Metabonomic Investigations of Aging and Caloric Restriction in a Life

Apr 6, 2007 - Top-Down Systems Biology Modeling of Host Metabotype−Microbiome Associations in Obese Rodents. Alison Waldram , Elaine Holmes ...
2 downloads 0 Views 3MB Size
Download PDF
Metabonomic Investigations of Aging and Caloric Restriction in a Life-Long Dog Study Yulan Wang,† Dennis Lawler,‡ Brian Larson,‡ Ziad Ramadan,§ Sunil Kochhar,§ Elaine Holmes,† and Jeremy K. Nicholson*,† Department of Biomolecular Medicine, SORA Division, Faculty of Medicine, Imperial College London, Sir Alexander Fleming Building, South Kensington, London SW7 2AZ, United Kingdom, Nestle´ Research Center, Lausanne, NESTEC Ltd., Vers-Chez-Les-Blanc, 1000 Lausanne 26, Switzerland, and The Nestle Research CentersSt. Louis, St. Louis, Missouri 63164 Received December 20, 2006

Long-term restriction of energy intake without malnutrition is a robust intervention that has been shown to prolong life and delay age-related morbidity. A 1H NMR-based metabonomic strategy was used to monitor urinary metabolic profiles throughout the lifetimes of control-fed and diet-restricted dogs. Urinary metabolic trajectories were constructed for each dog, and metabolic variation was found to be predominantly influenced by age. Urinary excretion of creatinine increased with age, reaching a maximum between ages 5 and 9 years and declining thereafter. Excretion of mixed glycoproteins was noted at earlier ages, which may be a reflection of growth patterns. In addition, consistent metabolic variation related to diet was also characterized, and energy-associated metabolites, such as creatine, 1-methylnicotinamide, lactate, acetate, and succinate, were depleted in urine from diet-restricted dogs. Both aging and diet restriction altered activities of the gut microbiotia, manifested by variation of aromatic metabolites and aliphatic amine compounds. This analysis allowed the metabolic response to two different physiological processes to be monitored throughout the lifetime of the canine population and may form part of a strategy to monitor and reduce the impact of age related diseases in the dog, as well as providing more general insights into extension of longevity in higher mammals. Keywords: metabonomics • aging • restricted diet • multivariate data analysis • 1H NMR spectroscopy

Introduction Improving the quality of life and delaying the onset of aging in man and pet animals are important scientific and social topics. Energy restriction or dietary restriction (DR) is a robust intervention that favorably impacts the life span of species as diverse as nematodes, spiders, rotifers, water fleas, fruit flies, fish, hamsters, mice, rats, and dogs.1,2 Energy restriction also modulates expression of late-life diseases, and studies in rodents, dogs, and primates have demonstrated that food (energy) restriction delays age-related morbidity from speciesand strain-specific causes.1-7 The underlying mechanism of the extended life span by means of dietary restriction is still a matter of debate.8 The hypotheses include a decrease in the basal metabolic rate,9 reduced free-radical-mediated damage,10,11 attenuated DRinduced aging-associated declines of protein turnover,12 and increased activities of NAD(+)-dependent histone deacetylase, Sir2, which plays a critical role in transcriptional silencing, genome stability, and longevity.13,14 The main metabolic changes * To whom correspondence should [email protected]. † Imperial College London. ‡ The Nestle Research CentersSt. Louis. § NESTEC Ltd.

1846

be

addressed.

Journal of Proteome Research 2007, 6, 1846-1854

Published on Web 04/06/2007

E-mail:

associated with DR include decreased levels of blood insulin, glucose, growth factors, and triglycerides.15 Several studies have also investigated and characterized the impact of diet on the aging processes. In one study, aging was accompanied by increased excretion of creatinine, taurine, and hippurate and reduced excretion of tricarboxylic acid cycle intermediates, such as 2-oxoglutarate.16 However, many of the published investigations were carried out using very young or old rats,17 which may not adequately represent other species or be representative of the total life span. In the current study, we followed 48 Labrador Retriever dogs throughout their lives to evaluate the metabolic effects of aging and the impact of life-long 25% food restriction in pair feeding studies. The complexity of characterizing the response to dietary intervention superimposed on the background of dynamic biological processes such as aging presents a substantial challenge, not the least because of the large volume of data collected over the entire life span. To address this problem, we applied a metabonomic approach to monitor and model multiple metabolic parameters simultaneously in a retrospective study on stored samples. Metabonomics involves the study of multivariate metabolic responses of complex organisms to physiological and/or pathological stressors, including the consequent disruption of systems regulation.18-21 Technically, metabonomics usually involves the collection and analysis of data from NMR spectros10.1021/pr060685n CCC: $37.00

