Comparative metabolomics elucidates postprandial metabolic

Subscriber access provided by NEW MEXICO STATE UNIV

Article

Comparative metabolomics elucidates postprandial metabolic modifications in plasma of obese individuals with metabolic syndrome Mengyang Xu, Fanyi Zhong, Richard S. Bruno, Kevin D. Ballard, Jing Zhang, and Jiangjiang Zhu J. Proteome Res., Just Accepted Manuscript • DOI: 10.1021/acs.jproteome.8b00315 • Publication Date (Web): 05 Jul 2018 Downloaded from http://pubs.acs.org on July 6, 2018

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 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 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.

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 30 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

Comparative metabolomics elucidates postprandial metabolic modifications in plasma of obese individuals with metabolic syndrome

Mengyang Xu1#, Fanyi Zhong1#, Richard S. Bruno2, Kevin D. Ballard3, Jing Zhang4, and Jiangjiang Zhu1* 1. Department of Chemistry and Biochemistry, Miami University, Oxford, OH, 45056 2. Human Nutrition Program, The Ohio State University, Columbus, OH, 43210 3. Department of Kinesiology and Health, Miami University, Oxford, OH, 45056 4. Department of Statistics, Miami University, Oxford, OH, 45056

# These two authors contribute equally to this work * Corresponding author, Email: [email protected] Tel: 513-529-3998

1 ACS Paragon Plus Environment

Journal of Proteome Research 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

Abstract Although higher intakes of dairy milk are associated with a lower risk of metabolic syndrome (MetS), the underlying protective mechanism remains unclear. This study investigated the dynamic metabolic profile shift following the ingestion of low-fat milk or an isocaloric volume of rice milk in obese individuals with metabolic syndrome (MetS). In a randomized, doubleblind, crossover study, postprandial plasma samples (n = 266) were collected from 19 MetS participants. Plasma samples were analyzed by a targeted metabolomics platform which specifically detects 117 metabolites from 25 metabolic pathways. The comprehensive timecourse metabolic profiling in MetS participants indicated that the postprandial metabolic profiles distinguish low-fat milk and rice milk consumption in a time-dependent manner. Metabolic biomarkers, such as orotate, leucine/isoleucine and adenine, showed significantly different trends in the two test beverages. Bayesian statistics identified 12 metabolites associated with clinical characteristics of postprandial vascular endothelial function, such as flow-mediated dilation (FMD), postprandial plasma markers of oxidative stress and NO status. Furthermore, metabolic pathway analysis based on these metabolite data indicated the potential utility of metabolomics to provide mechanistic insights of dietary interventions to regulate postprandial metabolic excursions.

Key words: Targeted metabolic profiling, nutrimetabolomics, low-fat milk, rice milk, Bayesian LASSO


Page 2 of 30

Page 3 of 30 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 Metabolic syndrome (MetS) and its associated morbidities, including obesity, hypertension, and dyslipidemia, increase cardiovascular disease (CVD)-related mortality and is a worldwide public health concern.1 Dietary modification is an effective strategy to manage cardiometabolic risk factors associated with MetS. Indeed, greater intakes of dairy products are associated with lower risk of MetS and CVD-related morbidity.2 While these epidemiological and observational studies are promising, the mechanisms by which ingestion of dairy foods mitigate MetS risk remain unclear. Mechanisms potentially explaining the association between greater dairy consumption and lower CVD risk have been reviewed.3 However, the lack of sensitive and/or robust techniques to detect and monitor physiological conversion of bioactive constituents in dairy into various metabolites, and/or poor understanding as to how these metabolites interact to mediate clinical outcomes, may explain the conflicting evidence regarding the health benefits of dairy. Metabolomics is the analysis of all or a subset of metabolites in a biological system, whereas nutrimetabolomics refers to the application of metabolomics in nutritional sciences. Nutrimetabolomics was developed to identify the complex relationships between dietary consumption and health outcomes and to define endogenous metabolic shifts in response to dietary interventions in both preclinical and clinical models.4-6 In the present study, we applied a mass spectrometry (MS)-based targeted metabolic profiling approach to detect the metabolic profile of MetS participants who previously completed a randomized clinical trial investigating the postprandial effects of low-fat dairy milk or rice milk ingestion on vascular endothelial function (VEF).7 We found that acute ingestion of low-fat milk, but not rice milk, maintained VEF by limiting postprandial hyperglycemia (PPH)-dependent increases in oxidative stress and reductions in nitric oxide (NO) bioavailability in 19 obese adults with MetS.7 We hypothesized



