phase.
We propose that calorically restricted cells entered G0 more efficiently
because the free glucose was exhausted as cells reached maximum density (Table
2.2). In contrast, the S. pombe cells cultured under normal conditions reached
maximum density in the presence of a glucose concentration sufficient to support
73 cell growth (Table 2.2). Consequently, under normal conditions, cells have
presumably stopped growing because of some other limiting nutrient(s), but the
medium still contains substantial amounts of glucose that signals cells to grow.
Thus, the remaining glucose may shorten lifespan by providing an extracellular
signal that prevents arrest in a quiescent state. In the case of calorically
restricted cells, glucose depletion and consequent loss of growth-inducing
signaling allow them to enter quiescence more efficiently, resulting in a longer
CLS. These S. pombe results are consistent with results from S. cerevisiae
where cells grown in low glucose medium enter G0 more efficiently with longer
CLS [372], and G0 cells placed in media lacking all nutrients except glucose
attempt to enter the cell cycle and show a dramatic loss in cell viability [100,
101].
We therefore suggest that a longer CLS is due in part to efficient cell cycle
exit in the absence of growth inducing stimuli. In the case of S. pombe, one such
stimulus is the presence of free glucose in the medium, and another appears to be
a function of nutrient-sensing kinases, such as AKT and TOR. This general
conclusion is relevant to human aging in that signaling terminally differentiated human cells to proliferate has been linked to shortened lifespan and cell death
even if the post-mitotic state is maintained by other mechanisms (e.g. [191, 285,
357, 382]). As humans and S. pombe use similar evolutionarily conserved
pathways to respond to extracellular stresses and stimuli (e.g. [44]), the S. pombe
CLS assay described here should be useful in identifying the mechanisms
modulating the lifespan of post-mitotic cells.
74
Chapter 3
Construction and characterization of a barcode-tagged insertion
mutant library in the fission yeast Schizosaccharomyces pombe
75 Abstract
Aging research in the budding yeast Saccharomyces cerevisiae has thrived
in part due to the available reagents for genome-wide studies (e.g. the ORF
deletion mutant collection). The fission yeast Schizosaccharomyces pombe is an
evolutionarily distinct yeast species that possesses many functions (e.g. RNA
interference, the splicing mechanism, requirement of mitochondrial DNA for wild
type cell survival) that are similar to multicellular organisms and absent in S.
cerevisiae. For these reasons, fission yeast has become a popular model
organism to study gene functions and molecular pathways. An S. pombe
deletion mutant collection has only recently become available and the essential
information for genetic analysis was not released until May 2010. We therefore
constructed and characterized a barcode-tagged insertion mutant library in S.
pombe. An insertion vector composed of the selectable marker ura4+, a random
barcode sequence flanked by restriction enzyme Sfi I recognition sites and other functional elements was used to generate insertion mutations in a wild type strain.
Phenotypic analysis of 3,600 mutants from the library suggests that the insertions
are distributed broadly throughout the genome. Mutants of this library are stored as pools of mixed mutants for large-scale parallel selection and in 384-well arrays for high-throughput screen on individual mutants. While the insertion structures of some mutants are unexpectedly complex (e.g. tandem integrations of the vector
and co-integration mitochondrial DNA), the insertion sites of several mutants
could be determined by thermal asymmetric interlaced (TAIL)-PCR.
76 Introduction
The approach to identifying genes in a biological pathway has shifted from
analyzing limited number of individual mutants with a predicted phenotype to whole-genome approaches. Monitoring changes in gene expression in response to genetic alteration (e.g. gene mutation) or environmental stimuli (e.g. drug treatment) using methods such as differential display [196] and microarray [305] can reveal genes with altered expression levels that may be functional components in genetic or molecular pathways. However, gene expression does not always correlate with specific biological functions. For example, deletion of a gene up-regulated in response to high osmotic pressure does not necessarily sensitize the mutant cells to the same condition [91]. An alternative approach of
studying gene function is to screen for phenotypes in a defined set of mutants or a
pool of random mutants. For example, Martin-Castellanos et al. deleted 175
fission yeast genes with elevated expression during meiosis and identified 7 novel
genes with crucial function in meiosis [216]. Similarly, Sajiki et al. analyzed
610 temperature-sensitive fission yeast mutants to seek for genes essential for
survival in quiescence [299]. Although these studies have provided new insights
in the corresponding research fields, these approaches are often time-consuming
and laborious due to the mutagenesis procedure, the identification of random
mutations and the requirement of investigating the mutants individually.
Recent efforts have been put towards the development of biological reagents
and selection strategies. One significant advance has been to create barcode
(also called “signature”)-tagged mutations that allow researchers to study many
77 mutants in a mixed population and isolate the desired mutants quickly after the
selection, an approach now referred to as “parallel selection”. This technique
was first applied in the bacterium Salmonella typhimurium to identify virulence
genes [126]. Hensel et al. created a S. typhimurium mutant bank by transposon- mediated mutagenesis, in which the transposons carried 40 bp unique sequence tags that provided each of their mutant strains a specific identify. After infecting mice with pools of “signature-tagged” bacterial mutants, they were able to
discover strains with non-virulent mutations, which diminished quickly in the host, by monitoring the abundance of the signature/barcode tags of all mutants post-infection by DNA colony blot hybridization [126]. The important aspects of this study are the ability to performing parallel analysis and a negative selection for bacterial mutants that were lost during the experiment. This approach would not be possible without the specific barcode tags on the mutants.
This idea has been extended into many other model organisms. For example, Singhi et al. developed a technique called selection-subtraction approach (SSA), which utilizes a 80 bp barcode-tagged cDNA expression library constructed on a retroviral vector [325]. Genes that suppress cell growth in some human tumor cell lines upon ectopic expression could be identified by the reduced abundance of their corresponding barcodes in a cell population after
infection [325]. In a similar approach, Ngo et al. used a loss-of-function RNA
interference screen to look for genes required for the survival of specific
subgroups of B-cell lymphoma [250]. They created a retroviral vector-based
small hairpin RNA (shRNA) library, in which individual shRNA clones were
78 tagged by unique 60 bp barcodes. After shRNA transfection, cells expressing shRNAs targeting genes essential for growth and survival of specific lymphoma
cells diminished in the population. The reduction of barcodes carried in these
toxic shRNA-expressing cells could be revealed by microarrays of barcode
sequences. Using this approach, they identified potential molecular targets for cancer therapy [250].
The budding yeast Saccharomyces cerevisiae is the model organism where barcode-tagged mutagenesis has been exploited the best. Taking advantage of the relatively small genome size, the well-annotated genome sequence and the efficient gene deletion technique, a collection of complete open reading frame
(ORF) deletion mutants was made by an international consortium [91, 375].
Each of the deletion mutants was tagged by two unique barcodes bordered by identical priming sites, which can be used to follow the proportion of each mutant in a culture (Figure 3.1). These barcodes can be amplified en masse by PCR to generate probes for commercially available high-density microarrays to take a
“census” of the relative abundance of each mutant in the culture under a variety of growth conditions.
A major advantage of this deletion mutant library is that the function of genes can be directly tested by observing the phenotypes of these mutants (i.e. fitness in different culturing conditions) rather than making assumptions as to how changes in the expression level of genes will affect phenotypes. When comparing the gene expression profiles with mutant fitness profiles in four culturing conditions (i.e. growth in media with 1M NaCl, 1.5M sorbitol, pH 8 or
79 galactose), a surprising conclusion is that only less than 7% of genes that showed
increased mRNA expression were required for growth in these conditions and the
remaining 93% of the genes induced in response to these stimuli was dispensable.
In addition, many genes necessary to maintain normal cell fitness under these treatments did not have altered expression level [91]. These results strongly demonstrate the significance of barcode-tagged mutants in genome-wide analysis of gene function as a complement to the traditional gene expression profiling.
Figure 3.1. The deletion and tagging strategy used in the S. cerevisiae and S. pombe gene deletion mutant collections. The deletion construct used in both gene deletion mutant collections contains a selectable marker KanMX4 and two 20-mer barcode tags, the up-tag (U-tag, green) and down-tag (D-tag, orange), each of which is bordered by two priming sites identical in all strains (black and dark green for U-tag, dark blue and pink for D-tag) to allow PCR amplification of barcodes. The deletion construct is flanked by sequences homologous to the 5’ and 3’ UTRs of the targeting ORF. Gene deletion was achieved by homologous recombination between the ends of deletion construct and corresponding sequences on the chromosome and replacement of the wild ORF with KanMX4 marker.
The fission yeast Schizosaccharomyces pombe has powerful genetic tools
and well-annotated genomic information similar to S. cerevisiae, but a barcode-
tagged deletion strain set was only developed three years ago by a private
company Bioneer, Inc. They employed a similar deletion and barcode tagging
strategy that was used in the budding yeast (Figure 3.1) [176]. Each of these
fission yeast deletion mutants also carries two barcodes whose sequences were
80 not made available to the public until May, 2010 [176]. Therefore, application
of this mutant library was previously limited to analysis of small subsets of
mutants individually or re-created mutant arrays [288, 393]. Parallel analysis
where multiple mutants are grown together in a single culture was not possible.
An alternative to gene deletion mutants are insertion mutants. In S.
cerevisiae, random insertion mutants can be generated by transposon-mediated
mutagenesis which allows quick identification of the insertion locations [32, 331].
In S. pombe, a different approach using non-homologous linear DNA targeting
vector had been demonstrated as means to generate random insertion mutants [42,
52]. Several studies have indicated that in the absence of significant sequence
homology, non-homologous/illegitimate recombination occurs at a higher
frequency in the fission yeast than in the budding yeast, allowing broad
distributions of the non-homologous linear DNA vector in the genome [93, 130,
293].
To create an alternative mutant collection for genome-wide study in S.
pombe, we have taken an insertion mutagenesis approach to construct a barcode-
tagged fission yeast mutant library. In our mutant library, an insertion DNA
cassette carrying an ura4+ selectable marker was used to transform a yeast strain
where the corresponding ura4+ sequence in the genome had been deleted. The
insertion vector, which also carried a unique barcode, had no sequence homology with the genome, and was expected to integrate randomly. Our barcode tagging strategy was designed to allow parallel analysis without pre-fabricated microarray
or even knowledge of the barcode sequences prior to doing the experiment.
81 Thus, the barcodes were constructed so that the abundance of barcodes, which represent the proportion of the corresponding mutants, can be determined by simple and routine techniques (e.g. PCR, restriction enzyme digestion, ligation and sequencing) after selection. This chapter presents the design of the barcode tagging strategy, the construction of this mutant library and characterization of a subset of individual mutants.
82 Materials and Methods
Strains and media
The E.coli electrocompetent cells NEB 5-alpha (Cat# C2989K, NEB) were
used for the construction of the bacterial barcode-tagged insertion DNA vector
library. The auxotrophic fission yeast wild type strain KRP1 [39] (h- ade6-M216 ura4-D18 leu1-32 his7-366, originally designated CHP429 from C. Hoffman [4]) was used to construct the fission yeast insertion mutant library.
Standard fission yeast growth media were used in this study. Unless otherwise specified, yeast extract + 225 mg/l of supplements (YES) contains 3% glucose, and Edinburgh minimal medium (EMM) contains 225 mg/l of supplements and 2% glucose [237]. Minimal medium agar (MMA) has 1%
glucose and 225 mg/l of supplements [131]. EMM + FOA contains 1 g/l of 5-
FOA (5-fluoroorotic acid), 2% glucose, 50.25 mg/l of uracil and 225 mg/l of
supplements for auxotrophic mutations other than ura4-D18 [82]. For EMM,
MMA and EMM + FOA media, the regular supplements were replaced by the
yeast complete supplements (YC - uracil, Table 3.1) in some experiments to allow the growth of additional auxotrophic mutants generated in this work.
Table 3.1. The supplements and their concentrations in the YC - uracil medium used in the selection of insertion mutants Amt Amt Amt Amt Suppl. Suppl. Suppl. Suppl. (mg/l) (mg/l) (mg/l) (mg/l) lysine 150 tryptophan 150 isoleucine 75 adenine 225 arginine 150 tyrosine 75 asparagine 75 leucine 225 threonine 150 methionine 75 proline 75 histidine 225 cysteine 150 valine 75 phenylalanine 75
83 Construction of the bacterial barcode-tagged insertion DNA vector library
A. Construction of the targeting DNA vector
A bacterial phage DNA fragment was designed as protective buffer
sequence to prevent deletion of the barcode after transformation into yeast. This
buffer was created by PCR with 9 overlapping oligonucleotides (hsplam1-9) and
S.pombeLmbdBrcd primer (Table 3.2). This construction removed all ATG
codons in the transcribed strand, inserted a mutated human HSP70 promoter [279]
and introduced a Blp I site at one end of the final fragment. The ura4+ selection
marker was first amplified by PCR with primer InvU4S and InvU4-AS and
extended by an overlapping PCR with a short double-stranded DNA formed from
two oligonucleotides lox71-InvU4S1 and lox71-InvU4AS1 to include a mutated
loxP site, lox71 [5], and a Sfi I site. The buffer fragment (ATGless -HSP70)
and extended ura4+ marker (lox71-InvU4) were individually cloned to pCR2.1-
TOPO vector (Invitrogen) and sequenced. The correct ATGless-HSP70
fragment was isolated from pCR2.1 vector as a Hind III-EcoR V fragment and
ligated to pCR2.1-lox71-InvU4 cut with Spe I, rendered blunt by treatment with
Klenow fragment of E. coli DNA polymerase I and cut with Hind III to make the
final insertion vector construct pCR2.1-ATGless -HSP70-lox71-InvU4, referred
to as “pInsertion-ura4”.
B. Insertion of the stuffer fragment and vector preparation
A 0.9 kb fragment containing part of the sck1+ coding sequence was generated by PCR using primers Stuffer 5’ Blp I and Stuffer 3’ Sfi I. After
84 Table 3.2. Oligonucleotides used in this study Oligonucleotide name Sequence Construction of ATGless buffer-HSP70 sequence hsplam1 GCTGAGCGCTAGCTACTGTATGTACATACAGTACCTCATCGAGCTCGGTT hsplam2 GGAGAGTCGACTTTTCCCTTCTGACTGCCTAACCGAGCTCGATGAGGTAC hsplam3 AAGGGAAAAGTCGACTCTCCGTGACGACTTATAAAAGCCCAGGGGCAAG hsplam4 CACCGTTAGCCGTTATCCGGACCGCTTGCCCCTGGGCTTTTATAAG hsplam5 GGATAACGGCTAACGGTGTACGTCAGCCCGGAAAAGTGCATATCCAG hsplam6 GTACCTAATATTTTCACGATGTTCTGCTGGATATGCACTTTTCCGGGCTG hsplam7 CATCGTGAAAATATTAGGTACTGTAAAAGCGGTGCCAGTCGGCATAC hsplam8 GCCGGGATGTGATCCACGGAGTATGCCGACTGGCACCGCTTTTAC hsplam9 TCCGTGGATCACATCCCGGCAAGCTTGGCACGCCAGTCGGGT S.pombeLmbdBrcd TCCGGATCCGTTTCTGCGGGAAAG
Construction of lox71-InvU4 lox71-InvU4S1 CCATGGCCTCCCTGGCCTACCGTTCGTATAGCATACATTATACGAAG lox71-InvU4AS1 CACAAATGCATACATATAGCCAGTGGATAACTTCGTATAATGTATGCTATAC InvU4S CACTGGCTATATGTATGCATTTGTG InvU4AS TCCGGATCCCGAAACTTTTTGACATCTAATTTATTCTGTTCC
Generation of sck1 stuffer Stuffer 5' Blp I TACAGCTCAGCTTCACAAAGAACAGG Stuffer 3' Sfi I TGAGGCCAGGGAGGCCTACGGACCAAATAAACTTGCCC
Generation of double strand barcode inserts Barcode_P_3-07 AGGCCCGGGCGAGTGT Barcode_3-07B† GCCTGGCCTCCCTGGCCANNNANANNNANANNNANANNNANANNNANANNNA NANNNACACTCGCCCGGGCCTCCC
TAIL PCR primers TAIL-LB2 CTCCATTAAGTAACAAATTCCTATTTAGAGAAAGAATGCTGAGTA TAIL-LB LOX71 AGCCAGTGGATAACTTCGTATAATGTATGCTATACGAACGGTA TAIL AD1† NGTCGASWGANAWGAA TAIL AD2† TGWGNAGSANCASAGA TAIL AD3† AGWGNAGWANCAWAGG TAIL AD4† STTGNTASTNCTNTGC TAIL AD5† NTCGASTWTSGWGTT TAIL AD6† WGTGNAGWANCANAGA
Other primers to determine insertion mutation sites BarcodePCR(888r) CACGACATGTGCAGAGATGCCGACGAAGCA InversePCR1 GGAAGGCATATCAGCAAAGACTTTCTCAGC InversePCR2 ACATGCTCCTACAACATTACCACAATCT Ade7_5'S (DH8C10) CTCCTGTTATCACGAAGCATGAAGAGT Ade7_3'AS (DH8C10) GGTCAACCAGTTTCGAAGATATTGCTTATC SPCA167.07c_'5S AACAAGCTAACGGTTATGCCCAGCCGTT (DH8H5)
85 Table 3.2. Oligonucleotides used in this study (continued) Oligonucleotide name Sequence SPAC167.97c_3'AS CACGACATTGTCTAGATAGCCTTGC (DH8H5) SPCC1442.04c_5'S AGATTCCACAAGTGAAGCCGAAACCCGA (10e1-36668) SPCC1442.04c_3'AS GATTCCTCAAGGTCGTTATCCCGCAAT (10e1-36668) Alp4 insert 5' (6a2- AGTTGCTCTGGCCTTCAACCTTCGATG 35750) Alp4 insert 3' (6a2- CTCAATCACGCGGCATACTGCCGACTT 35750) clg1 del 5' (1a7-4033) AGAACCTGCGCACAACCAACCACCGTAAAATT clg1 del 3' (1a7-4033) GAAGAGGGCGATGGTGGTGCGTTGGTGCT P12-SD-08-4031AS AAGAAAATGAATACACAGAGTCAGAGAAAGAGATAGAAGGA (1a7-4033) Brcd G1-08-4033 (1a7- TGCTATCCCTGTACTTTTCCTTATCCTTTTTCCTTTGGG 4033) SPAC24B11.08, 09 5' ACACAGGATCCGGCAACTTGTGCGATTTG del (1a8-4032) SPAC24B11.08, 09 3' AGCTAGCTTTTATTTGAAGATTAGGATGGCGT del (1a8-4032) † N = A, T, C or G; W = A or T; S = G or C
digestion with Blp I and Sfi I, the sck1 stuffer fragment was inserted at the corresponding sites on pInsertion-ura4. The resulting stuffer plasmid
(pInsertion-ura4-sck1) was then cut with Blp I, blunt ended with Klenow polymerase prior to Sfi I digestion. The digested DNA was resolved on 0.7% agarose gel to separate double-digested DNA (~6 kb) from the sck1 stuffer (~0.9 kb) and partially-digested DNA (~6.9 kb), which had a slower migration due to the presence of the 0.9 kb stuffer. The double-digested DNA was then purified from agarose gel by QIAGEN Gel Extraction Kit.
C. Insertion of the barcodes and bacterial library preparation
The double-stranded barcode inserts were generated from two oligo-
86 nucleotides, Barcode_3-07B and Barcode_P_3-07. Both oligonucleotides (20
μM) were phosphorylated by heating at 70oC for 5 minutes, chilling quickly on
ice and treatment with T4 polynucleotide kinase at 37oC for 1 hour. The two
oligonucleotides were then annealed together by slow cooling on a thermal cycler
using the following program: 95oC 1’30’’ (-1oC/cycle, 15 cycles), 80oC 2’ (-
0.5oC/ cycle, 70 cycles), 45oC 1’30’’ (-0.5oC/cycle, 66 cycles) and 12oC ∞.
The annealed oligonucleotides were converted to double strand DNA by treatment with the Klenow fragment of E. coli DNA polymerase I (3’-5’ exo-) at 37oC for 1
hour, followed by 75oC incubation for 20 minutes to inactivate the polymerase.
The blunt ends and CCC overhangs generated after Klenow treatment allowed
ligation of the inserts to the filled-in Blp I and Sfi I site of pInsertion-ura4.
The vector and insert DNA were ligated together at molar ratios of 1:1 (800 ng: 2.5 ng) or 1:3 (800 ng : 7.5 ng) with T4 DNA ligase (NEB) at 16oC for 16
hours. Ligated DNA (20 ng or 40 ng) was transformed to 25l of E. coli electrocompetent cells NEB 5-alpha in a 1 mm electroporation cuvette on BioRad
Gene Pulser II using the setting 1.7 kV, 200 Ω and 25 F. After electroporation,
975 l of SOC was added to the transformed cells, which were recovered at 37oC
with 250 rpm shaking for 1 hour. To determine the titer of transformation, 3 l
of the recovering culture was plated on LB + ampicillin (100 mg/l) plates in
duplicate. For the remaining cells, aliquots of 100 l were plated on the same
plates (10 plates per transformation) and grown at 37oC for overnight. Cells
grown on these plates were scraped off and grown in 100 ml of LB + ampicillin
medium at 37oC for 4 hours for plasmid DNA and freezer stock preparation.
87 A total of 37 transformations were performed to create 37 bacterial barcode
sub-libraries. The number of clones in each bacterial sub-library ranged from
8.5 104 to 2.6 105 and the total clones collected from all 37 transformations were 6.99 106.
Construction of the fission yeast barcode-tagged insertion mutant library
The linear insertion vector DNA was obtained by digesting the pInsertion- ura4-barcode library DNA with BamH I and gel purifying the 2.1 kb fragment followed by re-purification with phenol/chloroform/isoamyl alcohol (25:24:1; volume: volume: volume) extraction. For each transformation, 1 g of purified linear insertion DNA was used to transform 50 l of frozen KRP1 competent cells using the protocol published by Suga and Hatakeyama [340].
Transformed cells were plated on MMA + YC - uracil + 5-FOA (0.1 g/l)
(MMA + low FOA), where “YC - uracil” is the complete yeast supplements without uracil to allow recover auxotrophic mutations (Table 3.1), and grown at
30oC for at least 5 days. Colonies growing on MMA + low FOA plates were gridded on EMM + YC - uracil plates, grown for 3 days and replica plated on
YES plates. After growth on YES plates for 2 days, cells were replica plated to
EMM + YC - uracil and EMM + 5-FOA (1 g/l) + YC - uracil supplemented with a
low concentration uracil (50.25 mg/l) (EMM + FOA). Cells that grew on EMM
+ YC - uracil but not on EMM + FOA plates were inoculated in 96-well plates
with 200 l of YES medium per well. To confirm stable integration, selected
transformants from four 96-well plates were used to assemble 384-colony arrays
88 on EMM + YC - uracil and EMM + FOA omni plates using a 96-solid-pin
replicator (V&P Scientific, Inc). Unstable transformants (i.e. cells growing on
EMM + FOA omni plates) revealed in this step were removed from EMM + YC - uracil omni plates by pipetting cells out of the well.
Stable integrants were transferred from EMM + YC - uracil omni plates to
40 l of YES + 15% glycerol or EMM + YC - uracil + 15% glycerol medium in
384-well plates using a 384-solid-pin replicator (V&P Scientific, Inc). These
384-well plates were incubated at 30oC for two days before being stored at -80oC.
After transferring mutants to 384-well plates, cells left on 5 EMM + YC - uracil
omni plates were scraped off and grown in 50 ml of EMM + YC - uracil medium
for 4 hours at 30oC before aliquots were frozen as mixed library pools.
Determination of the size of S. pombe barcode-tagged insertion mutant library
and bacterial barcode library
We chose to generate 10,000 S. pombe insertion mutants as a balance
between the size of the yeast mutant library and the probability of obtaining a
mutant in every protein coding gene, which was calculated by the sampling
equation P = 1 – (1-f)N where P is the probability of finding any genes in the
genome, f is the fraction of a gene in the genome (gene size/genome size) and N is the number of insertion mutants generated [301]. Assuming that an insertion
in the 5’ UTR, introns, exons or the 3’ UTR can produce a mutant phenotype, and the average size of a S. pombe gene is 2 kb in a genome of ~14,000 kb, the
probability of finding at least a mutation in each individual gene in 10,000
89 random mutants P is 1 – (1 – 2/14,000)10,000 = 0.76 (or 76%).
To increase the chance of tagging individual S. pombe insertion mutants
with unique, non-repeated barcodes, only 250 to 1,500 S. pombe mutants were
generated from each bacterial barcode sub-library (the average number of barcode
clones in each sub-library is ~1.86 105). Thus, the number of S. pombe
mutants generated from each sub-library corresponds to less than 1% of the
available barcodes, and provides a 95% chance that all of the barcodes are
unique. A total of 18 bacterial sub-libraries were used to generate and tag
10,000 S. pombe insertion mutants.
Phenotypic assays to assess mutation diversity
Cells for phenotypic assays were first grown in 40 l of YES medium in
384-well plates for two days. For the temperature sensitivity test, cells were
transferred to two YES omni plates by a 384-solid-pin replicator, and one plate
was incubated at 30oC and the other at 36oC. Temperature sensitive mutants were scored as those grew at 30oC but not at 36oC after incubation for 4 days.
To identify auxotrophic mutations, cells were transferred from 384-well plates to
EMM + adenine, histidine, leucine and EMM + YC - uracil omni plates. Cells
with auxotrophic mutations grew on EMM + YC - uracil but not EMM + adenine,
histidine, leucine plates. For the identification of adenine biosynthesis
mutations, cells were transferred to YES - adenine omni plates to look for mutants
with altered colony color on this medium. To isolate petite-positive mutations,
cells grown in YES liquid medium were first transferred to 40 l of YES + 2%
90 potassium acetate + 12.5 mg/l of ethidium bromide (YES + EtBr) liquid medium
in 384-well plates for 2 days before printing these mutants on YES + EtBr omni
plates. Petite-positive mutations were scored as the ability to grow on the YES +
EtBr omni plates. For the above three assays, cells were grown at 30oC for 5
days before scoring the phenotypes.
Mutants identified from the 384-colony array assays were individually
verified by re-growing these mutants on the respective selectable media or
temperature as patches on regular Petri dish plates.
Identification of insertion sites by thermal asymmetric interlaced (TAIL)-PCR
TAIL-PCR uses alternate high and low annealing temperatures and
combinations of a set of arbitrary degenerate (AD) and three nested insertion
DNA-specific primers to amplify a small region of insertion DNA and the
adjacent genomic sequence. Three rounds of PCR, using alternate annealing
temperatures, the same degenerate primers and one of the three nested specific
primers in each round, yield one or a few bands in the tertiary PCR (Figure 3.2)
[324]. The primary PCR used the genomic DNA of S. pombe insertion mutants
as the template. One μl of 50-fold diluted products from the primary and the
secondary PCR was used as the template in the secondary and the tertiary PCR,
respectively. The insertion DNA-specific primers used were TAIL-LB LOX71,
TAIL-LB2 and haplam3. The degenerate primer (AD1 - AD6) and PCR cycles
were the same as described in [324]. TAIL-PCR products were treated with
exonuclease I and shrimp alkaline phosphatase (Exo-SAP, USB) to remove free
91 dNTPs and primers prior to sequencing. The products from the secondary and the tertiary PCR were sequenced with primers hsplam5 and hsplam7, respectively.
Figure 3.2. Amplification of insertion vector-genomic DNA junction by TAIL-PCR. (A) A single TAIL-PCR consists of three consecutive reactions. The first two reactions utilize alternating high and low annealing temperatures to allow insertion-DNA specific primers with high annealing temperatures and arbitrary degenerate (AD) primers with low annealing temperatures to bind to the templates. The tertiary cycle uses only a low annealing temperature to amplify products from the secondary cycle (this figure is adapted from [324]). (B) A set of AD primers and three nested insertion vector- specific primers are used in primary (green arrow), secondary (pink arrow) and tertiary (blue arrow) PCR reactions, respectively. Products from secondary or tertiary PCR reactions can be used in sequencing reactions to determine the genomic DNA adjacent to the insertion vector.
92 Results
Design of a linear barcode-tagged DNA vector for insertion mutagenesis in the fission yeast S. pombe
To create a fission yeast mutant library that allows one to study gene
function by analyzing these mutants both individually and in parallel as cultures
of mixed mutants, we chose to use insertion mutagenesis to generate random mutations tagged with unique barcode sequences. One potential advantage of this approach is the variety of the mutations that could be created. Insertions in the coding sequences can produce truncated proteins with no function, partial function or altered function. Insertions in the 5’ untranslated regions (UTRs) can change protein expression by inactivating promoters or binding sites of transcriptional regulators. Insertions in the 3’ UTRs may result in decreased mRNA stability and therefore lead to reduced expression of the encoded proteins
[242, 310, 331].
The insertion mutagenesis approach used in this work relies on non- homologous recombination-based integration of a linear DNA vector in the genome of a fission yeast strain that has no homology to the sequence on the insertion DNA vector. Two previous studies characterized two major classes of transformants that could be obtained when transforming linear DNA with no or very limited homology to the genome into S. pombe cells [106, 348]. In one type of transformants, the linear DNA vector was stably inserted in the genome with a broad distribution as single or tandem copies [106, 348]. Deletions on both ends of the insertion DNA vector and the adjacent genome sequences were
93 also detected [348]. The other type of transformants had the transfected DNA vector maintained as unstable and self-replicating circular DNA despite of the absence of an autonomously replicating sequence on the selectable marker ura4+ used in both studies [106, 348]. This kind of unstable cells was found to constitute the majority of the transformants in later studies [42, 52, 53].
This type of non-homologous recombination-dependent integration had been used as a mutagenesis method to create random mutants for genetic screens by
Chua et al. and Davidson et al [42, 52]. They chose to transform a ~1.7 kb ura4+ gene fragment to strains with the ura4-D18 mutation, which removes the ura4+ gene in the genome. Similar to the early work in S. pombe non- homologous recombination [106, 348], both studies isolated stable insertion mutants and unstable transformants, which were the major population and required laborious replica plating procedures on non-selective medium to be identified. One intriguing observation in these two experiments was that in most of the stable insertion mutants, the ura4+ DNA was inserted at a single locus in each transformant. This “one insertion per transformant” property is important for generating random mutants so that the phenotypes of selected mutants can be easily traced back to actual affected genes.