 2007 American Chemical Society

research articles

Metabolic Effects of Diet and Aging

copy or mass spectrometry using appropriate multivariate statistical techniques. 1H NMR is capable of simultaneously detecting a wide range of small-molecule metabolites, thus providing a “metabolic fingerprint” of the major extracellular components exchanged in the blood and other biofluids and tissues.20,22 Since 1H NMR spectra contain a high-density of metabolic information, a series of multivariate statistical analyses, such as principal component analysis (PCA) and orthogonal signal correction projection to latent structure discriminant analysis (O-PLS-DA),23,24 were applied to decode the changes relating to the biological perturbations of interest. This metabolic profiling strategy has been applied to the study of biochemical consequences of diet-disease interactions,25-27 mammal-parasite interactions,28 and subtle metabolic signatures of dietary preferences such as soy consumption, chamomile ingestion, and alcohol intake.29-31 In this study we have evaluated the systemic metabolic consequences of aging and of diet restriction throughout the lives of dogs.

Methods Animal Handling and Urine Collection. This study was conducted under the supervision of the Institutional Animal Care and Use Committee. A total of 48 Labrador dogs from 7 litters were allotted to a paired feeding design.1,4,32 The dogs were paired by gender and weaning weight within a litter and then were assigned randomly to control feeding (CF) or diet restriction (DR). They were housed under the same environmental conditions for life (2 × 19 m2 kennel runs, with one or two pairs per pen). Free indoor-outdoor access was available, and activity was not restricted. The feeding protocol was initiated at age 8 weeks.1,4 All dogs were fed the same extruded diet.1,4 Each DR dog was fed a quantity equal to 75% of the amount of food that was consumed on the previous day by the respective CF pair mate. When the dogs were 3 years old, two adjustments were incorporated into the feeding protocol to prevent insidious development of obesity among all of the dogs: (a) All dogs were switched from a 27% protein puppy growth formula to a 21% protein adult formula. (b) The amount of food given to CF dogs was fed at a constant amount of 62.1 kcal of metabolizable energy/kg of ideal body mass. DR pair mates continued to receive 75% of the consumption of their corresponding CF pair mates. All dogs received standard vaccinations and antiparasite treatments and were monitored daily throughout life for signs of illness. When necessary, appropriate therapeutic measures were instituted under veterinary supervision in a manner consistent with the management of the entire colony population. Similar disease conditions among these dogs were managed as uniformly as medically possible. Fasting urine samples were collected from each dog via catheterization or cystocentesis at ages 13, 18, and 32 weeks, ages 1.0, 1.5, and 2.0 years, and annually after age 5 years until death. Samples were stored at -20 °C until analyses were performed. NMR Spectroscopy. Urine samples were prepared by mixing 400 µL of urine with 200 µL of phosphate buffer (pH 7.4) containing 10% D2O as a field frequency lock and 0.05% sodium 3-(trimethylsilyl)propionate-2,2,3,3-d4 as a chemical shift reference. 1H NMR spectra were acquired for each sample using a Bruker DRX 600 NMR spectrometer operating at 600.13 MHz for 1H equipped with a 5 mm triple-resonance inverse detection

probe (Bruker, Germany). A standard one-dimensional NMR spectrum was acquired using the first increment of the pulse sequence [RD-90°-t1-90°-tm-90°-ACQ]. Water suppression was achieved with an irradiation on the water peak during the recycle delay (RD ) 2 s) and a mixing time, tm, of 100 ms. t1 was set to 3 µs. The 90° pulse length was adjusted to approximately 11.5 µs, and 128 transients were collected into 32K data points for each spectrum with a spectral width of 20 ppm. All free induction decays were multiplied by an exponential function equivalent to a 1 Hz line-broadening factor prior to Fourier transformation. For assignment purposes, standard two-dimensional (2D) correlation spectroscopy (COSY) and total correlation spectroscopy (TOCSY) NMR spectra were also acquired for a selected sample following the method described by Hurd33 and Bax.34 Data Analysis. 1H NMR spectra were corrected for phase and baseline distortion using XWINNMR 3.5 (Bruker). The spectra over the range δ 0.5-10.0 were digitized using a Matlab script developed in-house (Dr. O. Cloarec, Imperial College London). The region δ 4.5-5.35 was removed to avoid the effects of imperfect water suppression. Two normalization strategies were applied to the NMR data prior to data analyses to partially compensate for urinary concentration differences: (i) normalization to the sum of the spectrum; (ii) normalization to the integral of the creatinine resonance (δ 4.04), which also partially compensates for differences in muscle mass. O-PLS-DA of the NMR spectral data also was carried out after applying scaling to unit variance in a MATLAB 7.0 environment with a MATLAB script developed in-house (Dr. O. Cloarec, Imperial College London).23 The O-PLS-DA algorithm is derived from the projection to latent structures regression method.24 O-PLS-DA models were constructed using the NMR data as the X variables and the age or dietary intervention as the Y variable.23 This procedure involved the combination of orthogonal signal correction and PLS discriminant analysis and utilized back-transformation of the loadings incorporated with the weight of the variable contributing the discrimination in the models for displaying the loadings. The statistical total correlation spectroscopy (STOCSY)35 method developed with in-house software METATOOLS (Dr. O. Cloarec, Imperial College London) was also applied to spectral peaks found to be significantly different between dietary regimens and variation in aging for the purpose of structural elucidation of metabolites and to establish possible association with other metabolic components.