that the targeted metabolomics can be used as a novel approach to define MetS patients’ metabolic regulatory system response to a food-based intervention. Utilizing plasma specimens archived from our previous clinical trial,7 the objective of the present study was to apply a targeted MS-based metabolomics platform to determine the relationship between postprandial responses of host metabolic changes to help elucidate how dairy milk ingestion preserves VEF and lowers CVD risk. The information obtained from the analysis of these postprandial samples will assist in guiding novel dietary recommendations aimed at mitigating CVD risk in individuals with MetS. Materials and Methods Participants, study design and clinical characteristics. The protocol for this study was approved by the Institutional Review Board at the University of Connecticut. Written informed consent was obtained from all participants before enrolling, and plasma samples were completely deidentified prior to analysis. A detailed study design schematic is shown in Figure 1 and has been described previously.7 Briefly, participants were non-diabetic, non-smokers, and were not using any medications or dietary supplements during the study. Obese (body mass index (BMI) ≥30 kg/m2) men and pre-menopausal women were screened for the presence of ≥3 of the following established risk factors for MetS8 : waist circumference ≥102 cm for men and ≥88 cm for women, fasting plasma triglycerides (TG) ≥150 mg/dL, fasting plasma glucose ≥100 mg/dL, resting systolic (≥130 mmHg) and diastolic (≥85 mmHg) blood pressure, and fasting plasma HDL-cholesterol 1.3. These data suggest a time-dependent metabolic shift following consumption of beverages differing in carbohydrate and protein content. To demonstrate the most significantly different time point, the detailed box plot of nine significantly altered metabolites (i.e., orotate, leucine/isoleucine, mesoxalate, asparagine, citrulline, methionine, allantoin, ornithine and tyrosine) in low-fat milk vs. rice milk comparison at 120 min is shown in Figure 2. The metabolite level of reported amino acids in the low-fat milk trial was higher compared to the rice milk trial, which matched the higher protein content of low-fat milk (Table S1). Furthermore, multivariate statistical analyses were also applied to understand the trend of time/treatment-dependent metabolic profiles. As shown in Figure 3, partial least squarediscriminant analysis (PLS-DA) did not distinguish the two groups of participants at baseline (red dots = low-fat milk trial; green dots = rice milk trial), which was expected as the beverages have not been consumed. The separation between trials occurred at 60 min and was maintained until 120 min, after which the metabolic profiles of the two beverages merged again, likely due to the complete digestion/absorption of the test beverages. This observation indicates that our specific metabolic profile can be used to detect time-dependent postprandial changes following the ingestion of test beverages differing in protein and carbohydrate content. To define the impact of low-fat milk and rice milk on time/treatment-dependent changes in metabolite concentrations, two-way analysis of variance (ANOVA) was performed to consider 9 ACS Paragon Plus Environment