To test the feasibility of using the similar insertion mutagenesis to generate barcode-tagged insertion mutations in S. pombe, an initial barcoded ura4+ insertion vector was made by PCR amplification of ura4+ gene using ura4+- specific primers, of which the 3’ primer also contained a random barcode sequence in addition to the ura4+ 3’ sequence (Figure 3.3A). The Ura4p protein
94 has the activity of orotidine 5'-phosphate decarboxylase that can convert the
pyrimidine analog 5-FOA to toxic 5-fluorouracil. Ura4- mutant cells bearing
extrachromosomal ura4+ genes can lose ura4+ and are resistant to 5-FOA while
cells stably inheriting ura4+ on chromosomes die in the presence of 5-FOA [23,
107]. Therefore, using the ura4+ gene as the selectable marker provides a means
for quickly distinguishing stable and unstable transformants. Following
transformation to the wild type yeast, transformants were first selected on EMM +
YC - uracil to look for Ura+ transformants and then screened for failure to grow
on EMM + FOA as the toxic 5-FOA will kill the stable insertion mutants that cannot lose the ura4+ gene. It was found that barcodes were absent in most of
the stable insertion mutants (~95%), indicating that successful integration was
accompanied with deletion of barcodes. This result is consistent with previous
observations of microdeletion at the ends of the inserted DNA [42, 52, 348].
The insertion DNA vector was re-constructed to protect the barcode from degradation and increase the versatility of the final mutant library (Figure 3.3B).
At the 5’ end of the ura4+ gene, a 27 bp degenerate sequence with 12 intermittent
A’s to exclude BamH I and Sfi I restriction sites was added as the barcode tag.
Barcode tags are flanked by two Sfi I restriction sites so the barcodes can be
amplified by PCR, cut with Sfi I and oligomerized in a unique direction. In
addition, a mutated human HSP70 mini promoter [279] with a LexA binding site,
which allows activation of the promoter in the presence of the LexA-VP16 trans-
activator, was inserted at the 5’ end of the barcode, and a mutated loxP sequence,
lox71, was added between the barcode sequence and the 5’ end of ura4+ gene.
95 The lox71 mutant site allows integration of a plasmid bearing the lox66 mutant site, and the recombined mutant site is a poor substrate for further recombination due to its reduced binding affinity to Cre recombinase [5]. This mutant loxP system permits stable and quick modification of the mutant library. To prevent deletion of the barcodes during integration of the vector DNA, a 250 bp phage buffer sequence was placed in front of the barcode and HSP70 promoter. The buffer was assembled by overlapping PCR to remove all ATG codons on the transcribed strand. Therefore, when the HSP70 promoter is activated, translation will not start within the buffer sequence but in the adjacent genomic sequence.
Figure 3.3. The schematic of the insertion vector used to construct the barcode-tagged insertion mutant library. (A) The initial insertion vector consists of a selectable marker ura4+ gene and a random barcode (15 degenerated nucleotides with 8 intermittent T’s) directly following the 3’ UTR of ura4+. (B) The insertion vector used to construct the S. pombe insertion mutant library is composed of a selectable marker ura4+ gene, a random barcode (27 degenerated nucleotides with 14 intermittent A’s), a lox71 site for one-way integration, a mutated human HSP70 mini-promoter with a LexA binding site allowing induction of adjacent gene upon LexA-VP16 expression and a modified phage sequences, ATGless-, to protect the sequences 3’ the phage fragment.
96 Construction of the bacterial barcode-tagged insertion DNA library
The pInsertion-ura4 vector was used to create a bacterial barcode insertion
DNA library by ligating the double-stranded barcode inserts to the vector. The
DNA barcode used in this work was designed with 27 random bases, with a total
of 427 or 1.8 1016 possible barcode sequences. Our goal was to generate
10,000 yeast mutants with unique and non-redundant barcodes. We generated a
library of 6.99 106 bacterial clones as 37 sub-libraries of 1 to 2 105 clones per
sub-library. These sub-libraries were produced to help maintain the high complexity of barcodes in the event that some of the random sequences would alter the selection properties of the plasmid and bias the pool. In general, the
number of barcode-tagged fission yeast mutant produced per sub-library was
about 1% (1,500) of the total number of barcode clones in that sub-library.
Construction of the fission yeast barcode-tagged insertion mutant library
It had been shown that when linear DNA is transformed into S. pombe cells, it can be converted to unstable and self-replicating circular DNA in majority of transformants (70-90%) [42, 52, 53, 106, 348]. On regular 5-FOA selection medium that contains low amounts of uracil, these cells (with endogenous ura4+ gene mutated or deleted) are able to survive by losing the circular ura4+ DNA
(Figure 3.4). In the other 10-30% of transformants, the linear DNA can stably
integrate as a monomer or in concatomeric arrays at a single or multiple loci in
the genome [42, 52]. Cells with stable ura4+ integration constantly produce
Ura4p and are sensitive to 5-FOA (Figure 3.4).
97
Figure 3.4. Discrimination of different products of linear insertion DNA with an ura4+ marker after transformation into fission yeast. After transformed into fission yeast cells, the linear insertion DNA with an ura4+ gene can be circularized and maintained as multi-copy extrachromosomal DNA circles or integrate in the genome as single or multiple tandem copies. Cells carrying circular insertion DNA can lose ura4+- containing circles and grow on 5-FOA medium. However, cells with a stably integrated insertion vector are sensitive to 5-FOA and represent the desired stable insertion mutants which were collected in the mutant library.
Cells bearing the ura4+ circles are expected to have multiple ura4+ genes
per cell, produce a higher level of Ura4 protein, and thus be more sensitive to
immediate exposure to low concentrations of 5-FOA (Figure 3.5A). However,
stable integrants with one or few copies of integrated ura4+ DNA produce less
Ura4 protein and are less sensitive (Figure 3.5B). To help distinguish these two
types of cells and reduce the number of unstable transformants, 5-FOA was added
in the selection medium without uracil at a low concentration (0.1g/l), which
allowed wild type cells with a single copy of ura4+ gene to grow (data not shown) and was expected to be toxic to cells carrying multiple copies of circular ura4+
DNA due to increasing 5-flurouracil at high levels of Ura4 protein.
98
Figure 3.5. A selection strategy for cells containing a single copy and multiple copies of ura4+ genes based on the hypothetical metabolic outcome of altered Ura4p levels and low concentrations of 5-FOA on cell survival. (A) In cells with multiple copies of ura4+ genes, increased levels of Ura4p allow conversion of low dosage of 5-FOA to toxic 5- fluorouracil. These cells die in the medium supplied with a low concentration of 5-FOA (i.e 0.1 g/l). (B) In the presence of a single copy of ura4+ gene, endogenous orotidine-5- phosophate (orotidine-5P) may outcompete 5-FOA supplied in low concentrations as the preferred substrate of the limited amount of Ura4p, prevent Ura4p to metabolize 5-FOA to 5-fluorouracil and allow such cells to grow in medium with low 5-FOA.
Eighteen of the 37 bacterial barcode sub-libraries were used to generate the
initial transformants on MMA + low FOA plates, and these transformants were
verified for stable integration by growing these cells on regular EMM and EMM
+ FOA plates as individual patches. Stable insertion mutants (i.e. transformants
that grew on regular EMM but not EMM+ FOA plates) were collected and grown
in 200 l of YES medium 96-well plates (no selection for ura4+), and were then
used to make 384-colony arrays on omni plates. To further exclude false stable
insertion mutants, the 384-colony arrays were made on both EMM + YC - uracil
and regular EMM + FOA omni plates to confirm the Ura+ and 5-FOA-sensitive
phenotypes. Mutants that behaved as true integrants from five 384-colony arrays
99 were stored as a library pool of ~1,800 mutants or individually in 384-well plates
(Figure 3.6). A total of five pools of ~1,800 mutants per pool and one pool of
~1,100 mutants were constructed.
Figure 3.6. The flowchart of the fission yeast barcode-tagged insertion mutant library construction. The linear insertion DNA (Figure 3.3B) was transformed into the wild type strain KRP1 to obtain Ura+ transformants, which were tested for stable integration by 5- FOA sensitivity. Stable transformants (i.e. 5-FOA sensitive cells) were inoculated in YES medium without selection for ura4+ in 96-well plates, followed by printing 4 such plates to an EMM + YC – uracil and an EMM + 5-FOA omni plate to form 384-colony arrays for the second 5-FOA sensitivity test. Unstable transformants escaping the first selection were removed before these mutants were stored as 384-well mutant arrays or mixed mutant pools of ~1,800 mutants.
The insertion mutant library contains multiple diverse mutations
To evaluate the ability to isolate different mutants from this insertion mutant library and to assess the randomness of mutations, four easily scored phenotypic
100 assays were performed on 3,600 mutants. These assays isolated temperature- sensitive mutations (lack of growth at 36oC), adenine biosynthesis mutations
(altered colony color when grown on YES - adenine medium), auxotrophic
mutations (lack of growth on minimum medium), and petite-positive mutations
(growth in the presence of ethidium bromide (EtBr)).
We identified mutants in each assay and most of the mutants had only one
phenotype (Table 3.3). Expected mutation frequencies were calculated as the
ratio of the number of genes in each biological pathway (estimated from Gene
Ontology (GO) [308] or Kyoto Encyclopedia of Genes and Genomes (KEGG)
[169] databases) and the total number of S. pombe protein coding genes. Based
on the GO-predicted mutant frequency, 86 auxotrophic mutants should be present
in this group of 3,600 mutants; however only 30 were identified. Since only
~60% of the genome is protein coding [377], it is possible that half of the
insertions fell into intergenic regions, did not generate phenotypes that could be
detected in this assay, and therefore produced fewer mutants than predicted that
failed to grow on minimal medium. Only two mutations are known to cause red
pigment accumulation in cells: ade6+ and ade7+. One dark red mutant on YES-
adenine medium was identified as an ade7- insertion mutant. The frequency of auxotrophic and ade7- dark red mutants was consistent with nearly random
integration of the insertion vector in the S. pombe genome.
In addition, 27 temperature-sensitive mutants that failed to grow at 36oC and
13 petite-positive mutants that could grow on EtBr-containing medium were
identified, demonstrating that the library contained a diverse range of mutants.
101 Table 3.3. Characterization of the diversity of mutations in the barcode-tagged insertion mutants by 4 phenotypic analyses. GO or Expected Actual Mutant Phenotype KEGG number of number of Class expectation isolatesa isolates
Auxotrophic No or slow growth 2.4% 86 30 mutation on minimal medium (123b/5027c)
Color change (from NDd NDd 18 (whitee) Adenine pale pink to white
biosynthesis or red on low 0.04% mutation f c 1 (red) adenine medium) (2 /5027 ) 1-2
Petite- Growth on medium positive NDb NDb 13 with EtBr mutation
Temperature No or slow growth NDb NDb 27 sensitivity at 36oC
a. The estimated number of isolates is calculated as “the total number of mutants tested (3581)” “GO or KEGG expectation”. b. Estimated number is from AmiGO (Gene Ontology) database on Sanger Centre S. pombe page [308] with keywords “amino acid biosynthesis” and “nucleobase biosynthesis”. c. Total gene number in S. pombe genome (as of 5/6/2008) = 5027. d. Not determinable. e. White colony color can also result from mitochondrial defects [61, 177]. f. The estimated number is from KEGG [169] (ade6+ and ade7+).
Identification of the insertion sites by TAIL-PCR
Unlike the budding yeast gene deletion collection, this barcode-tagged
insertion mutant library was generated without any knowledge of the actual
mutations. Therefore, an approach that allows efficient and quick determination
of the insertion sites is required. It has been shown that TAIL-PCR could be successfully applied to large-scale determination of T-DNA insertions in
Arabidopsis thaliana mutant library [206, 324]. To determine if the same technique also works as efficiently in the fission yeast, TAIL-PCR was used to identify the insertion sites of mutants discovered in the four phenotypic assays
102 and an independent screen for long-lived mutants.
Among the 36 mutants tested, TAIL-PCR successfully amplified the
genomic DNA adjacent to the insertion vector from 10 mutants (Figure 3.7A, B and Table 3.4). There are 6 insertions located within protein coding sequences, one insertion found in a nuclear RNA gene and three insertions in intergenic regions. These results are consistent with the observed mutant frequencies found in the phenotypic screens.
Similar to previous studies, deletions were observed at both ends of the insertion vector and detailed analysis of some of the insertion mutants also revealed deletion of genomic sequences at the integration sites, ranging from 5 to
450 bp (Table 3.4) [42]. In some mutants, ura4+ DNA sequence was identified
at the 5’ end of the buffer sequence, an indication of tandem repeats of the
insertion vector that was also reported in a similar mutagenesis approach (Table
3.4 and Figure 3.7C and D) [52]. Among the 36 mutants analyzed,
mitochondrial DNA was also found co-integrated in the genome with the insertion
vector in 9 mutants (Figure 3.7E and data not shown). Fusion of mitochondrial
DNA to DNA introduced by transformation in this type of non-homologous
recombination was also reported previously [53].
103
Figure 3.7. Schematics of insertion mutations identified in this study. Gray color indicates intergenic regions; blue boxes are genes; represents the insertion vector; the only orange box in (E) symbolizes mitochondrial DNA. (A) Some insertions were identified in intergenic regions with no annotated protein or RNA coding genes (e.g. mutants DH6B20, 1a8-6535 and 3c2-6969 in Table 3.4). (B) In some mutants, a single insertion vector integrated in the middle of a gene (e.g. mutants DH9M24 and 1a8-4032). (C, D) Two (C, DH8C10 and 1a7-4033) or more (D, not shown in Table 3.4) insertion vectors were also found co-integrated in the genome as tandem (C) or reverse (D) repeats in certain mutants. Large deletions were found in additional copies of insertion vectors. (E) In several characterized insertions, mitochondrial DNA was found adjacent to the insertion vector. In these mutants with co-integrated mitochondrial DNA, TAIL-PCR was not able to extend the amplified products to the nuclear chromosomal sequences.
104 Table 3.4. Insertion sites, detailed structures of the integrated insertion vector and adjacent chromosome sequences determined by TAIL-PCR and sequencing
+ Base pairs deleted from ura4 Strain name Insertion sitea tandem repeat 5’ vector 3’ vector Chromosome
212 bp 5’ of SPBC1198.03c b; DH6B20b 11 NDc NDc NDc 294 bp 3’ of zas1+
392 (1st copy) 1093 (1st copy) DH8C10d 351 bp of ade7+e NDc Yes 13 (2nd copy) NDc (2nd copy)
DH8H5f 2419 bp of SPAC167.07c 11 5 5 Yesg
DH9M24f 1 bp of sin1+ 10 NDc NDc No
10e1-36668h 776 bp of SPCC1442.04c 9 6 0 Yesg
1a8-6535h 23 bp 5’ of trs130+ 10 NDc NDc NDc
238 bp 3’ of tbf2+ i; 3c2-6969h 22 NDc NDc NDc 237 bp 5’ of SPBC31F10.02
11 (1st copy); 5 (1st copy) 1a7-4033h 590 bp of clg1+ 4 Yes 10 (2nd copy) 976 (2nd copy)
1a8-4032h 905 bp of SPNCRNA.142 36 17 0 No
a. For intergenic insertions (DH6B20, 1a8-6535 and 3c2-6969), the distances between the insertion vectors to the surrounding genes are relative to start (5’) or the stop (3’) codons; for ORF insertions (the rest of the mutants), the numbers indicate the distance between ATG of the corresponding genes and the closest end of the insertion vector. b. An auxotrophic mutant that could not grow in minimum medium. The distance of the insertion site to SPBC1198.03c was estimated with the assumption that no deletion occurred on the chromosome. c. ND = Not determined. d. Adenine biosynthesis mutant; red colony on YES-adenine medium. e. A gene, SPNCRNA.568, producing the antisense transcript overlapping with 95% of ade7+ ORF is annotated on the complementary strand. f. Petite-positive mutants that could grow in ethidium bromide medium g. The size of amplified mutated alleles is larger than that of a single-copy-insertion prediction. h. Mutants were identified in a genetic screen for long-lived mutants. i. The distance of insertion site to tbf2+ was estimated with the assumption that no deletion occurred on the chromosome.
105 Discussion
The availability of barcode-tagged mutant libraries in different model systems has greatly facilitated the genome-wide investigation of gene interactions and analysis of biological process with complex phenotypes. In this study, we extended the advantage of this approach to the fission yeast S. pombe by creating an insertion mutant library in which all the mutants were tagged with unique barcode sequences. Although insertion mutagenesis had already been successfully used in fission yeast, the mutants created in previous studies were only allowed for those specific selections and could not be used in later screens without re-making the mutant pools. Furthermore, these previous insertion mutagenesis did not include barcode tags, could only be used in positive selection for mutants with distinct phenotypes and required laborious work on identification of individual mutants [42, 52].
In this study, the insertion mutants were tagged with unique barcodes and stored as pools of 1,800 mutants as well as 384-well arrays of individual mutants.
The 384-well mutant arrays allows genetic screens on individual mutants and can be extended to genetic approaches such as synthetic genetic array [289, 351].
The barcode tag in each mutant also allows quick identification of mutations of interest in parallel selections with thousands of mutants in a single culture. After identifying the most abundant barcodes in a selected culture, these barcodes can be used as specific primers to isolate the corresponding mutants from the 384-well plates or mini-arrays composed of the resulting selected cells. This approach provides significant advantages for genetic screens with difficult or weak
106 phenotypes (e.g. chronological aging). In addition, the barcodes were designed
in a way that only requires routine molecular biology techniques such as PCR,
restriction enzyme digestion, ligation and sequencing to analyze the final
composition of selected barcodes, which reveals the proportion of the
corresponding mutants after selection. Therefore, the more expensive and
sometimes not readily available microarray approach is not necessary for this
barcoded mutant library.
It has been shown that linear DNA preferentially integrates into
chromosomes through non-homologous recombination in fission yeast unless
significantly long, homologous sequences are present on the ends of the targeting
vector DNA [106, 348]. Taking advantage of this property, we designed an
insertion vector with an ura4+ gene as the selectable marker and used it to create
insertion mutations in a wild type strain that has the ura4-D18 deletion mutation, which removes the corresponding ura4+ sequence on the chromosome. Due to
the lack of homology to the genome sequence, the insertion DNA vector was
expected to have a broad distribution in the genome and create diverse mutations
such as loss-of-function (insertions at ORFs) and mis-regulation (insertions at
regulatory elements) mutations.
One of the additional features of this insertion vector is the inclusion of a
LexA-VP16 transactivator-inducible HSP70 mini-promoter that may drive the
expression of genes adjacent to the insertions. The 94-bp HSP70 mini-promoter
contains the region from -80 to +14 of human HSP70 5’ UTR (the numbers are
relative to the dominant transcription initiation site of the promoter when assayed
107 in mouse BALB/c 373 cells and fission yeast) and includes the transcription start site, HSP70 TATA box, and two upstream promoter elements, CCAAT and Sp1 binding site. Prentice and Kingston showed that this promoter is active in fission yeast cells and mutations in the CCAAT element and Sp1 binding site can significantly reduce its activity [103, 279]. A LexA protein binding site placed
5’ of the TATA box region should allow activation of this mutated promoter to drive the expression of genes adjacent to the insertions. Because of the presence of the transcription initiation site on this mutated promoter, when the insertion vector is in-frame integrated in or near open reading frames, expressing LexA-
VP16 can create additional phenotypes in such mutants by producing truncated proteins or altering the expression level of proteins.
This insertion vector also may overcome some of the challenges in the study of essential genes in haploid microorganisms. In diploid model organisms, creating heterozygous deletion mutations in essential genes allows the effect of haploinsufficiency to be examined [55, 92]. Otherwise, viable mutations in essential genes can only be generated by chemical or radiation mutagenesis and isolated by certain selection criteria such as temperature sensitivity. However, identification of the mutated genes could be challenging and time-consuming.
Such mutants are also not available in haploid ORF deletion mutant banks such as the one recently released by Bioneer, Inc [176]. A potential advantage of the insertion mutant library made in this work is that some insertions could result in viable mutations in essential genes (e.g. insertions at 5’ or 3’ UTRs of essential genes would only change their expression) and the presence of the insertion
108 vector allows rapid discovery of the mutated sequences. The inclusion of these
mutants may allow functional analysis of essential genes in a variety of biological
processes.
Transformation of linear ura4+ DNA in fission yeast cells carrying a
complete ura4+ deletion (ura4-D18) can either result in integration of the ura4+
DNA into the genome or circularization of the same DNA to form extra-
chromosomal plasmids that can be replicated in the absence of a known
autonomous replicating sequence [53, 106]. It has been suggested that the free
ends of ura4+ DNA were recognized as double strand breaks and repaired by
NHEJ or MMEJ to form circular ura4+ DNA. The same study also showed that
these ura4+ circles contained mitochondrial DNA insertion at the repaired junctions and indicated that this kind of extranuclear DNA insertion appears to be a common byproduct of non-homologous integration in fission yeast [53]. In the transformants analyzed here, ~90% of Ura+ transformants turned out to be non- stable transformants as revealed by vigorous growth on EMM + FOA plates.
These unstable transformants might be cells carrying circularized ura4+ DNA.
Although these unstable transformants were not further examined, there were
mitochondrial DNA fragments of various sizes identified at the insertion vector
and genomic DNA junctions in many stable ura4+ integrants. These data suggest that insertion of the barcode-tagged ura4+ DNA in this study may involve
similar mechanisms (i.e. NHEJ, MMEJ). In some stable integrants, multiple
copies of ura4+ DNA were found to integrate at the same locus as tandem repeats,
a phenomenon also reported previously [52].
109 One consequence of the extra mitochondrial and ura4+ DNA insertions is
that they impede the identification of the insertion sites by TAIL-PCR in three possible ways. First, co-integration of long mitochondrial or extra copies of insertion vector could prevent the PCR product to be made. Second, the presence of mitochondrial DNA could provide sequences annealing to the degenerate primers and prevent the synthesis of PCR product starting from the nuclear genomic sequence. Third, tandem integration of the insertion vector
DNA can also limit the PCR amplification within the insertion vector sequence.
It is important to note that all of the insertions are tagged by ura4+, a unique
DNA sequence in the cell that is also a selectable marker that can be followed in yeast and E. coli (by complementation of the pyrF mutation). Therefore, insertions that can not be identified by TAIL-PCR can still be identified by cloning the DNA fragment that bears ura4+. For this reason and aforementioned
analyses, this S. pombe barcode-tagged insertion mutant library provides an
important resource for high-throughput aging studies. As described in the next
chapter, a proof-of-principle experiment with a subset of this library was
successful in isolating a long-lived mutant from a pool of random mutants.
110 Chapter 4
A novel parallel selection approach identified a cyclin/CDK
complex, Clg1p/Pef1p, whose inactivation leads to lifespan
extension in the fission yeast Schizosaccharomyces pombe
111 Abstract
Genome-wide studies in the budding yeast Saccharomyces cerevisiae have
uncovered several evolutionarily conserved genes and pathways that regulate
aging and longevity. The current approaches rely on analysis of individual ORF
deletion mutants from a large, pre-defined library or microarray analysis of the
barcodes associated with these deletions. The construction of this barcode-
tagged ORF deletion strain set was a major effort that is difficult to extend to
other model systems. This chapter describes how our novel, barcoded
Schizosaccharomyces pombe insertion library allowed for the parallel selection of
mutants with extended chronological lifespan using only routine molecular
biology techniques and without prior knowledge of the barcode sequences that mark the mutations. Chronological lifespan is essentially a fitness test that measures the length of time cells can survive in stationary phase. In a culture of several thousand mutants, the mutants with longer lifespans, and their associated barcodes, increase in proportion as the cells with shorter lifespans die. By amplifying, oligomerizing and sequencing the barcodes from the viable cells late in the lifespan, we identified a long-lived insertion mutation in the cyclin-coding gene clg1+/mug80+. Complete deletion of the clg1+ ORF also extended lifespan.
Pas1p, Clg1p and Psl1p are three cyclins that associate with the cyclin-dependent kinase Pef1p, but only loss of Clg1p extended lifespan. This increased longevity required the Pef1p-associated kinase Cek1p. These results demonstrate that long-lived mutants affecting evolutionarily conserved pathways can be directly selected from a pool of random S. pombe mutants, and provides an example of
112 how to generate and use similar mutant libraries in other microbial systems without prior knowledge of the barcode sequences that tag the individual
mutations.
113 Introduction
It has been more than half a century since individual cells were
experimentally shown to have finite lifespans [122, 239]. These observations
prompted studies using a variety of model organisms, such as the budding yeast
Saccharomyces cerevisiae, the nematode Caenorhabditis elegans and the fruit fly
Drosophila melanogaster to uncover genetics that regulate lifespan expectancy
and soon led to the identification of the first single genes whose de-regulation or mutation caused lifespan extension in yeast, worms and flies [41, 87, 203].
Among these popular model organisms, the budding yeast S. cerevisiae has
proven to be quite valuable and versatile in characterizing complex phenotypes and molecular mechanisms of aging and longevity due to its sequenced, well annotated genome, powerful molecular genetics and relatively short lifespan.
Many lifespan-extending interventions observed in S. cerevisiae (e.g. caloric restriction, inactivation of nutrient sensing signaling TORC1, PKA and AKT) have also been shown to be conserved in other model organisms [81, 151].
Therefore, extended studies in yeast aging may facilitate identifying new genes and pathways that control aging in mammals and help the development of therapeutic agents for aging-associated diseases.
Although studies on individual genes have uncovered some key regulators of yeast aging, genome-wide analyses on genes important for the maintenance of reproductive ability and survival in quiescence and long lifespan in yeast should
provide more comprehensive understanding of aging and longevity. Several
pioneering large-scale studies using the barcode-tagged ORF deletion mutant set
114 or randomly generated temperature-sensitive mutants have demonstrated success in exploring the effects of different caloric restriction methods on aging, isolating long-lived mutations affecting TOR signaling, and identifying genes required for survival in quiescence [160, 245, 278, 299]. However, one challenge for this type of work is to create a large set of barcode-tagged ORF deletion mutants or random mutants that allow efficient isolation of lifespan-extending or shortening mutations. Also, the related approaches usually require extensive labor or a specialized equipment to score the complex phenotype of aging (i.e. the length of time of stationary phase survival or number of maximum divisions of hundreds to thousands individual mutants) [160, 245, 299]. Another approach is to pre-select for easily distinguished phenotypes that strongly associate with aging or longevity, such as stress sensitivity. Although a longevity mutation in the budding yeast AKT-coding gene SCH9 was discovered by this type of method
[77], lifespan-regulating genes that are not involved in stress response would not be identified. For these reasons, novel methods that allow efficient and unbiased identification of long-lived mutations will facilitate aging research.
The fission yeast S. pombe has recently emerged as another useful model organism to study molecular aspects of eukaryotic cell aging [39, 295].
Molecular phylogenetic comparisons indicate that S. pombe is as evolutionarily
distant from the budding yeast S. cerevisiae as humans are from C. elegans [124], suggesting that S. pombe is a distinct model system for biological research. S.
pombe also appears to be closer to the common ancestor of metazoans and fungi,
and shares some molecular aspects with humans that are absent in the budding
115 yeast, such as the presence of an RNAi machinery, a complex centromere
structure, similar RNA splicing mechanisms and the inability to lose the
mitochondrial genome in wild type cells [326, 377]. Therefore, aging research
in S. pombe has the potential of discovering new mechanisms that are shared in
fission yeast and mammals but not in budding yeast.
To date, evolutionarily conserved interventions that extend lifespan in other
organisms have also been shown to extend lifespan in S. pombe, such as caloric
restriction, deletion of PKA-coding gene pka1+ and AKT kinase orthologs sck1+ and sck2+ [39, 297]. As a new model system for aging research, S. pombe
lacked many of the extensive tools in the S. cerevisiae system that allow genome-
wide and high-throughput analyses (e.g. the barcoded ORF deletion mutant collection [91, 375], synthetic genetic array (SGA) [351]). A private company,
Bioneer, Inc., has created collections of S. pombe haploid and diploid gene
deletion mutants that employed a barcode-tagging strategy similar to that used in
the budding yeast ORF deletion mutant collection and started to distribute these
mutant sets to the S. pombe research community in 2007. The barcode
sequences in these mutants were only released in mid-2010, which made it
impossible to perform the large-scale parallel analyses possible in S. cerevisiae.
Furthermore, many of the deleted genes are only missing a part of the coding
sequences. For some genes, the deleted portion can be less then 10 % of the
coding sequence [176]. Whether these partial gene deletions are true null
mutations are unknown and they could also make interpretation of the
experimental outcome difficult (e.g. loss-of-function versus dominant-negative).
116 To generate a tool to allow high-throughput analysis of aging in S. pombe,
we constructed a mutant bank by non-homologous recombination-based integration of insertion DNA vectors that carry unique 27-bp molecular barcodes.
The design of the barcodes allows parallel selection of a large pool of mutants without prior knowledge of the barcode sequences and requires only routine molecular biology techniques (e.g. PCR, restriction enzyme digestion, ligation, sequencing) to analyze selected mutants.
We used this mutant library to perform an unbiased screen for long-lived mutants and isolated a strain bearing a mutation in the cyclin-coding gene clg1+/mug80+. The fission yeast Clg1p is homologous to the budding yeast
Clg1p, which belongs to a cyclin family of 10 members that all associate with the
cyclin-dependent kinase (CDK) Pho85p. These Pho85p-associating cyclins
(Pcls) are generally divided into two groups – Pcl1p, 2p-like subfamily (Pcl1p,
Pcl2p, Clg1p, Pcl5p, Pcl9p) and Pho80p-like subfamily (Pho80p, Pc16p, Pcl7p,
Pcl8p, Pcl10p), and regulate a wide variety of biological functions in S.
cerevisiae, including cell cycle progression, morphogenesis, carbon source utilization, glycogen and phosphate metabolism [38, 226]. Consequently, deletion of the PHO85 gene results in pleiotrophic phenotypes [38].
