Send Orders of Reprints at [email protected] 558 Current Genomics, 2012, 13, 558-565 Systems Biology in Aging: Linking the Old and the Young Lei Hou1,2, Jialiang Huang1,2, Christopher D. Green1, Jerome Boyd-Kirkup1, Wei Zhang1,2, Xiaoming Yu1,2, Wenxuan Gong1, Bing Zhou1,2 and Jing-Dong J. Han1,* 1Chinese Academy of Sciences Key Laboratory of Computational Biology, Chinese Academy of Sciences-Max Planck Partner Institute for Computational Biology, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, China; 2Center of Molecular Systems Biology, Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing, China Abstract: Aging can be defined as a process of progressive decline in the physiological capacity of an organism, mani- fested by accumulated alteration and destabilization at the whole system level. Systems biology approaches offer a prom- ising new perspective to examine the old problem of aging. We begin this review by introducing the concepts of systems biology, and then illustrate the application of systems biology approaches to aging research, from gene expression profil- ing to network analysis. We then introduce the network that can be constructed using known lifespan and aging regula- tors, and conclude with a look forward to the future of systems biology in aging research. In summary, systems biology is not only a young field that may help us understand aging at a higher level, but also an important platform that can link dif- ferent levels of knowledge on aging, moving us closer to a more comprehensive control of systematic decline during aging. Received on: March 25, 2012- Revised on: June 11, 2012- Accepted on: July 25, 2012 Keywords: Systems biology, Genomics, Proteomics, Longevity, Aging, Network analysis. 1. INTRODUCTION seem to play a central role by integrating upstream signals and regulating downstream processes. The crosstalk between Aging has fascinated researchers since ancient times. The these pathways and the combination of regulatory signals hugely complicated process that has been revealed may be ultimately affect the systems level phenotype of aging and interpreted from different aspects, such as the accumulation aging-associated functional decline. From this aspect, sys- of oxidative damage, shortening of telomeres, the costs of tems biology may be critical to integrate the various aging reproduction, metabolic rates, cellular senescence, etc., and regulatory signals, and predict and interpret the systems level these have in turn given rise to diverse theories of aging [1]. output, i.e. the aging-associated phenotypes. However, thanks to forward and reverse genetic technolo- gies, researchers in the recent decades have established that Systems biology examines a biological individual as a despite its complexity, a single or a few key genes in a few living system, consisting of components of different layers key pathways can modulate the aging rate. The most impor- Fig. (2A): 1) the chromosomal level including genomic se- tant players would appear to be those in nutrient sensing quences and the epigenetic states of chromosomes, such as pathways or stress response pathways, such as DAF- DNA methylation and histone modifications, 2) the RNA 2/IGF1R and DAF-16/FOXO in the Insulin/IGF like signal- level including mRNA and non-coding RNAs, 3) the protein ing pathway, AAK-2/AMPK in another nutrient sensing level including proteins and their different modification pathway, JNK in the stress response pathway, LET- states, and 4) the metabolite level. The cellular states at these 363/mTOR as an inhibitor of autophagy and activator of different layers can be monitored by high-throughput ex- translation and SIRT1/SIR2 in genome stability mainte- perimental approaches such as next generation sequencing nance, to name a few [2, 3]. In addition to genetic perturba- and de novo assembly for genomic sequences, chromatin tions, dietary perturbations, such as diet restriction (DR) are immunoprecipitation followed by microarray (ChIP-chip) or known to significantly extend lifespan in most organisms deep sequencing (ChIP-seq) for histone modifications, bisul- examined from yeasts to primates, although different path- fite sequencing for DNA methylation, microarray or RNA- ways may act under different DR conditions, and alternative seq for mRNA and for miRNA, protein microarray or mass DR strategies also effect C.elegans lifespan in different ways spectrometry (MS) for proteins, and MS or nuclear magnetic [3, 4]. The main pathways revealed under different DR resonance (NMR) for metabolites. Molecular interactions regimens are summarized in Fig. (1). In this small, convo- within and between these layers give rise to the observed luted DR response network, DAF-16 and ceTOR/LET-363 states of the whole system, manifested as phenotypes Fig. (2). All of these layers, as well as other potentially unidenti- fied ones, can be perturbed or regulated by physiological *Address correspondence to this author at the Chinese Academy of Sci- signals, as well as environmental cues. Unlike classical ap- ences, 320 Yue Yang Road, Shanghai, 200031, China; Tel: 86-21- proaches, which focus on only specific molecules or path- 54920458; Fax: 86-21-54920451; E-mail: [email protected] ways, systems biology takes advantage of these high- §These authors contributed equally to this work. throughput approaches to study components in each layer 1389-2029/12 $58.00+.00 ©2012 Bentham Science Publishers Systems Biology in Aging: Linking the Old and the Young Current Genomics, 2012, Vol. 13, No. 7 559 Fig. (1). A brief summary of the effects of dietary restriction (DR) on aging in C.elegans. lDR (liquid food DR), IF (intermittent fasting), DD (dietary deprivation), sDR (solid food DR), bDR (bacterial DR) and eat-2 are different DR strategies or mimics, as indicated in [4, 5]. Dashed lines denote known relationships between the two nodes which have not been demonstrated in DR. Fig. (2). Different layers of molecular interactions in a biological system. A) Life as a system is composed of information from different layers, modified based on [6]. B) Different approaches for mapping interactions within and between different layers. 3C (Chromosome Con- formation Capture), Hi-C (High-throughput Chromosome Capture), FISH (Fluorescence In Situ Hybridization), AGO-IP (Argonaute Im- munoprecipitation), BiFC (Bimolecular Fluorescence Complementation), Y2H (Yeast Two-Hybrid), TAP (Tandem Affinity Purification), Co-IP (Co- Immunoprecipitation), EMSA (Electrophoretic Mobility Shift Assay), Y1H (Yeast One-Hybrid), ChIP (Chromatin Immunopre- cipitation), RIP (RNA Immunoprecipitation), CLIP (Cross Linking and Immunoprecipitation). 560 Current Genomics, 2012, Vol. 13, No. 7 Hou et al. and their interactions at multiple layers during a time course lectively deposited in the AGEMAP database, and identified or after perturbation. some genes that have different age-related patterns in differ- ent groups of tissues. Rather than focus on spatial differ- 2. DETECTING MOLECULAR AGING PROFILES ences, others have investigated temporal differences in more USING HIGH-THROUGHPUT APPROACHES detail with additional time points, and have extended the studies into humans. Lu et al. [15] provided the first com- To examine the changes of individual genes during ag- prehensive aging-related map of the human brain transcrip- ing, experiments have been designed either to compare tome changes using 30 postmortem samples from age 26 to young and old samples, or to monitor a time series during 106, and identified that, among many other molecular and aging. In some cases, different conditions (different genetic functional changes, DNA repair related genes significantly backgrounds, tissues, diets, stimuli, etc.) are also taken into increase with age, suggesting that there is DNA damage consideration. These data are then analyzed to identify dif- stress during human brain aging. Also, using time course ferentially expressed genes, age-related gene expression data analyses, Somel et al. [16, 17] found a close relation- changes, and functional enrichment among the changed ship between development and aging, and a delay in the tim- genes. ing of the aging transition in humans compared with other As early as 1999, Lee et al. [7] showed an increase of primates. Recently using 1340 tissue samples from 57 devel- expression in stress response genes and a decrease in meta- oping and adult brain samples, Kang et al. [18] quantified bolic and biosynthetic genes during aging. Samples were expression trajectories at both the gene-level and exon-level, collected from skeletal muscle of mice to compare old and and provide a rich resource on both temporal and spatial young animals under normal and DR conditions [7]. They changes in human brain. found that DR could postpone aging by inhibiting macromo- In order to find common pathways affected by different lecular damage and promoting protein turnover. Pioneering dietary interventions that modulate aging and lifespan, we microarray studies prior to 2004,suggested that the effects of have obtained midlife hepatic gene expression profiles of aging, DR and exercise are characterized by transcriptional mice on high-fat diet, high-fat diet with CR, high-fat diet changes [8]. Lund et al., [9] using worms from different ge- with voluntary exercise, low-fat diet, low-fat diet with CR netic backgrounds, found that the expression of heat shock and low-fat diet with voluntary exercises. We found that genes decreased while certain transposases increased during pathways whose gene expression levels are correlated with C. elegans aging. Their work also suggested that the aging the mean lifespan under these six conditions are enriched for process and the transition from the reproductive to the dauer aging regulatory pathways, suggesting different dietary in- stage share some similar alterations in gene expression. tervention regimens may target a set of common lifespan Pletcher et al. [10] showed that genes related to stress re- modulating pathways and functions [19]. Similar to gene sponse and oogenesis seem to change during Drosophila expression profiles for coding genes, aging-related changes melanogaster aging, and those involved in cell growth, me- in miRNAs have also been profiled for C. elegans [20, 21], tabolism and reproduction are down-regulated by DR. This monkey and the human brain [17, 20, 21]. data argued against