Metabolomics Reveals that the Type of Protein in a High-Fat Meal Modulates Postprandial Mitochondrial Overload and Incomplete Substrate Oxidation in Healthy Overweight Men

Metabolomics Reveals that the Type of Protein in a High-Fat Meal Modulates Postprandial... Abstract Background A meal rich in saturated fatty acids induces a postprandial metabolic challenge. The type of dietary protein may modulate postprandial metabolism. Objective We studied the effect of dietary protein type on postprandial changes in the metabolome after a high-fat meal. Methods In a 3-period, crossover, postprandial study, 10 healthy overweight men with an elevated waist circumference (>94 cm) ingested high-fat meals made up of cream fat (70% of energy), sucrose (15% energy), and protein (15% energy) from either casein (CAS), whey protein (WHE), or α-lactalbumin-enriched whey protein (LAC). Urine collected immediately before and 2, 4, and 6 h after the meal was analyzed for metabolomics, a secondary outcome of the clinical study. We used mixed-effect models, partial least-square regression, and pathway enrichment analysis. Results At 4 and 6 h after the meal, the postprandial metabolome was found to be fully discriminated according to protein type. We identified 17 metabolites that significantly explained the effect of protein type on postprandial metabolomic changes (protein-time interaction). Among this signature, acylcarnitines and other acylated metabolites related to fatty acid or amino acid oxidation were the main discriminant features. The difference in metabolic profiles was mainly explained by urinary acylcarnitines and some other acylated products (protein type, Ps < 0.0001), with a dramatically greater increase (100- to 1000-fold) after WHE, and to a lesser extent after LAC, as compared with CAS. Pathway enrichment analysis confirmed that the type of protein had modified fatty acid oxidation (P < 0.05). Conclusion Taken together, our results indicate that, in healthy overweight men, the type of protein in a high-fat meal interplays with fatty acid oxidation with a differential accumulation of incomplete oxidation products. A high-fat meal containing WHE, but not CAS, resulted in this outpacing of the tricarboxylic acid cycle. This study was registered at clinicaltrials.gov as NCT00931151. high-fat meal, dietary protein, postprandial, metabolomics, acylcarnitines, humans, overweight men, urine, BCAA Introduction During the postprandial period, the metabolism is challenged by the sudden influx of energy nutrients, resulting in an allostatic load (1–5). In healthy individuals, a high-saturated-fat, high-sucrose meal is considered a metabolic challenge, and has been associated with postprandial low-grade inflammatory activation, endothelial dysfunction, and impaired insulin action (6–8). This paradigm has been much studied in order to decipher the causal relations between metabolic and physiologic changes that occur during the postprandial period (9–11). These phenomena are exacerbated in overweight, obese, or cardiometabolic individuals (2, 5, 12). Unlike saturated fat and sugar, the third energy macronutrient, namely protein, does not contribute to this postprandial metabolic challenge, but little is known about how different types of dietary protein affect postprandial metabolism (13, 14). There are many ways in which dietary proteins and the resulting amino acids may affect postprandial metabolism. Firstly, amino acids may rapidly stimulate insulin secretion and action (15–17). Dietary proteins and amino acids may also affect the metabolic routing of fatty acids and sugars (18). In particular, after a meal, when amino acids and fatty acid oxidative intermediates largely fuel the tricarboxylic acid (TCA) cycle (19), the amino acid composition of the protein may affect the final postprandial changes in metabolic fluxes through its effect on the intermediary metabolism of amino acids. This could be especially true in a situation of compromised insulin action, such as in overweight individuals, in whom the expression of an altered intermediary metabolism of fats and amino acids (20) could be heightened in the postprandial period. Furthermore, fatty acid oxidation is known to downregulate BCAA oxidation at the branched-chain α-ketoacid dehydrogenase complex (20). Although this interplay between fatty acids and amino acid metabolism is not BCAA specific, BCAA metabolism in obesity and overnutrition has been the subject of much investigation in relation to the initiation of insulin resistance (20, 21). However, it remains largely unknown how dietary protein affects postprandial intermediary metabolism, particularly during the postprandial period or following short-term exposure to a high-fat diet. Metabolomic approaches are highly appropriate to study the trajectories of complex changes in intricate metabolic pathways such as those that occur during the postprandial phase (22–24). Since the plasma metabolome is tempered by homeostatic forces, whereas postprandial changes in metabolic fluxes should be integrated in the urinary metabolomic output, urinary metabolomics is expected to be more discriminating (25). Clinical studies involving repeated postprandial challenges offer opportunities to identify the effect of meal composition on postprandial metabolism using metabolomics (26, 27), by characterizing overall changes and identifying the most salient contributors to these changes (4). Postprandial metabolomic studies of the effect of protein intake are scarce. In this study, we analyzed the postprandial urinary metabolomics of healthy overweight men receiving high-fat meals differing in the type of dietary protein: either casein (CAS), whey protein (WHE), or α-lactalbumin-enriched whey protein (LAC). We chose dietary proteins that are of practical interest and have contrasting nutritional characteristics, including differences in their amino acid (e.g., leucine) content. Our aim was to understand the degree to which each type of protein affects postprandial changes in metabolomics that are driven by a high fat load and to try and identify the set of metabolites that are the most important contributors. Methods Subjects and study design Ten healthy overweight men who had participated in a postprandial study designed to examine the acute effects of different dietary proteins in high-fat meals on postprandial TGs and some physiological markers were selected for our study on urinary metabolomics. Details of this randomized, 3-phase crossover study were reported previously (14). Urinary metabolomics was a secondary outcome of this clinical study. All the subjects were healthy men (age range 21–50 y) who were overweight [BMI (kg/m2) >25] and had an enlarged waist circumference (>94 cm). Exclusion criteria included any established disease or regular use of medication, hypertension, excessive alcohol intake (>2 units of alcohol per day), moderate or high use of tobacco products (>9 cigarettes or equivalent per day), use of any nutritional supplements, moderate or vigorous physical activity (taken as >4 h of moderate/vigorous physical activity per week), and blood hemoglobin <13 g/dL. The characteristics of these subjects were as follows (mean ± SD): age, 34 ± 9 y; height, 178 ± 3 cm; weight, 96 ± 6 kg; BMI, 30.2 ± 1.5; body fat, 24.3% ± 2.0%; waist circumference, 96.3 ± 3.2 cm. All participants gave their written informed consent prior to enrollment. The study was approved by the Institutional Review Board for Saint-Germain-en-Laye Hospital, authorized by the French Ministry of Health, and registered at clinicaltrials.gov as NCT00931151. All subjects completed 3 treatment sessions consisting of a postprandial study, which were separated by ≥2 wk. The subjects were required to consume 1 of the 3 test meals during each treatment session in a randomized order, according to a Latin square. The participants were asked to empty their bladders of night/early-morning urine. Urine was then collected immediately before the meal and every 2 h thereafter (0, 2, 4, and 6 h after the meal). Blood was sampled before the meal and 0.5, 1, 1.5, 2, 3, 4, and 6 h after the meal. Test meals The composition of the meals was as follows: energy, 1200 kcal; fat, 93.3 g (i.e., 70% energy); carbohydrates, 45 g (i.e., 15% energy); crude protein, 45 g (i.e., 15% energy). The test meals consisted of a mixture of 233 g 40%-fat cream, 45 g sucrose, 160 mL water, and protein isolates of either CAS, WHE, or LAC, in quantities (54, 55, and 49 g, respectively) adjusted to yield the same amount of protein (45 g). The cream and sugar were standard commercial products and the protein isolates were supplied by Ingredia (CAS and WHE) and Armor Protein (LAC). Because the amounts of remaining carbohydrates and minerals were not the same in these protein-rich powders, the meals were equilibrated with the addition of lactose, calcium phosphate, potassium phosphate, and magnesium phosphate. TGs and apoB-48 in plasma TGs were assayed with the use of an enzymatic colorimetric method with commercial kits (RANDOX Laboratories), and apoB-48 were assayed with the use of an ELISA assay (Biovendor), as previously described (14) Metabolomic analysis of urine samples The urine samples were defrosted at room temperature, centrifuged at 7000 × g for 5 min at 4°C, and then diluted 4-fold with distilled water. Urinary metabolic profiles were determined with the use of an untargeted metabolomic approach that could cover a broad range of metabolites. Analyses were performed following a procedure described in Morio et al. (25), based on a Waters Acquity UPLC chromatographic system (Waters Corporation) coupled to a Waters Qtof-Micro equipped with an electrospray source and a lockmass sprayer to ensure accuracy. MS data were collected in continuum full-scan mode with a mass-to-charge ratio (m/z) of 70–1000 from 0 to 22 min, in positive mode. To prevent possible differences between the sample batches, a Latin square method was performed to obtain a randomized list of samples for analysis. For analysis of each sample, 6 µL of diluted urine was injected into a 100 × 2.1 mm, 1.7 µm BEH Shield RP18 column at 30°C. The mobile phase components were 1% formic acid (A) and acetonitrile with 1% formic acid (B). The flow rate was set at 400 µL/min. The raw data were transformed to centroid mode and mass corrected before being analyzed with the use of the XCMS platform (28, 29). The LC-MS data were peak detected and noise reduced for both the LC and MS components. Each peak in the resulting 3-dimensional data set was represented by a retention time m/z and its ion intensity in each sample. The matrix obtained was then exported for statistical analysis. Since this study was closely controlled for fluid intake during the experimental sessions, the urinary volume did not vary between conditions [the mean urinary volumes after the CAS, WHE, and LAC meals were 206 ± 24, 188 ± 24 and 254 ± 30 mL (mean ± SE), respectively]. The metabolites were expressed as raw (log-transformed) relative intensities and not quantified as concentrations. Accordingly, we did not normalize by urinary volume or creatinine concentration, nor did we calculate total urinary output. Metabolites contributing to the discrimination of the different phenotypes were first identified with the use of an in-house database containing the reference spectra of >1000 authentic standard compounds. The remaining unknown compounds were then identified on the basis of their exact masses which were compared with those registered in the Human Metabolome Database (30). Metabolites were classified according to the method described by Sumner et al. (31), which is based on levels of confidence in the identification process, as follows: identified (confirmed by a standard), putatively annotated (based on physicochemical properties or spectral similarity with public/commercial spectral libraries), and unknown compounds. Statistical analyses A statistical workflow based on univariate and multivariate analyses was used for the metabolomic data. All ion intensities were log transformed before analysis. Mixed-effect models were built with the lme4 R package v. 3.1.1 (32). Time, protein, time-protein interaction, and batch were included in the models. A Benjamini Hochberg P value adjustment (33) was used for multiple testing, and when a fixed effect was significant, post-hoc comparisons with Bonferroni corrections were also performed. Significance was set at a corrected P value <0.05. In order to reveal the specificity of features with a significant protein or interaction effect, Venn diagrams were generated from the internet application JVENN (34). Multivariate analyses by partial least square (PLS) were performed to rank the combinatorial predictive ability of the candidate biomarkers for proteins, time, and interactions. SIMCA-P+ v. 13 (Umetrics) was used. The analysis was performed first on all the variables, then on variables indicating a significant protein effect at a given time point after the meal, and finally on variables with a significant interaction effect at a given time point after the meal. The overall quality of the models was assessed with the use of the cumulative R2 (R2Ycum) and cumulative Q2 (Q2cum) criteria. Moreover, in order to verify that PLS components could not lead to classifications by chance, a permutation test based on 100 random permutations was performed. Variable importance on projection (VIP) values were used to assess the importance of each ion in the PLS discriminant analysis (PLS-DA) model. R software was used to perform Spearman correlations (and tests of significance) with and without the meal protein in the model, between ions and plasma TGs, and between ions and plasma apoB-48. Plasma concentrations were log transformed before analysis. Metabolite pathway enrichment analysis Through the use of MetExplore (35, 36), the metabolites identified as significantly explaining the effect of protein type on postprandial changes (protein-time interaction) were mapped on a genome-scale model of human metabolism [RECON 2.04, (37)], in order to identify the corresponding metabolic pathways/subnetworks that were significantly altered by pathway enrichment. Multiple comparisons across metabolic subnetworks were determined from Bonferroni corrections. Results Metabolomics reveals differences in postprandial kinetics Under positive ionization, 5225 ions were extracted from the raw data. Univariate statistical methods based on the use of a mixed-effect model revealed 1150 ions that were significantly modulated according to ≥1 of the following 3 criteria: 139 ions were dependent on the protein type (i.e., CAS, WHE, LAC), 1078 on time (i.e., 0, 2, 4, 6 h after the meal), and 36 on the time-protein interaction. Post-hoc comparisons indicated that ∼85% of variables with a protein effect differed significantly between CAS and WHE and between LAC and WHE, and only 