the test beverage, time, and their interactive effects. As shown in Figure 4A, 13 metabolites showed a significant treatment effect, 44 metabolites showed a significant time effect, and 13 metabolites showed a significant interaction effect. Eight of these metabolites were found in all three comparisons (Table S3). Figure 4B shows that by ANOVA-simultaneous component analysis (ASCA), 10 metabolites (represented by circles in the bottom right quadrant) can be considered as suitable model metabolites to reflect the beverage-specific trend in this clinical nutrition study. The detailed list of these metabolites is presented in Table S4. In crosscomparison of Table S3 and S4, we identified four metabolites that were found in both lists. The four metabolites were orotate, leucine/isoleucine (due to technical constraints, these isomers cannot be separated), adenine and methionine. The treatment- and time-dependent trend for these four metabolites are shown in Figure 5. Time-dependent trend and significantly altered profiles between low-fat and rice milk consumption can be seen for all four metabolites, indicated that these metabolites can potentially be used to differentiate low-fat milk verse rice milk ingestion during the postprandial period. Plasma samples were also analyzed for insulin, glucose, arginine, and MDA concentrations. These results have been previously reported7 and their beverage- and time-dependent trends are shown in Figure S 8a-d. Brachial artery FMD was assessed as a measure of VEF and is shown in Figure S 8e. We investigated the association between metabolite biomarkers and plasma insulin, glucose, arginine, and MDA concentrations and brachial artery FMD. To investigate these associations, a multiplex multivariate statistical analysis approach (Bayesian hierarchical models with Least Absolute Selection and Shrinkage Operator (LASSO) variable selection) was applied. PCA screening at each time point for both trials was first conducted and the frequency of metabolites associated with the 20 highest absolute values of loadings in the selected PCAs was


Page 10 of 30

Page 11 of 30 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


found. The most frequent metabolites that passed the screening stage were then used in the analysis and 32 metabolites passed this level of selection. The findings of this analysis enabled us to test if these metabolites correlated with postprandial plasma insulin, glucose, MDA and arginine concentrations and brachial FMD responses in the low-fat milk and rice milk trials. As shown in Table 1, 12 metabolites were significantly associated with postprandial insulin, glucose, MDA and arginine concentrations and brachial FMD responses independent of beverage ingestion. The point estimator of the effect (PE) and Bayesian credible interval (CI) values indicated the potential utility of these 12 metabolites to monitor postprandial alterations in specific cardiometabolic measures in MetS participants, regardless of the beverage ingested. Positive PE values suggest a positive relationship between the corresponding metabolite and the postprandial response, whereas negative PE values suggest a negative relationship between the corresponding metabolite and the postprandial response. The 90% CI indicated the range of the effect of the metabolite to the corresponding postprandial response. When the 90% CI does not include zero, the corresponding parameter was considered statistically significant in the Bayesian analysis. To further understand the metabolic connections of our detected metabolites in biologically important metabolic pathways, we performed a manual curation of metabolic pathways by searching these important metabolites from Table 1 and Figure 5 in KEGG website (http://www.genome.jp/kegg/). Next, we constructed a simplified metabolic pathway map to understand the biological implication of the two-way ANOVA, ASCA and Bayesian LASSO results. For example, as demonstrated in Figure 6A, the urea cycle metabolite citrulline, an amino acid compound made from ornithine and carbamoyl phosphate, is reported to be also produced from arginine as a by-product of the reaction catalyzed by nitric oxide system (NOS)



family.17 In this reaction, arginine is first oxidized into N-hydroxyl-arginine, which is then further oxidized to citrulline concomitant with the release of nitric oxide. Other metabolic pathway connections, such as the metabolism of arginine to trans-4-hydroxyproline and creatine, and then further metabolized to serine and homoserine/N-acetylserine, respectively, also indicated the possible mechanistic insights of dietary intervention to the systematic metabolic reprogramming in these MetS participants. Figure 6B demonstrated that at the most significant time point of 120 mins, three impacted metabolic pathways in comparison of low-fat milk vs. rice milk ingestion (determined by having impact score >0.5 and -log(p) value > 2), and all three pathways, i.e., alanine, aspartate and glutamate metabolism; glycine, serine and threonine metabolism; and arginine and proline metabolism, are detected at significantly higher level in low-fat milk group in comparison to rice milk group. Discussion In plasma samples obtained during a previously described randomized clinical trial in individuals with MetS,7 we applied an HPLC-MS/MS-based targeted metabolomics approach to elucidate differences in the metabolism of low-fat milk and an isocaloric volume of rice milk during the postprandial period. Taking advantage of the double-blind crossover study design, in this study each individual participant involved in the study can serve as a control for himself/herself in the comparison of low-fat milk and rice milk consumption. We observed significant differences in plasma metabolite profiles between test beverages at 60-120 min post-ingestion. Univariate and multivariate statistical analyses identified a group of detected metabolites, including orotate, leucine/isoleucine, and adenine in this study, that may be utilized as beverage-specific metabolic biomarkers. Additionally, a multiplex statistical model using Bayesian LASSO was developed using our metabolite data together with the time and dietary factors. This model effectively used 12 ACS Paragon Plus Environment