the hypothesis that aging is due to tran- scriptional changes in some specific genomic regions, but Other high-throughput data has helped to globally char- rather support it as the result of increasing disregulation of acterize aging from different perspectives. As protein trans- gene expression. Kayo et al. [11] using samples from the lation plays a regulatory role in aging and lifespan, recent muscles of monkeys, found an increase in inflammation and deep sequence-based polysome-RNA profiling has been used oxidative stress as well as a decrease in mitochondrial elec- to detect translationally regulated genes through eukaryotic tron transport and oxidative phosphorylation during aging. translation initiation factor 4G (eIF4G) knock-down, known Bronikowski et al. [12] demonstrated that, in mouse heart, to extend lifespan, and has revealed that many lifespan regu- exercise could delay the aging-related expression changes, lators, in particular genes encoding respiratory chain compo- which involve inflammatory response, stress response, signal nents are controlled at the translation level [22, 23]. transduction and energy metabolism genes. Studies during These high throughput profiling experiments have gener- this period also provided the first glimpse of aging-related ated large amounts of data for meta-analysis [24], which can global transcriptomic changes. They indicated that some compare molecular functions and expression patterns that common changes, such as the involvement of stress re- change during aging in different systems. However, such sponse, energy metabolism and mitochondrial genes, are studies are far from exhaustive, as they only describe the shared by different species during aging and modified by molecular changes during aging, which could in fact be the DR. consequence of aging, rather than the cause of aging. Thus to Since 2004, many more factors have been considered explore the causal factors for aging, studies are increasingly when designing microarray experiments. For example, to devoted to the identification of aging and lifespan regulators. compare age-related changes between different species, McCarroll et al. [13] carried out microarray analyses for 3. INFERRING AGING REGULATORS both C.elegans and D. melanogaster, and by examining The majority of the known aging/lifespan regulators, for changes in orthologous genes between the two species they example, age-1 [25], daf-2 and daf-16 [26], were identified detected similar age-related changes in mitochondria and through genetics approaches in model organisms such as DNA repair genes in both organisms. To examine tissue spe- yeast, worm and fruitfly. Most of these genes are so defined cific changes, Zahn et al. [14] generated expression profiles because their perturbation (knock out, knock down or over- for 16 different mouse tissues during aging, which are col- expression) can extend lifespan. A few genome-wide RNAi Systems Biology in Aging: Linking the Old and the Young Current Genomics, 2012, Vol. 13, No. 7 561 screens have also identified such genes (see the accompany- 3.3. Network-Based Approaches ing review by Bennett et al. [27] in this issue). Searching for Based on these large-scale molecular interactions data, aging regulators is important to reduce the complexity of such as protein-protein interactions (PPIs), genetic interac- aging down to single genes. This is crucial to delineating tions, TF-target interactions, and miRNA-target interactions, how aging is regulated by specific signals, such as signals molecular networks can be used to visualize the relationships from nutrient or stress sensing, metabolism, translation, re- among a gene set, with genes represented as nodes and their production, telomeres, etc. [3]. Systems biology can build on molecular interactions as edges. Topological features of a these findings and predict the critical players and combinato- network can often reveal the most critical regulators as hubs, rial effects. Large-scale approaches are generally used to or nodes with the most links, and the functional monitor candidate genes at the same time, often with or units/neighborhood among genes as the network modules, without perturbation to the system. within which nodes are densely connected and in between which the nodes are relatively loosely connected. 3.1. Genetics-Based Approaches Managbanag et al. [37] have found that the genes linking Two independent genome-wide RNAi screens for lon- the known longevity genes in Saccharomyces cerevisiae gevity genes in C.elegans have found 89 and 23 longevity through PPI shortest paths are more likely to be aging regu- genes respectively, from more than ten thousand clones [28, lators. Utilizing a human PPI network, Bell et al. [38] found 29], but there is still a significant false negative ratio [29]. In that genes within a subnetwork consisting of human aging addition, given that these screens do not identify any gene regulators, human homologs of aging regulators identified in whose function promotes longevity, the number should invertebrates, and their one-step neighbors are more likely to probably be at least double. Candidate gene-based or ge- be hubs and more connected to each other