Amino Acid and Fatty Acid Levels Affect Hepatic Phosphorus Metabolite Content in Metabolically Healthy Humans

Amino Acid and Fatty Acid Levels Affect Hepatic Phosphorus Metabolite Content in Metabolically... Abstract Objective Hepatic energy metabolism negatively relates to insulin resistance and liver fat content in patients with type 2 diabetes, but its role in metabolically healthy humans is unclear. We hypothesized that intrahepatocellular γ-adenosine triphosphate (γATP) and inorganic phosphate (Pi) concentrations exhibit similar associations with insulin sensitivity in nondiabetic, nonobese volunteers. Design A total of 76 participants underwent a four-point sampling, 75-g oral glucose tolerance test (OGTT), as well as in vivo31P/1H magnetic resonance spectroscopy. In 62 of them, targeted plasma metabolomic profiling was performed. Pearson correlation analyses were performed for the dependent variables γATP and Pi. Results Adjusted for age, sex, and body mass index (BMI), hepatic γATP and Pi related to 2-hour OGTT glucose (r = 0.25 and r = 0.27, both P < 0.05), and Pi further associated with nonesterified fatty acids (NEFAs; r = 0.28, P < 0.05). However, neither γATP nor Pi correlated with several measures of insulin sensitivity. Hepatic γATP correlated with circulating leucine (r = 0.42, P < 0.001) and Pi with C16:1 fatty acids palmitoleic acid and C16:1w5 (r = 0.28 and 0.30, respectively, P < 0.01), as well as with δ-9-desaturase index (r = 0.33, P < 0.05). Only the association of γATP with leucine remained important after correction for multiple testing. Leucine and palmitoleic acid, together with age, sex, and BMI, accounted for 26% and for 15% of the variabilities in γATP and Pi, respectively. Conclusions Specific circulating amino acids and NEFAs, but not measures of insulin sensitivity, partly affect hepatic phosphorus metabolites, suggesting mutual interaction between hepatic energy metabolism and circulating metabolites in nondiabetic humans. Hepatic energy metabolism is altered in insulin-resistant and/or type 2 diabetes (T2D) patients with nonalcoholic fatty liver disease (NAFLD) (1, 2). Patients with T2D present with lower hepatic adenosine triphosphate (ATP), inorganic phosphate (Pi) concentrations, and ATP synthase flux (2, 3). On the other hand, obese individuals without nonalcoholic steatohepatitis exhibit increased hepatic oxidative capacity (1). These conflicting results may result from adaptive upregulation of oxidative phosphorylation in obesity (1), along with enhanced production of reactive oxygen species. Increased reactive oxygen species production, which is also linked to hyperglycemia and hyperlipidemia, can induce mitochondrial damage and thereby, contribute to lower mitochondrial function in T2D and nonalcoholic steatohepatitis (1, 3, 4). In exercising skeletal muscle, ATP production is mainly regulated by training intensity, whereas in the liver, ATP-consuming metabolic processes challenge hepatic mitochondria (5, 6). Several factors, such as body mass index (BMI), glycemia, and hepatic insulin sensitivity (1, 2, 7), affect skeletal muscle and hepatic phosphorus metabolites in metabolic diseases, but there are no data on the factors regulating energy metabolism in nonobese nondiabetic persons. Cellular metabolic processes require substrate supply and signaling, particularly by branched-chain amino acids (BCAAs) (8, 9) and nonesterified fatty acids (NEFAs) (10). Although BCAAs have been suggested to improve energy metabolism (8), BCAAs are also increased in insulin-resistant states, such as obesity and T2D, and can induce insulin resistance (8, 9). In metabolic diseases, impaired BCAA metabolism could lead to accumulation of toxic BCAA metabolites, which would, in turn, impair mitochondrial function and give rise to stress signaling associated with insulin resistance (8). Likewise, chronic increases in NEFA can also induce insulin resistance (4) and have been linked to abnormal mitochondrial function (4). Although BCAA and NEFA likely reflect the risk for metabolic diseases (4, 8), their physiological roles for hepatic phosphorus metabolism remain unclear. We have previously developed noninvasive 31P magnet resonance spectroscopy (MRS) techniques, allowing for exact quantification of hepatic 31P metabolite concentrations in a clinical setting (11, 12). With the use of these methods, we now aimed at identifying metabolic determinants of hepatic γ-ATP (γATP) and Pi under physiological conditions in a cohort of middle-aged metabolically healthy volunteers. Participants and Methods Study population The study was performed as a feasibility study in the context of the national cohort; its protocol is in line with the Declaration of Helsinki (version 2008) and approved by the Ethics Board of the Bavarian Medical Association and of the Heinrich-Heine University Düsseldorf. All participants gave their written, informed consent before inclusion into the study. From the general population with residency in the Düsseldorf region, 496 persons, aged 20 to 70 years, were recruited from July 2011 to December 2012 by advertisement or via local residents’ registration office (Supplemental Fig. S1). Of those, 270 persons without known diabetes mellitus underwent an extended examination, including standardized interview, anthropometry, and blood sampling, after 10 hours fast, as well as a 75-g oral glucose tolerance test (OGTT). Magnetic resonance imaging and MRS were performed in 120 of these individuals. Parts of these data have already been published (11, 13). Clinical examination Anthropometric data were measured, as reported previously (14). OGTT A standardized 75-g OGTT (ACCU-CHECK® Dextro O.G.T.; Roche, Basel, Switzerland) was performed after 10 hours overnight fasting with blood sampling, before and 30, 60, and 120 minutes after start. Dysglycemia was classified according to criteria of the American Diabetes Association. Laboratory measurements Measurements of glucose, alanine aminotransferase, aspartate aminotransferase, γ-glutamyl transpeptidase, high-density lipoprotein-cholesterol, low-density lipoprotein-cholesterol, triglyceride, insulin, and C-peptide were performed, as reported previously (14). Hemoglobin A1c was determined on a VARIANT-IITM Analyzer (Bio-Rad Laboratories, Munich, Germany). Plasma NEFAs were determined using the NEFA-HR kit (Wako Chemicals, Neuss, Germany). Metabolome analyses Sodium-heparinate plasma samples were rapidly frozen and stored at −80°C. Targeted metabolic profiling of 294 blood metabolites was performed with the X MetaDis/DQTM kit at Biocrates Life Sciences (Innsbruck, Austria) (13). Fatty acid C18:0 was excluded from further analyses because of analytical errors. BCAAs were defined by the sum of isoleucine, leucine, and valine and δ-9-desaturase (D9D) index by the ratio of palmitoleic and palmitic acid (13). Magnetic resonance imaging and MRS All examinations were performed on a clinical 3-Tesla whole-body magnet (X-series Achieva; Philips, Best, The Netherlands), equipped with a 16-channel Torso XL phased-array receiver coil for 1H MRS and a 14-cm circular 31P surface coil (transmit-receive coil) using standardized procedures, as described (11). Liver triglyceride concentration [hepatocellular lipid content (HCL)] was assessed by 1H MRS using the single voxel-stimulated echo acquisition mode (14); fat fraction was calculated as percentage of the ratio of lipids/(water + lipids), as described (11). For liver 31P MRS, scout images were taken with the 1H body coil in three orientations, followed by transverse T2-weighted images with multislice two-dimensional spin echo images and respiratory triggering to identify liver position clearly and place the volume of interest (6 × 6 × 6 cm) within the liver, avoiding skeletal muscle. Localized liver spectra were obtained using image-selected spectroscopy, with specifications, and analyzed as described before (11). Measures of insulin sensitivity From glucose, insulin, and C-peptide values, obtained during the OGTT, the following indices have been calculated for insulin sensitivity: oral glucose insulin-sensitivity index (OGIS), representing total glucose clearance or whole-body dynamic insulin sensitivity (15), and quantitative insulin-sensitivity check index (QUICKI) to assess fasting (hepatic) insulin sensitivity, calculated as 1/[log(G0) + log(I0)], where G0 and I0 are fasting glucose and insulin, respectively (16). The hepatic insulin resistance index (17) was calculated as the following: glucose area under the curve (AUC)Gluc0-30 (g/dL) × AUCI0-30 (U/dL) × min2. Liver insulin-resistance index (relationship between insulin sensitivity and cardiovascular disease) (18) was calculated as the following: −0.091 + [log AUCI0-120 (pmol/L) × 0.400] + [log fat mass (%) × 0.346] – [log high-density lipoprotein-cholesterol (mg/dL) × 0.408] + [log BMI (kg/m2) × 0.435]. The adipose tissue insulin resistance (Adipo-IR) index was calculated as the following: fasting NEFA (mmol/L) × I0 (pmol/L) (19). De novo lipogenesis (DNL) was calculated as reported previously (13). Statistical analyses Normally distributed parameters are presented as means ± standard deviation and skewed distributed parameters as median (interquartile range). Pearson correlation analyses were performed for the dependent variables γATP and Pi. To account for potential confounders, such as age, sex, and BMI, partial (adjusted) Pearson correlations were calculated also. To improve normality, skewed distributed data were natural log transformed before analyses, as indicated (see Tables 1, 2, and 3 ). P < 0.05 was considered to indicate statistically significant differences. The main analyses (associations between hepatic γATP or Pi with markers of insulin resistance/amino acids/fatty acids) were adjusted for multiple testing using the Bonferroni method, as indicated in the respective table footnotes. Missing data were excluded from the analyses. No imputation methods were performed. Analyses were done using SAS Version 9.4 (SAS Institute, Cary, NC). Table 1. Correlation of Hepatic Phosphorus Metabolites With Clinical Parameters and Insulin Sensitivity Indices   Unadjusted   Adjusted for Age, Sex, and BMI   Adjusted for Age, Sex, BMI, and Pi or γATP     γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L    Pearson Correlation Coefficient  BMI, kg/m2  0.11  0.07          HCL, % (ln)  0.10  0.04  0.08  0.03  0.07  −0.007  NEFA, µmol/L (ln)  0.19  0.28a  0.18  0.28a  0.08  0.22  TG, mg/dL (ln)  0.10  0.03  0.07  0.009  0.07  −0.02  Fasting glucose, mg/dL  0.07  0.05  0.08  0.05  0.06  0.03  2-h Glucose, mg/dL  0.33b  0.33b  0.25a  0.27a  0.15  0.19  Fasting insulin, mU/L (ln)  −0.13  −0.12  −0.17  −0.14  −0.12  −0.08  HbA1c, %  0.19  0.06  0.12  −0.007  0.13  −0.06  OGIS, au  −0.10  −0.08  −0.002  −0.02  0.007  −0.02  QUICKI (ln)  0.11  0.09  0.15  0.12  0.11  0.06  Hepatic-IR index  −0.04  −0.12  −0.08  −0.14  −0.02  −0.12  RISC  0.11  −0.03  0.03  −0.06  0.05  −0.07  Adipo-IR index  0.03  0.08  0.003  0.06  −0.02  0.06  DNL index  −0.13  −0.008  −0.09  0.06  −0.13  0.11    Unadjusted   Adjusted for Age, Sex, and BMI   Adjusted for Age, Sex, BMI, and Pi or γATP     γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L    Pearson Correlation Coefficient  BMI, kg/m2  0.11  0.07          HCL, % (ln)  0.10  0.04  0.08  0.03  0.07  −0.007  NEFA, µmol/L (ln)  0.19  0.28a  0.18  0.28a  0.08  0.22  TG, mg/dL (ln)  0.10  0.03  0.07  0.009  0.07  −0.02  Fasting glucose, mg/dL  0.07  0.05  0.08  0.05  0.06  0.03  2-h Glucose, mg/dL  0.33b  0.33b  0.25a  0.27a  0.15  0.19  Fasting insulin, mU/L (ln)  −0.13  −0.12  −0.17  −0.14  −0.12  −0.08  HbA1c, %  0.19  0.06  0.12  −0.007  0.13  −0.06  OGIS, au  −0.10  −0.08  −0.002  −0.02  0.007  −0.02  QUICKI (ln)  0.11  0.09  0.15  0.12  0.11  0.06  Hepatic-IR index  −0.04  −0.12  −0.08  −0.14  −0.02  −0.12  RISC  0.11  −0.03  0.03  −0.06  0.05  −0.07  Adipo-IR index  0.03  0.08  0.003  0.06  −0.02  0.06  DNL index  −0.13  −0.008  −0.09  0.06  −0.13  0.11  Pearson and partial Pearson correlations of γATP and Pi with the respective variable are given. The Bonferroni-adjusted significance level is 0.0018 [0.05/(14 metabolic parameters × 2 hepatic phosphorus metabolites)]. After Bonferroni adjustment for multiple comparisons, none of the observed correlations remained statistically significant. Abbreviations: HbA1c, hemoglobin A1c; ln, natural log transformed; RISC, relationship between insulin sensitivity and cardiovascular disease; TG, triglyceride. a P < 0.05. b P < 0.01. View Large Table 2. Correlation of Hepatic Phosphorus Metabolites With Specific Amino Acids   