Pho80p was the first Pho85p cyclin discovered [162, 257]. A recent study on Pho80p/Pho85p complex identified the protein Rim15p as its new substrate
and suggested an intriguing role of Pho80p/Pho85p complex as a negative
regulator of quiescence [369]. It has also been shown that Rim15p is required
for survival in stationary phase and lifespan extension by mutations in SCH9,
117 CYR1, TOR1 and caloric restriction by reducing glucose levels [77, 369, 371].
Here we show that in S. pombe, Clg1p associates with the CDK Pef1p and
regulates lifespan in a manner similar to the Pho80p/Pho85p complex. Of the
two S. pombe genes whose proteins have the strongest sequence similarity to the
S. cerevisiae Rim15p, Cek1p appears to be the Rim15p ortholog based upon its
association with Pef1p and its requirement in clg1-mediated lifespan extension.
These results also demonstrate that our barcoded insertion mutant library can be used to select for long-lived mutants in evolutionarily conserved lifespan- extending pathways.
118 Materials and Methods
Strains and media
The fission yeast strain KRP1 was used to construct most of the mutants
used in this study (Table 4.1). KRP34 was made by transforming a 3.2 kb ura4+
fragment that converted the ura4-D18 deletion mutation in KRP1 to the wild type
ura4+ allele. KRP83 and KRP84 were constructed by transforming a 0.3 kb
leu1+ gene fragment that overlaps the region containing the leu1-32 mutation into
KRP1 and KRP34, respectively, to convert them to Leu+. KRP70 (cek1) is
from a haploid deletion mutant set purchased from Bioneer, Inc. (strain number
BG4355H). K566-11, expressing C-terminally 3HA-tagged Pef1p, was
described in Tanaka and Okayama [344]. The fission yeast strains used in this
study are listed in Table 4.1. The budding yeast strain Y187 (MATα, ura3-52, his3-200, ade2-101, trp1-901, leu2-3, 112, gal4Δ, met-, gal80Δ, MEL1, URA3::
GAL1UAS-GAL1TATA-lacz; Clontech) was used in the two-hybrid assay. The
barcode-tagged fission yeast insertion mutant library used in the long-lived
mutant screen was described in Chapter 3. Mutants from pool#1 and #2 (~3600
mutants in total) were used in this screen.
Standard fission yeast growth media were used in this study. Unless
otherwise specified, yeast extract + 225 mg/l of supplements (YES) [237] and
synthetic dextrose + 150 mg/l of supplements (SD) [39, 291] contain 3% glucose;
Edinburgh minimal medium (EMM) and EMM glutamate (EMMG) contain 2%
glucose and 225 mg/l of supplements [237].
119 Table 4.1. Fission yeast strains used in this study Strain Genotype Comment - KRP1 h ade6-M216 ura4-D18 leu1-32 his7-366 Wild type, [4, 39] - + KRP34 h ade6-M216 leu1-32 his7-366 Wild type, ura4 , derived from KRP1 - + KRP83 h ade6-M216 ura4-D18 his7-366 Wild type, leu1 , derived from KRP1 - + + KRP84 h ade6-M216 his7-366 Wild type, leu1 ura4 , derived from KRP34 - KRP40 h ade6-M216 ura4-D18 leu1-32 his7-366 Re-constructed mutant spncrna.142::barcode-ura4+† - KRP42 h ade6-M216 ura4-D18 leu1-32 his7-366 Re-constructed mutant clg1:: barcode- ura4+† - KRP43 h ade6-M216 ura4-D18 leu1-32 his7-366 Original mutant isolate sprrna.47::barcode-ura4+† - KRP44 h ade6-M216 ura4-D18 leu1-32 his7-366 Original mutant isolate clg1:: barcode-ura4+† - KRP45 h ade6-M216 ura4-D18 leu1-32 his7-366 Original mutant isolate spncrna.142:: barcode-ura4+† - + KRP87 h ade6-M216 ura4-D18 leu1-32 his7-366 clg1::ura4 + + KRP88 h ade6-M210 ura4-D18 leu1-32 his7-366 clg1::ura4 - + KRP131 h ade6-M216 ura4-D18 his7-366 pef1::ura4 - + KRP138 h ade6-M216 leu1-32 his7-366 clg1::leu1 - + KRP139 h ade6-M216 ura4-D18 leu1-32 his7-366 clg1::leu1 pef1::ura4+ - + KRP92 h ade6-M216 ura4-D18 leu1-32 his7-366 ppk18::ura4 - ‡ + KRP109 h ade6-M216? ura4-D18 leu1-32 his7-366 clg1::ura4 Spores from KRP88 (KRP111) ppk18::ura4+ KRP92 + * KRP70 h ade6-M210? ura4-D18 leu1-32 cek1::KanMX BG4355H of Bioneer S. pombe haploid deletion set version1 - * + KRP171 h ade6-M210? ura4-D18 leu1-32 his7-366 clg1::ura4 Spores from KRP70 nd cek1::KanMX KRP87 (2 backcross) - * KRP173 h ade6-M210? leu1-32 his7-366 cek1::KanMX Spores from KRP70 KRP87 (2nd backcross) - + KRP102 h ade6-M216 ura4-D18 leu1-32 his7-366 psl1::ura4 - + KRP103 h ade6-M216 ura4-D18 leu1-32 his7-366 pas1::ura4 - + + K566-11 h ura4-D18 pef1<< pef1HA -ura4 leu1-32 From [344] † These genes were disrupted by the insertion vectors containing both ura4+ and additional DNA sequences as described in the text and Chapter 3. ‡ The ade6 marker was not tested after tetrad dissection. * The strains derived from the Bioneer S. pombe haploid deletion set may be ade6-M210 or ade6- M216, and the exact genotype is not specified in the user manual for each strain.
120 Chronological aging assays
Chronological aging assays were performed as described in Chapter 2 as
well as Chen and Runge [39]. Briefly, growing cells were seeded into 30 ml of
SD medium with 3% glucose and 150 mg/ml of adenine, histidine, leucine, uracil at the initial cell density of 5 104 cells/ml in 125 ml flasks to maintain a constant medium/flask volume ratio, and grown in an enclosed air platform shaker rotating at 220 rpm at 30°C. Cultures were grown until they reached the maximum cell density (2 days) and this time point was designated as Day 0. Aliquots of cultures were taken on the days indicated, and multiple dilutions were plated on
YES plates in duplicate and grown at 30oC for 4 days. Colonies formed on the
YES plates were counted and used to calculate the number of colony forming
units per ml (CFU/ml) of culture. For each experiment, CFUs were monitored
until they reached < 10/ml. Each experiment was done at least twice with two
independent gene deletion transformants.
A parallel selection for long-lived mutants using PCR-mediated barcode
sequencing
For the long-lived mutant screen, S. pombe barcode-tagged insertion
mutants from pool#1 and #2 freezer stocks were thawed on ice, plated on EMM +
YC -uracil plates (~105 cells/plate, 12 plates for each mutant pool) and grown at
30oC for 5 days. Revived cells were scraped off the plates and resuspended into sterile milliQ water to inoculate the selecting cultures at the density of 5 104
cells/ml in 1000-ml flasks in duplicate with 240 ml of SD medium with 3%
121 glucose, 150 mg/l of adenine, leucine, histidine and uracil. The CFUs of the cultures were monitored at 30oC with 220rpm shaking for 15 days.
To identify the long-lived mutants, 600 surviving cells were recovered from
the day 14 culture. To facilitate the isolation of mutants enriched after the
selection, these 600 mutants were divided to 6 groups for genomic DNA
preparation and subsequent analyses. DNA fragments (~ 750 bp) containing
barcodes were amplified from genomic DNA of the surviving cells by PCR using
primers hsplam6 and BarcodePCR (888r) (primers used in this study are listed in
Table 4.2). The amplified DNA was digested with Sfi I and separated on a 2 %
low melting agarose gel. The gel slice (~0.3 2 cm) containing the Sfi I
overhang-bordered barcode DNA (66 bp) was melted at 65oC with 100 l of 1X
TE, 70 l of 3M sodium acetate, pH 5.2 and then mixed with 0.6 ml of TE-
saturated phenol prior to centrifugation at 13,200 rpm for 5 minutes. The
aqueous phase was re-extracted with 0.6 ml of phenol/chloroform/isoamyl alcohol
(25:24:1; volume: volume: volume), followed by extraction with 0.6 ml of
chloroform/isoamyl alcohol (24:1). The final aqueous phase solution (~ 0.5 ml)
was precipitated with 50 l of 3M sodium acetate, pH 5.2, 1.1 ml of 100% ethanol
at -80oC overnight and the precipitated DNA was washed with 1 ml of 70%
ethanol. The resulting barcode DNA was dissolved in 30 l of 10mM Tris-HCl,
pH 8.0. Barcode DNA (~1 g) was oligomerized by T4 DNA ligase (400,000
units/ml, NEB; 600 units at the beginning of the reaction and adding another 400
units after 8 hours) with 15% polyethylene glycol (PEG) 3350 in a 20-l reaction at 16oC for 16 hours. The oligomerized barcode DNA was purified by a
122 QIAGEN PCR Purification column first to remove PEG and then resolved on a
2% low-melting agarose gel. Barcode oligomers with the size between 0.3 and 1 kb were purified from gel using the method described above. The purified long barcode oligomers were ligated to Sfi I-digested and alkaline phosphatase (CIP)-
treated pInsertion-ura4 vector (reviewed in Chapter 3) and transformed to E. coli.
Table 4.2. Oligonucleotides used in this study Oligonucleotide name Sequence Barcode amplification and sequencing hsplam6 GTACCTAATATTTTCACGATGTTCTGCTGGATATGCACTTTTCCG GGCTG BarcodePCR(888r) CACGACATGTGCAGAGATGCCGACGAAGCA TAIL-LB LOX71 AGCCAGTGGATAACTTCGTATAATGTATGCTATACGAACGGTA
TAIL PCR hsplam3 AAGGGAAAAGTCGACTCTCCGTGACGACTTATAAAAGCCCAGGGGCAAG hsplam5 GGATAACGGCTAACGGTGTACGTCAGCCCGGAAAAGTGCATATCCAG hsplam7 CATCGTGAAAATATTAGGTACTGTAAAAGCGGTGCCAGTCGGCATAC TAIL-LB2 CTCCATTAAGTAACAAATTCCTATTTAGAGAAAGAATGCTGAGTA TAIL-LB LOX71 AGCCAGTGGATAACTTCGTATAATGTATGCTATACGAACGGTA TAIL AD1† NGTCGASWGANAWGAA TAIL AD2† TGWGNAGSANCASAGA TAIL AD3† AGWGNAGWANCAWAGG TAIL AD4† STTGNTASTNCTNTGC TAIL AD6† WGTGNAGWANCANAGA
Barcode primers Barcode_4030 CATTGTTCTTGTTGCTTTATATTTGTCTATTGCTATCAT Barcode_4031 CCTTCTATCTCTTTCTCTGACTCTGTGTATTCATTTTCT Barcode_4032 TCATCTTTCTCTTCCTGTTACTTTCCGTCTCTCTATCCC Barcode_4033 TGCTATCCCTGTACTTTTCCTTATCCTTTTTCCTTTGGG Barcode_4034 TTTTATCGCTATCAATCTCACTCTTTTTATCGCTTTCGG Barcode_4035 TGTTGTTATTCTATGTTTTTTTATCATTCTCCTTATCCT
Splinkerette PCR SPLK_A GAAGAGTAACCGTTGCTAGGAGAGACCGTGGCTGAATGAGACTGGTGTCG ACACTAGTGG SPLK_B_Spe/I Xba I CTAGCCACTAGTGTCGACACCAGTCTCTAATTTTTTTTTTCAAAAAAA SPLKFwd_1 GAAGAGTAACCGTTGCTAGGAGAGACC SPLKFwd_2 GTGGCTGAATGAGACTGGTGTCGAC Barcode 08-4030AS AATGATAGCAATAGACAAATATAAAGCAACAAGAACAATGA
123 Table 4.2. Oligonucleotides used in this study (continued) Oligonucleotide name Sequence Generation of KRP40 (re-constructed mutant 1a2-4032) SPAC24B11.08, 09_5'_del ACACAGGATCCGGCAACTTGTGCGATTTG SPAC24B11.08, 09_3'_del AGCTAGCTTTTATTTGAAGATTAGGATGGCGT
External confirmation primers of KRP40 SPAC24B11.08, AGTAGGTTAATAGACAGAGGTGCAAAGAGG 09_5'_del_confir SPAC24B11.08, TCCACACGAATATCGATGCTAAGTTGGT 09_3'_del_confir Generation of KRP42 (re-constructed mutant 1a7-4033) clg1 del 5' AGAACCTGCGCACAACCAACCACCGTAAAATT clg1 del 3' GAAGAGGGCGATGGTGGTGCGTTGGTGCT Clg1 5'+ InvU4-ASS GAACAGAATAAATTAGATGTCAAAAAGTTTCGTCAGTCTCGGAGG TAGTAGTGGCGGTAT TAIL-LB2 CTCCATTAAGTAACAAATTCCTATTTAGAGAAAGAATGCTGAGTA
External confirmation primers to KRP42 clg1 del 5' confirm TCTTGCTGAATAATCTTCAACGCTACGC clg1 del 3' confirm CCAATAGATAACAAAGCCACACGAGAAAGCA
Construction of KRP87 (clg1+ complete ORF deletion mutant) Clg1 5'_del_CspC I CAAATCGAGTGGCGTTTCACTCCAAATTGTCTCGAA Clg1_del_U4_5'R AACAGAATAAATTAGATGTCTGCTGAAGAATTTCGAAAAGGAAAGCAAAT Clg1 3'_del_CspC I CAAATCGAGTGGAGATAGATGCAAACGCTTCGCCGCT Clg1_del_U4_3'F ATGCATACATATAGCCAGTGTGATCGTGCTTACTTTCTGCGGGTGTTGTT
External confirmation primers to KRP87 Clg1 5'_confirm-2 CTGCCAATACACTCTTCACTCAGTGCT Clg1 3'_comform-3 ACAACTGATCAGGATCTAGGTAAGG
Construction of KRP131 (pef1+ complete ORF deletion mutant) Pef1 5'_del_CspC I CAAATCGAGTGGAGACTAGCGTATAAAGGTTAGTCT Pef1_del_U4_5' ATGCATACATATAGCCAGTGGTTGTGCGATGGAAATATCAAGTGCCT Pef1 3'_del_CspC I CAAATCGAGTGGTTAGTCACCTCCTACCATTGATGCC Pef1_del_U4_3' AACAGAATAAATTAGATGTCGCTGTAAATTAACATCAAAAAGTTTAACCA
External confirmation primers to KRP131 Pef1 5'_confirm-2 ATGCTTAATTGCTACATGCTTCGGAT Pef1 3'_confirm-2 TAACTTACGGTCTCCGTTAGAGGATC
Construction of KRP103 (pas1+ complete ORF deletion mutant) Pas1 5'_del_EcoR I ACTGCAATATCACGCCTCTTTGAATTC Pas1 5'_del_U4 ATGCATACATATAGCCAGTGCGTGTAGAGAAATATCTAAGATGAAA Pas1 3'_del_Nde I CCATTCATAGAACAGCATATGTGGGAT Pas1 3'_del_U4 AACAGAATAAATTAGATGTCTTATCAAGCCCAATCTGCGCTGTC
124 Table 4.2. Oligonucleotides used in this study (continued) Oligonucleotide name Oligonucleotide name External confirmation primers to KRP103 Pas1 5'_confirm CCATTCATAGAACAGCATATGTGGGAT Pas1 3'_confirm GCTCGAAGTAAGCTCGGACCATCCTA
Construction of KRP102 (SPBC20F10.10 complete ORF deletion mutant) SPBC20F10.10 5'_del_CspCI CAAATCGAGTGGCTTACTATTCAAGTTATCTCTTCCAGC SPBC20F10.10 5'_del_U4 ATGCATACATATAGCCAGTGAGGACTATTATAAAAATGTTTGGTAAATGG SPBC20F10.10 3'_del_CspCI CAAATCGAGTGGTAAGCTATCCACGTCTTCCTGCTT SPBC20F10.10 3'_del_U4 AACAGAATAAATTAGATGTCATTTACTACTGTGTTTGGTCCTCC External confirmation primers to KRP102 SPBC20F10.10 5'_confirm AGCTCATACTGTGTGCACTCCATAT SPBC20F10.10 3'_confirm AGAAATAGCAGCTTCTGCTAGTAG
Construction of KRP92 (ppk18+ complete ORF deletion mutant) Ppk18 5'_del_CspCI CAAATCGAGTGGGTGAGCATATTGAATGATGGGGTGTCT Ppk18 5'_del_U4 ATGCATACATATAGCCAGTGTAACTTTATAACTCCATTCAGATTTTG Ppk18 3'_del_CspCI CAAATCGAGTGGATTCGTTCCTCTGTTCAGAAACTA Ppk18 3'_del_U4 AACAGAATAAATTAGATGTCGTTCAAACGTTTGTTAATCGTGAT
External confirmation primers to KRP92 Ppk18 5'_confirm GTGGTAGTGTACACGCTATACACTC Ppk18 3'_confirm CCATTTGCCTGTCCGGCTGGAAAA
Primers for two-hybrid assay plasmid construction and sequencing GBD_pef1_5' GGCGGATCCGTATGAACTACCAAAGGCTTGAAAAGTTAGGAGAGGGAACA TATGCGCATG Pef1_2_exon GAGAGGGAACATATGCGCATGTTTATAAGGG Pef1_ORF_3' GGCGTCGACTATGCGGTTAAAAACCAAGCATGTTGA GAD_clg1_5' GGCGTCGACCTATGTCGTTTCCTTATCAGCACACTTCACGT Clg1_ORF_3' GGCAGACTCTAAGTCATTGCATAGCGATTGTACACT GDBD_Seq TCATCGGAAGAGAGTAGT GAD_Seq TACCACTACAATGGATGA
Construction of pREP41-FLAG vector pREP1_FLAG_S TATGGACTACAAAGACGATGACGACAAGGGAGGTGCGGCCGCTGTCGACG CTAGCG pREP1_FLAG_AS GATCCGCTAGCGTCGACAGCGGCCGCACCTCCCTTGTCGTCATCGTCTTT GTAGTCCA
Amplification of ORFs for pREP41-FLAG expression plasmid construction SPBC20F10.10_5'_Not I GCCGCGGCCGCTATGTCCTTGGCGTTTACGCTT SPBC20F10.10_3'_BamH I GCCGGATCCGGAGGACCAAACACAGTAGTAAAT Clg1_5'_Not I GCCGCGGCCGCTATGTCGTTTCCTTATCAGCACACT Clg1_3'_BamH I GCCGGATCCTCACTAAGTCATTGCATAGCGATT Clg1N_3'_BamH I GCCGGATCCCTAATCAGTCTCGGAGGTAGTAGTGGC Cek1_5'_Not I GCCGCGGCCGCTATGAAGCATATAAAAAACGAACGCG
125 Table 4.2. Oligonucleotides used in this study (continued) Oligonucleotide name Oligonucleotide name Cek1_3'_BamH I GCCGGATCCTCAAGAGTGCCAAACATCTAACTT Ppk18_5'_Not I GCCGCGGCCGCTATGGTAATGCAAGAACGCAATTCC Ppk18_3'_Nhe I GCCGCTAGCTCATTTATTGCAAAGCCGGGCTAT Ppk31_5'_Not I GCCGCGGCCGCTATGACTAATCCAGAGCAATTGAAG Ppk31_3'_Nhe I GCCGCTAGCCATGTTATGTCTATGGTTTTACATGATG pREP_FLAG_5'_Seq ACCGGATAATGGACCTGTTAATCG † N = A, T, C or G; W = A or T; S = G or C
Bacterial clones with large barcode inserts were first screened by extracting the total bacterial DNA from cells with phenol/chloroform/isoamyl alcohol
(25:24:1) and 1X DNA loading dye, and examining the aqueous phase, which contained bacterial genomic DNA and barcode-containing plasmids, by agarose gel electrophoresis to compare the electrophoretic mobility of the plasmids containing barcode inserts with the control plasmid without insert (pInsertion- ura4) on a 0.7% agarose gel. Plasmid DNA was prepared from cells with large inserts and verified by digestion with BamH I. Plasmids with long barcode inserts were sequenced with primer TAIL-LB LOX71 to determine the barcode sequences.
Identification of insertion sites by thermal asymmetric interlaced (TAIL)-PCR and splinkerette PCR
TAIL-PCR, which uses combinations of a set of arbitrary degenerate and three nested insertion DNA-specific primers to amplify a small region of insertion
DNA and adjacent genomic sequence with alternate high and low annealing temperatures in three rounds of PCR, was used to determined the insertion sites in clg1- (1a7-4033) and spncrna.142- (1a2-4032) mutants. The PCR reactions were
126 performed as described in Chapter 3 as well as Singer and Burke [324] (reviewed in Figure 3.2), except the insertion DNA-specific primers used in this study are
TAIL-LB LOX71, TAIL-LB2 and hsplam3. The products from the secondary or
the tertiary PCR reactions were sequenced with primer hsplam5 and hsplam7,
respectively.
Splinkerette PCR [56] was used to identify the insertion site in the 28S
rRNA mutant (1a2-4030). A double-stranded splinkerette adaptor with a
hairpin-forming sequence and a Spe I/Xba I overhang (made by annealing oligos
SPLK_A and SPLK_B_SpeI/XbaI) was ligated to 1a2-4040 genomic DNA
digested with Spe I and Xba I, and the ligation product was used as the template in the first PCR reaction with primers SPLKFwd_1 and Barcode 08-4030AS. The product from the first PCR reaction was diluted 50-fold, and 1 μl of the diluted
DNA was used as the template in the second PCR with primers SPLKFwd_2 and
Barcode 08-4030AS. The product generated in the second PCR reaction was
sequenced by the SPLKFwd_2 primer.
TAIL- and splinkerette PCR products were treated with shrimp alkaline
phosphatase (SAP) and exonuclease I (ExoI) (Exo-SAP, USB) to remove free
dNTPs and primers prior to sequencing.
Construction of fission yeast mutant strains for lifespan analysis
In clg1- mutant 1a7-4033, the tandem copies of ura4+ insertion vector
impeded direct PCR amplification of the mutated allele, probably due to
recombination of the repeated ura4+ sequences in denaturing and annealing steps.
127 To avoid directly amplifying the template with ura4+ repeats, two mega primers
were made first using primers TAIL-LB2 and clg1_del_3’ ( product size 1.3 kb)
as well as primers Clg1 5’+InvU4-ASS and clg1_del_5’ (product size 1 kb) with
KRP1 genomic DNA as the template. These two mega primers partially overlap
with the untranslated regions (UTRs) of the ura4+ gene (the overlapping
sequences in 5’ and 3’ UTRs of ura4+ are 99 and 44 bp, respectively) and were
used to amplify the ura4+ gene to create a deletion vector with flanking 5’ and 3’
ORF sequences of clg1+. The insertion position and 4-bp deletion of clg1+ ORF
were kept the same as the original insertion mutation in the mutant 1a7-4033,
except the duplicated insertion vector DNA was replaced as a single copy of
ura4+ gene. The PCR product generated by the mega primers (~ 4 kb) was further amplified with clg1_del_5’ and clg1_del_3’ primers and the amplified
DNA was used to transform the wild type strain KRP1.
To re-construct the spncrna.142- mutant 1a8-4032, the disrupted allele in
this mutant was amplified by PCR with primers SPAC24B11.08, 09_5'_del and
SPAC24B11.08, 09 _3'_del and used to transform KRP1 cells.
To construct the complete ORF deletion mutants of clg1+, pef1+, pas1+,
psl1+/SPBC20F10.10 and ppk18+, two mega primers consisting of the 5’ and 3’
UTRs of the target gene (~0.7 to 1 kb immediately proceeding ATG or following
the stop codon) and ura4+-overlapping sequences were first made by PCR using
the genomic DNA of KRP1 as the template. The resulting mega primers were
used in the subsequent PCR reaction with the 1.7 kb ura4+ gene (corresponding to
the Hind III genomic fragment) as the template to assemble the final targeting
128 vector composed of 5’ and 3’ UTRs of the target gene and the ura4+ marker.
The PCR product was further amplified by the 5’ and 3’ gene-specific primers,
followed by cloning the amplified DNA to pCR2.1-TOPO vector (Invitrogen),
except for pef1::ura4+, which was cloned to pBluescript II SK. Restriction
enzyme CspC I recognition sites were engineered in the 5’ and 3’ gene-specific primers so the deletion vector is bordered by these sites (EocR I and Nde I for
pas1). These vectors were fully sequenced to verify the absence of mutations
and digested by CspC I (or EcoR I and Nde I for pCR2.1-pas1::ura4+) to release
the targeting DNA from the vector. Purified linear targeting vector DNA was used to transform the wild type strain KRP1 or KRP83.
The transfected cells were plated on EMMG with proper supplements to the auxotrophic mutations and grown at 30oC for 5-7 days. Correct deletions were
confirmed by PCR using primers external to the sequences for targeting the
integration and the internal primers for the ura4+ marker.
The double deletion mutant strain clg1 pef1 (KRP139) was made by transforming clg1:: leu1+ targeting vector DNA (constructed as described above
except replacing ura4+ with leu1+) to the pef1::ura4+ strain (KRP95) and
selecting for Leu+ Ura+ transformants. To generate the clg1 cek1 (KRP171) and clg1 ppk18 (KRP109) double deletion mutants, the clg1 single deletion strain (KRP88) was mated with the cek1 (KRP70) and ppk (KRP92) single
deletion strain, respectively, and sporulated on ME plates for tetrad dissection.
Spores from complete tetrads with four spores were examined for the presence of
both deleted alleles by monitoring the inheritance of the ura4+ and KanMX4
129 markers and confirmation PCR with gene-specific primers. Two backcrosses
were performed to obtain the clg1 cek1 double deletion mutant (KRP171) to reduce potential difference in genetic backgrounds between our lab strain and the
strain from Bioneer, Inc. Two cek1spores were isolated in the second
backcross for the same reason.
Expression plasmids construction
To construct plasmids for the two-hybrid assay, the complete clg1+ coding sequence or a partial clg1+ fragment corresponding to the first 590 nucleotides of
clg1+ ORF was made by PCR, cut with Sal I and cloned to the Sal I and filled-in
Bgl II sites on pGAD424. PCR amplified pef1+ was cut with BamH I and Sal I and cloned to the same sites on pGBT9. The cloned ORFs were verified by sequencing using GAD_seq and GDBD_seq primers.
To express FLAG-tagged Clg1p, Cek1p or Ppk18p in fission yeast, the vector pREP41-FLAG was first generated by ligating a double-stranded DNA insert that has a FLAG epitope-coding sequence and a peptide linker (2 glycines and 3 alanines) [298] (made by annealing oligonucleotides pREP1_ FLAG_S and
pREP1_FLAG_AS) to the Nde I and BamH I sites on pREP41 vector [13]. The
ORF of clg1+ or cek1+ was generated by PCR, digested with Not I and BamH I
and ligated to the same sites on pREP41-FLAG. The ORF of ppk18+, also
generated by PCR, was cloned at Not I and Nhe I sites of pREP41-FLAG. All expression constructs were verified by sequencing with pREP_FLAG_5’_seq primer and gene specific primers.
130 Yeast two hybrid assay
The pGBT9- and pGAD424-derived plasmids were transformed into the
budding yeast strain Y187 and transformants were selected on complete medium
plates without leucine and tryptophan at 30oC for 3 days. To detect the
expression of the reporter gene lacZ, 5 colonies of each transformation were
patched on the same medium plates and grown at 30oC for 2 days. Cell patches
were lifted on Whatman CHR paper and submerged into liquid nitrogen for 30
seconds. CHR paper with frozen cell patches were thawed at room temperature
for 3 to 5 miuntes and overlaid on another CHR paper soaked in 3 ml of Z
buffer/X-gal/-mercaptoethanol (60 mM Na2HPO4, 40 mM NaH2PO4, 10 mM
KCl, 1 mM MgSO4, 1 mg/ml X-gal, pH 7.0, 39 mM -mercaptoethanol). The
reaction was kept in the dark and allowed to proceed for at least one hour at room
temperature.
Protein extraction, immunoprecipitation and Western blotting
Protein induction and extraction were performed as described [13, 237].
Briefly, a single colony of pREP41-FLAG transformed cells was inoculated in 2
ml of EMM + adenine, leucine (225 mg/l), 2% glucose, 5 g/ml of thiamine and
grown on a roller drum for 2 days at 30oC. Cells were washed with 1 ml of
sterile milliQ H2O before diluted in 50 ml of the same medium without thiamine
at 5 105 cells/ml and grown in an air platform shaker rotating at 220 rpm at
30oC for 24 hours to allow protein expression. Cells were then spun down and washed with 5 ml of pre-chilled stop buffer (150 mM NaCl, 50 mM NaF, 10 mM
131 EDTA, 1 mM NaN3) and 1 ml of pre-chilled HB buffer (25 mM MOPS pH7.2, 15
mM MgCl2, 15 mM EGTA, 1 mM DTT, 1% Triton-X100, 60 mM -
glycerophosphate, 15 mM p-nitrophenylphosphate, 0.1 mM sodium vanadate, 1
mM PMSF, 1X protease inhibitor cocktail (Roche)). Cells were lysed in 2-ml
polypropylene screw cap tubes with 100 l of HB buffer by zirconia/silica beads
(added to ~ 2 mm below the meniscus) beating for 30 seconds for 4 times on
Mini-Beadbeater-16 (Biospec). The samples were placed on ice for 1 minute
after each beating to reduce heat generated during the process. The beads were
then washed with 500 l of pre-chilled HB buffer, spun at 13,200 rpm at 4oC for 5
minutes, and the supernatant was transferred to new tubes, followed by incubation
on ice for 20 minutes. After a final centrifugation at 13,200 rpm at 4oC for 15
minutes, the soluble fraction was transferred to new tubes and protein
concentration was determined by Bradford method (BioRad).
For immunoprecipitation (IP), 4 to 5 mg of protein lysate in HB buffer was
used along with 3 g of mouse monoclonal anti-HA (F-7, Santa Cruz) or anti-
FLAG (M2, Sigma) antibody-conjugated protein G sepharose (GE Healthcare) in
a final reaction volume of 500 l. NaCl was added to each sample to a final concentration of 150 mM. The immunoprecipitation was performed on a roller at 4oC for 4 hours. IP samples were washed with 1 ml of pre-chilled HB buffer
containing 150 mM NaCl for 3 times prior to SDS-PAGE analysis.