than expected for nome-wide association study (GWAS) have identified single the rest of the PPI network. Budovsky et al. [39] also nucleotide polymorphism (SNPs) in FOXO1A and FOXO3A showed that hubs in the PPI network connecting human ho- contributing to the longevity of human populations [30-33]. mologs of lifespan modifiers were often associated with age- A further genetic approach to predicting possible genetic related diseases. Focusing on the modularity of the ag- cause of aging uses the identification of expression quantita- ing/longevity network, we have shown that the subnetwork tive trait loci (eQTL), and has identified associations be- of genes with positively or negatively correlated transcrip- tween SNPs and genes with expression changes during aging tional changes during aging largely exist in a small number [18]. of network modules. In particular, the modules showing negative expression correlation with each other during aging 3.2. Expression Profile-Based Approaches correspond to alternative temporal cellular states, such as A different, but straightforward, approach to predicting proliferation versus differentiation and reductive metabolic aging regulators uses the profiles of transcriptional or trans- versus oxidative metabolic, and the genes connecting mod- lational changes induced by genetic or environmental pertur- ules tend to be enriched for transcriptional regulators and bations that alter lifespan. Comparison of transcriptional lifespan/aging modifiers [40, 41]. profiles between long-lived daf-2 or age-1 mutants and wild- Network analyses additionally revealed systems level type or the daf-16; daf-2 double mutant showed that poten- relationships between age-related diseases and the aging tial downstream targets of daf-16 are also aging regulators regulators. Miller et al. [42] used a weighted gene co- like daf-16 itself. Additional comparisons of time course expression network to identify transcriptional networks in data using daf-2 RNAi, daf-2 and daf-16 double RNAi and Alzheimer's disease (AD) and found a significant association wild type reduced the false positive rate in the prediction of between gene expression changes during the progression of aging regulators [34]. A similar experiment that measured AD and those during normal aging. Wang et al. [43] con- translational profile changes by ifg-1 RNAi, the C. elegans structed a human disease-aging network to study the rela- ortholog of eIF4G, found that factors that suppress the lon- tionships between aging genes and genetic disease genes. gevity induced by ifg-1 RNAi can be predicted by detecting This study showed that disease genes located close to aging differentially translated genes [23]. genes have central positions in the PPI network. Another common procedure is based on the assumption Network approaches are instrumental in discerning that the upstream regulators responsible for transcriptional global properties of aging/lifespan regulators, making com- changes during aging potentially regulate aging. Through putational predictions and inferring the modularity and rela- searching differentially expressed genes, enriched transcrip- tionships of various aging regulators. However, they should tion factor (TF) binding motifs enriched on the promoters be applied with great caution as to avoid bias introduced by these genes, NF-kB and elt-3/elt-5/elt-6 GATA transcrip- the literature, the lack of spatial and temporal information, or tional circuit have been identified as aging regulators in the limited coverage of the network [44]. mammals and C. elegans, respectively [35, 36]. Approaches based on transcriptional profiles are quick to 4. EPIGENETIC REGULATION OF AGING implement, but have certain limitations, for example, the In addition to gene expression changes, the states of epi- overrepresentation of TF motifs does not necessarily indicate genetic modifications have emerged to be significantly im- a causal relationship. Even when the predicted factors are portant in modulating lifespan (see the accompanying review necessary for the expression change, they are not necessarily by Liu and Zhou in this issue [45]). Epigenetic modifications sufficient to cause the change. include DNA and histone modifications that are potentially 562 Current Genomics, 2012, Vol. 13, No. 7 Hou et al. heritable and reversible without changing the genetic code ing the H3K27me3 states of genes in the insulin/IGF-1 sig- [46]. With the application of recent high-throughput ap- naling pathway [57]. Together, these works indicate that proaches, such as bisulfite sequencing, ChIP-seq or ChIP- histone modifications can be equally important as gene chip, etc. (Section 1), epigenetic controls have become well- expression as both markers and modifiers of aging. recognized as important regulatory mechanisms during the Unlike gene expression profiling, comprehensive profil- lifetime of an organism [46, 47]. For example, using the anti- ing of the epigenetic landscape during aging has been hin- O-GlcNAc ChIP-on-chip whole-genome tiling arrays on dered by the identification of numerous epigenetic marks and C.elegans, Love et al. [48] found 800 genes displaying dif- their combinational codes [58-60]. Currently, experimental ferential cycling of O-GlcNAc which have functions closely limitations allow only for the selection of a specific histone related to aging. By examining DNA methylation at CpG or DNA modification of interest, for example, whose modi- sites throughout the human genome, Hernandez et al. [49] fication enzyme is known to affect aging/lifespan, and gen- identified hundreds of CpG sites with levels of DNA methy- erate a genome-wide profile for the particular modification. lation in the human brain highly correlated with chronologi- Future high-throughput technologies that can simultaneously cal age. probe multiple or all epigenetic modifications will greatly Many regulators of histone modifications have been expedite our understanding of the aging epigenome. In the found to be associated with longevity in worms and fruitflies meantime, focus on the epigenetic changes that directly de- [28, 50, 51]. Recently, several excellent studies further high- termine gene expression [61] may greatly simplify the task light the link between histone modifications and the aging of mapping the aging epigenome. process Fig. (3). In particular, Peleg et al. [52] found that 5. FUTURE DIRECTIONS deregulated acetylation of histone H4 lysine 12 (H4K12) may represent an early biomarker of an impaired genome Systems biology has made great progress during the last environment and reduced cognitive ability in the aging decade, mainly as a platform to study complex biological mouse brain. Additionally, blocking the loss of H4K12 ace- problems. Systems biology has also made great achieve- tylation incurred by increased histone deacetylase 2 activity ments in aging through profiling mRNA, miRNA and pro- in the brain of Alzheimer mouse model could effectively tein changes during aging, inferring the relationships block the onset of the disease [53]. The Sir2 histone deacety- among aging-associated genes and predicting unknown lase regulates the replicative lifespan of yeast by acting on aging regulators. However, it is far from enough. histone H4K16 at subtelomeric regions and near the ribo- At least two aspects need to be addressed using a sys- somal DNA [54]. Increasing levels of SIRT1, a mammalian tem biology approach in aging research. First, although homolog of Sir2, promoted genomic stability and delayed many different pathways, compartments or processes are aging-related gene expression changes in mice, in part, by known to be closely related to aging, such as the IIS decreasing histone acetylation [55]. The genome-wide bind- pathway, autophagy, mitochondria, oxidative stress re- ing profiles of the SIRT1, analyzed by ChIP-chip, revealed sponse and so on, it remains unclear as to how they inter- that genes disassociated with SIRT1 upon DNA damage in act, are co-regulated and balanced during aging. To pro- mouse embryo stem cells significantly overlap with genes vide a glimpse of this problem, we visualized the network derepressed in the aging mouse brains, suggesting a role of communities among the known aging regulators based on SIRT1 in keeping gene silenced in the young cells [55]. entries in the GenAge database [62, 63] Fig. (4). Utilizing Greer et al. have shown that members of the H3K4 trimethy- either protein interaction network data Fig. (4A) or litera- lation complex regulated lifespan in a germline-dependent ture co-citation data Fig. (4B) as a network template, the manner in C. elegans [56]. We have found that a H3K27 aging regulator network appears to robustly consist of demethylase UTX-1 regulated C. elegans lifespan by chang- four major parts: 1) signaling pathways sensing nutrients Fig. (3). A summary of data linking epigenome to aging [52, 54-57]. Systems Biology in Aging: Linking the Old and the Young Current Genomics, 2012, Vol. 13, No. 7 563 Fig. (4). Network communities among known aging regulators in human and model organisms based on two different interactome datasets. Nodes include human aging regulators and human homologs of aging regulators in worm, fly and mouse from GenAge [62, 63]. Clusters or communities in the networks were extracted by the MCL algorithm [64] and only top clusters with more than 10 genes for each network are shown, and different clusters with similar functional enrichment are merged. (A) The network based on a protein functional in- teraction network [65]. (B) The edges in the network represent cocitation of the two genes together in at least 2 PubMed abstracts under the context of aging, i.e. also co-cited with “aging”, “ageing”, “lifespan”, “life span” as calculated by Cociter (http:// www.picb.ac.cn/ han- lab/cociter). In both graphs, the enriched functions within the gene clusters are coded by the colors of the nodes: green - signaling pathways, red - DNA damage response, yellow - mitochondria function and oxidative stress response, blue - ribosome and translation related genes, and purple - protein localization, transport and autophagy. and controlling growth and proliferation (green nodes), Sciences (Grant #KSCX2-EW-R-02 and KSCX2-EW-J- DNA damage response for maintaining integrity of the 15) and stem cell leading project XDA01010303 J.D.J.