Results 1H

NMR Spectra of Dog Urine. Typical examples of 600 MHz NMR spectra of urine obtained from a dog in the CF group at ages 13 weeks (A), 72 weeks (B), and 9 years (C) are shown in Figure 1. Resonance assignments were made with the aid of existing literature,28,36 two-dimensional NMR spectroscopy (COSY and TOCSY). Prominent urinary metabolites noted from NMR spectra include amino acids, organic acids, creatinine, and methylamines. Two broad resonances at 1.95 and 2.05 ppm were putatively assigned as N-acetylglycoproteins.37 STOCSY was used to support the structural confirmation and resonances such as those for the N-acetylglycoproteins by confirming correlations between the two broad N-acetylglycoproteins at 1.95 and 2.05 ppm and resonances from the mixed glycoproteins at core amino acid signals 0.8-0.9, 1.4, 1.7, 2.3, and 4.34.5 ppm. Age-related variations in the urinary profiles were evident. For example, relatively high levels of N-acetylglycoproteins were observed in young dogs, whereas taurine and

1H

Journal of Proteome Research • Vol. 6, No. 5, 2007 1847

research articles

Wang et al.

Figure 1. Three typical 600 MHz 1H NMR spectra of urine obtained from a dog with a control diet at ages 13 weeks (A), 1.5 years (B), and 9 years (C). Compared to the chemical shift range at δ 3.0-4.5, the spectra in the regions δ 6.5-9 and 0.7-3.0 are displayed at 4× magnification. Key: 1, R-ketobutyrate; 2, 2-propanol; 3, lactate; 4, alanine; 5, acetate; 6, 7, mixed N-acetylglycoproteins; 8, succinate; 9, citrate; 10, dimethylamine; 11, trimethylamine; 12, dimethylglycine; 13, creatinine; 14, taurine; 15, trimethylamine N-oxide; 16, glycine; 17, (4-hydroxyphenyl)propionic acid; 18, (phenylacetyl)glycine; 19, N-methylnicotinamide; 20, hippurate; 21, formate; 22, polyols.

glycerol concentrations were higher in urine samples obtained at age 1.5 years. Overview of Aging and Dietary Intervention. To understand globally how dietary restriction is affecting aging, O-PLS-DA analysis was performed on data normalized to the total unit area. Here data extracted from NMR spectra of urine samples were used as X variables, and class information, i.e., age and feeding regime, was used as Y variables. A total of one orthogonal and five PLS components were calculated for the model, and the trajectory plot of the cross-validated scores PC1 versus PC2 illustrating variations in the metabolic space at different stages of growth, and the influence of restricted diet, is shown in Figure 2. Here each coordinate represents an average of scores for all dogs in a feeding group at a particular age, and the bars denote the standard deviation derived from the scores. The trajectory demonstrated an initial rapid shifting of urine metabolic profiles before age 1 year, after which the metabolic signature stabilized between ages 1 and 2 years. A second metabolic shift was observed between ages 5 and 9 years. Urine profiles obtained from dogs of over age 10 years presented a third metabolic transformation (Figure 2). Urine profiles obtained from CF and DR dogs followed the same trajectory globally and were generally dominated by agerelated changes. For example, the relative concentration of creatinine was highest between ages 5 and 9 years, followed by decreasing concentrations after age 10 years. In addition to the age-related changes, the more subtle influences of DR on the urine metabolic profiles also were evident. To uncover additional metabolites which could vary with aging or dietary intervention, the same strategy was applied to the NMR data normalized with respect to the integrated creatinine resonance at δ 4.04, thereby removing some of the influences attributed solely to muscle mass. The trajectory plot obtained from the data normalized on creatinine is shown in Figure 3. The trajectory plots obtained from both normalization methods were similar. The loadings suggested that concentrations of trimethylamine N-oxide (TMAO) and N-acetylglyco1848