47.5% differed significantly different between CAS and LAC. On Venn diagrams (Figure 1), around half of the variables with a protein effect were specific to WHE, and 27 ions were discriminant for the 3 meals. In the same way, 17 ions with an interaction effect were specific to WHE, and 19 ions were found to differ significantly between the 3 proteins (Figure 1). Post-hoc comparisons of interactions revealed that 2, 4, or 6 h after the meal, 75–100% of variables with an interaction differed significantly between CAS and WHE and between LAC and WHE, and <50% differed significantly between CAS and LAC. FIGURE 1 View largeDownload slide Venn diagrams showing the numbers of ions significantly modulated as a function of protein type (A, 139 ions) and the protein-time interaction (B, 36 ions), together with respective histograms representing the numbers of significant ions for each dietary protein type (C, D) in 10 healthy overweight men after a high-fat meal that included CAS, WHE, or LAC. CAS, casein; LAC, α-lactalbumin-enriched whey protein; WHE, whey protein. FIGURE 1 View largeDownload slide Venn diagrams showing the numbers of ions significantly modulated as a function of protein type (A, 139 ions) and the protein-time interaction (B, 36 ions), together with respective histograms representing the numbers of significant ions for each dietary protein type (C, D) in 10 healthy overweight men after a high-fat meal that included CAS, WHE, or LAC. CAS, casein; LAC, α-lactalbumin-enriched whey protein; WHE, whey protein. PLS-DA analyses performed on all extracted variables (Supplemental Figure 1) enabled the prediction of protein type with the use of a model including 5 components (R2Ycum = 0.863 and Q2cum = 0.662). A total of 1855 ions had a VIP >1. In particular, 110 ions among the 139 ions with a significant global protein effect were within the first quarter of VIP, and 35 ions among the 36 features with a significant global interaction effect were within the first 6% of VIP. PLS-DA models were then built at each time point (i.e., 0, 2, 4, or 6 h after the meal), introducing only the 139 variables with a significant protein effect to rank the combinatorial predictive ability of these candidate biomarkers. No valid model was found before the meal and 2 h after the meal. By contrast, at 4 and 6 h after the meal, the models with 2 components were specific and predictive with R2Ycum values of 0.867 and 0.875, respectively, and Q2cum values of 0.757 and 0.739, respectively (Figure 2). Twenty-six of the 36 ions with a significant global interaction effect were always within the first quarter of VIP at 4 or 6 h after the meals. In terms of these features, the urine metabolome was much more markedly modified after WHE than after the other proteins (Figure 3). FIGURE 2 View largeDownload slide Discrimination of protein type in the meal based on the scores plot of PLS-DA performed on the 139 variables with a significant protein effect 4 h after the meal (A) and 6 h after the meal (B) and 100-permutation tests of PLS-DA models 4 h after the meal (C) and 6 h after the meal (D), in 10 healthy overweight men after a high-fat meal that included CAS, WHE, or LAC. The permutation plots (C and D) show the Pearson correlation coefficient between the original y variable and the permuted y variable (x axis) versus the cumulative R2 and Q2 (y axis), and the regression line. CAS, casein; LAC, α-lactalbumin-enriched whey protein; PLS-DA, partial least-squares discriminant analysis; t[1], partial least-squares component 1; t[2], partial least-squares component 2; WHE, whey protein. FIGURE 2 View largeDownload slide Discrimination of protein type in the meal based on the scores plot of PLS-DA performed on the 139 variables with a significant protein effect 4 h after the meal (A) and 6 h after the meal (B) and 100-permutation tests of PLS-DA models 4 h after the meal (C) and 6 h after the meal (D), in 10 healthy overweight men after a high-fat meal that included CAS, WHE, or LAC. The permutation plots (C and D) show the Pearson correlation coefficient between the original y variable and the permuted y variable (x axis) versus the cumulative R2 and Q2 (y axis), and the regression line. CAS, casein; LAC, α-lactalbumin-enriched whey protein; PLS-DA, partial least-squares discriminant analysis; t[1], partial least-squares component 1; t[2], partial least-squares component 2; WHE, whey protein. FIGURE 3 View largeDownload slide Box plots representing postprandial changes to mean ion intensities (after unit-variance scaling) for 25 of the 36 significant features of the protein-time interaction [after the removal of analytical redundancy (11 isotopes, adducts and fragments)], in 10 healthy overweight men after a high-fat meal that included either CAS (A), LAC (B), or WHE (C). Boxes show median and 25th/75th percentiles, whiskers show 5th/95th percentiles, and open circles show outliers. AU, arbitrary units; CAS, casein; LAC, α-lactalbumin-enriched whey protein; WHE, whey protein. FIGURE 3 View largeDownload slide Box plots representing postprandial changes to mean ion intensities (after unit-variance scaling) for 25 of the 36 significant features of the protein-time interaction [after the removal of analytical redundancy (11 isotopes, adducts and fragments)], in 10 healthy overweight men after a high-fat meal that included either CAS (A), LAC (B), or WHE (C). Boxes show median and 25th/75th percentiles, whiskers show 5th/95th percentiles, and open circles show outliers. AU, arbitrary units; CAS, casein; LAC, α-lactalbumin-enriched whey protein; WHE, whey protein. Out of the 36 ions that were significant for the protein-time interaction, 14 were found to be parent ions (the others being isotopes, adducts, and in-source fragments) and the corresponding metabolites were identified from databases combined with high-resolution MS (Table 1). Four metabolites remained unknown, but their polarity, masses, and in-source fragmentation were suggestive of modified peptides (especially acetylated species). Interestingly, in the urine metabolomes, acylcarnitines were the metabolites most discriminant of protein type after ingestion. Amino acids and derivatives, as well as peptides, were also found to be important to discrimination. Regarding the individual kinetics of these principal discriminant metabolites (Figure 4), the greatest amplitudes in variation were found after the consumption of WHE, with postprandial increase ≤3 orders of magnitude higher than with CAS. FIGURE 4 View largeDownload slide Postprandial kinetics of ion intensities of the principal confirmed identity metabolites (A–I) for the protein-time interaction, in healthy overweight men after a high-fat meal that included CAS, WHE, or LAC. Data are means ± SDs, n = 10. Protein type and the protein type × time interaction were significant for all variables (all Ps < 0.05) according to mixed models. AU, arbitrary units; CAS, casein; GE, Glycyl-glutamate; LAC, α-lactalbumin-enriched whey protein; VI, Valyl-isoleucine; WHE, whey protein. FIGURE 4 View largeDownload slide Postprandial kinetics of ion intensities of the principal confirmed identity metabolites (A–I) for the protein-time interaction, in healthy overweight men after a high-fat meal that included CAS, WHE, or LAC. Data are means ± SDs, n = 10. Protein type and the protein type × time interaction were significant for all variables (all Ps < 0.05) according to mixed models. AU, arbitrary units; CAS, casein; GE, Glycyl-glutamate; LAC, α-lactalbumin-enriched whey protein; VI, Valyl-isoleucine; WHE, whey protein. TABLE 1 Identification of metabolites that are significant regarding the protein-time interaction in 10 healthy overweight men after a high-fat meal that included casein, whey protein, or α-lactalbumin-enriched whey protein1 PLS-DA on protein (139 ions) Statistics2 PLS-DA on protein (all ions) 4 h after meal 6 h after meal Identification Experimental mass-to-charge ratio, m:z Retention time, min Protein type onVenn diagram TimeP value Protein typeP value Protein × timeP value VIP value VIP rank VIP value VIP rank VIP value VIP rank N-Acetyl-l-aspartic acid3 176.103 1.7 CAS, LAC, and WHE 2.8 × 10−11 3.5 × 10−08 2.7 × 10−04 1.4 313 0.7 77 0.8 67 Tryptophan3 188.164 4.9 CAS, LAC, and WHE 5.8 × 10−12 1.9 × 10−07 1.1 × 10−02 1.9 83 1.2 31 1.4 23 Indolebutyric acid3 204.159 6.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.8 8 1.9 6 1.9 7 Propionylcarnitine3 218.174 8.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.5 26 1.8 8 1.6 13 Capryloylglycine 202.177 8.7 CAS, LAC, and WHE 1.2 × 10−06 4.9 × 10−06 2.6 × 10−02 1.8 125 1.0 56 1.1 43 R-Butyryl carnitine3 232.191 9.3 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.1 2 2.3 2 2.2 2 Tiglylcarnitine3 228.179 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 1.1 × 10−04 1.6 188 1.0 49 1.0 48 Valyl-isoleucine3 231.168 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 4.5 × 10−03 1.6 172 1.0 50 1.0 51 Unknown 354.012 10.2 WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.0 3 2.2 5 1.9 6 Hexanoylcarnitine3 260.221 10.9 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.3 1 2.3 1 2.3 1 Unknown 339.164 13.1 WHE 1.3 × 10−2 <2.2 × 10−16 1.6 × 10−4 2.2 43 1.3 28 1.4 27 Unknown 441.087 13.6 WHE 3.1 × 10−3 1.5 × 10−8 4.6 × 10−4 1.5 220 1.4 23 1.3 33 Glycyl-glutamate3 409.068 14.9 WHE NS 2.6 × 10−11 5.8 × 10−4 2.3 35 1.3 25 1.4 26 Unknown 455.094 14.9 WHE NS 9.9 × 10−12 1.4 × 10−5 2.0 71 1.4 20 1.6 17 PLS-DA on protein (139 ions) Statistics2 PLS-DA on protein (all ions) 4 h after meal 6 h after meal Identification Experimental mass-to-charge ratio, m:z Retention time, min Protein type onVenn diagram TimeP value Protein typeP value Protein × timeP value VIP value VIP rank VIP value VIP rank VIP value VIP rank N-Acetyl-l-aspartic acid3 176.103 1.7 CAS, LAC, and WHE 2.8 × 10−11 3.5 × 10−08 2.7 × 10−04 1.4 313 0.7 77 0.8 67 Tryptophan3 188.164 4.9 CAS, LAC, and WHE 5.8 × 10−12 1.9 × 10−07 1.1 × 10−02 1.9 83 1.2 31 1.4 23 Indolebutyric acid3 204.159 6.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.8 8 1.9 6 1.9 7 Propionylcarnitine3 218.174 8.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.5 26 1.8 8 1.6 13 Capryloylglycine 202.177 8.7 CAS, LAC, and WHE 1.2 × 10−06 4.9 × 10−06 2.6 × 10−02 1.8 125 1.0 56 1.1 43 R-Butyryl carnitine3 232.191 9.3 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.1 2 2.3 2 2.2 2 Tiglylcarnitine3 228.179 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 1.1 × 10−04 1.6 188 1.0 49 1.0 48 Valyl-isoleucine3 231.168 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 4.5 × 10−03 1.6 172 1.0 50 1.0 51 Unknown 354.012 10.2 WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.0 3 2.2 5 1.9 6 Hexanoylcarnitine3 260.221 10.9 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.3 1 2.3 1 2.3 1 Unknown 339.164 13.1 WHE 1.3 × 10−2 <2.2 × 10−16 1.6 × 10−4 2.2 43 1.3 28 1.4 27 Unknown 441.087 13.6 WHE 3.1 × 10−3 1.5 × 10−8 4.6 × 10−4 1.5 220 1.4 23 1.3 33 Glycyl-glutamate3 409.068 14.9 WHE NS 2.6 × 10−11 5.8 × 10−4 2.3 35 1.3 25 1.4 26 Unknown 455.094 14.9 WHE NS 9.9 × 10−12 1.4 × 10−5 2.0 71 1.4 20 1.6 17 1VIP obtained from PLS-DA analyses. CAS, casein; LAC, α-lactalbumin-enriched whey protein; PLS-DA, partial-least square discriminant analysis; VIP, variable importance on projection value; WHE, whey protein. 2Results from mixed-effect models with Benjamini Hochberg adjustments. NS, P > 0.05. 3Compounds identified that were confirmed by standards. View Large TABLE 1 Identification of metabolites that are significant regarding the protein-time interaction in 10 healthy overweight men after a high-fat meal that included casein, whey protein, or α-lactalbumin-enriched whey protein1 PLS-DA on protein (139 ions) Statistics2 PLS-DA on protein (all ions) 4 h after meal 6 h after meal Identification Experimental mass-to-charge ratio, m:z Retention time, min Protein type onVenn diagram TimeP value Protein typeP value Protein × timeP value VIP value VIP rank VIP value VIP rank VIP value VIP rank N-Acetyl-l-aspartic acid3 176.103 1.7 CAS, LAC, and WHE 2.8 × 10−11 3.5 × 10−08 2.7 × 10−04 1.4 313 0.7 77 0.8 67 Tryptophan3 188.164 4.9 CAS, LAC, and WHE 5.8 × 10−12 1.9 × 10−07 1.1 × 10−02 1.9 83 1.2 31 1.4 23 Indolebutyric acid3 204.159 6.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.8 8 1.9 6 1.9 7 Propionylcarnitine3 218.174 8.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.5 26 1.8 8 1.6 13 Capryloylglycine 202.177 8.7 CAS, LAC, and WHE 1.2 × 10−06 4.9 × 10−06 2.6 × 10−02 1.8 125 1.0 56 1.1 43 R-Butyryl carnitine3 232.191 9.3 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.1 2 2.3 2 2.2 2 Tiglylcarnitine3 228.179 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 1.1 × 10−04 1.6 188 1.0 49 1.0 48 Valyl-isoleucine3 231.168 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 4.5 × 10−03 1.6 172 1.0 50 1.0 51 Unknown 354.012 10.2 WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.0 3 2.2 5 1.9 6 Hexanoylcarnitine3 260.221 10.9 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.3 1 2.3 1 2.3 1 Unknown 339.164 13.1 WHE 1.3 × 10−2 <2.2 × 10−16 1.6 × 10−4 2.2 43 1.3 28 1.4 27 Unknown 441.087 13.6 WHE 3.1 × 10−3 1.5 × 10−8 4.6 × 10−4 1.5 220 1.4 23 1.3 33 Glycyl-glutamate3 409.068 14.9 WHE NS 2.6 × 10−11 5.8 × 10−4 2.3 35 1.3 25 1.4 26 Unknown 455.094 14.9 WHE NS 9.9 × 10−12 1.4 × 10−5 2.0 71 1.4 20 1.6 17 PLS-DA on protein (139 ions) Statistics2 PLS-DA on protein (all ions) 4 h after meal 6 h after meal Identification Experimental mass-to-charge ratio, m:z Retention time, min Protein type onVenn diagram TimeP value Protein typeP value Protein × timeP value VIP value VIP rank VIP value VIP rank VIP value VIP rank N-Acetyl-l-aspartic acid3 176.103 1.7 CAS, LAC, and WHE 2.8 × 10−11 3.5 × 10−08 2.7 × 10−04 1.4 313 0.7 77 0.8 67 Tryptophan3 188.164 4.9 CAS, LAC, and WHE 5.8 × 10−12 1.9 × 10−07 1.1 × 10−02 1.9 83 1.2 31 1.4 23 Indolebutyric acid3 204.159 6.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.8 8 1.9 6 1.9 7 Propionylcarnitine3 218.174 8.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.5 26 1.8 8 1.6 13 Capryloylglycine 202.177 8.7 CAS, LAC, and WHE 1.2 × 10−06 4.9 × 10−06 2.6 × 10−02 1.8 125 1.0 56 1.1 43 R-Butyryl carnitine3 232.191 9.3 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.1 2 2.3 2 2.2 2 Tiglylcarnitine3 228.179 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 1.1 × 10−04 1.6 188 1.0 49 1.0 48 Valyl-isoleucine3 231.168 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 4.5 × 10−03 1.6 172 1.0 50 1.0 51 Unknown 354.012 10.2 WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.0 3 2.2 5 1.9 6 Hexanoylcarnitine3 260.221 10.9 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.3 1 2.3 1 2.3 1 Unknown 339.164 13.1 WHE 1.3 × 10−2 <2.2 × 10−16 1.6 × 10−4 2.2 43 1.3 28 1.4 27 Unknown 441.087 13.6 WHE 3.1 × 10−3 1.5 × 10−8 4.6 × 10−4 1.5 220 1.4 23 1.3 33 Glycyl-glutamate3 409.068 14.9 WHE NS 2.6 × 10−11 5.8 × 10−4 2.3 35 1.3 25 1.4 26 Unknown 455.094 14.9 WHE NS 9.9 × 10−12 1.4 × 10−5 2.0 71 1.4 20 1.6 17 1VIP obtained from PLS-DA analyses. CAS, casein; LAC, α-lactalbumin-enriched whey protein; PLS-DA, partial-least square discriminant analysis; VIP, variable importance on projection value; WHE, whey protein. 