Physiol. Genomics AJP citation statistics
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Physiol. Genomics 27: 187-200, 2006. First published August 1, 2006; doi:10.1152/physiolgenomics.00084.2006
1094-8341/06 $8.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Supplemental Figures and Tables
Right arrow All Versions of this Article:
27/3/187    most recent
00084.2006v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (5)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Selman, C.
Right arrow Articles by Withers, D. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Selman, C.
Right arrow Articles by Withers, D. J.
Received 10 May 2006; accepted in final form 28 July 2006.
Physiological Genomics 27:187-200 (2006)
1094-8341/06 $8.00 © 2006 American Physiological Society

Call For Papers: Comparative Genomics

Coordinated multitissue transcriptional and plasma metabonomic profiles following acute caloric restriction in mice

Colin Selman1,*, Nicola D. Kerrison2,*, Anisha Cooray3, Matthew D. W. Piper4, Steven J. Lingard1, Richard H. Barton3, Eugene F. Schuster2, Eric Blanc2, David Gems4, Jeremy K. Nicholson3, Janet M. Thornton2, Linda Partridge4 and Dominic J. Withers1

1 Centre for Diabetes and Endocrinology, Department of Medicine, University College London, Rayne Institute, London
2 European Molecular Biology Laboratory (EMBL)-European Bioinformatics Institute (EBI), Wellcome Trust Genome Campus, Hinxton, Cambridge
3 Biological Chemistry Section, Department of Biomolecular Medicine, Division of Surgery, Oncology, Reproductive Biology and Anaesthetics, Faculty of Medicine, Imperial College London, South Kensington, London
4 UCL Centre for Ageing Research, Department of Biology, University College London, London, United Kingdom


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Caloric restriction (CR) increases healthy life span in a range of organisms. The underlying mechanisms are not understood but appear to include changes in gene expression, protein function, and metabolism. Recent studies demonstrate that acute CR alters mortality rates within days in flies. Multitissue transcriptional changes and concomitant metabolic responses to acute CR have not been described. We generated whole genome RNA transcript profiles in liver, skeletal muscle, colon, and hypothalamus and simultaneously measured plasma metabolites using proton nuclear magnetic resonance in mice subjected to acute CR. Liver and muscle showed increased gene expressions associated with fatty acid metabolism and a reduction in those involved in hepatic lipid biosynthesis. Glucogenic amino acids increased in plasma, and gene expression for hepatic gluconeogenesis was enhanced. Increased expression of genes for hormone-mediated signaling and decreased expression of genes involved in protein binding and development occurred in hypothalamus. Cell proliferation genes were decreased and cellular transport genes increased in colon. Acute CR captured many, but not all, hepatic transcriptional changes of long-term CR. Our findings demonstrate a clear transcriptional response across multiple tissues during acute CR, with congruent plasma metabolite changes. Liver and muscle switched gene expression away from energetically expensive biosynthetic processes toward energy conservation and utilization processes, including fatty acid metabolism and gluconeogenesis. Both muscle and colon switched gene expression away from cellular proliferation. Mice undergoing acute CR rapidly adopt many transcriptional and metabolic changes of long-term CR, suggesting that the beneficial effects of CR may require only a short-term reduction in caloric intake.

aging; microarray; gene expression


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
CALORIC RESTRICTION (CR) extends mean and maximum life span and retards various age-associated pathologies in a range of vertebrate and invertebrate species (24, 41, 46, 61). However, the mechanisms through which CR extends life span are not understood. CR is accompanied by changes in gene and protein expression and resultant alterations in both cellular and whole organism metabolism that are thought to underlie its effects on aging.

RNA transcript profiles have been useful for defining changes in gene expression following CR, leading to the formulation of experimentally testable hypotheses. Long-term CR in rodents attenuates many age-associated transcriptional changes, including expression of genes involved with cellular stress, in skeletal muscle (36), hypothalamus (20), and liver (11, 16). However, to date, most microarray mRNA expression studies in mice have been confined to single tissues and have profiled only part of the genome. In addition, single-tissue studies have varied widely in experimental design, bioinformatic analysis, and mouse strains used (29), thus making direct comparisons between tissues difficult. Furthermore, altered transcript profiles have not been related to concomitant plasma metabolic changes.

Changes in gene expression following initiation of CR (or dietary restriction; DR) appear rapid in both Drosophila and mice. Significant shifts in transcript profile in Drosophila occurred within 3 days following DR (47). In studies of hepatic gene expression in mice, a rapid and cumulative shift mirroring a long-term (12 mo) CR transcriptional profile was observed after 2, 4, or 8 wk of CR. A return to ad libitum (AD) levels reversed 90% of this long-term CR profile after 8 wk (16). This suggests that these changes in gene expression during acute CR may induce rapid beneficial effects on life span. In Drosophila, a dietary shift to DR altered mortality rate within days (40). However, accurate measurements of mortality rates during acute CR have not been measured in mice, although CR initiated at 12 mo of age extended mean and maximum life span, decreased age-associated mortality, and reduced cancer incidence (16, 49).

In this study, we examined acute whole genome transcriptional changes following short-term CR in liver, skeletal muscle, hypothalamus, and colon of 16-wk-old male C57BL/6 mice subjected to a stepwise 16-day protocol of CR terminating in a 48-h period of 30% CR (see supplemental data; Supplemental Materials are available in the online version of this article). In parallel, we employed a nuclear magnetic resonance (NMR) metabonomic strategy capable of identifying subtle biochemical changes (56) to determine plasma metabolite profiles during CR in the same animals subjected to transcriptional analysis, as plasma should relate directly to global metabolic changes in the concentration of metabolites. Our experimental design therefore permitted simultaneous examination of tissue-specific and common changes in gene expression together with relevant metabolic end-points.

We show that both liver and muscle switch away from energetically expensive biosynthetic processes toward fatty acid metabolism, ß-oxidation, and gluconeogenesis as the dominant energy pathways during acute CR. This enhanced gluconeogenesis appears fuelled by skeletal muscle (and perhaps colon) catabolism, with genes associated with growth downregulated in both tissues, and an associated increase in plasma of several glucogenic amino acids. A simultaneous decrease in plasma levels of glucose and several lipids was also observed.