Journal of Proteome Research • Vol. 6, No. 5, 2007

proteins are relatively high in the urine of young dogs and decrease with age. Metabolic Changes during Aging. To investigate age-related metabolic effects, O-PLS-DA of data normalized on the total area was performed. 1H NMR spectra of urine samples obtained from dogs of over age 10 years were excluded from this analysis due to potentially confounding health problems or drug effects. For example, some of the dogs have developed osteoarthritis at this age, which is common in Labrador Retrievers, and were treated accordingly. The urine metabolic profiles at three time points from both the CF and DR groups were selected to further explore the age-induced changes in metabolism. The ages that were chosen were 13 weeks, 1.5 years, and 9 years (Figure 2 and 3). O-PLS-DA comparison between spectra obtained from the three selected age groups was carried out with a unit variance scaling. One orthogonal component and two PLS components were calculated for the model. The quality of the model was described by the cross-validation parameter Q2, indicating the predictability of the model, and R2, which represents the total explained variation for the X matrix (Table 1). Clear separation was achieved between dogs at age 13 weeks and those from the other two age groups with 97% prediction, while overlap between dogs at ages 1.5 years (85%) and 9 years (75%) was observed as indicated in the cross-validated scores plots (Figure 4). The metabolites associated with each of the age groups with respect to the other two age groups can be found in the corresponding PLS coefficient plots (Figure 5). The direction of the peaks in the PLS coefficient plot indicates the changes in the relative concentration of the metabolites associated with a particular age group, upward peaks indicating increased concentration and downward peaks decreased concentration with respect to those of the other age groups. The color associated with signals in the coefficient plot indicates the significance of metabolites in characterizing the NMR data for a given class as defined by the color scaling map on the righthand side of each coefficient plot. The coefficients for the

research articles

Metabolic Effects of Diet and Aging

Figure 2. Cross-validated scores mean trajectory derived from O-PLS-DA of 1H NMR spectra of dog’s urine normalized on the sum of the spectrum indicating metabolic changes associated with aging and dietary restriction. (B) is an expansion of (A). Key: blue line, control diet; red line, restricted diet.

metabolites associated with a particular age group are given in Table 1. For this analysis, the coefficient was considered to be significant when higher than 0.30, which corresponds to the critical value of a correlation coefficient of 5% (p ) 0.05). A number of metabolites were found to vary with age, among which urine creatinine was most marked. Urine obtained from dogs at a very young age (13 weeks) presented relatively higher amounts of mixed N-acetylglycoproteins and dimethylamine (DMA), but lower amounts of 2-oxobutyrate, 2-propanol, dimethylglycine (DMG), creatinine, hippurate, and 3-(hydroxyphenyl)propionic acid (3-HPPA). The urine samples from growing dogs (1.5 years old) contained higher concentrations of taurine, hippurate, creatinine, and 3-HPPA and lower concentrations of acetate, DMA, mixed glycoproteins, 1-methylnicotinamide, lactate, succinate, and TMAO, with respect to the samples from other age groups. Metabolites separating dogs aged 9 years from the younger dogs included elevated levels of R-ketobutyrate, 2-propanol, lactate, alanine, succinate, creatinine, DMG, and trimethylamine and lower concentrations of mixed glycoproteins. Metabolic Changes due to Diet Restriction. The effects of DR were evaluated by O-PLS-DA of 1H NMR profiles of urine

obtained from CF and DR groups at the same ages. NMR data normalized to the total area were used in this analysis. The coefficient plots exemplifying the impact of DR at different stages of life are plotted in Figure 6. The dominant metabolites contributing to the separation between the CF and DR groups, together with Q2 and R2 values, are listed in Table 2. Because the number of samples in each class had been reduced by attrition to 20-23 by this point in the analysis, a coefficient above 0.4 was considered to be significant for this analysis. Compared to CF dogs, DR dogs had consistently elevated urine levels of aromatic metabolites, including hippurate, (phenylacetyl)glycine (PAG), 4-hydroxyphenylacetic acid, and 3-HPPA (Table 2). These differences were accompanied by increased concentration of creatine/creatinine and acetate. In addition, concentrations of lactate, succinate, and 1-methylnicotinamide were increased in the urine of DR dogs at age 9 years.

Discussion The aim of the present study was to investigate urinary metabolic changes over the life spans of CF and DR dogs. Both the CF and DR groups followed the same aging trajectory, suggesting that age-related changes dominated urinary metaJournal of Proteome Research • Vol. 6, No. 5, 2007 1849

research articles

Wang et al.