Unadjusted   Adjusted for Age, Sex, and BMI   Adjusted for Age, Sex, BMI, and Pi or γATP     γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L    Pearson Correlation Coefficient  Alanine (Ala; ln)  0.16  0.04  0.08  −0.002  0.08  −0.03  Arginine (Arg; ln)  0.03  0.19  0.01  0.19  −0.06  0.20  Citrulline (Cit; ln)  0.16  −0.007  0.12  −0.08  0.16  −0.14  Glycine (Gly; ln)  0.14  −0.03  0.08  −0.10  0.13  −0.14  Histidine (His; ln)  −0.002  −0.06  0.05  −0.02  0.07  −0.05  Isoleucine (Ile; ln)  0.19  −0.09  0.32a  −0.005  0.35b  −0.14  Leucine (Leu; ln)  0.28a  0.06  0.42c  0.16  0.40b  −0.004  Lysine (Lys; ln)  0.22  0.13  0.15  0.07  0.13  0.01  Methionine (Met; ln)  0.05  −0.09  0.18  0.006  0.19  −0.07  Ornithine (Orn; ln)  0.35b  0.14  0.28a  0.05  0.28a  −0.06  Phenylalanine (Phe; ln)  0.23  −0.02  0.24  −0.02  0.26a  −0.12  Proline (Pro; ln)  −0.01  −0.08  0.06  −0.02  0.07  −0.04  Serine (Ser; ln)  0.15  −0.04  0.12  −0.05  0.15  −0.10  Threonine (Thr; ln)  −0.22  −0.28a  −0.17  −0.23  −0.10  −0.18  Tryptophane (Trp; ln)  0.23  −0.004  0.28a  0.05  0.29a  −0.07  Tyrosine (Tyr; ln)  0.25  −0.07  0.21  −0.10  0.26a  −0.20  Valine (Val; ln)  0.15  0.005  0.25  0.06  0.24  −0.03  BCAA (Ile/Leu/Val; ln)  0.22  −0.007  0.37b  0.09  0.37b  −0.06    Unadjusted   Adjusted for Age, Sex, and BMI   Adjusted for Age, Sex, BMI, and Pi or γATP     γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L    Pearson Correlation Coefficient  Alanine (Ala; ln)  0.16  0.04  0.08  −0.002  0.08  −0.03  Arginine (Arg; ln)  0.03  0.19  0.01  0.19  −0.06  0.20  Citrulline (Cit; ln)  0.16  −0.007  0.12  −0.08  0.16  −0.14  Glycine (Gly; ln)  0.14  −0.03  0.08  −0.10  0.13  −0.14  Histidine (His; ln)  −0.002  −0.06  0.05  −0.02  0.07  −0.05  Isoleucine (Ile; ln)  0.19  −0.09  0.32a  −0.005  0.35b  −0.14  Leucine (Leu; ln)  0.28a  0.06  0.42c  0.16  0.40b  −0.004  Lysine (Lys; ln)  0.22  0.13  0.15  0.07  0.13  0.01  Methionine (Met; ln)  0.05  −0.09  0.18  0.006  0.19  −0.07  Ornithine (Orn; ln)  0.35b  0.14  0.28a  0.05  0.28a  −0.06  Phenylalanine (Phe; ln)  0.23  −0.02  0.24  −0.02  0.26a  −0.12  Proline (Pro; ln)  −0.01  −0.08  0.06  −0.02  0.07  −0.04  Serine (Ser; ln)  0.15  −0.04  0.12  −0.05  0.15  −0.10  Threonine (Thr; ln)  −0.22  −0.28a  −0.17  −0.23  −0.10  −0.18  Tryptophane (Trp; ln)  0.23  −0.004  0.28a  0.05  0.29a  −0.07  Tyrosine (Tyr; ln)  0.25  −0.07  0.21  −0.10  0.26a  −0.20  Valine (Val; ln)  0.15  0.005  0.25  0.06  0.24  −0.03  BCAA (Ile/Leu/Val; ln)  0.22  −0.007  0.37b  0.09  0.37b  −0.06  Pearson and partial Pearson correlations of γATP and Pi with specific amino acids are given, respectively. The Bonferroni-adjusted significance level is 0.0014 [0.05/(18 amino acid parameters × 2 hepatic phosphorus metabolites)]. After Bonferroni adjustment for multiple comparisons, only the association of γATP and leucine remained statistically significant. Abbreviation: ln, natural log transformed. a P < 0.05. b P < 0.01. c P < 0.001. View Large Table 3. Correlation of Hepatic Phosphorus Metabolites With Free Fatty Acid Species   Unadjusted   Adjusted for Age, Sex, and BMI   Adjusted for Age, Sex, BMI, and Pi or γATP     γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L    Pearson Correlation Coefficient  SFA               C14:0 (myristic; ln)  0.02  0.13  −0.05  0.09  −0.09  0.11   C15:0 (pentadecanoic; ln)  −0.13  0.07  −0.13  0.08  −0.17  0.14   C16:0 (palmitic; ln)  −0.06  0.09  −0.19  0.02  −0.21  0.10   C17:0 (heptadecanoic; ln)  0.0007  0.04  −0.08  −0.03  −0.08  0.004   C18:0 (stearic; ln)               C19:0 (nonadecanoic; ln)  0.02  0.02  −0.16  −0.14  −0.12  −0.09   C20:0 (eicosanoic; ln)  −0.24  −0.18  −0.22  −0.16  −0.17  −0.09  MUFA               cC14:1w5 (myristoleic; ln)  0.13  0.28a  0.08  0.28a  −0.02  0.27a   cC15:1w5, c10 (pentadecenoic; ln)  −0.14  0.10  −0.19  0.07  −0.23  0.15   cC16:1w5 (ln)  −0.04  0.32a  −0.14  0.30a  −0.28a  0.38b   cC16:1w7 (palmitoleic; ln)  0.09  0.33b  −0.03  0.28a  −0.15  0.31a   cC16:1w10 (sapienic; ln)  0.01  0.16  −0.13  0.10  −0.18  0.16   cC17:1w7, c10 (heptadecenoic; ln)  −0.16  −0.004  −0.23  −0.07  −0.22  0.02   cC17:1w8 (ln)  0.10  0.21  −0.03  0.11  −0.08  0.14   cC18:1w7 (vaccenic; ln)  −0.01  0.20  −0.17  0.10  −0.22  0.18   cC18:1w9 (oleic; ln)  0.01  0.12  −0.13  0.02  −0.14  0.07   cC20:1w9, c11 (eicosenoic; ln)  −0.02  0.13  −0.12  0.05  −0.15  0.10  PUFA               cC18:2w6 (linoleic; ln)  0.05  0.08  −0.07  −0.04  −0.06  −0.01   cC18:3w3 (linolenic; ln)  0.03  0.13  −0.12  0.02  −0.13  0.06   cC18:3w6 (gamma linolenic; ln)  0.09  0.18  −0.005  0.13  −0.06  0.15   cC18:4w3 (stearidonic; ln)  0.03  0.20  −0.05  0.14  −0.11  0.17   cC20:2w6, c11, 14 (eicosadienoic; ln)  −0.02  0.16  −0.12  0.08  −0.17  0.14   cC20:3w6, c8, 11, 14 (eicosatrienoic; ln)  −0.05  −0.01  −0.05  0.02  −0.06  0.04   cC20:3w9 (mead; ln)  0.06  0.04  0.03  0.003  0.03  −0.009   cC20:4w6 (arachidonic; ln)  −0.17  0.04  −0.18  0.05  −0.22  0.13   cC20:5w3 (EPA; ln)  0.004  0.07  −0.13  −0.05  −0.12  0.003   cC22:4w6 (adrenic; ln)  −0.15  0.26a  −0.25  0.22  −0.37b  0.35b   cC22:5w3 (DPA; ln)  0.04  0.24  −0.12  0.13  −0.19  0.19   cC22:6w3 (DHA; ln)  0.03  0.14  −0.12  0.05  −0.14  0.10   Q_cC16_1__cC16_0  0.15  0.37b  0.08  0.33a  −0.05  0.33a    Unadjusted   Adjusted for Age, Sex, and BMI   Adjusted for Age, Sex, BMI, and Pi or γATP     γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L    Pearson Correlation Coefficient  SFA               C14:0 (myristic; ln)  0.02  0.13  −0.05  0.09  −0.09  0.11   C15:0 (pentadecanoic; ln)  −0.13  0.07  −0.13  0.08  −0.17  0.14   C16:0 (palmitic; ln)  −0.06  0.09  −0.19  0.02  −0.21  0.10   C17:0 (heptadecanoic; ln)  0.0007  0.04  −0.08  −0.03  −0.08  0.004   C18:0 (stearic; ln)               C19:0 (nonadecanoic; ln)  0.02  0.02  −0.16  −0.14  −0.12  −0.09   C20:0 (eicosanoic; ln)  −0.24  −0.18  −0.22  −0.16  −0.17  −0.09  MUFA               cC14:1w5 (myristoleic; ln)  0.13  0.28a  0.08  0.28a  −0.02  0.27a   cC15:1w5, c10 (pentadecenoic; ln)  −0.14  0.10  −0.19  0.07  −0.23  0.15   cC16:1w5 (ln)  −0.04  0.32a  −0.14  0.30a  −0.28a  0.38b   cC16:1w7 (palmitoleic; ln)  0.09  0.33b  −0.03  0.28a  −0.15  0.31a   cC16:1w10 (sapienic; ln)  0.01  0.16  −0.13  0.10  −0.18  0.16   cC17:1w7, c10 (heptadecenoic; ln)  −0.16  −0.004  −0.23  −0.07  −0.22  0.02   cC17:1w8 (ln)  0.10  0.21  −0.03  0.11  −0.08  0.14   cC18:1w7 (vaccenic; ln)  −0.01  0.20  −0.17  0.10  −0.22  0.18   cC18:1w9 (oleic; ln)  0.01  0.12  −0.13  0.02  −0.14  0.07   cC20:1w9, c11 (eicosenoic; ln)  −0.02  0.13  −0.12  0.05  −0.15  0.10  PUFA               cC18:2w6 (linoleic; ln)  0.05  0.08  −0.07  −0.04  −0.06  −0.01   cC18:3w3 (linolenic; ln)  0.03  0.13  −0.12  0.02  −0.13  0.06   cC18:3w6 (gamma linolenic; ln)  0.09  0.18  −0.005  0.13  −0.06  0.15   cC18:4w3 (stearidonic; ln)  0.03  0.20  −0.05  0.14  −0.11  0.17   cC20:2w6, c11, 14 (eicosadienoic; ln)  −0.02  0.16  −0.12  0.08  −0.17  0.14   cC20:3w6, c8, 11, 14 (eicosatrienoic; ln)  −0.05  −0.01  −0.05  0.02  −0.06  0.04   cC20:3w9 (mead; ln)  0.06  0.04  0.03  0.003  0.03  −0.009   cC20:4w6 (arachidonic; ln)  −0.17  0.04  −0.18  0.05  −0.22  0.13   cC20:5w3 (EPA; ln)  0.004  0.07  −0.13  −0.05  −0.12  0.003   cC22:4w6 (adrenic; ln)  −0.15  0.26a  −0.25  0.22  −0.37b  0.35b   cC22:5w3 (DPA; ln)  0.04  0.24  −0.12  0.13  −0.19  0.19   cC22:6w3 (DHA; ln)  0.03  0.14  −0.12  0.05  −0.14  0.10   Q_cC16_1__cC16_0  0.15  0.37b  0.08  0.33a  −0.05  0.33a  Pearson and partial Pearson correlations of γATP and Pi with specific amino acids are given, respectively. The Bonferroni-adjusted significance level is 0.00083 [0.05/(30 fatty acid parameters × 2 hepatic phosphorus metabolites)]. After Bonferroni adjustment for multiple comparisons, only the association of Pi and palmitoleic acid remained statistically significant. Abbreviations: DHA, docosahexaenoic acid; DPA, docosapentaenoic acid; EPA, eicosapentaenoic acid; ln, natural log transformed; MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids; SFA, saturated fatty acids. a P < 0.05. b P < 0.01. View Large Results Study population Of 120 volunteers undergoing OGTT and MRS examinations, 76 entered the final analysis after exclusion of participants with missing measures and/or metabolic disease (Supplemental Fig. S1). Participants (26 men, 50 women) were middle aged (54 ± 13 years), nonobese (BMI 24.8 ± 3.2 kg/m2), insulin sensitive (OGIS 461.2 ± 61.2 arbitrary units (au), QUICKI 0.46 [0.43;0.50]), and had low HCL (1.00 [0.41;2.76]%) and transaminases (alanine aminotransferase 18 [15;24] U/L, aspartate aminotransferase 23 [20;27] U/L). Fasting and 2-hour OGTT glucose were 4.2 ± 0.5 and 5.0 ± 1.3 mmol/L, respectively. Fasting NEFA and insulin were 516 [404;714] and 42.6 [30.3;62.0] pmol/L, respectively. Hepatic γATP, Pi concentrations, and the Pi/γATP ratio were 2.73 ± 0.54, 2.02 ± 0.49, and 0.75 ± 0.18 mmol/L, respectively. Absolute concentrations of the measured amino acids and lipid metabolites are shown in Supplemental Tables S1 and S2. Correlation analyses for hepatic phosphorus metabolites with glycemia and insulin sensitivity Pearson correlation analyses revealed positive relationships of hepatic γATP concentrations with 2-hour OGTT glucose and of Pi with 2-hour OGTT glucose and NEFA, respectively (Fig. 1A–1D; Table 1). After adjustment for age, sex, and BMI, 2-hour OGTT glucose still related to both γATP and Pi content and NEFA levels to Pi concentrations. The adjustment of the analyses for Pi or γATP abolished these associations, pointing to the impact of the concentration of the respective other phosphorus metabolite on these relationships (Table 1). No associations with γATP or Pi content were found for any calculated measure of insulin sensitivity in unadjusted and adjusted analyses (Table 1). No association remained statistically significant after correction for multiple testing. Figure 1. View largeDownload slide Correlations of γATP and Pi with plasma metabolites. (A, C, and E) Correlations of γATP with 2-hour OGTT glucose, NEFA, and leucine, respectively. (B, D, and F) Correlations of Pi with 2-hour OGTT glucose, NEFA, and palmitoleic acid, respectively. Before correlation analyses, NEFA and metabolites were log transformed to achieve normal distribution. Metabolites were adjusted further for plate of measurement (shown as au). Figure 1. View largeDownload slide Correlations of γATP and Pi with plasma metabolites. (A, C, and E) Correlations of γATP with 2-hour OGTT glucose, NEFA, and leucine, respectively. (B, D, and F) Correlations of Pi with 2-hour OGTT glucose, NEFA, and palmitoleic acid, respectively. Before correlation analyses, NEFA and metabolites were log transformed to achieve normal distribution. Metabolites were adjusted further for plate of measurement (shown as au). Correlation analyses for hepatic phosphorus metabolites with amino acids Hepatic γATP was positively associated with plasma leucine (Fig. 1E) and ornithine. Hepatic Pi related negatively to threonine (Table 2). After adjustment for age, sex, and BMI, γATP still positively correlated with leucine and ornithine. Moreover, γATP associated with isoleucine, tryptophan, and BCAAs. The association of Pi with threonine disappeared after adjustment for age, sex, and BMI (Table 2). After additional adjustment for Pi, the correlation of γATP with isoleucine, leucine, ornithine, tryptophane, and BCAA remained, whereas additional correlations of γATP with tyrosine and phenylalanine were found (Table 2). After correction for multiple testing, only the association between γATP and leucine remained statistically significant (Table 2). Correlation analyses for hepatic phosphorus metabolites with fatty acids species Neither γATP nor Pi related to any analyzed saturated fatty acid species. Furthermore, γATP did not associate with any unsaturated fatty acid species (Table 3). Pi related positively to myristoleic acid (cC14:1w5), cC16:1w5, palmitoleic acid (cC16:1w7; Fig. 1F), adrenic acid (cC22:4w6), and the D9D index. After adjustment for age, sex, and BMI, only the relationships of Pi with myristoleic acid, cC16:1w5, palmitoleic acid, and D9D index remained (Table 3). Further adjustment for Pi revealed negative relationships of γATP with cC16:1w5 and adrenic acid. After adjustment for γATP, Pi related positively to myristoleic acid, cC16:1w5, palmitoleic acid, adrenic acid, and D9D index (Table 3). When corrected for multiple testing, none of the associations remained statistically significant (Table 3). In multivariate regression models, age, sex, and BMI explained 9.8% of γATP and 7.8% of Pi variability, respectively. Additional adjustment