Metabolomics Insights into the Modulatory Effects ... - ACS Publications

Subscriber access provided by UNIV OF CAMBRIDGE

Article

Metabolomics insights into the modulatory effects of long term low calorie intake in mice Junfang Wu, Liu Yang, Shoufeng Li, Ping Huang, Yong Liu, Yulan Wang, and Huiru Tang J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.6b00336 • Publication Date (Web): 07 Jun 2016 Downloaded from http://pubs.acs.org on June 11, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of Proteome Research is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Journal of Proteome Research

Title page: Metabolomics insights into the modulatory effects of long term low calorie intake in mice

Junfang Wu†, Liu Yang‡, Shoufeng Li‡, Ping Huang‡, Yong Liu‡, *, Yulan Wang†, §, *, Huiru Tang†, #, * †

Key Laboratory of Magnetic Resonance in Biological Systems, State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, Wuhan Centre for Magnetic Resonance, Wuhan Institute of Physics and Mathematics, Chinese Academy of Sciences, Wuhan, 430071, P. R. China ‡ Key Laboratory of Nutrition and Metabolism, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai, 200031, P.R. China § Collaborative Innovation Center for Diagnosis and Treatment of Infectious Diseases, Zhejiang University, Hangzhou 310058, P. R. China. # State Key Laboratory of Genetic Engineering, Collaborative Innovation Center for Genetics and Development, Metabolomics and Systems Biology Laboratory, School of Life Sciences, Fudan University, Shanghai, 200433, P.R. China. Email address: Junfang Wu: [email protected] Liu Yang: [email protected] Shoufeng Li: [email protected] Ping Huang: [email protected] Yong Liu: [email protected] Huiru Tang: [email protected] Yulan Wang: [email protected]

*

Authors for correspondence:

Huiru Tang: E-mail: [email protected] Yulan Wang: E-mail: [email protected] Yong Liu: E-mail: [email protected]

1

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 34

ABSTRACT There is increasing evidence that calorie restriction (CR) without malnutrition can extend longevity and delay the onset of age associated disorders. Identifying the biochemical perturbations associated with different dietary habits would provide valuable insights into associations between metabolism and longevity. In order to reveal the effects of long-term dietary interventions on the metabolic perturbations, we investigated the serum and urinary metabolic changes induced by interactive high/low fat diet in combination with/without reduced caloric intake over a life span in mice using NMR based metabonomics. We found that the high calorie dietary regime disturbed lipid metabolism, suppressed glycolysis and TCA cycles, stimulated oxidative stress, promoted nucleotide metabolism and gluconeogenesis, and perturbed gut microbiota-host interactions. Such changes could be modified by long term low calorie intake. Most importantly, we found that the calorie intake index exerts a dominant effect on metabolic perturbations, irrespective of dietary regime. Our investigation provides a holistic view on the metabolic impact of long term dietary interventions, which are important for detecting physiological changes and dietary effects on mammalian metabolism.

Keywords: Calorie restriction, High fat diet, NMR, Metabonomics, Long-term diet interventions

2


Page 3 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


INTRODUCTION The beneficial nutrients contained in food are essential for maintaining human health and improving the quality of life. A healthy diet has been the focus of attention in both scientific and social communities for decades1, however overconsumption of high fat dietary components is an increasing issue in both developed and developing countries. High fat diets are responsible for changes in the expressions of several important genes 2, 3, and a range of proteins and metabolites 4, 5 involved in energy and lipid metabolism, and are known to contribute to a number of chronic diseases such as obesity, diabetes mellitus, coronary heart disease, fatty liver disease and cancer 1, 6, 7. Calorie restriction (CR) or energy restriction without malnutrition has long been recognized as a natural interventional strategy for promoting health and extend longevity ability to ameliorate cardiac dysfunction

10

8, 9

, and has a significant

and reduce the risk of atherosclerosis

11

by reducing

reactive oxygen species in CR. This has been demonstrated in a number of species including yeast, worm, rodents, dogs and humans

12-14

. CR has also been shown to modulate the levels of hormones

(triiodothyronine and insulin) 12 and lipid metabolites (triacylglycerol, free fatty acids, ceramide and phosphatidylcholine) 13, 15, 16, as well as alter energy metabolism (creatine and succinate) 17. However, the biochemical alterations as a result of CR have not been fully established. In addition, the effects of long-term CR on the regulation of homeostatic control system remain unknown, and in particular, the interactive metabolic changes of CR with low fat diet/high fat diet. RNA and protein play important roles in controlling the flux of metabolites through biochemical pathway, and the effects of nutritional interventions have been assessed on a molecular level, using different platforms, including transcriptomics

18

, proteomics and metabonomics

19

. Whilst

transcriptomics and proteomics monitor processes of biological perturbations, metabonomics 3