2Results from mixed-effect models with Benjamini Hochberg adjustments. NS, P > 0.05. 3Compounds identified that were confirmed by standards. View Large Integration of biochemical and metabolomic data Overall, all protein types taken together, 9 of the 14 metabolites with significant protein-time interactions were correlated with TG and apoB-48, with a mean Spearman correlation coefficient of ∼0.3. More precisely, and interestingly, the strength of the correlation was indeed dependent on the protein type. The highest correlation coefficient estimates were obtained after LAC, with a positive Spearman coefficient between TG and acylcarnitines ranging from 0.32 (for propionylcarnitine) to 0.55 (for hexanoylcartinine) after LAC, and from 0.23 (for propionylcarnitine) to 0.38 (for capryloylglycine) after WHE, whereas no such significant correlations were found after the CAS meal (data not shown). Pathway analysis Based on metabolite pathway enrichment analysis, we found that the fatty acid oxidation pathway was significantly affected by the type of protein source (P < 0.05). There was also a trend towards a change in the alanine and aspartate metabolism pathway (raw P < 0.01), but this was not significant after correction for overall type I error. Discussion The effect of protein type on the postprandial metabolome Our first finding was that the kinetics of the postprandial metabolome could be very clearly discriminated according to protein type even though protein only contributed 15% to the energy in the meal. This discrimination was important and pertained to the type of the protein in general, since most of the metabolites that were modulated significantly by proteins were not specific to a particular type but rather common to the 3 types taken together. With the use of metabolomics, it has recently been reported that individuals receiving long-term protein supplementation could be strongly discriminated based on the type of protein they had received, indicating that supplementation with proteins that differ in composition affects broader metabolic systems (38). Our study reveals that this occurs as early as the postprandial period. Acylcarnitines and other acylated metabolites are found to be dramatically affected by protein type The second and most important finding was that acylcarnitines and other acylated metabolites were the most important discriminant molecules in terms of the effect of protein type, and were found at dramatically differing levels in the urine as a function of protein type. New light has recently been thrown on the role of acylcartinines in the pathophysiology of metabolic diseases (39–42). Acylcarnitines are carnitine esters derived from fatty acids or amino acids that are transferred into the mitochondria. Acylcarnitine synthesis may increase when β-oxidation rates are in excess of complete oxidation to CO2 via the TCA cycle (43). Acylcartinines are found to be elevated in the plasma and skeletal muscle of individuals with insulin resistance, and they have recently been reported as being a component of the metabolomic signature in obesity (19, 39, 44). Interestingly, acylcarnitines are most often part of a metabolic profile that also features plasma amino acids such as BCAAs (43, 45–48). In the obese, higher plasma acylcarnitines have been ascribed to a higher fatty acid influx from the diet and an incomplete fatty acid oxidation in the muscle resulting from the outpacing of the TCA cycle (40). Similar results have been reported in prediabetes and diabetes (39, 49). However, fewer studies have focused on acute changes such as those that occur after a high-fat meal, and more generally on postprandial metabolomics (23, 24). Acylcarnitine modifications form part of the changes observed in the metabolome after a carbohydrate-rich meal, with a reduction in postprandial acylcarnitines in both lean and obese subjects that can be explained by a switch from β oxidation to glycolysis (22, 50, 51). By contrast, plasma acylcarnitines increase after a high-fat diet in obese individuals (and decrease in lean individuals), as was shown by Baker, Boyle and colleagues (19, 52). These authors reported an increase in medium-chain acylcarnitines in the muscle in the fed state after 5 d of high-fat feeding, indicative of a limitation in the fatty acid flux in β oxidation, and found an increase in short-chain acylcarnitines, suggesting increased anaplerosis for TCA cycle intermediates (19). They did not, however, study the changes occurring in the postprandial period itself (19). During our study, the acylcarnitines related to the protein type were of both the short-chain (i.e., C3 and C4 carnitine) and medium-chain (i.e., C6) type. Medium-chain carnitines are intermediates produced by β oxidation, whereas short-chain acylcarnitines are distal products of the degradation of amino acids (40, 53). Other acylated metabolites such as capryloylglycine are usually minor compounds formed from acyl-CoA, primarily originating from β oxidation. Acetyl-aspartic acid is also a product of acylation from acetyl-CoA. Tiglylcarnitine is a product of isoleucine catabolism. Taken together, we interpret this signature as indicating that the type of dietary protein modified the efficiency of distal β oxidation and, most importantly, regulated the overall outpacing of the TCA cycle that occurs during the postprandial period. After the meal containing WHE, when compared to that containing CAS, we found evidence of a dramatic increase in acylcarnitines and other acylated metabolites in the urine, starting 2 h after the meal and culminating ∼4 h after ingestion. It is important to note that the urinary concentrations of these metabolites were 100- to 1000-fold higher after WHE than after CAS. Postprandial changes to these metabolites after LAC were intermediate between the response to CAS and WHE. Many acylcarnitines and acylated metabolites were significantly correlated to TGs and apoB-48 after the WHE and LAC meals, but not after the CAS meal. Possible mechanisms involved in the effect of meal amino acids on intermediary metabolism The reasons for this marked differential effect of protein type remain speculative. First, it would be tempting to interpret some of our results as simply being due to a slower influx of fatty acids after the CAS meal than after the other meals, as previously shown (14). However, even if a threshold effect were postulated, the ∼20% lower TG AUC after CAS as compared with the other proteins (14) cannot explain differences in responses of 2–3 orders of magnitude. Furthermore, the differences in responses appeared quite early in the urinary metabolome (starting 2h after the meal), whereas plasma TGs remained at very similar levels with all the meals at 3 h postprandially (Supplemental Figure 2). Likewise, there were no differences in plasma nonesterified fatty acid concentrations between meals (Supplemental Figure 2). In contrast to TGs, plasma amino acids had already peaked at 1 h after meal (14), with similar kinetics between meals, confirming that potential mechanisms could rather involve the amino acid composition of the protein meals. Firstly, the oxidation of amino acids that occurred in addition to the oxidation of meal fatty acids may have directly contributed to an overflow of the TCA cycle during the postprandial phase by fueling the acetyl-CoA pool (e.g., leucine, isoleucine, tryptophan). The oxidation of amino acids also provides intermediates that share the fatty acid oxidation pathways (e.g., valine, isoleucine, lysine, tryptophan, leucine, phenylalanine, and tyrosine at the level of the enoyl-CoA hydratase or the aceto-acetyl-CoA transferase). However, and especially as compared with the high amount of fatty acids (90 g), the meals did not contain very different amounts of these amino acids (Table 2). This may explain why Hoefle et al. (54) reported similar postprandial increases in plasma short-chain acylcarnitines when adding 50 g WHE and 50 g CAS to a carbohydrate load. Lastly, tiglylcarnitine, which is specific to the metabolism of isoleucine, was differentially affected by the protein meals in our study, whereas all meals contained the same amount of isoleucine. This tends to rule out that a higher intake of certain amino acids (with WHE) was contributing as substrates to the overflow of the TCA cycle. TABLE 2 Amounts of amino acids in the high-fat meals that included CAS, WHE, or LAC1 CAS, g/meal2 WHE, g/meal2 LAC, g/meal2 Leucine 4.32 5.43 4.33 Isoleucine 2.30 2.25 2.29 Valine 2.79 2.43 2.37 Lysine 3.56 4.45 4.08 Phenylalanine 2.07 1.57 1.80 Tyrosine 2.16 1.66 1.75 Tryptophan 0.59 1.21 1.18 Threonine 1.85 2.16 2.33 Methionine 1.13 1.35 0.69 Cystine 0.32 1.71 2.32 Arginine 1.49 1.08 1.31 Histidine 1.17 0.85 1.14 Alanine 1.40 2.07 1.35 Aspartic acid 3.29 5.03 5.59 Glutamic acid 9.45 7.28 6.86 Glycine 0.81 0.81 1.18 Proline 4.19 1.89 2.08 Serine 2.16 1.80 2.37 CAS, g/meal2 WHE, g/meal2 LAC, g/meal2 Leucine 4.32 5.43 4.33 Isoleucine 2.30 2.25 2.29 Valine 2.79 2.43 2.37 Lysine 3.56 4.45 4.08 Phenylalanine 2.07 1.57 1.80 Tyrosine 2.16 1.66 1.75 Tryptophan 0.59 1.21 1.18 Threonine 1.85 2.16 2.33 Methionine 1.13 1.35 0.69 Cystine 0.32 1.71 2.32 Arginine 1.49 1.08 1.31 Histidine 1.17 0.85 1.14 Alanine 1.40 2.07 1.35 Aspartic acid 3.29 5.03 5.59 Glutamic acid 9.45 7.28 6.86 Glycine 0.81 0.81 1.18 Proline 4.19 1.89 2.08 Serine 2.16 1.80 2.37 1CAS, casein; LAC, α-lactalbumin-enriched whey protein; WHE, whey protein. 2Amounts calculated from the percentage of each individual amino acid to the sum of the amino acids measured multiplied by the crude protein content (N × 6.25) in the meals, i.e., 45 g protein. View Large TABLE 2 Amounts of amino acids in the high-fat meals that included CAS, WHE, or LAC1 CAS, g/meal2 WHE, g/meal2 LAC, g/meal2 Leucine 4.32 5.43 4.33 Isoleucine 2.30 2.25 2.29 Valine 2.79 2.43 2.37 Lysine 3.56 4.45 4.08 Phenylalanine 2.07 1.57 1.80 Tyrosine 2.16 1.66 1.75 Tryptophan 0.59 1.21 1.18 Threonine 1.85 2.16 2.33 Methionine 1.13 1.35 0.69 Cystine 0.32 1.71 2.32 Arginine 1.49 1.08 1.31 Histidine 1.17 0.85 1.14 Alanine 1.40 2.07 1.35 Aspartic acid 3.29 5.03 5.59 Glutamic acid 9.45 7.28 6.86 Glycine 0.81 0.81 1.18 Proline 4.19 1.89 2.08 Serine 2.16 1.80 2.37 CAS, g/meal2 WHE, g/meal2 LAC, g/meal2 Leucine 4.32 5.43 4.33 Isoleucine 2.30 2.25 2.29 Valine 2.79 2.43 2.37 Lysine 3.56 4.45 4.08 Phenylalanine 2.07 1.57 1.80 Tyrosine 2.16 1.66 1.75 Tryptophan 0.59 1.21 1.18 Threonine 1.85 2.16 2.33 Methionine 1.13 1.35 0.69 Cystine 0.32 1.71 2.32 Arginine 1.49 1.08 1.31 Histidine 1.17 0.85 1.14 Alanine 1.40 2.07 1.35 Aspartic acid 3.29 5.03 5.59 Glutamic acid 9.45 7.28 6.86 Glycine 0.81 0.81 1.18 Proline 4.19 1.89 2.08 Serine 2.16 1.80 2.37 1CAS, casein; LAC, α-lactalbumin-enriched whey protein; WHE, whey protein. 2Amounts calculated from the percentage of each individual amino acid to the sum of the amino acids measured multiplied by the crude protein content (N × 6.25) in the meals, i.e., 45 g protein. View Large Alternatively, it could be hypothesized that WHE (as opposed to CAS) may have increased β oxidation in such a way that this resulted in exceeding the capacity of the TCA cycle to handle oxidative products from both fatty acids and amino acids. If the possible role of meal amino acids as substrates were ruled out, it is necessary to consider that they may have mediated the results by means of a signaling effect. Indeed, this hypothesis seems the most likely because, on the one hand, amino acids have long been studied as modulators of insulin secretion and action in the postprandial setting (15) and, on the other hand, the metabolic switching and adaptability of muscle for fat oxidation is largely dependent on insulin sensitivity (55, 56). It is nevertheless difficult to identify candidate amino acids that could readily explain the orientation of the effects evidenced here, and the underlying mechanisms remain speculative. In conclusion, our study has revealed that protein type regulates the oxidative pathways of fatty acids and amino acids after a high-fat meal. An increase in acylcarnitines and other acylated products was very marked with WHE, but blunted with CAS. Our study lends credence to the importance of the interplay between amino acids and fatty acid metabolism as proposed in the context of chronic high-fat feeding (21, 43, 57). These findings may also have a pathophysiologic relevance, since it has been proposed that β oxidation of fatty acids that exceeds the capacity of the TCA cycle yields incomplete fat oxidation and mitochondrial distress, which are essential events in the pathogenesis of insulin resistance (40, 58, 59). This warrants further studies in order to determine how dietary protein may modify intermediary metabolism and the initiation of insulin resistance and inflammatory pathways as early as during the postprandial period. Acknowledgments The authors’ responsibilities were as follows—FM, EP-G, and J-FH: designed the research; EP-G, MB-B, CJ, J-FM, J-FH, and FM: conducted the research; EP-G, MB-B, and HF: analyzed the data; EP-G, MB-B, HF, J-FH, DD, and FM: interpreted the results; EP-G, MB-B, and FM: drafted the manuscript; EP-G and FM: had primary responsibility for the final content; and all authors: read and approved the final manuscript. Notes Supported in part by a grant from the National Interprofessional Centre of the Dairy Industry and the French Agency for Research and Technology (National Program for Research in Food and Human Nutrition 2006, SURPROL project). Author disclosures: EP-G, MB-B, HF, CJ, JFM, JFH, DD, and FM, no conflicts of interest. Supplemental Figures 1 and 2 are available from the “Supplementary data” link in the online posting of the article and from the same link in the online table of contents at https://academic.oup.com/jn/. Abbreviations used: CAS, casein; LAC, α-lactalbumin-enriched whey protein; PLS, partial least square; PLS-DA, PLS discriminant analysis; TCA, tricarboxylic acid; VIP, variable importance on projection; WHE, whey protein. References 1. Olza J , Calder PC . Metabolic and inflammatory responses to different caloric loads of a high-fat meal are distinct between normal-weight and obese individuals . J Nutr 2014 ; 144 : 1493 – 4 . Google Scholar CrossRef Search ADS PubMed 2. Schwander F , Kopf-Bolanz KA , Buri C , Portmann R , Egger L , Chollet M , McTernan PG , Piya MK , Gijs MA , Vionnet N et al. A dose-response strategy reveals differences between normal-weight and obese men in their metabolic and inflammatory responses to a high-fat meal . J Nutr 2014 ; 144 : 1517 – 23 . Google Scholar CrossRef Search ADS PubMed 3. Schulkin J . Rethinking homeostasis: allostatic regulation in physiology and pathophysiology . Cambridge (MA) : MIT Press ; 2003 . 