To reduce degradation of overexpressed FLAG-Cek1p and FLAG-Ppk18p,
these immunoprecipitates were denatured by incubating the washed IP samples in
1X SDS sample buffer with -mercaptoethanol at room temperature for 20
132 minutes before gel loading.
Western blotting/protein transfer was performed using Mini Trans-Blot
Electrophoretic Transfer Cell (BioRad) as described by the manufacturer.
Primary antibodies used were rabbit polyclonal anti-HA (Y-11, Santa Cruz) and rabbit polyclonal anti-FLAG (F7425, Sigma). Donkey anti-rabbit HRP- conjugated secondary antibody (Santa Cruz) was used in all experiments.
Analysis of CLS assay data
Comparison of different chronological lifespan (CLS) was performed as described in [39] and Chapter 2. Statistical comparisons were performed using
the Wilcoxon signed rank test in Prism 4 (Graphpad Software).
133 Results
An unbiased parallel selection for long-lived barcode-tagged insertion mutants using a novel sequencing strategy
To perform an unbiased and genome-wide genetic selection for long-lived mutants, one would need to screen as many mutations as possible and assay the long-lived phenotype directly. In principle, mutants with longer-than-wild type lifespans are easily distinguishable from cells with normal lifespans if they are monitored in separated cultures. Some mutants may have both longer median and maximum lifespans while some have “rectangularized” lifespans in which they have high viability for a longer period of time (longer median lifespan) but all cells die at the same time as wild type cells (the same maximum lifespan,
Figure 4.1A).
In a parallel screen which has hundreds of thousands of random mutants in the initial culture, long-lived mutants may only constitute a small portion in the total mutant population (Figure 4.1B). One way to select for mutants with long lifespans is to age cells until all cells with normal lifespan die and the remaining surviving cells would be the long-lived mutants. However, this method will only identify mutants with longer median and maximum lifespans, and mutants that only have longer median lifespan will be missed. To ensure the recovery of both types of long-lived mutants, the sampling time point of such screen has to be early enough so that viable cells of both kinds of mutants are available. Although the fraction of long-lived mutants may increase, cells with normal lifespans will still constitute the majority of the surviving cells at this point (Figure 4.1B, the gray
134 bar). Thus, an efficient method that allows quick isolation of these long-lived mutants from the surviving cells is required. Tagging each mutant with a unique barcode, even in the absence of the prior knowledge of the barcode sequences, provides such a method.
Figure 4.1. Hypothetical survival curves of long-lived mutants with different median and maximum lifespans. The gray bar labeled “Sampling time” indicates one of the time points in a CLS assay that viability is determined and viable cells are collected. (A) Long-lived mutants with different lifespans can easily be distinguished from cells with normal lifespans (the black curve) when they are monitored in separated cultures. Some long-lived mutants have both longer median and maximum lifespans (the green curve), while others with a “rectangularized” lifespan curve only have extended median lifespans (the red curve). (B) In a CLS assay of pooled random mutants, the initial proportion of desired long-lived mutants can be very small (e.g. 1/106 of total population). Aging the culture near the end of the lifespan, but not until the time all cells with normal lifespan die, can amplify the fraction of different kinds of long-lived mutants (e.g. from 1 1/106 to ~1% of total surviving population). Additional approaches are required to retrieve these long-lived mutants from a high background of viable cells with normal lifespans.
To test the feasibility of unbiased parallel selection for lifespan-extending mutations from a pool of random mutants, ~3,600 barcoded insertion mutants were seeded in 240 ml of SD medium supplemented with 3% glucose, 150 mg/l
135 of adenine, leucine, histidine, uracil in 1000-ml flasks and aged for 15 days.
Colonies of 600 surviving cells on day 14, which had ~800 colony forming units per ml of the culture, were patched on YES plates to generate master plates for genomic DNA preparation and subsequent mutant isolation. As mentioned in
Chapter 3, the barcodes in this mutant library are flanked by two restriction enzyme Sfi I recognition sites which allow efficient barcode oligomerization and sequencing (Figure 4.2).
Figure 4.2. Using the S. pombe barcode-tagged insertion mutant library and a novel barcode sequencing strategy to identify lifespan-extending mutations. About 3,600 barcode-tagged insert mutants were pooled together and aged for 15 days in a single culture. Genomic DNA was made from surviving cells on day 14 to amplify barcodes with flanking Sfi I recognition sites. After Sfi I digestion, barcodes with Sfi I overhangs were oligomerized by standard ligation reaction to produce long barcode oligomers. Sequencing these long barcode oligomers provided information regarding the variety and frequency of barcodes and corresponding mutants in the selected culture on day 14.
136 To generate barcodes with Sfi I enzyme overhangs, primer hsplam6 and
BarcodePCR(888r) were used in PCR reactions with genomic DNA of day 14
surviving cells as the template to amplify partial insertion vector fragments that
have ~200 bp and ~500 bp extension on the 5’ and 3’ ends, respectively, of the 66
bp Sfi I flanked-barcodes. The ~750 bp PCR product was digested with Sfi I
enzyme, which cut the full length fragment into three small fragments of 66, ~200
and ~500 bp. The size difference allowed the barcodes to be separated from the
flanking fragments and the purified barcode monomers were ligated together to
make long barcode oligomers for sequencing.
A total of 405 barcode sequences were determined to represent the barcode
frequencies of the selected culture (Table 4.3). There were 6 kinds of barcodes
isolated at frequencies of more than 10 times, ranging from 12 to 95 (the bottom 6
rows of Table 4.3). On the other hand, a total of 32 types of barcodes were only identified 4 or fewer times (the top 4 rows of Table 4.3). These results are consistent with the interpretation that the proportion of the mutants carrying these
6 barcodes gradually increased due to their increased viability in quiescence compared to cells with normal lifespans. To rule out the possibility that enrichment of these barcodes resulted from a biased barcode representation at the beginning of the experiment, barcodes of the initial mutant pool were also sampled and sequenced in the same way. As shown in Table 4.4, there were no specific barcodes that were dominant in the initial mutant pool. Moreover, the barcode compositions in the initial and the selected cultures were significantly changed as no barcodes were identified in both cultures. These data indicate that
137 the increased frequencies of the 6 barcodes in Table 4.3 must occur as a result of the selection on these mutants.
Table 4.3. Barcode sequencing of the surviving mutants after selection for 14 days in stationary phase and insertion mutations identified in the overrepresented mutants. Frequencies of Types of % of total Inserted and affected gene(s) barcode (A) barcode (B) (A B / 405) 1 22 5.4 ND 2 5 2.5 ND 3 3 2.2 ND 4 2 2.0 ND † * 12 1 (4031) 3.0 590 bp of clg1+/mug80+ (cyclin homolog) ORF † * 19 1(4033) 4.7 590 bp of clg1+/mug80+ (cyclin homolog) ORF
69 1(4032) * 17.0 905 bp of SPNCRNA.142
‡ * # 73 1(4035) 18.0 SPRRNA.47, 48, 49
‡ * # 88 1(4034) 21.7 SPRRNA.47, 48, 49
‡ * # 95 1(4030) 23.4 SPRRNA.47, 48, 49 The total number of determined barcode sequences is 405. ND = not determined † Two barcodes in the clg1-/mug80- mutant. ‡ Three barcodes in the sprrna.22- mutant. * Representative barcode number # Insertion was located in the 28S ribosomal RNA coding gene that exists as three identical copies on both ends of chromosome III and detailed insertion structure was not determined.
Table 4.4. Barcode sequencing of pool#1 and #2 mutants in the initial culture (day 0) for long-lived mutant selection.
Frequencies of barcode (A) Types of barcode (B) % of total (A B / 143) 1 105 73.5 2 16 22.3 3 2 4.2 The total number of sequenced barcodes is 143.
138 To isolate the mutants bearing these enriched barcodes, these 6 barcodes were first used as the specific primers in conjunction with a universal primer on the insertion vector to identify the corresponding mutants by PCR from the master plates containing the day 14 surviving cells. Two barcodes (4031 and 4033) always identified the same cell patches on the master plates, suggesting that these
2 barcodes co-existed in the same mutant. Similarly, barcodes 4030, 4034, and
4035 were also found in the same patches independent of those carrying barcodes
4031 and 4033. Therefore, the 6 classes of enriched barcodes identified 3 mutants that increased in proportion after the mutant pool was aged for 14 days in the SD medium (Table 4.3).
These 6 barcodes were also used to determine the insertion sites. TAIL-
PCR was first used and successfully located mutations in two strains. The strain
(1a7-4033) bearing barcodes 4031 and 4033 had an insertion at the 590th bp of the
clg1+/mug80+ ORF (full length = 1386 bp, Figure 4.3A). Consistent with the
presence of two barcodes in this mutant, two tandem copies of the insertion vector
were found at the insertion site with a copy of the insertion vector having a large
truncation. This insertion also resulted in 4 bp deletion of the clg1+ ORF.
An insertion at the 905th bp of SPNCRNA.142 (full length = 978 bp) was
identified by TAIL-PCR in the mutant (1a8-4032) carrying the barcode 4032.
The two protein coding genes closest to the insertion are erv46+ and
SPAC24B11.09, with their start codons located 1540 bp and 1736 bp from the
insertion site, respectively (Figure 4.3B). There was no deletion found in the
chromosomal sequence surrounding the insertion vector.
139 Figure 4.3. Insertion sites determined in two of the isolated mutants from the day 14 culture. (A) In the mutant tagged with barcode 4031 and 4033, the insertion was found in protein coding gene clg1+ (also known as mug80+). The insertion resulted in a 4 bp deletion of the coding sequence. Tandem integration of two insertion vectors was detected (one full-length copy and one truncated copy, see Table 3.4). (B) In the mutant tagged with barcode 4032, insertion site was located in 3’ end of the non-coding RNA SPNCRNA.142. The closest protein-coding genes (erv46+ and SPAC24B11.09) are located at least 1.5 kb from each end of the insertion site.
For the mutant bearing barcodes 4030, 4034, and 4035, TAIL-PCR only
identified ura4+ sequences, an indication of co-integration of multiple copies of the insertion vector (Figure 4.4). An alternative approach, splinkerette PCR
[56], was then used to locate the insertion. After ligating the Spe I- and Xba I- digested genomic DNA to the splinkerette adaptor with CTAG overhang, a PCR reaction using the ligation product as the template and specific primers on the insertion vector and the splinkerette adaptor was able to generate a ~1.2 kb product (Figure 4.4). Sequencing this PCR product determined that one end of the insertion is within the 28S ribosomal RNA (rRNA) coding gene. In S.
pombe, the 5.8S, 18S and 28S rRNA genes co-exist as repeated arrays on the ends
of chromosome III [123]. Therefore, the sequencing result could not determine
the exact location of the insertion.
140
Figure 4.4. Characterization of the 28S ribosomal RNA (rRNA) insertion mutant by splinkerette PCR. Three copies of insertion vectors with barcodes 4030, 4034 and 4035 were identified in this mutant. To determine the insertion location by splinkerette PCR, genomic DNA of this mutant was digested with restriction enzymes Spe I and Xba I and ligated to the splinkerette adaptor with the compatible overhang (CTAG). The ligation product was used as the template in the first PCR with primers specific to the splinkerette adaptor (purple arrow) and 4030 insertion vector (green arrow). The product from the first PCR reaction was diluted and used as the template in the second PCR with the same 4030 insertion vector-specific primer (green arrow) and a nested primer on the splinkerette adaptor (also shown as a purple arrow). Sequencing of the product from the second PCR identified the sequence encoding 28S rRNA.
CLS of clg1-, spncrna.142- and 28S rRNA mutants
To verify whether the enrichment of the above three mutants indeed resulted
from improved survival in stationary phase, the lifespan of these mutants were re-
determined. Considering that secondary mutations might occur during long-term
selection in stationary phase and change lifespans of the initial mutants, the
insertion mutations discovered in the clg1- (1a7-4033) and spncrna.142- (1a8-
4032) mutant were re-constructed in the parental wild type strain KRP1. The
lifespans of these re-constructed mutants and the original mutant isolates were
both re-examined at the same time. If secondary mutations arising during the
141 selection were the cause of lifespan extension in these mutants, the long-lived
phenotype will only be repeated in the original mutant isolates but not in the new
re-created mutant strains. Because the insertion in the complex and repetitive
structure of rRNA coding gene arrays prevented recreating the exact insertion
structure, only the original 28S rRNA mutant isolate (1a2-4030) was re-tested.
Among the five mutant strains examined, the original (KRP44) and re-
constructed (KRP42) clg1- insertion mutants both had a lifespan longer than that
of the wild type control, indicating that the increased abundance of this mutation
during the selection was due to the lifespan-extending effect of the insertion
mutation in the clg1+ ORF (Figure 4.5A).
Contrary to the clg1- mutants, both original (KRP45) and re-constructed
(KRP40) spncrna.142- insertion mutant showed no lifespan extension compared
to the wild type control, indicating that the enrichment of this mutant in the
selection appears to be unrelated to increased lifespan (Figure 4.5B). It was
unclear why this mutant was enriched, and the isolation of this mutant may reflect
improved survival in a culture of many different mutants that is independent of
longevity regulation.
The original 28S rRNA mutant KRP43 still retained the lifespan-extending phenotype (Figure 4.5C). This mutant was not further analyzed in this study.
142 Figure 4.5. The chronological lifespan of the three mutants isolated from the parallel selection for long-lived mutants. (A) The clg1- insertion mutation was recreated in the parental wild type strain KRP1 to make the strain KRP42. Both the new insertion mutant KRP42 and the original mutant isolate from day 14 culture (KRP44) exhibited a longer-than-wild type lifespan although the original isolate KRP44 exhibited large variations between the two duplicate cultures. (B) The spncrna.142- insertion mutation was re-constructed in the same parental wild type strain to make KRP40. Both the original (KRP45) and the re-constructed (KRP40) spncrna.142- mutants had lifespans similar to that of the wild type KRP34. (C) The original 28S ribosomal RNA mutant isolate retained better viability in all sampling time points and had a longer maximum lifespan. Note: in KRP40 and KRP42, the corresponding genes were disrupted by insertion vectors that contain not only ura4+ marker, but also a barcode and other DNA sequences as described in the text and Chapter 3.
143 Establishment of a working model for Clg1p-dependent lifespan regulation
The clg1+/mug80+ (meiotically up-regulated gene) gene was first identified
as one of the 552 genes co-induced in the middle of meiosis [220]. Despite its
increased expression in meiosis, its function in meiotic regulation or other
biological aspects is unknown. In the fission yeast genome database
(Schozosaccharomyces pombe GeneDB [308]), the budding yeast S. cerevisiae cyclin-coding gene CLG1 is listed as an ortholog to the fission yeast clg1+/
mug80+. The budding yeast Clg1p is a member of a cyclin family that associates
with the cyclin-dependent kinase (CDK) Pho85p. A BLASTP search of S.
pombe proteins using Pho85p primary sequence as the query identified Pef1p as a
fission yeast Pho85p ortholog.
The budding yeast CDK Pho85p associates with a total of 10 Pcl cyclins
(Pho85p associating cyclins), including Clg1p and the founding cyclin member
Pho80p [38]. The S. cerevisiae Pho80p/Pho85p cyclin/CDK complex has been
implicated as a negative regulator in the transition between proliferation and
quiescence through a protein kinase Rim15p [369]. The identification of the
fission yeast orthologous cyclin Clg1p and CDK Pef1p drove the hypothesis that
Clg1p/Pef1p has similar activity in restraining entry into quiescence, which is
important for maintaining chronological lifespan. This hypothesis is consistent
with the lifespan-extending phenotype of the clg1- insertion and predicts a working model in which Clg1p associates with Pef1p to regulate S. pombe
chronological aging through a Rim15p ortholog (Figure 4.6).
144 Figure 4.6. A working model for lifespan regulation by the Clg1p/Pef1p cyclin/CDK complex and the S. pombe Rim15p ortholog. Based on studies of S. cerevisiae Pho80p/Pho85p and the identification of S. pombe CDK ortholog Pef1p, Clg1p and Pef1p were predicted to form a complex that phosphorylates and promotes nuclear exclusion of a S. pombe Rim15p ortholog in conditions that favor proliferation (left). The lifespan extension phenotype of clg1- insertion together with work in S. cerevisiae suggested that inactivation of the Pef1p/Clg1p complex by depletion of Clg1p and/or Pef1p prevents phosphorylation of the S. pombe Rim15p ortholog and extends lifespan (middle). This lifespan extension phenotype depends on the S. pombe Rim15p ortholog, and removal of this protein reverses lifespan extension induced by clg1 (right). To test this working model, four predictions were made and examined in this chapter: (1) Pef1p physically interacts with Clg1p. (2) Deletion of pef1+ extends lifespan and the lifespan of clg1 pef1 double deletion mutant should be the same as either single deletion mutant. (3) The Clg1p/Pef1p complex should physically interact with the S. pombe Rim15p ortholog. (4) Deletion of this S. pombe Rim15p ortholog should reverse the lifespan extension by deletion of clg1+.
Clg1p physically interacts with the CDK Pef1p
The prerequisite of this prediction is a physical interaction between Clg1p and Pef1p. A filter-based yeast two hybrid assay was first used to test this prediction (Figure 4.7A). A positive interaction in this assay was observed when
Gal4 activation domain (AD)-Pef1p and Gal4 DNA binding domain (DBD)-
Clg1p were co-expressed (Figure 4.7B). To test whether the Clg1p N-terminal
145 truncated fragment encoded by the clg1- insertion mutant (from the first 590 bps of clg1+ coding sequence) retains the ability to interact with Pef1p, this truncated
N-terminal protein (Clg1(N)p) was fused with Gal4 DBD and co-expressed with
Gal4 AD-Pef1p. No interaction with Pef1p was detected by this method (Figure
4.7B, bottom).
Figure 4.7. The two hybrid assay to detect physical interaction between Gal4 DNA binding domain (DBD)-Pef1p and Gal4 activation domain (AD)-Clg1p. (A) If Pef1p interacts with Clg1p, this interaction will bring Gal4 AD to the lacZ gene promoter and drive the expression of -galactosidase, which can be detected by the formation of blue color in permeablized cells on filters soaked in X-gal. (B) Cells expressing both Gal4 DBD-Pef1p and Gal4 AD-Clg1p showed positive interaction (2nd row from the bottom). Co-expressing Gal4 DBD-Pef1p with a Gal4 AD-truncated Clg1p N-terminal fragment (Clg1(N)p; encoded by the first 590 bps of clg1+ coding sequence) resulted in no interaction. Controls expressing neither or only one of these fusion proteins all did not drive the expression of the reporter protein -galactosidase.
To test whether Clg1p and Pef1p also interact in fission yeast cells, FLAG- tagged Clg1p was expressed in cells which already expressed Pef1p-3HA from the pef1+ locus. Immunoprecipitates (IPs) bound to anti-FLAG antibody coupled to protein G sepharose were analyzed by SDS-PAGE and Western
146 blotting with anti-HA antibody. Pef1p-3HA was only detected in the IP from
cells co-expressing FLAG-Clg1p but not in control cells expressing endogenous
untagged Clg1p (Figure 4.8A). In a reciprocal IP, FLAG-Clg1p was expressed
in cells expressing 3HA-tagged or untagged Pef1p and precipitated by anti-HA
antibody-coupled protein G sepharose. FLAG-Clg1p was only found in the
sample that also contained Pef1p-3HA (Figure 4.8B). The two hybrid and IP
experiments show that Clg1p physically associates with Pef1p.
Figure 4.8. Co-immunoprecipitation (IP) identifies a physical interaction between Pef1p and Clg1p in fission yeast. (A) The control plasmid pREP41-FLAG or pREP41-FLAG- clg1+ was transformed into K566-11 which expressed Pef1p-3HA from the endogenous pef1+ locus. Transformants were grown in EMM medium to induce protein expression. Cell lysate made from the EMM cultures were incubated with anti-FLAG (M2) antibody- bound protein G sepharose. IP samples and 100 g of lysate were analyzed by Western blotting (WB) using anti-HA antibody (Y-11) or anti-FLAG antibody (F7425). Pef1p- 3HA was only detected in the IP sample containing FLAG-Clg1p. (B) The pREP41- FLAG-clg1+ plasmid was transfected in KRP34 (untagged Pef1p) or K566-11 (Pef1p- 3HA). Cell lysate of these two types of transformants were incubated with anti-HA (F7) antibody-coupled protein G sepharose. IP samples and 100 g of lysate were analyzed as in (A). FLAG-Clg1p was only recovered in the IP sample containing Pef1p-3HA. The signal of Pef1p-3HA in the lysate was very weak due to the poor expression in this experiment but was enhanced in the IP sample.
Clg1p and the CDK Pef1p control CLS through the same pathway
Based on the result that the truncated Clg1p N-terminal fragment failed to interact with Pef1p in the two hybrid assay, one can hypothesize that the insertion
147 in clg1+ eliminated the CDK-activating function of Clg1p and resulted in loss of the function/activity of this cyclin/CDK complex and lifespan extension (Figure
4.6). To test whether inactivation of the Clg1p/Pef1p complex extends lifespan, the coding sequence for clg1+ or pef1+ was completely deleted and the lifespan of
these complete ORF deletion strains were determined. Similar to the clg1::ura4+ insertion mutant (Figure 4.5A), the clg1::ura4+ ORF deletion mutant also had a
longer lifespan compared to the wild type control (Figure 4.9A). This result is
consistent with the prediction that the original insertion mutation results in a loss-
of-function phenotype. Deletion of pef1+ also showed a longer-than-wild type
lifespan (Figure 4.9B) as predicted in the working model that inactivation of the
CDK Pef1p extends lifespan (Figure 4.6).
To examine whether Clg1p and Pef1p regulated lifespan through a common
pathway, the lifespan of the clg1 pef1 double mutant was measured. The
prediction was that if these proteins work independently, the lifespan of the
double deletion mutant would be distinct from both single deletion mutants. On
the other hand, if they function in the same pathway, the lifespan of the double
mutant should be the same as one of the single deletion mutants (Figure 4.6).
Consistent with the latter prediction, the clg1 pef1 double-deletion mutant also
had a longer lifespan than the wild type control and its lifespan curve overlapped
that of pef1 cells (p > 0.2, Figure 4.9D). These results indicate that Pef1p and
Clg1p regulate S. pombe chronological lifespan through the same genetic
pathway. Their physical interaction also suggests that Pef1p and Clg1p regulate
lifespan in fission yeast as a complex.
148 Figure 4.9. Deletion of clg1+ or pef1+ extends S. pombe chronological lifespan. (A, B) Both clg1 KRP138 and pef1 KRP131 single deletion mutants showed an extended lifespan compared to the wild type control (p = 0.0003 for clg1KRP138; p = 0.0045 for pef1KRP131). (C) When both genes were deleted, the double deletion mutant strain KRP139 still had a lifespan significantly longer than that of the wild type (p = 0.0029). (D) The lifespan curves of the double deletion mutant KRP139 and the single deletion mutant clg1 KRP138 and pef1 KRP131 significantly overlapped and were indistinguishable from each other before day 13 (p > 0.2 for all comparisons).
Identification of other Pcl-like cyclins in S. pombe
Because of the large family size of Pcl cyclins in S. cerevisiae, it appeared likely that there is more than just one Pcl-like cyclin in S. pombe. Two major cyclin domains are present in the 10 Pcl cyclins in the Pfam 24.0 database (as of
March 2010; database updated in October 2009 and contained 11912 families
[274]): cyclin (Pfam ID: PF08613) and cyclin_N (Pfam ID: PF00134). The search of S. pombe proteins containing cyclin_N (PF00134) in Pfam 24.0
149 database resulted in 11 sequences. All except one of these genes have known functions in mitotic and meiotic cell cycle progression, and regulation of RNA polymerase II complex. However, none of these 11 fission yeast sequences showed significant sequence similarity to the budding yeast Pho85p-associating cyclins. Thus, the cyclin_N domain is not conserved in S. pombe Pef1p- associating cyclins such as Clg1p.
When searching the Pfam 24.0 database for S. pombe protein containing the cyclin domain (Pfam#: PF08613), three proteins were identified, including Pas1p,
Clg1p and an uncharacterized protein Spbc20f10.10p. Unlike the 11 proteins containing the cyclin_N domain, the budding yeast Pcl cyclins homologous to these three putative cyclins were identified with Pcl5p most similar to Pas1p,
Clg1p (S. c.) to Clg1p (S. p.) and Pcl7p to Spbc20f10.10p (Table 4.5).
Table 4.5. Identification of S. cerevisiae orthologs of S. pombe Spbc20f10.10p by BLASTP searcha Query Protein Accession Max score Total score E-valueb coverage Pcl7p NP_012214.1 162 162 76% 4e-41 Pcl6p NP_010980.1 103 103 71% 2e-23 Pho80p NP_014642.1 74.3 74.3 57% 2e-14 a. Protein blast search in NCBI using the dataset “non-redundant protein sequences” including organism “S. cerevisiae S288c (taxid: 559292)” sequences only (as of June, 2010). b. Only proteins with an E-value smaller than 1e-10 are showed.
Pas1p had been previously shown to associate with Pef1p kinase to regulate
G1/S transition through activation of the Res2p-Cdc10p transcriptional regulator complex by Tanaka and Okayama [344]. We showed that Clg1p associates with
Pef1p in a two hybrid assay and co-IP experiments in this study. To test whether
Spbc20f10.10p associates with Pef1p, FLAG-tagged Spbc20f10.10p was
150 expressed in the strain expressing Pef1p-3HA. Monoclonal anti-FLAG-coupled sepharose G was used to precipitate FLAG-Spbc20f10.10p and Western blotting analysis of the protein precipitates with anti-HA antibody showed that Pef1p-3HA
was only detected in lysate that also had FLAG-Spbc20f10.10p. The IP sample
from control cells transformed with the pREP41-FLAG empty plasmid did not
contain Pef1p-3HA (Figure 4.10). This result shows that Spbc20f10.10p also
interacts with Pef1p kinase and there are at least three Pef1p-associating cyclins in
S. pombe. Based on these data, the SPBC20F10.10 gene has been given the
common name psl1+ for Pcl Seven Like cyclin.
Figure 4.10. A co-IP assay indicates that Pef1p interacts with the cyclin homolog Spbc20f10.10p. The control plasmid pREP41-FLAG or pREP41-FLAG-SPBC20F10.10 was transformed into K566-11 cells that co-expressed Pef1p-3HA. Cell lysate prepared from the transformed cells was incubated with anti-FLAG (M2) antibody-coupled protein G sepharose. Western blotting was used to detect the presence of Pef1p-3HA or FLAG- Spbc20f10.10p in the IP samples or crude lysate (100 g per lane). Pef1p was only detected in the IP sample that also contained FLAG-Spbc20f10.10p but not in the control. FLAG-Spbc20f10.10p signal was too weak to be detected in the lysate due to the poor expression level (red asterisk in the lane of Pef1p-3HA “+” and FLAG-Spbc20f10.10p “+”) but was detectable in the anti-FLAG antibody IP sample. Based on these data and the sequence similarity between Spbc20f10.10p and S. cerevisiae Pcl7p, Spbc20f10.10p is now called Psl1p for Pc1 Seven Like cyclin.
151 Deletion of psl1+ or pas1+ does not extend CLS
Some of the budding yeast Pcl cyclins have been shown to be functionally redundant [38]. The identification of additional Pef1p-associating cyclins in S. pombe raised the possibility that Pef1p may also regulate lifespan through its interaction with the other two cyclins. To test this hypothesis, psl1+ or pas1+
was deleted and the lifespan of individual single deletion strains was determined.
In contrast to clg1, deletion of psl1+ consistently shortened lifespan (Figure
4.11A) and the lifespan curve of pas1 strain overlapped with that of the wild type control (Figure 4.11B). Therefore, lifespan extension is a unique phenotype regulated by Clg1p and Pef1p but not other two Pef1p-associating cyclins.
Figure 4.11. Deletion of cyclin homologs psl1+ or pas1+ does not increase chronological lifespan. (A) Deletion of psl1+ resulted in a shortened lifespan (KRP102). (B) Deletion of pas1+ did not significantly affect chronological aging as the lifespan of pas1 KRP103 overlapped with that of the wild type in majority of the time in the aging assay.
Identification of potential S. pombe Rim15p orthologs
In S. cerevisiae, the protein kinase Rim15p mediates lifespan extension by
inactivation of TORC1, Sch9p and PKA [77, 371]. In conditions favoring
proliferation, these nutrient-sensing pathways promote Rim15p phosphorylation
152 directly or indirectly, which results in Rim15p cytoplasmic retention or
inactivation [270, 283, 369]. Rim15p is also a substrate of the Pho80p/Pho85p
cyclin/CDK complex, which controls Rim15p subcellular localization and
Rim15p-dependent gene expression by phosphorylation [369]. To test whether
the fission yeast Clg1p/Pef1p also regulates lifespan through a Rim15p-like
effector, potential fission yeast Rim15p orthologs were sought first by BLASTP
search of S. pombe proteins in the non-redundant protein sequences dataset of
BLAST/NCBI using S. cerevisiae Rim15p primary sequence as the query. Two
S. pombe serine/threonine kinases, Cek1p and Ppk18p, with a total score greater
than 500, a query coverage over 50% and an E-value smaller than 5 10-37 were
identified (Table 4.6).
In addition to the sequence similarity, budding yeast Rim15p, fission yeast
Cek1p and Ppk18p have split kinase domains that are interrupted between
subdomain VII and VIII by a short insert sequence (~190 amino acids in Rim15p;
~120 amino acids in Cek1p and Ppk18p) [302, 362]. Besides the atypical kinase domain structure, Cameroni et al. also reported that an unconventional evolutionarily conserved PAS domain exists in budding yeast Rim15p and fission yeast Cek1p and Ppk18p [34]. Thus, Cek1p and Ppk18p are legitimate candidate
Rim15p orthologs.
