Metabolic Response to Decaﬀeinated Green Tea Extract during Rest and Moderate-Intensity Exercise Doris M. Jacobs,*,§ Adrian B. Hodgson,† Rebecca K. Randell,† Krishna Mahabir-Jagessar-T,§ Ursula Garczarek,§ Asker E. Jeukendrup,† David J. Mela,§ and Silvina Lotito# †
School of Sport and Exercise Sciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom Unilever R&D, Olivier van Noortlaan 120, 3130 AC Vlaardingen, The Netherlands # Unilever R&D, Colworth Science Park, Sharnbrook, Bedfordshire MK44 1LQ, United Kingdom §
S Supporting Information *
ABSTRACT: We previously reported that a 7 day ingestion of caﬀeinated green tea extract (cGTE) induced marked metabolic diﬀerences during rest and exercise. Here, we report the metabolic eﬀects of 1, 7, and 28 day ingestions of decaﬀeinated GTE (dGTE). In this crossover placebo-controlled study, 19 healthy males ingested dGTE or placebo (PLA) for 28 days, separated by a 28 day wash-out period. On days 1, 7, and 28, participants completed a 30 min cycling exercise 2 h after the ingestion of dGTE or PLA. Blood samples were collected at rest (t = 0 and 120 min) and during exercise (t = 150 min). Plasma was analyzed using untargeted four-phase metabolite proﬁling and targeted proﬁling of catecholamines and catechins. dGTE abolished several metabolic eﬀects when compared to our previous study with cGTE. However, following 7 and 28 day dGTE ingestions, increases in 3-hydroxybutyrate, a metabolic marker of fat oxidation, were observed at t = 0 min. dGTE ingestion did not induce signiﬁcant acute or acute-on-chronic eﬀects on endogenous metabolites just prior to and during exercise. KEYWORDS: green tea, catechin, caﬀeine, metabolomics, exercise, fat oxidation
INTRODUCTION Green tea (GT) has been reported to induce beneﬁcial eﬀects on weight management,1,2 glucose control,3 and cardiovascular risk factors.4,5 In particular, chronic ingestion of green tea extract (GTE) has been shown to promote weight loss,6 possibly due to elevated fat oxidation and total energy expenditure. Although a number of studies have shown increases in fat oxidation at rest7−10 and during exercise11,12 after GT intake, the current evidence from the literature still is inconclusive,13 as some studies have also reported no signiﬁcant eﬀects.14−16 The inconsistent study outcomes may be due to various factors, such as the duration of GTE intake, dosing, formulation and composition of GTE, sample size, population as well as variations in the measurements of fat oxidation parameters.13 With regard to the GTE composition, the catechins, especially (−)-epigallocatechin3-gallate (EGCG), are thought be the major bioactive components as demonstrated in many cell culture, animal, and human studies.17 Moreover, the amount of caﬀeine present in GTE is assumed to contribute to altering fat oxidation in humans. Several studies have shown that caﬀeine stimulates fat oxidation in humans at rest.18−20 Yet, the thermogenic eﬀect of GTE containing caﬀeine has been found to be greater than that of an equivalent amount of caﬀeine, suggesting that GT catechins also stimulate energy expenditure at rest.9 Synergistic eﬀects between caﬀeine and catechins have even been hypothesized.13,21 The potential molecular mechanisms of action aﬀecting energy expenditure, fat oxidation, fat absorption, and energy intake after ingestion of catechin- and caﬀeine-rich tea have been reviewed elsewhere.22 This study follows our previous study16 in which a 7 day supplementation of caﬀeinated GTE (cGTE) was not found to © 2014 American Chemical Society
signiﬁcantly alter whole body fat oxidation rates during exercise in healthy, physically active males. Yet, cGTE induced increased lactate concentrations during exercise,23 possibly indicating higher glycolytic activity, which is known to be associated with reduced fat oxidation.24 Considering that caﬀeine ingestion has been observed to increase lactate in arterial blood during exercise,25 we speculated that caﬀeine could have counteracted the up-regulation of fat metabolism during physical activity in our previous study. For this reason and because of the results of another study showing that decaﬀeinated GTE (dGTE) was able to augment fat oxidation during exercise,12 we had hypothesized that dGTE is more eﬀective than cGTE in increasing fat oxidation during exercise. We