4. van Ommen B , Keijer J , Heil SG , Kaput J . Challenging homeostasis to define biomarkers for nutrition related health . Mol Nutr Food Res 2009 ; 53 : 795 – 804 . Google Scholar CrossRef Search ADS PubMed 5. Kardinaal AF , van Erk MJ , Dutman AE , Stroeve JH , van de Steeg E , Bijlsma S , Kooistra T , van Ommen B , Wopereis S . Quantifying phenotypic flexibility as the response to a high-fat challenge test in different states of metabolic health . FASEB J 2015 ; 29 : 4600 – 13 . Google Scholar CrossRef Search ADS PubMed 6. van Oostrom AJ , Rabelink TJ , Verseyden C , Sijmonsma TP , Plokker HW , De Jaegere PP , Cabezas MC . Activation of leukocytes by postprandial lipemia in healthy volunteers . Atheroscler Suppl 2004 ; 177 : 175 – 82 . Google Scholar CrossRef Search ADS 7. Kolovou GD , Anagnostopoulou KK , Daskalopoulou SS , Mikhailidis DP , Cokkinos DV . Clinical relevance of postprandial lipaemia . Curr Med Chem 2005 ; 12 : 1931 – 45 . Google Scholar CrossRef Search ADS PubMed 8. Hernandez EA , Kahl S , Seelig A , Begovatz P , Irmler M , Kupriyanova Y , Nowotny B , Nowotny P , Herder C , Barosa C et al. Acute dietary fat intake initiates alterations in energy metabolism and insulin resistance . J Clin Invest 2017 ; 127 : 695 – 708 . Google Scholar CrossRef Search ADS PubMed 9. Bae JH , Bassenge E , Kim KB , Kim YN , Kim KS , Lee HJ , Moon KC , Lee MS , Park KY , Schwemmer M . Postprandial hypertriglyceridemia impairs endothelial function by enhanced oxidant stress . Atheroscler Suppl 2001 ; 155 : 517 – 23 . Google Scholar CrossRef Search ADS 10. Erridge C , Attina T , Spickett CM , Webb DJ . A high-fat meal induces low-grade endotoxemia: evidence of a novel mechanism of postprandial inflammation . Am J Clin Nutr 2007 ; 86 : 1286 – 92 . Google Scholar CrossRef Search ADS PubMed 11. Deveaux A , Pham I , West SG , Andre E , Lantoine-Adam F , Bunouf P , Sadi S , Hermier D , Mathe V , Fouillet H et al. l-Arginine supplementation alleviates postprandial endothelial dysfunction when baseline fasting plasma arginine concentration is low: a randomized controlled trial in healthy overweight adults with cardiometabolic risk factors . J Nutr 2016 ; 146 : 1330 – 40 . Google Scholar CrossRef Search ADS PubMed 12. Blackburn P , Despres JP , Lamarche B , Tremblay A , Bergeron J , Lemieux I , Couillard C . Postprandial variations of plasma inflammatory markers in abdominally obese men . Obesity (Silver Spring) 2006 ; 14 : 1747 – 54 . Google Scholar CrossRef Search ADS PubMed 13. Dandona P , Ghanim H , Chaudhuri A , Dhindsa S , Kim SS . Macronutrient intake induces oxidative and inflammatory stress: potential relevance to atherosclerosis and insulin resistance . Exp Mol Med 2010 ; 42 : 245 – 53 . Google Scholar CrossRef Search ADS PubMed 14. Mariotti F , Valette M , Lopez C , Fouillet H , Famelart MH , Mathe V , Airinei G , Benamouzig R , Gaudichon C , Tome D et al. Casein compared with whey proteins affects the organization of dietary fat during digestion and attenuates the postprandial triglyceride response to a mixed high-fat meal in healthy, overweight men . J Nutr 2015 ; 145 : 2657 – 64 . Google Scholar CrossRef Search ADS PubMed 15. Gannon MC , Nuttall FQ . Amino acid ingestion and glucose metabolism—a review . IUBMB Life 2010 ; 62 : 660 – 8 . Google Scholar CrossRef Search ADS PubMed 16. Calbet JA , MacLean DA . Plasma glucagon and insulin responses depend on the rate of appearance of amino acids after ingestion of different protein solutions in humans . J Nutr 2002 ; 132 : 2174 – 82 . Google Scholar CrossRef Search ADS PubMed 17. Tremblay F , Lavigne C , Jacques H , Marette A . Role of dietary proteins and amino acids in the pathogenesis of insulin resistance . Annu Rev Nutr 2007 ; 27 : 293 – 310 . Google Scholar CrossRef Search ADS PubMed 18. Wolf N , Newsome SD , Peters J , Fogel ML . Variability in the routing of dietary proteins and lipids to consumer tissues influences tissue-specific isotopic discrimination . Rapid Commun Mass Spectrom 2015 ; 29 : 1448 – 56 . Google Scholar CrossRef Search ADS PubMed 19. Baker PR II , Boyle KE , Koves TR , Ilkayeva OR , Muoio DM , Houmard JA , Friedman JE . Metabolomic analysis reveals altered skeletal muscle amino acid and fatty acid handling in obese humans . Obesity (Silver Spring) 2015 ; 23 : 981 – 8 . Google Scholar CrossRef Search ADS PubMed 20. Adams SH . Emerging perspectives on essential amino acid metabolism in obesity and the insulin-resistant state . Adv Nutr 2011 ; 2 : 445 – 56 . Google Scholar CrossRef Search ADS PubMed 21. Newgard CB . Interplay between lipids and branched-chain amino acids in development of insulin resistance . Cell Metab 2012 ; 15 : 606 – 14 . Google Scholar CrossRef Search ADS PubMed 22. Badoud F , Lam KP , Perreault M , Zulyniak MA , Britz-McKibbin P , Mutch DM . Metabolomics reveals metabolically healthy and unhealthy obese individuals differ in their response to a caloric challenge . PLoS One 2015 ; 10 : e0134613 . Google Scholar CrossRef Search ADS PubMed 23. Zivkovic AM , Wiest MM , Nguyen U , Nording ML , Watkins SM , German JB . Assessing individual metabolic responsiveness to a lipid challenge using a targeted metabolomic approach . Metabolomics 2009 ; 5 : 209 – 18 . Google Scholar CrossRef Search ADS PubMed 24. Pellis L , van Erk MJ , van Ommen B , Bakker GC , Hendriks HF , Cnubben NH , Kleemann R , van Someren EP , Bobeldijk I , Rubingh CM et al. Plasma metabolomics and proteomics profiling after a postprandial challenge reveal subtle diet effects on human metabolic status . Metabolomics 2012 ; 8 : 347 – 59 . Google Scholar CrossRef Search ADS PubMed 25. Morio B , Comte B , Martin JF , Chanseaume E , Alligier M , Junot C , Lyan B , Boirie Y , Vidal H , Laville M et al. Metabolomics reveals differential metabolic adjustments of normal and overweight subjects during overfeeding . Metabolomics 2015 ; 11 : 920 – 38 . Google Scholar CrossRef Search ADS 26. Karimpour M , Surowiec I , Wu J , Gouveia-Figueira S , Pinto R , Trygg J , Zivkovic AM , Nording ML . Postprandial metabolomics: a pilot mass spectrometry and NMR study of the human plasma metabolome in response to a challenge meal . Anal Chim Acta 2016 ; 908 : 121 – 31 . Google Scholar CrossRef Search ADS PubMed 27. Meikle PJ , Barlow CK , Mellett NA , Mundra PA , Bonham MP , Larsen A , Cameron-Smith D , Sinclair A , Nestel PJ , Wong G . Postprandial plasma phospholipids in men are influenced by the source of dietary fat . J Nutr 2015 ; 145 : 2012 – 8 . Google Scholar CrossRef Search ADS PubMed 28. Benton HP , Wong DM , Trauger SA , Siuzdak G . XCMS2: processing tandem mass spectrometry data for metabolite identification and structural characterization . Anal Chem 2008 ; 80 : 6382 – 9 . Google Scholar CrossRef Search ADS PubMed 29. Smith CA , Want EJ , O'Maille G , Abagyan R , Siuzdak G . XCMS: processing mass spectrometry data for metabolite profiling using nonlinear peak alignment, matching, and identification . Anal Chem 2006 ; 78 : 779 – 87 . Google Scholar CrossRef Search ADS PubMed 30. Wishart DS . Human Metabolome Database: completing the ‘human parts list’ . Pharmacogenomics 2007 ; 8 : 683 – 6 . Google Scholar CrossRef Search ADS PubMed 31. Sumner LW , Amberg A , Barrett D , Beale MH , Beger R , Daykin CA , Fan TW , Fiehn O , Goodacre R , Griffin JL et al. Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI) . Metabolomics 2007 ; 3 : 211 – 21 . Google Scholar CrossRef Search ADS PubMed 32. Bates D , Mächler M , Bolker B , Walker S . Fitting linear mixed-effects models using lme4 . J Stat Softw 2015 ; 67 . doi: 10.18637/jss.v067.i01 . 33. Benjamini Y , Hochberg Y . Controlling the false discovery rate—a practical and powerful approach to multiple testing . J R Stat Soc Series B Stat Methodol 1995 ; 57 : 289 – 300 . 34. Bardou P , Mariette J , Escudie F , Djemiel C , Klopp C . jvenn: an interactive Venn diagram viewer . BMC Bioinformatics 2014 ; 15 : 293 . Google Scholar CrossRef Search ADS PubMed 35. Cottret L , Wildridge D , Vinson F , Barrett MP , Charles H , Sagot MF , Jourdan F . MetExplore: a web server to link metabolomic experiments and genome-scale metabolic networks . Nucleic Acids Res 2010 ; 38 ( Web Server issue ): W132 – 7 . Google Scholar CrossRef Search ADS PubMed 36. Poupin N , Jourdan F . Analysing human metabolic networks using metabolomics: understanding the impact of diet on health . In: Brennan L, , Sébédio J-L , editors. Metabolomics as a tool in nutrition research . Cambridge : Woodhead Publishing ; 2015 . p. 85 – 114 . Google Scholar CrossRef Search ADS 37. Thiele I , Swainston N , Fleming RM , Hoppe A , Sahoo S , Aurich MK , Haraldsdottir H , Mo ML , Rolfsson O , Stobbe MD et al. A community-driven global reconstruction of human metabolism . Nat Biotechnol 2013 ; 31 : 419 – 25 . Google Scholar CrossRef Search ADS PubMed 38. Piccolo BD , Comerford KB , Karakas SE , Knotts TA , Fiehn O , Adams SH . Whey protein supplementation does not alter plasma branched-chained amino acid profiles but results in unique metabolomics patterns in obese women enrolled in an 8-week weight loss trial . J Nutr 2015 ; 145 : 691 – 700 . Google Scholar CrossRef Search ADS PubMed 39. Adams SH , Hoppel CL , Lok KH , Zhao L , Wong SW , Minkler PE , Hwang DH , Newman JW , Garvey WT . Plasma acylcarnitine profiles suggest incomplete long-chain fatty acid beta-oxidation and altered tricarboxylic acid cycle activity in type 2 diabetic African-American women . J Nutr 2009 ; 139 : 1073 – 81 . Google Scholar CrossRef Search ADS PubMed 40. Schooneman MG , Vaz FM , Houten SM , Soeters MR . Acylcarnitines: reflecting or inflicting insulin resistance? Diabetes 2013 ; 62 : 1 – 8 . Google Scholar CrossRef Search ADS PubMed 41. Muoio DM , Noland RC , Kovalik JP , Seiler SE , Davies MN , DeBalsi KL , Ilkayeva OR , Stevens RD , Kheterpal I , Zhang J et al. Muscle-specific deletion of carnitine acetyltransferase compromises glucose tolerance and metabolic flexibility . Cell Metab 2012 ; 15 : 764 – 77 . Google Scholar CrossRef Search ADS PubMed 42. McCoin CS , Knotts TA , Adams SH . Acylcarnitines—old actors auditioning for new roles in metabolic physiology . Nat Rev Endocrinol 2015 ; 11 : 617 – 25 . Google Scholar CrossRef Search ADS PubMed 43. Rauschert S , Uhl O , Koletzko B , Hellmuth C . Metabolomic biomarkers for obesity in humans: a short review . Ann nutr metab 2014 ; 64 : 314 – 24 . Google Scholar CrossRef Search ADS PubMed 44. Mihalik SJ , Goodpaster BH , Kelley DE , Chace DH , Vockley J , Toledo FG , DeLany JP . Increased levels of plasma acylcarnitines in obesity and type 2 diabetes and identification of a marker of glucolipotoxicity . Obesity (Silver Spring) 2010 ; 18 : 1695 – 700 . Google Scholar CrossRef Search ADS PubMed 45. Boulet MM , Chevrier G , Grenier-Larouche T , Pelletier M , Nadeau M , Scarpa J , Prehn C , Marette A , Adamski J , Tchernof A . Alterations of plasma metabolite profiles related to adipose tissue distribution and cardiometabolic risk . Am J Physiol Endocrinol Metab 2015 ; 309 : E736 – 46 . Google Scholar CrossRef Search ADS PubMed 46. Newgard CB , An J , Bain JR , Muehlbauer MJ , Stevens RD , Lien LF , Haqq AM , Shah SH , Arlotto M , Slentz CA et al. A branched-chain amino acid-related metabolic signature that differentiates obese and lean humans and contributes to insulin resistance . Cell Metab 2009 ; 9 : 311 – 26 . Google Scholar CrossRef Search ADS PubMed 47. Lynch CJ , Adams SH . Branched-chain amino acids in metabolic signalling and insulin resistance . Nat Rev Endocrinol 2014 ; 10 : 723 – 36 . Google Scholar CrossRef Search ADS PubMed 48. Polakof S , Remond D , Bernalier-Donadille A , Rambeau M , Pujos-Guillot E , Comte B , Dardevet D , Savary-Auzeloux I . Metabolic adaptations to HFHS overfeeding: how whole body and tissues postprandial metabolic flexibility adapt in Yucatan mini-pigs . Eur J Nutr 2018 ; 57 : 119 – 35 . Google Scholar CrossRef Search ADS PubMed 49. Mai M , Tonjes A , Kovacs P , Stumvoll M , Fiedler GM , Leichtle AB . Serum levels of acylcarnitines are altered in prediabetic conditions . PLoS One 2013 ; 8 : e82459 . Google Scholar CrossRef Search ADS PubMed 50. Thompson DK , Sloane R , Bain JR , Stevens RD , Newgard CB , Pieper CF , Kraus VB . Daily variation of serum acylcarnitines and amino acids . Metabolomics 2012 ; 8 : 556 – 65 . Google Scholar CrossRef Search ADS PubMed 51. Shrestha A , Mullner E , Poutanen K , Mykkanen H , Moazzami AA . Metabolic changes in serum metabolome in response to a meal . Eur J Nutr 2017 ; 56 : 671 – 81 . Google Scholar CrossRef Search ADS PubMed 52. Boyle KE , Canham JP , Consitt LA , Zheng D , Koves TR , Gavin TP , Holbert D , Neufer PD , Ilkayeva O , Muoio DM et al. A high-fat diet elicits differential responses in genes coordinating oxidative metabolism in skeletal muscle of lean and obese individuals . J Clin Endocrinol Metab 2011 ; 96 : 775 – 81 . Google Scholar CrossRef Search ADS PubMed 53. Berg JM , Tymoczko JL , Stryer L . Carbon atoms of degraded amino acids emerge as major metabolic intermediates . In: Biochemistry . 5th ed . New-York : W H Freeman ; 2002 . 54. Hoefle AS , Bangert AM , Stamfort A , Gedrich K , Rist MJ , Lee YM , Skurk T , Daniel H . Metabolic responses of healthy or prediabetic adults to bovine whey protein and sodium caseinate do not differ . J Nutr 2015 ; 145 : 467 – 75 . Google Scholar CrossRef Search ADS PubMed 55. Kelley DE . Skeletal muscle fat oxidation: timing and flexibility are everything . J Clin Invest 2005 ; 115 : 1699 – 702 . Google Scholar CrossRef Search ADS PubMed 56. Ukropcova B , McNeil M , Sereda O , de Jonge L , Xie H , Bray GA , Smith SR . Dynamic changes in fat oxidation in human primary myocytes mirror metabolic characteristics of the donor . J Clin Invest 2005 ; 115 : 1934 – 41 . Google Scholar CrossRef Search ADS PubMed 57. Morris C , O'Grada C , Ryan M , Roche HM , Gibney MJ , Gibney ER , Brennan L . The relationship between BMI and metabolomic profiles: a focus on amino acids . Proc Nutr Soc 2012 ; 71 : 634 – 8 . Google Scholar CrossRef Search ADS PubMed 58. Koves TR , Ussher JR , Noland RC , Slentz D , Mosedale M , Ilkayeva O , Bain J , Stevens R , Dyck JR , Newgard CB et al. Mitochondrial overload and incomplete fatty acid oxidation contribute to skeletal muscle insulin resistance . Cell Metab 2008 ; 7 : 45 – 56 . Google Scholar CrossRef Search ADS PubMed 59. Watt MJ , Hevener AL . Fluxing the mitochondria to insulin resistance . Cell Metab 2008 ; 7 : 5 – 6 . Google Scholar CrossRef Search ADS PubMed © 2018 American Society for Nutrition. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Nutrition Oxford University Press

Metabolomics Reveals that the Type of Protein in a High-Fat Meal Modulates Postprandial Mitochondrial Overload and Incomplete Substrate Oxidation in Healthy Overweight Men

Loading next page...