If Clg1p/Pef1p regulates S. pombe chronological lifespan through Cek1p or
Ppk18p, deletion of cek1+ or ppk18+ should block lifespan extension caused by
the clg1 mutation (Figure 4.6). Therefore, the lifespans of the cek1 and ppk18 single deletion mutants and the corresponding clg1 double deletion
153 Table 4.6. Comparison of S. cerevisiae Rim15p and its probable S. pombe orthologs Cek1p and Ppk18p identified by BLASTP search PAS Kinase Max Total Query Protein Accession E-valuec domaina insertb scorec scorec coveragec Rim15p NP_116620.1 YES YES - - - - Cek1p NP_588310.1 YES YES 231 547 50% 5e-61 Ppk18p NP_001018270.1 YES YES 212 551 59% 5e-55 a. See [34] b. See [302, 362]; the kinase insert between subdomain VII and VIII were also identified in a domain search on Pfam 24.0 database. c. Protein blast search in NCBI using the dataset “non-redundant protein sequences” including organism “S. pombe (taxid:4896)” sequences only (as of May, 2010).
mutants were determined. The clg1 single mutant consistently had a longer
lifespan (Figure 4.12A). As shown in Figure 4.12B, the cek1 single deletion mutant had a lifespan comparable to that of the wild type control. When the same deletion mutation was introduced in the long-lived clg1 mutant, not only the lifespan-extending phenotype of clg1 was abolished, the lifespan of cek1 clg1 double-deletion mutant became very similar to that of the cek1 single mutant and the wild type strain (Figure 4.12C, D). These data indicate that lifespan extension by clg1+ deletion requires Cek1p.
Deletion of ppk18+ resulted in shortened lifespan (Figure 4.13B), which indicates that Ppk18p may be required for survival in stationary phase. The
ppk18 clg1 double deletion mutant also had a lifespan shorter than that of the
wild type control (Figure 4.13C). Unlike the cek1 clg1 mutant, the ppk18 clg1 double deletion mutant consistently showed an intermediate lifespan between the short-lived ppk18 and the long-lived clg1 (Figure 4.13D), suggesting that Ppk18p and Clg1p affect lifespan through different genetic pathways. As the lifespan curves of the cek1 and cek1 clg1 strains (Figure
154 4.12D) are more similar than the analogous curves of ppk18 and ppk18 clg1
(Figure 4.13D), these data indicate that only Cek1p acts in the lifespan regulation pathway of the Clg1p/Pef1p complex.
Figure 4.12. Deletion of cek1+ in clg1 background abolishes the lifespan-extending effect of clg1. (A) Deletion of clg1+ extended lifespan (KRP87). (B) Deletion of cek1+ did not affect lifespan significantly as the lifespan curves of cek1 KRP173 and the wild type control overlapped on most of the sampling points (p > 0.06). (C) The lifespan of the clg1 cek1 double deletion mutant KRP171 was also not significantly different from that of the wild type (p > 0.6). (D) Deletion of cek1+ in clg1 background abolished the lifespan-extending phenotype of the clg1 mutant KRP87 and the lifespan of the clg1 cek1 double deletion mutant KRP171 was not significantly different from that of cek1 single deletion mutant KRP173 (p > 0.2).
155 Figure 4.13. Deletion of ppk18+ shortens chronological lifespan in both wild type and the clg1 strains. (A) Deletion of clg1+ ORF consistently increased lifespan. (B) Deletion of ppk18+ (KRP92) resulted in a shortened lifespan in wild type and (C) in the clg1 background (KRP109). (D) The clg1 ppk18 double deletion mutant had an intermediate lifespan which was different from that of clg1 KRP87 and distinguishable from that of ppk18 KRP92 (p > 0.4316, day 1 to day 9).
Cek1p physically interacts with Pef1p
In budding yeast, Rim15p is a direct substrate of the Pho80p/Pho85p
complex. To test whether Cek1p physically interacts with the Clg1p/Pef1p
complex, FLAG-tagged Cek1p was co-expressed with Pef1p-3HA for co-IP
analysis. IP of cell lysate with anti-FLAG antibody followed by Western
blotting with anti-HA antibody only detected Pef1p-3HA in cells bearing
pREP41-FLAG-cek1+, but not in cells bearing pREP41-FLAG empty vector or
pREP41-FLAG-ppk18+ (Figure 4.14). Similar to previous results with S.
cerevisiae Rim15p [283, 362], both FLAG-Cek1p and FLAG-Ppk18p showed
156 significant degradation (Figure 4.14). In addition, FLAG-Ppk18p was expressed at lower levels than FLAG-Cek1p. Consequently, a lower level of full-length
FLAG-Ppk18p association with Clg1p/Pef1p cannot be excluded. However, the strong association detected between Cek1p and Pef1p is consistent with the above genetic analysis and indicates that Cek1p is a downstream effector of the
Clg1p/Pef1p complex for lifespan regulation.
Figure 4.14. Pef1p interacts with Cek1p in fission yeast cells. The control vector pREP41-FLAG, pREP41-FLAG-cek1+ or pREP41-FLAG-ppk18+ was transformed into K566-11 cells that expressed Pef1p- 3HA. Cell lysate prepared from the transformed cells was immunoprecipitated with anti-FLAG (M2) antibody-coupled protein G sepharose. Western blotting was used to detect the presence of Pef1p- 3HA, FLAG-Cek1p or FLAG-Ppk18p in the IP samples or crude lysate (100 g per lane). Pef1-3HA was strongly detected in the IP sample that also had FLAG- Cek1p, but only background levels were detected in the vector control and FLAG- Ppk18p. Both FLAG-Cek1p and FLAG- Ppk18p underwent significant degradation during lysate preparation, shown as the multiple bands on the film. The full-length FLAG-Cek1p and FLAG-Ppk18p are indicated by red and yellow arrow heads, respectively.
A proposed model of chronological lifespan control in S. pombe by Clg1p, Pef1p and Cek1p
Based on the lifespan analysis on the clg1, pef1, clg1 pef1, cek1 mutants and physical interactions between Pef1p, Clg1p and Pef1p, Cek1p, we proposed a working model in which these three proteins may regulate the chronological lifespan of the fission yeast S. pombe. In the wild type cells, the
157 Clg1p/Pef1p cyclin/CDK complex may promote cell proliferation and maintain a
constant lifespan through interacting and phosphorylating Cek1p, possibly
promoting phosphor-Cek1p nuclear exclusion similar to the behavior of Rim15p
in budding yeast [369] (Figure 4.15, left). In the absence of the activity of
Pef1p/Clg1p complex (e.g. deletion of clg1+ or pef1+), lifespan can be extended
through Cek1p (Figure 4.15, middle) as depletion of Cek1p in the long-lived mutant without Pef1p/Clg1p activity abolishes the longevity phenotype (Figure
4.15, right). The nuclear-cytoplasmic shuttling of Cek1p and phosphorylation of
Cek1p by Pef1p/Clg1p remains to be tested, as well as the different transcriptional consequences of altering the phosphorylation state of Cek1p.
Figure 4.15. A proposed model of CLS regulation by Clg1p, Pef1p, and Cek1p. In wild type cells, the cyclin/CDK complex Clg1p/Pef1p interacts with the protein kinase Cek1p, and may regulate the activity or subcellular localization of Cek1p by phosphorylation to maintain normal lifespan (left). The loss of Clg1p/Pef1p activity results in S. pombe CLS extension, presumably through activation of Cek1p and its downstream targets (middle). Consistently, elimination of Cek1p abolishes CLS extension by loss of Clg1p (and likely loss of Pef1p).
158 Discussion
In this study, a novel S. pombe insertion mutant library was used to demonstrate the proof of principle for performing parallel selection on a group of mixed mutants to identify long-lived strains. The S. pombe insertion mutants
used in the selection are tagged by random barcodes which can serve as specific
primers to determine the insertion mutations. These barcodes are flanked by two
Sfi I sites and after PCR amplification and Sfi I digestion, the barcodes can be
oligomerized in a unique orientation by a routine ligation reaction to facilitate
efficient barcode sequencing. Therefore, this approach only requires commonly
available techniques (e.g. PCR, restriction enzyme digestion, ligation, cloning,
and sequencing). Additional advantages of a library of this kind are that a
variety of parallel selections can be performed and the prior knowledge of the
barcode identities in the mutants is not required as they can be determined after
selection by barcode oligomerization and sequencing.
A genetic screen was carried out by selecting ~3,600 barcode-tagged
insertion mutants for lifespan-extending mutations using the S. pombe aging assay
described in Chapter 2 and [39]. Barcodes amplified and sequenced from 600
surviving cells sampled from the day 14 aged culture showed distinct frequencies.
While many barcodes were only identified very few times (< 5 times), 6 different
kinds of barcodes showed conspicuous enrichment after the selection. These 6
barcodes identified 3 mutants, 2 of which had longer lifespans than the wild type
cells (Figure 4.5). Thus, one can directly select for long-lived mutants using this
type of random barcoded insertion mutant library.
159 The efficiency of barcode sequencing and identification of the selected
mutants with the predicted phenotypes proved to be two important rate-limiting
steps in this approach. To ensure efficient and cost-effective barcode
sequencing, one needs to be able to obtain significantly long barcode oligomers
and clone these barcode oligomers into the sequencing vector with a high ligation
frequency. An important point for barcode oligomerization was having
sufficient starting material (i.e. barcode monomers) for the reaction, so that the larger oligomers could be gel isolated from monomers and shorter ones in
sufficient amounts for ligation. The high rate of ligation was found to be
achievable by CIP treatment at 50oC to efficiently remove the phosphate group on
the ends of the digested vector with 3’ extensions to prevent vector self-ligation.
Because the identity of the barcode and insertion site in each mutant is
unknown, isolation of selected mutants that may have the predicted phenotypes
from this S. pombe insertion mutant library currently depends on identifying the
barcode of the selected mutant and using that sequence as a specific primer in
specialized PCR approaches (e.g. TAIL-PCR, splinkerette PCR) to map the
insertion sites in the genome. This procedure was time-consuming as the
number of total selected mutants was high and insertion structures were
sometimes complex. This difficulty stems from the rearrangement of the
insertion vector transformed into S. pombe and the capture of other sequences (i.e.
mitochondrial DNA) at the insertion site. The variable sequences at the
junctions prevent the use of any high-throughput methods to determine insertion
sites. Thus, while our barcoded insertion vector approach overcame the
160 problems that thwarted previous insertion libraries such as large numbers of extrachromosomal vectors [53] and tandem integrants whose insertion sites could not be mapped [52], it is not a high-throughput library at this point. These issues highlight the differences between S. pombe and other unicellular eukaryotes such as S. cerevisiae and Dictyostelium discoideum where transfection of insertion vectors results mostly in insertion of one copy of the vector [186, 215, 307]. It is notable that the types of insertions we detected resemble those seen in mammalian tissue cultures [293], suggesting strong similarities in DNA repair in S. pombe and mammalian cells.
While enrichment of the barcodes suggests better survival of the corresponding mutants under this selection, one mutant (1a8-4032; insertion in
SPNCRNA.142) did not showed a longer-than-wild type lifespan in the CLS assay to reconfirm the long-lived phenotype. One explanation for the increased proportion of this mutant in the selected culture is that the mutation might increase growth rate, which could increase the overall fraction of this mutant in the proliferating phase. Because this mutant died at a similar rate as the wild type cells (Figure 4.5), the increased growth rate would have made the proportion of this mutant in the day 0 culture similar to that in the day 14 culture (i.e. ~17% of cells in the day 0 culture were this specific mutant). However, barcode sequencing of the day 0 culture did not detect this barcode (barcode 4032) and does not support this assumption. The other possibility is that the improved survival of this mutant in quiescence resulted from and depended on an interaction(s) with other mutants in the aged culture (e.g. factors released from
161 other dead or lysed mutants). Therefore, in a culture where all cells were derived from this mutant background, the long lifespan phenotype could not be repeated due to the absence of such interaction.
The identification of a long-lived mutant with an insertion in the clg1+/ mug80+ gene shows that this approach can identify lifespan-controlling genes in conserved longevity pathways. The budding yeast ortholog of clg1+/mug80+ coded protein, Clg1p, is a member of a Pcl cyclin family whose interacting CDK is Pho85p. Although budding yeast Clg1p has not been directly implicated in yeast lifespan regulation, it has been shown that Pho85p together with one of the
Pcl cyclins, Pho80p, regulate the transition from proliferation to stationary phase
[369]. In conditions favoring cell growth, Pho80p/Pho85p phosphorylates protein kinase Rim15p, promotes Rim15p interaction with the 14-3-3 protein
Bmh2p, which results in cytoplasmic export and retention of Rim15p [369].
RIM15 was first identified as a glucose-repressible gene whose expression is important for early meiotic gene expression in S. cerevisiae [362]. Rim15p is also required for proper expression of genes up-regulated before entering into stationary phase, glycogen and trehalose production, thermal tolerance and cell survival in stationary phase [34, 283]. These results suggest that the
Pho80p/Pho85p cyclin/CDK complex controls entry into a quiescent state through
Rim15p. As we had previously found that efficient entry into a quiescent state increased S. pombe CLS ([39] and Chapter 3), we hypothesized that S. pombe
Clg1p controlled CLS in a manner similar to Pho80p/Pho85p and Rim15p.
The potential S. pombe orthologs of Pho80p include a group of three Pcl-
162 like cyclins, and the S. pombe Pho85p CDK ortholog appears to be Pef1p. Pef1p
was first identified as a kinase associated with the cyclin Pas1p and this complex
regulates the G1/S transition through the Res2p/Cdc10p complex [344].
However, the lifespan regulating function of Pef1p appears to be specifically
associated with Clg1p, and is independent of Pas1p and the third Pef1p-
associating cyclin Psl1p/Spbc20f10.10p because deletion of these two cyclin-
coding genes did not increase lifespan (Figure 4.11).
The potential Rim15p orthologs suggested by sequence similarity were
Ppk18p and Cek1p (Table 4.6). While no known biological function has been assigned to Ppk18p, Cek1p was linked to the control of cell cycle progression and
DNA damage checkpoint [182, 302]. Epistasis analysis between clg1 and cek1 or ppk18 indicated that Cek1p seems more likely to be the downstream effector of Pef1p/Clg1p based the results that the cek1 clg1 double deletion mutant had a lifespan very similar to the cek1 single deletion mutant, and deletion of the cek1+ ORF in the clg1mutant abolished the longevity phenotype
of this mutant. On the other hand, ppk18 clg1 had an intermediate lifespan
between that of clg1 and ppk18, indicating that these two genes alter lifespan
through different pathways.
Evolutionary conservation of lifespan controlling pathways from yeast to
humans has been demonstrated in the TOR, PKA, AKT, S6K and CR pathways,
and our results coupled with those from budding yeast suggest that Pho80p/
Pho85p and Clg1p/Pef1p may define another broadly conserved pathway [81,
135]. Mammalian Cdk5, which is known for its role in regulating neuronal
163 migration and nurite growth [89], has been shown to be a functional ortholog of
Pho85p in two independent studies. Ectopic expression of rat or mouse Cdk5 in budding yeast complemented several defects in the pho85 mutant [136, 251].
When expressed in budding yeast cells, mouse Cdk5 associated with at least 5 Pcl
cyclins and was able to phosphorylated Pho4p (a Pho85p substrate) in an in vitro
kinase assay [251].
Another intriguing question that lies ahead is whether a common mediator
that integrates and transmits signals from different pathways to regulate cell fates
between proliferation and quiescence is also evolutionarily conserved. In
budding yeast, Rim15p kinase has been shown to function in this manner and
coordinate signals from TOR, PKA, AKT (Sch9p), Pho80p/Pho85p and caloric
restriction. In this study, the identification of the lifespan-extending mutation in
clg1+ led to the discovery of Cek1p as a Rim15p ortholog in S. pombe. Cek1p
shows sequence homology to two protein kinase families in mice and human,
MAST (microtubule associated serine/threonine kinase) and LATS (large tumor
suppressor homolog) kinases, which also seem to have sequence homologs in
worms and flies based on a BLASTP search at NCBI. Some members of MAST
and LATS also contain the atypical interrupted kinase domains that are found in
Rim15p and Cek1p. Examining the roles of these genes in lifespan regulation
and their genetic and/or physical interactions with the currently known longevity
pathways in other multicellular model systems should provide more information
regarding how conserved this regulatory mechanism is and identify important
regulators of quiescence and proliferation in complex eukaryotes.
164 Chapter 5
General discussion and future directions
165 Using yeast as a model to study aging: advantages and limitations
Although the ultimate outcome of aging is loss of viability, aging-related disorders such as cancers, cardiovascular and neurodegenerative diseases appear to be the most “bothersome byproducts” of aging. Understanding the genetic determinants and molecular mechanisms of aging and aging-associating disorders can potentially extend life expectancy. More importantly, the same information will help the development of interventions and cures that can delay and alleviate the onset and symptoms of aging-related diseases to produce a healthier physiological state in old age. However, aging research in humans and mammalian model organisms is a challenging task due to the complexity of aging phenotypes and intricate genetic and environmental interactions that affect aging.
For example, even though the effect of dietary restriction was first recognized in laboratory rats over 80 years ago, the genetic explanation for this phenomenon did not begin to surface until the last two decades or so. Therefore, scientists have turned to simple model organisms like yeast, flies and worms, which possess genetic and molecular techniques to study aging at both cellular and organismal levels.
Among the commonly used model organisms, yeast has been one of the most productive. Despite the conspicuous difference between yeast and mammals (e.g. size, gene number, physiology), history has taught us that many important fundamental properties of life are essentially the same in yeast and mammals (e.g. cyclin-dependent kinase controlled cell cycle progression [256]), and studies in yeast greatly facilitate our understanding of these subjects in
166 complex organisms. In addition to serving as a source of knowledge on evolutionarily conserved aspects of life, there are also many successful examples of using yeast as a model to elucidate the genetic basis of certain human diseases
and the effect of therapeutic agents to cancers [232, 265, 273, 321, 363].
Well-established yeast genetics and detailed knowledge in yeast molecular
and cell biology provide a solid advantage for researchers to study aging and
longevity in yeast. Additionally, the relatively short lifespan of yeast makes it
very suitable for aging research, in which the major phenotype to score is the
length of time that organisms remain alive. Moreover, in multicellular
organisms, the overall lifespan is determined by the aging processes of two cell
types, post-mitotic cells and proliferating cells. These two types of aging can
not be easily separated when analyzing the lifespan of multicellular organisms.
As a unicellular organism, the lifespan of yeast can be measured as replicative
lifespan (RLS) of dividing cells in the proliferation phase and chronological lifespan (CLS) of quiescent cells in the stationary phase. Therefore, yeast provides opportunities to study genetic and molecular mechanisms specific to and common in each type of aging. Many genetic and non-genetic interventions shown to increase yeast lifespan have now been found to have similar effects on lifespan or alleviate aging-associated pathologies in mice [81]. This conservation of known lifespan-regulating pathways further supports yeast as an excellent model for exploring the core regulatory mechanisms that are evolutionarily conserved in mammals.
Of course, one needs to be aware that the evolutionary distance between
167 yeast and metazoans has set some limitations in the power of yeast genetics.
Genes and molecular pathways that evolved and appeared in metazoans after their divergence from yeast are not possible to study in yeast. One of such examples is the insulin/IGF-1 pathway, which regulates energy metabolism, nutrient homeostasis and lifespan in animals as simple as C. elegans and D. melanogaster.
Although some downstream signal transducers exist in yeast (e.g. the AKT orthologs Sch9p in budding yeast and Sck1p, Sck2p in fission yeast), many components in this pathway are absent. The same limitation also applies to other hormone and endocrine regulatory pathways which are known to affect aging in animals [11]. Therefore, the simplicity of yeast prevents examination of metazoan-specific genes and pathways. Also, tissue-specific overexpression or inactivation of specific genes has been shown to increase lifespan in worms and flies [175]. Although the detailed mechanisms regarding how localized genetic changes can affect the overall lifespan of the entire organisms are unclear, communications between cells, tissues and organs are necessarily involved. As a unicellular organism, the yeast system cannot address many of these complexities.
Even when a conserved gene is implicated in regulating aging in both yeast and multicellular organisms, the underlying mechanisms can be different. In budding yeast, the sirtuin Sir2p is proposed to extend RLS in part by suppressing the production of extracellular ribosomal circles (ERCs) [158]. The requirement of sirtuins for long lifespan has also been proven in invertebrate model organisms, such as flies and worms [210]. In mice, sirtuins have been implicated in
168 regulating symptoms and phenotypes that are closely associated with aging as well [111]. However, in both invertebrate models and mice, sirtuin-dependent lifespan extension and retardation of aging phenotypes appear to be unrelated to
ERC formation. This demonstrates that even though sirtuins in different species
have similar primary sequences and biochemical properties, they regulate lifespan
through distinct mechanisms. It is conceivable that there are other conserved
genes that regulate organismal aging through different mechanisms.
The fission yeast Schizosaccharomyces pombe as an emerging yeast aging model
Yeast aging research began with studies in cell divisions in 1950s [12, 239]
and started to bloom in 1990s. The budding yeast S. cerevisiae was the sole
yeast model used in aging research until Barker and Walmsley published the first paper describing replicative aging in the fission yeast S. pombe [9]. Due to the
technical difficulty on discriminating the mother and new-born daughter cells in fission yeast, the field had stayed quiet until almost a decade later, asymmetric distribution of oxidatively damaged proteins was observed in dividing fission yeast cells and proposed as a determinant of their RLS [66, 234]. Interestingly, the partition of damaged proteins in the fission yeast depends on Sir2p and functional cytoskeleton as in the budding yeast, suggesting that the underlying mechanism is evolutionarily conserved [1, 66, 205].
Despite the early initiation of S. pombe replicative aging research, the genetic and molecular mechanisms controlling fission yeast CLS has drawn the most attention in the recent years [295]. This field broke the ground by
169 revisiting and recapitulating some early findings in budding yeast and other model organisms, such as caloric restriction and mutations in AKT kinase-coding genes and PKA pathways [39, 296, 297]. In another study, Zuin et al. isolated several
mutants defective in mitochondrial function and established a correlation between reduced CLS and impaired electron transport chain and respiration in fission yeast
[393].
There were also a few new findings. Fatty acid metabolism was suggested as a determinant of lifespan based on studies in which deletions of long-chain fatty acyl-CoA synthetase lcl1+ and lcl2+ resulted in lifespan shortening and extension, respectively [88, 264]. Additionally, while essentially all studies in yeast and other model organisms examine the effects of activity-abolishing mutations (e.g. deletion, RNAi) on lifespan, Ohtsuka et al. took a different approach and identified three small proteins called Ecls (extended chronological
lifespan) whose overexpression promoted lifespan [259, 260]. Recently, the
same approach was used to identify 4 genes (i.e. sds23+, cka1+, adh1+, rpb10+)
whose overexpression could suppress the respiration defects in a constitutively active G protein G subunit mutant gpa2R167H and extend CLS in this mutant and
wild type cells [294].
Current approaches for large-scale genetic screens for genes determining yeast
lifespan
The availability of the budding yeast deletion mutant collection greatly
facilitates large-scale isolation of mutants with extended lifespan. Earlier efforts
170 focused on analyzing RLS or CLS of mutants in this collection individually and
successfully discovered mutants in pathways involved in longevity regulation
[160, 278]. The first large-scale RLS screen was not trivial because it required tremendous amounts of time and labor in micromanipulation of cells under microscopes to even determine only one eighth of the available haploid deletion mutants [160]. It will certainly be more challenging if this approach is to be utilized in fission yeast. Therefore, a robust assay that can automatically eliminate new-born daughter cells in a RLS experiment and specifically determine the number of mitotic divisions the mother cells undergo will facilitate large-scale identification of RLS-determining genes.
Two independent studies provided potential solutions. Jarolim et al. replaced the wild type CDC6 gene with an engineered CDC6 allele under the control of the promoter of the mother-specific HO endonuclease gene. The
CDC6 gene is essential for cell growth and Cdc6p expression driven by HO promoter only occurs in mother cells but not the new born cells [143]. On the other hand, Lindstrom and Gottschling created a “mother enrichment program
(MEP)” that can specifically and inducibly eliminate two essential genes, UBC9 and CDC20, in the daughter cells by an estradiol-dependent Cre recombinase driven by the promoter of daughter cell-specific gene SCW11 [204]. Therefore, new born daughter cells in both assays will not be able to grown and divide due to the lack of essential proteins (i.e. Cdc6p, Ubc9p, Cdc20p) and the lifespans of a group of mother/founding cells grown in a liquid culture can be determined based on the optical density of the final culture, which correlates with the dividing
171 potential of a mutant and the total number of progenies generated by the input
mother cells [143, 204]. To conduct genome-wide screening, these genetic
systems need to be introduced into the yeast deletion strain set using existing
high-throughput techniques such as SGA [351], or large-scale mutagenesis that
allows rapid identification of the mutated gene (e.g. transposon mutagenesis [32,
331]).
The above approaches rely on the asymmetric expression of certain genes in budding yeast. Although fission yeast undergoes morphologically symmetric division in majority of the lifespan and genes with expression patterns similar to
that of HO endonuclease or SCW11 have not been found, the phenotype of
asymmetric damaged protein segregation in both yeast species indicates that there
may be genes whose expression is cell lineage-specific in fission yeast as well.
The discovery of such genes in combination with the above described approaches
should create an opportunity to systematically study genetic determinants of
fission yeast RLS.
The protocols which are suitable for high-throughput analysis CLS in
budding yeast deletion mutant collection have also been described [72, 221, 245,
278]. The initial studies involved aging individual ORF deletion mutants in multi-well plates and transferring aged cells to nutrient-rich liquid medium. The fraction of viable cells at each sampling time point decided the outgrowth efficiency of the aged culture in the fresh medium and the optical density of the re-inoculated cultures after a constant period time of incubation. By compiling the optical densities of outgrowth cultures at different time points, the lifespan of
172 individual mutants could be determined. A mutant with an extended lifespan
would have more viable cells at most of the sampling time points, retain better outgrowth efficiency and produce higher optical densities in the outgrowth cultures. The initial study successfully identified several components in the
TOR pathway whose mutations extended lifespan. However, in the process of aging of initial cultures and outgrowth of aged cells, cells were allowed to settle to the bottom of wells [278]. This type of culturing method will likely affect nutrient distribution and aeration, which becomes important when yeast cells reach stationary phase and gradually shift metabolism to respiration and oxidative phosphorylation. The second approach utilized the Bioscreen C MBR machine
(Growth Curve USA), a computer-controlled incubator/shaker/OD reader that provides better culture conditions (e.g. evaporation prevention, constant shaking of cultures in 100-well culture plates) and more precisely determines the growth kinetics of aged cells at multiple time points of during outgrowth [245].
Although this machine has been demonstrated to be able to monitor aging and other experiments required accurate quantification of cell growth, the high accuracy and design for proper growth in small-well culture plates do come with a high price tag (i.e. ~$ 35,000) and the 100-well plates specific to this machine can not be easily adapted to the conventional 96- or 384-well plates used in most assays [245]. Nonetheless, the Bioscreen C MBR machine-based assay can be an efficient tool for applications to genome-wide screens for budding and fission yeast mutants with altered lifespan.
Recently, the Longo and Smith groups took the advantage of characterized
173 barcode sequences and commercially available barcode microarrays of the S.
cerevisiae ORF deletion set to perform parallel selections for CLS-regulating
genes [72, 221]. This microarray-based approach allowed both studies to identify long- and short-lived mutants. Among the short-lived mutants, both studies independently identified mutants defective in autophagy, consistent with the previously known roles of autophagy in organismal longevity [48]. In accordance with the role of autophagy in degradation of macromolecules and organelles, deletions of genes involved in mitochondrial degradation and vacuolar function were also found to shorten lifespan [72]. Novel conserved lifespan- extending mutations were also identified in both studies. The Smith group identified long-lived mutants involved in the de novo purine biosynthesis pathway and the Longo group isolated mutants involved in fatty acid transport, biosynthesis and tRNA methylation [72, 221].
Developing a S. pombe CLS assay and a novel genetic screen for long-lived fission yeast mutants
As S. pombe is emerging as a valuable aging model, it is important to have a proper assay, a versatile mutant library and an efficient selection approach for large-scale screen for lifespan-regulating genes. In this dissertation, a S. pombe aging assay which can be used to determine the lifespans of individual mutants and to survey the long-term survival profile of thousands of mutants in a mixed culture was developed.
Although some fission yeast chronological aging assays directly employed
174 the protocols from the budding yeast experiments, the CLS assay presented in this dissertation was created with two criteria in mind. The first criterion is that a legitimate CLS assay should recapitulate evolutionarily conserved features of aging, and this was confirmed by the lifespan-extending effects of CR and deletions of sck1+, sck2+ and tor1+ detected by this S. pombe aging assay. The
second concerns about how the aging assay should be controlled to allow proper
comparison between the wild type and mutant cells. In most budding yeast CLS
assays, to avoid the effects of different auxotrophic mutations between the wild
type control and mutants of interest on cell growth and lifespan, extra amounts of
supplements to the auxotrophic mutations were added to the medium. Although
no concerns have been raised, it has been reported that the concentrations of
amino acid and nucleobase supplements can affect yeast lifespan [146, 221]. To
avoid complications that might arise from the difference in supplement levels and
auxotrophic mutations, the S. pombe CLS assays used in this project employed
different wild type strains with auxotrophic markers matching the mutants of
interest as the controls. This consideration allows unambiguous determination
of the effects of gene mutations on lifespan.
As described in Chapter 2, a fission yeast aging assay that can monitor the
viability of a population of cells from 107 to108 cells/ml to < 10 cells/ml was
developed. The selection approach itself has a low tendency of generating new/spontaneous mutations or epigenetic variants which promote cell growth, survival and change the genetic composition in a culture during long-term incubation in stationary phase, a phenomenon that was observed in budding yeast
175 chronological aging [24, 74, 76]. An aging assay that does not promote adaptive mutations or epigenetic variants is important because it will not interfere with the selection and identification of lifespan-extending mutations from defined pools of known mutants.
This assay in combination with the barcode-tagged insertion mutant library described in Chapter 3 enabled a novel parallel selection that allowed monitoring and isolation of long-lived mutants from a pool of mixed mutants. The presence of barcodes in these insertion mutants makes it possible to select for thousands of mutants in a single culture and subsequently recover mutants enriched in the selection by determining the frequencies of different barcodes. An advantage specific to this mutant library is that analysis of the final barcode composition does not require microarray as parallel selections using the budding yeast deletion mutant collection do. Instead, only routine molecular biological techniques such as PCR, ligation and the conventional sequencing method are needed to determine the sequences of Sfi I sites-flanked barcodes. The proof-of-principle experiment described in Chapter 4 shows that this method successfully detected mutants and barcodes which were enriched after 14-day selection in stationary phase.