   METHODS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 
Study Design
Twenty male C57BL/6 mice were purchased from a commercial breeder (Harlan) at 4 wk of age and housed at ~22°C under a 12:12-h light-dark cycle (lights on from 0700 to 1900; Imperial College, London, UK). Mice were maintained in groups of five (see Ref. 28) under specific pathogen-free conditions within individually ventilated cages (Techniplast), with AD access to normal chow (2018 Teklad Global 18% Protein Rodent Diet, Harlan Teklad) and water. The manufacturers (Harlan Teklad) report that, by proximate analysis, the chow diet consists of crude protein (18.90%), crude oil (5.70%), crude fiber (3.80%), ash (5.90%), nitrogen-free extract (55.70%), carbohydrate (57.33%), starch (41.24%), and sugar (4.93%) plus predefined minerals, amino acids, vitamins, and fatty acids. The main saturated, mono-unsaturated, and polyunsaturated fatty acid components (g/kg diet) of the diet were, respectively, palmitic acid C16:0 (7.64 g/kg), oleic acid C18:1w9 (12.59 g/kg), and linoleic acid C18:2w6 (31.35 g/kg). The total levels of digestible and metabolizable energy in the diet were 3.4 and 3.3 kcal/g, respectively. At 12 wk of age, mice in different cages did not significantly differ from one another in body mass, and each cage was randomly assigned to the CR or AD feeding regime. The experiment commenced at 14 wk of age, with the time of feeding of the CR mice being 0900 each day, in agreement with feeding time in other studies (e.g., Ref. 16). Food was not removed from either the CR or the AD mice at any time during the day, with the hoppers of the CR mice always being empty by 0900 of the following day. Food intake of CR mice was adjusted according to the intake of AD mice measured the preceding week. CR mice underwent a step-down regime (see supplemental data) similar to protocols described elsewhere (43), with daily food intake reduced to 90% of AD mice at 14 wk, 80% at 15 wk, and 70% at 16 wk of age, i.e., 30% CR relative to AD controls. CR mice were not fed in the morning immediately before death. Neither group underwent a 24-h (16) or 48-h (11) fast before death, as we wished to avoid inducing acute starvation conditions that may transcriptionally mirror those seen during CR (8). Exactly 48 h after the initiation of 30% CR, 10 mice per group (CR or AD; CR10 or AD10, respectively) were killed by terminal anesthesia (ip injection of fentanyl-fluanisone; Hypnorm, Janssen Animal Health) and benzodiazepine mix (Midazolam, Roche) between 0900 and 1200. Individual mice in each group were killed alternately to minimize any circadian effects, and the left lobe of liver, the gastrocnemius, the hypothalamus (block containing dorsomedial, ventromedial, arcuate, and paraventricular hypothalamus), and a 2-cm portion of the upper colon were removed, flash-frozen in liquid nitrogen, and stored at –80°C until RNA extraction. To minimize the potential for hierarchies to develop within our group-housed mice, we purchased mice that had been weaned together from 4 wk of age. These same mice were then maintained in groups of five until death. In addition, all large food pellets were broken into smaller pieces to further avoid the potential for any one mouse to monopolize the food supply. We observed no evidence of hierarchies in our group-housed CR or AD mice, either in terms of body mass differences within cages or in terms of fighting. All mice were monitored daily, and, in agreement with Ikeno et al. (28), all CR mice were able to feed simultaneously at the hopper with no evidence that any one mouse interfered with the feeding behavior of any other mouse within a cage. No signs of pathology were observed in any mice, and all procedures followed local ethical guidelines and were conducted under licence from the United Kingdom Home Office.

Tissue Preparation for DNA Microarray Analysis
Individual samples were pooled in triplicate (AD mice or CR mice 1–3, 4–6, and 7–9) before RNA extraction with the remaining sample (AD10 or CR10) equally distributed among each of the three groups, with ~100 mg (Ohaus, 0.0001 g) of tissue taken from each individual. The hypothalamus weighed ~100 mg; therefore, samples from individual AD10/CR10 were divided equally among the three pooled groups. Several recent studies suggest that pooling individual samples gives an accurate measure of the average expression (30, 31). The pooling strategy adopted was partially based on the methods suggested by Kendziorski et al. (30) and was employed to help reduce the effects of biological variability, i.e., to maximize our ability to detect the effects of experimental treatment. Samples were prepared according to Affymetrix protocols (Affymetrix, http://www.affymetrix.com). In brief, samples were homogenized on ice by use of a Positron homogenizer in TRIzol reagent (Invitrogen), followed by purification using RNeasy columns (Qiagen, http://www1.qiagen.com), with RNA quality and concentration determined using an Agilent Bioanalyzer 2100 (Agilent Technologies).

Microarray Measurement of Gene Expression
Changes in transcript abundance were measured using mouse whole genome oligonucleotide microarrays (Mouse Genome 430 2.0 Array, Affymetrix), with all protocols undertaken at the Institute of Child Health (ICH) Gene Microarray Centre, University College London, UK. Three biological replicates were analyzed for each condition (acute CR or AD control). The cRNA probe generation, washing, labeling, and scanning followed Affymetrix and ICH Gene Microarray Centre standard protocols.

Statistical Analysis of Transcript Representation
Raw image files were converted to probe-level data files (.cel files). Normalization, statistical testing, and multiple testing correction of data files used the R (http://www.R-project.org) package Goldenspike (13). A false discovery rate cutoff of 0.05 was used to generate lists of genes whose expression was significantly altered by CR for all tissues except colon, as a q-value cutoff of 0.05 resulted in only 22 up- and 11 downregulated probe sets. Because changes in the colon were more subtle than in other tissues, the stringency was lowered to reflect this. We chose the value of 0.11, an optimal point in the sense of maximizing the gene list length while keeping the q-value cutoff as low as possible. Probes were mapped to transcript sequences from the Ensembl mouse genome (version 27.33c.1). Acceptable matches were defined as having at least 7 of the 11 probes in the probe set matching the gene transcript with no more than 2 mismatches, and no other gene matched by >3 probes in the probe set. This procedure mapped ~47% of the probe sets on the mouse genome 430 2.0 array. Gene Ontology (Gene Ontology Consortium), Interpro, and Locuslink annotation for the genes was retrieved from Ensembl. Kegg annotation was mapped to genes via Locuslink identifiers. These annotations, together with expression analysis systematic explorer (EASE) (25), were used to identify overrepresented annotation categories within the gene lists and scored using the EASE score, and significance was calculated using a modified Fisher’s exact test. For analyses not involving EASE, unmapped probe sets were annotated with information from Affymetrix’s NetAffx site (39). In our EASE comparison with the long-term CR data set (16), we reported those genes that mapped directly to Ensembl gene lists. Raw microarray data are available in the EBI (EMBL-EBI) ArrayExpress repository (accession no. E-MEXP-748).

In our tissue comparative approach, we recorded the number of differentially expressed probe sets in each tissue and the number of probe sets differentially expressed in several tissues. The significance of overlaps between tissues in probe sets differentially expressed during CR was tested with the null hypothesis that each probe was randomly and independently selected from the pool of all probe sets represented on the microarray. Using Fisher’s exact test, we calculated the probability of an overlap occurring by chance. The probabilities calculated must be treated with some caution because, in reality, genes are not randomly and independently expressed. However, these probabilities provide an indication of similarities between responses of tissues to CR. We examined overlaps of EASE-generated lists overrepresented (P < 0.1) from the differentially expressed probe sets. Probability estimates were adjusted for multiple testing using a Bonferroni correction, with adjusted scores of P < 0.05 deemed significant.

Metabonomic Sample Preparation, 1H-NMR Spectroscopy, and Multivariate Data Analysis
Blood was collected under terminal anesthesia by cardiac puncture and centrifuged (8,000 rpm for 15 min at 4°C), and the resulting plasma was stored at –80°C until analysis. Two hundred-microliter aliquots of each sample were transferred into 5-mm NMR tubes and diluted with 300 µl of extender solution, comprised of isotonic saline (0.9% wt/vol), 3 mM sodium azide, and 20% (vol/vol) D2O. 1H-NMR spectroscopy and data reduction are described in detail elsewhere (9). All resulting 1H-NMR spectra were manually phased and baseline corrected using XWINNMR software (Bruker Biospin), and the glucose anomeric proton signal at 5.23 parts/min (ppm) was used as a frequency reference to calibrate the position of spectra. Subsequent automatic data reduction and normalization to unit total for each spectral integral were then carried out (9), and the reduced and normalized NMR spectral data were exported to SIMCA-P (version 10.5; Umetrics, Umeå, Sweden) and mean centered without further scaling. 1H-NMR spectroscopy generates large data sets containing a range of metabolic information, so to extract meaningful and clear information related to pathophysiological stimuli, data sets were simplified using principal component analysis (PCA) (63) and partial least squares (PLS) (64) methods. It is important to note that these data report concentrations of metabolites in plasma and do not represent measurements of metabolic flux.