for leucine, as a significant independent predictor, explained 26% of γATP variances, and the adjustment for age, sex, BMI, and palmitoleic acid accounted for 15% of Pi variability. A multivariate regression analysis containing variables age, sex, BMI, palmitoleic acid, and cC16:1w5 did not improve R2 (0.16). Predictors of hepatic phosphorus metabolite concentrations These analyses identified leucine and C16:1 fatty acid palmitoleic acid as predictors of γATP and Pi content, respectively. We further confirmed that additional adjustment for selected parameters, indicative of insulin resistance and glucose intolerance, did not affect these relationships (Supplemental Table S3). The close association of γATP and leucine was not affected by adjustment for any of the chosen parameters. However, the relation of Pi and palmitoleic acid disappeared by adjustment for 2-hour OGTT glucose and NEFA levels, respectively. Discussion This study found that specific essential amino acids and fatty acid species are independent predictors of hepatic γATP and Pi content in metabolically healthy humans, whereas measures of insulin sensitivity neither related to hepatic γATP nor to Pi content. This cohort had low HCL, normal glucose tolerance, and insulin sensitivity, based on the OGTT analysis (15). Moreover, absolute concentrations of hepatic γATP and Pi were similar to nondiabetic cohorts with comparable baseline characteristics (2, 12). Thus, this cohort represents an appropriate group for investigation of hepatic energy status under physiological conditions. Absolute hepatic concentrations of Pi and ATP have been reported to be positively associated with hepatic but not with peripheral insulin sensitivity and negatively with HCL in a mixed collective of patients with T2D and nondiabetic humans (2). Another study reported a negative correlation of ATP levels with BMI (20). Of note, our results do not confirm these associations in a collective of nondiabetic humans, as we did not observe any association of γATP and Pi with indices of fasting (QUICKI), dynamic peripheral (OGIS), or hepatic (Hepatic-IR index, relationship between insulin sensitivity and cardiovascular disease) and adipose tissue (Adipo-IR index) insulin sensitivity, HCL, or BMI, respectively. This might result from the fact that our collective excluded participants with hyperglycemia or hepatic steatosis. Furthermore, we used fasting and OGTT-derived measures of insulin sensitivity, but OGIS and M-values calculated from OGTT and hyperinsulinemic-euglycemic clamps, respectively, are closely correlated (7, 15). Unexpectedly, we observed a positive relationship of both γATP and Pi with 2-hour OGTT glucose levels, which may serve as a rough predictor of insulin resistance (21). However, factors other than insulin resistance, such as duration of fasting and the amount of carbohydrate intake on the previous days, may influence OGTT glucose levels (22, 23). In T2D, the negative relationship between hepatic γATP and ATP synthase flux and insulin resistance was determined, at least in part, by fasting plasma glucose and hemoglobin A1c, respectively (2, 3). Thus, the observed positive relationships of hepatic γATP and Pi with 2-hour OGTT glucose observed in this nonobese nondiabetic cohort are unlikely a result of insulin resistance. Of note, nondiabetic but obese persons even show greater hepatic oxidative capacity, despite peripheral insulin resistance (1). We found a positive relation of Pi with circulating NEFA. In patients with T2D, no correlation was found for ATP flux and NEFA (3). Previous studies reported no association of γATP and plasma NEFA levels in metabolically healthy and diseased humans (12, 24), whereas Pi associated negatively with NEFA levels in a combined collective of healthy, obese, and T2D volunteers (12). Moreover, Pi did not correlate with the Adipo-IR index, suggesting that Adipo-IR cannot exclusively explain the relationship of Pi and NEFA. However, a discrete effect cannot be resolved by our study as a result of the limited sample size. Thus, in healthy humans with intact metabolic flexibility, higher substrate availability of either glucose or NEFA may charge hepatic phosphorus pools and/or stimulate energy-generating processes (25). Of note, further adjustment for γATP or Pi, respectively, abolished the previous associations, indicating that the relationship of 2-hour OGTT glucose with γATP and Pi, as well as of NEFA concentrations with Pi, also depends on the prevailing Pi and γATP levels, respectively. Absolute fasting γATP and Pi levels, as well as Pi/ATP ratios of our study, differed, in part, from those of the respective studies, which might also explain the observed differences in the correlation analyses (2, 3, 12). However, none of the associations of γATP or Pi with 2-hour OGTT glucose and of NEFA concentrations with Pi was statistically significant when corrected for multiple testing. Interestingly, this study found a positive association of γATP with some essential amino acids and of Pi with some fatty acid species, respectively. People with insulin resistance or T2D typically present with higher levels of combined BCAA (leucine, isoleucine, valine) and/or other amino acids (phenylalanine, tryptophane, tyrosine, alanine, citrulline) and lower levels of threonine, glycine, and glutamine (9, 13). In our metabolically healthy cohort, γATP related strongest to leucine, independent of changes in Pi content. Leucine and isoleucine were reported to be elevated in obese compared with lean humans (9). In our cohort, γATP still related closely to leucine even after adjustment for age, sex, and BMI. Further investigation of parameters affecting the relation of γATP and leucine showed no effect on this interaction. In obese mice, leucine increases ATP concentrations in brown adipose tissue as a result of stimulation of mitochondrial biogenesis (26), and in skeletal muscle cells, leucine stimulates fatty acid oxidation, as well as oxygen consumption (27). Moreover, leucine provides carbon skeletons to the tricarboxylic acid (TCA) cycle at the level of acetyl-coenzyme A, which may enhance TCA cycle flux (28). Thus, higher plasma leucine levels likely contribute to improved mitochondrial performance rather than reflect insulin resistance or impaired mitochondrial function in our collective of healthy humans. This study observed relevant correlations of Pi with unsaturated fatty acid species, such as myristoleic acid, C16:1w5, palmitoleic acid, and adrenic acid. The absolute concentrations of these fatty acids were rather low, as previously reported (29). Whereas our assay does not distinguish between cis and trans fatty acids, the low contribution of trans fatty acids to the total circulating fatty acid pool, amounting to <2% (30), suggests that the observed correlations are mainly driven by the respective cis isomers. Myristoleic acid usually accounts for only small amounts of total fatty acids in animal tissues but is abundant in milk fat and may also raise palmitoleic acid levels (31). It is therefore conceivable that the circulating palmitoleic and myristoleic acid levels, at least partly, result from dietary intake. We did not detect a correlation of the DNL index and γATP or Pi, respectively. The main fatty acid products of DNL include palmitic acid, palmitoleic acid, vaccenic acid, stearic acid, and oleic acid (32). Except for palmitoleic acid, we did not find a correlation of those fatty acids with Pi, further suggesting that the relationship of Pi and palmitoleic acid does not result from DNL. Nevertheless, palmitoleic acid per se and the D9D index have been proposed as markers of DNL (33). The strong interaction between Pi and palmitoleic acid disappeared upon adjustments for fasting NEFA and 2-hour OGTT glucose levels. Fasting NEFA levels mainly reflect adipose tissue lipolysis, which also is tightly regulated by the glucose-regulating hormones insulin and catecholamines (34). This suggests that both NEFA supply to the liver and regulatory hormones contribute to this interaction. Palmitoleic acid has originally gained attention as an anti-inflammatory adipokine (35). In a murine model, chronic palmitoleic acid feeding enhances hepatic glucose uptake and fatty acid oxidation for energy production instead of storage through activation of adenosine 5′-monophosphate-activated protein kinase and fibroblast growth factor 21 -21, dependent on peroxisome proliferator-activated receptor α (10). Of note, cis monounsaturated fatty acids also strongly inhibit Ca2+ uptake, thereby inducing net Ca2+ efflux and higher extramitochondrial Ca2+ concentrations (36). Effects of palmitoleic acid and its ester on mitochondrial metabolism have also been reported in other tissues (37). Taken together, these relationships of palmitoleic and myristoleic acids with hepatic Pi might result directly or indirectly from their interaction with mitochondrial Ca2+ handling and reduction in the mitochondrial membrane potential. As to possible hepatic effects of the other unsaturated fatty acids, adrenic acid may act as an inflammation enhancer in NAFLD (38), whereas there is currently no information on the physiological relevance of C16:1w5. The strength of this study is the investigation of hepatic energy metabolism in a larger cohort of well-phenotyped, glucose-tolerant humans. Up until now, hepatic ATP and Pi content have mainly been studied in the context of metabolic diseases, such as T2D or NAFLD (2, 3, 6). This approach might help to differentiate among factors regulating ATP and Pi in nondiabetic, nonobese, non-NAFLD persons and in those with metabolic disease. This study also has some limitations. The main weakness resides in the inclusion of only healthy, nonobese individuals. Nevertheless, relationships of amino and fatty acid species with hepatic phosphorus metabolites have, to our knowledge, not been reported before, and this study provides that information. Although measurement of absolute hepatic concentrations of ATP and Pi is reliable and highly reproducible, it covers only one aspect of hepatic energy metabolism—the flux through hepatic ATP synthase (3)—but does not address other features of mitochondrial function, such as TCA cycle activity and oxidative capacity, which may be differently affected in health and disease (1, 39, 40). Furthermore, the observed effect of leucine and palmitoleic acid on ATP and Pi levels is only modest, implying that other factors also regulate hepatic energy metabolism in healthy humans. Such factors might be less relevant in the presence of hepatic steatosis, insulin resistance, or hepatocellular injury (2, 6). Our data, from the examined cohort, suggest that hepatic concentrations of ATP and Pi relate to specific circulating amino acids and fatty acid species, but analyses could be considered exploratory as a result of the size of our cohort and the lack of statistical significance after Bonferroni correction for many associations. However, the strongest correlation found for hepatic γATP levels with leucine was still present after correction for multiple testing. In conclusion, hepatic γATP and Pi do not relate to measures of insulin sensitivity in metabolically healthy nonobese humans. However, specific circulating amino acids and NEFAs partly affect the availability of hepatic phosphorus metabolites. These results point to a mutual interaction between hepatic energy metabolism and circulating metabolites and may therefore contribute to the better understanding of the differences in hepatic energy metabolism between glucose-tolerant and insulin-resistant humans. Abbreviations: Adipo-IR adipose tissue insulin resistance ATP adenosine triphosphate au arbitrary unit(s) AUC area under the curve BCAA branched-chain amino acid BMI body mass index D9D δ-9-desaturase DNL de novo lipogenesis HCL hepatocellular lipid content MRS magnet resonance spectroscopy NAFLD nonalcoholic fatty liver disease NEFA nonesterified fatty acid OGIS oral glucose insulin-sensitivity index OGTT oral glucose tolerance test Pi inorganic phosphate QUICKI quantitative insulin-sensitivity check index T2D type 2 diabetes TCA tricarboxylic acid γATP γ-adenosine triphosphate. Acknowledgments We thank Nicole Achterath, Irena Latta, Andrea Nagel, Birgit Platzbecker, Daniela Seeger, Dominik Scheibelhut, and Rita Schreiner (all from the Institute for Clinical Diabetology, German Diabetes Center, Germany) for excellent technical assistance and care of the participants. The study design of the National Cohort was done by the German National Cohort Epidemiologic Planning Committee and Project Management Team and approved by Dr. Karl-Heinz Jöckel, chairman of the Scientific Board of the German National Cohort. Financial Support: This study was conducted within the pretest studies of the German National Cohort (www.nationale-kohorte.de), which were funded by the Federal Ministry of Education and Research (BMBF), Förderkennzeichen 01ER1001A-I, and supported by the participating universities, institutes of the Leibniz Association. This study was further supported, in part, by the German Diabetes Association (DDG); Schmutzler Stiftung; ICEMED Helmholtz-Alliance and the German Center for Diabetes Research (DZD e.V.); German Federal Ministry of Health (Germany); and Ministry of Innovation, Science, and Research of the State North Rhine-Westphalia (Germany). Author Contributions: M.R. designed the study and headed the clinical experiments. S.K., B.N., B.K., A.B., P.J.N., F.W., J.