The barcode oligomerization and sequencing strategy created in this study depend on the barcode-flanking Sfi I sites and allow efficient determination of barcode abundance and composition without the prior knowledge of barcode. In principle, a similar strategy can also be applied to barcoded mutants whose barcodes are not designed with flanking Sfi I or other restriction enzyme sites, not available as microarrys and barcode sequences are unknown. In this case,
176 barcodes can be amplified with primers containing Sfi I recognition sequences, other restriction enzyme sites or specialized sequences that can be processed to generate specific single-strand overhangs on the ends of the barcodes (e.g. uracil- specific excision reagent (USER) [19]). These barcodes with unique single- stranded overhangs can be analyzed in the same way as described in Chapter 4.
Limitations of the current barcode-tagged insertion mutant library and probable alternatives
Despite aforementioned advantages and the successful proof-of-principle experiment described in Chapter 4, this barcode-tagged insertion mutant library does have certain limitations. First of all, the non-homologous recombination- based integration method produced unexpected complex insertions which increased the chance of tagging the same strain with multiple barcodes and impeded high-throughput determination of the insertion sites. Second, although this method has the benefit of low cost and requirement of only routine techniques, the throughput of this method is no match to conventional microarray or high-throughput sequencing technology. In addition, while in principle both positive and negative selection can be performed in this barcode-tagged insertion mutant library, the lack of knowledge of the barcodes and insertion sites in these mutants reduces the feasibility of negative selection and efficient mutant isolation.
Even with the aforesaid limitations, the higher mutation variety of an insertion mutant library compared to a defined collection of ORF deletions still makes an insertion mutant library an attractive tool for aging and other biological
177 studies. Therefore, alternative mutagenesis methods that produce simple and unbiased insertions will be needed. Transposon-mediated mutagenesis is a likely candidate approach. Bacterial transposon Tn3 insertion into genomic libraries followed by their homologous integration into the S. cerevisiae genome has been used as a large-scale mutagenesis approach [32, 316]. Similarly,
Evertts et al. recently demonstrated the use of a DNA transposon hermes from house fly Musca domestica as mutagen to generate random insertion mutants in S. pombe [69]. The advantage of this approach is that the single, simplified insertions occur in most of insertion mutants and the transposition events have a broad distribution across the genome (i.e. equal insertion frequencies in coding and non-coding sequences) [69].
With the advancement of the current genomic technologies, it will not be difficult to pre-determine the short-sequence barcodes in hundreds of thousands of barcode-tagged mutants arranged as mutant arrays by high-throughput deep sequencing technologies in combination with molecular marker anchoring, multi- dimensional pooling tactic [230, 378]. This approach has recently been adopted in budding yeast (i.e. barcode analysis by sequencing, also called “Bar-seq”) to re-determine and verify the barcode sequences in the deletion mutant collections with high sensitivity and accuracy [327, 328]. Large-scale insertion site determination will also be possible for a new barcoded insertion library created by transposon-mediated mutagenesis. The simpler integration structures produced by transposons should allow inverse PCR or TAIL-PCR to efficiently amplify genomic DNA adjacent to the insertion vector for high-throughput sequencing.
178 A second generation fission yeast barcode-tagged insertion mutant library with
equivalent information to the budding yeast ORF deletion mutant sets will greatly
facilitate the discovery of genes that limit and promote lifespan. Construction of such a library is currently underway in our laboratory.
In addition to re-sequencing barcodes in the budding yeast deletion mutant collection, the same deep-sequencing technique has been used as a quantitative
phenotyping tool that successfully determined the relative sensitivity of different
mutants to chemotherapeutic drugs. This approach also allows analyzing
multiple experiments in a single sequencing reaction [327, 328]. When applied
to parallel selections for mutants with altered lifespan, the multiplexing capacity
of deep sequencing will allow one to monitor CLS of thousands mutants at
multiple time points and under various selecting conditions in a single sequencing reaction.
A Proof-of-principle parallel genetic screen identified the Clg1p-Pef1p-Cek1p
lifespan-regulating pathway
To test the feasibility of parallel selection in fission yeast aging research, a
proof-of-principle genetic screen was carried out with a group of barcode-tagged
insertion mutants and identified the Clg1p-Pef1p-Cek1p pathway that regulates
CLS in S. pombe. Two similar screens in budding yeast using the barcode-
tagged ORF deletion set also identified novel lifespan-regulating genes [72, 221].
These results clearly demonstrated that this type of approach can directly select
mutants with altered lifespan. Compared to the methods used by Powers et al.
179 and Murakami et al. which assayed mutants separately [245, 278], the major advantage of parallel selection is the reduced need for labor-intensive culturing
processes (i.e. monitoring thousands of mutants in a single culture versus multiple
96-well or 100-well plates). Despite the improved convenience, it is important
to be aware of some aspects that may affect the interpretation and subsequent
analysis of the results from such batch selections. As the relative lifespan is
measured as the number of each remaining viable mutant and its associated
barcode in the batch culture, the growth rate and fluctuation in the initial cell
number of each mutant strain will affect the presentation of the final
mutant/barcode composition after selection (e.g. under-presentation of a long-
lived mutant with a slow growth rate or over-presentation of a mutant with normal
lifespan inoculated at a higher initial cell number). Also, with increasing
chronological age, it has been reported that aged budding yeast cells could
produce substances such as ethanol, acetic acid or low-molecular-weight proteins
that influenced the viability of the surviving cells in the culture [33, 71, 127, 221].
The medium composition of an aged culture with many different mutants is likely
to be different from that of an aged culture with only one single mutant strain.
Therefore, a mutant with a longer lifespan in a culture with thousands of different
mutants might not have the same phenotype when analyzed alone in an individual
culture due to the lack of lifespan promoting or inhibiting factors generated by
other mutant cells.
Nonetheless, the proof-of-principle selection in Chapter 4 identified a long-
lived mutant with a mutation in the cyclin coding gene clg1+. Our genetic and
180 biochemical data propose a model of Clg1p-dependent lifespan regulation, in
which the CDK Pef1p and the cyclin Clg1p negatively regulate longevity by
inhibiting the kinase Cek1p (Figure 5.1). Elimination of Clg1p, Pef1p, or both
proteins results in S. pombe chronological lifespan extension which can be
reversed by depletion of the Cek1p kinase.
Figure 5.1. A model of Clg1p-Pef1p-Cek1p-dependent CLS regulation. In wild type cells, the cyclin/CDK complex Clg1p/Pef1p interacts with the protein kinase Cek1p, and may regulate the activity or subcellular localization of Cek1p by phosphorylation to maintain normal lifespan (left). The loss of Clg1p/Pef1p activity results in S. pombe CLS extension, presumably through activation of Cek1p and its downstream targets (middle). Consistently, elimination of Cek1p abolishes CLS extension by loss of Clg1p (and likely loss of Pef1p).
The budding yeast ortholog of Pef1p, Pho85p, has 10 associating cyclins
similar to the 3 corresponding S. pombe cyclins examined in this work, and its
kinase activity is closely regulated by nutrition levels. For example, in response
to inorganic phosphate starvation, Pho80p cyclin-associated Pho85p becomes
inactive through the binding of inhibitor Pho81p to promote phosphate
181 metabolism [135]. Inhibition of the activity of Pho80p/ Pho85p by deletion
mutations has also been shown to induce the quiescent program and slightly enhanced survival early in stationary phase [369]. The above observations are consistent with the cellular response to nutrient limitation and the subsequent inactivation of Pho85p/cyclin complexes to prepare for a physiological transition from proliferation to quiescence. As the proper exit of the proliferating cycle correlates with longevity in yeast [39, 372], inactivation of S. pombe Clg1p/Pef1p may increase CLS through a similar mechanism. Although the closest S. pombe
Clg1p ortholog is the S. cerevisiae Clg1p, but not Pho80p, the reduced number of the cyclin members in S. pombe raises the possibility that the S. pombe
Clg1p/Pef1p complex may have acquired activities of multiple S. cerevisiae
Pcl/Pho85p complexes during evolution.
The Pho80p/Pho85p complex regulates the quiescent program through the
protein kinase Rim15p [369], the fission yeast ortholog of Cek1p. The
experiments presented in Chapter 4 indicate that Cek1p is required for lifespan
extension by inactivation of Clg1p/Pef1p, consistent with the activity of budding yeast Rim15p. Both Rim15p and Cek1p have non-continuous kinase domains which are interrupted between kinase subdomain VII and VIII by a kinase insert of ~100 amino acids. In the case of Rim15p, the kinase insert harbors a 14-3-3 protein binding site which can be phosphorylated by Pho80p/Pho85p and TORC1 to regulate the interaction between Rim15p and the 14-3-3 protein Bmh2p and subcellular localization of Rim15p [369]. Although no 14-3-3 protein binding sequence is predicted in Cek1p, it was speculated that the kinase insert region
182 may also regulate Cek1p subcellular localization similar to the kinase intervening
sequence of S. pombe Dsk1p kinase, which controls Dsk1p nuclear import in a
cell cycle- and phosphorylation-dependent manner [302]. Cameroni et al. also
predicted a non-conventional N-terminal PAS (Per Arnt Sim) domain, a sensor to
many intracellular cues, in both Rim15p and Cek1p. They speculated that the
PAS domain of Rim15p may involve in sensing stress levels or the redox status
[34]. It will be intriguing to determine if the PAS domain of Cek1p affects its kinase activity and/or subcellular localization and if mutations in PAS domain alone will abolish the lifespan-extending effect of Pef1p/Clg1p inactivation.
Cell cycle, senescence and aging
The relation of cell cycle and aging has been broadly studied and examined
in the context of replicative or cellular senescence. In this scenario, aging or
senescence is regarded as cells losing the ability to divide due to accumulation of
irreparable DNA damage or critically short telomeres, subsequent activation of
DNA damage checkpoint proteins, and the eventual induction of cell cycle
inhibitors and cell cycle arrest [36]. In contrast to quiescent cells which can re-
enter cell proliferation cycle upon exposure to necessary nutrients, irreversible
cell cycle arrest occurs in some senescent cells (e.g. cells with activated p16-pRB
signaling) and inactivation of checkpoint signaling is required for cells that have
undergone reversible senescence (e.g. cells with activated p53) to re-enter the cell
cycle [35]. Accumulation of these senescent cells is thought to be one of the
causes of functional decline in tissues and organs with increasing age and the
183 phenotypes of aging.
In contrast to cellular senescence which addresses aging in dividing cells,
the work presented in this dissertation provides direct evidence that cell cycle
control can be coupled to the lifespan of non-dividing, quiescent cells by showing
that efficient cell cycle arrest is correlated to lifespan extension by caloric
restriction. The correlation between long lifespan and cell cycle arrest has also
been observed in several budding yeast mutants with extended chronological
lifespan [372]. It is proposed that efficient cell cycle arrest in response to
nutrition exhaustion help establish a quiescent state quickly and prevent the
initiation subsequent cell cycle events (e.g. protein synthesis for cell growth in G1
phase and DNA replication in S phase) in the absence of essential nutrients.
Accordingly, it was found that budding yeast under calorie restriction or carrying
lifespan-extending mutations have reduced replicative stress upon enter stationary
phase [372]. Consistent with this view, stimulation of cell cycle progression in normally quiescent cell populations has been shown to trigger cell death in budding yeast and mammalian cells [100, 101, 191, 285, 357, 382].
Interestingly, the lifespan-extending mutant identified in the unbiased selection described in chapter 4 has a mutation in the cyclin coding gene clg1+.
Additional experiments also showed that deletion of the gene encoding Clg1p-
associating CDK Pef1p also extends lifespan. Although it is not known if Clg1p
or the other new Pef1p-associating cyclin Psl1p may regulate cell division cycle in S. pombe, the previously known Pef1p cyclin Pas1p has been shown to regulate
G1/S transition through the Pef1p kinase [344] and studies in S. cerevisiae also
184 indicate that some of the budding yeast orthologous proteins participate in G1/S
phase progression [68, 225]. Therefore, it will be interesting to know whether
and how the fission yeast Pef1p and associating cyclins control cell cycle and the longevity phenotypes of pef1 and clg1 associate with an altered cell cycle state.
Future directions
The budding yeast Rim15p is required for lifespan extension by CR and inactivation of TORC1, Sch9 (AKT) and PKA [77, 371], suggesting an important role of Rim15p in integrating and transmitting signals from these various pathways to regulate proliferation and longevity. Similar genetic and non- genetic interventions in S. pombe have also been shown to increase chronological lifespan (Chapter 2 and [39, 297]). Epistasis tests of cek1 with CR and mutants in the aforementioned genetic pathways will help establish whether Cek1p also functions as a common mediator of different signaling pathways in fission yeast to coordinate cell growth and lifespan.
As to the downstream signaling of Cek1p that ultimately promotes longevity, several hints are also available from the studies of budding yeast
Rim15p. One of the mechanisms by which Rim15p promotes stationary phase survival is up-regulating gene activities that are required for the transition from proliferation to stationary phase. Cameroni et al. showed that Rim15p was required for expression of 5 classes of genes (i.e. carbohydrate metabolism, general stress response, oxidative stress response, respiration, lipid and fatty acid metabolism) that are generally induced upon entry to stationary phase. In a
185 recent study, Rim15p, Pho85p and some of the Pho85p-associating cyclins were also demonstrated as regulators of autophagy [383], a conserved process
associated with long lifespan in many organisms [48]. While Rim15p clearly
acts as a positive regulator of autophagy, Pho85p and its associating cyclins are
context-dependent activators or inhibitors of autophagy [383].
To elucidate the function of Cek1p in S. pombe longevity, one approach is to
monitor genes and processes that are regulated by Rim15p in S. cerevisiae. For
example, the expression of orthologs of the Rim15p-dependent genes mentioned
above can be examined in the wild type and cek1 cells. Genes whose
expression is reduced in the cek1 mutant can be deleted or over-expressed in
cek1, clg1 or cek1 pef1 mutants to determine their roles in Clg1p-Pef1p-
Cek1p-dependent lifespan regulation. The role of autophagy in S. pombe
longevity has only started being examined recently [343]. It would be
interesting to determine whether the Clg1p-Pef1p-Cek1p pathway also regulates
autophagy in fission yeast as Rim15p and Pho85p-cyclin complexes do in
budding yeast, and if autophagy is required for lifespan extension by inactivation
of the Clg1p/Pef1p complex.
The above mentioned approaches, however, will only guarantee a limited
understanding of Cek1-dependent longevity regulation. A genome-wide
epistasis test between cek1+ and other S. pombe genes should allow identification
of more Cek1p downstream regulators and comprehensive understanding of this
signaling pathway. To perform this test, it is required to introduce cek1 to a
new or existing barcode-tagged deletion or insertion mutant library to create
186 double mutants whose lifespan can be quickly determined by high-throughput
assays. High-throughput double mutant generation can be achieved by the two
strategies similar to the budding yeast SGA approach developed by Roguev et al.
[289]. These methods allow large-scale generation of haploid double mutants from heterozygous diploid mutant cells by selectable markers expressed under mating type-specific promoters [289]. The same approach can also be applied to explore additional targets and effectors of Clg1p and Pef1p.
In summary, this study created a new path to use S. pombe as a model to
study molecular and genetic pathways that control aging and longevity in
eukaryotes. The barcode tagging strategy and insertion mutagenesis show
promises to comprehensively examine different aspects in fission yeast aging.
Although the ligation and conventional sequencing-based barcode profiling
method appears to have limited power compared to microarray and high- throughput sequencing, it provides an easy and low cost alternative that can be performed in most laboratories. A new barcode-tagged fission yeast insertion mutant library with simplified integration outcome and well-annotated barcode and mutation information will provide a major benefit, and complements the commercially available deletion mutant library to aid aging research and other biological disciplines.
187 Bibliography
1. Aguilaniu H, Gustafsson L, Rigoulet M, Nystrom T. 2003. Asymmetric inheritance of oxidatively damaged proteins during cytokinesis. Science 299: 1751-1753 2. Alvers AL, Wood MS, Hu D, Kaywell AC, Dunn WA, Jr., Aris JP. 2009. Autophagy is required for extension of yeast chronological life span by rapamycin. Autophagy 5: 847-849 3. Anderson RM, Shanmuganayagam D, Weindruch R. 2009. Caloric restriction and aging: studies in mice and monkeys. Toxicol Pathol 37: 47- 51 4. Apolinario E, Nocero M, Jin M, Hoffman CS. 1993. Cloning and manipulation of the Schizosaccharomyces pombe his7+ gene as a new selectable marker for molecular genetic studies. Curr Genet 24: 491-495 5. Araki K, Araki M, Yamamura K. 1997. Targeted integration of DNA using mutant lox sites in embryonic stem cells. Nucleic Acids Res 25: 868- 872 6. Arking R, Buck S, Berrios A, Dwyer S, Baker GT, 3rd. 1991. Elevated paraquat resistance can be used as a bioassay for longevity in a genetically based long-lived strain of Drosophila. Dev Genet 12: 362-370 7. Bai X, Jiang Y. 2010. Key factors in mTOR regulation. Cell Mol Life Sci 67: 239-253 8. Banks AS, Kon N, Knight C, Matsumoto M, Gutierrez-Juarez R, et al. 2008. SirT1 gain of function increases energy efficiency and prevents diabetes in mice. Cell Metab 8: 333-341 9. Barker MG, Walmsley RM. 1999. Replicative ageing in the fission yeast Schizosaccharomyces pombe. Yeast 15: 1511-1518 10. Barros MH, Bandy B, Tahara EB, Kowaltowski AJ. 2004. Higher respiratory activity decreases mitochondrial reactive oxygen release and increases life span in Saccharomyces cerevisiae. J Biol Chem 279: 49883- 49888 11. Bartke A. 2005. Minireview: role of the growth hormone/insulin-like growth factor system in mammalian aging. Endocrinology 146: 3718-3723 12. Barton AA. 1950. Some aspects of cell division in saccharomyces cerevisiae. J Gen Microbiol 4: 84-86 13. Basi G, Schmid E, Maundrell K. 1993. TATA box mutations in the Schizosaccharomyces pombe nmt1 promoter affect transcription efficiency but not the transcription start point or thiamine repressibility. Gene 123: 131-136 14. Baur JA, Pearson KJ, Price NL, Jamieson HA, Lerin C, et al. 2006. Resveratrol improves health and survival of mice on a high-calorie diet. Nature 444: 337-342 15. Benit P, Slama A, Cartault F, Giurgea I, Chretien D, et al. 2004. Mutant NDUFS3 subunit of mitochondrial complex I causes Leigh syndrome. J Med Genet 41: 14-17
188 16. Berdichevsky A, Viswanathan M, Horvitz HR, Guarente L. 2006. C. elegans SIR-2.1 interacts with 14-3-3 proteins to activate DAF-16 and extend life span. Cell 125: 1165-1177 17. Bhaskar PT, Hay N. 2007. The two TORCs and Akt. Dev Cell 12: 487- 502 18. Bishop NA, Guarente L. 2007. Genetic links between diet and lifespan: shared mechanisms from yeast to humans. Nat. Rev. Genet. 8: 835-844 19. Bitinaite J, Rubino M, Varma KH, Schildkraut I, Vaisvila R, Vaiskunaite R. 2007. USER friendly DNA engineering and cloning method by uracil excision. Nucleic Acids Res 35: 1992-2002 20. Bjedov I, Toivonen JM, Kerr F, Slack C, Jacobson J, et al. 2010. Mechanisms of life span extension by rapamycin in the fruit fly Drosophila melanogaster. Cell Metab 11: 35-46 21. Bland JM, Altman DG. 1996. Transforming data. Bmj 312: 770 22. Bluher M, Kahn BB, Kahn CR. 2003. Extended longevity in mice lacking the insulin receptor in adipose tissue. Science 299: 572-574 23. Boeke JD, LaCroute F, Fink GR. 1984. A positive selection for mutants lacking orotidine-5'-phosphate decarboxylase activity in yeast: 5-fluoro- orotic acid resistance. Mol Gen Genet 197: 345-346 24. Boer VM, Amini S, Botstein D. 2008. Influence of genotype and nutrition on survival and metabolism of starving yeast. Proc Natl Acad Sci U S A 105: 6930-6935 25. Bokov A, Chaudhuri A, Richardson A. 2004. The role of oxidative damage and stress in aging. Mech Ageing Dev 125: 811-826 26. Bonawitz ND, Chatenay-Lapointe M, Pan Y, Shadel GS. 2007. Reduced TOR signaling extends chronological life span via increased respiration and upregulation of mitochondrial gene expression. Cell Metab 5: 265-277 27. Bonawitz ND, Rodeheffer MS, Shadel GS. 2006. Defective mitochondrial gene expression results in reactive oxygen species-mediated inhibition of respiration and reduction of yeast life span. Mol Cell Biol 26: 4818-4829 28. Botstein D, Fink GR. 1988. Yeast: an experimental organism for modern biology. Science 240: 1439-1443 29. Broughton S, Partridge L. 2009. Insulin/IGF-like signalling, the central nervous system and aging. Biochem J 418: 1-12 30. Brummendorf TH, Balabanov S. 2006. Telomere length dynamics in normal hematopoiesis and in disease states characterized by increased stem cell turnover. Leukemia 20: 1706-1716 31. Bryk M, Banerjee M, Murphy M, Knudsen KE, Garfinkel DJ, Curcio MJ. 1997. Transcriptional silencing of Ty1 elements in the RDN1 locus of yeast. Genes Dev 11: 255-269 32. Burns N, Grimwade B, Ross-Macdonald PB, Choi EY, Finberg K, et al. 1994. Large-scale analysis of gene expression, protein localization, and gene disruption in Saccharomyces cerevisiae. Genes Dev 8: 1087-1105 33. Burtner CR, Murakami CJ, Kennedy BK, Kaeberlein M. 2009. A molecular mechanism of chronological aging in yeast. Cell Cycle 8: 1256- 1270
189 34. Cameroni E, Hulo N, Roosen J, Winderickx J, De Virgilio C. 2004. The novel yeast PAS kinase Rim 15 orchestrates G0-associated antioxidant defense mechanisms. Cell Cycle 3: 462-468 35. Campisi J. 2005. Senescent cells, tumor suppression, and organismal aging: good citizens, bad neighbors. Cell 120: 513-522 36. Campisi J, d'Adda di Fagagna F. 2007. Cellular senescence: when bad things happen to good cells. Nat Rev Mol Cell Biol 8: 729-740 37. Campuzano V, Montermini L, Molto MD, Pianese L, Cossee M, et al. 1996. Friedreich's ataxia: autosomal recessive disease caused by an intronic GAA triplet repeat expansion. Science 271: 1423-1427 38. Carroll AS, O'Shea EK. 2002. Pho85 and signaling environmental conditions. Trends Biochem Sci 27: 87-93 39. Chen BR, Runge KW. 2009. A new Schizosaccharomyces pombe chronological lifespan assay reveals that caloric restriction promotes efficient cell cycle exit and extends longevity. Exp Gerontol 44: 493-502 40. Chen D, Steele AD, Lindquist S, Guarente L. 2005. Increase in activity during calorie restriction requires Sirt1. Science 310: 1641 41. Chen JB, Sun J, Jazwinski SM. 1990. Prolongation of the yeast life span by the v-Ha-RAS oncogene. Mol Microbiol 4: 2081-2086 42. Chua G, Taricani L, Stangle W, Young PG. 2000. Insertional mutagenesis based on illegitimate recombination in Schizosaccharomyces pombe. Nucleic Acids Res 28: E53 43. Clancy DJ, Gems D, Harshman LG, Oldham S, Stocker H, et al. 2001. Extension of life-span by loss of CHICO, a Drosophila insulin receptor substrate protein. Science 292: 104-106 44. Clotet J, Posas F. 2007. Control of cell cycle in response to osmostress: lessons from yeast. Methods Enzymol. 428: 63-76 45. Cohen HY, Miller C, Bitterman KJ, Wall NR, Hekking B, et al. 2004. Calorie restriction promotes mammalian cell survival by inducing the SIRT1 deacetylase. Science 305: 390-392 46. Colman RJ, Anderson RM, Johnson SC, Kastman EK, Kosmatka KJ, et al. 2009. Caloric restriction delays disease onset and mortality in rhesus monkeys. Science 325: 201-204 47. Costello G, Rodgers L, Beach D. 1986. Fission yeast enters the stationary phase G0 state from either mitotic G1 or G2. Curr. Genet. 11: 119-125 48. Cuervo AM. 2008. Autophagy and aging: keeping that old broom working. Trends Genet 24: 604-612 49. Cuervo AM, Bergamini E, Brunk UT, Droge W, Ffrench M, Terman A. 2005. Autophagy and aging: the importance of maintaining "clean" cells. Autophagy 1: 131-140 50. Curran SP, Ruvkun G. 2007. Lifespan regulation by evolutionarily conserved genes essential for viability. PLoS Genet 3: e56 51. Dang W, Steffen KK, Perry R, Dorsey JA, Johnson FB, et al. 2009. Histone H4 lysine 16 acetylation regulates cellular lifespan. Nature 459: 802-807 52. Davidson MK, Young NP, Glick GG, Wahls WP. 2004. Meiotic
190 chromosome segregation mutants identified by insertional mutagenesis of fission yeast Schizosaccharomyces pombe; tandem-repeat, single-site integrations. Nucleic Acids Res 32: 4400-4410 53. Decottignies A. 2005. Capture of extranuclear DNA at fission yeast double-strand breaks. Genetics 171: 1535-1548 54. DePinho RA. 2000. The age of cancer. Nature 408: 248-254 55. Deutschbauer AM, Jaramillo DF, Proctor M, Kumm J, Hillenmeyer ME, et al. 2005. Mechanisms of haploinsufficiency revealed by genome-wide profiling in yeast. Genetics 169: 1915-1925 56. Devon RS, Porteous DJ, Brookes AJ. 1995. Splinkerettes--improved vectorettes for greater efficiency in PCR walking. Nucleic Acids Res 23: 1644-1645 57. Diaz-Troya S, Perez-Perez ME, Florencio FJ, Crespo JL. 2008. The role of TOR in autophagy regulation from yeast to plants and mammals. Autophagy 4: 851-865 58. Dillin A, Crawford DK, Kenyon C. 2002. Timing requirements for insulin/IGF-1 signaling in C. elegans. Science 298: 830-834 59. Dilova I, Easlon E, Lin SJ. 2007. Calorie restriction and the nutrient sensing signaling pathways. Cell Mol Life Sci 64: 752-767 60. Doonan R, McElwee JJ, Matthijssens F, Walker GA, Houthoofd K, et al. 2008. Against the oxidative damage theory of aging: superoxide dismutases protect against oxidative stress but have little or no effect on life span in Caenorhabditis elegans. Genes Dev 22: 3236-3241 61. Dujon B. 1981. Mitochondrial genetics and functions. In The Molecular Biology of the Yeast Saccharomyces, ed. JN Strathern, Jones, EW, Broach, JR, pp. 505-635. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press 62. Easlon E, Tsang F, Dilova I, Wang C, Lu SP, et al. 2007. The dihydrolipoamide acetyltransferase is a novel metabolic longevity factor and is required for calorie restriction-mediated life span extension. J Biol Chem 282: 6161-6171 63. Eigenmann JE, Patterson DF, Froesch ER. 1984. Body size parallels insulin-like growth factor I levels but not growth hormone secretory capacity. Acta Endocrinol (Copenh) 106: 448-453 64. Eisenberg T, Knauer H, Schauer A, Buttner S, Ruckenstuhl C, et al. 2009. Induction of autophagy by spermidine promotes longevity. Nat Cell Biol 11: 1305-1314 65. Elchuri S, Oberley TD, Qi W, Eisenstein RS, Jackson Roberts L, et al. 2005. CuZnSOD deficiency leads to persistent and widespread oxidative damage and hepatocarcinogenesis later in life. Oncogene 24: 367-380 66. Erjavec N, Cvijovic M, Klipp E, Nystrom T. 2008. Selective benefits of damage partitioning in unicellular systems and its effects on aging. Proc Natl Acad Sci U S A 105: 18764-18769 67. Erjavec N, Nystrom T. 2007. Sir2p-dependent protein segregation gives rise to a superior reactive oxygen species management in the progeny of Saccharomyces cerevisiae. Proc Natl Acad Sci U S A 104: 10877-10881