Microarray Data Validation by Real-Time PCR
Real-time PCR (RT-PCR) was performed as previously described (14). TaqMan gene expression assays (symbol, assay ID) for Forkhead box 01/Forkhead transcription factor-1 (Foxo1, Mm00490672_m1), peroxisome proliferator-activated receptor-{alpha} (Ppara, Mm00440939_m1), DNA damage-inducible transcript-4 (Ddit4, Mm00512503_g1), diacylglycerol-O-acyltransferase-2 (Dgat2, Mm00499530_m1), and insulin receptor substrate-1 (Irs1, Mm00439720_s1) were run in duplicate. Transcription elongation factor A (SII)-1 (Tcea1, Mm00815387_s1) was amplified as a control, as its expression was unaffected by CR in both this study and in others (e.g., Ref. 16). We tested 3 pooled liver samples, as described earlier, in both CR and AD mice and an additional 20 unpooled samples consisting of the 10 individual mice per group (CR1:10 and AD1:10) from which our pools were derived. We employed the comparative threshold cycle (CT) method, with fold changes calculated according to the following expression: 2–{Delta}{Delta}CT. P values refer to a one-tailed t-test between normalized CT values (CT target gene – CT Tcea1) of CR vs. AD mice (1).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
DISCUSSION
 GRANTS
 REFERENCES
 
Acute CR Reduced Body Mass
CR mice showed a 12% reduction in body mass (mean ± SE; CR 26.5 ± 0.6 g, control 30.0 ± 0.6 g, F = 16.2, P < 0.01) compared with AD mice. Body mass during the experiment (final body mass – initial body mass before CR) increased in AD mice by 1.1 ± 0.3 g and decreased in CR mice by 1.4 ± 0.2 g. Every single mouse in the CR group (n = 10) decreased in body mass during the period of CR. One mouse from the AD group (n = 10) decreased in mass (–0.4 g) during this same period of time.

Acute CR Altered Plasma Metabolite Profile
NMR spectroscopy clearly distinguished the plasma metabolite profiles of acute CR and AD control mice (Fig. 1; Table 1). The Euclidean sum of principal components 1 and 3 in Table 1 indicates the direction and relative change in plasma metabolites between the CR and AD groups, as seen by the position of chemical shift variables in the PLS-discriminant analysis loadings scatter plot. These results indicate increases in plasma lactate, 3-hydroxybutyrate, cholesterol, and associated low-density lipoprotein levels under acute CR, in addition to increased plasma creatine and several glucogenic amino acids (methionine, glutamine, alanine, valine), suggesting a metabolic switch toward energy conservation and gluconeogenesis. In agreement, plasma concentrations of glucose and choline were decreased during acute CR, as were the resonances of lipids including very low-density lipoproteins (VLDL; 1.25–1.29 ppm). The results indicate that the magnitude of change in some plasma metabolites is greater than for others. For example, relative to AD controls, the increase observed in plasma levels of lactate in CR mice is greater than the increase in plasma alanine (Table 1).


Figure 1
View larger version (25K):
[in this window]
[in a new window]

  		Fig. 1. Partial least squares-discriminant analysis (PLS-DA) of plasma samples. Initial principal component analysis (PCA) separated plasma samples according to class [ad libitum (AD) and acute caloric restriction (CR)], so PLS-DA was used to maximize this separation and identify differences in metabolites (chemical shifts) that are the basis of this class separation. The score scatter plot (A) illustrates that the CR replicates (red) were well resolved from the AD replicates (black). The major variables responsible for class separation are revealed by the PLS-DA loadings scatter plot (B). This represents the chemical shift regions important to the intergroup differences within the discriminating model, and the relative magnitude of these differences in plasma metabolites between the 2 experimental groups. The metabolites responsible for the separation of classes are summarized in Table 1. Group separation was primarily explained by principal component 1 (PC1), while PC3 explained primarily within-group variation, especially within the CR group. PC2 also showed class separation, but is not reported, as it was based solely on differences in lactate between groups.

 

View this table:
[in this window]
[in a new window]

  		Table 1. Plasma metabolite profile

 
Acute CR Significantly Altered Transcript Profiles in Individual Tissues
A total of 1,960 probe sets (see supplemental data) showed significantly altered expression in liver (987 higher and 973 lower in CR relative to AD controls, respectively) as did 520 (213 higher and 307 lower) in skeletal muscle, 309 (152 higher and 157 lower) in colon, and 145 (64 higher and 81 lower) in the hypothalamus. The biological relevance of these changes was assigned using EASE analysis (Table 2 and supplemental data; see GoFig. 3 for summary).


View this table:
[in this window]
[in a new window]

  		Table 2. Main overrepresented tissue-specific gene classes determined by EASE

 

Figure 2
View larger version (36K):
[in this window]
[in a new window]

  		Fig. 2. Main tissue-specific expression analysis systematic explorer (EASE) categories. Overview of the main tissue-specific EASE categories altered following acute CR. Upregulated gene classes are indicated in red and downregulated genes in black. Note that the organs depicted are human for ease of recognition. No implication is made about the validity of these results in humans.

 

Figure 3
View larger version (14K):
[in this window]
[in a new window]

  		Fig. 3. No. of shared genes up- and downregulated across tissues. Comparative tissue response to acute CR. Upregulated genes are indicated in red and downregulated genes in black. Significant relationships determined by overlaps in probe set lists across tissues using Fishers exact probability test. No. denotes amount of genes shared between tissues. P < 0.001 denoted by dotted line. P < 0.0001 denoted by solid line.

 
Acute CR effects on liver.
A metabolic switch toward gluconeogenesis, ß-oxidation, and fatty acid metabolism and away from lipogenesis occurred in the liver (Table 2; Fig. 2). During acute CR, genes involved in macromolecular biosynthetic processes, such as steroid, sterol, and cholesterol biosynthesis but also amino acid, amino acid derivative, and amine metabolism, were downregulated. EASE categories in the upregulated list included lipid metabolism, palmitoyl-CoA hydrolase activity, and acyl-CoA metabolism, with genes identified including Mte1, Cte1, Acadl (initiating gene of ß-oxidation), Ppara (Ppar-{alpha}), and Hnf-4{alpha}. Pgc-1{alpha} expression was upregulated during acute CR, and this protein plays a key role in gluconeogenesis and ß-oxidation, coactivates Ppar-{alpha} and Hnf-4{alpha}, and induces lipolysis. Pgc-1{alpha} also modulates energetically expensive biological processes, such as hepatic gluconeogenesis, by stimulating mitochondrial biogenesis and oxidative metabolism. This enhancement of mitochondrial biogenesis and oxidative metabolism was further suggested by the identification of the EASE category mitochondrion, which included the genes succinate dehydrogenase (complex II), cytochrome-c oxidase (complex IV), several ATP synthases, carnitine carrier/transporter proteins, and the mitochondrial transcription factor Tfam.

Acute CR effects on skeletal muscle.
A reduction in expression of genes involved in muscle growth and an increase in those involved in fatty acid metabolism occurred in skeletal muscle following acute CR (Table 2, Fig. 2). Downregulated EASE classes included genes associated with protein metabolism, protein transport, cell growth regulation, blood vessel development, collagen and macromolecule biosynthesis. In the upregulated gene list, EASE categories involved in energy metabolism, including carbohydrate metabolism, carboxylic acid metabolism and fatty acid metabolism, were identified. Gene classes involved in MAPK signaling and the JNK pathway were upregulated, and these are important in growth and development.