-H.H., and B.H. researched the data. S.K. analyzed the data and wrote the manuscript. G.P. calculated indices of insulin sensitivity. K.S. and G.G. performed the statistical analyses. All of the authors contributed substantially to aspects of study design or the acquisition of data, contributed to drafting of the article or revised it critically for important intellectual content, and gave final approval to the version to be published. M.R. is responsible for the integrity of the work as a whole. Disclosure Summary: The authors have nothing to disclose. References 1. Koliaki C, Szendroedi J, Kaul K, Jelenik T, Nowotny P, Jankowiak F, Herder C, Carstensen M, Krausch M, Knoefel WT, Schlensak M, Roden M. Adaptation of hepatic mitochondrial function in humans with non-alcoholic fatty liver is lost in steatohepatitis. Cell Metab . 2015; 21( 5): 739– 746. Google Scholar CrossRef Search ADS PubMed  2. Szendroedi J, Chmelik M, Schmid AI, Nowotny P, Brehm A, Krssak M, Moser E, Roden M. Abnormal hepatic energy homeostasis in type 2 diabetes. Hepatology . 2009; 50( 4): 1079– 1086. Google Scholar CrossRef Search ADS PubMed  3. Schmid AI, Szendroedi J, Chmelik M, Krssák M, Moser E, Roden M. Liver ATP synthesis is lower and relates to insulin sensitivity in patients with type 2 diabetes. Diabetes Care . 2011; 34( 2): 448– 453. Google Scholar CrossRef Search ADS PubMed  4. Jelenik T, Roden M. Mitochondrial plasticity in obesity and diabetes mellitus. Antioxid Redox Signal . 2013; 19( 3): 258– 268. Google Scholar CrossRef Search ADS PubMed  5. Kemp GJ, Taylor DJ, Radda GK. Control of phosphocreatine resynthesis during recovery from exercise in human skeletal muscle. NMR Biomed . 1993; 6( 1): 66– 72. Google Scholar CrossRef Search ADS PubMed  6. Cortez-Pinto H, Chatham J, Chacko VP, Arnold C, Rashid A, Diehl AM. Alterations in liver ATP homeostasis in human nonalcoholic steatohepatitis: a pilot study. JAMA . 1999; 282( 17): 1659– 1664. Google Scholar CrossRef Search ADS PubMed  7. Szendroedi J, Kaul K, Kloock L, Straßburger K, Schmid AI, Chmelik M, Kacerovsky M, Kacerovsky-Bielesz G, Prikoszovich T, Brehm A, Krssák M, Gruber S, Krebs M, Kautzky-Willer A, Moser E, Pacini G, Roden M. Lower fasting muscle mitochondrial activity relates to hepatic steatosis in humans. Diabetes Care . 2014; 37( 2): 468– 474. Google Scholar CrossRef Search ADS PubMed  8. Lynch CJ, Adams SH. Branched-chain amino acids in metabolic signalling and insulin resistance. Nat Rev Endocrinol . 2014; 10( 12): 723– 736. Google Scholar CrossRef Search ADS PubMed  9. Newgard CB, An J, Bain JR, Muehlbauer MJ, Stevens RD, Lien LF, Haqq AM, Shah SH, Arlotto M, Slentz CA, Rochon J, Gallup D, Ilkayeva O, Wenner BR, Yancy WS, Jr, Eisenson H, Musante G, Surwit RS, Millington DS, Butler MD, Svetkey LP. A branched-chain amino acid-related metabolic signature that differentiates obese and lean humans and contributes to insulin resistance. Cell Metab . 2009; 9( 4): 311– 326. Google Scholar CrossRef Search ADS PubMed  10. de Souza CO, Teixeira AAS, Biondo LA, Lima Junior EA, Batatinha HAP, Rosa Neto JC. Palmitoleic acid improves metabolic functions in fatty liver by PPARα-dependent AMPK activation. J Cell Physiol . 2017; 232( 8): 2168– 2177. Google Scholar CrossRef Search ADS PubMed  11. Laufs A, Livingstone R, Nowotny B, Nowotny P, Wickrath F, Giani G, Bunke J, Roden M, Hwang JH. Quantitative liver 31P magnetic resonance spectroscopy at 3T on a clinical scanner. Magn Reson Med . 2014; 71( 5): 1670– 1675. Google Scholar CrossRef Search ADS PubMed  12. Fritsch M, Koliaki C, Livingstone R, Phielix E, Bierwagen A, Meisinger M, Jelenik T, Strassburger K, Zimmermann S, Brockmann K, Wolff C, Hwang JH, Szendroedi J, Roden M. Time course of postprandial hepatic phosphorus metabolites in lean, obese, and type 2 diabetes patients. Am J Clin Nutr . 2015; 102( 5): 1051– 1058. Google Scholar CrossRef Search ADS PubMed  13. Knebel B, Strassburger K, Szendroedi J, Kotzka J, Scheer M, Nowotny B, Müssig K, Lehr S, Pacini G, Finner H, Klüppelholz B, Giani G, Al-Hasani H, Roden M; German Diabetes Study Group. Specific metabolic profiles and their relationship to insulin resistance in recent-onset type 1 and type 2 diabetes. J Clin Endocrinol Metab . 2016; 101( 5): 2130– 2140. Google Scholar CrossRef Search ADS PubMed  14. Kahl S, Straßburger K, Nowotny B, Livingstone R, Klüppelholz B, Keßel K, Hwang JH, Giani G, Hoffmann B, Pacini G, Gastaldelli A, Roden M. Comparison of liver fat indices for the diagnosis of hepatic steatosis and insulin resistance. PLoS One . 2014; 9( 4): e94059. Google Scholar CrossRef Search ADS PubMed  15. Mari A, Pacini G, Murphy E, Ludvik B, Nolan JJ. A model-based method for assessing insulin sensitivity from the oral glucose tolerance test. Diabetes Care . 2001; 24( 3): 539– 548. Google Scholar CrossRef Search ADS PubMed  16. Katz A, Nambi SS, Mather K, Baron AD, Follmann DA, Sullivan G, Quon MJ. Quantitative insulin sensitivity check index: a simple, accurate method for assessing insulin sensitivity in humans. J Clin Endocrinol Metab . 2000; 85( 7): 2402– 2410. Google Scholar CrossRef Search ADS PubMed  17. Abdul-Ghani MA, Matsuda M, Balas B, DeFronzo RA. Muscle and liver insulin resistance indexes derived from the oral glucose tolerance test. Diabetes Care . 2007; 30( 1): 89– 94. Google Scholar CrossRef Search ADS PubMed  18. Vangipurapu J, Stančáková A, Kuulasmaa T, Paananen J, Kuusisto J, Ferrannini E, Laakso M; EGIR-RISC Study Group. A novel surrogate index for hepatic insulin resistance. Diabetologia . 2011; 54( 3): 540– 543. Google Scholar CrossRef Search ADS PubMed  19. Gastaldelli A, Harrison SA, Belfort-Aguilar R, Hardies LJ, Balas B, Schenker S, Cusi K. Importance of changes in adipose tissue insulin resistance to histological response during thiazolidinedione treatment of patients with nonalcoholic steatohepatitis. Hepatology . 2009; 50( 4): 1087– 1093. Google Scholar CrossRef Search ADS PubMed  20. Nair S, P Chacko V, Arnold C, Diehl AM. Hepatic ATP reserve and efficiency of replenishing: comparison between obese and nonobese normal individuals. Am J Gastroenterol . 2003; 98( 2): 466– 470. Google Scholar PubMed  21. Saxena P, Prakash A, Nigam A. Efficacy of 2-hour post glucose insulin levels in predicting insulin resistance in polycystic ovarian syndrome with infertility. J Hum Reprod Sci . 2011; 4( 1): 20– 22. Google Scholar CrossRef Search ADS PubMed  22. Wang P, Lu MC, Yu CW, Yan YH. Influence of the time of day and fasting duration on glucose level following a 1-hour, 50-gram glucose challenge test in pregnant women. PLoS One . 2014; 9( 11): e112526. Google Scholar CrossRef Search ADS PubMed  23. Szathmary EJ. The impact of low carbohydrate consumption on glucose tolerance, insulin concentration and insulin response to glucose challenge in Dogrib Indians. Med Anthropol . 1989; 11( 3): 329– 350. Google Scholar CrossRef Search ADS PubMed  24. Chavez AO, Kamath S, Jani R, Sharma LK, Monroy A, Abdul-Ghani MA, Centonze VE, Sathyanarayana P, Coletta DK, Jenkinson CP, Bai Y, Folli F, Defronzo RA, Tripathy D. Effect of short-term free fatty acids elevation on mitochondrial function in skeletal muscle of healthy individuals. J Clin Endocrinol Metab . 2010; 95( 1): 422– 429. Google Scholar CrossRef Search ADS PubMed  25. Goodpaster BH, Sparks LM. Metabolic flexibility in health and disease. Cell Metab . 2017; 25( 5): 1027– 1036. Google Scholar CrossRef Search ADS PubMed  26. Li H, Xu M, Lee J, He C, Xie Z. Leucine supplementation increases SIRT1 expression and prevents mitochondrial dysfunction and metabolic disorders in high-fat diet-induced obese mice. Am J Physiol Endocrinol Metab . 2012; 303( 10): E1234– E1244. Google Scholar CrossRef Search ADS PubMed  27. Sun X, Zemel MB. Leucine modulation of mitochondrial mass and oxygen consumption in skeletal muscle cells and adipocytes. Nutr Metab (Lond) . 2009; 6( 1): 26. Google Scholar CrossRef Search ADS PubMed  28. Tatpati LL, Irving BA, Tom A, Bigelow ML, Klaus K, Short KR, Nair KS. The effect of branched chain amino acids on skeletal muscle mitochondrial function in young and elderly adults. J Clin Endocrinol Metab . 2010; 95( 2): 894– 902. Google Scholar CrossRef Search ADS PubMed  29. Abdelmagid SA, Clarke SE, Nielsen DE, Badawi A, El-Sohemy A, Mutch DM, Ma DW. Comprehensive profiling of plasma fatty acid concentrations in young healthy Canadian adults [published correction appears in PLoS One. 2015;10(5):e0128167]. PLoS One . 2015; 10( 2): e0116195. Google Scholar CrossRef Search ADS PubMed  30. Chavarro JE, Stampfer MJ, Campos H, Kurth T, Willett WC, Ma J. A prospective study of trans-fatty acid levels in blood and risk of prostate cancer. Cancer Epidemiol Biomarkers Prev . 2008; 17( 1): 95– 101. Google Scholar CrossRef Search ADS PubMed  31. Rioux V, Catheline D, Bouriel M, Legrand P. Dietary myristic acid at physiologically relevant levels increases the tissue content of C20:5 n-3 and C20:3 n-6 in the rat. Reprod Nutr Dev . 2005; 45( 5): 599– 612. Google Scholar CrossRef Search ADS PubMed  32. Wu JH, Lemaitre RN, Imamura F, King IB, Song X, Spiegelman D, Siscovick DS, Mozaffarian D. Fatty acids in the de novo lipogenesis pathway and risk of coronary heart disease: the Cardiovascular Health Study. Am J Clin Nutr . 2011; 94( 2): 431– 438. Google Scholar CrossRef Search ADS PubMed  33. Lee JJ, Lambert JE, Hovhannisyan Y, Ramos-Roman MA, Trombold JR, Wagner DA, Parks EJ. Palmitoleic acid is elevated in fatty liver disease and reflects hepatic lipogenesis. Am J Clin Nutr . 2015; 101( 1): 34– 43. Google Scholar CrossRef Search ADS PubMed  34. Bézaire V, Langin D. Regulation of adipose tissue lipolysis revisited. Proc Nutr Soc . 2009; 68( 4): 350– 360. Google Scholar CrossRef Search ADS PubMed  35. Cao H, Gerhold K, Mayers JR, Wiest MM, Watkins SM, Hotamisligil GS. Identification of a lipokine, a lipid hormone linking adipose tissue to systemic metabolism. Cell . 2008; 134( 6): 933– 944. Google Scholar CrossRef Search ADS PubMed  36. Rustenbeck I, Lenzen S. Regulation of transmembrane ion transport by reaction products of phospholipase A2. II. Effects of arachidonic acid and other fatty acids on mitochondrial Ca2+ transport. Biochim Biophys Acta . 1989; 982( 1): 147– 155. Google Scholar CrossRef Search ADS PubMed  37. Criddle DN, Murphy J, Fistetto G, Barrow S, Tepikin AV, Neoptolemos JP, Sutton R, Petersen OH. Fatty acid ethyl esters cause pancreatic calcium toxicity via inositol trisphosphate receptors and loss of ATP synthesis. Gastroenterology . 2006; 130( 3): 781– 793. Google Scholar CrossRef Search ADS PubMed  38. Horas H Nababan S, Nishiumi S, Kawano Y, Kobayashi T, Yoshida M, Azuma T. Adrenic acid as an inflammation enhancer in non-alcoholic fatty liver disease. Arch Biochem Biophys . 2017; 623-624: 64– 75. Google Scholar CrossRef Search ADS PubMed  39. Sunny NE, Parks EJ, Browning JD, Burgess SC. Excessive hepatic mitochondrial TCA cycle and gluconeogenesis in humans with nonalcoholic fatty liver disease. Cell Metab . 2011; 14( 6): 804– 810. Google Scholar CrossRef Search ADS PubMed  40. Befroy DE, Perry RJ, Jain N, Dufour S, Cline GW, Trimmer JK, Brosnan J, Rothman DL, Petersen KF, Shulman GI. Direct assessment of hepatic mitochondrial oxidative and anaplerotic fluxes in humans using dynamic 13C magnetic resonance spectroscopy. Nat Med . 2014; 20( 1): 98– 102. Google Scholar CrossRef Search ADS PubMed  Copyright © 2018 Endocrine Society http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Clinical Endocrinology and Metabolism Oxford University Press

Loading next page...
 
/lp/ou_press/amino-acid-and-fatty-acid-levels-affect-hepatic-phosphorus-metabolite-c3GjlVagtZ
Publisher
Endocrine Society
Copyright
Copyright © 2018 Endocrine Society
ISSN
0021-972X
eISSN
1945-7197
D.O.I.