found, however, that neither acute (1 day) nor 7 nor 28 day dGTE ingestion signiﬁcantly changed whole body fat oxidation rates during exercise when compared to placebo (PLA).26 Nevertheless, we were interested in assessing the metabolic response in plasma similar to our previous study, in which our metabolomics approach has revealed changes in metabolites related to energy metabolism, yet not to adrenergic stimulation, at rest and during exercise.27 These eﬀects were subtle and therefore may not have reached a level that would result in signiﬁcantly altered whole body fat oxidation rates. Yet, it is conceivable that the metabolic response in plasma may be a sensitive measure at an early onset before any physiological eﬀects become apparent. In addition, metabolomics is a comprehensive and unbiased approach and therefore particularly Received: Revised: Accepted: Published: 9936
June 10, 2014 August 20, 2014 September 7, 2014 September 8, 2014 dx.doi.org/10.1021/jf502764r | J. Agric. Food Chem. 2014, 62, 9936−9943
Journal of Agricultural and Food Chemistry
Figure 1. Study design: (A, B) study arms. PLA, placebo; dGTE, decaﬀeinated green tea extract; w, week; D01, D07, D28, day 1, 7, or 28 of the respective supplementation period. All participants gave written informed consent to participate in the study. The study was approved by the Life and Sciences Ethical Review Committee at the University of Birmingham. All blood samples were collected in EDTA-containing tubes and stored on ice for no longer than 35 min. Subsequently, plasma was separated by centrifugation (3500 rpm, 15 min, 4 °C), aliquoted in 1 mL samples, and stored at −80 °C. Data Acquisition. Plasma Analysis of Polyphenols. Polyphenol concentrations in plasma from 19 subjects were analyzed at t = 0 min and t = 120 min on all six experimental days. In total, 225 samples (3 samples were missing) were measured by high-performance liquid chromatography multiple-reaction monitoring mass spectrometry (HPLC-MRMMS) with prior enzymatic deconjugation of glucuronides and sulfates, as described previously.23 In total, 9 phenolic compounds were quantiﬁed by means of 10-point calibration curves using external standards. They included (−)-catechin (C), (−)-epicatechin (EC), (−)-gallocatechin (GC), (−)-epigallocatechin (EGC), (−)-epicatechin gallate (ECG), (−)-epigallocatechin gallate (EGCG), 3/4-O-methylgallic acid (3/4OMGA), 5-(3′,4′-dihydroxyphenyl)-γ-valerolactone (3,4-diOH-VL), and 5-(3′-methoxy-4′-hydroxyphenyl)-γ-valerolactone (3-MeO-4-OHVL). The concentrations of 3-O-methylgallic acid (3-OMGA) and 4-Omethylgallic acid (4-OMGA) were combined to 3/4-OMGA because they could not be fully separated. Similarly, combined ECG values were reported for ECG and CG. Metabolite Proﬁling. Plasma metabolite proﬁles from 19 subjects were acquired from samples collected at t = 0, 120, and 150 min on four experimental days (dGTE_D01, dGTE7_D07, dGTE_D28, PLA_D28), respectively. Four-phase metabolite proﬁling and quantiﬁcation of catecholamines were performed on human plasma samples at Metanomics Health GmbH, Berlin, Germany. Three types of mass spectrometry analyses were applied. GC-MS (gas chromatography−mass spectrometry; Agilent 6890 GC coupled to an Agilent 5973 MS-System, Agilent, Waldbronn, Germany) and LC-MS/MS (liquid chromatography-MS/ MS; Agilent 1100 HPLC-System (Agilent) coupled to an Applied Biosystems API4000 MS/MS-System (Applied Biosystems, Darmstadt, Germany)) were used for broad proﬁling, as described by van Ravenzwaay et al.29 SPE-LC-MS/MS (solid phase extraction-LC-MS/ MS; Symbiosis Pharma (Spark, Emmen, The Netherlands) coupled to an Applied Biosystems API4000 MS/MS-System (Applied Biosystems, Darmstadt, Germany) was used for the determination of catecholamine concentrations. Two hundred and one metabolites fulﬁlled the quality criteria for relative quantiﬁcation, and absolute quantiﬁcation was
useful in capturing the multifactorial and subtle inﬂuences of complex dietary ingredients on overall metabolism. Therefore, in the present study, we applied GC-MS- and LCMS-based untargeted four-phase metabolite proﬁling and targeted proﬁling of catecholamines to human plasma to compare the metabolic eﬀects (i) between cGTE and dGTE supplementation and (ii) between 1, 7, and 28 days of dGTE ingestion at rest and during exercise.