Acute CR effects on colon.
A decreased expression of genes associated with cellular proliferation but an increase in cellular transport genes was observed in colon following acute CR (Table 2; Fig. 2). EASE categories associated with cell growth and cell maintenance and proliferation were downregulated in the colon following acute CR, including several cyclin-dependent protein serine kinases and serine/threonine protein kinases implicated in tumor formation. In the upregulated gene list, EASE categories of ion transport, rhythmic behavior, and nitric oxide-mediated signal transduction were identified. The category transport/ion transport, which included the intestinal-specific epithelial glucose transporter Glut5 (Slc2a5) and Klf15, suggested that transport of metabolic intermediates is enhanced in the colon during acute CR.

Acute CR effects on hypothalamus.
In the hypothalamus, relatively few EASE categories were identified as being significantly altered following acute CR (Table 2, Fig. 2). The gene category development was identified from our downregulated gene list and included the genes Ccnd2 (cyclin D2), and forkhead-related brain factor Bf-1. Dgkz (diacylglycerol kinase, zeta) expression was also down and may play a role in mammalian target of rapamycin (mTOR) signaling via phosphatidic acid production, which subsequently induces p70S6K phosphorylation. The category binding was also identified, with several genes involved in protein and calcium binding downregulated. In the upregulated gene list, there were four significant overrepresented categories: hormone-mediated signaling, Zn-finger, basement membrane processes, and binding, in particular nucleic acid binding.

Significant EASE Categorical Overlap Across Tissues in Response to Acute CR
We tested the significance of tissue overlaps between our differentially expressed probe set lists using Fisher’s exact test. Clear biological similarities in our upregulated (up in CR) probe set lists were identified (Fig. 3), with significant overlap among liver, colon, and muscle and between hypothalamus and liver. In downregulated probe set lists, liver, colon, and hypothalamus showed significant overlap with one another, with muscle showing overlap only with colon. The greatest overlap (50) in probe sets was between the downregulated muscle and the upregulated liver probe set list. Three significant EASE categories, containing two genes in each, were identified as upregulated in liver and downregulated in muscle, including a protein carrier EASE category (NM_026775, NM_028876). The EASE category 14-3-3 protein contained the genes Ywhae and Ywhag. Ywhae increases resistance to apoptosis via Bcl-2 inhibition, with decreased resistance to apoptosis, potentially being important in muscle catabolism observed during CR. The final significant EASE category was ubiquitin-activating enzyme repeat, containing Ube1c (Nedd8 activating) and Uble1b (Sumo-1 activating). Ube1c inhibits steroid receptor-induced transcription, with hepatic steroid biosynthesis decreased during acute CR. Both are important in cell cycle progression and morphogenesis, possibly explaining decreased expression in muscle during acute CR.

No shared EASE categories were identified across the four tissues in our downregulated gene lists. However, several were identified in our upregulated gene lists (Table 3), including transporter activity, ion homeostasis, amino transferase, and zinc ion homeostasis/nitric oxide-mediated signal transduction shared across liver and colon. A simultaneous and enhanced requirement for fatty acid metabolism was seen in liver and muscle, with categories including acyl-CoA metabolism, acyl-CoA thioesterase activity, and palmitoyl-CoA hydrolase activity (Table 3) identified following acute CR. Our findings demonstrate that shared and common transcriptional responses occur across tissues following acute CR, particularly between liver and colon, and between liver and muscle.


View this table:
[in this window]
[in a new window]

  		Table 3. Main overrepresented EASE categories in the upregulated gene list when comparing across tissues following acute CR

 
Overlap Between Tissues in Gene Expression in Response to Acute CR
Fifty genes were significantly downregulated in two or more tissues (see supplemental data) following acute CR. Several genes involved in chaperone activity/stress response were downregulated in two or more tissues, particularly in colon and liver, as were several immune response genes. Seventy-nine genes were upregulated following acute CR across two or more tissues, although none were upregulated across all four tissues (see supplemental data), with the majority shared between colon and liver (30) or muscle and liver (27).

Similar Transcriptional Effects of Acute and Long-Term (27 Mo) CR
We compared our acute hepatic transcriptional data with long-term CR hepatic data derived from 24-h fasted male B6C3F1 mice, which were group housed, maintained on a defined diet, and undergoing a long-term (27 mo) 44% decrease in caloric intake (16). Significant overlap in EASE categories (Fig. 4) between acute and long-term CR was observed, with the likelihood of overlaps occurring by chance (Fisher’s exact probability test) being 3 x 10–19 for the upregulated and 8 x 10–9 for the downregulated gene categories. Eleven of forty-six significant EASE categories identified and downregulated by long-term CR were shared with acute CR, including lipid metabolism, endoplasmic reticulum, and cytochrome P450 (E class groups I and IV). In the upregulated gene lists, 20 EASE categories of 42 significantly overrepresented in long-term CR were shared with acute CR, including mitochondrion, electron transport, carboxylic acid metabolism, and fatty acid metabolism.


Figure 4
View larger version (13K):
[in this window]
[in a new window]

  		Fig. 4. Comparison of EASE categories generated from long-term and acute CR. The no. of significant (P < 0.1) shared and unique hepatic EASE categories generated from the upregulated and downregulated gene lists from acute CR (this study) and long-term CR (16). Upregulated EASE classes are shown in red and downregulated EASE categories in black.

 
EASE Classes Identified Following Long-Term CR Only
Twenty-two EASE categories were overrepresented and upregulated following long-term CR but not during acute CR, including amino acid catabolism, aromatic compound catabolism, L-phenylalanine metabolism, L-phenylalanine catabolism, and oxidoreductase activity. Thirty-five categories were downregulated following long-term but not short-term CR, including biosynthesis of secondary metabolites, proteasome {alpha}-subunit, lipocalin-related protein, odorant and pheromone binding.

EASE Classes Identified Following Acute CR Only
EASE categories changing under acute CR, but not long-term CR, were more numerous, with 97 downregulated and 128 upregulated. Downregulated categories included lipid biosynthesis, glutathione S-transferase (mu class), tRNA metabolism, and tRNA ligase activity, and upregulated categories included cell/ion homeostasis, glycerol/triacylglycerol metabolism, acyl-CoA metabolism, fatty acid ß-oxidation, and rhythmic behavior.

Discordant Effects of Long-Term and Acute CR on EASE Classes
Categories for short-chain dehydrogenase/reductase, glucose/ribitol dehydrogenase, and B class cytochrome P450 were upregulated during acute CR but downregulated during long-term CR, including B class cytochrome P450s and short-chain dehydrogenases/reductases. Gene categories associated with amino acid, aromatic compound, and amine metabolism showed upregulation following long-term CR but downregulation following acute CR.

Validation of Microarray Results by RT-PCR
RT-PCR, using five individual genes, was used to independently verify some of our microarray results in each of the three pooled liver samples derived from our CR and AD mice. All genes reliably produced significant (P ≤ 0.05) expression changes (Table 4), with a strong correlation (r2 = 0.982, P < 0.01) between fold changes generated by our microarray and RT-PCR analyses. RT-PCR run on the 20 individual liver samples (AD1:10, CR1:10) from which we derived our RNA pools and using the same five genes showed close agreement between the pooled and unpooled liver RT-PCR and microarray data (Table 4 and supplemental data).