Page 12 of 30

Page 13 of 30 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


a group of detected metabolites to predict postprandial VEF of MetS individuals, which can be reflected by the strength of the metabolite related signal in the variable selection step. Therefore, the present data support the potential utility of our targeted metabolomics approach in future studies aiming to establish the link between functional foods and cardiometabolic health. Nutrition research generally focuses on improving the health of populations and individuals through systematic manipulation of dietary intake. The fields of nutrition- and health-related research are beginning to unveil that specific components of foods interact with numerous human metabolic pathways that influence health and the risk of chronic disease.5 Poor dietary habits are considered a major factor contributing to the rapid increase in the incidence of metabolic disorders, such as obesity, diabetes, and cardiovascular disease.18 Robust dietary assessment methods may provide inroads to better understand the linkage between diet and chronic disease profiles. Conventional methods used to collect quantitative information on dietary intakes (e.g., food diaries and food frequency questionnaires) are subject to possible biases and have been shown to be unreliable for characterizing and quantifying dietary habits.19 In addition, these methods are unreliable for certain groups such as the elderly or obese people, whose self-reported energy intakes tend to be underestimated as assessed by energy expenditure measurements.20 To overcome the problems with using self-reported methods to measure dietary exposure, nutritional scientists have begun to examine metabolic biomarkers to measure individuals’ dietary intake and nutrient status21. The utility of metabolic biomarkers has been suggested to provide a more accurate and objective measurement of dietary intake in comparison to traditional food surveys, as the measurement of metabolic biomarkers account for the bioconversion and metabolism of food/beverage components.6, 21 One common application of these food-related biomarkers is to utilize them as references to assess the validity of dietary



intervention measures. In the present study, we found that the metabolic plasma profiles of individuals with MetS administered different test beverages displayed significantly different patterns during the postprandial period. We consider this finding an advance of metabolomics application in nutrition research. For example, we found higher levels of plasma orotate following ingestion of low-fat milk compared to rice milk (Figure 2). Orotate, formerly recognized as vitamin B-13, is a minor dietary constituent and is abundant in cow’s milk.22 Thus, it was not surprising that orotate was one of the significant distinguishing factors between the two test beverages. Furthermore, our targeted metabolomics approach, in combination with advanced statistical analyses, can provide new insights into the connection between different metabolites and different pathways in chronic disease patients (Figure 6). Further, we suggest that our approach may assist in better understanding how nutritional interventions influence metabolic activities, and how distinct metabolic activities (e.g., the detected level of thiamine monophosphate, thymine, leucine/isoleucine, pyroglutamic acid, N-acetylserine, glyceraldehyde, creatine, and serine) are associated with VEF (Table 1). It is generally acknowledged that the development of robust food-related biomarkers will help to better classify individuals’ dietary intake, and in turn will improve the ability to predict the relationship between dietary intake and chronic diseases, such as MetS. Due to its highsensitivity, high-throughput and rapid turnaround time, targeted MS-based metabolomics should be utilized in future studies to evaluate the impact of dietary interventions on patients with chronic disease. We found that low-fat milk and rice milk ingestion induced different postprandial metabolic profiles in individuals with MetS. For example, at 120 minutes, three metabolic pathways, i.e., alanine, aspartate and glutamate metabolism; glycine, serine and threonine metabolism; and arginine and proline metabolism, are detected at a significantly higher