MATERIALS AND METHODS
Study Design. The study was designed as a double-blind, crossover, counterbalanced study (Figure 1) and has been described previously.26 In brief, 20 healthy, physically active, habitual caﬀeine-consuming, male participants [mean ± SD age, 21 ± 2 years; weight, 75.0 ± 7.0 kg; body mass index (BMI), 23.2 ± 2.2 kg/m2; maximal oxygen consumption (V̇ O2max), 55.4 ± 4.6 mL·mL·kg−1·min−1] completed two 28 day supplementation periods of dGTE and PLA ingestion, separated by a 28 day wash-out period. One subject dropped out of the study. Each supplementation period included three experimental days, namely, on the ﬁrst day (dGTE_D01 or PLA_D01), after 7 days (dGTE_D07 or PLA_D07), and after 28 days (dGTE_D28 or PLA_D28) of supplementation. Before each experimental day, the subjects followed a 24 h controlled diet (consisting of three meals each containing ∼50% carbohydrate, ∼35% fat, and ∼15% protein, total energy intake = 2700 kcal) and were asked to refrain from any physical activity and not consume alcohol- or caﬀeine-based beverages. On each experimental day, a resting blood sample (5 mL) was taken at baseline (t = 0 min) after a 12 h overnight fast. Then the participants consumed two capsules of dGTE or PLA with 200 mL of water, rested for 2 h in a seated position and subsequently completed a 30 min cycle exercise bout at 50% Wmax (55 V̇ O2max). Blood samples were taken before the exercise bout commenced (t = 120 min) and at 30 min during the exercise bout. Participants ingested four capsules per day containing dGTE or PLA for a total of 28 days, respectively. Two capsules were consumed an hour before lunch (or 2 h before the exercise bout on the experimental days), and two additional capsules were consumed an hour before dinner. The dGTE was Sunphenon 90 DCF-T (lot 003191) obtained from Taiyo Europe (Fiderstadt, Germany). Each dGTE capsule contained 156 ± 3 mg of EGCG, 284 ± 6 mg of total catechins, and ∼3 mg of caﬀeine. (The catechin composition has been reported in ref 28.) The total catechin ingestion was 1136 ± 24 mg per day. The PLA capsule contained 273 ± 25 mg of cellulose. 9937
dx.doi.org/10.1021/jf502764r | J. Agric. Food Chem. 2014, 62, 9936−9943
Journal of Agricultural and Food Chemistry
performed for additional 10 metabolites. From a total of 211 metabolites, 184 were known metabolites and 27 were not chemically identiﬁed with suﬃcient certainty (i.e., thus considered in the present study to be unknown analytes). The known metabolites belonged to diﬀerent metabolite (ontology) classes: 22 amino acids; 17 metabolites related to amino acids; 12 carbohydrates and related metabolites; 84 complex lipids, fatty acids, and related metabolites; 16 metabolites related to energy metabolism; 13 hormones, signal substances, and related metabolites; 4 nucleobases and related metabolites; 7 vitamin, cofactors, and related metabolites; and 9 miscellaneous metabolites. Technical reference samples were measured in parallel with the study samples to allow the relative quantiﬁcation of metabolites in the study samples. These technical reference samples were generated by pooling aliquots of plasma from all study samples. Relative quantiﬁcation for each metabolite was obtained by normalizing peak intensity in the study samples to the median peak intensity of the corresponding metabolite in the technical reference samples measured in the same batch. For the metabolite proﬁling by GC-MS and LC-MS/MS, proteins were removed from plasma samples (60 μL) by precipitation. Subsequently, polar and nonpolar plasma fractions were separated for both GC-MS and LC-MS/MS analysis by adding water and a mixture of ethanol and dichloromethane. For GC-MS analyses, the nonpolar fraction was treated with methanol under acidic conditions to yield the fatty acid methyl esters derived from both free fatty acids and hydrolyzed complex lipids. The polar and nonpolar fractions were further derivatized with O-methyl-hydroxyamine hydrochloride (20 mg/mL in pyridine, 50 μL) to convert oxo groups to O-methyloximes and subsequently with a silylating agent (MSTFA, 50 μL) before GC-MS analysis.30 For LC-MS/MS analyses, both fractions were reconstituted in appropriate solvent mixtures. High-performance LC (HPLC) was performed by gradient elution using methanol/water/formic acid on reversed phase separation columns. Mass spectrometric detection technology was applied as described in U.S. patent 7196323, which allows targeted and high-sensitivity “multiple reaction monitoring” proﬁling in parallel to a full screen analysis. For the lipid phase, the broad proﬁling technology determines, for example, fatty acid concentrations after acid/methanol treatment, which is essential for derivatization preceding GC-MS analysis. As a consequence, complex