View this table:
[in this window]
[in a new window]

  		Table 4. Independent validation by real-time PCR of 5 genes significantly altered in the liver by acute CR

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
GRANTS
 REFERENCES
 
Our main findings demonstrating transcriptional responses across multiple tissues, with congruent plasma metabolite changes during acute CR, are summarized in Table 4. Although changes in RNA transcript abundance will not always be associated with a corresponding change in protein abundance or activity, we suggest that systematic changes in expression of genes involved in the same EASE category imply the presence of a functional change. In addition, it is feasible that some of the genes and pathways reported may be regulated specifically by CR and AD in C57BL/6 mice, since it appears that "genotype-regulated genes" exist between different strains of mice that may affect their response to CR (29).

Acute CR Switched Gene Expression Away From Lipid Biosynthesis and Toward Fatty Acid Metabolism and Gluconeogenesis
In both the liver and muscle, we saw a simultaneous metabolic shift in transcription toward fatty acid catabolism, ß-oxidation, and gluconeogenesis and away from lipid biosynthetic processes. Decreased hepatic lipid biosynthetic gene expression was observed following acute CR. Similar reductions in expression of lipid biosynthetic genes, particularly cholesterol biosynthesis, have also previously been described in long-lived Snell dwarf mice (10). Long-term CR reduces the plasma levels of total cholesterol and plasma triglycerides (18, 37). Our metabonomic analysis indicates that this decreased lipid biosynthesis gene transcription coincides with a reduction in several plasma lipid metabolites. Our data indicate that the improved lipid profiles seen after long-term CR and associated with various health benefits (18) occur rapidly on acute CR, suggesting that even short-term reductions in calorie intake may be beneficial to health. However, plasma cholesterol and associated low-density lipoprotein metabolite levels were increased by acute CR, despite a clear reduction in cholesterol biosynthetic gene expression and in contrast to longer-term CR (37). This apparent paradox of high plasma cholesterol but reduced biosynthesis, particularly in liver, may result from transient cholesterol mobilization from white adipose tissue following initiation of CR (Fig. 5). These observations suggest that the low plasma cholesterol seen in long-term CR, but not acute CR, is the net result of significantly reduced but steady-state adipose depots and reduced biosynthesis. Enhanced fatty acid metabolism in liver and muscle, in association with reduced plasma glucose, may help explain the elevated plasma levels of 3-hydroxybutyrate during acute CR. The elevated levels of this ketone body during acute CR suggest an additional requirement for gluconeogenesis to produce C6 components, when C4 substrates required for the TCA cycle (e.g., oxaloacetate) may be in deficit.


Figure 5
View larger version (33K):
[in this window]
[in a new window]

  		Fig. 5. Global summary of metabolite and transcript changes following acute CR in mice. Main tissue-specific EASE categories reported for liver, muscle, hypothalamus, and colon, with arrows denoting up- ({uparrow}) or downregulation ({downarrow}) in acute CR mice relative to AD controls. Plasma metabolites are identified in blue, with arrows, as above, denoting direction of change. EASE category overlap across tissues is identified by italics. We suggest that acute CR, a period of negative energy balance, results in a hypothalamic drive for increased food intake and enhanced energy assimilation in the gastrointestinal tract. Plasma glucose and several plasma lipids decrease, while transcripts involved in glucose metabolism, glucose transport, and insulin sensitivity are altered simultaneously across several tissues, and there is a marked decrease in hepatic lipid biosynthesis. We propose that white adipose tissue catabolism elevates plasma cholesterol, and there is increased liver and muscle ß-oxidation, with metabolism rapidly switching from glycolysis toward lipolysis. Ketones, such as plasma 3-hydroxybutyrate, are indicative of enhanced ß-oxidation and can be utilized as an alternative energy source when glucose is limited, as during acute CR. However, we suggest that fat stores are depleted rapidly at this time, ultimately leading to a switch to gluconeogenesis as the primary energy-generating pathway. Muscle (and possibly colon) breakdown provides amino acid precursors for gluconeogenesis, with elevated plasma glutamine also helping to maintain acid-base balance, thus avoiding ketoacidosis. Elevated lactate suggests partial metabolism of glucose during acute CR, thereby returning C3 compounds (e.g., lactate) to the liver for gluconeogenesis. Note that the organs depicted are human for ease of recognition. No implication is made about the validity of these results in humans.

 
Enhanced Hepatic Mitochondrial Gene Expression Following Acute CR
Genes involved with electron transport and EASE classes linked to mitochondrial function were upregulated in the liver. We suggest that shifts in cellular metabolic pathways observed during acute CR, such as gluconeogenesis and ß-oxidation, are energetically expensive and require enhanced mitochondrial function to provide ATP. It has recently been shown that CR (3- or 12-mo duration) promotes mitochondrial biogenesis in several tissues via endothelial nitric oxide synthase (eNOS) induction, leading to elevated oxygen consumption, increased mitochondrial DNA levels (marker of mitochondrial content), and ATP production (45). While no significant expression changes in any isoform of NOS was observed, several genes central to mitochondrial biogenesis and function, including Pgc-1{alpha}, Tfam, and cytochrome c (Cycs), were significantly upregulated in liver following acute CR.

Acute CR Increased Protein Catabolism in Skeletal Muscle
EASE categories linked to ubiquitination, protein targeting and transport, collagen deposition, blood vessel development, and cell growth were all down in muscle during acute CR. Lean body mass, which includes muscle mass, is lost rapidly in humans under acute CR (19), and muscle mass in rats subjected to 8 wk of 40% CR was significantly reduced (54). The elevation of the branched amino acid valine in plasma in our study is consistent with increased muscle turnover. Muscle catabolism during CR may provide gluconeogenic precursors (57). Gpt2 and Tkt, involved in carbon metabolism, and glutamine synthase (Glul), which is important in disposal of the products of protein catabolism, were all upregulated in muscle. Elevated plasma levels of the glucogenic amino acids methionine, valine, glutamine, and alanine and also creatine were seen in acute CR mice, consistent with enhanced gluconeogenesis. In addition, the elevated plasma levels of lactate and alanine suggest that glucose-utilizing tissues may only partially catabolize glucose, thus sparing carbon skeletons that can then fuel hepatic gluconeogenesis. We suggest that an increased requirement for hepatic gluconeogenesis, in conjunction with muscle catabolism, during acute CR may help explain the elevated plasma methionine levels observed.

Altered Transcript Profile Relevant to Insulin Sensitivity During Acute CR
Both long-term (3) and short-term CR improve insulin sensitivity in muscle (12) and decrease both plasma insulin and glucose (2, 4, 22), which may play a key role in life span extension (20). We saw reduced plasma glucose in mice following acute CR and upregulation of glucose homeostatic genes, including Pgc-1{alpha}, Ppar-{alpha}, Hnf-4{alpha}, and Foxo1, particularly in liver and muscle. Several genes implicated in insulin resistance were downregulated, particularly in muscle, including protein kinase C activator Cish, protein kinase C substrate Marcks and Socs3. In the hypothalamus, the receptor for adiponectin (Adipor2), which promotes insulin sensitivity, was upregulated. The observed increase in plasma lactate following acute CR appears counterintuitive but may be associated with the reduced plasma glucose observed at this time. Lactate production is not only a consequence of anaerobic metabolism in contracting skeletal muscle but is formed continuously under aerobic conditions and is, as mentioned previously, an important C3 metabolite produced in the brain and utilized during hepatic gluconeogenesis in periods of reduced energy availability. Lactate uptake in mitochondria and pyruvate-lactate release in peroxisomes is important in reoxidation of NADH and anaplerotic reactions and is essential to both gluconeogenesis and ß-oxidation, both of which appear transcriptionally upregulated by acute CR in liver and muscle.