- genome (red nodes), mitochondria and oxidative stress H., CAS foreign young investigator fellowships to C.D.G. response (yellow nodes), and ribosome and translation and J.B.-K. and NSFC training grant to C.D.G. (blue nodes). It is obvious that the first two are inten- REFERENCES sively linked and closely entangled, while the latter two are relatively independent processes with only few links [1] Weinert, B.T.; Timiras, P.S. Invited review: Theories of aging. J connected to the first two processes. Also, it is interesting Appl Physiol, 2003, 95(4), 1706-16. [2] Antebi, A. Genetics of aging in Caenorhabditis elegans. PLoS to note that, by comparing the molecular interaction-based Genet, 2007, 3(9), 1565-71. network with the co-citation network, the role of auto- [3] Kenyon, C.J. The genetics of ageing. Nature, 2010, 464(7288), phagy and protein transport in aging might be either over- 504-12. estimated due to study bias or under-estimated by the in- [4] Greer, E.L.; Brunet, A. Different dietary restriction regimens extend lifespan by both independent and overlapping genetic completeness of the molecular interactions among these pathways in C. elegans. Aging Cell, 2009, 8(2), 113-27. genes. [5] Honjoh, S.; Yamamoto, T.; Uno, M.; Nishida, E. Signalling through RHEB-1 mediates intermittent fasting-induced longevity in Second, although high-throughput data on different lay- C. elegans. Nature, 2009, 457(7230), 726-30. ers of the living system Fig. (2) can now be easily obtained, [6] Han, J.D. Understanding biological functions through molecular it remains obscure as to how information flows or ex- networks. Cell Res, 2008, 18(2), 224-37. changes across these layers to arrive at the alternative [7] Lee, C.K.; Klopp, R.G.; Weindruch, R.; Prolla, T.A. Gene expression profile of aging and its retardation by caloric restriction. “old/aging” state of the molecular network from the young Science, 1999, 285(5432), 1390-3. state, what events cause the state transition and what are [8] Melov, S.; Hubbard, A. Microarrays as a tool to investigate the the network circuitry and epigenetic events locking the biology of aging: a retrospective and a look to the future. Sci Aging network in the aging state. Knowledge Environ, 2004, 2004(42), re7. [9] Lund, J.; Tedesco, P.; Duke, K.; Wang, J.; Kim, S.K.; Johnson, T.E. Transcriptional profile of aging in C. elegans. Curr Biol, 2002, CONFLICT OF INTEREST 12(18), 1566-73. [10] Pletcher, S.D.; Macdonald, S.J.; Marguerie, R.; Certa, U.; Stearns, The author(s) confirm that this article content has no S.C.; Goldstein, D.B.; Partridge, L. Genome-wide transcript conflicts of interest. profiles in aging and calorically restricted Drosophila melanogaster. Curr Biol, 2002, 12(9), 712-23. ACKNOWLEDGEMENTS [11] Kayo, T.; Allison, D.B.; Weindruch, R.; Prolla, T.A. Influences of aging and caloric restriction on the transcriptional profile of The authors are supported by grants from the China skeletal muscle from rhesus monkeys. Proc Natl Acad Sci U S A, Natural National Science Foundation (Grant #30890033 2001, 98(9), 5093-8. [12] Bronikowski, A.M.; Carter, P.A.; Morgan, T.J.; Garland, T., Jr.; Ung, and 91019019), Chinese Ministry of Science and Tech- N.; Pugh, T.D.; Weindruch, R.; Prolla, T.A. Lifelong voluntary nology (Grant #2011CB504206) and Chinese Academy of exercise in the mouse prevents age-related alterations in gene 564 Current Genomics, 2012, Vol. 13, No. 7 Hou et al. expression in the heart. Physiol Genomics, 2003, 12(2), 129-38. [31] Kojima, T.; Kamei, H.; Aizu, T.; Arai, Y.; Takayama, M.; [13] McCarroll, S.A.; Murphy, C.T.; Zou, S.; Pletcher, S.D.; Chin, C.S.; Nakazawa, S.; Ebihara, Y.; Inagaki, H.; Masui, Y.; Gondo, Y.; Jan, Y.N.; Kenyon, C.; Bargmann, C.I.; Li, H. Comparing genomic Sakaki, Y.; Hirose, N. Association analysis between longevity in expression patterns across species identifies shared transcriptional the Japanese population and polymorphic variants of genes profile in aging. Nat Genet, 2004, 36(2), 197-204. involved in insulin and insulin-like growth factor 1 signaling [14] Zahn, J.M.; Poosala, S.; Owen, A.B.; Ingram, D.K.; Lustig, A.; pathways. Exp Gerontol, 2004, 39(11-12), 1595-8. Carter, A.; Weeraratna, A.T.; Taub, D.D.; Gorospe, M.; Mazan- [32] Flachsbart, F.; Caliebe, A.; Kleindorp, R.; Blanche, H.; von Eller- Mamczarz, K.; Lakatta, E.G.; Boheler, K.R.; Xu, X.; Mattson, Eberstein, H.; Nikolaus, S.; Schreiber, S.; Nebel, A. Association of M.P.; Falco, G.; Ko, M.S.; Schlessinger, D.; Firman, J.; FOXO3A variation with human longevity confirmed in German Kummerfeld, S.K.; Wood, W.H., 3rd; Zonderman, A.B.; Kim, centenarians. Proc Natl Acad Sci U S A, 2009, 106(8), 2700-5. S.K.; Becker, K.G. AGEMAP: a gene expression database for [33] Anselmi, C.V.; Malovini, A.; Roncarati, R.; Novelli, V.; Villa, F.; aging in mice. PLoS Genet, 2007, 3(11), e201. Condorelli, G.; Bellazzi, R.; Puca, A.A. Association of the [15] Lu, T.; Pan, Y.; Kao, S.Y.; Li, C.; Kohane, I.; Chan, J.; Yankner, FOXO3A locus with extreme longevity in a southern Italian B.A. Gene regulation and DNA damage in the ageing human brain. centenarian study. Rejuvenation Res, 2009, 12(2), 95-104. Nature, 2004, 429(6994), 