191 68. Espinoza FH, Ogas J, Herskowitz I, Morgan DO. 1994. Cell cycle control by a complex of the cyclin HCS26 (PCL1) and the kinase PHO85. Science 266: 1388-1391 69. Evertts AG, Plymire C, Craig NL, Levin HL. 2007. The hermes transposon of Musca domestica is an efficient tool for the mutagenesis of Schizosaccharomyces pombe. Genetics 177: 2519-2523 70. Fabrizio P, Battistella L, Vardavas R, Gattazzo C, Liou LL, et al. 2004. Superoxide is a mediator of an altruistic aging program in Saccharomyces cerevisiae. J. Cell Biol. 166: 1055-1067 71. Fabrizio P, Gattazzo C, Battistella L, Wei M, Cheng C, et al. 2005. Sir2 blocks extreme life-span extension. Cell 123: 655-667 72. Fabrizio P, Hoon S, Shamalnasab M, Galbani A, Wei M, et al. 2010. Genome-wide screen in Saccharomyces cerevisiae identifies vacuolar protein sorting, autophagy, biosynthetic, and tRNA methylation genes involved in life span regulation. PLoS Genet 6: e1001024. doi:1001010.1001371/journal.pgen.1001024 73. Fabrizio P, Liou LL, Moy VN, Diaspro A, Valentine JS, et al. 2003. SOD2 functions downstream of Sch9 to extend longevity in yeast. Genetics 163: 35-46 74. Fabrizio P, Longo VD. 2003. The chronological life span of Saccharomyces cerevisiae. Aging Cell 2: 73-81 75. Fabrizio P, Longo VD. 2008. Chronological aging-induced apoptosis in yeast. Biochim. Biophys. Acta 1783: 1280-1285 76. Fabrizio P, Pletcher SD, Minois N, Vaupel JW, Longo VD. 2004. Chronological aging-independent replicative life span regulation by Msn2/Msn4 and Sod2 in Saccharomyces cerevisiae. Febs Lett 557: 136- 142 77. Fabrizio P, Pozza F, Pletcher SD, Gendron CM, Longo VD. 2001. Regulation of longevity and stress resistance by Sch9 in yeast. Science 292: 288-290 78. Finkel T, Holbrook NJ. 2000. Oxidants, oxidative stress and the biology of ageing. Nature 408: 239-247 79. Firestein R, Blander G, Michan S, Oberdoerffer P, Ogino S, et al. 2008. The SIRT1 deacetylase suppresses intestinal tumorigenesis and colon cancer growth. PLoS ONE 3: e2020 80. Fontana L, Klein S. 2007. Aging, adiposity, and calorie restriction. Jama 297: 986-994 81. Fontana L, Partridge L, Longo VD. 2010. Extending healthy life span-- from yeast to humans. Science 328: 321-326 82. Forsburg Lab webpage. http://www-bcf.usc.edu/~forsburg/drugs.html#FOA. 83. Forsburg SL. 2001. The art and design of genetic screens: yeast. Nat Rev Genet 2: 659-668 84. Forsburg SL. 2005. The yeasts Saccharomyces cerevisiae and Schizosaccharomyces pombe: models for cell biology research. Gravit Space Biol Bull 18: 3-9 85. Forsburg SL, Rhind N. 2006. Basic methods for fission yeast. Yeast 23:
192 173-183 86. Fredickson AG, Stephanopoulos G. 1981. Microbial competition. Science 213: 972-979 87. Friedman DB, Johnson TE. 1988. A mutation in the age-1 gene in Caenorhabditis elegans lengthens life and reduces hermaphrodite fertility. Genetics 118: 75-86 88. Fujita Y, Mita S, Ohtsuka H, Aiba H. 2007. Identification of a fatty acyl- CoA synthetase gene, lcf2+, which affects viability after entry into the stationary phase in Schizosaccharomyces pombe. Biosci Biotechnol Biochem 71: 3041-3047 89. Gao CY, Zelenka PS. 1997. Cyclins, cyclin-dependent kinases and differentiation. Bioessays 19: 307-315 90. Gems D, Partridge L. 2008. Stress-response hormesis and aging: "that which does not kill us makes us stronger". Cell Metab 7: 200-203 91. Giaever G, Chu AM, Ni L, Connelly C, Riles L, et al. 2002. Functional profiling of the Saccharomyces cerevisiae genome. Nature 418: 387-391 92. Giaever G, Shoemaker DD, Jones TW, Liang H, Winzeler EA, et al. 1999. Genomic profiling of drug sensitivities via induced haploinsufficiency. Nat Genet 21: 278-283 93. Goedecke W, Pfeiffer P, Vielmetter W. 1994. Nonhomologous DNA end joining in Schizosaccharomyces pombe efficiently eliminates DNA double-strand-breaks from haploid sequences. Nucleic Acids Res 22: 2094-2101 94. Goffeau A, Barrell BG, Bussey H, Davis RW, Dujon B, et al. 1996. Life with 6000 genes. Science 274: 546, 563-547 95. Goldberg AA, Bourque SD, Kyryakov P, Gregg C, Boukh-Viner T, et al. 2009. Effect of calorie restriction on the metabolic history of chronologically aging yeast. Exp Gerontol 44: 555-571 96. Gottlieb S, Esposito RE. 1989. A new role for a yeast transcriptional silencer gene, SIR2, in regulation of recombination in ribosomal DNA. Cell 56: 771-776 97. Gottschal JC. 1993. Growth kinetics and competition--some contemporary comments. Antonie Van Leeuwenhoek 63: 299-313 98. Gottschling DE, Aparicio OM, Billington BL, Zakian VA. 1990. Position effect at S. cerevisiae telomeres: reversible repression of Pol II transcription. Cell 63: 751-762 99. Graham BH, Waymire KG, Cottrell B, Trounce IA, MacGregor GR, Wallace DC. 1997. A mouse model for mitochondrial myopathy and cardiomyopathy resulting from a deficiency in the heart/muscle isoform of the adenine nucleotide translocator. Nat Genet 16: 226-234 100. Granot D, Snyder M. 1991. Glucose induces cAMP-independent growth- related changes in stationary-phase cells of Saccharomyces cerevisiae. Proc Natl Acad Sci U S A 88: 5724-5728 101. Granot D, Snyder M. 1993. Carbon source induces growth of stationary phase yeast cells, independent of carbon source metabolism. Yeast 9: 465- 479
193 102. Gray JV, Petsko GA, Johnston GC, Ringe D, Singer RA, Werner- Washburne M. 2004. "Sleeping beauty": quiescence in Saccharomyces cerevisiae. Microbiol. Mol. Biol. Rev. 68: 187-206 103. Greene JM, Larin Z, Taylor IC, Prentice H, Gwinn KA, Kingston RE. 1987. Multiple basal elements of a human hsp70 promoter function differently in human and rodent cell lines. Mol Cell Biol 7: 3646-3655 104. Greer EL, Brunet A. 2009. Different dietary restriction regimens extend lifespan by both independent and overlapping genetic pathways in C. elegans. Aging Cell 8: 113-127 105. Greer KA, Canterberry SC, Murphy KE. 2007. Statistical analysis regarding the effects of height and weight on life span of the domestic dog. Res Vet Sci 82: 208-214 106. Grimm C, Kohli J. 1988. Observations on integrative transformation in Schizosaccharomyces pombe. Mol Gen Genet 215: 87-93 107. Grimm C, Kohli J, Murray J, Maundrell K. 1988. Genetic engineering of Schizosaccharomyces pombe: a system for gene disruption and replacement using the ura4 gene as a selectable marker. Mol Gen Genet 215: 81-86 108. Guarente L. 2005. Calorie restriction and SIR2 genes--towards a mechanism. Mech. Ageing Dev. 126: 923-928 109. Guarente L. 2008. Mitochondria--a nexus for aging, calorie restriction, and sirtuins? Cell 132: 171-176 110. Gutteridge JM, Halliwell B. 1990. The measurement and mechanism of lipid peroxidation in biological systems. Trends Biochem Sci 15: 129-135 111. Haigis MC, Guarente LP. 2006. Mammalian sirtuins--emerging roles in physiology, aging, and calorie restriction. Genes Dev 20: 2913-2921 112. Haigis MC, Sinclair DA. 2010. Mammalian sirtuins: biological insights and disease relevance. Annu Rev Pathol 5: 253-295 113. Hamilton B, Dong Y, Shindo M, Liu W, Odell I, et al. 2005. A systematic RNAi screen for longevity genes in C. elegans. Genes Dev 19: 1544-1555 114. Han ES, Hickey M. 2005. Microarray evaluation of dietary restriction. J Nutr 135: 1343-1346 115. Hands SL, Proud CG, Wyttenbach A. 2009. mTOR's role in ageing: protein synthesis or autophagy? Aging (Albany NY) 1: 586-597 116. Hansen M, Chandra A, Mitic LL, Onken B, Driscoll M, Kenyon C. 2008. A role for autophagy in the extension of lifespan by dietary restriction in C. elegans. PLoS Genet 4: e24 117. Hansen M, Taubert S, Crawford D, Libina N, Lee SJ, Kenyon C. 2007. Lifespan extension by conditions that inhibit translation in Caenorhabditis elegans. Aging Cell 6: 95-110 118. Harman D. 1956. Aging: a theory based on free radical and radiation chemistry. J Gerontol 11: 298-300 119. Harrison DE, Strong R, Sharp ZD, Nelson JF, Astle CM, et al. 2009. Rapamycin fed late in life extends lifespan in genetically heterogeneous mice. Nature 460: 392-395 120. Hartwell LH. 2002. Nobel Lecture. Yeast and cancer. Biosci Rep 22: 373-
194 394 121. Hayashi T, Hatanaka M, Nagao K, Nakaseko Y, Kanoh J, et al. 2007. Rapamycin sensitivity of the Schizosaccharomyces pombe tor2 mutant and organization of two highly phosphorylated TOR complexes by specific and common subunits. Genes Cells 12: 1357-1370 122. Hayflick L, Moorhead PS. 1961. The serial cultivation of human diploid cell strains. Exp Cell Res 25: 585-621 123. Hayles J, Nurse P. 1992. Genetics of the fission yeast Schizosaccharomyces pombe. Annu Rev Genet 26: 373-402 124. Hedges SB. 2002. The origin and evolution of model organisms. Nat Rev Genet 3: 838-849 125. Heitman J, Movva NR, Hall MN. 1991. Targets for cell cycle arrest by the immunosuppressant rapamycin in yeast. Science 253: 905-909 126. Hensel M, Shea JE, Gleeson C, Jones MD, Dalton E, Holden DW. 1995. Simultaneous identification of bacterial virulence genes by negative selection. Science 269: 400-403 127. Herker E, Jungwirth H, Lehmann KA, Maldener C, Frohlich KU, et al. 2004. Chronological aging leads to apoptosis in yeast. J Cell Biol 164: 501-507 128. Heydari AR, Unnikrishnan A, Lucente LV, Richardson A. 2007. Caloric restriction and genomic stability. Nucleic Acids Res 35: 7485-7496 129. Hinnebusch AG. 2005. Translational regulation of GCN4 and the general amino acid control of yeast. Annu Rev Microbiol 59: 407-450 130. Hinnen A, Hicks JB, Fink GR. 1978. Transformation of yeast. Proc Natl Acad Sci U S A 75: 1929-1933 131. Hirashima K, Iwaki T, Takegawa K, Giga-Hama Y, Tohda H. 2006. A simple and effective chromosome modification method for large-scale deletion of genome sequences and identification of essential genes in fission yeast. Nucleic Acids Res 34: e11 132. Hoffman CS. 2005. Glucose sensing via the protein kinase A pathway in Schizosaccharomyces pombe. Biochem. Soc. Trans. 33: 257-260 133. Hoffman CS, Winston F. 1990. Isolation and characterization of mutants constitutive for expression of the fbp1 gene of Schizosaccharomyces pombe. Genetics 124: 807-816 134. Howitz KT, Bitterman KJ, Cohen HY, Lamming DW, Lavu S, et al. 2003. Small molecule activators of sirtuins extend Saccharomyces cerevisiae lifespan. Nature 425: 191-196 135. Huang D, Friesen H, Andrews B. 2007. Pho85, a multifunctional cyclin- dependent protein kinase in budding yeast. Mol Microbiol 66: 303-314 136. Huang D, Patrick G, Moffat J, Tsai LH, Andrews B. 1999. Mammalian Cdk5 is a functional homologue of the budding yeast Pho85 cyclin- dependent protein kinase. Proc Natl Acad Sci U S A 96: 14445-14450 137. Hunt T. 2002. Nobel Lecture. Protein synthesis, proteolysis, and cell cycle transitions. Biosci Rep 22: 465-486 138. Hwangbo DS, Gershman B, Tu MP, Palmer M, Tatar M. 2004. Drosophila dFOXO controls lifespan and regulates insulin signalling in brain and fat
195 body. Nature 429: 562-566 139. Ikeda K, Morigasaki S, Tatebe H, Tamanoi F, Shiozaki K. 2008. Fission yeast TOR complex 2 activates the AGC-family Gad8 kinase essential for stress resistance and cell cycle control. Cell Cycle 7: 358-364 140. Ishii N, Fujii M, Hartman PS, Tsuda M, Yasuda K, et al. 1998. A mutation in succinate dehydrogenase cytochrome b causes oxidative stress and ageing in nematodes. Nature 394: 694-697 141. Jacinto E, Loewith R, Schmidt A, Lin S, Ruegg MA, et al. 2004. Mammalian TOR complex 2 controls the actin cytoskeleton and is rapamycin insensitive. Nat Cell Biol 6: 1122-1128 142. Jang YC, Perez VI, Song W, Lustgarten MS, Salmon AB, et al. 2009. Overexpression of Mn superoxide dismutase does not increase life span in mice. J Gerontol A Biol Sci Med Sci 64: 1114-1125 143. Jarolim S, Millen J, Heeren G, Laun P, Goldfarb DS, Breitenbach M. 2004. A novel assay for replicative lifespan in Saccharomyces cerevisiae. FEMS Yeast Res 5: 169-177 144. Jia K, Chen D, Riddle DL. 2004. The TOR pathway interacts with the insulin signaling pathway to regulate C. elegans larval development, metabolism and life span. Development 131: 3897-3906 145. Jia K, Levine B. 2007. Autophagy is required for dietary restriction- mediated life span extension in C. elegans. Autophagy 3: 597-599 146. Jiang JC, Wawryn J, Shantha Kumara HM, Jazwinski SM. 2002. Distinct roles of processes modulated by histone deacetylases Rpd3p, Hda1p, and Sir2p in life extension by caloric restriction in yeast. Exp Gerontol 37: 1023-1030 147. Johnson TE, Hartman PS. 1988. Radiation effects on life span in Caenorhabditis elegans. J Gerontol 43: B137-141 148. Jones W. 1884. Longevity in a fasting spider. Science 3: 4 149. Jung CH, Ro SH, Cao J, Otto NM, Kim DH. 2010. mTOR regulation of autophagy. Febs Lett 584: 1287-1295 150. Kaeberlein M. 2009. Spermidine surprise for a long life. Nat Cell Biol 11: 1277-1278 151. Kaeberlein M. 2010. Lessons on longevity from budding yeast. Nature 464: 513-519 152. Kaeberlein M. 2010. Resveratrol and rapamycin: are they anti-aging drugs? Bioessays 32: 96-99 153. Kaeberlein M, Andalis AA, Fink GR, Guarente L. 2002. High osmolarity extends life span in Saccharomyces cerevisiae by a mechanism related to calorie restriction. Mol Cell Biol 22: 8056-8066 154. Kaeberlein M, Hu D, Kerr EO, Tsuchiya M, Westman EA, et al. 2005. Increased life span due to calorie restriction in respiratory-deficient yeast. PLoS Genet 1: e69 155. Kaeberlein M, Kennedy BK. 2007. Protein translation, 2007. Aging Cell 6: 731-734 156. Kaeberlein M, Kennedy BK. 2008. Protein translation, 2008. Aging Cell 7: 777-782
196 157. Kaeberlein M, Kirkland KT, Fields S, Kennedy BK. 2004. Sir2- independent life span extension by calorie restriction in yeast. PLoS Biol 2: E296 158. Kaeberlein M, McVey M, Guarente L. 1999. The SIR2/3/4 complex and SIR2 alone promote longevity in Saccharomyces cerevisiae by two different mechanisms. Genes Dev 13: 2570-2580 159. Kaeberlein M, Powers RW, 3rd. 2007. Sir2 and calorie restriction in yeast: a skeptical perspective. Ageing Res Rev 6: 128-140 160. Kaeberlein M, Powers RW, 3rd, Steffen KK, Westman EA, Hu D, et al. 2005. Regulation of yeast replicative life span by TOR and Sch9 in response to nutrients. Science 310: 1193-1196 161. Kaeberlein TL, Smith ED, Tsuchiya M, Welton KL, Thomas JH, et al. 2006. Lifespan extension in Caenorhabditis elegans by complete removal of food. Aging Cell 5: 487-494 162. Kaffman A, Herskowitz I, Tjian R, O'Shea EK. 1994. Phosphorylation of the transcription factor PHO4 by a cyclin-CDK complex, PHO80-PHO85. Science 263: 1153-1156 163. Kanfi Y, Shalman R, Peshti V, Pilosof SN, Gozlan YM, et al. 2008. Regulation of SIRT6 protein levels by nutrient availability. Febs Lett 582: 543-548 164. Kapahi P, Chen D, Rogers AN, Katewa SD, Li PW, et al. 2010. With TOR, less is more: a key role for the conserved nutrient-sensing TOR pathway in aging. Cell Metab 11: 453-465 165. Kapahi P, Zid BM, Harper T, Koslover D, Sapin V, Benzer S. 2004. Regulation of lifespan in Drosophila by modulation of genes in the TOR signaling pathway. Curr Biol 14: 885-890 166. Kaufer NF, Potashkin J. 2000. Analysis of the splicing machinery in fission yeast: a comparison with budding yeast and mammals. Nucleic Acids Res 28: 3003-3010 167. Kaufer NF, Simanis V, Nurse P. 1985. Fission yeast Schizosaccharomyces pombe correctly excises a mammalian RNA transcript intervening sequence. Nature 318: 78-80 168. Kawahara TL, Michishita E, Adler AS, Damian M, Berber E, et al. 2009. SIRT6 links histone H3 lysine 9 deacetylation to NF-kappaB-dependent gene expression and organismal life span. Cell 136: 62-74 169. KEGG. http://www.genome.jp/kegg/. 170. Kennedy BK, Austriaco NR, Jr., Zhang J, Guarente L. 1995. Mutation in the silencing gene SIR4 can delay aging in S. cerevisiae. Cell 80: 485-496 171. Kennedy BK, Gotta M, Sinclair DA, Mills K, McNabb DS, et al. 1997. Redistribution of silencing proteins from telomeres to the nucleolus is associated with extension of life span in S. cerevisiae. Cell 89: 381-391 172. Kennedy BK, Kaeberlein M. 2009. Hot topics in aging research: protein translation, 2009. Aging Cell 8: 617-623 173. Kenyon C. 2005. The plasticity of aging: insights from long-lived mutants. Cell 120: 449-460 174. Kenyon C, Chang J, Gensch E, Rudner A, Tabtiang R. 1993. A C. elegans
197 mutant that lives twice as long as wild type. Nature 366: 461-464 175. Kenyon CJ. 2010. The genetics of ageing. Nature 464: 504-512 176. Kim DU, Hayles J, Kim D, Wood V, Park HO, et al. 2010. Analysis of a genome-wide set of gene deletions in the fission yeast Schizosaccharomyces pombe. Nat Biotechnol 177. Kim G, Sikder H, Singh KK. 2002. A colony color method identifies the vulnerability of mitochondria to oxidative damage. Mutagenesis 17: 375- 381 178. Kimura KD, Tissenbaum HA, Liu Y, Ruvkun G. 1997. daf-2, an insulin receptor-like gene that regulates longevity and diapause in Caenorhabditis elegans. Science 277: 942-946 179. Kirby K, Hu J, Hilliker AJ, Phillips JP. 2002. RNA interference-mediated silencing of Sod2 in Drosophila leads to early adult-onset mortality and elevated endogenous oxidative stress. Proc Natl Acad Sci U S A 99: 16162-16167 180. Kirkwood TB. 2005. Understanding the odd science of aging. Cell 120: 437-447 181. Kirkwood TB, Austad SN. 2000. Why do we age? Nature 408: 233-238 182. Kosoy A, Calonge TM, Outwin EA, O'Connell MJ. 2007. Fission yeast Rnf4 homologs are required for DNA repair. J Biol Chem 282: 20388- 20394 183. Krishnan KJ, Reeve AK, Samuels DC, Chinnery PF, Blackwood JK, et al. 2008. What causes mitochondrial DNA deletions in human cells? Nat Genet 40: 275-279 184. Kuranda MJ, Robbins PW. 1987. Cloning and heterologous expression of glycosidase genes from Saccharomyces cerevisiae. Proc. Natl. Acad. Sci. U S A 84: 2585-2589 185. Kurzhals RL, Tie F, Stratton CA, Harte PJ. 2008. Drosophila ESC-like can substitute for ESC and becomes required for Polycomb silencing if ESC is absent. Dev. Biol. 313: 293-306 186. Kuspa A, Loomis WF. 1992. Tagging developmental genes in Dictyostelium by restriction enzyme-mediated integration of plasmid DNA. Proc Natl Acad Sci U S A 89: 8803-8807 187. Lagouge M, Argmann C, Gerhart-Hines Z, Meziane H, Lerin C, et al. 2006. Resveratrol improves mitochondrial function and protects against metabolic disease by activating SIRT1 and PGC-1alpha. Cell 127: 1109- 1122 188. Landis GN, Tower J. 2005. Superoxide dismutase evolution and life span regulation. Mech Ageing Dev 126: 365-379 189. Lebovitz RM, Zhang H, Vogel H, Cartwright J, Jr., Dionne L, et al. 1996. Neurodegeneration, myocardial injury, and perinatal death in mitochondrial superoxide dismutase-deficient mice. Proc Natl Acad Sci U S A 93: 9782-9787 190. Lee GD, Wilson MA, Zhu M, Wolkow CA, de Cabo R, et al. 2006. Dietary deprivation extends lifespan in Caenorhabditis elegans. Aging Cell 5: 515-524
198 191. Lee H-G, Casadesus G, Zhu X, Castellani RJ, McShea A, et al. 2009. Cell cycle re-entry mediated neurodegeneration and its treatment role in the pathogenesis of Alzheimer's disease. Neurochem. Int. doi: 10.1016/j.neuint.2008.10.013 192. Lee SS, Kennedy S, Tolonen AC, Ruvkun G. 2003. DAF-16 target genes that control C. elegans life-span and metabolism. Science 300: 644-647 193. Lee SS, Lee RY, Fraser AG, Kamath RS, Ahringer J, Ruvkun G. 2003. A systematic RNAi screen identifies a critical role for mitochondria in C. elegans longevity. Nat Genet 33: 40-48 194. Lepperdinger G, Berger P, Breitenbach M, Frohlich KU, Grillari J, et al. 2008. The use of genetically engineered model systems for research on human aging. Front Biosci 13: 7022-7031 195. Li Y, Huang TT, Carlson EJ, Melov S, Ursell PC, et al. 1995. Dilated cardiomyopathy and neonatal lethality in mutant mice lacking manganese superoxide dismutase. Nat Genet 11: 376-381 196. Liang P, Pardee AB. 1992. Differential display of eukaryotic messenger RNA by means of the polymerase chain reaction. Science 257: 967-971 197. Lin K, Dorman JB, Rodan A, Kenyon C. 1997. daf-16: An HNF- 3/forkhead family member that can function to double the life-span of Caenorhabditis elegans. Science 278: 1319-1322 198. Lin K, Hsin H, Libina N, Kenyon C. 2001. Regulation of the Caenorhabditis elegans longevity protein DAF-16 by insulin/IGF-1 and germline signaling. Nat Genet 28: 139-145 199. Lin SJ, Defossez PA, Guarente L. 2000. Requirement of NAD and SIR2 for life-span extension by calorie restriction in Saccharomyces cerevisiae. Science 289: 2126-2128 200. Lin SJ, Ford E, Haigis M, Liszt G, Guarente L. 2004. Calorie restriction extends yeast life span by lowering the level of NADH. Genes Dev 18: 12- 16 201. Lin SJ, Guarente L. 2006. Increased life span due to calorie restriction in respiratory-deficient yeast. PLoS Genet 2: e33; author reply e34 202. Lin SJ, Kaeberlein M, Andalis AA, Sturtz LA, Defossez PA, et al. 2002. Calorie restriction extends Saccharomyces cerevisiae lifespan by increasing respiration. Nature 418: 344-348 203. Lin YJ, Seroude L, Benzer S. 1998. Extended life-span and stress resistance in the Drosophila mutant methuselah. Science 282: 943-946 204. Lindstrom DL, Gottschling DE. 2009. The mother enrichment program: a genetic system for facile replicative life span analysis in Saccharomyces cerevisiae. Genetics 183: 413-422, 411SI-413SI 205. Liu B, Larsson L, Caballero A, Hao X, Oling D, et al. 2010. The polarisome is required for segregation and retrograde transport of protein aggregates. Cell 140: 257-267 206. Liu YG, Whittier RF. 1995. Thermal asymmetric interlaced PCR: automatable amplification and sequencing of insert end fragments from P1 and YAC clones for chromosome walking. Genomics 25: 674-681 207. Lodish H, Berk A, Zipursky SL, Matsudaira P, Kaiser CA, et al. 2004.
199 Molecular Cell Biology. New York: W. H. Freeman & Co 208. Loewith R, Jacinto E, Wullschleger S, Lorberg A, Crespo JL, et al. 2002. Two TOR complexes, only one of which is rapamycin sensitive, have distinct roles in cell growth control. Mol Cell 10: 457-468 209. Longo VD, Gralla EB, Valentine JS. 1996. Superoxide dismutase activity is essential for stationary phase survival in Saccharomyces cerevisiae. Mitochondrial production of toxic oxygen species in vivo. J Biol Chem 271: 12275-12280 210. Longo VD, Kennedy BK. 2006. Sirtuins in aging and age-related disease. Cell 126: 257-268 211. Lopez-Lluch G, Hunt N, Jones B, Zhu M, Jamieson H, et al. 2006. Calorie restriction induces mitochondrial biogenesis and bioenergetic efficiency. Proc Natl Acad Sci U S A 103: 1768-1773 212. Ma XM, Blenis J. 2009. Molecular mechanisms of mTOR-mediated translational control. Nat Rev Mol Cell Biol 10: 307-318 213. MacLean M, Harris N, Piper PW. 2001. Chronological lifespan of stationary phase yeast cells; a model for investigating the factors that might influence the ageing of postmitotic tissues in higher organisms. Yeast 18: 499-509 214. Maclean MJ, Aamodt R, Harris N, Alseth I, Seeberg E, et al. 2003. Base excision repair activities required for yeast to attain a full chronological life span. Aging Cell 2: 93-104 215. Manivasakam P, Schiestl RH. 1998. Nonhomologous end joining during restriction enzyme-mediated DNA integration in Saccharomyces cerevisiae. Mol Cell Biol 18: 1736-1745 216. Martin-Castellanos C, Blanco M, Rozalen AE, Perez-Hidalgo L, Garcia AI, et al. 2005. A large-scale screen in S. pombe identifies seven novel genes required for critical meiotic events. Curr Biol 15: 2056-2062 217. Masoro EJ. 1998. Influence of caloric intake on aging and on the response to stressors. J Toxicol Environ Health B Crit Rev 1: 243-257 218. Masoro EJ. 2005. Overview of caloric restriction and ageing. Mech. Ageing Dev. 126: 913-922 219. Masoro EJ. 2007. The role of hormesis in life extension by dietary restriction. Interdiscip. Top. Gerontol. 35: 1-17 220. Mata J, Lyne R, Burns G, Bahler J. 2002. The transcriptional program of meiosis and sporulation in fission yeast. Nat Genet 32: 143-147 221. Matecic M, Smith DL, Pan X, Maqani N, Bekiranov S, et al. 2010. A microarray-based genetic screen for yeast chronological aging factors. PLoS Genet 6: e1000921 222. Matsuo T, Kubo Y, Watanabe Y, Yamamoto M. 2003. Schizosaccharomyces pombe AGC family kinase Gad8p forms a conserved signaling module with TOR and PDK1-like kinases. EMBO J 22: 3073-3083 223. Matsuo T, Otsubo Y, Urano J, Tamanoi F, Yamamoto M. 2007. Loss of the TOR kinase Tor2 mimics nitrogen starvation and activates the sexual development pathway in fission yeast. Mol Cell Biol 27: 3154-3164
200 224. McCay CM, Crowell MF, Maynard LA. 1935. The effect of retarded growth upon the length of life span and upon the ultimate body size. J. Nutr. 10: 63-79 225. Measday V, Moore L, Ogas J, Tyers M, Andrews B. 1994. The PCL2 (ORFD)-PHO85 cyclin-dependent kinase complex: a cell cycle regulator in yeast. Science 266: 1391-1395 226. Measday V, Moore L, Retnakaran R, Lee J, Donoviel M, et al. 1997. A family of cyclin-like proteins that interact with the Pho85 cyclin- dependent kinase. Mol Cell Biol 17: 1212-1223 227. Melov S, Lithgow GJ, Fischer DR, Tedesco PM, Johnson TE. 1995. Increased frequency of deletions in the mitochondrial genome with age of Caenorhabditis elegans. Nucleic Acids Res 23: 1419-1425 228. Merry BJ. 2004. Oxidative stress and mitochondrial function with aging-- the effects of calorie restriction. Aging Cell 3: 7-12 229. Metcalfe NB, Monaghan P. 2003. Growth versus lifespan: perspectives from evolutionary ecology. Exp. Gerontol. 38: 935-940 230. Metzker ML. 2010. Sequencing technologies - the next generation. Nat Rev Genet 11: 31-46 231. Michishita E, McCord RA, Berber E, Kioi M, Padilla-Nash H, et al. 2008. SIRT6 is a histone H3 lysine 9 deacetylase that modulates telomeric chromatin. Nature 452: 492-496 232. Millson SH, Prodromou C, Piper PW. 2010. A simple yeast-based system for analyzing inhibitor resistance in the human cancer drug targets Hsp90alpha/beta. Biochem Pharmacol 79: 1581-1588 233. Milne JC, Lambert PD, Schenk S, Carney DP, Smith JJ, et al. 2007. Small molecule activators of SIRT1 as therapeutics for the treatment of type 2 diabetes. Nature 450: 712-716 234. Minois N, Frajnt M, Dolling M, Lagona F, Schmid M, et al. 2006. Symmetrically dividing cells of the fission yeast schizosaccharomyces pombe do age. Biogerontology 7: 261-267 235. Mochida S, Yanagida M. 2006. Distinct modes of DNA damage response in S. pombe G0 and vegetative cells. Genes Cells 11: 13-27 236. Morck C, Pilon M. 2006. C. elegans feeding defective mutants have shorter body lengths and increased autophagy. BMC Dev Biol 6: 39 237. Moreno S, Klar A, Nurse P. 1991. Molecular Biology of the Fission Yeast Schizosacchromyces pombe. Meth.Enzym. 194: 795-823 238. Morselli E, Galluzzi L, Kepp O, Criollo A, Maiuri MC, et al. 2009. Autophagy mediates pharmacological lifespan extension by spermidine and resveratrol. Aging (Albany NY) 1: 961-970 239. Mortimer RK, Johnston JR. 1959. Life span of individual yeast cells. Nature 183: 1751-1752 240. Moskovitz J, Yim MB, Chock PB. 2002. Free radicals and disease. Arch Biochem Biophys 397: 354-359 241. Motulsky HJ. 2003. Prism 4 Statistics Guide – Statistical analysis for laboratory and clinical researchers. San Diego, CA: GraphPad Software, Inc.