Alterations in Energy Balance During Acute CR
The colon and hypothalamus are crucial to energy balance, playing key roles in energy assimilation and energy sensing and the regulation of food intake and energy expenditure, respectively (6). These tissues are likely to be vital for successful adaptation to CR. In contrast to longer-term CR (20), we observed increased hypothalamic pro-MCH precursor (pro-melanin-concentrating hormone) and orexin precursor gene expression, suggesting that there is a drive for enhanced food intake during acute CR. In contrast to Fu et al. (20), we saw no downregulation of stress response, protein folding, or heat response EASE categories (20), although Hspa1b and Hspb1 were downregulated. In the colon, we saw downregulation of genes involved in cellular proliferation and maintenance. These transcriptional changes may contribute to the reduced incidence of colon cancer seen during long-term CR (51) and may also reflect an acute energy-saving mechanism or reduced nutrient transit through the gut at this time. The colon could also act as an additional energy source for hepatic gluconeogenesis. Upregulation of genes involved in cellular transport was observed and may improve extraction and utilization of available energy from the diet, which may be vital to successful adaptation to CR.

Potential Longevity Assurance Genes Identified During Acute CR and Associated With Insulin/Insulin-Like Growth Factor Signaling
Considerable recent interest has focused on mutations in insulin/insulin-like growth factor signaling (IIS) and on how CR and aging impinge on this pathway (32, 41). Both treatments ultimately encroach on nutrient sensing. An elegant study (58) comparing CR and Ames mice indicated that both groups shared an overlapping group of genes, but also both treatments additively affected a panel of genes, suggesting some shared transcriptional responses. Our cross-tissue comparison revealed several IIS pathway genes, particularly in liver and muscle, differentially expressed following acute CR. Expression of insulin receptor substrate protein-1 (Irs1) was downregulated in both liver and muscle following acute CR. Short-term CR (20 days) reduced IRS1 protein expression (22) in mouse skeletal muscle. However, during longer-term CR (14 mo), basal tyrosine-phosphorylated IRS1 protein levels in mouse skeletal muscle appeared unchanged (2, 3), although a nonsignificant trend to increase (~170%) was reported (3). IRS1 knockout mice share some phenotypic similarities with long-lived dwarf mice (7, 62), and life span in chico (dIRS) null flies is enhanced (15).

The forkhead transcription factor Foxo1 (Fkhr) was upregulated in both liver and muscle following acute CR, as was Foxo3a (Fkhrl1) in the liver. The life span extension in daf-2 Caenorhabditis elegans mutants requires daf-16 (Foxo) activity (38), although it appears nonessential for the life span effects of DR (26, 35). Overexpression of dFoxo in the adult Drosophila fat body increased life span and reduced fecundity in female flies (23, 27). Foxo1 also appears to regulate various stress response genes in mammals, including Gadd45, which was also upregulated in liver and muscle (34). In addition, Ppar-{alpha}, upregulated in the liver (and colon), is a powerful coregulator of Foxo1, and both appear central to de novo glucose synthesis (50).

The expression of Redd1 (Ddit4/Rtp801), a hypoxia-inducible factor-responsive gene, was enhanced in liver and muscle. The Drosophila Redd1 paralogs Scylla and Charybdis are stress-induced negative regulators of growth and appear critical for survival under hypoxia and low-nutrient conditions (52). Overexpression of these paralogs downregulates S6 kinase signaling via the TOR signaling pathway (52). Interestingly, global deletion of S6K1 in mice causes mild growth retardation, enhanced insulin sensitivity, and reduced fat mass (60), all features shared by CR rodents and to an extent by Ames dwarf mice (17). In addition, the translational repressor 4e-bp1 (Eif4ebp1) was upregulated in both liver and muscle following acute CR, with an increase in protein levels of 4E-BP1 reported in the liver of Ames mice compared with control mice (55).

Potential Longevity Assurance Genes Identified During Acute CR and Associated With Circadian Rhythm
Circadian rhythm influences many physiological activities, including sleep-wake cycles, endocrine function, body temperature, hepatic metabolism, and detoxification (53). Several period circadian transcriptional factors (Per1, Per2, Per3, Usp2) were upregulated in liver, muscle, and colon following acute CR, as were associated circadian proline- and acid-rich (PAR) basic leucine zipper (bZip) transcription factors (Dbp, Hlf, Tef) in the liver. Arntl/Bmal1 was downregulated in liver, muscle, and colon. Mice with combined deletion of Dpb, Hlf, and Tef show accelerated aging and premature death (21), and mice homozygous for Clock have altered diurnal feeding and a complex hyperphagic, hyperlipidemic, hyperglycemic, hypoinsulinemic, and obese phenotype (59). These findings, and our data, suggest circadian rhythm genes may be important in mammalian energy balance. While feeding time per se does not impinge on life span extension during CR (44), changes in circadian gene transcription following acute CR may be relevant to life span extension, as age-related impairments in sleep patterns are well established (48), alterations in circadian rhythm impinge negatively on life span in various organisms (5), and shift work in humans is associated with elevated plasma glucose and lipid levels and a greater risk of cardiovascular disease, cancer, diabetes, and overall mortality (33).

Mice Undergoing Acute CR Rapidly Adopt Many Transcriptional Changes of Long-Term CR
Several hepatic EASE categories altered after long-term CR (16) were also seen in acute CR, including an upregulation of categories involved in electron transport, mitochondria, and fatty acid metabolism. Interestingly, several categories showed discordance between long-term and acute CR. The categories B class cytochrome P450 and short-chain dehydrogenase/reductase were upregulated during acute CR but downregulated by long-term CR and appear important in C. elegans longevity assurance (42).

Our cross-tissue comparison identified several potential longevity assurance genes (and pathways) altered in common across tissues. We suggest our simultaneous analysis of transcriptional and metabolite alterations is a powerful technique that may narrow down and help identify mechanisms through which CR extends life span, and that the benefits associated with CR may require only a short-term reduction in caloric intake.


   GRANTS
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
REFERENCES
 
This work was performed under the Functional Genomics of Ageing Consortium supported by the Wellcome Trust Functional Genomics Programme (to D. J. Withers, L. Partridge, J. M. Thornton, and D. Gems), Research into Ageing and The Rosetrees Trust (to D. J. Withers, L. Partridge, and D. Gems), the Biotechnology and Biological Sciences Research Council and EMBL (to J. M. Thornton).


   ACKNOWLEDGMENTS
 
We thank Andrew Craig, Roger Tatoud, and Olivier Cloarec for bioinformatic advice. We are grateful to Mike Hubank and Nipurna Jina at the Institute of Child Health Gene Microarray Centre, University College London, United Kingdom, for microarray and bioinformatics support.


   FOOTNOTES
 
Address for reprint requests and other correspondence: C. Selman, Centre for Diabetes and Endocrinology, Dept. of Medicine, Univ. College London, Rayne Institute, 5 University St., London WC1E 6JJ, UK (e-mail: c.selman{at}ucl.ac.uk).

Article published online before print. See web site for date of publication (http://physiolgenomics.physiology.org).