Figure 3. Cross-validated scores mean trajectory derived from O-PLS-DA of 1H NMR spectra of dog’s urine normalized on the integration of the creatinine resonance indicating metabolic changes associated with aging and dietary restriction. (B) is an expansion of (A). Key: blue line, control diet; red line, restricted diet. Table 1. Changes of Metabolites Associated with Aging (Q2 ) 0.74, R2X ) 0.32)a chemical shift (δ, ppm) 13 weeks 1.5 years

alanine acetate N-acetylglycoproteins citrate creatinine DMA DMG hippurate 3-HHPA R-ketobutyrate lactate N-methylnicotinamide 2-propanol succinate TMA TMAO taurine

1.48 1.924 2.04 2.55 3.0465 2.724 2.94 7.831 7.313(s) 1.07 1.3285 9.28 1.1 2.41 2.88 3.27 3.437

NSDb NSD 0.9308 NSD -0.4731 0.942 -0.6631 -0.3717 -0.6303 -0.7143 NSD NSD -0.6932 NSD NSD NSD NSD

+0.3701 -0.352 -0.6026 NSD +0.3796 -0.5582 NSD +0.3043 +0.6004 NSD -0.2994 -0.3221 +0.5152 -0.3448 NSD -0.3949 +0.5322

9 years

+0.3573 NSD -0.4087 +0.3751 +0.4175 NSD +0.6561 NSD NSD +0.7166 +0.5904 NSD +0.6485 +0.3637 +0.4594 NSD NSD

a A positive value indicates an increase in the concentration of metabolites, and a negative value incicates a decrease in the concentration of metabolites. b No significant difference.

bolic profiles, with aging exerting a greater effect on the metabolism than dietary restriction. 1850

Journal of Proteome Research • Vol. 6, No. 5, 2007

Figure 4. Cross-validated scores plot derived from O-PLS-DA of 1H NMR spectra of urine obtained from dogs at ages 13 weeks (green asterisks), 1.5 years (blue asterisks), and 9 years (red asterisks). Each asterisk represents the metabolic profile obtained from each dog.

Creatinine was unsurprisingly the metabolite that exerted the highest leverage on the overall metabolic variation with age. Creatinine excretion increased up to the age 5-9 years and then declined at later stages of life. The change in creatinine

Metabolic Effects of Diet and Aging

research articles

Figure 5. Coefficient plot obtained from O-PLS-DA of 1H NMR spectra of urine obtained from dogs at ages 13 weeks (A), 1.5 years (B), and 9 years (C).

concentration with age could arise from two sources: (a) muscle mass and (b) renal function and structure. Urine creatinine is an index of muscle mass and is directly correlated with body mass.38 It is therefore a straightforward explanation of the initial increase in urine creatinine associated with the growing and increased muscle mass in dogs. Renal plasma flow and glomerular filtration rates tend to decrease with aging and therefore may also contribute to increased urinary excretion of creatinine with age.39,40 However, serial clinical evaluations, clinical chemistry profiles, and finally postmortem examination indicated that diminished renal function was not a viable explanation for the age-related changes in urine creatinine observed here.1,4 Therefore, the decline of creatinine excretion with advancing age was most likely associated with progressive changes in muscle performance and lean mass,41,42 which is consistent with studies performed in rats showing decreased creatinine excretion after age 20 months.16 The intensities of acetylglycoprotein signals were markedly stronger in urines obtained at age 13 weeks, compared to older ages. Since domestic dogs are adolescent at this age, this observation likely reflects growth processes more universally than it does any specific maturation. A considerable increase in urinary hippurate and 3-HPPA also were associated with dogs at age 1.5 years. Microbes can modulate urine concentrations of hippurate and other aromatic metabolic species. For example, Nicholls et al.36 reported that urinary hippurate excretion in rats depended on gut microbiotal activities and

that a stable gut microbiota was achieved up to 21 days after germ-free rats were introduced into a standard laboratory environment. In other studies, hippurate and 3-HPPA excretions were correlated with the microflora composition of the colon.43 Rapid changes in urine hippurate and 3-HPPA also were observed in young rats.44 It is likely that the changes that we observed in aromatic metabolic species in dog urine are associated with establishment of a stable gut microbiota. In supporting this, we observed changes in urinary levels of aliphatic amines, including trimethylamine (TMA), DMA, DMG, and TMAO. For example, the DMA concentration was high at age 13 weeks, whereas relative concentrations of TMA and DMG were high at age 9 years. These aliphatic amines result from degradation of dietary choline by the gut microbiota, and fluctuation of these metabolites as well as certain aromatic metabolites (described above) likely indicates time-related changes in the activities or populations of the gut microbiota.45 These findings assume new significance in the light of recent studies highlighting the importance of gut microbial activities in the development of insulin resistance48 and in the altered balance of gut Bacteroidetes to Firmicutes ratio in obese humans (who have lower Bacteroidetes levels).49 Thus, these data are consistent with the growing realization that diet-related changes in the gut microbiota may be influential in the development of human obesity (and by inference also in pet animals) as they can potentially increase the calorific bioavailability from the diet.49,50 Journal of Proteome Research • Vol. 6, No. 5, 2007 1851

research articles

Wang et al.