10.1210/jc.2017-01773
Publisher site
See Article on Publisher Site

Abstract

Abstract Objective Hepatic energy metabolism negatively relates to insulin resistance and liver fat content in patients with type 2 diabetes, but its role in metabolically healthy humans is unclear. We hypothesized that intrahepatocellular γ-adenosine triphosphate (γATP) and inorganic phosphate (Pi) concentrations exhibit similar associations with insulin sensitivity in nondiabetic, nonobese volunteers. Design A total of 76 participants underwent a four-point sampling, 75-g oral glucose tolerance test (OGTT), as well as in vivo31P/1H magnetic resonance spectroscopy. In 62 of them, targeted plasma metabolomic profiling was performed. Pearson correlation analyses were performed for the dependent variables γATP and Pi. Results Adjusted for age, sex, and body mass index (BMI), hepatic γATP and Pi related to 2-hour OGTT glucose (r = 0.25 and r = 0.27, both P < 0.05), and Pi further associated with nonesterified fatty acids (NEFAs; r = 0.28, P < 0.05). However, neither γATP nor Pi correlated with several measures of insulin sensitivity. Hepatic γATP correlated with circulating leucine (r = 0.42, P < 0.001) and Pi with C16:1 fatty acids palmitoleic acid and C16:1w5 (r = 0.28 and 0.30, respectively, P < 0.01), as well as with δ-9-desaturase index (r = 0.33, P < 0.05). Only the association of γATP with leucine remained important after correction for multiple testing. Leucine and palmitoleic acid, together with age, sex, and BMI, accounted for 26% and for 15% of the variabilities in γATP and Pi, respectively. Conclusions Specific circulating amino acids and NEFAs, but not measures of insulin sensitivity, partly affect hepatic phosphorus metabolites, suggesting mutual interaction between hepatic energy metabolism and circulating metabolites in nondiabetic humans. Hepatic energy metabolism is altered in insulin-resistant and/or type 2 diabetes (T2D) patients with nonalcoholic fatty liver disease (NAFLD) (1, 2). Patients with T2D present with lower hepatic adenosine triphosphate (ATP), inorganic phosphate (Pi) concentrations, and ATP synthase flux (2, 3). On the other hand, obese individuals without nonalcoholic steatohepatitis exhibit increased hepatic oxidative capacity (1). These conflicting results may result from adaptive upregulation of oxidative phosphorylation in obesity (1), along with enhanced production of reactive oxygen species. Increased reactive oxygen species production, which is also linked to hyperglycemia and hyperlipidemia, can induce mitochondrial damage and thereby, contribute to lower mitochondrial function in T2D and nonalcoholic steatohepatitis (1, 3, 4). In exercising skeletal muscle, ATP production is mainly regulated by training intensity, whereas in the liver, ATP-consuming metabolic processes challenge hepatic mitochondria (5, 6). Several factors, such as body mass index (BMI), glycemia, and hepatic insulin sensitivity (1, 2, 7), affect skeletal muscle and hepatic phosphorus metabolites in metabolic diseases, but there are no data on the factors regulating energy metabolism in nonobese nondiabetic persons. Cellular metabolic processes require substrate supply and signaling, particularly by branched-chain amino acids (BCAAs) (8, 9) and nonesterified fatty acids (NEFAs) (10). Although BCAAs have been suggested to improve energy metabolism (8), BCAAs are also increased in insulin-resistant states, such as obesity and T2D, and can induce insulin resistance (8, 9). In metabolic diseases, impaired BCAA metabolism could lead to accumulation of toxic BCAA metabolites, which would, in turn, impair mitochondrial function and give rise to stress signaling associated with insulin resistance (8). Likewise, chronic increases in NEFA can also induce insulin resistance (4) and have been linked to abnormal mitochondrial function (4). Although BCAA and NEFA likely reflect the risk for metabolic diseases (4, 8), their physiological roles for hepatic phosphorus metabolism remain unclear. We have previously developed noninvasive 31P magnet resonance spectroscopy (MRS) techniques, allowing for exact quantification of hepatic 31P metabolite concentrations in a clinical setting (11, 12). With the use of these methods, we now aimed at identifying metabolic determinants of hepatic γ-ATP (γATP) and Pi under physiological conditions in a cohort of middle-aged metabolically healthy volunteers. Participants and Methods Study population The study was performed as a feasibility study in the context of the national cohort; its protocol is in line with the Declaration of Helsinki (version 2008) and approved by the Ethics Board of the Bavarian Medical Association and of the Heinrich-Heine University Düsseldorf. All participants gave their written, informed consent before inclusion into the study. From the general population with residency in the Düsseldorf region, 496 persons, aged 20 to 70 years, were recruited from July 2011 to December 2012 by advertisement or via local residents’ registration office (Supplemental Fig. S1). Of those, 270 persons without known diabetes mellitus underwent an extended examination, including standardized interview, anthropometry, and blood sampling, after 10 hours fast, as well as a 75-g oral glucose tolerance test (OGTT). Magnetic resonance imaging and MRS were performed in 120 of these individuals. Parts of these data have already been published (11, 13). Clinical examination Anthropometric data were measured, as reported previously (14). OGTT A standardized 75-g OGTT (ACCU-CHECK® Dextro O.G.T.; Roche, Basel, Switzerland) was performed after 10 hours overnight fasting with blood sampling, before and 30, 60, and 120 minutes after start. Dysglycemia was classified according to criteria of the American Diabetes Association. Laboratory measurements Measurements of glucose, alanine aminotransferase, aspartate aminotransferase, γ-glutamyl transpeptidase, high-density lipoprotein-cholesterol, low-density lipoprotein-cholesterol, triglyceride, insulin, and C-peptide were performed, as reported previously (14). Hemoglobin A1c was determined on a VARIANT-IITM Analyzer (Bio-Rad Laboratories, Munich, Germany). Plasma NEFAs were determined using the NEFA-HR kit (Wako Chemicals, Neuss, Germany). Metabolome analyses Sodium-heparinate plasma samples were rapidly frozen and stored at −80°C. Targeted metabolic profiling of 294 blood metabolites was performed with the X MetaDis/DQTM kit at Biocrates Life Sciences (Innsbruck, Austria) (13). Fatty acid C18:0 was excluded from further analyses because of analytical errors. BCAAs were defined by the sum of isoleucine, leucine, and valine and δ-9-desaturase (D9D) index by the ratio of palmitoleic and palmitic acid (13). Magnetic resonance imaging and MRS All examinations were performed on a clinical 3-Tesla whole-body magnet (X-series Achieva; Philips, Best, The Netherlands), equipped with a 16-channel Torso XL phased-array receiver coil for 1H MRS and a 14-cm circular 31P surface coil (transmit-receive coil) using standardized procedures, as described (11). Liver triglyceride concentration [hepatocellular lipid content (HCL)] was assessed by 1H MRS using the single voxel-stimulated echo acquisition mode (14); fat fraction was calculated as percentage of the ratio of lipids/(water + lipids), as described (11). For liver 31P MRS, scout images were taken with the 1H body coil in three orientations, followed by transverse T2-weighted images with multislice two-dimensional spin echo images and respiratory triggering to identify liver position clearly and place the volume of interest (6 × 6 × 6 cm) within the liver, avoiding skeletal muscle. Localized liver spectra were obtained using image-selected spectroscopy, with specifications, and analyzed as described before (11). Measures of insulin sensitivity From glucose, insulin, and C-peptide values, obtained during the OGTT, the following indices have been calculated for insulin sensitivity: oral glucose insulin-sensitivity index (OGIS), representing total glucose clearance or whole-body dynamic insulin sensitivity (15), and quantitative insulin-sensitivity check index (QUICKI) to assess fasting (hepatic) insulin sensitivity, calculated as 1/[log(G0) + log(I0)], where G0 and I0 are fasting glucose and insulin, respectively (16). The hepatic insulin resistance index (17) was calculated as the following: glucose area under the curve (AUC)Gluc0-30 (g/dL) × AUCI0-30 (U/dL) × min2. Liver insulin-resistance index (relationship between insulin sensitivity and cardiovascular disease) (18) was calculated as the following: −0.091 + [log AUCI0-120 (pmol/L) × 0.400] + [log fat mass (%) × 0.346] – [log high-density lipoprotein-cholesterol (mg/dL) × 0.408] + [log BMI (kg/m2) × 0.435]. The adipose tissue insulin resistance (Adipo-IR) index was calculated as the following: fasting NEFA (mmol/L) × I0 (pmol/L) (19). De novo lipogenesis (DNL) was calculated as reported previously (13). Statistical analyses Normally distributed parameters are presented as means ± standard deviation and skewed distributed parameters as median (interquartile range). Pearson correlation analyses were performed for the dependent variables γATP and Pi. To account for potential confounders, such as age, sex, and BMI, partial (adjusted) Pearson correlations were calculated also. To improve normality, skewed distributed data were natural log transformed before analyses, as indicated (see Tables 1, 2, and 3 ). P < 0.05 was considered to indicate statistically significant differences. The main analyses (associations between hepatic γATP or Pi with markers of insulin resistance/amino acids/fatty acids) were adjusted for multiple testing using the Bonferroni method, as indicated in the respective table footnotes. Missing data were excluded from the analyses. No imputation methods were performed. Analyses were done using SAS Version 9.4 (SAS Institute, Cary, NC). Table 1. Correlation of Hepatic Phosphorus Metabolites With Clinical Parameters and Insulin Sensitivity Indices   Unadjusted   Adjusted for Age, Sex, and BMI   Adjusted for Age, Sex, BMI, and Pi or γATP     γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L    Pearson Correlation Coefficient  BMI, kg/m2  0.11  0.07          HCL, % (ln)  0.10  0.04  0.08  0.03  0.07  −0.007  NEFA, µmol/L (ln)  0.19  0.28a  0.18  0.28a  0.08  0.22  TG, mg/dL (ln)  0.10  0.03  0.07  0.009  0.07  −0.02  Fasting glucose, mg/dL  0.07  0.05  0.08  0.05  0.06  0.03  2-h Glucose, mg/dL  0.33b  0.33b  0.25a  0.27a  0.15  0.19  Fasting insulin, mU/L (ln)  −0.13  −0.12  −0.17  −0.14  −0.12  −0.08  HbA1c, %  0.19  0.06  0.12  −0.007  0.13  −0.06  OGIS, au  −0.10  −0.08  −0.002  −0.02  0.007  −0.02  QUICKI (ln)  0.11  0.09  0.15  0.12  0.11  0.06  Hepatic-IR index  −0.04  −0.12  −0.08  −0.14  −0.02  −0.12  RISC  0.11  −0.03  0.03  −0.06  0.05  −0.07  Adipo-IR index  0.03  0.08  0.003  0.06  −0.02  0.06  DNL index  −0.13  −0.008  −0.09  0.06  −0.13  0.11    Unadjusted   Adjusted for Age, Sex, and BMI   Adjusted for Age, Sex, BMI, and Pi or γATP     γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L    Pearson Correlation Coefficient  BMI, kg/m2  0.11  0.07          HCL, % (ln)  0.10  0.04  0.08  0.03  0.07  −0.007  NEFA, µmol/L (ln)  0.19  0.28a  0.18  0.28a  0.08  0.22  TG, mg/dL (ln)  0.10  0.03  0.07  0.009  0.07  −0.02  Fasting glucose, mg/dL  0.07  0.05  0.08  0.05  0.06  0.03  2-h Glucose, mg/dL  0.33b  0.33b  0.25a  0.27a  0.15  0.19  Fasting insulin, mU/L (ln)  −0.13  −0.12  −0.17  −0.14  −0.12  −0.08  HbA1c, %  0.19  0.06  0.12  −0.007  0.13  −0.06  OGIS, au  −0.10  −0.08  −0.002  −0.02  0.007  −0.02  QUICKI (ln)  0.11  0.09  0.15  0.12  0.11  0.06  Hepatic-IR index  −0.04  −0.12  −0.08  −0.14  −0.02  −0.12  RISC  0.11  −0.03  0.03  −0.06  0.05  −0.07  Adipo-IR index  0.03  0.08  0.003  0.06  −0.02  0.06  DNL index  −0.13  −0.008  −0.09  0.06  −0.13  0.11  Pearson and partial Pearson correlations of γATP and Pi with the respective variable are given. The Bonferroni-adjusted significance level is 0.0018 [0.05/(14 metabolic parameters × 2 hepatic phosphorus metabolites)]. After Bonferroni adjustment for multiple comparisons, none of the observed correlations remained statistically significant. Abbreviations: HbA1c, hemoglobin A1c; ln, natural log transformed; RISC, relationship between insulin sensitivity and cardiovascular disease; TG, triglyceride. a P < 0.05. b P < 0.01. View Large Table 2. Correlation of Hepatic Phosphorus Metabolites With Specific Amino Acids   Unadjusted   Adjusted for Age, Sex, and BMI   Adjusted for Age, Sex, BMI, and Pi or