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 34

holistically captures the system end point responses to these biological perturbations by measuring the chemical compositions of biological samples, such as biofluids and tissues 20. Metabonomics could therefore be a useful tool for measuring the metabolic changes related to the combinations of high/low-fat diets and CR, by analyzing the metabolic profiles of plasma or urine. Previously, NMR/MS based metabonomics investigations have revealed that a high fat diet induced an adaptive regulation of lipid storage 21, 22, perturbations of nicotinamide, fatty acid and amino acid metabolisms, and gut microflora activities, as compared with low fat diet 23-25. CR has also been found to partially prevent such alterations, particularly by rebalancing fatty acid metabolism and modulating bile acid biosynthesis

17, 26

. Despite these studies, there has been less focus on the way in which combining

life-long CR with high/low fat diet impacts on metabolism. Importantly, it is also unclear whether the metabolic changes related to a long-term high fat diet could be reversed in a calorie restricted or low fat diet. In this study, we used a mouse model, and NMR based metabonomics combined with multivariate data analysis to identify the biochemical processes that arise from combining long-term CR with a high/low fat diet. The objective of our study was to define the metabolic responses to different levels of long term energy intake. Our study provides insight into the metabolic processes that are important for understanding dietary modulatory effects on mammalian metabolism, and the associations between metabolism and longevity.

MATERIALS AND METHODS Animal Handling and Sample Collection The present study was carried out in accordance with the approved guidelines for experimental 4


Page 5 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


research established by Institutional Animal Care and Use committee of Institute of Nutritional Sciences, Chinese Academy of Sciences, China. All experimental procedures and protocols were approved by Shanghai Institute for Nutritional Sciences, Chinese Academy of Sciences. Male C57BL/6J mice at 4 weeks of age were purchased from Shanghai Animal Co, Ltd and housed in an environmentally controlled facility (23 ± 3°C temperature; 35 ± 5% relative humidity; 12 h:12 h dark/light cycle). After one week of acclimation with a low-fat diet (Research Diets), the animals were randomly divided into four dietary intervention groups (n = 30 for each group): low-fat diet (LF, 10% calories from fat, 20% calories from protein, 70% calories from carbohydrate, D12450B, Research Diets, Inc.), low-fat diet with 30% reduced calorie intake (LR); high-fat diet (HF, 60% calories from fat, 20% calories from protein, 20% calories from carbohydrate, D12492, Research Diets, Inc.) and high-fat diet with 30% reduced calorie intake (HR) (Figure S1). The formulas of the low fat diet and high fat diet were listed in Supplemental Table S1, as well as typical fatty acid composition of fats in both diets was listed in Table S2. All mice were housed individually during the study. The daily consumption of food in LF and HF groups was recorded and averaged once a week to calculate the amount of food for the reduced calorie intake groups in the following week 27, 28. At 58 weeks of dietary intervention, eight mice of each group were randomly sacrificed and the blood was collected via heart puncture. After centrifugation at 4,000 g for 30 min, serum was collected and stored at -80 °C. Urine samples of the remaining mice were collected at 74 weeks of dietary intervention (n=22 for each group) and stored at -80°C. The mice were individually housed in metabolic cage for collecting the overnight urine (6:30 PM-6:30 AM). The 0.02% sodium azide was used as preservative. The urine samples contaminated with food particles and/or feces were excluded 5



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 34

from the study.

1

H NMR Spectroscopy Serum samples were prepared by adding 200 µL of serum (or topped up to 200 µL with saline)

to 400 µL 0.9% saline solution (VD2O:VH2O=1:1) containing 45 mM phosphate buffer (pH 7.4). The serum-buffer mixture was then centrifuged at 10000 g for 10 min and 550 µL of the supernatant was transferred to a 5 mm NMR tube. The serum NMR spectra were acquired at 298K under automatic sample changer system (BACS, Bruker Biospin, Germany) using a Bruker AVIII 600 MHz spectrometer equipped with an inverse cryoprobe (operating at 600.13MHz for 1H). One dimensional (1D)

1

H NMR spectra were acquired for each serum sample using water-suppressed

Carr-Purcell-Meiboom-Gill (CPMG) spin-echo pulse sequence (RD-90°-(τ-180°-τ)n-ACQ) for suppressing signals from macromolecules. Spin-echo loop time (2nτ) was set to 70 ms with a recycle delay of 2.0 s. In addition, a water-suppressed diffusion-edited pulse sequence (RD-90°-G1-τ-180°G2-τ-90°-∆-90°-G3-τ-180°-G4-τ-90°-Te-90°-ACQ) was used for suppressing peaks from low molecular weight metabolites in serum. The diffusion-edited spectrum was obtained using a recycle delay of 2.0 s, eddy delay τ of 5 ms, and a diffusion delay ∆ of 200 ms, during which small molecules were allowed to diffuse. The 90° pulse length was adjusted to approximately 10 µs, and 64 transients were collected into 32 k data points for each spectrum with a spectral width of 20 ppm. Urine samples were prepared by mixing 550 µL of urine with 55 µL of 1.5 M of deuterated phosphate buffer (NaH2PO4 and K2HPO4, pD 7.47) containing 0.01% TSP (sodium 3-(trimethylsilyl) propionate-2,2,3,3-d4) to serve as a chemical shift reference (δ0.0) 29. The urine-buffer mixture was left to stand for 5 min at room temperature and then centrifuged at 10,000 g for 10min. The 6