Figure 6. Coefficient plot obtained from O-PLS-DA of 1H NMR spectra of urine obtained from dogs fed with control and restricted diets at ages 13 weeks (A), 1.5 years (B), and 9 years (C). Table 2. Changes of Metabolites Associated with Restricted Dietary Interventionsa

alanine acetate N-acetylglycoproteins creatine DMA hippurate 4-hydroxyphenylacetic acid lactate N-methylnicotinamide PAG succinate b

chemical shift(δ, ppm)

13 weeks (Q2 ) 0.48, R2X ) 0.31)

1.5 years (Q2 ) 0.19, R2X ) 0.26)

9 years (Q2 ) 0.48, R2X ) 0.23)

1.48 1.924 2.1 3.0465 2.724 7.831 7.22 1.3285 9.28 7.36 (3.76) 2.41

+0.3917 NSDb -0.5693 -0.4179 NSD NSD +0.3888 NSD NSD +0.3713 NSD

NSD -0.3968 NSD NSD +0.4218 +0.4378 NSD NSD NSD NSD NSD

NSD -0.4474 NSD -0.5593 NSD +0.5304 +0.5687 -0.4939 -0.6429 +0.4926 -0.4453

a A positive value indicates an increase in the concentration of metabolites, and a negative value incicates a decrease in the concentration of metabolites. No significant difference.

The concentration of urine lactate was increased in older dogs. In older humans, increased lactate dehydrogenase activity is more pronounced in the cohort that is aged beyond 90 years,46 which parallels the observation of elevated urinary lactate in older dogs. In the same study of older humans, a marked increase in γ-glutamyltransferase activity was observed in both older men and women,46 suggesting changes in liver function or in dynamics of enzyme metabolism. Although aging dominated urine metabolic profiles of dogs in our study, DR also had an indisputable impact on metabo1852

Journal of Proteome Research • Vol. 6, No. 5, 2007

lism. The most prominent difference due to DR was the progressively increasing concentrations of aromatic metabolites, such as hippurate, PAG, and 4-HPPA. These aromatic metabolites are strongly associated with the gut microbiota, and elevation of these metabolites in urine suggests that DR modified the gut microbiota. Furthermore, this modification was a lengthy and dynamic process, as suggested by the observations of different metabolites at different stages of the dogs’ lives. It is interesting that similar results were obtained in rats fed a standard diet, compared with rats fed a casein-

Metabolic Effects of Diet and Aging

rich diet.16 Moreover, reduced concentrations of creatine, 1-methylnicotinamide, lactate, acetate, and succinate were observed in urine of dogs fed with RD compared with those on the CR regime. These metabolites are associated with energy metabolism and are the end products of compounds involved in energy metabolism. For example, 1-methylnicotinamide is metabolized from nicotinamide adenine dinucleotide and nicotinamide adenine dinucleotide phosphate, while lactate and acetate are end products of glycolysis and fatty acid β-oxidation, respectively. The results observed here are indicative of reduced energy expenditure in calorie-restricted animals and may be integrally associated with the delayed aging in calorie-restricted animals.9,47 Clearly, due to the complexity of the study, which lasted the life span of the dogs (over 14 years), selective analysis of data from three age groups only approximately captured changes caused by aging and calorie restriction. A more comprehensive approach to the data analysis is necessary to fully encapsulate biological processes of aging, which is currently under way.

Conclusion Changes in urinary metabolic signatures were investigated for the duration of the lives of paired sibling dogs fed as controls or with 25% DR. Age was the dominating factor influencing the metabolic trajectory and was mainly associated with the increased excretion of creatinine up to adulthood, followed by a decrease in later life that occurred roughly in parallel with declining lean body composition.1,4 In addition, a relatively high excretion of mixed glycoproteins was noted in dogs at early ages. Changes in gut microbiotal metabolism were associated with both aging and dietary restriction, and these may be of significance in relation to development of obesity. Additional effects of dietary restriction were associated with reduced energy expenditure manifested by depleted levels of creatine, 1-methylnicotinamide, lactate, acetate, and succinate in urine of dogs fed with DR. This study has also highlighted the benefits of using a metabonomic strategy for the detection of subtle physiological changes and dietary effects on mammalian metabolism. The role which gut microflora plays in longevity and quality-of-life responses to DR is potentially important and merits further investigation. Abbreviations: CF, control feeding; COSY, correlation spectroscopy; DMA, dimethylamine; DMG, dimethylglycine; HPPA, hydroxyphenylpropionic acid; NMR, nuclear magnetic resonance; O-PLS-DA, orthogonal projection on latent structure discriminant analysis; PAG, phenylacetylglycine; RD, restricted diet; STOCSY, statistical total correlation spectroscopy: TOCSY, total correlation spectroscopy; TMA, trimethylamine; TMAO, trimethylamine N-oxide.