883-91. [34] Murphy, C.T.; McCarroll, S.A.; Bargmann, C.I.; Fraser, A.; [16] Somel, M.; Franz, H.; Yan, Z.; Lorenc, A.; Guo, S.; Giger, T.; Kamath, R.S.; Ahringer, J.; Li, H.; Kenyon, C. Genes that act Kelso, J.; Nickel, B.; Dannemann, M.; Bahn, S.; Webster, M.J.; downstream of DAF-16 to influence the lifespan of Caenorhabditis Weickert, C.S.; Lachmann, M.; Paabo, S.; Khaitovich, P. elegans. Nature, 2003, 424(6946), 277-83. Transcriptional neoteny in the human brain. Proc Natl Acad Sci U [35] Adler, A.S.; Sinha, S.; Kawahara, T.L.; Zhang, J.Y.; Segal, E.; S A, 2009, 106(14), 5743-8. Chang, H.Y. Motif module map reveals enforcement of aging by [17] Somel, M.; Guo, S.; Fu, N.; Yan, Z.; Hu, H.Y.; Xu, Y.; Yuan, Y.; continual NF-kappaB activity. Genes Dev, 2007, 21(24), 3244-57. Ning, Z.; Hu, Y.; Menzel, C.; Hu, H.; Lachmann, M.; Zeng, R.; [36] Budovskaya, Y.V.; Wu, K.; Southworth, L.K.; Jiang, M.; Tedesco, Chen, W.; Khaitovich, P. MicroRNA, mRNA, and protein P.; Johnson, T.E.; Kim, S.K. An elt-3/elt-5/elt-6 GATA expression link development and aging in human and macaque transcription circuit guides aging in C. elegans. Cell, 2008, 134(2), brain. Genome Res, 2010, 20(9), 1207-18. 291-303. [18] Kang, H.J.; Kawasawa, Y.I.; Cheng, F.; Zhu, Y.; Xu, X.; Li, M.; [37] Managbanag, J.R.; Witten, T.M.; Bonchev, D.; Fox, L.A.; Sousa, A.M.; Pletikos, M.; Meyer, K.A.; Sedmak, G.; Guennel, T.; Tsuchiya, M.; Kennedy, B.K.; Kaeberlein, M. Shortest-path Shin, Y.; Johnson, M.B.; Krsnik, Z.; Mayer, S.; Fertuzinhos, S.; network analysis is a useful approach toward identifying genetic Umlauf, S.; Lisgo, S.N.; Vortmeyer, A.; Weinberger, D.R.; Mane, determinants of longevity. PLoS One, 2008, 3(11), e3802. S.; Hyde, T.M.; Huttner, A.; Reimers, M.; Kleinman, J.E.; Sestan, [38] Bell, R.; Hubbard, A.; Chettier, R.; Chen, D.; Miller, J.P.; Kapahi, N. Spatio-temporal transcriptome of the human brain. Nature, P.; Tarnopolsky, M.; Sahasrabuhde, S.; Melov, S.; Hughes, R.E. A 2011, 478(7370), 483-9. human protein interaction network shows conservation of aging [19] Zhou, B.; Yang, L.; Li, S.; Huang, J.; Chen, H.; Hou, L.; Wang, J.; processes between human and invertebrate species. PLoS Genet, Green, C.D.; Yan, Z.; Huang, X.; Kaeberlein, M.; Zhu, L.; Xiao, 2009, 5(3), e1000414. H.; Liu, Y.; Han, J.D. Midlife gene expressions identify modulators [39] Budovsky, A.; Abramovich, A.; Cohen, R.; Chalifa-Caspi, V.; of aging through dietary interventions. Proc Natl Acad Sci U S A, Fraifeld, V. Longevity network: construction and implications. 2012, 109(19), E1201-9. Mech Ageing Dev, 2007, 128(1), 117-24. [20] Ibanez-Ventoso, C.; Yang, M.; Guo, S.; Robins, H.; Padgett, R.W.; [40] Xia, K.; Xue, H.; Dong, D.; Zhu, S.; Wang, J.; Zhang, Q.; Hou, L.; Driscoll, M. Modulated microRNA expression during adult Chen, H.; Tao, R.; Huang, Z.; Fu, Z.; Chen, Y.G.; Han, J.D. lifespan in Caenorhabditis elegans. Aging Cell, 2006, 5(3), 235-46. Identification of the proliferation/differentiation switch in the [21] de Lencastre, A.; Pincus, Z.; Zhou, K.; Kato, M.; Lee, S.S.; Slack, cellular network of multicellular organisms. PLoS Comput Biol, F.J. MicroRNAs both promote and antagonize longevity in C. 2006, 2(11), e145. elegans. Curr Biol, 2010, 20(24), 2159-68. [41] Xue, H.; Xian, B.; Dong, D.; Xia, K.; Zhu, S.; Zhang, Z.; Hou, L.; [22] Zid, B.M.; Rogers, A.N.; Katewa, S.D.; Vargas, M.A.; Kolipinski, Zhang, Q.; Zhang, Y.; Han, J.D. A modular network model of M.C.; Lu, T.A.; Benzer, S.; Kapahi, P. 4E-BP extends lifespan aging. Mol Syst Biol, 2007, 3, 147. upon dietary restriction by enhancing mitochondrial activity in [42] Miller, J.A.; Oldham, M.C.; Geschwind, D.H. A systems level Drosophila. Cell, 2009, 139(1), 149-60. analysis of transcriptional changes in Alzheimer's disease and [23] Rogers, A.N.; Chen, D.; McColl, G.; Czerwieniec, G.; Felkey, K.; normal aging. J Neurosci, 2008, 28(6), 1410-20. Gibson, B.W.; Hubbard, A.; Melov, S.; Lithgow, G.J.; Kapahi, P. [43] Wang, J.; Zhang, S.; Wang, Y.; Chen, L.; Zhang, X.S. Disease- Life Span Extension via eIF4G Inhibition Is Mediated by aging network reveals significant roles of aging genes in Posttranscriptional Remodeling of Stress Response Gene connecting genetic diseases. PLoS Comput Biol, 2009, 5(9), Expression in C. elegans. Cell Metab, 2011, 14(1), 55-66. e1000521. [24] de Magalhaes, J.P.; Curado, J.; Church, G.M. Meta-analysis of age- [44] Han, J.D.; Dupuy, D.; Bertin, N.; Cusick, M.E.; Vidal, M. Effect of related gene expression profiles identifies common signatures of sampling on topology predictions of protein-protein interaction aging. Bioinformatics, 2009, 25(7), 875-81. networks. Nat Biotechnol, 2005, 23(7), 839-44. [25] Johnson, T.E. Increased life-span of age-1 mutants in [45] Liu, B.; Zhou, Z. Chromatin remodeling, DNA damage repair and Caenorhabditis elegans and lower Gompertz rate of aging. Science, aging. Curr Genomics, 2012, This issue. 1990, 249(4971), 908-12. [46] Feinberg, A.P. Epigenetics at the epicenter of modern medicine. [26] Lin, K.; Dorman, J.B.; Rodan, A.; Kenyon, C. daf-16: An HNF- Jama, 2008, 299(11), 1345-50. 