* C. Selman and N. D. Kerrison contributed equally to this work. Back


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 METHODS
 RESULTS
 DISCUSSION
 GRANTS
 REFERENCES
 

  1. Amador-Noguez D, Yagi K, Venable S, and Darlington G. Gene expression profile of long-lived Ames dwarf mice and Little mice. Aging Cell 3: 423–441, 2004.[CrossRef][ISI][Medline]
  2. Argentino DP, Dominici FP, Al-Regaiey K, Bonkowski MS, Bartke A, and Turyn D. Effects of long-term caloric restriction on early steps of the insulin-signaling system in mouse skeletal muscle. J Gerontol A Biol Sci Med Sci 60: 28–34, 2005.[Abstract/Free Full Text]
  3. Argentino DP, Dominici FP, Munoz MC, Al-Regaiey K, Bartke A, and Turyn D. Effects of long-term caloric restriction on glucose homeostasis and on the first steps of the insulin signaling system in skeletal muscle of normal and Ames dwarf (Prop1df/Prop1df) mice. Exp Gerontol 40: 27–35, 2005.[CrossRef][ISI][Medline]
  4. Argentino DP, Munoz MC, Rocha JS, Bartke A, Turyn D, and Dominici FP. Short-term caloric restriction does not modify the in vivo insulin signaling pathway leading to Akt activation in skeletal muscle of Ames dwarf (Prop1(df)/Prop1(df)) mice. Horm Metab Res 37: 672–679, 2005.[CrossRef][ISI][Medline]
  5. Aujard F, Dkhissi-Benyahya O, Fournier I, Claustrat B, Schilling A, Cooper HM, and Perret M. Artificially accelerated aging by shortened photoperiod alters early gene expression (Fos) in the suprachiasmatic nucleus and sulfatoxymelatonin excretion in a small primate, Microcebus murinus. Neuroscience 105: 403–412, 2001.[CrossRef][ISI][Medline]
  6. Badman MK and Flier JS. The gut and energy balance: visceral allies in the obesity wars. Science 307: 1909–1914, 2005.[Abstract/Free Full Text]
  7. Bartke A, Wright JC, Mattison JA, Ingram DK, Miller RA, and Roth GS. Extending the lifespan of long-lived mice. Nature 414: 412, 2001.[CrossRef][Medline]
  8. Bauer M, Hamm AC, Bonaus M, Jacob A, Jaekel J, Schorle H, Pankratz MJ, and Katzenberger JD. Starvation response in mouse liver shows strong correlation with life-span-prolonging processes. Physiol Genomics 17: 230–244, 2004.[Abstract/Free Full Text]
  9. Beckwith-Hall BM, Brindle JT, Barton RH, Coen M, Holmes E, Nicholson JK, and Antti H. Application of orthogonal signal correction to minimise the effects of physical and biological variation in high resolution 1H NMR spectra of biofluids. Analyst 127: 1283–1288, 2002.[CrossRef][Medline]
  10. Boylston WH, Gerstner A, DeFord JH, Madsen M, Flurkey K, Harrison DE, and Papaconstantinou J. Altered cholesterologenic and lipogenic transcriptional profile in livers of aging Snell dwarf (Pit1dw/dwJ) mice. Aging Cell 3: 283–296, 2004.[CrossRef][ISI][Medline]
  11. Cao SX, Dhahbi JM, Mote PL, and Spindler SR. Genomic profiling of short- and long-term caloric restriction effects in the liver of aging mice. Proc Natl Acad Sci USA 98: 10630–10635, 2001.[Abstract/Free Full Text]
  12. Cartee GD and Dean DJ. Glucose transport with brief dietary restriction: heterogenous responses in muscles. Am J Physiol Endocrinol Metab 266: E946–E952, 1994.[Abstract/Free Full Text]
  13. Choe SE, Boutros M, Michelson AM, Church GM, and Halfon MS. Preferred analysis methods for Affymetrix GeneChips revealed by a wholly defined control dataset. Genome Biol 6: R16, 2005.[CrossRef][Medline]
  14. Choudhury AI, Heffron H, Smith MA, Al-Qassab H, Xu AW, Selman C, Simmgen M, Clements M, Claret M, Maccoll G, Bedford DC, Hisadome K, Diakonov I, Moosajee V, Bell JD, Speakman JR, Batterham RL, Barsh GS, Ashford ML, and Withers DJ. The role of insulin receptor substrate 2 in hypothalamic and beta cell function. J Clin Invest 115: 940–950, 2005.[CrossRef][ISI][Medline]
  15. Clancy DJ, Gems D, Harshman LG, Oldham S, Stocker H, Hafen E, Leevers SJ, and Partridge L. Extension of life-span by loss of CHICO, a Drosophila insulin receptor substrate protein. Science 292: 104–106, 2001.[Abstract/Free Full Text]
  16. Dhahbi JM, Kim HJ, Mote PL, Beaver RJ, and Spindler SR. Temporal linkage between the phenotypic and genomic responses to caloric restriction. Proc Natl Acad Sci USA 101: 5524–5529, 2004.[Abstract/Free Full Text]
  17. Dominici FP, Hauck S, Argentino DP, Bartke A, and Turyn D. Increased insulin sensitivity and upregulation of insulin receptor, insulin receptor substrate (IRS)-1 and IRS-2 in liver of Ames dwarf mice. J Endocrinol 173: 81–94, 2002.[Abstract]
  18. Fontana L, Meyer TE, Klein S, and Holloszy JO. Long-term calorie restriction is highly effective in reducing the risk for atherosclerosis in humans. Proc Natl Acad Sci USA 101: 6659–6663, 2004.[Abstract/Free Full Text]
  19. Friedlander AL, Braun B, Pollack M, MacDonald JR, Fulco CS, Muza SR, Rock PB, Henderson GC, Horning MA, Brooks GA, Hoffman AR, and Cymerman A. Three weeks of caloric restriction alters protein metabolism in normal-weight, young men. Am J Physiol Endocrinol Metab 289: E446–E455, 2005.[Abstract/Free Full Text]
  20. Fu C, Xi L, McCarter R, Hickey M, and Han ES. Early hypothalamic response to age-dependent gene expression by calorie restriction. Neurobiol Aging 27: 1315–1325, 2006.[CrossRef][ISI][Medline]
  21. Gachon F, Fonjallaz P, Damiola F, Gos P, Kodama T, Zakany J, Duboule D, Petit B, Tafti M, and Schibler U. The loss of circadian PAR bZip transcription factors results in epilepsy. Genes Dev 18: 1397–1412, 2004.[Abstract/Free Full Text]
  22. Gazdag AC, Dumke CL, Kahn CR, and Cartee GD. Calorie restriction increases insulin-stimulated glucose transport in skeletal muscle from IRS-1 knockout mice. Diabetes 48: 1930–1936, 1999.[Abstract]
  23. Giannakou ME, Goss M, Junger MA, Hafen E, Leevers SJ, and Partridge L. Long-lived Drosophila with overexpressed dFOXO in adult fat body. Science 305: 361, 2004.[Abstract/Free Full Text]
  24. Guarente L. Calorie restriction and SIR2 genes–towards a mechanism. Mech Ageing Dev 126: 923–928, 2005.[CrossRef][ISI][Medline]
  25. Hosack DA, Dennis G Jr, Sherman BT, Lane HC, and Lempicki RA. Identifying biological themes within lists of genes with EASE. Genome Biol 4: R70, 2003.[CrossRef][Medline]