Page 7 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


supernatant (550 µL) was then transferred to a 5 mm NMR tube. Urine spectra were also randomly acquired using a BACS automation system with a Bruker AVIII 600 MHz spectrometer equipped with an inverse cryoprobe. The 1D 1H NMR spectra of the urine samples were also acquired at 298K using the first increment of NOESY pulse sequence (RD-90°-t1-90°-tm-90°-ACQ). Water suppression was achieved with an irradiation on the water peak during the recycle delay of 2.0 s and a mixing time, tm, of 100 ms. t1 was set to 4 µs. The 90° pulse length was adjusted to approximately 10 µs and 32 transients were collected into 32 k data points for each spectrum with a spectral width of 20 ppm. Standard 2D 1H-1H COSY, 1H-1H TOCSY, 1H-13C HMBC, 1H-13C HSQC and J-resolved spectra were also acquired for the purpose of metabolite resonance assignment from selected samples.

Data Processing and multivariate data analysis All free induction decays were multiplied by an exponential function equivalent to a 1Hz line-broadening factor prior to Fourier transformation. The 1H NMR spectra were manually corrected for phase and baseline distortion using Topspin 2.1. The spectra of urine samples were referenced to TSP at δ 0.00, and the chemical shifts of serum spectra were referenced to the internal lactate CH3 resonance at δ 1.33. After peak alignment of some metabolites (citrate, taurine and creatine) with in-house Matlab script, 1H NMR spectrum (δ10.0-0.5) of urine sample were then automatically reduced to 3995 integral segments of equal length (0.002 ppm) using AMIX (v3.8, Bruker Biospin). The regions of water resonance (δ5.06-4.67), urea resonance (δ6.33-5.28) were removed to eliminate baseline effects of imperfect water suppression and urea signals. In addition, signals of tartarate (δ4.40-4.32) and choline bitartrate originating from extraneous dietary sources were also excluded 30. 7



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 34

Urinary data were normalized to the total sum of spectrum for further multivariate data analysis. The serum CPMG spectrum over the range of δ9.0-0.5 was also divided into 0.002 ppm per bucket size. The regions of water resonance (δ5.17-4.36), urea resonance (δ6.10-5.60) and ethanol contamination from sterilization procedure (δ3.69-3.63 and δ1.21-1.15) were removed. For diffusion-edited serum spectrum, a region of δ5.50-0.50 excluding water resonance (δ5.22-4.38) were integrated with the same bucket length. All serum spectra were integrated without normalization for further analysis. Principal Component Analysis (PCA) was performed using a mean-centered approach, using SIMCA-P 12.0 (Umetrics, Sweden) to view the group separation and find possible outliers. Data were visualized by the principal component (PC) score plots. Each point on the scores plot is defined by the spectrum of an individual sample. O-PLS-DA was performed to maximize separation between different groups by using the pareto-scaled approach. The quality of the O-PLS-DA model used for dietary pair-wise comparison was described by cross validation parameter Q2 and R2 and ensured with a CV-ANOVA approach (p < 0.05) 31. The correlation coefficients from cross validated model represent the importance of the NMR variables in discriminating different specific dietary regions. Such coefficients were color-coded and plotted together with back-transformed weight of the variables in order to enhance interpretability of the model 32. Thus, metabolites responsible for the differences detected in the scores plot were extracted from the coefficient plot. In our study, a correlation coefficient (|r|) higher than 0.666 (for serum samples) or 0.456 (for urine samples) was taken as statistical significant (p < 0.05). For further in-depth understanding of the lipid peroxidation after a high fat diet, the percentage of fatty acid composition in the serum profile, including unsaturated fatty acids (UFA%, δ5.499-5.223), polyunsaturated fatty acids (PUFA%, δ2.853-2.731), total fatty acids (TFA%, 8


Page 9 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


δ1.033-0.809), monounsaturated fatty acids (MUFA%), saturated fatty acids (SFA%) and the ratio of PUFA to MUFA, were also calculated from diffusion-edited NMR spectra 33. The effects of different dietary intervention groups were presented as mean ± SEM and analyzed by two-way ANOVA (p < 0.05).