Page 14 of 30

Page 15 of 30 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


level in low-fat milk group in comparison to rice milk group. In combination with advanced data analysis techniques, we also demonstrated that our MS-based metabolomics approach can connect metabolite information with clinical characteristics of MetS, and the metabolic profiles are in good correlation with postprandial VEF in these patients. With independent validation and increased number of study participants down to the road, findings from the present study can be integrated into additional metabolic pathway analyses to assist future studies mechanistically investigating the impact of both acute and chronic dietary interventions on indices of cardiometabolic health in patients with chronic disease. ASSOCIATED CONTENT Supporting Information Additional Materials and Method section. Table S1. Comparison of isocaloric low-fat milk and rice milk nutritional composition Table S2. HPLC-MS/MS detection method for targeted metabolites in this study, two SRM transitions for each compound were used for most metabolites to ensure confident detection. Table S3. List of 8 metabolites that shown significant treatment, time, and interaction effects. Table S4. List of 10 metabolites having significant p-value in the ANOVA-simultaneous component analysis (ASCA) time-treatment interaction analysis. Table S5. Pathway analysis results in comparison of low-fat milk and rice milk groups at 120 mins. Figure S1. Heatmap of targeted metabolites of 19 participants at 0 minutes (before beverage administration). L: low-fat milk trial; R: rice milk trial. Figure S2. Heatmap of targeted metabolites of 19 MetS participants at 30 minutes post beverage administration. L: low-fat milk trial; R: rice milk trial.



Figure S3. Heatmap of targeted metabolites of 19 MetS participants at 60 minutes post beverage administration. L: low-fat milk trial; R: rice milk trial. Figure S4. Heatmap of targeted metabolites of 19 MetS participants at 90 minutes post beverage administration. L: low-fat milk trial; R: rice milk trial. Figure S5. Heatmap of targeted metabolites of 19 MetS participants at 120 minutes post beverage administration. L: low-fat milk trial; R: rice milk trial. Figure S6. Heatmap of targeted metabolites of 19 MetS participants at 150 minutes post beverage administration. L: low-fat milk trial; R: rice milk trial. Figure S7. Heatmap of targeted metabolites of 19 MetS participants at 180 minutes post beverage administration. L: low-fat milk trial; R: rice milk trial. Figure S8. Time course measurement of clinical characteristics that showing different time trends in low-fat (L) and rice milk (R) groups. (A) Plasma Insulin (pmol/l), (B) Plasma Glucose (mmol/l), (C) Plasma Arginine (µmol/L), (D) Plasma MDA (µmol/L) and (E) Brachial artery FMD (%). This material is available free of charge via the Internet at http://pubs.acs.org.

AUTHOR INFORMATION Corresponding Author Tel: +1 513 529 3998, Fax: +1 513 529 5715, Email address: [email protected] Author Contributions #

F.Z. and M.X. contributed equally to this project.

Conflict of interest The authors claim no conflict of interest.


Page 16 of 30

Page 17 of 30 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


ACKNOWLEDGMENTS This study was supported by Miami University (startup fund to JZ) and a grant from the National Dairy Council (to RSB). We’d like to extend our gratitude to the study participants.



References: 1.

Grundy, S. M., Metabolic Syndrome Pandemic. Atertio. Thromb. Vasc. Biol. 2008, 28 (4), 629-

636. 2.

Lutsey, P. L.; Steffen, L. M.; Stevens, J., Dietary Intake and the Development of the Metabolic

Syndrome. The Atherosclerosis Risk in Communities Study 2008, 117 (6), 754-761. 3.

Ballard, K. D.; Bruno, R. S., Protective role of dairy and its constituents on vascular function

independent of blood pressure–lowering activities. Nutr. Rev. 2015, 73 (1), 36-50. 4.

Zhu, Y.; Wang, P.; Sha, W.; Sang, S., Urinary Biomarkers of Whole Grain Wheat Intake

Identified by Non-targeted and Targeted Metabolomics Approaches. 2016, 6, 36278. 5.

O'Gorman, A.; Gibbons, H.; Brennan, L., Metabolomics in the identification of biomarkers of

dietary intake. Computational and Structural Biotechnology Journal 2013, 4, e201301004. 6.