 
/lp/ou_press/metabolomics-reveals-that-the-type-of-protein-in-a-high-fat-meal-LrToB0LNuy
Publisher
Oxford University Press
Copyright
© 2018 American Society for Nutrition.
ISSN
0022-3166
eISSN
1541-6100
D.O.I.
10.1093/jn/nxy049
Publisher site
See Article on Publisher Site

Abstract

Abstract Background A meal rich in saturated fatty acids induces a postprandial metabolic challenge. The type of dietary protein may modulate postprandial metabolism. Objective We studied the effect of dietary protein type on postprandial changes in the metabolome after a high-fat meal. Methods In a 3-period, crossover, postprandial study, 10 healthy overweight men with an elevated waist circumference (>94 cm) ingested high-fat meals made up of cream fat (70% of energy), sucrose (15% energy), and protein (15% energy) from either casein (CAS), whey protein (WHE), or α-lactalbumin-enriched whey protein (LAC). Urine collected immediately before and 2, 4, and 6 h after the meal was analyzed for metabolomics, a secondary outcome of the clinical study. We used mixed-effect models, partial least-square regression, and pathway enrichment analysis. Results At 4 and 6 h after the meal, the postprandial metabolome was found to be fully discriminated according to protein type. We identified 17 metabolites that significantly explained the effect of protein type on postprandial metabolomic changes (protein-time interaction). Among this signature, acylcarnitines and other acylated metabolites related to fatty acid or amino acid oxidation were the main discriminant features. The difference in metabolic profiles was mainly explained by urinary acylcarnitines and some other acylated products (protein type, Ps < 0.0001), with a dramatically greater increase (100- to 1000-fold) after WHE, and to a lesser extent after LAC, as compared with CAS. Pathway enrichment analysis confirmed that the type of protein had modified fatty acid oxidation (P < 0.05). Conclusion Taken together, our results indicate that, in healthy overweight men, the type of protein in a high-fat meal interplays with fatty acid oxidation with a differential accumulation of incomplete oxidation products. A high-fat meal containing WHE, but not CAS, resulted in this outpacing of the tricarboxylic acid cycle. This study was registered at clinicaltrials.gov as NCT00931151. high-fat meal, dietary protein, postprandial, metabolomics, acylcarnitines, humans, overweight men, urine, BCAA Introduction During the postprandial period, the metabolism is challenged by the sudden influx of energy nutrients, resulting in an allostatic load (1–5). In healthy individuals, a high-saturated-fat, high-sucrose meal is considered a metabolic challenge, and has been associated with postprandial low-grade inflammatory activation, endothelial dysfunction, and impaired insulin action (6–8). This paradigm has been much studied in order to decipher the causal relations between metabolic and physiologic changes that occur during the postprandial period (9–11). These phenomena are exacerbated in overweight, obese, or cardiometabolic individuals (2, 5, 12). Unlike saturated fat and sugar, the third energy macronutrient, namely protein, does not contribute to this postprandial metabolic challenge, but little is known about how different types of dietary protein affect postprandial metabolism (13, 14). There are many ways in which dietary proteins and the resulting amino acids may affect postprandial metabolism. Firstly, amino acids may rapidly stimulate insulin secretion and action (15–17). Dietary proteins and amino acids may also affect the metabolic routing of fatty acids and sugars (18). In particular, after a meal, when amino acids and fatty acid oxidative intermediates largely fuel the tricarboxylic acid (TCA) cycle (19), the amino acid composition of the protein may affect the final postprandial changes in metabolic fluxes through its effect on the intermediary metabolism of amino acids. This could be especially true in a situation of compromised insulin action, such as in overweight individuals, in whom the expression of an altered intermediary metabolism of fats and amino acids (20) could be heightened in the postprandial period. Furthermore, fatty acid oxidation is known to downregulate BCAA oxidation at the branched-chain α-ketoacid dehydrogenase complex (20). Although this interplay between fatty acids and amino acid metabolism is not BCAA specific, BCAA metabolism in obesity and overnutrition has been the subject of much investigation in relation to the initiation of insulin resistance (20, 21). However, it remains largely unknown how dietary protein affects postprandial intermediary metabolism, particularly during the postprandial period or following short-term exposure to a high-fat diet. Metabolomic approaches are highly appropriate to study the trajectories of complex changes in intricate metabolic pathways such as those that occur during the postprandial phase (22–24). Since the plasma metabolome is tempered by homeostatic forces, whereas postprandial changes in metabolic fluxes should be integrated in the urinary metabolomic output, urinary metabolomics is expected to be more discriminating (25). Clinical studies involving repeated postprandial challenges offer opportunities to identify the effect of meal composition on postprandial metabolism using metabolomics (26, 27), by characterizing overall changes and identifying the most salient contributors to these changes (4). Postprandial metabolomic studies of the effect of protein intake are scarce. In this study, we analyzed the postprandial urinary metabolomics of healthy overweight men receiving high-fat meals differing in the type of dietary protein: either casein (CAS), whey protein (WHE), or α-lactalbumin-enriched whey protein (LAC). We chose dietary proteins that are of practical interest and have contrasting nutritional characteristics, including differences in their amino acid (e.g., leucine) content. Our aim was to understand the degree to which each type of protein affects postprandial changes in metabolomics that are driven by a high fat load and to try and identify the set of metabolites that are the most important contributors. Methods Subjects and study design Ten healthy overweight men who had participated in a postprandial study designed to examine the acute effects of different dietary proteins in high-fat meals on postprandial TGs and some physiological markers were selected for our study on urinary metabolomics. Details of this randomized, 3-phase crossover study were reported previously (14). Urinary metabolomics was a secondary outcome of this clinical study. All the subjects were healthy men (age range 21–50 y) who were overweight [BMI (kg/m2) >25] and had an enlarged waist circumference (>94 cm). Exclusion criteria included any established disease or regular use of medication, hypertension, excessive alcohol intake (>2 units of alcohol per day), moderate or high use of tobacco products (>9 cigarettes or equivalent per day), use of any nutritional supplements, moderate or vigorous physical activity (taken as >4 h of moderate/vigorous physical activity per week), and blood hemoglobin <13 g/dL. The characteristics of these subjects were as follows (mean ± SD): age, 34 ± 9 y; height, 178 ± 3 cm; weight, 96 ± 6 kg; BMI, 30.2 ± 1.5; body fat, 24.3% ± 2.0%; waist circumference, 96.3 ± 3.2 cm. All participants gave their written informed consent prior to enrollment. The study was approved by the Institutional Review Board for Saint-Germain-en-Laye Hospital, authorized by the French Ministry of Health, and registered at clinicaltrials.gov as NCT00931151. All subjects completed 3 treatment sessions consisting of a postprandial study, which were separated by ≥2 wk. The subjects were required to consume 1 of the 3 test meals during each treatment session in a randomized order, according to a Latin square. The participants were asked to empty their bladders of night/early-morning urine. Urine was then collected immediately before the meal and every 2 h thereafter (0, 2, 4, and 6 h after the meal). Blood was sampled before the meal and 0.5, 1, 1.5, 2, 3, 4, and 6 h after the meal. Test meals The composition of the meals was as follows: energy, 1200 kcal; fat, 93.3 g (i.e., 70% energy); carbohydrates, 45 g (i.e., 15% energy); crude protein, 45 g (i.e., 15% energy). The test meals consisted of a mixture of 233 g 40%-fat cream, 45 g sucrose, 160 mL water, and protein isolates of either CAS, WHE, or LAC, in quantities (54, 55, and 49 g, respectively) adjusted to yield the same amount of protein (45 g). The cream and sugar were standard commercial products and the protein isolates were supplied by Ingredia (CAS and WHE) and Armor Protein (LAC). Because the amounts of remaining carbohydrates and minerals were not the same in these protein-rich powders, the meals were equilibrated with the addition of lactose, calcium phosphate, potassium phosphate, and magnesium phosphate. TGs and apoB-48 in plasma TGs were assayed with the use of an enzymatic colorimetric method with commercial kits (RANDOX Laboratories), and apoB-48 were assayed with the use of an ELISA assay (Biovendor), as previously described (14) Metabolomic analysis of urine samples The urine samples were defrosted at room temperature, centrifuged at 7000 × g for 5 min at 4°C, and then diluted 4-fold with distilled water. Urinary metabolic profiles were determined with the use of an untargeted metabolomic approach that could cover a broad range of metabolites. Analyses were performed following a procedure described in Morio et al. (25), based on a Waters Acquity UPLC chromatographic system (Waters Corporation) coupled to a Waters Qtof-Micro equipped with an electrospray source and a lockmass sprayer to ensure accuracy. MS data were collected in continuum full-scan mode with a mass-to-charge ratio (m/z) of 70–1000 from 0 to 22 min, in positive mode. To prevent possible differences between the sample batches, a Latin square method was performed to obtain a randomized list of samples for analysis. For analysis of each sample, 6 µL of diluted urine was injected into a 100 × 2.1 mm, 1.7 µm BEH Shield RP18 column at 30°C. The mobile phase components were 1% formic acid (A) and acetonitrile with 1% formic acid (B). The flow rate was set at 400 µL/min. The raw data were transformed to centroid mode and mass corrected before being analyzed with the use of the XCMS platform (28, 29). The LC-MS data were peak detected and noise reduced for both the LC and MS components. Each peak in the resulting 3-dimensional data set was represented by a retention time m/z and its ion intensity in each sample. The matrix obtained was then exported for statistical analysis. Since this study was closely controlled for fluid intake during the experimental sessions, the urinary volume did not vary between conditions [the mean urinary volumes after the CAS, WHE, and LAC meals were 206 ± 24, 188 ± 24 and 254 ± 30 mL (mean ± SE), respectively]. The metabolites were expressed as raw (log-transformed) relative intensities and not quantified as concentrations. Accordingly, we did not normalize by urinary volume or creatinine concentration, nor did we calculate total urinary output. Metabolites contributing to the discrimination of the different phenotypes were first identified with the use of an in-house database containing the reference spectra of >1000 authentic standard compounds. The remaining unknown compounds were then identified on the basis of their exact masses which were compared with those registered in the Human Metabolome Database (30). Metabolites were classified according to the method described by Sumner et al. (31), which is based on levels of confidence in the identification process, as follows: identified (confirmed by a standard), putatively annotated (based on physicochemical properties or spectral similarity with public/commercial spectral libraries), and unknown compounds. Statistical analyses A statistical workflow based on univariate and multivariate analyses was used for the metabolomic data. All ion intensities were log transformed before analysis. Mixed-effect models were built with the lme4 R package v. 3.1.1 (32). Time, protein, time-protein interaction, and batch were included in the models. A Benjamini Hochberg P value adjustment (33) was used for multiple testing, and when a fixed effect was significant, post-hoc comparisons with Bonferroni corrections were also performed. Significance was set at a corrected P value <0.05. In order to reveal the specificity of features with a significant protein or interaction effect, Venn diagrams were generated from the internet application JVENN (34). Multivariate analyses by partial least square (PLS) were performed to rank the combinatorial predictive ability of the candidate biomarkers for proteins, time, and interactions. SIMCA-P+ v. 13 (Umetrics) was used. The analysis was performed first on all the variables, then on variables indicating a significant protein effect at a given time point after the meal, and finally on variables with a significant interaction effect at a given time point after the meal. The overall quality of the models was assessed with the use of the cumulative R2 (R2Ycum) and cumulative Q2 (Q2cum) criteria. Moreover, in order to verify that PLS components could not lead to classifications by chance, a permutation test based on 100 random permutations was performed. Variable importance on projection (VIP) values were used to assess the importance of each ion in the PLS discriminant analysis (PLS-DA) model. R software was used to perform Spearman correlations (and tests of significance) with and without the meal protein in the model, between ions and plasma TGs, and between ions and plasma apoB-48. Plasma concentrations were log transformed before analysis. Metabolite pathway enrichment analysis Through the use of MetExplore (35, 36), the metabolites identified as significantly explaining the effect of protein type on postprandial changes (protein-time interaction) were mapped on a genome-scale model of human metabolism [RECON 2.04, (37)], in order to identify the corresponding metabolic pathways/subnetworks that were significantly altered by pathway enrichment. Multiple comparisons across metabolic subnetworks were determined from Bonferroni corrections. Results Metabolomics reveals differences in postprandial kinetics Under positive ionization, 5225 ions were extracted from the raw data. Univariate statistical methods based on the use of a mixed-effect model revealed 1150 ions that were significantly modulated according to ≥1 of the following 3 criteria: 139 ions were dependent on the protein type (i.e., CAS, WHE, LAC), 1078 on time (i.e., 0, 2, 4, 6 h after the meal), and 36 on the time-protein interaction. Post-hoc comparisons indicated that ∼85% of variables with a protein effect differed significantly between CAS and WHE and between LAC