RESULTS NMR spectra of mouse serum and urine Typical 1H NMR spectra of urine and serum were obtained from low fat diet (LF) and low fat diet with reduced calorie intake (LR) mice (Figure 1A-D). Resonance peaks were assigned to specific metabolites based on literature data with further confirmation by 2D NMR spectra

34, 35

.A

number of low molecular weight metabolites were observed in the LR and LF mice urinary spectra, such as tricarboxylic acid cycle (TCA) related metabolites (citrate, 2-oxo-glutamate (2-OG), succinate and fumarate), gut-microbial related metabolites (hippurate, p-cresol sulfate (p-CS), p-cresol glucuronide (p-CG), phenylacetylglycine (PAG), p-hydroxyphenylacetate, DMA, TMA and TMAO),

as

well

as

nicotinate-related

N1-methyl-2-pyridone-5-carboxamide

metabolites

(2-PY),

(N-methylnicotinamide

3-ureidopropionate

(3-UP),

(NMND), creatine,

N1-methyl-4-pyridone-3-carboxamide (4-PY)), and indoxylsulfate (Figure 1A and B). Metabolites identified from serum spectra were glucose, citrate, creatine, and membrane component-related metabolites, such as phosphorylcholine (PC) and glyceryl phosphorylcholine (GPC) as well as a range of amino acids (valine, leucine, isoleucine, alanine and lysine ) (Figure 1C and D). In addition, serum obtained from the LF group displayed higher levels of unsaturated fatty acids (UFA), polyunsaturated fatty acids (PUFA) and N-acetyl-glycoproteins than those from the LR group. The 9



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 34

relative levels for different classes of fatty acids in plasma were calculated from the diffusion-edited spectra as previously reported

33

. These calculations were based on the spectral integral areas for

–CH=CH- (from UFA, δ5.499-5.223), =CH-CH2-CH= (from PUFA, δ2.853-2.731) and -CH3 (from all fatty acids, δ1.033-0.809) taking into consideration of proton numbers.

Long term reduced caloric intake induced metabolic changes The effects of CR in combination with different dietary habits (low or high fat diet) on the metabolic profiles of urine were investigated using O-PLS-DA (Figure 2A and B, Table 1). The coefficient plots showed that under a calorie restricted diet, both mice on the high fat (HF) and low fat diet had comparable changes to their urinary metabolic profiles, showing that reduced calorie intake alone affects the metabolic profile of urine. Reduced caloric intake resulted in elevated levels of TCA cycle intermediates (citrate and cis-aconitate), depleted levels of nucleotide-related metabolites (NMND, 2-PY and 4-PY) and alterations of some gut-microbial related metabolites (hippurate, indoxylsulfate and p-hydroxyphenylacetate). O-PLS-DA of serum profiles showed that CR is able to reduce the levels of lipoproteins in serum, which is independent of dietary habits, i.e. high or low fat diets (Figure 2C and D, Table 2). In addition, decreased levels of glucose, TMA and TMAO were only displayed in the group of CR in combination with high fat diet, whilst down-regulation of p-CG and p-CS and β-oxidation of lipid products, such as acetone and 3-hydroxybutyrate (3-HB), was only noted in the CR plus low fat diet group (Figure 2, Table 2).

Long term high fat diet induced metabolic changes The effects on metabolic profiles of high fat versus low fat diet within dietary restriction or 10


Page 11 of 34

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


normal diet groups were also investigated by O-PLS-DA for urine (Figure 3A and B) and serum (Figure 3C and D). The urinary profiles as a result of a high fat diet in normal food intake (Figure 4A, Table 1) showed elevations in the levels of glucose, creatine and gut microbial related metabolites, such as p-CG and p-CS, together with depletions in the levels of indoxylsulfate, citrate and succinate (Figure. 3A). For calorie restricted animals (Figure 3B), a high fat diet induced extra elevations of NMND, 2-PY, 4-PY and depleted levels of 2-OG, acetate and lactate, as compared with a low fat diet (Table 1). In serum, we observed marked higher levels of lipoprotein lipids, such as PC and GPC, and lower levels of citrate in the high fat diet group as compared with the low fat diet group. These changes were noted in both normal and calorie restricted high fat diets. In addition, the calorie restricted high fat diet group showed higher concentrations of glucose, choline and triglycerides, and lower concentrations of succinate as compared to the calorie restricted low fat diet group (Figure 3D, Table 2).

Correlation between metabolites and calorie intake index During the study, we recorded the average energy intake and body weight for each dietary intervention (Supplemental Table S3 and Table S4) 28. Of note, the LR group had the lowest energy intake, while the HF group had the highest energy intake, at all-time points. The energy intake was highly correlated with body weight (r=0.79, p

Metabolomics Insights into the Modulatory Effects ... - ACS Publications

Recommend Documents