γATP     γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L    Pearson Correlation Coefficient  Alanine (Ala; ln)  0.16  0.04  0.08  −0.002  0.08  −0.03  Arginine (Arg; ln)  0.03  0.19  0.01  0.19  −0.06  0.20  Citrulline (Cit; ln)  0.16  −0.007  0.12  −0.08  0.16  −0.14  Glycine (Gly; ln)  0.14  −0.03  0.08  −0.10  0.13  −0.14  Histidine (His; ln)  −0.002  −0.06  0.05  −0.02  0.07  −0.05  Isoleucine (Ile; ln)  0.19  −0.09  0.32a  −0.005  0.35b  −0.14  Leucine (Leu; ln)  0.28a  0.06  0.42c  0.16  0.40b  −0.004  Lysine (Lys; ln)  0.22  0.13  0.15  0.07  0.13  0.01  Methionine (Met; ln)  0.05  −0.09  0.18  0.006  0.19  −0.07  Ornithine (Orn; ln)  0.35b  0.14  0.28a  0.05  0.28a  −0.06  Phenylalanine (Phe; ln)  0.23  −0.02  0.24  −0.02  0.26a  −0.12  Proline (Pro; ln)  −0.01  −0.08  0.06  −0.02  0.07  −0.04  Serine (Ser; ln)  0.15  −0.04  0.12  −0.05  0.15  −0.10  Threonine (Thr; ln)  −0.22  −0.28a  −0.17  −0.23  −0.10  −0.18  Tryptophane (Trp; ln)  0.23  −0.004  0.28a  0.05  0.29a  −0.07  Tyrosine (Tyr; ln)  0.25  −0.07  0.21  −0.10  0.26a  −0.20  Valine (Val; ln)  0.15  0.005  0.25  0.06  0.24  −0.03  BCAA (Ile/Leu/Val; ln)  0.22  −0.007  0.37b  0.09  0.37b  −0.06    Unadjusted   Adjusted for Age, Sex, and BMI   Adjusted for Age, Sex, BMI, and Pi or γATP     γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L    Pearson Correlation Coefficient  Alanine (Ala; ln)  0.16  0.04  0.08  −0.002  0.08  −0.03  Arginine (Arg; ln)  0.03  0.19  0.01  0.19  −0.06  0.20  Citrulline (Cit; ln)  0.16  −0.007  0.12  −0.08  0.16  −0.14  Glycine (Gly; ln)  0.14  −0.03  0.08  −0.10  0.13  −0.14  Histidine (His; ln)  −0.002  −0.06  0.05  −0.02  0.07  −0.05  Isoleucine (Ile; ln)  0.19  −0.09  0.32a  −0.005  0.35b  −0.14  Leucine (Leu; ln)  0.28a  0.06  0.42c  0.16  0.40b  −0.004  Lysine (Lys; ln)  0.22  0.13  0.15  0.07  0.13  0.01  Methionine (Met; ln)  0.05  −0.09  0.18  0.006  0.19  −0.07  Ornithine (Orn; ln)  0.35b  0.14  0.28a  0.05  0.28a  −0.06  Phenylalanine (Phe; ln)  0.23  −0.02  0.24  −0.02  0.26a  −0.12  Proline (Pro; ln)  −0.01  −0.08  0.06  −0.02  0.07  −0.04  Serine (Ser; ln)  0.15  −0.04  0.12  −0.05  0.15  −0.10  Threonine (Thr; ln)  −0.22  −0.28a  −0.17  −0.23  −0.10  −0.18  Tryptophane (Trp; ln)  0.23  −0.004  0.28a  0.05  0.29a  −0.07  Tyrosine (Tyr; ln)  0.25  −0.07  0.21  −0.10  0.26a  −0.20  Valine (Val; ln)  0.15  0.005  0.25  0.06  0.24  −0.03  BCAA (Ile/Leu/Val; ln)  0.22  −0.007  0.37b  0.09  0.37b  −0.06  Pearson and partial Pearson correlations of γATP and Pi with specific amino acids are given, respectively. The Bonferroni-adjusted significance level is 0.0014 [0.05/(18 amino acid parameters × 2 hepatic phosphorus metabolites)]. After Bonferroni adjustment for multiple comparisons, only the association of γATP and leucine remained statistically significant. Abbreviation: ln, natural log transformed. a P < 0.05. b P < 0.01. c P < 0.001. View Large Table 3. Correlation of Hepatic Phosphorus Metabolites With Free Fatty Acid Species   Unadjusted   Adjusted for Age, Sex, and BMI   Adjusted for Age, Sex, BMI, and Pi or γATP     γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L    Pearson Correlation Coefficient  SFA               C14:0 (myristic; ln)  0.02  0.13  −0.05  0.09  −0.09  0.11   C15:0 (pentadecanoic; ln)  −0.13  0.07  −0.13  0.08  −0.17  0.14   C16:0 (palmitic; ln)  −0.06  0.09  −0.19  0.02  −0.21  0.10   C17:0 (heptadecanoic; ln)  0.0007  0.04  −0.08  −0.03  −0.08  0.004   C18:0 (stearic; ln)               C19:0 (nonadecanoic; ln)  0.02  0.02  −0.16  −0.14  −0.12  −0.09   C20:0 (eicosanoic; ln)  −0.24  −0.18  −0.22  −0.16  −0.17  −0.09  MUFA               cC14:1w5 (myristoleic; ln)  0.13  0.28a  0.08  0.28a  −0.02  0.27a   cC15:1w5, c10 (pentadecenoic; ln)  −0.14  0.10  −0.19  0.07  −0.23  0.15   cC16:1w5 (ln)  −0.04  0.32a  −0.14  0.30a  −0.28a  0.38b   cC16:1w7 (palmitoleic; ln)  0.09  0.33b  −0.03  0.28a  −0.15  0.31a   cC16:1w10 (sapienic; ln)  0.01  0.16  −0.13  0.10  −0.18  0.16   cC17:1w7, c10 (heptadecenoic; ln)  −0.16  −0.004  −0.23  −0.07  −0.22  0.02   cC17:1w8 (ln)  0.10  0.21  −0.03  0.11  −0.08  0.14   cC18:1w7 (vaccenic; ln)  −0.01  0.20  −0.17  0.10  −0.22  0.18   cC18:1w9 (oleic; ln)  0.01  0.12  −0.13  0.02  −0.14  0.07   cC20:1w9, c11 (eicosenoic; ln)  −0.02  0.13  −0.12  0.05  −0.15  0.10  PUFA               cC18:2w6 (linoleic; ln)  0.05  0.08  −0.07  −0.04  −0.06  −0.01   cC18:3w3 (linolenic; ln)  0.03  0.13  −0.12  0.02  −0.13  0.06   cC18:3w6 (gamma linolenic; ln)  0.09  0.18  −0.005  0.13  −0.06  0.15   cC18:4w3 (stearidonic; ln)  0.03  0.20  −0.05  0.14  −0.11  0.17   cC20:2w6, c11, 14 (eicosadienoic; ln)  −0.02  0.16  −0.12  0.08  −0.17  0.14   cC20:3w6, c8, 11, 14 (eicosatrienoic; ln)  −0.05  −0.01  −0.05  0.02  −0.06  0.04   cC20:3w9 (mead; ln)  0.06  0.04  0.03  0.003  0.03  −0.009   cC20:4w6 (arachidonic; ln)  −0.17  0.04  −0.18  0.05  −0.22  0.13   cC20:5w3 (EPA; ln)  0.004  0.07  −0.13  −0.05  −0.12  0.003   cC22:4w6 (adrenic; ln)  −0.15  0.26a  −0.25  0.22  −0.37b  0.35b   cC22:5w3 (DPA; ln)  0.04  0.24  −0.12  0.13  −0.19  0.19   cC22:6w3 (DHA; ln)  0.03  0.14  −0.12  0.05  −0.14  0.10   Q_cC16_1__cC16_0  0.15  0.37b  0.08  0.33a  −0.05  0.33a    Unadjusted   Adjusted for Age, Sex, and BMI   Adjusted for Age, Sex, BMI, and Pi or γATP     γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L  γATP, mmol/L  Pi, mmol/L    Pearson Correlation Coefficient  SFA               C14:0 (myristic; ln)  0.02  0.13  −0.05  0.09  −0.09  0.11   C15:0 (pentadecanoic; ln)  −0.13  0.07  −0.13  0.08  −0.17  0.14   C16:0 (palmitic; ln)  −0.06  0.09  −0.19  0.02  −0.21  0.10   C17:0 (heptadecanoic; ln)  0.0007  0.04  −0.08  −0.03  −0.08  0.004   C18:0 (stearic; ln)               C19:0 (nonadecanoic; ln)  0.02  0.02  −0.16  −0.14  −0.12  −0.09   C20:0 (eicosanoic; ln)  −0.24  −0.18  −0.22  −0.16  −0.17  −0.09  MUFA               cC14:1w5 (myristoleic; ln)  0.13  0.28a  0.08  0.28a  −0.02  0.27a   cC15:1w5, c10 (pentadecenoic; ln)  −0.14  0.10  −0.19  0.07  −0.23  0.15   cC16:1w5 (ln)  −0.04  0.32a  −0.14  0.30a  −0.28a  0.38b   cC16:1w7 (palmitoleic; ln)  0.09  0.33b  −0.03  0.28a  −0.15  0.31a   cC16:1w10 (sapienic; ln)  0.01  0.16  −0.13  0.10  −0.18  0.16   cC17:1w7, c10 (heptadecenoic; ln)  −0.16  −0.004  −0.23  −0.07  −0.22  0.02   cC17:1w8 (ln)  0.10  0.21  −0.03  0.11  −0.08  0.14   cC18:1w7 (vaccenic; ln)  −0.01  0.20  −0.17  0.10  −0.22  0.18   cC18:1w9 (oleic; ln)  0.01  0.12  −0.13  0.02  −0.14  0.07   cC20:1w9, c11 (eicosenoic; ln)  −0.02  0.13  −0.12  0.05  −0.15  0.10  PUFA               cC18:2w6 (linoleic; ln)  0.05  0.08  −0.07  −0.04  −0.06  −0.01   cC18:3w3 (linolenic; ln)  0.03  0.13  −0.12  0.02  −0.13  0.06   cC18:3w6 (gamma linolenic; ln)  0.09  0.18  −0.005  0.13  −0.06  0.15   cC18:4w3 (stearidonic; ln)  0.03  0.20  −0.05  0.14  −0.11  0.17   cC20:2w6, c11, 14 (eicosadienoic; ln)  −0.02  0.16  −0.12  0.08  −0.17  0.14   cC20:3w6, c8, 11, 14 (eicosatrienoic; ln)  −0.05  −0.01  −0.05  0.02  −0.06  0.04   cC20:3w9 (mead; ln)  0.06  0.04  0.03  0.003  0.03  −0.009   cC20:4w6 (arachidonic; ln)  −0.17  0.04  −0.18  0.05  −0.22  0.13   cC20:5w3 (EPA; ln)  0.004  0.07  −0.13  −0.05  −0.12  0.003   cC22:4w6 (adrenic; ln)  −0.15  0.26a  −0.25  0.22  −0.37b  0.35b   cC22:5w3 (DPA; ln)  0.04  0.24  −0.12  0.13  −0.19  0.19   cC22:6w3 (DHA; ln)  0.03  0.14  −0.12  0.05  −0.14  0.10   Q_cC16_1__cC16_0  0.15  0.37b  0.08  0.33a  −0.05  0.33a  Pearson and partial Pearson correlations of γATP and Pi with specific amino acids are given, respectively. The Bonferroni-adjusted significance level is 0.00083 [0.05/(30 fatty acid parameters × 2 hepatic phosphorus metabolites)]. After Bonferroni adjustment for multiple comparisons, only the association of Pi and palmitoleic acid remained statistically significant. Abbreviations: DHA, docosahexaenoic acid; DPA, docosapentaenoic acid; EPA, eicosapentaenoic acid; ln, natural log transformed; MUFA, monounsaturated fatty acids; PUFA, polyunsaturated fatty acids; SFA, saturated fatty acids. a P < 0.05. b P < 0.01. View Large Results Study population Of 120 volunteers undergoing OGTT and MRS examinations, 76 entered the final analysis after exclusion of participants with missing measures and/or metabolic disease (Supplemental Fig. S1). Participants (26 men, 50 women) were middle aged (54 ± 13 years), nonobese (BMI 24.8 ± 3.2 kg/m2), insulin sensitive (OGIS 461.2 ± 61.2 arbitrary units (au), QUICKI 0.46 [0.43;0.50]), and had low HCL (1.00 [0.41;2.76]%) and transaminases (alanine aminotransferase 18 [15;24] U/L, aspartate aminotransferase 23 [20;27] U/L). Fasting and 2-hour OGTT glucose were 4.2 ± 0.5 and 5.0 ± 1.3 mmol/L, respectively. Fasting NEFA and insulin were 516 [404;714] and 42.6 [30.3;62.0] pmol/L, respectively. Hepatic γATP, Pi concentrations, and the Pi/γATP ratio were 2.73 ± 0.54, 2.02 ± 0.49, and 0.75 ± 0.18 mmol/L, respectively. Absolute concentrations of the measured amino acids and lipid metabolites are shown in Supplemental Tables S1 and S2. Correlation analyses for hepatic phosphorus metabolites with glycemia and insulin sensitivity Pearson correlation analyses revealed positive relationships of hepatic γATP concentrations with 2-hour OGTT glucose and of Pi with 2-hour OGTT glucose and NEFA, respectively (Fig. 1A–1D; Table 1). After adjustment for age, sex, and BMI, 2-hour OGTT glucose still related to both γATP and Pi content and NEFA levels to Pi concentrations. The adjustment of the analyses for Pi or γATP abolished these associations, pointing to the impact of the concentration of the respective other phosphorus metabolite on these relationships (Table 1). No associations with γATP or Pi content were found for any calculated measure of insulin sensitivity in unadjusted and adjusted analyses (Table 1). No association remained statistically significant after correction for multiple testing. Figure 1. View largeDownload slide Correlations of γATP and Pi with plasma metabolites. (A, C, and E) Correlations of γATP with 2-hour OGTT glucose, NEFA, and leucine, respectively. (B, D, and F) Correlations of Pi with 2-hour OGTT glucose, NEFA, and palmitoleic acid, respectively. Before correlation analyses, NEFA and metabolites were log transformed to achieve normal distribution. Metabolites were adjusted further for plate of measurement (shown as au). Figure 1. View largeDownload slide Correlations of γATP and Pi with plasma metabolites. (A, C, and E) Correlations of γATP with 2-hour OGTT glucose, NEFA, and leucine, respectively. (B, D, and F) Correlations of Pi with 2-hour OGTT glucose, NEFA, and palmitoleic acid, respectively. Before correlation analyses, NEFA and metabolites were log transformed to achieve normal distribution. Metabolites were adjusted further for plate of measurement (shown as au). Correlation analyses for hepatic phosphorus metabolites with amino acids Hepatic γATP was positively associated with plasma leucine (Fig. 1E) and ornithine. Hepatic Pi related negatively to threonine (Table 2). After adjustment for age, sex, and BMI, γATP still positively correlated with leucine and ornithine. Moreover, γATP associated with isoleucine, tryptophan, and BCAAs. The association of Pi with threonine disappeared after adjustment for age, sex, and BMI (Table 2). After additional adjustment for Pi, the correlation of γATP with isoleucine, leucine, ornithine, tryptophane, and BCAA remained, whereas additional correlations of γATP with tyrosine and phenylalanine were found (Table 2). After correction for multiple testing, only the association between γATP and leucine remained statistically significant (Table 2). Correlation analyses for hepatic phosphorus metabolites with fatty acids species Neither γATP nor Pi related to any analyzed saturated fatty acid species. Furthermore, γATP did not associate with any unsaturated fatty acid species (Table 3). Pi related positively to myristoleic acid (cC14:1w5), cC16:1w5, palmitoleic acid (cC16:1w7; Fig. 1F), adrenic acid (cC22:4w6), and the D9D index. After adjustment for age, sex, and BMI, only the relationships of Pi with myristoleic acid, cC16:1w5, palmitoleic acid, and D9D index remained (Table 3). Further adjustment for Pi revealed negative relationships of γATP with cC16:1w5 and adrenic acid. After adjustment for γATP, Pi related positively to myristoleic acid, cC16:1w5, palmitoleic acid, adrenic acid, and D9D index (Table 3). When corrected for multiple testing, none of the associations remained statistically significant (Table 3). In multivariate regression models, age, sex, and BMI explained 9.8% of γATP and 7.8% of Pi variability, respectively. Additional adjustment