and WHE, and only 47.5% differed significantly different between CAS and LAC. On Venn diagrams (Figure 1), around half of the variables with a protein effect were specific to WHE, and 27 ions were discriminant for the 3 meals. In the same way, 17 ions with an interaction effect were specific to WHE, and 19 ions were found to differ significantly between the 3 proteins (Figure 1). Post-hoc comparisons of interactions revealed that 2, 4, or 6 h after the meal, 75–100% of variables with an interaction differed significantly between CAS and WHE and between LAC and WHE, and <50% differed significantly between CAS and LAC. FIGURE 1 View largeDownload slide Venn diagrams showing the numbers of ions significantly modulated as a function of protein type (A, 139 ions) and the protein-time interaction (B, 36 ions), together with respective histograms representing the numbers of significant ions for each dietary protein type (C, D) in 10 healthy overweight men after a high-fat meal that included CAS, WHE, or LAC. CAS, casein; LAC, α-lactalbumin-enriched whey protein; WHE, whey protein. FIGURE 1 View largeDownload slide Venn diagrams showing the numbers of ions significantly modulated as a function of protein type (A, 139 ions) and the protein-time interaction (B, 36 ions), together with respective histograms representing the numbers of significant ions for each dietary protein type (C, D) in 10 healthy overweight men after a high-fat meal that included CAS, WHE, or LAC. CAS, casein; LAC, α-lactalbumin-enriched whey protein; WHE, whey protein. PLS-DA analyses performed on all extracted variables (Supplemental Figure 1) enabled the prediction of protein type with the use of a model including 5 components (R2Ycum = 0.863 and Q2cum = 0.662). A total of 1855 ions had a VIP >1. In particular, 110 ions among the 139 ions with a significant global protein effect were within the first quarter of VIP, and 35 ions among the 36 features with a significant global interaction effect were within the first 6% of VIP. PLS-DA models were then built at each time point (i.e., 0, 2, 4, or 6 h after the meal), introducing only the 139 variables with a significant protein effect to rank the combinatorial predictive ability of these candidate biomarkers. No valid model was found before the meal and 2 h after the meal. By contrast, at 4 and 6 h after the meal, the models with 2 components were specific and predictive with R2Ycum values of 0.867 and 0.875, respectively, and Q2cum values of 0.757 and 0.739, respectively (Figure 2). Twenty-six of the 36 ions with a significant global interaction effect were always within the first quarter of VIP at 4 or 6 h after the meals. In terms of these features, the urine metabolome was much more markedly modified after WHE than after the other proteins (Figure 3). FIGURE 2 View largeDownload slide Discrimination of protein type in the meal based on the scores plot of PLS-DA performed on the 139 variables with a significant protein effect 4 h after the meal (A) and 6 h after the meal (B) and 100-permutation tests of PLS-DA models 4 h after the meal (C) and 6 h after the meal (D), in 10 healthy overweight men after a high-fat meal that included CAS, WHE, or LAC. The permutation plots (C and D) show the Pearson correlation coefficient between the original y variable and the permuted y variable (x axis) versus the cumulative R2 and Q2 (y axis), and the regression line. CAS, casein; LAC, α-lactalbumin-enriched whey protein; PLS-DA, partial least-squares discriminant analysis; t[1], partial least-squares component 1; t[2], partial least-squares component 2; WHE, whey protein. FIGURE 2 View largeDownload slide Discrimination of protein type in the meal based on the scores plot of PLS-DA performed on the 139 variables with a significant protein effect 4 h after the meal (A) and 6 h after the meal (B) and 100-permutation tests of PLS-DA models 4 h after the meal (C) and 6 h after the meal (D), in 10 healthy overweight men after a high-fat meal that included CAS, WHE, or LAC. The permutation plots (C and D) show the Pearson correlation coefficient between the original y variable and the permuted y variable (x axis) versus the cumulative R2 and Q2 (y axis), and the regression line. CAS, casein; LAC, α-lactalbumin-enriched whey protein; PLS-DA, partial least-squares discriminant analysis; t[1], partial least-squares component 1; t[2], partial least-squares component 2; WHE, whey protein. FIGURE 3 View largeDownload slide Box plots representing postprandial changes to mean ion intensities (after unit-variance scaling) for 25 of the 36 significant features of the protein-time interaction [after the removal of analytical redundancy (11 isotopes, adducts and fragments)], in 10 healthy overweight men after a high-fat meal that included either CAS (A), LAC (B), or WHE (C). Boxes show median and 25th/75th percentiles, whiskers show 5th/95th percentiles, and open circles show outliers. AU, arbitrary units; CAS, casein; LAC, α-lactalbumin-enriched whey protein; WHE, whey protein. FIGURE 3 View largeDownload slide Box plots representing postprandial changes to mean ion intensities (after unit-variance scaling) for 25 of the 36 significant features of the protein-time interaction [after the removal of analytical redundancy (11 isotopes, adducts and fragments)], in 10 healthy overweight men after a high-fat meal that included either CAS (A), LAC (B), or WHE (C). Boxes show median and 25th/75th percentiles, whiskers show 5th/95th percentiles, and open circles show outliers. AU, arbitrary units; CAS, casein; LAC, α-lactalbumin-enriched whey protein; WHE, whey protein. Out of the 36 ions that were significant for the protein-time interaction, 14 were found to be parent ions (the others being isotopes, adducts, and in-source fragments) and the corresponding metabolites were identified from databases combined with high-resolution MS (Table 1). Four metabolites remained unknown, but their polarity, masses, and in-source fragmentation were suggestive of modified peptides (especially acetylated species). Interestingly, in the urine metabolomes, acylcarnitines were the metabolites most discriminant of protein type after ingestion. Amino acids and derivatives, as well as peptides, were also found to be important to discrimination. Regarding the individual kinetics of these principal discriminant metabolites (Figure 4), the greatest amplitudes in variation were found after the consumption of WHE, with postprandial increase ≤3 orders of magnitude higher than with CAS. FIGURE 4 View largeDownload slide Postprandial kinetics of ion intensities of the principal confirmed identity metabolites (A–I) for the protein-time interaction, in healthy overweight men after a high-fat meal that included CAS, WHE, or LAC. Data are means ± SDs, n = 10. Protein type and the protein type × time interaction were significant for all variables (all Ps < 0.05) according to mixed models. AU, arbitrary units; CAS, casein; GE, Glycyl-glutamate; LAC, α-lactalbumin-enriched whey protein; VI, Valyl-isoleucine; WHE, whey protein. FIGURE 4 View largeDownload slide Postprandial kinetics of ion intensities of the principal confirmed identity metabolites (A–I) for the protein-time interaction, in healthy overweight men after a high-fat meal that included CAS, WHE, or LAC. Data are means ± SDs, n = 10. Protein type and the protein type × time interaction were significant for all variables (all Ps < 0.05) according to mixed models. AU, arbitrary units; CAS, casein; GE, Glycyl-glutamate; LAC, α-lactalbumin-enriched whey protein; VI, Valyl-isoleucine; WHE, whey protein. TABLE 1 Identification of metabolites that are significant regarding the protein-time interaction in 10 healthy overweight men after a high-fat meal that included casein, whey protein, or α-lactalbumin-enriched whey protein1 PLS-DA on protein (139 ions) Statistics2 PLS-DA on protein (all ions) 4 h after meal 6 h after meal Identification Experimental mass-to-charge ratio, m:z Retention time, min Protein type onVenn diagram TimeP value Protein typeP value Protein × timeP value VIP value VIP rank VIP value VIP rank VIP value VIP rank N-Acetyl-l-aspartic acid3 176.103 1.7 CAS, LAC, and WHE 2.8 × 10−11 3.5 × 10−08 2.7 × 10−04 1.4 313 0.7 77 0.8 67 Tryptophan3 188.164 4.9 CAS, LAC, and WHE 5.8 × 10−12 1.9 × 10−07 1.1 × 10−02 1.9 83 1.2 31 1.4 23 Indolebutyric acid3 204.159 6.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.8 8 1.9 6 1.9 7 Propionylcarnitine3 218.174 8.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.5 26 1.8 8 1.6 13 Capryloylglycine 202.177 8.7 CAS, LAC, and WHE 1.2 × 10−06 4.9 × 10−06 2.6 × 10−02 1.8 125 1.0 56 1.1 43 R-Butyryl carnitine3 232.191 9.3 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.1 2 2.3 2 2.2 2 Tiglylcarnitine3 228.179 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 1.1 × 10−04 1.6 188 1.0 49 1.0 48 Valyl-isoleucine3 231.168 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 4.5 × 10−03 1.6 172 1.0 50 1.0 51 Unknown 354.012 10.2 WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.0 3 2.2 5 1.9 6 Hexanoylcarnitine3 260.221 10.9 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.3 1 2.3 1 2.3 1 Unknown 339.164 13.1 WHE 1.3 × 10−2 <2.2 × 10−16 1.6 × 10−4 2.2 43 1.3 28 1.4 27 Unknown 441.087 13.6 WHE 3.1 × 10−3 1.5 × 10−8 4.6 × 10−4 1.5 220 1.4 23 1.3 33 Glycyl-glutamate3 409.068 14.9 WHE NS 2.6 × 10−11 5.8 × 10−4 2.3 35 1.3 25 1.4 26 Unknown 455.094 14.9 WHE NS 9.9 × 10−12 1.4 × 10−5 2.0 71 1.4 20 1.6 17 PLS-DA on protein (139 ions) Statistics2 PLS-DA on protein (all ions) 4 h after meal 6 h after meal Identification Experimental mass-to-charge ratio, m:z Retention time, min Protein type onVenn diagram TimeP value Protein typeP value Protein × timeP value VIP value VIP rank VIP value VIP rank VIP value VIP rank N-Acetyl-l-aspartic acid3 176.103 1.7 CAS, LAC, and WHE 2.8 × 10−11 3.5 × 10−08 2.7 × 10−04 1.4 313 0.7 77 0.8 67 Tryptophan3 188.164 4.9 CAS, LAC, and WHE 5.8 × 10−12 1.9 × 10−07 1.1 × 10−02 1.9 83 1.2 31 1.4 23 Indolebutyric acid3 204.159 6.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.8 8 1.9 6 1.9 7 Propionylcarnitine3 218.174 8.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.5 26 1.8 8 1.6 13 Capryloylglycine 202.177 8.7 CAS, LAC, and WHE 1.2 × 10−06 4.9 × 10−06 2.6 × 10−02 1.8 125 1.0 56 1.1 43 R-Butyryl carnitine3 232.191 9.3 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.1 2 2.3 2 2.2 2 Tiglylcarnitine3 228.179 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 1.1 × 10−04 1.6 188 1.0 49 1.0 48 Valyl-isoleucine3 231.168 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 4.5 × 10−03 1.6 172 1.0 50 1.0 51 Unknown 354.012 10.2 WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.0 3 2.2 5 1.9 6 Hexanoylcarnitine3 260.221 10.9 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.3 1 2.3 1 2.3 1 Unknown 339.164 13.1 WHE 1.3 × 10−2 <2.2 × 10−16 1.6 × 10−4 2.2 43 1.3 28 1.4 27 Unknown 441.087 13.6 WHE 3.1 × 10−3 1.5 × 10−8 4.6 × 10−4 1.5 220 1.4 23 1.3 33 Glycyl-glutamate3 409.068 14.9 WHE NS 2.6 × 10−11 5.8 × 10−4 2.3 35 1.3 25 1.4 26 Unknown 455.094 14.9 WHE NS 9.9 × 10−12 1.4 × 10−5 2.0 71 1.4 20 1.6 17 1VIP obtained from PLS-DA analyses. CAS, casein; LAC, α-lactalbumin-enriched whey protein; PLS-DA, partial-least square discriminant analysis; VIP, variable importance on projection value; WHE, whey protein. 2Results from mixed-effect models with Benjamini Hochberg adjustments. NS, P > 0.05. 3Compounds identified that were confirmed by standards. View Large TABLE 1 Identification of metabolites that are significant regarding the protein-time interaction in 10 healthy overweight men after a high-fat meal that included casein, whey protein, or α-lactalbumin-enriched whey protein1 PLS-DA on protein (139 ions) Statistics2 PLS-DA on protein (all ions) 4 h after meal 6 h after meal Identification Experimental mass-to-charge ratio, m:z Retention time, min Protein type onVenn diagram TimeP value Protein typeP value Protein × timeP value VIP value VIP rank VIP value VIP rank VIP value VIP rank N-Acetyl-l-aspartic acid3 176.103 1.7 CAS, LAC, and WHE 2.8 × 10−11 3.5 × 10−08 2.7 × 10−04 1.4 313 0.7 77 0.8 67 Tryptophan3 188.164 4.9 CAS, LAC, and WHE 5.8 × 10−12 1.9 × 10−07 1.1 × 10−02 1.9 83 1.2 31 1.4 23 Indolebutyric acid3 204.159 6.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.8 8 1.9 6 1.9 7 Propionylcarnitine3 218.174 8.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.5 26 1.8 8 1.6 13 Capryloylglycine 202.177 8.7 CAS, LAC, and WHE 1.2 × 10−06 4.9 × 10−06 2.6 × 10−02 1.8 125 1.0 56 1.1 43 R-Butyryl carnitine3 232.191 9.3 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.1 2 2.3 2 2.2 2 Tiglylcarnitine3 228.179 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 1.1 × 10−04 1.6 188 1.0 49 1.0 48 Valyl-isoleucine3 231.168 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 4.5 × 10−03 1.6 172 1.0 50 1.0 51 Unknown 354.012 10.2 WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.0 3 2.2 5 1.9 6 Hexanoylcarnitine3 260.221 10.9 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.3 1 2.3 1 2.3 1 Unknown 339.164 13.1 WHE 1.3 × 10−2 <2.2 × 10−16 1.6 × 10−4 2.2 43 1.3 28 1.4 27 Unknown 441.087 13.6 WHE 3.1 × 10−3 1.5 × 10−8 4.6 × 10−4 1.5 220 1.4 23 1.3 33 Glycyl-glutamate3 409.068 14.9 WHE NS 2.6 × 10−11 5.8 × 10−4 2.3 35 1.3 25 1.4 26 Unknown 455.094 14.9 WHE NS 9.9 × 10−12 1.4 × 10−5 2.0 71 1.4 20 1.6 17 PLS-DA on protein (139 ions) Statistics2 PLS-DA on protein (all ions) 4 h after meal 6 h after meal Identification Experimental mass-to-charge ratio, m:z Retention time, min Protein type onVenn diagram TimeP value Protein typeP value Protein × timeP value VIP value VIP rank VIP value VIP rank VIP value VIP rank N-Acetyl-l-aspartic acid3 176.103 1.7 CAS, LAC, and WHE 2.8 × 10−11 3.5 × 10−08 2.7 × 10−04 1.4 313 0.7 77 0.8 67 Tryptophan3 188.164 4.9 CAS, LAC, and WHE 5.8 × 10−12 1.9 × 10−07 1.1 × 10−02 1.9 83 1.2 31 1.4 23 Indolebutyric acid3 204.159 6.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.8 8 1.9 6 1.9 7 Propionylcarnitine3 218.174 8.1 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 2.5 26 1.8 8 1.6 13 Capryloylglycine 202.177 8.7 CAS, LAC, and WHE 1.2 × 10−06 4.9 × 10−06 2.6 × 10−02 1.8 125 1.0 56 1.1 43 R-Butyryl carnitine3 232.191 9.3 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.1 2 2.3 2 2.2 2 Tiglylcarnitine3 228.179 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 1.1 × 10−04 1.6 188 1.0 49 1.0 48 Valyl-isoleucine3 231.168 9.6 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 4.5 × 10−03 1.6 172 1.0 50 1.0 51 Unknown 354.012 10.2 WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.0 3 2.2 5 1.9 6 Hexanoylcarnitine3 260.221 10.9 CAS, LAC, and WHE <2.2 × 10−16 <2.2 × 10−16 <2.2 × 10−16 3.3 1 2.3 1 2.3 1 Unknown 339.164 13.1 WHE 1.3 × 10−2 <2.2 × 10−16 1.6 × 10−4 2.2 43 1.3 28 1.4 27 Unknown 441.087 13.6 WHE 3.1 × 10−3 1.5 × 10−8 4.6 × 10−4 1.5 220 1.4 23 1.3 33 Glycyl-glutamate3 409.068 14.9 WHE NS 2.6 × 10−11 5.8 × 10−4 2.3 35 1.3 25 1.4 26 Unknown 455.094 14.9 WHE NS 9.9 × 10−12 1.4 × 10−5 2.0 71 1.4 20 1.6 17 1VIP obtained from PLS-DA analyses. CAS, casein; LAC, α-lactalbumin-enriched whey protein; PLS-DA, partial-least square discriminant analysis; VIP, variable importance on projection value; WHE, whey protein. 