lipids are hydrolyzed to components of the lipid backbone (i.e., glycerol) and fatty acids. Hence, the concentration of a fatty acid determined by this procedure represents the sum of its occurrence in free and in lipid-bound form. Components of the backbone can be recognized by the term (“lipid fraction“) added to the metabolite name. As an example “glycerol, lipid fraction” represents glycerol liberated from complex lipids; in contrast, “glycerol, polar fraction” represents glycerol that had been present originally in the biological sample. The use of “additional” (add) indicates that quantiﬁcation can be aﬀected by the co-occurrence of metabolites exhibiting identical characteristics in the analytical methods. Literature data and/or comparison with alternative methods (e.g., LC-MS/MS, GC-MS) suggest that such metabolites are present at minor concentrations only. Catecholamines and their related metabolites were measured by online SPE-LC-MS/MS, as described by Yamada et al.31 Quantiﬁcation was performed using stable isotope-labeled standards. Data Analysis of Metabolite Proﬁles. The data set included 215 metabolite proﬁles. (One subject was excluded because of protocol violation. Another sample was missing.) Analysis of the four-phase metabolite proﬁles and the catecholamines were based on pool-normalized ratios and on absolute concentrations in nanograms per milliliter, respectively. For the multivariate and univariate statistical analysis, the data were log-transformed to better match normal distribution. Principal component analysis (PCA) was performed using SIMCA P + version 12 software (Umetrics, Umea, Sweden). Data were centered and scaled to unit variance to introduce a common scale for all metabolites independent of their absolute variance. The explorative unsupervised multivariate analysis method PCA was used for the detection of trends, patterns, and groupings among samples and variables. PCA revealed outliers that belonged to samples from one subject and thus were excluded from further analysis.
Single-metabolite analysis of variance (ANOVA) analysis was performed using the R-software package nlme.32,33 For this analysis, additional six samples were excluded, because their plasma catechin concentrations signiﬁcantly deviated from average concentrations. The analysis was performed for in total 211 metabolites including all SQ metabolites from four-phase metabolite proﬁling and all catecholamines quantiﬁed by targeted proﬁling. A mixed-eﬀects ANOVA model was built using the factors treatment (PLA_D28, dGTE_D01, dGTE_D07, dGTE_D28), time (0, 120, 150), study arm (A, B) and subject. All possible secondary and tertiary interactions between ﬁxed factors were evaluated in the model reﬁnement process. Model diagnostics was performed to ensure adequacy of the model structure. To this end, residuals were inspected visually by scatter plots of standardized residuals versus ﬁtted values. Additionally, factors or interactions resulting in a number of signiﬁcantly changed metabolites beneath the expected false-positive rate were excluded. Inclusion of all other factors and interactions ensured that residuals did not correlate with ﬁtted values and showed homogeneous variance. No interaction between the factors time and study arm was considered. Thus, the following model was used (R annotation):
fixed: ∼( treatment + time)2 + study arm random: ∼ 1|subject The t statistics results of the ANOVA models comprised estimates, standard deviations, t values, and p values and q values corrected according to the methods of Benjamini-Hochberg (q value) or Bonferroni (qval.Bonf). The resulting numbers of signiﬁcantly changed metabolites were evaluated by binomial test to account for false positives due to multiple hypotheses testing.
RESULTS Eﬀect of dGTE on Exogenous Metabolites. To check the compliance, the concentrations of several catechins were measured in plasma after deconjugation of glucuronides and sulfates at baseline (t = 0 min) and 2 h after the ingestion of two capsules of dGTE or PLA (t = 120 min) on D01, D07, and D28, respectively. Six individual samples were identiﬁed as outliers, and thus removed, because they did not conﬁrm dGTE exposure in the dGTE condition or indicated dGTE exposure in the PLA condition. The concentrations of the remaining 219 samples are shown as box plots in Figure 2. As expected, the concentrations of EGCG, EGC, EC, and ECG/CG signiﬁcantly increased 2 h after dGTE ingestion. These catechins accumulated in plasma following 7 days of dGTE ingestion, as shown by the increased baseline levels at D07 when compared to D01. However, there was no further accumulation, because the baseline levels after 7 and 28 days dGTE intake were similar. The plasma concentrations of GC, C, and GCG were