for leucine, as a significant independent predictor, explained 26% of γATP variances, and the adjustment for age, sex, BMI, and palmitoleic acid accounted for 15% of Pi variability. A multivariate regression analysis containing variables age, sex, BMI, palmitoleic acid, and cC16:1w5 did not improve R2 (0.16). Predictors of hepatic phosphorus metabolite concentrations These analyses identified leucine and C16:1 fatty acid palmitoleic acid as predictors of γATP and Pi content, respectively. We further confirmed that additional adjustment for selected parameters, indicative of insulin resistance and glucose intolerance, did not affect these relationships (Supplemental Table S3). The close association of γATP and leucine was not affected by adjustment for any of the chosen parameters. However, the relation of Pi and palmitoleic acid disappeared by adjustment for 2-hour OGTT glucose and NEFA levels, respectively. Discussion This study found that specific essential amino acids and fatty acid species are independent predictors of hepatic γATP and Pi content in metabolically healthy humans, whereas measures of insulin sensitivity neither related to hepatic γATP nor to Pi content. This cohort had low HCL, normal glucose tolerance, and insulin sensitivity, based on the OGTT analysis (15). Moreover, absolute concentrations of hepatic γATP and Pi were similar to nondiabetic cohorts with comparable baseline characteristics (2, 12). Thus, this cohort represents an appropriate group for investigation of hepatic energy status under physiological conditions. Absolute hepatic concentrations of Pi and ATP have been reported to be positively associated with hepatic but not with peripheral insulin sensitivity and negatively with HCL in a mixed collective of patients with T2D and nondiabetic humans (2). Another study reported a negative correlation of ATP levels with BMI (20). Of note, our results do not confirm these associations in a collective of nondiabetic humans, as we did not observe any association of γATP and Pi with indices of fasting (QUICKI), dynamic peripheral (OGIS), or hepatic (Hepatic-IR index, relationship between insulin sensitivity and cardiovascular disease) and adipose tissue (Adipo-IR index) insulin sensitivity, HCL, or BMI, respectively. This might result from the fact that our collective excluded participants with hyperglycemia or hepatic steatosis. Furthermore, we used fasting and OGTT-derived measures of insulin sensitivity, but OGIS and M-values calculated from OGTT and hyperinsulinemic-euglycemic clamps, respectively, are closely correlated (7, 15). Unexpectedly, we observed a positive relationship of both γATP and Pi with 2-hour OGTT glucose levels, which may serve as a rough predictor of insulin resistance (21). However, factors other than insulin resistance, such as duration of fasting and the amount of carbohydrate intake on the previous days, may influence OGTT glucose levels (22, 23). In T2D, the negative relationship between hepatic γATP and ATP synthase flux and insulin resistance was determined, at least in part, by fasting plasma glucose and hemoglobin A1c, respectively (2, 3). Thus, the observed positive relationships of hepatic γATP and Pi with 2-hour OGTT glucose observed in this nonobese nondiabetic cohort are unlikely a result of insulin resistance. Of note, nondiabetic but obese persons even show greater hepatic oxidative capacity, despite peripheral insulin resistance (1). We found a positive relation of Pi with circulating NEFA. In patients with T2D, no correlation was found for ATP flux and NEFA (3). Previous studies reported no association of γATP and plasma NEFA levels in metabolically healthy and diseased humans (12, 24), whereas Pi associated negatively with NEFA levels in a combined collective of healthy, obese, and T2D volunteers (12). Moreover, Pi did not correlate with the Adipo-IR index, suggesting that Adipo-IR cannot exclusively explain the relationship of Pi and NEFA. However, a discrete effect cannot be resolved by our study as a result of the limited sample size. Thus, in healthy humans with intact metabolic flexibility, higher substrate availability of either glucose or NEFA may charge hepatic phosphorus pools and/or stimulate energy-generating processes (25). Of note, further adjustment for γATP or Pi, respectively, abolished the previous associations, indicating that the relationship of 2-hour OGTT glucose with γATP and Pi, as well as of NEFA concentrations with Pi, also depends on the prevailing Pi and γATP levels, respectively. Absolute fasting γATP and Pi levels, as well as Pi/ATP ratios of our study, differed, in part, from those of the respective studies, which might also explain the observed differences in the correlation analyses (2, 3, 12). However, none of the associations of γATP or Pi with 2-hour OGTT glucose and of NEFA concentrations with Pi was statistically significant when corrected for multiple testing. Interestingly, this study found a positive association of γATP with some essential amino acids and of Pi with some fatty acid species, respectively. People with insulin resistance or T2D typically present with higher levels of combined BCAA (leucine, isoleucine, valine) and/or other amino acids (phenylalanine, tryptophane, tyrosine, alanine, citrulline) and lower levels of threonine, glycine, and glutamine (9, 13). In our metabolically healthy cohort, γATP related strongest to leucine, independent of changes in Pi content. Leucine and isoleucine were reported to be elevated in obese compared with lean humans (9). In our cohort, γATP still related closely to leucine even after adjustment for age, sex, and BMI. Further investigation of parameters affecting the relation of γATP and leucine showed no effect on this interaction. In obese mice, leucine increases ATP concentrations in brown adipose tissue as a result of stimulation of mitochondrial biogenesis (26), and in skeletal muscle cells, leucine stimulates fatty acid oxidation, as well as oxygen consumption (27). Moreover, leucine provides carbon skeletons to the tricarboxylic acid (TCA) cycle at the level of acetyl-coenzyme A, which may enhance TCA cycle flux (28). Thus, higher plasma leucine levels likely contribute to improved mitochondrial performance rather than reflect insulin resistance or impaired mitochondrial function in our collective of healthy humans. This study observed relevant correlations of Pi with unsaturated fatty acid species, such as myristoleic acid, C16:1w5, palmitoleic acid, and adrenic acid. The absolute concentrations of these fatty acids were rather low, as previously reported (29). Whereas our assay does not distinguish between cis and trans fatty acids, the low contribution of trans fatty acids to the total circulating fatty acid pool, amounting to <2% (30), suggests that the observed correlations are mainly driven by the respective cis isomers. Myristoleic acid usually accounts for only small amounts of total fatty acids in animal tissues but is abundant in milk fat and may also raise palmitoleic acid levels (31). It is therefore conceivable that the circulating palmitoleic and myristoleic acid levels, at least partly, result from dietary intake. We did not detect a correlation of the DNL index and γATP or Pi, respectively. The main fatty acid products of DNL include palmitic acid, palmitoleic acid, vaccenic acid, stearic acid, and oleic acid (32). Except for palmitoleic acid, we did not find a correlation of those fatty acids with Pi, further suggesting that the relationship of Pi and palmitoleic acid does not result from DNL. Nevertheless, palmitoleic acid per se and the D9D index have been proposed as markers of DNL (33). The strong interaction between Pi and palmitoleic acid disappeared upon adjustments for fasting NEFA and 2-hour OGTT glucose levels. Fasting NEFA levels mainly reflect adipose tissue lipolysis, which also is tightly regulated by the glucose-regulating hormones insulin and catecholamines (34). This suggests that both NEFA supply to the liver and regulatory hormones contribute to this interaction. Palmitoleic acid has originally gained attention as an anti-inflammatory adipokine (35). In a murine model, chronic palmitoleic acid feeding enhances hepatic glucose uptake and fatty acid oxidation for energy production instead of storage through activation of adenosine 5′-monophosphate-activated protein kinase and fibroblast growth factor 21 -21, dependent on peroxisome proliferator-activated receptor α (10). Of note, cis monounsaturated fatty acids also strongly inhibit Ca2+ uptake, thereby inducing net Ca2+ efflux and higher extramitochondrial Ca2+ concentrations (36). Effects of palmitoleic acid and its ester on mitochondrial metabolism have also been reported in other tissues (37). Taken together, these relationships of palmitoleic and myristoleic acids with hepatic Pi might result directly or indirectly from their interaction with mitochondrial Ca2+ handling and reduction in the mitochondrial membrane potential. As to possible hepatic effects of the other unsaturated fatty acids, adrenic acid may act as an inflammation enhancer in NAFLD (38), whereas there is currently no information on the physiological relevance of C16:1w5. The strength of this study is the investigation of hepatic energy metabolism in a larger cohort of well-phenotyped, glucose-tolerant humans. Up until now, hepatic ATP and Pi content have mainly been studied in the context of metabolic diseases, such as T2D or NAFLD (2, 3, 6). This approach might help to differentiate among factors regulating ATP and Pi in nondiabetic, nonobese, non-NAFLD persons and in those with metabolic disease. This study also has some limitations. The main weakness resides in the inclusion of only healthy, nonobese individuals. Nevertheless, relationships of amino and fatty acid species with hepatic phosphorus metabolites have, to our knowledge, not been reported before, and this study provides that information. Although measurement of absolute hepatic concentrations of ATP and Pi is reliable and highly reproducible, it covers only one aspect of hepatic energy metabolism—the flux through hepatic ATP synthase (3)—but does not address other features of mitochondrial function, such as TCA cycle activity and oxidative capacity, which may be differently affected in health and disease (1, 39, 40). Furthermore, the observed effect of leucine and palmitoleic acid on ATP and Pi levels is only modest, implying that other factors also regulate hepatic energy metabolism in healthy humans. Such factors might be less relevant in the presence of hepatic steatosis, insulin resistance, or hepatocellular injury (2, 6). Our data, from the examined cohort, suggest that hepatic concentrations of ATP and Pi relate to specific circulating amino acids and fatty acid species, but analyses could be considered exploratory as a result of the size of our cohort and the lack of statistical significance after Bonferroni correction for many associations. However, the strongest correlation found for hepatic γATP levels with leucine was still present after correction for multiple testing. In conclusion, hepatic γATP and Pi do not relate to measures of insulin sensitivity in metabolically healthy nonobese humans. However, specific circulating amino acids and NEFAs partly affect the availability of hepatic phosphorus metabolites. These results point to a mutual interaction between hepatic energy metabolism and circulating metabolites and may therefore contribute to the better understanding of the differences in hepatic energy metabolism between glucose-tolerant and insulin-resistant humans. Abbreviations: Adipo-IR adipose tissue insulin resistance ATP adenosine triphosphate au arbitrary unit(s) AUC area under the curve BCAA branched-chain amino acid BMI body mass index D9D δ-9-desaturase DNL de novo lipogenesis HCL hepatocellular lipid content MRS magnet resonance spectroscopy NAFLD nonalcoholic fatty liver disease NEFA nonesterified fatty acid OGIS oral glucose insulin-sensitivity index OGTT oral glucose tolerance test Pi inorganic phosphate QUICKI quantitative insulin-sensitivity check index T2D type 2 diabetes TCA tricarboxylic acid γATP γ-adenosine triphosphate. Acknowledgments We thank Nicole Achterath, Irena Latta, Andrea Nagel, Birgit Platzbecker, Daniela Seeger, Dominik Scheibelhut, and Rita Schreiner (all from the Institute for Clinical Diabetology, German Diabetes Center, Germany) for excellent technical assistance and care of the participants. The study design of the National Cohort was done by the German National Cohort Epidemiologic Planning Committee and Project Management Team and approved by Dr. Karl-Heinz Jöckel, chairman of the Scientific Board of the German National Cohort. Financial Support: This study was conducted within the pretest studies of the German National Cohort (www.nationale-kohorte.de), which were funded by the Federal Ministry of Education and Research (BMBF), Förderkennzeichen 01ER1001A-I, and supported by the participating universities, institutes of the Leibniz Association. This study was further supported, in part, by the German Diabetes Association (DDG); Schmutzler Stiftung; ICEMED Helmholtz-Alliance and the German Center for Diabetes Research (DZD e.V.); German Federal Ministry of Health (Germany); and Ministry of Innovation, Science, and Research of the State North Rhine-Westphalia (Germany). Author Contributions: M.R. designed the study and headed the clinical experiments. S.K., B.N., B.K., A.B., P.J.N., F.W., J.