Zamora-Ros, R.; Rabassa, M.; Llorach, R.; González, C. A.; Andres-Lacueva, C., Application of

Dietary Phenolic Biomarkers in Epidemiology: Past, Present, and Future. J. Agric. Food. Chem. 2012, 60 (27), 6648-6657. 7.

Ballard, K. D.; Mah, E.; Guo, Y.; Pei, R.; Volek, J. S.; Bruno, R. S., Low-fat milk ingestion

prevents postprandial hyperglycemia-mediated impairments in vascular endothelial function in obese individuals with metabolic syndrome. The Journal of nutrition 2013, 143 (10), 1602-1610. 8.

Alberti, K. G. M. M.; Eckel, R. H.; Grundy, S. M.; Zimmet, P. Z.; Cleeman, J. I.; Donato, K. A.;

Fruchart, J.-C.; James, W. P. T.; Loria, C. M.; Smith, S. C., Harmonizing the Metabolic Syndrome. A Joint Interim Statement of the International Diabetes Federation Task Force on Epidemiology and Prevention; National Heart, Lung, and Blood Institute; American Heart Association; World Heart Federation; International Atherosclerosis Society; and International Association for the Study of Obesity 2009, 120 (16), 1640-1645. 9.

Mah, E.; Noh, S. K.; Ballard, K. D.; Matos, M. E.; Volek, J. S.; Bruno, R. S., Postprandial

Hyperglycemia Impairs Vascular Endothelial Function in Healthy Men by Inducing Lipid Peroxidation


Page 18 of 30

Page 19 of 30 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


and Increasing Asymmetric Dimethylarginine:Arginine. The Journal of Nutrition 2011, 141 (11), 19611968. 10.

Zhu, J.; Djukovic, D.; Deng, L.; Gu, H.; Himmati, F.; Chiorean, E. G.; Raftery, D., Colorectal

Cancer Detection Using Targeted Serum Metabolic Profiling. J. Proteome Res. 2014, 13 (9), 4120-4130. 11.

Zhu, J.; Djukovic, D.; Deng, L.; Gu, H.; Himmati, F.; Abu Zaid, M.; Chiorean, E.; Raftery, D.,

Targeted serum metabolite profiling and sequential metabolite ratio analysis for colorectal cancer progression monitoring. Anal. Bioanal. Chem. 2015, 407 (26), 7857-7863. 12.

Schelli, K.; Rutowski, J.; Roubidoux, J.; Zhu, J., Staphylococcus aureus methicillin resistance

detected by HPLC-MS/MS targeted metabolic profiling. J. Chromatogr. B 2017, 1047, 124-130. 13.

Bajad, S. U.; Lu, W.; Kimball, E. H.; Yuan, J.; Peterson, C.; Rabinowitz, J. D., Separation and

quantitation of water soluble cellular metabolites by hydrophilic interaction chromatography-tandem mass spectrometry. J. Chromatogr. 2006, 1125 (1), 76-88. 14.

Yuan, M.; Breitkopf, S. B.; Yang, X.; Asara, J. M., A positive/negative ion–switching, targeted

mass spectrometry–based metabolomics platform for bodily fluids, cells, and fresh and fixed tissue. Nat. Protoc. 2012, 7 (5), 872-881. 15.

Xia, J.; Sinelnikov, I. V.; Han, B.; Wishart, D. S., MetaboAnalyst 3.0—making metabolomics

more meaningful. Nucleic Acids Res. 2015, 43 (W1), W251-W257. 16.

Smilde, A. K.; Jansen, J. J.; Hoefsloot, H. C. J.; Lamers, R.-J. A. N.; van der Greef, J.;

Timmerman, M. E., ANOVA-simultaneous component analysis (ASCA): a new tool for analyzing designed metabolomics data. Bioinformatics 2005, 21 (13), 3043-3048. 17.

Cynober, L. A., Plasma amino acid levels with a note on membrane transport: characteristics,

regulation, and metabolic significance. Nutrition 2002, 18 (9), 761-766. 18.

Ba, S.; I, C.; Jc, S.; Wpt, J., Diet, nutrition and the prevention of excess weight gain and obesity.

Public Health Nutr. 2007, 7 (1a), 123-146.