  26. Houthoofd K, Braeckman BP, Johnson TE, and Vanfleteren JR. Life extension via dietary restriction is independent of the Ins/IGF-1 signalling pathway in Caenorhabditis elegans. Exp Gerontol 38: 947–954, 2003.[CrossRef][ISI][Medline]
  27. Hwangbo DS, Gershman B, Tu MP, Palmer M, and Tatar M. Drosophila dFOXO controls lifespan and regulates insulin signalling in brain and fat body. Nature 429: 562–566, 2004.[CrossRef][Medline]
  28. Ikeno Y, Hubbard GB, Lee S, Richardson A, Strong R, Diaz V, and Nelson JF. Housing density does not influence the longevity effect of calorie restriction. J Gerontol A Biol Sci Med Sci 60: 1510–1517, 2005.[Abstract/Free Full Text]
  29. Kaput J, Klein KG, Reyes EJ, Kibbe WA, Cooney CA, Jovanovic B, Visek WJ, and Wolff GL. Identification of genes contributing to the obese yellow Avy phenotype: caloric restriction, genotype, diet x genotype interactions. Physiol Genomics 18: 316–324, 2004.[Abstract/Free Full Text]
  30. Kendziorski C, Irizarry RA, Chen KS, Haag JD, and Gould MN. On the utility of pooling biological samples in microarray experiments. Proc Natl Acad Sci USA 102: 4252–4257, 2005.[Abstract/Free Full Text]
  31. Kendziorski CM, Zhang Y, Lan H, and Attie AD. The efficiency of pooling mRNA in microarray experiments. Biostatistics 4: 465–477, 2003.[Abstract]
  32. Kenyon C. The plasticity of aging: insights from long-lived mutants. Cell 120: 449–460, 2005.[CrossRef][ISI][Medline]
  33. Knutsson A, Hammar N, and Karlsson B. Shift workers’ mortality scrutinized. Chronobiol Int 21: 1049–1053, 2004.[CrossRef][ISI][Medline]
  34. Kobayashi Y, Furukawa-Hibi Y, Chen C, Horio Y, Isobe K, Ikeda K, and Motoyama N. SIRT1 is critical regulator of FOXO-mediated transcription in response to oxidative stress. Int J Mol Med 16: 237–243, 2005.[ISI][Medline]
  35. Lakowski B and Hekimi S. The genetics of caloric restriction in Caenorhabditis elegans. Proc Natl Acad Sci USA 95: 13091–13096, 1998.[Abstract/Free Full Text]
  36. Lee CK, Klopp RG, Weindruch R, and Prolla TA. Gene expression profile of aging and its retardation by caloric restriction. Science 285: 1390–1393, 1999.[Abstract/Free Full Text]
  37. Liepa GU, Masoro EJ, Bertrand HA, and Yu BP. Food restriction as a modulator of age-related changes in serum lipids. Am J Physiol Endocrinol Metab 238: E253–E257, 1980.[Abstract/Free Full Text]
  38. Lin K, Hsin H, Libina N, and Kenyon C. Regulation of the Caenorhabditis elegans longevity protein DAF-16 by insulin/IGF-1 and germline signaling. Nat Genet 28: 139–145, 2001.[CrossRef][ISI][Medline]
  39. Liu G, Loraine AE, Shigeta R, Cline M, Cheng J, Valmeekam V, Sun S, Kulp D, and Siani-Rose MA. NetAffx: Affymetrix probesets and annotations. Nucleic Acids Res 31: 82–86, 2003.[Abstract/Free Full Text]
  40. Mair W, Goymer P, Pletcher SD, and Partridge L. Demography of dietary restriction and death in Drosophila. Science 301: 1731–1733, 2003.[Abstract/Free Full Text]
  41. Masoro EJ. Overview of caloric restriction and ageing. Mech Ageing Dev 126: 913–922, 2005.[CrossRef][ISI][Medline]
  42. McElwee JJ, Schuster E, Blanc E, Thomas JH, and Gems D. Shared transcriptional signature in Caenorhabditis elegans Dauer larvae and long-lived daf-2 mutants implicates detoxification system in longevity assurance. J Biol Chem 279: 44533–44543, 2004.[Abstract/Free Full Text]
  43. Miller RA, Chang Y, Galecki AT, Al-Regaiey K, Kopchick JJ, and Bartke A. Gene expression patterns in calorically restricted mice: partial overlap with long-lived mutant mice. Mol Endocrinol 16: 2657–2666, 2002.[Abstract/Free Full Text]
  44. Nelson W. Food restriction, circadian disorder and longevity of rats and mice. J Nutr 118: 286–289, 1988.[Medline]
  45. Nisoli E, Tonello C, Cardile A, Cozzi V, Bracale R, Tedesco L, Falcone S, Valerio A, Cantoni O, Clementi E, Moncada S, and Carruba MO. Calorie restriction promotes mitochondrial biogenesis by inducing the expression of eNOS. Science 310: 314–317, 2005.[Abstract/Free Full Text]
  46. Partridge L, Piper MD, and Mair W. Dietary restriction in Drosophila. Mech Ageing Dev 126: 938–950, 2005.[CrossRef][ISI][Medline]
  47. Pletcher SD, Macdonald SJ, Marguerie R, Certa U, Stearns SC, Goldstein DB, and Partridge L. Genome-wide transcript profiles in aging and calorically restricted Drosophila melanogaster. Curr Biol 12: 712–723, 2002.[CrossRef][ISI][Medline]
  48. Prinz PN. Age impairments in sleep, metabolic and immune functions. Exp Gerontol 39: 1739–1743, 2004.[CrossRef][ISI][Medline]
  49. Pugh TD, Oberley TD, and Weindruch R. Dietary intervention at middle age: caloric restriction but not dehydroepiandrosterone sulfate increases lifespan and lifetime cancer incidence in mice. Cancer Res 59: 1642–1648, 1999.[Abstract/Free Full Text]
  50. Puigserver P, Rhee J, Donovan J, Walkey CJ, Yoon JC, Oriente F, Kitamura Y, Altomonte J, Dong H, Accili D, and Spiegelman BM. Insulin-regulated hepatic gluconeogenesis through FOXO1-PGC-1alpha interaction. Nature 423: 550–555, 2003.[CrossRef][Medline]
  51. Raju J and Bird RP. Energy restriction reduces the number of advanced aberrant crypt foci and attenuates the expression of colonic transforming growth factor beta and cyclooxygenase isoforms in Zucker obese (fa/fa) rats. Cancer Res 63: 6595–6601, 2003.[Abstract/Free Full Text]
  52. Reiling JH and Hafen E. The hypoxia-induced paralogs Scylla and Charybdis inhibit growth by down-regulating S6K activity upstream of TSC in Drosophila. Genes Dev 18: 2879–2892, 2004.[Abstract/Free Full Text]
  53. Schibler U. The daily rhythms of genes, cells and organs. EMBO Rep 6: S9–S13, 2005.[CrossRef][ISI][Medline]
  54. Selman C, Phillips T, Staib JL, Duncan JS, Leeuwenburgh C, and Speakman JR. Energy expenditure of calorically restricted rats is higher than predicted from their altered body composition. Mech Ageing Dev 126: 783–793, 2005.[CrossRef][ISI][Medline]
  55. Sharp ZD and Bartke A. Evidence for down-regulation of phosphoinositide 3-kinase/Akt/mammalian target of rapamycin (PI3K/Akt/mTOR)-dependent translation regulatory signaling pathways in Ames dwarf mice. J Gero