Acknowledgment. We thank Dr. O. Cloarec for allowing us to use the O-PLS-DA MATLAB tool for data analysis and helpful discussions. This investigation received financial support from Nestle´. References (1) Kealy, R. D.; Lawler, D. F.; Ballam, J. M.; Mantz, S. L.; Biery, D. N.; Greeley, E. H.; Lust, G.; Segre, M.; Smith, G. K.; Stowe, H. D. J. Am. Vet. Med. Assoc. 2002, 220 1315-1320. (2) Masoro, E. J. J. Gerontol. 1988, 43, B59-B64. (3) Li, D. Y.; Sun, F.; Wang, K. Y. Biochem. Biophys. Res. Commun. 2003, 309, 457-463. (4) Lawler, D. E.; Evans, R. H.; Larson, B. T.; Spitznagel, E. L.; Ellersieck, M. R.; Kealy, R. D. J. Am. Vet. Med. Assoc. 2005, 226, 225-231.

research articles (5) Lane, M. A.; Ingram, D. K.; Roth, G. S. Toxicol. Sci. 1999, 52, 4148. (6) Hansen, B. C.; Bodkin, N. L.; Ortmeyer, H. K. Toxicol. Sci. 1999, 52, 56-60. (7) Kemnitz, J. W.; Roecker, E. B.; Weindruch, R.; Elson, D. F.; Baum, S. T.; Bergman, R. N. Am. J. Physiol. 1994, 266, E540-E547. (8) Sinclair, D. A. Mech. Ageing Dev. 2005, 126, 987-1002. (9) Roth, G. S.; Ingram, D. K.; Black, A.; Lane, M. A. Eur. J. Clin. Nutr. 2000, 54, S15-S20. (10) Barja, G. Biol. Rev. 2004, 79, 235-251. (11) Merry, B. J. Int. J. Biochem. Cell Biol. 2002, 34, 1340-1354. (12) Del Roso, A.; Vittorini, S.; Cavallini, G.; Donati, A.; Gori, Z.; Masini, M.; Pollera, M.; Bergamini, E. Exp. Gerontol. 2003, 38, 519-527. (13) Bitterman, K. J.; Anderson, R. M.; Cohen, H. Y.; Latorre-Esteves, M.; Sinclair, D. A. J. Biol. Chem. 2002, 277, 45099-45107. (14) Lin, S. J.; Defossez, P. A.; Guarente, L. Science 2000, 289, 21262128. (15) Hunt, N. D.; Hyun, D. H.; Allard, J. S.; Minor, R. K.; Mattson, M. P.; Ingram, D. K.; de Cabo, R. Ageing Res. Rev. 2006, 5, 125-143. (16) Bell, J. D.; Sadler, P. J.; Morris, V. C.; Levander, O. A. Magn. Reson. Med. 1991, 17, 414-422. (17) Williams, R. E.; Lenz, E. A.; Evans, J. A.; Wilson, I. D.; Granger, J. H.; Plumb, R. S.; Stumpf, C. L. J. Pharm. Biomed. Anal. 2005, 38, 465-471. (18) Nicholson, J. K.; Lindon, J. C.; Holmes, E. Xenobiotica 1999, 29, 1181-1189. (19) Nicholson, J. K.; Connelly, J.; Lindon, J. C.; Holmes, E. Nat. Rev. Drug Discovery 2002, 1, 153-161. (20) Nicholson, J. K.; Wilson, I. D. Nat. Rev. Drug Discovery 2003, 2, 668-676. (21) Nicholson, J. K.; Holmes, E.; Lindon, J. C.; Wilson, I. D. Nat. Biotechnol. 2004, 22, 1268-1274. (22) Gartland, K. P. R.; Beddell, C. R.; Lindon, J. C.; Nicholson, J. K. J. Pharm. Biomed. Anal. 1990, 8, 963-968. (23) Cloarec, O.; Dumas, M. E.; Trygg, J.; Craig, A.; Barton, R. H.; Lindon, J. C.; Nicholson, J. K.; Holmes, E. Anal. Chem. 2005, 77, 517-526. (24) Trygg, J. J. Chemom. 