19.

Kipnis, V.; Midthune, D.; Freedman, L.; Bingham, S.; Day, N. E.; Riboli, E.; Ferrari, P.; Carroll,

R. J., Bias in dietary-report instruments and its implications for nutritional epidemiology. Public Health Nutr. 2006, 5 (6a), 915-923. 20.

Seale, J. L.; Klein, G.; Friedmann, J.; Jensen, G. L.; Mitchell, D. C.; Smiciklas-Wright, H.,

Energy expenditure measured by doubly labeled water, activity recall, and diet records in the rural elderly. Nutrition 2002, 18 (7), 568-573. 21.

Jenab, M.; Slimani, N.; Bictash, M.; Ferrari, P.; Bingham, S. A., Biomarkers in nutritional

epidemiology: applications, needs and new horizons. Hum. Genet. 2009, 125 (5), 507-525. 22.

Brosnan, M. E.; Brosnan, J. T., Orotic Acid Excretion and Arginine Metabolism. The Journal of

Nutrition 2007, 137 (6), 1656S-1661S.


Page 20 of 30

Page 21 of 30 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


Table 1. Bayesian LASSO selected metabolites that have strong correlation to FMD, postprandial plasma markers of oxidative stress and NO status. Metabolite

Insulin

Glucose

MDA

Arginine

Thiamine monophosphate

90% CI: [-1.18, -0.22] PE: -1.31 90% CI: [-1.62, -0.39] PE: -1.26 90% CI: [-1.78, -1.24] PE: 1.08 90% CI: [0.72, 1.80] PE: -0..52 90% CI: [-1.48, -0.39]

Thymine

leucine/isoleucine

Pyroglutamic acid

N-Acetylserine PE: 0.16 90% CI: [0.05, 0.27]

hydroxyproline PE: 0.27 90% CI: [0.05, 0.49]

3-Aminoisobutanoate

PE: -1.52 90% CI: [-1.75, -0.52] PE: 1.47 90% CI: [0.87, 1.85]

Glyceraldehyde

Creatine

PE: 0.23 90% CI: [0.08, 0.38]

Homoserine

Serine

Citrulline

FMD PE: -0.67

PE: -0.57 PE: -0.60 90% CI: 90% CI: [-0.77, -0.36] [-0.82, -0.39]

PE: -0.40 90% CI: [-0.64, -0.15]

PE: 0.18 90% CI: [0.07, 0.29] PE: 0.14 90% CI: [0.03, 0.25] PE: 0.13 90% CI: [0.03, 0.24]

Note: PE, point estimator of the effect; CI, Bayesian credible interval.


PE: -0.91 90% CI: [-1.28, -0.81]


Figure legends: Figure 1. A schematic workflow of this study. Figure 2. Box plots showing nine metabolites with a significant p-value (1.3) in comparison of low-fat (left and red) and rice milk groups (right and green) from plasma samples collected at 120 minutes. Figure 3. PLS-DA showing metabolic profile based separation from low-fat and rice milk consumption groups. Each panel represents metabolic profile from blood draw at (A). 0 minute; (B). 30 mins; (C). 60 mins; (D). 90 mins; (E). 120 mins; (F). 150 mins; and (G). 180 mins. Figure 4. (A) Two-way ANOVA analysis results demonstrate a different number of metabolites that have significant p-value (0.5 and -log(p) value > 2), i.e., alanine, aspartate and glutamate metabolism; glycine, serine and threonine metabolism; and arginine and proline metabolism, are labeled. The details of the other pathways displayed in this figure, their impact score and -log (p) value can be seen in Table S5.



Figure 1


Page 24 of 30

Page 25 of 30 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


Figure 2



Figure 3 26 ACS Paragon Plus Environment

Page 26 of 30

Page 27 of 30 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


Figure 4



Figure 5


Page 28 of 30

Page 29 of 30 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


Figure 6 29 ACS Paragon Plus Environment


For TOC only


Page 30 of 30

Comparative metabolomics elucidates postprandial metabolic

Recommend Documents