2Results from mixed-effect models with Benjamini Hochberg adjustments. NS, P > 0.05. 3Compounds identified that were confirmed by standards. View Large Integration of biochemical and metabolomic data Overall, all protein types taken together, 9 of the 14 metabolites with significant protein-time interactions were correlated with TG and apoB-48, with a mean Spearman correlation coefficient of ∼0.3. More precisely, and interestingly, the strength of the correlation was indeed dependent on the protein type. The highest correlation coefficient estimates were obtained after LAC, with a positive Spearman coefficient between TG and acylcarnitines ranging from 0.32 (for propionylcarnitine) to 0.55 (for hexanoylcartinine) after LAC, and from 0.23 (for propionylcarnitine) to 0.38 (for capryloylglycine) after WHE, whereas no such significant correlations were found after the CAS meal (data not shown). Pathway analysis Based on metabolite pathway enrichment analysis, we found that the fatty acid oxidation pathway was significantly affected by the type of protein source (P < 0.05). There was also a trend towards a change in the alanine and aspartate metabolism pathway (raw P < 0.01), but this was not significant after correction for overall type I error. Discussion The effect of protein type on the postprandial metabolome Our first finding was that the kinetics of the postprandial metabolome could be very clearly discriminated according to protein type even though protein only contributed 15% to the energy in the meal. This discrimination was important and pertained to the type of the protein in general, since most of the metabolites that were modulated significantly by proteins were not specific to a particular type but rather common to the 3 types taken together. With the use of metabolomics, it has recently been reported that individuals receiving long-term protein supplementation could be strongly discriminated based on the type of protein they had received, indicating that supplementation with proteins that differ in composition affects broader metabolic systems (38). Our study reveals that this occurs as early as the postprandial period. Acylcarnitines and other acylated metabolites are found to be dramatically affected by protein type The second and most important finding was that acylcarnitines and other acylated metabolites were the most important discriminant molecules in terms of the effect of protein type, and were found at dramatically differing levels in the urine as a function of protein type. New light has recently been thrown on the role of acylcartinines in the pathophysiology of metabolic diseases (39–42). Acylcarnitines are carnitine esters derived from fatty acids or amino acids that are transferred into the mitochondria. Acylcarnitine synthesis may increase when β-oxidation rates are in excess of complete oxidation to CO2 via the TCA cycle (43). Acylcartinines are found to be elevated in the plasma and skeletal muscle of individuals with insulin resistance, and they have recently been reported as being a component of the metabolomic signature in obesity (19, 39, 44). Interestingly, acylcarnitines are most often part of a metabolic profile that also features plasma amino acids such as BCAAs (43, 45–48). In the obese, higher plasma acylcarnitines have been ascribed to a higher fatty acid influx from the diet and an incomplete fatty acid oxidation in the muscle resulting from the outpacing of the TCA cycle (40). Similar results have been reported in prediabetes and diabetes (39, 49). However, fewer studies have focused on acute changes such as those that occur after a high-fat meal, and more generally on postprandial metabolomics (23, 24). Acylcarnitine modifications form part of the changes observed in the metabolome after a carbohydrate-rich meal, with a reduction in postprandial acylcarnitines in both lean and obese subjects that can be explained by a switch from β oxidation to glycolysis (22, 50, 51). By contrast, plasma acylcarnitines increase after a high-fat diet in obese individuals (and decrease in lean individuals), as was shown by Baker, Boyle and colleagues (19, 52). These authors reported an increase in medium-chain acylcarnitines in the muscle in the fed state after 5 d of high-fat feeding, indicative of a limitation in the fatty acid flux in β oxidation, and found an increase in short-chain acylcarnitines, suggesting increased anaplerosis for TCA cycle intermediates (19). They did not, however, study the changes occurring in the postprandial period itself (19). During our study, the acylcarnitines related to the protein type were of both the short-chain (i.e., C3 and C4 carnitine) and medium-chain (i.e., C6) type. Medium-chain carnitines are intermediates produced by β oxidation, whereas short-chain acylcarnitines are distal products of the degradation of amino acids (40, 53). Other acylated metabolites such as capryloylglycine are usually minor compounds formed from acyl-CoA, primarily originating from β oxidation. Acetyl-aspartic acid is also a product of acylation from acetyl-CoA. Tiglylcarnitine is a product of isoleucine catabolism. Taken together, we interpret this signature as indicating that the type of dietary protein modified the efficiency of distal β oxidation and, most importantly, regulated the overall outpacing of the TCA cycle that occurs during the postprandial period. After the meal containing WHE, when compared to that containing CAS, we found evidence of a dramatic increase in acylcarnitines and other acylated metabolites in the urine, starting 2 h after the meal and culminating ∼4 h after ingestion. It is important to note that the urinary concentrations of these metabolites were 100- to 1000-fold higher after WHE than after CAS. Postprandial changes to these metabolites after LAC were intermediate between the response to CAS and WHE. Many acylcarnitines and acylated metabolites were significantly correlated to TGs and apoB-48 after the WHE and LAC meals, but not after the CAS meal. Possible mechanisms involved in the effect of meal amino acids on intermediary metabolism The reasons for this marked differential effect of protein type remain speculative. First, it would be tempting to interpret some of our results as simply being due to a slower influx of fatty acids after the CAS meal than after the other meals, as previously shown (14). However, even if a threshold effect were postulated, the ∼20% lower TG AUC after CAS as compared with the other proteins (14) cannot explain differences in responses of 2–3 orders of magnitude. Furthermore, the differences in responses appeared quite early in the urinary metabolome (starting 2h after the meal), whereas plasma TGs remained at very similar levels with all the meals at 3 h postprandially (Supplemental Figure 2). Likewise, there were no differences in plasma nonesterified fatty acid concentrations between meals (Supplemental Figure 2). In contrast to TGs, plasma amino acids had already peaked at 1 h after meal (14), with similar kinetics between meals, confirming that potential mechanisms could rather involve the amino acid composition of the protein meals. Firstly, the oxidation of amino acids that occurred in addition to the oxidation of meal fatty acids may have directly contributed to an overflow of the TCA cycle during the postprandial phase by fueling the acetyl-CoA pool (e.g., leucine, isoleucine, tryptophan). The oxidation of amino acids also provides intermediates that share the fatty acid oxidation pathways (e.g., valine, isoleucine, lysine, tryptophan, leucine, phenylalanine, and tyrosine at the level of the enoyl-CoA hydratase or the aceto-acetyl-CoA transferase). However, and especially as compared with the high amount of fatty acids (90 g), the meals did not contain very different amounts of these amino acids (Table 2). This may explain why Hoefle et al. (54) reported similar postprandial increases in plasma short-chain acylcarnitines when adding 50 g WHE and 50 g CAS to a carbohydrate load. Lastly, tiglylcarnitine, which is specific to the metabolism of isoleucine, was differentially affected by the protein meals in our study, whereas all meals contained the same amount of isoleucine. This tends to rule out that a higher intake of certain amino acids (with WHE) was contributing as substrates to the overflow of the TCA cycle. TABLE 2 Amounts of amino acids in the high-fat meals that included CAS, WHE, or LAC1 CAS, g/meal2 WHE, g/meal2 LAC, g/meal2 Leucine 4.32 5.43 4.33 Isoleucine 2.30 2.25 2.29 Valine 2.79 2.43 2.37 Lysine 3.56 4.45 4.08 Phenylalanine 2.07 1.57 1.80 Tyrosine 2.16 1.66 1.75 Tryptophan 0.59 1.21 1.18 Threonine 1.85 2.16 2.33 Methionine 1.13 1.35 0.69 Cystine 0.32 1.71 2.32 Arginine 1.49 1.08 1.31 Histidine 1.17 0.85 1.14 Alanine 1.40 2.07 1.35 Aspartic acid 3.29 5.03 5.59 Glutamic acid 9.45 7.28 6.86 Glycine 0.81 0.81 1.18 Proline 4.19 1.89 2.08 Serine 2.16 1.80 2.37 CAS, g/meal2 WHE, g/meal2 LAC, g/meal2 Leucine 4.32 5.43 4.33 Isoleucine 2.30 2.25 2.29 Valine 2.79 2.43 2.37 Lysine 3.56 4.45 4.08 Phenylalanine 2.07 1.57 1.80 Tyrosine 2.16 1.66 1.75 Tryptophan 0.59 1.21 1.18 Threonine 1.85 2.16 2.33 Methionine 1.13 1.35 0.69 Cystine 0.32 1.71 2.32 Arginine 1.49 1.08 1.31 Histidine 1.17 0.85 1.14 Alanine 1.40 2.07 1.35 Aspartic acid 3.29 5.03 5.59 Glutamic acid 9.45 7.28 6.86 Glycine 0.81 0.81 1.18 Proline 4.19 1.89 2.08 Serine 2.16 1.80 2.37 1CAS, casein; LAC, α-lactalbumin-enriched whey protein; WHE, whey protein. 2Amounts calculated from the percentage of each individual amino acid to the sum of the amino acids measured multiplied by the crude protein content (N × 6.25) in the meals, i.e., 45 g protein. View Large TABLE 2 Amounts of amino acids in the high-fat meals that included CAS, WHE, or LAC1 CAS, g/meal2 WHE, g/meal2 LAC, g/meal2 Leucine 4.32 5.43 4.33 Isoleucine 2.30 2.25 2.29 Valine 2.79 2.43 2.37 Lysine 3.56 4.45 4.08 Phenylalanine 2.07 1.57 1.80 Tyrosine 2.16 1.66 1.75 Tryptophan 0.59 1.21 1.18 Threonine 1.85 2.16 2.33 Methionine 1.13 1.35 0.69 Cystine 0.32 1.71 2.32 Arginine 1.49 1.08 1.31 Histidine 1.17 0.85 1.14 Alanine 1.40 2.07 1.35 Aspartic acid 3.29 5.03 5.59 Glutamic acid 9.45 7.28 6.86 Glycine 0.81 0.81 1.18 Proline 4.19 1.89 2.08 Serine 2.16 1.80 2.37 CAS, g/meal2 WHE, g/meal2 LAC, g/meal2 Leucine 4.32 5.43 4.33 Isoleucine 2.30 2.25 2.29 Valine 2.79 2.43 2.37 Lysine 3.56 4.45 4.08 Phenylalanine 2.07 1.57 1.80 Tyrosine 2.16 1.66 1.75 Tryptophan 0.59 1.21 1.18 Threonine 1.85 2.16 2.33 Methionine 1.13 1.35 0.69 Cystine 0.32 1.71 2.32 Arginine 1.49 1.08 1.31 Histidine 1.17 0.85 1.14 Alanine 1.40 2.07 1.35 Aspartic acid 3.29 5.03 5.59 Glutamic acid 9.45 7.28 6.86 Glycine 0.81 0.81 1.18 Proline 4.19 1.89 2.08 Serine 2.16 1.80 2.37 1CAS, casein; LAC, α-lactalbumin-enriched whey protein; WHE, whey protein. 2Amounts calculated from the percentage of each individual amino acid to the sum of the amino acids measured multiplied by the crude protein content (N × 6.25) in the meals, i.e., 45 g protein. View Large Alternatively, it could be hypothesized that WHE (as opposed to CAS) may have increased β oxidation in such a way that this resulted in exceeding the capacity of the TCA cycle to handle oxidative products from both fatty acids and amino acids. If the possible role of meal amino acids as substrates were ruled out, it is necessary to consider that they may have mediated the results by means of a signaling effect. Indeed, this hypothesis seems the most likely because, on the one hand, amino acids have long been studied as modulators of insulin secretion and action in the postprandial setting (15) and, on the other hand, the metabolic switching and adaptability of muscle for fat oxidation is largely dependent on insulin sensitivity (55, 56). It is nevertheless difficult to identify candidate amino acids that could readily explain the orientation of the effects evidenced here, and the underlying mechanisms remain speculative. In conclusion, our study has revealed that protein type regulates the oxidative pathways of fatty acids and amino acids after a high-fat meal. An increase in acylcarnitines and other acylated products was very marked with WHE, but blunted with CAS. Our study lends credence to the importance of the interplay between amino acids and fatty acid metabolism as proposed in the context of chronic high-fat feeding (21, 43, 57). These findings may also have a pathophysiologic relevance, since it has been proposed that β oxidation of fatty acids that exceeds the capacity of the TCA cycle yields incomplete fat oxidation and mitochondrial distress, which are essential events in the pathogenesis of insulin resistance (40, 58, 59). This warrants further studies in order to determine how dietary protein may modify intermediary metabolism and the initiation of insulin resistance and inflammatory pathways as early as during the postprandial period. Acknowledgments The authors’ responsibilities were as follows—FM, EP-G, and J-FH: designed the research; EP-G, MB-B, CJ, J-FM, J-FH, and FM: conducted the research; EP-G, MB-B, and HF: analyzed the data; EP-G, MB-B, HF, J-FH, DD, and FM: interpreted the results; EP-G, MB-B, and FM: drafted the manuscript; EP-G and FM: had primary responsibility for the final content; and all authors: read and approved the final manuscript. Notes Supported in part by a grant from the National Interprofessional Centre of the Dairy Industry and the French Agency for Research and Technology (National Program for Research in Food and Human Nutrition 2006, SURPROL project). Author disclosures: EP-G, MB-B, HF, CJ, JFM, JFH, DD, and FM, no conflicts of interest. Supplemental Figures 1 and 2 are available from the “Supplementary data” link in the online posting of the article and from the same link in the online table of contents at https://academic.oup.com/jn/. Abbreviations used: CAS, casein; LAC, α-lactalbumin-enriched whey protein; PLS, partial least square; PLS-DA, PLS discriminant analysis; TCA, tricarboxylic acid; VIP, variable importance on projection; WHE, whey protein. References 1. Olza J , Calder PC . Metabolic and inflammatory responses to different caloric loads of a high-fat meal are distinct between normal-weight and obese individuals . J Nutr 2014 ; 144 : 1493 – 4 . Google Scholar CrossRef Search ADS PubMed 2. Schwander F , Kopf-Bolanz KA , Buri C , Portmann R , Egger L , Chollet M , McTernan PG , Piya MK , Gijs MA , Vionnet N et al. A dose-response strategy reveals differences between normal-weight and obese men in their metabolic and inflammatory responses to a high-fat meal . J Nutr 2014 ; 144 : 1517 – 23 . Google Scholar CrossRef Search ADS PubMed 3. Schulkin J . Rethinking homeostasis: allostatic regulation in physiology and pathophysiology . Cambridge (MA) : MIT Press ; 2003 . 