2002, 16, 283-293. (25) Go, V. L. W.; Wong, D. A.; Wang, Y.; Butrum, R. R.; Norman, H. A.; Wilkerson, L. J. Nutr. 2004, 134, 3513S-3516S. (26) Klaus, S. Curr. Drug Targets 2004, 5, 241-250. (27) Lindon, J. C.; Holmes, E.; Bollard, M. E.; Stanley, E. G.; Nicholson, J. K. Biomarkers 2004, 9, 1-31. (28) Wang, Y. L.; Holmes, E.; Nicholson, J. K.; Cloarec, O.; Chollet, J.; Tanner, M.; Singer, B. H.; Ultzinger, J. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 12676-12681. (29) Solanky, K. S.; Bailey, N. J. C.; Beckwith-Hall, B. M.; Davis, A.; Bingham, S.; Holmes, E.; Nicholson, J. K.; Cassidy, A. Anal. Biochem. 2003, 323, 197-204. (30) Teague, C.; Holmes, E.; Maibaum, E.; Nicholson, J.; Tang, H. R.; Chan, Q. N.; Elliott, P.; Wilson, I. Analyst 2004, 129, 259-264. (31) Wang, Y. L.; Tang, H. R.; Nicholson, J. K.; Hylands, P. J.; Sampson, J.; Holmes, E. J. Agric. Food Chem. 2005, 53, 191-196. (32) Lawler, D. F.; Ballam, J. M.; Meadow, R.; Larson, B. T.; Li, Q.; Stowe, H. D.; Kealy, R. D. Exp. Gerontol. 2007, 42, 204-214. (33) Hurd, R. E. J. Magn. Reson. 1990, 87, 422-428. (34) Bax, A.; Davis, D. G. J. Magn. Reson. 1985, 65, 355-360. (35) Cloarec, O.; Dumas, M. E.; Craig, A.; Barton, R. H.; Trygg, J.; Hudson, J.; Blancher, C.; Gauguier, D.; Lindon, J. C.; Holmes, E.; Nicholson, J. Anal. Chem. 2005, 77, 1282-1289. (36) Nicholls, A. W.; Mortishire-Smith, R. J.; Nicholson, J. K. Chem. Res. Toxicol. 2003, 16, 1395-1404. (37) Bell, J. D.; Brown, J. C. C.; Nicholson, J. K.; Sadler, P. J. FEBS Lett. 1987, 215, 311-315. (38) Davies, K. M.; Heaney, R. P.; Rafferty, K. Metabolism 2002, 51, 935-939. (39) Perrone, R. D.; Madias, N. E.; Levey, A. S. Clin. Chem. 1992, 38, 1933-1953. (40) Davies, D. F.; Shock N. W J. Clin. Invest. 1950, 29, 496-507. (41) Evans, W. J. J. Gerontol., A 1995, 50, 5-8. (42) Hurley, B. F. J. Gerontol., A 1995, 50, 41-44. (43) Williams, R. E.; Eyton-Jones, H. W.; Farnworth, M. J.; Gallagher, R.; Provan, W. M. Xenobiotica 2002, 32, 783-794. (44) Williams, R. E.; Lenz, E. M.; Lowden, J. S.; Rantalainen, M.; Wilson, I. D. Mol. Biosyst. 2005, 1, 166-175. (45) Zhang, A. Q.; Mitchell, S. C.; Smith, R. L. Food Chem. Toxicol. 1999, 37, 515-520. (46) Tietz, N. W.; Shuey, D. F.; Wekstein, D. R. Clin. Chem. 1992, 38, 1167-1185.

Journal of Proteome Research • Vol. 6, No. 5, 2007 1853

research articles (47) Weindruch, R.; Walford, R. L.; Fligiel, S.; Guthrie, D. J. Nutr. 1986, 116, 641-654. (48) Dumas, M. E.; Barton, R. H.; Toya, A.; Cloarec, O.; Blancher, C.; Rothwell, A.; Fearnside, J.; Tatoud, R.; Blanc, V.; Lindon, J. C.; Mitchell, S. C.; Holmes, E.; McCarthy, M. I.; Scott, J.; Gauquier, D.; Nicholson, J. K. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 1251112516.

1854

Journal of Proteome Research • Vol. 6, No. 5, 2007

Wang et al. (49) Ley, R. E.; Turnbaugh, P. J.; Klein, S.; Gordon, J. I. Nature 2006, 444, 1022-1023. (50) Turnbaugh, P. J.; Ley, R. E.; Mahowald, M. A.; Magrini, V.; Mardis, E. R.; Gordon, J. I. Nature 2006, 444, 1027-1031.

PR060685N