201 242. Muhlrad D, Parker R. 1999. Aberrant mRNAs with extended 3' UTRs are substrates for rapid degradation by mRNA surveillance. Rna 5: 1299-1307 243. Mukhopadhyay A, Oh SW, Tissenbaum HA. 2006. Worming pathways to and from DAF-16/FOXO. Exp. Gerontol. 41: 928-934 244. Munoz MJ, Riddle DL. 2003. Positive selection of Caenorhabditis elegans mutants with increased stress resistance and longevity. Genetics 163: 171- 180 245. Murakami CJ, Burtner CR, Kennedy BK, Kaeberlein M. 2008. A method for high-throughput quantitative analysis of yeast chronological life span. J Gerontol A Biol Sci Med Sci 63: 113-121 246. Murphy CT, McCarroll SA, Bargmann CI, Fraser A, Kamath RS, et al. 2003. Genes that act downstream of DAF-16 to influence the lifespan of Caenorhabditis elegans. Nature 424: 277-283 247. Mutoh N, Kitajima S. 2007. Accelerated chronological aging of a mutant fission yeast deficient in both glutathione and superoxide dismutase having Cu and Zn as cofactors and its enhancement by sir2 deficiency. Biosci Biotechnol Biochem 71: 2841-2844 248. Nakaseko Y, Adachi Y, Funahashi S, Niwa O, Yanagida M. 1986. Chromosome walking shows a highly homologous repetitive sequence present in all the centromere regions of fission yeast. EMBO J. 5: 1011- 1021 249. Narasimhan SD, Yen K, Tissenbaum HA. 2009. Converging pathways in lifespan regulation. Curr Biol 19: R657-666 250. Ngo VN, Davis RE, Lamy L, Yu X, Zhao H, et al. 2006. A loss-of- function RNA interference screen for molecular targets in cancer. Nature 441: 106-110 251. Nishizawa M, Kanaya Y, Toh EA. 1999. Mouse cyclin-dependent kinase (Cdk) 5 is a functional homologue of a yeast Cdk, pho85 kinase. J Biol Chem 274: 33859-33862 252. Nisoli E, Tonello C, Cardile A, Cozzi V, Bracale R, et al. 2005. Calorie restriction promotes mitochondrial biogenesis by inducing the expression of eNOS. Science 310: 314-317 253. Niwa O, Matsumoto T, Chikashige Y, Yanagida M. 1989. Characterization of Schizosaccharomyces pombe minichromosome deletion derivatives and a functional allocation of their centromere. EMBO J. 8: 3045-3052 254. Nunomura A, Chiba S, Kosaka K, Takeda A, Castellani RJ, et al. 2002. Neuronal RNA oxidation is a prominent feature of dementia with Lewy bodies. Neuroreport 13: 2035-2039 255. Nunomura A, Chiba S, Lippa CF, Cras P, Kalaria RN, et al. 2004. Neuronal RNA oxidation is a prominent feature of familial Alzheimer's disease. Neurobiol Dis 17: 108-113 256. Nurse PM. 2002. Nobel Lecture. Cyclin dependent kinases and cell cycle control. Biosci Rep 22: 487-499 257. Ogawa N, Oshima Y. 1990. Functional domains of a positive regulatory protein, PHO4, for transcriptional control of the phosphatase regulon in
202 Saccharomyces cerevisiae. Mol Cell Biol 10: 2224-2236 258. Ohtsuka H, Mita S, Ogawa Y, Azuma K, Ito H, Aiba H. 2008. A novel gene, ecl1(+), extends the chronological lifespan in fission yeast. FEMS Yeast Res. 8: 520-530 259. Ohtsuka H, Mita S, Ogawa Y, Azuma K, Ito H, Aiba H. 2008. A novel gene, ecl1(+), extends the chronological lifespan in fission yeast. FEMS Yeast Res 8: 520-530 260. Ohtsuka H, Ogawa Y, Mizuno H, Mita S, Aiba H. 2009. Identification of Ecl family genes that extend chronological lifespan in fission yeast. Biosci Biotechnol Biochem 73: 885-889 261. Oliver CN, Ahn BW, Moerman EJ, Goldstein S, Stadtman ER. 1987. Age- related changes in oxidized proteins. J Biol Chem 262: 5488-5491 262. Ooi SL, Shoemaker DD, Boeke JD. 2003. DNA helicase gene interaction network defined using synthetic lethality analyzed by microarray. Nat Genet 35: 277-286 263. Orr WC, Sohal RS. 1994. Extension of life-span by overexpression of superoxide dismutase and catalase in Drosophila melanogaster. Science 263: 1128-1130 264. Oshiro T, Aiba H, Mizuno T. 2003. A defect in a fatty acyl-CoA synthetase gene, lcf1+, results in a decrease in viability after entry into the stationary phase in fission yeast. Mol Genet Genomics 269: 437-442 265. Outeiro TF, Muchowski PJ. 2004. Molecular genetics approaches in yeast to study amyloid diseases. J Mol Neurosci 23: 49-60 266. Pamplona R. 2008. Membrane phospholipids, lipoxidative damage and molecular integrity: a causal role in aging and longevity. Biochim Biophys Acta 1777: 1249-1262 267. Pan X, Yuan DS, Xiang D, Wang X, Sookhai-Mahadeo S, et al. 2004. A robust toolkit for functional profiling of the yeast genome. Mol Cell 16: 487-496 268. Paradis S, Ruvkun G. 1998. Caenorhabditis elegans Akt/PKB transduces insulin receptor-like signals from AGE-1 PI3 kinase to the DAF-16 transcription factor. Genes Dev. 12: 2488-2498 269. Parrella E, Longo VD. 2010. Insulin/IGF-I and related signaling pathways regulate aging in nondividing cells: from yeast to the mammalian brain. ScientificWorldJournal 10: 161-177 270. Pedruzzi I, Dubouloz F, Cameroni E, Wanke V, Roosen J, et al. 2003. TOR and PKA signaling pathways converge on the protein kinase Rim15 to control entry into G0. Mol Cell 12: 1607-1613 271. Perez VI, Bokov A, Van Remmen H, Mele J, Ran Q, et al. 2009. Is the oxidative stress theory of aging dead? Biochim Biophys Acta 1790: 1005- 1014 272. Perez VI, Van Remmen H, Bokov A, Epstein CJ, Vijg J, Richardson A. 2009. The overexpression of major antioxidant enzymes does not extend the lifespan of mice. Aging Cell 8: 73-75 273. Perocchi F, Mancera E, Steinmetz LM. 2008. Systematic screens for human disease genes, from yeast to human and back. Mol Biosyst 4: 18-29
203 274. Pfam. http://pfam.sanger.ac.uk/. 275. Phillips JP, Campbell SD, Michaud D, Charbonneau M, Hilliker AJ. 1989. Null mutation of copper/zinc superoxide dismutase in Drosophila confers hypersensitivity to paraquat and reduced longevity. Proc Natl Acad Sci U S A 86: 2761-2765 276. Piper MD, Bartke A. 2008. Diet and aging. Cell Metab 8: 99-104 277. Piper MD, Partridge L. 2007. Dietary restriction in Drosophila: delayed aging or experimental artefact? PLoS Genet 3: e57 278. Powers RW, 3rd, Kaeberlein M, Caldwell SD, Kennedy BK, Fields S. 2006. Extension of chronological life span in yeast by decreased TOR pathway signaling. Genes Dev 20: 174-184 279. Prentice HL, Kingston RE. 1992. Mammalian promoter element function in the fission yeast Schizosaccharomyces pombe. Nucleic Acids Res 20: 3383-3390 280. Qin W, Yang T, Ho L, Zhao Z, Wang J, et al. 2006. Neuronal SIRT1 activation as a novel mechanism underlying the prevention of Alzheimer disease amyloid neuropathology by calorie restriction. J Biol Chem 281: 21745-21754 281. Raffaghello L, Lee C, Safdie FM, Wei M, Madia F, et al. 2008. Starvation-dependent differential stress resistance protects normal but not cancer cells against high-dose chemotherapy. Proc Natl Acad Sci U S A 105: 8215-8220 282. Rea SL, Ventura N, Johnson TE. 2007. Relationship between mitochondrial electron transport chain dysfunction, development, and life extension in Caenorhabditis elegans. PLoS Biol 5: e259 283. Reinders A, Burckert N, Boller T, Wiemken A, De Virgilio C. 1998. Saccharomyces cerevisiae cAMP-dependent protein kinase controls entry into stationary phase through the Rim15p protein kinase. Genes Dev 12: 2943-2955 284. Rine J, Herskowitz I. 1987. Four genes responsible for a position effect on expression from HML and HMR in Saccharomyces cerevisiae. Genetics 116: 9-22 285. Rodriguez R, Campa VM, Riera J, Carcedo MT, Ucker DS, et al. 2007. TNF triggers mitogenic signals in NIH 3T3 cells but induces apoptosis when the cell cycle is blocked. Eur. Cytokine Netw. 18: 172-180 286. Rogina B, Helfand SL. 2004. Sir2 mediates longevity in the fly through a pathway related to calorie restriction. Proc Natl Acad Sci U S A 101: 15998-16003 287. Rogina B, Helfand SL, Frankel S. 2002. Longevity regulation by Drosophila Rpd3 deacetylase and caloric restriction. Science 298: 1745 288. Roguev A, Bandyopadhyay S, Zofall M, Zhang K, Fischer T, et al. 2008. Conservation and rewiring of functional modules revealed by an epistasis map in fission yeast. Science 322: 405-410 289. Roguev A, Wiren M, Weissman JS, Krogan NJ. 2007. High-throughput genetic interaction mapping in the fission yeast Schizosaccharomyces pombe. Nat Methods 4: 861-866
204 290. Rohde JR, Bastidas R, Puria R, Cardenas ME. 2008. Nutritional control via Tor signaling in Saccharomyces cerevisiae. Curr Opin Microbiol 11: 153-160 291. Rose MD, Winston F, Hieter P. 1990. Methods in yeast genetics: A laboratory course manual: Cold Spring Harbor Laboratory 292. Rossignol R, Faustin B, Rocher C, Malgat M, Mazat JP, Letellier T. 2003. Mitochondrial threshold effects. Biochem J 370: 751-762 293. Roth D, Wilson J. 1988. In Genetic Recombination, ed. R Kucherlapati, G Simth, pp. 621-653: American Society for Microbiology, Washington, DC 294. Roux AE, Arseneault G, Chartrand P, Ferbeyre G, Rokeach LA. 2010. A screen for genes involved in respiration control and longevity in Schizosaccharomyces pombe. Ann N Y Acad Sci 1197: 19-27 295. Roux AE, Chartrand P, Ferbeyre G, Rokeach LA. 2010. Fission yeast and other yeasts as emergent models to unravel cellular aging in eukaryotes. J Gerontol A Biol Sci Med Sci 65: 1-8 296. Roux AE, Leroux A, Alaamery MA, Hoffman CS, Chartrand P, et al. 2009. Pro-aging effects of glucose signaling through a G protein-coupled glucose receptor in fission yeast. PLoS Genet 5: e1000408 297. Roux AE, Quissac A, Chartrand P, Ferbeyre G, Rokeach LA. 2006. Regulation of chronological aging in Schizosaccharomyces pombe by the protein kinases Pka1 and Sck2. Aging Cell 5: 345-357 298. Sabourin M, Tuzon CT, Fisher TS, Zakian VA. 2007. A flexible protein linker improves the function of epitope-tagged proteins in Saccharomyces cerevisiae. Yeast 24: 39-45 299. Sajiki K, Hatanaka M, Nakamura T, Takeda K, Shimanuki M, et al. 2009. Genetic control of cellular quiescence in S. pombe. J Cell Sci 122: 1418- 1429 300. Salminen A, Kaarniranta K. 2010. Insulin/IGF-1 paradox of aging: regulation via AKT/IKK/NF-kappaB signaling. Cell Signal 22: 573-577 301. Sambrook J, Russell DW. 2001. Molecular Cloning-A laboratory manual. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press. 4.1-4.92 pp. 302. Samejima I, Yanagida M. 1994. Identification of cut8+ and cek1+, a novel protein kinase gene, which complement a fission yeast mutation that blocks anaphase. Mol Cell Biol 14: 6361-6371 303. Sarbassov DD, Ali SM, Sengupta S, Sheen JH, Hsu PP, et al. 2006. Prolonged rapamycin treatment inhibits mTORC2 assembly and Akt/PKB. Mol Cell 22: 159-168 304. Sato A, Nakada K, Shitara H, Kasahara A, Yonekawa H, Hayashi J. 2007. Deletion-mutant mtDNA increases in somatic tissues but decreases in female germ cells with age. Genetics 177: 2031-2037 305. Schena M, Shalon D, Davis RW, Brown PO. 1995. Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science 270: 467-470 306. Scherz-Shouval R, Elazar Z. 2007. ROS, mitochondria and the regulation of autophagy. Trends Cell Biol 17: 422-427
205 307. Schiestl RH, Zhu J, Petes TD. 1994. Effect of mutations in genes affecting homologous recombination on restriction enzyme-mediated and illegitimate recombination in Saccharomyces cerevisiae. Mol Cell Biol 14: 4493-4500 308. Schizosaccharomyces pombe GeneDB. http://old.genedb.org/genedb/pombe/ 309. Schriner SE, Linford NJ, Martin GM, Treuting P, Ogburn CE, et al. 2005. Extension of murine life span by overexpression of catalase targeted to mitochondria. Science 308: 1909-1911 310. Schuldiner M, Collins SR, Thompson NJ, Denic V, Bhamidipati A, et al. 2005. Exploration of the function and organization of the yeast early secretory pathway through an epistatic miniarray profile. Cell 123: 507- 519 311. Schuldiner M, Collins SR, Weissman JS, Krogan NJ. 2006. Quantitative genetic analysis in Saccharomyces cerevisiae using epistatic miniarray profiles (E-MAPs) and its application to chromatin functions. Methods 40: 344-352 312. Schulz TJ, Zarse K, Voigt A, Urban N, Birringer M, Ristow M. 2007. Glucose restriction extends Caenorhabditis elegans life span by inducing mitochondrial respiration and increasing oxidative stress. Cell Metab 6: 280-293 313. Sedensky MM, Morgan PG. 2006. Mitochondrial respiration and reactive oxygen species in C. elegans. Exp Gerontol 41: 957-967 314. Selman C, Lingard S, Choudhury AI, Batterham RL, Claret M, et al. 2008. Evidence for lifespan extension and delayed age-related biomarkers in insulin receptor substrate 1 null mice. Faseb J 22: 807-818 315. Selman C, Tullet JM, Wieser D, Irvine E, Lingard SJ, et al. 2009. Ribosomal protein S6 kinase 1 signaling regulates mammalian life span. Science 326: 140-144 316. Seringhaus M, Kumar A, Hartigan J, Snyder M, Gerstein M. 2006. Genomic analysis of insertion behavior and target specificity of mini-Tn7 and Tn3 transposons in Saccharomyces cerevisiae. Nucleic Acids Res 34: e57 317. Sherman F. 1991. Getting started with yeast. Methods Enzymol. 194: 3-21 318. Shi T, Wang F, Stieren E, Tong Q. 2005. SIRT3, a mitochondrial sirtuin deacetylase, regulates mitochondrial function and thermogenesis in brown adipocytes. J Biol Chem 280: 13560-13567 319. Shimanuki M, Chung SY, Chikashige Y, Kawasaki Y, Uehara L, et al. 2007. Two-step, extensive alterations in the transcriptome from G0 arrest to cell division in Schizosaccharomyces pombe. Genes Cells 12: 677-692 320. Shuster EO, Guthrie C. 1990. Human U2 snRNA can function in pre- mRNA splicing in yeast. Nature 345: 270-273 321. Simon JA, Szankasi P, Nguyen DK, Ludlow C, Dunstan HM, et al. 2000. Differential toxicities of anticancer agents among DNA repair and checkpoint mutants of Saccharomyces cerevisiae. Cancer Res 60: 328-333 322. Sinclair DA. 2005. Toward a unified theory of caloric restriction and longevity regulation. Mech Ageing Dev 126: 987-1002
206 323. Sinclair DA, Guarente L. 1997. Extrachromosomal rDNA circles--a cause of aging in yeast. Cell 91: 1033-1042 324. Singer T, Burke E. 2003. High-throughput TAIL-PCR as a tool to identify DNA flanking insertions. Methods Mol Biol 236: 241-272 325. Singhi AD, Kondratov RV, Neznanov N, Chernov MV, Gudkov AV. 2004. Selection-subtraction approach (SSA): a universal genetic screening technique that enables negative selection. Proc Natl Acad Sci U S A 101: 9327-9332 326. Sipiczki M. 2004. Fission Yeast Phylogenesis and Evolution. In The Molecular Biology of Schizosaccharomyces pombe, ed. R Egel, pp. 431- 443. Heidelberg: Springer-Verlag 327. Smith AM, Heisler LE, Mellor J, Kaper F, Thompson MJ, et al. 2009. Quantitative phenotyping via deep barcode sequencing. Genome Res 19: 1836-1842 328. Smith AM, Heisler LE, St Onge RP, Farias-Hesson E, Wallace IM, et al. 2010. Highly-multiplexed barcode sequencing: an efficient method for parallel analysis of pooled samples. Nucleic Acids Res 329. Smith DL, Jr., McClure JM, Matecic M, Smith JS. 2007. Calorie restriction extends the chronological lifespan of Saccharomyces cerevisiae independently of the Sirtuins. Aging Cell 6: 649-662 330. Smith JS, Boeke JD. 1997. An unusual form of transcriptional silencing in yeast ribosomal DNA. Genes Dev 11: 241-254 331. Smith JS, Caputo E, Boeke JD. 1999. A genetic screen for ribosomal DNA silencing defects identifies multiple DNA replication and chromatin- modulating factors. Mol Cell Biol 19: 3184-3197 332. Sohal RS, Weindruch R. 1996. Oxidative stress, caloric restriction, and aging. Science 273: 59-63 333. Sokal RR, Rohlf FJ. 1995. Biometry: The Principals and Practice of Statistics in Biological Research. New York: W. H. Freeman and Company. 850 pp. 334. Soukas AA, Kane EA, Carr CE, Melo JA, Ruvkun G. 2009. Rictor/TORC2 regulates fat metabolism, feeding, growth, and life span in Caenorhabditis elegans. Genes Dev 23: 496-511 335. Stadtman ER. 2006. Protein oxidation and aging. Free Radic Res 40: 1250-1258 336. Stanfel MN, Shamieh LS, Kaeberlein M, Kennedy BK. 2009. The TOR pathway comes of age. Biochim Biophys Acta 1790: 1067-1074 337. Steffen KK, MacKay VL, Kerr EO, Tsuchiya M, Hu D, et al. 2008. Yeast life span extension by depletion of 60s ribosomal subunits is mediated by Gcn4. Cell 133: 292-302 338. Su SS, Tanaka Y, Samejima I, Tanaka K, Yanagida M. 1996. A nitrogen starvation-induced dormant G0 state in fission yeast: the establishment from uncommitted G1 state and its delay for return to proliferation. J Cell Sci 109 ( Pt 6): 1347-1357 339. Suga M, Goto A, Hatakeyama T. 2007. Electrically induced protein release from Schizosaccharomyces pombe cells in a hyperosmotic
207 condition during and following a high electropulsation. J Biosci Bioeng 103: 298-302 340. Suga M, Hatakeyama T. 2005. A rapid and simple procedure for high- efficiency lithium acetate transformation of cryopreserved Schizosaccharomyces pombe cells. Yeast 22: 799-804 341. Sun C, Zhang F, Ge X, Yan T, Chen X, et al. 2007. SIRT1 improves insulin sensitivity under insulin-resistant conditions by repressing PTP1B. Cell Metab 6: 307-319 342. Sutter NB, Bustamante CD, Chase K, Gray MM, Zhao K, et al. 2007. A single IGF1 allele is a major determinant of small size in dogs. Science 316: 112-115 343. Takeda K, Yoshida T, Kikuchi S, Nagao K, Kokubu A, et al. 2010. Synergistic roles of the proteasome and autophagy for mitochondrial maintenance and chronological lifespan in fission yeast. Proc Natl Acad Sci U S A 107: 3540-3545 344. Tanaka K, Okayama H. 2000. A pcl-like cyclin activates the Res2p- Cdc10p cell cycle "start" transcriptional factor complex in fission yeast. Mol Biol Cell 11: 2845-2862 345. Tanaka M, Chock PB, Stadtman ER. 2007. Oxidized messenger RNA induces translation errors. Proc Natl Acad Sci U S A 104: 66-71 346. Tatar M, Bartke A, Antebi A. 2003. The endocrine regulation of aging by insulin-like signals. Science 299: 1346-1351 347. Tatar M, Kopelman A, Epstein D, Tu MP, Yin CM, Garofalo RS. 2001. A mutant Drosophila insulin receptor homolog that extends life-span and impairs neuroendocrine function. Science 292: 107-110 348. Tatebayashi K, Kato J, Ikeda H. 1994. Structural analyses of DNA fragments integrated by illegitimate recombination in Schizosaccharomyces pombe. Mol Gen Genet 244: 111-119 349. Thierbach R, Schulz TJ, Isken F, Voigt A, Mietzner B, et al. 2005. Targeted disruption of hepatic frataxin expression causes impaired mitochondrial function, decreased life span and tumor growth in mice. Hum Mol Genet 14: 3857-3864 350. Tissenbaum HA, Guarente L. 2001. Increased dosage of a sir-2 gene extends lifespan in Caenorhabditis elegans. Nature 410: 227-230 351. Tong AH, Evangelista M, Parsons AB, Xu H, Bader GD, et al. 2001. Systematic genetic analysis with ordered arrays of yeast deletion mutants. Science 294: 2364-2368 352. Trifunovic A, Larsson NG. 2008. Mitochondrial dysfunction as a cause of ageing. J Intern Med 263: 167-178 353. Tu MP, Epstein D, Tatar M. 2002. The demography of slow aging in male and female Drosophila mutant for the insulin-receptor substrate homologue chico. Aging Cell 1: 75-80 354. Um SH, Frigerio F, Watanabe M, Picard F, Joaquin M, et al. 2004. Absence of S6K1 protects against age- and diet-induced obesity while enhancing insulin sensitivity. Nature 431: 200-205 355. Van Raamsdonk JM, Hekimi S. 2009. Deletion of the mitochondrial
208 superoxide dismutase sod-2 extends lifespan in Caenorhabditis elegans. PLoS Genet 5: e1000361 356. Vaquero A, Reinberg D. 2009. Calorie restriction and the exercise of chromatin. Genes Dev 23: 1849-1869 357. Varvel NH, Bhaskar K, Patil AR, Pimplikar SW, Herrup K, Lamb BT. 2008. Abeta oligomers induce neuronal cell cycle events in Alzheimer's disease. J. Neurosci. 28: 10786-10793 358. Vazquez-Manrique RP, Gonzalez-Cabo P, Ros S, Aziz H, Baylis HA, Palau F. 2006. Reduction of Caenorhabditis elegans frataxin increases sensitivity to oxidative stress, reduces lifespan, and causes lethality in a mitochondrial complex II mutant. Faseb J 20: 172-174 359. Vellai T, Takacs-Vellai K, Zhang Y, Kovacs AL, Orosz L, Muller F. 2003. Genetics: influence of TOR kinase on lifespan in C. elegans. Nature 426: 620 360. Ventura N, Rea S, Henderson ST, Condo I, Johnson TE, Testi R. 2005. Reduced expression of frataxin extends the lifespan of Caenorhabditis elegans. Aging Cell 4: 109-112 361. Ventura N, Rea SL. 2007. Caenorhabditis elegans mitochondrial mutants as an investigative tool to study human neurodegenerative diseases associated with mitochondrial dysfunction. Biotechnol J 2: 584-595 362. Vidan S, Mitchell AP. 1997. Stimulation of yeast meiotic gene expression by the glucose-repressible protein kinase Rim15p. Mol Cell Biol 17: 2688- 2697 363. Vitiello SP, Wolfe DM, Pearce DA. 2007. Absence of Btn1p in the yeast model for juvenile Batten disease may cause arginine to become toxic to yeast cells. Hum Mol Genet 16: 1007-1016 364. Wallace DC. 2005. A mitochondrial paradigm of metabolic and degenerative diseases, aging, and cancer: a dawn for evolutionary medicine. Annu Rev Genet 39: 359-407 365. Wang F, Nguyen M, Qin FX, Tong Q. 2007. SIRT2 deacetylates FOXO3a in response to oxidative stress and caloric restriction. Aging Cell 6: 505- 514 366. Wang L, Kao R, Ivey FD, Hoffman CS. 2004. Strategies for gene disruptions and plasmid constructions in fission yeast. Methods 33: 199- 205 367. Wang RH, Zheng Y, Kim HS, Xu X, Cao L, et al. 2008. Interplay among BRCA1, SIRT1, and Survivin during BRCA1-associated tumorigenesis. Mol Cell 32: 11-20 368. Wanke V, Cameroni E, Uotila A, Piccolis M, Urban J, et al. 2008. Caffeine extends yeast lifespan by targeting TORC1. Mol Microbiol 69: 277-285 369. Wanke V, Pedruzzi I, Cameroni E, Dubouloz F, De Virgilio C. 2005. Regulation of G0 entry by the Pho80-Pho85 cyclin-CDK complex. EMBO J 24: 4271-4278 370. Wawryn J, Krzepilko A, Myszka A, Bilinski T. 1999. Deficiency in superoxide dismutases shortens life span of yeast cells. Acta Biochim Pol
209 46: 249-253 371. Wei M, Fabrizio P, Hu J, Ge H, Cheng C, et al. 2008. Life span extension by calorie restriction depends on Rim15 and transcription factors downstream of Ras/PKA, Tor, and Sch9. PLoS Genet 4: e13 372. Weinberger M, Feng L, Paul A, Smith DL, Jr., Hontz RD, et al. 2007. DNA replication stress is a determinant of chronological lifespan in budding yeast. PLoS ONE 2: e748 373. Weindruch R, Walford RL, Fligiel S, Guthrie D. 1986. The retardation of aging in mice by dietary restriction: longevity, cancer, immunity and lifetime energy intake. J. Nutr. 116: 641-654 374. Werner-Washburne M, Braun EL, Crawford ME, Peck VM. 1996. Stationary phase in Saccharomyces cerevisiae. Mol Microbiol 19: 1159- 1166 375. Winzeler EA, Shoemaker DD, Astromoff A, Liang H, Anderson K, et al. 1999. Functional characterization of the S. cerevisiae genome by gene deletion and parallel analysis. Science 285: 901-906 376. Wood JG, Rogina B, Lavu S, Howitz K, Helfand SL, et al. 2004. Sirtuin activators mimic caloric restriction and delay ageing in metazoans. Nature 430: 686-689 377. Wood V, Gwilliam R, Rajandream MA, Lyne M, Lyne R, et al. 2002. The genome sequence of Schizosaccharomyces pombe. Nature 415: 871-880 378. Wu X, Zhong G, Findley SD, Cregan P, Stacey G, Nguyen HT. 2008. Genetic marker anchoring by six-dimensional pools for development of a soybean physical map. BMC Genomics 9: 28 379. Yamada K, Nakagawa CW, Mutoh N. 1999. Schizosaccharomyces pombe homologue of glutathione peroxidase, which does not contain selenocysteine, is induced by several stresses and works as an antioxidant. Yeast 15: 1125-1132 380. Yanagida M. 2002. The model unicellular eukaryote, Schizosaccharomyces pombe. Genome Biol 3: COMMENT2003 381. Yanagida M. 2009. Cellular quiescence: are controlling genes conserved? Trends Cell Biol 19: 705-715 382. Yang Y, Varvel NH, Lamb BT, Herrup K. 2006. Ectopic cell cycle events link human Alzheimer's disease and amyloid precursor protein transgenic mouse models. J. Neurosci. 26: 775-784 383. Yang Z, Geng J, Yen WL, Wang K, Klionsky DJ. 2010. Positive or negative roles of different cyclin-dependent kinase Pho85-cyclin complexes orchestrate induction of autophagy in Saccharomyces cerevisiae. Mol Cell 38: 250-264 384. Yashin AI, Cypser JR, Johnson TE, Michalski AI, Boyko SI, Novoseltsev VN. 2001. Ageing and survival after different doses of heat shock: the results of analysis of data from stress experiments with the nematode worm Caenorhabditis elegans. Mech Ageing Dev 122: 1477-1495 385. Yoritaka A, Hattori N, Uchida K, Tanaka M, Stadtman ER, Mizuno Y. 1996. Immunohistochemical detection of 4-hydroxynonenal protein adducts in Parkinson disease. Proc Natl Acad Sci U S A 93: 2696-2701
210 386. Yu BP. 1994. Cellular defenses against damage from reactive oxygen species. Physiol Rev 74: 139-162 387. Yui R, Ohno Y, Matsuura ET. 2003. Accumulation of deleted mitochondrial DNA in aging Drosophila melanogaster. Genes Genet Syst 78: 245-251 388. Zainal TA, Oberley TD, Allison DB, Szweda LI, Weindruch R. 2000. Caloric restriction of rhesus monkeys lowers oxidative damage in skeletal muscle. Faseb J 14: 1825-1836 389. Zarse K, Schulz TJ, Birringer M, Ristow M. 2007. Impaired respiration is positively correlated with decreased life span in Caenorhabditis elegans models of Friedreich Ataxia. Faseb J 21: 1271-1275 390. Zid BM, Rogers AN, Katewa SD, Vargas MA, Kolipinski MC, et al. 2009. 4E-BP extends lifespan upon dietary restriction by enhancing mitochondrial activity in Drosophila. Cell 139: 149-160 391. Zimmermann S, Martens UM. 2008. Telomeres, senescence, and hematopoietic stem cells. Cell Tissue Res. 331: 79-90 392. Zuin A, Carmona M, Morales-Ivorra I, Gabrielli N, Vivancos AP, et al. 2010. Lifespan extension by calorie restriction relies on the Sty1 MAP kinase stress pathway. EMBO J 29: 981-991 393. Zuin A, Gabrielli N, Calvo IA, Garcia-Santamarina S, Hoe KL, et al. 2008. Mitochondrial dysfunction increases oxidative stress and decreases chronological life span in fission yeast. PLoS ONE 3: e2842
211