3/forkhead family member that can function to double the life-span [47] Tollefsbol, T.O. Epigenetics of Aging. Springer, 2010. of Caenorhabditis elegans. Science, 1997, 278(5341), 1319-22. [48] Love, D.C.; Ghosh, S.; Mondoux, M.A.; Fukushige, T.; Wang, P.; [27] Bennett, C.F.; Yanos, M.; Kaeberlein, M. Genome-wide RNAi Wilson, M.A.; Iser, W.B.; Wolkow, C.A.; Krause, M.W.; Hanover, longevity screens in Caenorhabditis elegans. Curr Genomics, 2012, J.A. Dynamic O-GlcNAc cycling at promoters of Caenorhabditis This issue. elegans genes regulating longevity, stress, and immunity. Proc Natl [28] Hamilton, B.; Dong, Y.; Shindo, M.; Liu, W.; Odell, I.; Ruvkun, Acad Sci U S A, 2010, 107(16), 7413-8. G.; Lee, S.S. A systematic RNAi screen for longevity genes in C. [49] Hernandez, D.G.; Nalls, M.A.; Gibbs, J.R.; Arepalli, S.; van der elegans. Genes Dev, 2005, 19(13), 1544-55. Brug, M.; Chong, S.; Moore, M.; Longo, D.L.; Cookson, M.R.; [29] Hansen, M.; Hsu, A.L.; Dillin, A.; Kenyon, C. New genes tied to Traynor, B.J.; Singleton, A.B. Distinct DNA methylation changes endocrine, metabolic, and dietary regulation of lifespan from a highly correlated with chronological age in the human brain. Hum Caenorhabditis elegans genomic RNAi screen. PLoS Genet, 2005, Mol Genet, 2011, 20(6), 1164-72. 1(1), 119-28. [50] Li, J.; Ebata, A.; Dong, Y.; Rizki, G.; Iwata, T.; Lee, S.S. [30] Li, Y.; Wang, W.J.; Cao, H.; Lu, J.; Wu, C.; Hu, F.Y.; Guo, J.; Zhao, Caenorhabditis elegans HCF-1 functions in longevity maintenance L.; Yang, F.; Zhang, Y.X.; Li, W.; Zheng, G.Y.; Cui, H.; Chen, X.; as a DAF-16 regulator. PLoS Biol, 2008, 6(9), e233. Zhu, Z.; He, H.; Dong, B.; Mo, X.; Zeng, Y.; Tian, X.L. Genetic [51] Siebold, A.P.; Banerjee, R.; Tie, F.; Kiss, D.L.; Moskowitz, J.; association of FOXO1A and FOXO3A with longevity trait in Han Harte, P.J. Polycomb Repressive Complex 2 and Trithorax Chinese populations. Hum Mol Genet, 2009, 18(24), 4897-904. modulate Drosophila longevity and stress resistance. Proc Natl Systems Biology in Aging: Linking the Old and the Young Current Genomics, 2012, Vol. 13, No. 7 565 Acad Sci U S A, 2010, 107(1), 169-74. N.; Khaitovich, P.; Chen, C.D.; Zhang, H.; Jenuwein, T.; Han, J.D. [52] Peleg, S.; Sananbenesi, F.; Zovoilis, A.; Burkhardt, S.; Bahari- Histone Demethylase UTX-1 Regulates C. elegans Life Span by Javan, S.; Agis-Balboa, R.C.; Cota, P.; Wittnam, J.L.; Gogol- Targeting the Insulin/IGF-1 Signaling Pathway. Cell Metab, 2011, Doering, A.; Opitz, L.; Salinas-Riester, G.; Dettenhofer, M.; Kang, 14(2), 161-172. H.; Farinelli, L.; Chen, W.; Fischer, A. Altered histone acetylation [58] Strahl, B.D.; Allis, C.D. The language of covalent histone is associated with age-dependent memory impairment in mice. modifications. Nature, 2000, 403(6765), 41-5. Science, 2010, 328(5979), 753-6. [59] Jenuwein, T.; Allis, C.D. Translating the histone code. Science, [53] Graff, J.; Rei, D.; Guan, J.S.; Wang, W.Y.; Seo, J.; Hennig, K.M.; 2001, 293(5532), 1074-80. Nieland, T.J.; Fass, D.M.; Kao, P.F.; Kahn, M.; Su, S.C.; Samiei, [60] Tan, M.; Luo, H.; Lee, S.; Jin, F.; Yang, J.S.; Montellier, E.; A.; Joseph, N.; Haggarty, S.J.; Delalle, I.; Tsai, L.H. An epigenetic Buchou, T.; Cheng, Z.; Rousseaux, S.; Rajagopal, N.; Lu, Z.; Ye, blockade of cognitive functions in the neurodegenerating brain. Z.; Zhu, Q.; Wysocka, J.; Ye, Y.; Khochbin, S.; Ren, B.; Zhao, Y. Nature, 2012, 483(7388), 222-6. Identification of 67 histone marks and histone lysine crotonylation [54] Dang, W.; Steffen, K.K.; Perry, R.; Dorsey, J.A.; Johnson, F.B.; as a new type of histone modification. Cell, 2011, 146(6), 1016-28. Shilatifard, A.; Kaeberlein, M.; Kennedy, B.K.; Berger, S.L. [61] Yu, H.; Zhu, S.; Zhou, B.; Xue, H.; Han, J.D. Inferring causal Histone H4 lysine 16 acetylation regulates cellular lifespan. Nature, relationships among different histone modifications and gene 2009, 459(7248), 802-7. expression. Genome Res, 2008, 18(8), 1314-24. [55] Oberdoerffer, P.; Michan, S.; McVay, M.; Mostoslavsky, R.; Vann, [62] de Magalhaes, J.P.; Costa, J.; Toussaint, O. HAGR: the Human J.; Park, S.K.; Hartlerode, A.; Stegmuller, J.; Hafner, A.; Loerch, Ageing Genomic Resources. Nucleic Acids Res, 2005, 33(Database P.; Wright, S.M.; Mills, K.D.; Bonni, A.; Yankner, B.A.; Scully, issue), D537-43. R.; Prolla, T.A.; Alt, F.W.; Sinclair, D.A. SIRT1 redistribution on [63] de Magalhaes, J.P.; Budovsky, A.; Lehmann, G.; Costa, J.; Li, Y.; chromatin promotes genomic stability but alters gene expression Fraifeld, V.; Church, G.M. The Human Ageing Genomic during aging. Cell, 2008, 135(5), 907-18. Resources: online databases and tools for biogerontologists. Aging [56] Greer, E.L.; Maures, T.J.; Hauswirth, A.G.; Green, E.M.; Leeman, Cell, 2009, 8(1), 65-72. D.S.; Maro, G.S.; Han, S.; Banko, M.R.; Gozani, O.; Brunet, A. [64] Enright, A.J.; Van Dongen, S.; Ouzounis, C.A. An efficient Members of the H3K4 trimethylation complex regulate lifespan in algorithm for large-scale detection of protein families. Nucleic a germline-dependent manner in C. elegans. Nature, 2010, Acids Res, 2002, 30(7), 1575-84. 466(7304), 383-7. [65] Wu, G.; Feng, X.; Stein, L. A human functional protein interaction [57] Jin, C.; Li, J.; Green, C.D.; Yu, X.; Tang, X.; Han, D.; Xian, B.; network and its application to cancer data analysis. Genome Biol, Wang, D.; Huang, X.; Cao, X.; Yan, Z.; Hou, L.; Liu, J.; Shukeir, 2010, 11(5), R53.