-H.H., and B.H. researched the data. S.K. analyzed the data and wrote the manuscript. G.P. calculated indices of insulin sensitivity. K.S. and G.G. performed the statistical analyses. All of the authors contributed substantially to aspects of study design or the acquisition of data, contributed to drafting of the article or revised it critically for important intellectual content, and gave final approval to the version to be published. M.R. is responsible for the integrity of the work as a whole. Disclosure Summary: The authors have nothing to disclose. References 1. Koliaki C, Szendroedi J, Kaul K, Jelenik T, Nowotny P, Jankowiak F, Herder C, Carstensen M, Krausch M, Knoefel WT, Schlensak M, Roden M. Adaptation of hepatic mitochondrial function in humans with non-alcoholic fatty liver is lost in steatohepatitis. Cell Metab . 2015; 21( 5): 739– 746. Google Scholar CrossRef Search ADS PubMed  2. Szendroedi J, Chmelik M, Schmid AI, Nowotny P, Brehm A, Krssak M, Moser E, Roden M. Abnormal hepatic energy homeostasis in type 2 diabetes. Hepatology . 2009; 50( 4): 1079– 1086. Google Scholar CrossRef Search ADS PubMed  3. Schmid AI, Szendroedi J, Chmelik M, Krssák M, Moser E, Roden M. Liver ATP synthesis is lower and relates to insulin sensitivity in patients with type 2 diabetes. Diabetes Care . 2011; 34( 2): 448– 453. Google Scholar CrossRef Search ADS PubMed  4. Jelenik T, Roden M. Mitochondrial plasticity in obesity and diabetes mellitus. Antioxid Redox Signal . 2013; 19( 3): 258– 268. Google Scholar CrossRef Search ADS PubMed  5. Kemp GJ, Taylor DJ, Radda GK. Control of phosphocreatine resynthesis during recovery from exercise in human skeletal muscle. NMR Biomed . 1993; 6( 1): 66– 72. Google Scholar CrossRef Search ADS PubMed  6. Cortez-Pinto H, Chatham J, Chacko VP, Arnold C, Rashid A, Diehl AM. Alterations in liver ATP homeostasis in human nonalcoholic steatohepatitis: a pilot study. JAMA . 1999; 282( 17): 1659– 1664. Google Scholar CrossRef Search ADS PubMed  7. Szendroedi J, Kaul K, Kloock L, Straßburger K, Schmid AI, Chmelik M, Kacerovsky M, Kacerovsky-Bielesz G, Prikoszovich T, Brehm A, Krssák M, Gruber S, Krebs M, Kautzky-Willer A, Moser E, Pacini G, Roden M. Lower fasting muscle mitochondrial activity relates to hepatic steatosis in humans. Diabetes Care . 2014; 37( 2): 468– 474. Google Scholar CrossRef Search ADS PubMed  8. Lynch CJ, Adams SH. Branched-chain amino acids in metabolic signalling and insulin resistance. Nat Rev Endocrinol . 2014; 10( 12): 723– 736. Google Scholar CrossRef Search ADS PubMed  9. Newgard CB, An J, Bain JR, Muehlbauer MJ, Stevens RD, Lien LF, Haqq AM, Shah SH, Arlotto M, Slentz CA, Rochon J, Gallup D, Ilkayeva O, Wenner BR, Yancy WS, Jr, Eisenson H, Musante G, Surwit RS, Millington DS, Butler MD, Svetkey LP. A branched-chain amino acid-related metabolic signature that differentiates obese and lean humans and contributes to insulin resistance. Cell Metab . 2009; 9( 4): 311– 326. Google Scholar CrossRef Search ADS PubMed  10. de Souza CO, Teixeira AAS, Biondo LA, Lima Junior EA, Batatinha HAP, Rosa Neto JC. Palmitoleic acid improves metabolic functions in fatty liver by PPARα-dependent AMPK activation. J Cell Physiol . 2017; 232( 8): 2168– 2177. Google Scholar CrossRef Search ADS PubMed  11. Laufs A, Livingstone R, Nowotny B, Nowotny P, Wickrath F, Giani G, Bunke J, Roden M, Hwang JH. Quantitative liver 31P magnetic resonance spectroscopy at 3T on a clinical scanner. Magn Reson Med . 2014; 71( 5): 1670– 1675. Google Scholar CrossRef Search ADS PubMed  12. Fritsch M, Koliaki C, Livingstone R, Phielix E, Bierwagen A, Meisinger M, Jelenik T, Strassburger K, Zimmermann S, Brockmann K, Wolff C, Hwang JH, Szendroedi J, Roden M. Time course of postprandial hepatic phosphorus metabolites in lean, obese, and type 2 diabetes patients. Am J Clin Nutr . 2015; 102( 5): 1051– 1058. Google Scholar CrossRef Search ADS PubMed  13. Knebel B, Strassburger K, Szendroedi J, Kotzka J, Scheer M, Nowotny B, Müssig K, Lehr S, Pacini G, Finner H, Klüppelholz B, Giani G, Al-Hasani H, Roden M; German Diabetes Study Group. Specific metabolic profiles and their relationship to insulin resistance in recent-onset type 1 and type 2 diabetes. J Clin Endocrinol Metab . 2016; 101( 5): 2130– 2140. Google Scholar CrossRef Search ADS PubMed  14. Kahl S, Straßburger K, Nowotny B, Livingstone R, Klüppelholz B, Keßel K, Hwang JH, Giani G, Hoffmann B, Pacini G, Gastaldelli A, Roden M. Comparison of liver fat indices for the diagnosis of hepatic steatosis and insulin resistance. PLoS One . 2014; 9( 4): e94059. Google Scholar CrossRef Search ADS PubMed  15. Mari A, Pacini G, Murphy E, Ludvik B, Nolan JJ. A model-based method for assessing insulin sensitivity from the oral glucose tolerance test. Diabetes Care . 2001; 24( 3): 539– 548. Google Scholar CrossRef Search ADS PubMed  16. Katz A, Nambi SS, Mather K, Baron AD, Follmann DA, Sullivan G, Quon MJ. Quantitative insulin sensitivity check index: a simple, accurate method for assessing insulin sensitivity in humans. J Clin Endocrinol Metab . 2000; 85( 7): 2402– 2410. Google Scholar CrossRef Search ADS PubMed  17. Abdul-Ghani MA, Matsuda M, Balas B, DeFronzo RA. Muscle and liver insulin resistance indexes derived from the oral glucose tolerance test. Diabetes Care . 2007; 30( 1): 89– 94. Google Scholar CrossRef Search ADS PubMed  18. Vangipurapu J, Stančáková A, Kuulasmaa T, Paananen J, Kuusisto J, Ferrannini E, Laakso M; EGIR-RISC Study Group. A novel surrogate index for hepatic insulin resistance. Diabetologia . 2011; 54( 3): 540– 543. Google Scholar CrossRef Search ADS PubMed  19. Gastaldelli A, Harrison SA, Belfort-Aguilar R, Hardies LJ, Balas B, Schenker S, Cusi K. Importance of changes in adipose tissue insulin resistance to histological response during thiazolidinedione treatment of patients with nonalcoholic steatohepatitis. Hepatology . 2009; 50( 4): 1087– 1093. Google Scholar CrossRef Search ADS PubMed  20. Nair S, P Chacko V, Arnold C, Diehl AM. Hepatic ATP reserve and efficiency of replenishing: comparison between obese and nonobese normal individuals. Am J Gastroenterol . 2003; 98( 2): 466– 470. Google Scholar PubMed  21. Saxena P, Prakash A, Nigam A. Efficacy of 2-hour post glucose insulin levels in predicting insulin resistance in polycystic ovarian syndrome with infertility. J Hum Reprod Sci . 2011; 4( 1): 20– 22. Google Scholar CrossRef Search ADS PubMed  22. Wang P, Lu MC, Yu CW, Yan YH. Influence of the time of day and fasting duration on glucose level following a 1-hour, 50-gram glucose challenge test in pregnant women. PLoS One . 2014; 9( 11): e112526. Google Scholar CrossRef Search ADS PubMed  23. Szathmary EJ. The impact of low carbohydrate consumption on glucose tolerance, insulin concentration and insulin response to glucose challenge in Dogrib Indians. Med Anthropol . 1989; 11( 3): 329– 350. Google Scholar CrossRef Search ADS PubMed  24. Chavez AO, Kamath S, Jani R, Sharma LK, Monroy A, Abdul-Ghani MA, Centonze VE, Sathyanarayana P, Coletta DK, Jenkinson CP, Bai Y, Folli F, Defronzo RA, Tripathy D. Effect of short-term free fatty acids elevation on mitochondrial function in skeletal muscle of healthy individuals. J Clin Endocrinol Metab . 2010; 95( 1): 422– 429. Google Scholar CrossRef Search ADS PubMed  25. Goodpaster BH, Sparks LM. Metabolic flexibility in health and disease. Cell Metab . 2017; 25( 5): 1027– 1036. Google Scholar CrossRef Search ADS PubMed  26. Li H, Xu M, Lee J, He C, Xie Z. Leucine supplementation increases SIRT1 expression and prevents mitochondrial dysfunction and metabolic disorders in high-fat diet-induced obese mice. Am J Physiol Endocrinol Metab . 2012; 303( 10): E1234– E1244. Google Scholar CrossRef Search ADS PubMed  27. Sun X, Zemel MB. Leucine modulation of mitochondrial mass and oxygen consumption in skeletal muscle cells and adipocytes. Nutr Metab (Lond) . 2009; 6( 1): 26. Google Scholar CrossRef Search ADS PubMed  28. Tatpati LL, Irving BA, Tom A, Bigelow ML, Klaus K, Short KR, Nair KS. The effect of branched chain amino acids on skeletal muscle mitochondrial function in young and elderly adults. J Clin Endocrinol Metab . 2010; 95( 2): 894– 902. Google Scholar CrossRef Search ADS PubMed  29. Abdelmagid SA, Clarke SE, Nielsen DE, Badawi A, El-Sohemy A, Mutch DM, Ma DW. Comprehensive profiling of plasma fatty acid concentrations in young healthy Canadian adults [published correction appears in PLoS One. 2015;10(5):e0128167]. PLoS One . 2015; 10( 2): e0116195. Google Scholar CrossRef Search ADS PubMed  30. Chavarro JE, Stampfer MJ, Campos H, Kurth T, Willett WC, Ma J. A prospective study of trans-fatty acid levels in blood and risk of prostate cancer. Cancer Epidemiol Biomarkers Prev . 2008; 17( 1): 95– 101. Google Scholar CrossRef Search ADS PubMed  31. Rioux V, Catheline D, Bouriel M, Legrand P. Dietary myristic acid at physiologically relevant levels increases the tissue content of C20:5 n-3 and C20:3 n-6 in the rat. Reprod Nutr Dev . 2005; 45( 5): 599– 612. Google Scholar CrossRef Search ADS PubMed  32. Wu JH, Lemaitre RN, Imamura F, King IB, Song X, Spiegelman D, Siscovick DS, Mozaffarian D. Fatty acids in the de novo lipogenesis pathway and risk of coronary heart disease: the Cardiovascular Health Study. Am J Clin Nutr . 2011; 94( 2): 431– 438. Google Scholar CrossRef Search ADS PubMed  33. Lee JJ, Lambert JE, Hovhannisyan Y, Ramos-Roman MA, Trombold JR, Wagner DA, Parks EJ. Palmitoleic acid is elevated in fatty liver disease and reflects hepatic lipogenesis. Am J Clin Nutr . 2015; 101( 1): 34– 43. Google Scholar CrossRef Search ADS PubMed  34. Bézaire V, Langin D. Regulation of adipose tissue lipolysis revisited. Proc Nutr Soc . 2009; 68( 4): 350– 360. Google Scholar CrossRef Search ADS PubMed  35. Cao H, Gerhold K, Mayers JR, Wiest MM, Watkins SM, Hotamisligil GS. Identification of a lipokine, a lipid hormone linking adipose tissue to systemic metabolism. Cell . 2008; 134( 6): 933– 944. Google Scholar CrossRef Search ADS PubMed  36. Rustenbeck I, Lenzen S. Regulation of transmembrane ion transport by reaction products of phospholipase A2. II. Effects of arachidonic acid and other fatty acids on mitochondrial Ca2+ transport. Biochim Biophys Acta . 1989; 982( 1): 147– 155. Google Scholar CrossRef Search ADS PubMed  37. Criddle DN, Murphy J, Fistetto G, Barrow S, Tepikin AV, Neoptolemos JP, Sutton R, Petersen OH. Fatty acid ethyl esters cause pancreatic calcium toxicity via inositol trisphosphate receptors and loss of ATP synthesis. Gastroenterology . 2006; 130( 3): 781– 793. Google Scholar CrossRef Search ADS PubMed  38. Horas H Nababan S, Nishiumi S, Kawano Y, Kobayashi T, Yoshida M, Azuma T. Adrenic acid as an inflammation enhancer in non-alcoholic fatty liver disease. Arch Biochem Biophys . 2017; 623-624: 64– 75. Google Scholar CrossRef Search ADS PubMed  39. Sunny NE, Parks EJ, Browning JD, Burgess SC. Excessive hepatic mitochondrial TCA cycle and gluconeogenesis in humans with nonalcoholic fatty liver disease. Cell Metab . 2011; 14( 6): 804– 810. Google Scholar CrossRef Search ADS PubMed  40. Befroy DE, Perry RJ, Jain N, Dufour S, Cline GW, Trimmer JK, Brosnan J, Rothman DL, Petersen KF, Shulman GI. Direct assessment of hepatic mitochondrial oxidative and anaplerotic fluxes in humans using dynamic 13C magnetic resonance spectroscopy. Nat Med . 2014; 20( 1): 98– 102. Google Scholar CrossRef Search ADS PubMed  Copyright © 2018 Endocrine Society

Journal

Journal of Clinical Endocrinology and MetabolismOxford University Press

Published: Feb 1, 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 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

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