4. van Ommen B , Keijer J , Heil SG , Kaput J . Challenging homeostasis to define biomarkers for nutrition related health . Mol Nutr Food Res 2009 ; 53 : 795 – 804 . Google Scholar CrossRef Search ADS PubMed 5. Kardinaal AF , van Erk MJ , Dutman AE , Stroeve JH , van de Steeg E , Bijlsma S , Kooistra T , van Ommen B , Wopereis S . Quantifying phenotypic flexibility as the response to a high-fat challenge test in different states of metabolic health . FASEB J 2015 ; 29 : 4600 – 13 . Google Scholar CrossRef Search ADS PubMed 6. van Oostrom AJ , Rabelink TJ , Verseyden C , Sijmonsma TP , Plokker HW , De Jaegere PP , Cabezas MC . Activation of leukocytes by postprandial lipemia in healthy volunteers . Atheroscler Suppl 2004 ; 177 : 175 – 82 . Google Scholar CrossRef Search ADS 7. Kolovou GD , Anagnostopoulou KK , Daskalopoulou SS , Mikhailidis DP , Cokkinos DV . Clinical relevance of postprandial lipaemia . Curr Med Chem 2005 ; 12 : 1931 – 45 . Google Scholar CrossRef Search ADS PubMed 8. Hernandez EA , Kahl S , Seelig A , Begovatz P , Irmler M , Kupriyanova Y , Nowotny B , Nowotny P , Herder C , Barosa C et al. Acute dietary fat intake initiates alterations in energy metabolism and insulin resistance . J Clin Invest 2017 ; 127 : 695 – 708 . Google Scholar CrossRef Search ADS PubMed 9. Bae JH , Bassenge E , Kim KB , Kim YN , Kim KS , Lee HJ , Moon KC , Lee MS , Park KY , Schwemmer M . Postprandial hypertriglyceridemia impairs endothelial function by enhanced oxidant stress . Atheroscler Suppl 2001 ; 155 : 517 – 23 . Google Scholar CrossRef Search ADS 10. Erridge C , Attina T , Spickett CM , Webb DJ . A high-fat meal induces low-grade endotoxemia: evidence of a novel mechanism of postprandial inflammation . Am J Clin Nutr 2007 ; 86 : 1286 – 92 . Google Scholar CrossRef Search ADS PubMed 11. Deveaux A , Pham I , West SG , Andre E , Lantoine-Adam F , Bunouf P , Sadi S , Hermier D , Mathe V , Fouillet H et al. l-Arginine supplementation alleviates postprandial endothelial dysfunction when baseline fasting plasma arginine concentration is low: a randomized controlled trial in healthy overweight adults with cardiometabolic risk factors . J Nutr 2016 ; 146 : 1330 – 40 . Google Scholar CrossRef Search ADS PubMed 12. Blackburn P , Despres JP , Lamarche B , Tremblay A , Bergeron J , Lemieux I , Couillard C . Postprandial variations of plasma inflammatory markers in abdominally obese men . Obesity (Silver Spring) 2006 ; 14 : 1747 – 54 . Google Scholar CrossRef Search ADS PubMed 13. Dandona P , Ghanim H , Chaudhuri A , Dhindsa S , Kim SS . Macronutrient intake induces oxidative and inflammatory stress: potential relevance to atherosclerosis and insulin resistance . Exp Mol Med 2010 ; 42 : 245 – 53 . Google Scholar CrossRef Search ADS PubMed 14. Mariotti F , Valette M , Lopez C , Fouillet H , Famelart MH , Mathe V , Airinei G , Benamouzig R , Gaudichon C , Tome D et al. Casein compared with whey proteins affects the organization of dietary fat during digestion and attenuates the postprandial triglyceride response to a mixed high-fat meal in healthy, overweight men . J Nutr 2015 ; 145 : 2657 – 64 . Google Scholar CrossRef Search ADS PubMed 15. Gannon MC , Nuttall FQ . Amino acid ingestion and glucose metabolism—a review . IUBMB Life 2010 ; 62 : 660 – 8 . Google Scholar CrossRef Search ADS PubMed 16. Calbet JA , MacLean DA . Plasma glucagon and insulin responses depend on the rate of appearance of amino acids after ingestion of different protein solutions in humans . J Nutr 2002 ; 132 : 2174 – 82 . Google Scholar CrossRef Search ADS PubMed 17. Tremblay F , Lavigne C , Jacques H , Marette A . Role of dietary proteins and amino acids in the pathogenesis of insulin resistance . Annu Rev Nutr 2007 ; 27 : 293 – 310 . Google Scholar CrossRef Search ADS PubMed 18. Wolf N , Newsome SD , Peters J , Fogel ML . Variability in the routing of dietary proteins and lipids to consumer tissues influences tissue-specific isotopic discrimination . Rapid Commun Mass Spectrom 2015 ; 29 : 1448 – 56 . Google Scholar CrossRef Search ADS PubMed 19. Baker PR II , Boyle KE , Koves TR , Ilkayeva OR , Muoio DM , Houmard JA , Friedman JE . Metabolomic analysis reveals altered skeletal muscle amino acid and fatty acid handling in obese humans . Obesity (Silver Spring) 2015 ; 23 : 981 – 8 . Google Scholar CrossRef Search ADS PubMed 20. Adams SH . Emerging perspectives on essential amino acid metabolism in obesity and the insulin-resistant state . Adv Nutr 2011 ; 2 : 445 – 56 . Google Scholar CrossRef Search ADS PubMed 21. Newgard CB . Interplay between lipids and branched-chain amino acids in development of insulin resistance . Cell Metab 2012 ; 15 : 606 – 14 . Google Scholar CrossRef Search ADS PubMed 22. Badoud F , Lam KP , Perreault M , Zulyniak MA , Britz-McKibbin P , Mutch DM . Metabolomics reveals metabolically healthy and unhealthy obese individuals differ in their response to a caloric challenge . PLoS One 2015 ; 10 : e0134613 . Google Scholar CrossRef Search ADS PubMed 23. Zivkovic AM , Wiest MM , Nguyen U , Nording ML , Watkins SM , German JB . Assessing individual metabolic responsiveness to a lipid challenge using a targeted metabolomic approach . Metabolomics 2009 ; 5 : 209 – 18 . Google Scholar CrossRef Search ADS PubMed 24. Pellis L , van Erk MJ , van Ommen B , Bakker GC , Hendriks HF , Cnubben NH , Kleemann R , van Someren EP , Bobeldijk I , Rubingh CM et al. Plasma metabolomics and proteomics profiling after a postprandial challenge reveal subtle diet effects on human metabolic status . Metabolomics 2012 ; 8 : 347 – 59 . Google Scholar CrossRef Search ADS PubMed 25. Morio B , Comte B , Martin JF , Chanseaume E , Alligier M , Junot C , Lyan B , Boirie Y , Vidal H , Laville M et al. Metabolomics reveals differential metabolic adjustments of normal and overweight subjects during overfeeding . Metabolomics 2015 ; 11 : 920 – 38 . Google Scholar CrossRef Search ADS 26. Karimpour M , Surowiec I , Wu J , Gouveia-Figueira S , Pinto R , Trygg J , Zivkovic AM , Nording ML . Postprandial metabolomics: a pilot mass spectrometry and NMR study of the human plasma metabolome in response to a challenge meal . Anal Chim Acta 2016 ; 908 : 121 – 31 . Google Scholar CrossRef Search ADS PubMed 27. Meikle PJ , Barlow CK , Mellett NA , Mundra PA , Bonham MP , Larsen A , Cameron-Smith D , Sinclair A , Nestel PJ , Wong G . Postprandial plasma phospholipids in men are influenced by the source of dietary fat . J Nutr 2015 ; 145 : 2012 – 8 . Google Scholar CrossRef Search ADS PubMed 28. Benton HP , Wong DM , Trauger SA , Siuzdak G . XCMS2: processing tandem mass spectrometry data for metabolite identification and structural characterization . Anal Chem 2008 ; 80 : 6382 – 9 . Google Scholar CrossRef Search ADS PubMed 29. Smith CA , Want EJ , O'Maille G , Abagyan R , Siuzdak G . XCMS: processing mass spectrometry data for metabolite profiling using nonlinear peak alignment, matching, and identification . Anal Chem 2006 ; 78 : 779 – 87 . Google Scholar CrossRef Search ADS PubMed 30. Wishart DS . Human Metabolome Database: completing the ‘human parts list’ . Pharmacogenomics 2007 ; 8 : 683 – 6 . Google Scholar CrossRef Search ADS PubMed 31. Sumner LW , Amberg A , Barrett D , Beale MH , Beger R , Daykin CA , Fan TW , Fiehn O , Goodacre R , Griffin JL et al. Proposed minimum reporting standards for chemical analysis Chemical Analysis Working Group (CAWG) Metabolomics Standards Initiative (MSI) . Metabolomics 2007 ; 3 : 211 – 21 . Google Scholar CrossRef Search ADS PubMed 32. Bates D , Mächler M , Bolker B , Walker S . Fitting linear mixed-effects models using lme4 . J Stat Softw 2015 ; 67 . doi: 10.18637/jss.v067.i01 . 33. Benjamini Y , Hochberg Y . Controlling the false discovery rate—a practical and powerful approach to multiple testing . J R Stat Soc Series B Stat Methodol 1995 ; 57 : 289 – 300 . 34. Bardou P , Mariette J , Escudie F , Djemiel C , Klopp C . jvenn: an interactive Venn diagram viewer . BMC Bioinformatics 2014 ; 15 : 293 . Google Scholar CrossRef Search ADS PubMed 35. Cottret L , Wildridge D , Vinson F , Barrett MP , Charles H , Sagot MF , Jourdan F . MetExplore: a web server to link metabolomic experiments and genome-scale metabolic networks . Nucleic Acids Res 2010 ; 38 ( Web Server issue ): W132 – 7 . Google Scholar CrossRef Search ADS PubMed 36. Poupin N , Jourdan F . Analysing human metabolic networks using metabolomics: understanding the impact of diet on health . In: Brennan L, , Sébédio J-L , editors. Metabolomics as a tool in nutrition research . Cambridge : Woodhead Publishing ; 2015 . p. 85 – 114 . Google Scholar CrossRef Search ADS 37. Thiele I , Swainston N , Fleming RM , Hoppe A , Sahoo S , Aurich MK , Haraldsdottir H , Mo ML , Rolfsson O , Stobbe MD et al. A community-driven global reconstruction of human metabolism . Nat Biotechnol 2013 ; 31 : 419 – 25 . Google Scholar CrossRef Search ADS PubMed 38. Piccolo BD , Comerford KB , Karakas SE , Knotts TA , Fiehn O , Adams SH . Whey protein supplementation does not alter plasma branched-chained amino acid profiles but results in unique metabolomics patterns in obese women enrolled in an 8-week weight loss trial . J Nutr 2015 ; 145 : 691 – 700 . Google Scholar CrossRef Search ADS PubMed 39. Adams SH , Hoppel CL , Lok KH , Zhao L , Wong SW , Minkler PE , Hwang DH , Newman JW , Garvey WT . Plasma acylcarnitine profiles suggest incomplete long-chain fatty acid beta-oxidation and altered tricarboxylic acid cycle activity in type 2 diabetic African-American women . J Nutr 2009 ; 139 : 1073 – 81 . Google Scholar CrossRef Search ADS PubMed 40. Schooneman MG , Vaz FM , Houten SM , Soeters MR . Acylcarnitines: reflecting or inflicting insulin resistance? Diabetes 2013 ; 62 : 1 – 8 . Google Scholar CrossRef Search ADS PubMed 41. Muoio DM , Noland RC , Kovalik JP , Seiler SE , Davies MN , DeBalsi KL , Ilkayeva OR , Stevens RD , Kheterpal I , Zhang J et al. Muscle-specific deletion of carnitine acetyltransferase compromises glucose tolerance and metabolic flexibility . Cell Metab 2012 ; 15 : 764 – 77 . Google Scholar CrossRef Search ADS PubMed 42. McCoin CS , Knotts TA , Adams SH . Acylcarnitines—old actors auditioning for new roles in metabolic physiology . Nat Rev Endocrinol 2015 ; 11 : 617 – 25 . Google Scholar CrossRef Search ADS PubMed 43. Rauschert S , Uhl O , Koletzko B , Hellmuth C . Metabolomic biomarkers for obesity in humans: a short review . Ann nutr metab 2014 ; 64 : 314 – 24 . Google Scholar CrossRef Search ADS PubMed 44. Mihalik SJ , Goodpaster BH , Kelley DE , Chace DH , Vockley J , Toledo FG , DeLany JP . Increased levels of plasma acylcarnitines in obesity and type 2 diabetes and identification of a marker of glucolipotoxicity . Obesity (Silver Spring) 2010 ; 18 : 1695 – 700 . Google Scholar CrossRef Search ADS PubMed 45. Boulet MM , Chevrier G , Grenier-Larouche T , Pelletier M , Nadeau M , Scarpa J , Prehn C , Marette A , Adamski J , Tchernof A . Alterations of plasma metabolite profiles related to adipose tissue distribution and cardiometabolic risk . Am J Physiol Endocrinol Metab 2015 ; 309 : E736 – 46 . Google Scholar CrossRef Search ADS PubMed 46. Newgard CB , An J , Bain JR , Muehlbauer MJ , Stevens RD , Lien LF , Haqq AM , Shah SH , Arlotto M , Slentz CA et al. A branched-chain amino acid-related metabolic signature that differentiates obese and lean humans and contributes to insulin resistance . Cell Metab 2009 ; 9 : 311 – 26 . Google Scholar CrossRef Search ADS PubMed 47. Lynch CJ , Adams SH . Branched-chain amino acids in metabolic signalling and insulin resistance . Nat Rev Endocrinol 2014 ; 10 : 723 – 36 . Google Scholar CrossRef Search ADS PubMed 48. Polakof S , Remond D , Bernalier-Donadille A , Rambeau M , Pujos-Guillot E , Comte B , Dardevet D , Savary-Auzeloux I . Metabolic adaptations to HFHS overfeeding: how whole body and tissues postprandial metabolic flexibility adapt in Yucatan mini-pigs . Eur J Nutr 2018 ; 57 : 119 – 35 . Google Scholar CrossRef Search ADS PubMed 49. Mai M , Tonjes A , Kovacs P , Stumvoll M , Fiedler GM , Leichtle AB . Serum levels of acylcarnitines are altered in prediabetic conditions . PLoS One 2013 ; 8 : e82459 . Google Scholar CrossRef Search ADS PubMed 50. Thompson DK , Sloane R , Bain JR , Stevens RD , Newgard CB , Pieper CF , Kraus VB . Daily variation of serum acylcarnitines and amino acids . Metabolomics 2012 ; 8 : 556 – 65 . Google Scholar CrossRef Search ADS PubMed 51. Shrestha A , Mullner E , Poutanen K , Mykkanen H , Moazzami AA . Metabolic changes in serum metabolome in response to a meal . Eur J Nutr 2017 ; 56 : 671 – 81 . Google Scholar CrossRef Search ADS PubMed 52. Boyle KE , Canham JP , Consitt LA , Zheng D , Koves TR , Gavin TP , Holbert D , Neufer PD , Ilkayeva O , Muoio DM et al. A high-fat diet elicits differential responses in genes coordinating oxidative metabolism in skeletal muscle of lean and obese individuals . J Clin Endocrinol Metab 2011 ; 96 : 775 – 81 . Google Scholar CrossRef Search ADS PubMed 53. Berg JM , Tymoczko JL , Stryer L . Carbon atoms of degraded amino acids emerge as major metabolic intermediates . In: Biochemistry . 5th ed . New-York : W H Freeman ; 2002 . 54. Hoefle AS , Bangert AM , Stamfort A , Gedrich K , Rist MJ , Lee YM , Skurk T , Daniel H . Metabolic responses of healthy or prediabetic adults to bovine whey protein and sodium caseinate do not differ . J Nutr 2015 ; 145 : 467 – 75 . Google Scholar CrossRef Search ADS PubMed 55. Kelley DE . Skeletal muscle fat oxidation: timing and flexibility are everything . J Clin Invest 2005 ; 115 : 1699 – 702 . Google Scholar CrossRef Search ADS PubMed 56. Ukropcova B , McNeil M , Sereda O , de Jonge L , Xie H , Bray GA , Smith SR . Dynamic changes in fat oxidation in human primary myocytes mirror metabolic characteristics of the donor . J Clin Invest 2005 ; 115 : 1934 – 41 . Google Scholar CrossRef Search ADS PubMed 57. Morris C , O'Grada C , Ryan M , Roche HM , Gibney MJ , Gibney ER , Brennan L . The relationship between BMI and metabolomic profiles: a focus on amino acids . Proc Nutr Soc 2012 ; 71 : 634 – 8 . Google Scholar CrossRef Search ADS PubMed 58. Koves TR , Ussher JR , Noland RC , Slentz D , Mosedale M , Ilkayeva O , Bain J , Stevens R , Dyck JR , Newgard CB et al. Mitochondrial overload and incomplete fatty acid oxidation contribute to skeletal muscle insulin resistance . Cell Metab 2008 ; 7 : 45 – 56 . Google Scholar CrossRef Search ADS PubMed 59. Watt MJ , Hevener AL . Fluxing the mitochondria to insulin resistance . Cell Metab 2008 ; 7 : 5 – 6 . Google Scholar CrossRef Search ADS PubMed © 2018 American Society for Nutrition. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Journal of NutritionOxford University Press

Published: Jun 7, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off