Short-Term Exercise Training Does Not Stimulate Skeletal Muscle ATP Synthesis in Relatives of Humans with Type 2 Diabetes

Diabetes Publish Ahead of Print, published online March 5, 2009

Short-term exercise training does not stimulate skeletal muscle ATP synthesis in relatives of humans with type 2 diabetes

Gertrud Kacerovsky-Bielesz1,2, Marek Chmelik2,3, Charlotte Ling4, Rochus Pokan5, Julia Szendroedi1,2, Michaela Farukuoye2, Michaela Kacerovsky2, Albrecht I. Schmid2,3, Stephan Gruber3, Michael Wolzt6, Ewald Moser3, Giovanni Pacini7, Gerhard Smekal5, Leif Groop4, Michael Roden1,2,8

1 1. Medical Department, Hanusch Hospital, Vienna, Austria, 2 Karl-Landsteiner Institute for Endocrinology and Metabolism, Vienna, Austria, 3 MR Center of Excellence, Medical University of Vienna, Austria, 4 Department of Clinical Sciences, Lund University, Malmö, Sweden, 5 Department of Sports and Exercise Physiology, University of Vienna, Austria, 6 Department of Clinical Pharmacology, Medical University of Vienna, Austria, 7 Metabolic Unit, Institute of Biomedical Engineering, CNR, Padova, Italy 8 Institute for Clinical Diabetology, German Diabetes Center - Leibniz Center for Diabetes Research, Department of Medicine/Metabolic Diseases, Heinrich Heine University Düsseldorf, Düsseldorf, Germany

Corresponding Author: Michael Roden, MD Email: [email protected]

Clinical trial reg. No. NCT00710008, clinicaltrials.gov.

Additional information for this article can be found in an online appendix at http://diabetes.diabetesjournals.org

Submitted 6 September 2008 and accepted 27 February 2009.

This is an uncopyedited electronic version of an article accepted for publication in Diabetes. The American Diabetes Association, publisher of Diabetes, is not responsible for any errors or omissions in this version of the manuscript or any version derived from it by third parties. The definitive publisher-authenticated version will be available in a future issue of Diabetes in print and online at http://diabetes.diabetesjournals.org.

Background. We tested the hypothesis that short-term exercise training improves hereditary insulin resistance by stimulating ATP synthesis and investigated associations with gene polymorphisms.

Methods. 24 nonobese first-degree relatives of type-2 diabetic patients and 12 controls were studied at rest and 48 hours after three bouts of exercise. In addition to measurements of oxygen uptake and insulin sensitivity (OGTT), ectopic lipids and mitochondrial ATP synthesis were assessed using 1H and 31P magnetic resonance spectroscopy (MRS), respectively. They were genotyped for polymorphisms in genes regulating mitochondrial function, PPARGC1A (rs8192678) and NDUFB6 (rs540467).

Results. Relatives had slightly lower (p=0.012) insulin sensitivity than controls. In controls, ATP synthase flux rose by 18% (p=0.0001) being 23% higher (p=0.002) than in relatives after exercise training. Relatives responding to exercise training with increased ATP synthesis (+19%, p=0.009) showed improved insulin sensitivity (p=0.009), whereas “non-responders” failed to increase their insulin sensitivity. A polymorphism in the NDUFB6 gene from respiratory-chain complex I related to ATP synthesis (p=0.02) and insulin sensitivity response to exercise training (p=0.05) ATP synthase flux correlated with O2 uptake and insulin sensitivity.

Conclusions. The ability of short-term exercise to stimulate ATP production distinguished individuals with improved insulin sensitivity from those who did not improve their insulin sensitivity. In addition, the NDUFB6 gene polymorphism appeared to modulate this adaptation. This suggests that genes involved in mitochondrial function contribute to the response of ATP synthesis to exercise training. This trial has been registered at ClinicalTrials.gov (NCT 00710008). Exercise training and ATP production

ife style intervention is the (18, 19). However, little is known on time recommended strategy for prevention course and onset of changes in glucose and L of type 2 diabetes mellitus (T2DM). energy metabolism independently of acute First-degree relatives of patients with T2DM exercise effects. (REL) have an increased risk of insulin We employed magnetic resonance resistance and T2DM (1, 2). Inherited and spectroscopy (MRS) to measure in vivo flux environmental factors cause insulin resistance of inorganic phosphate (Pi) to ATP through via intracellular lipid and inflammatory ATP synthase (fATPase) as well as IMCL, mediators which interfere with insulin HCL before and after three bouts of cycling signaling leading to an impaired rise of training to test the following hypotheses: (i) glucose-6-phosphate (G6P) due to reduced increased fATPase is an early event in the glucose transport/phosphorylation (2, 3). response to short term exercise training, (ii) These alterations can coexist with excessive responses of fATPase and insulin sensitivity storage of intramyocellular or hepatocellular are different in REL compared with healthy lipids (IMCL, HCL) and impaired controls (CON), and (iii) the responses are mitochondrial function and/or number in modulated by polymorphisms in genes which insulin resistant states such as aging (4), free are mutually linked to exercise capacity, fatty acids (FFA) elevation (5) and in some energy metabolism and insulin sensitivity in (6-8) but not all (9, 10) humans at risk of or epidemiologic studies. with T2DM. Non-diabetic REL suffering from severe insulin resistance present with METHODS elevated FFA, IMCL and HCL along with Volunteers. We recruited healthy REL of one impaired ATP synthesis possibly due to (n=19) or two (n=5) parents with T2DM, reduced mitochondrial contents (11). which was confirmed by hyperglycemia, oral Inherited and acquired factors associate with antidiabetic medication or insulin use. Twelve gene expression of the respiratory chain CON matched for sex, age, body mass index components, NDUFB6 and COX7A1, and (BMI) and physical activity were recruited as their transcriptional co-activators, PGC-1α/β, controls (Table 1). The participants which determine maximum oxygen uptake underwent medical history and physical (VO2max) and insulin action (12, 13). It examination. All participants were weight- remains unclear whether altered ATP stable over the last 6 months prior to the synthesis results from increased availability of study. None of them was smoking or lipid and/or adipokines such as adiponectin, regularly performing intense exercise. the nicotinamide phosphoribosyltransferase. Study protocols. Volunteers gave written visfatin, and retinol binding protein-4 (RBP4) informed consent to the study which was (14). approved by the institutional ethical board It is further uncertain whether such and performed according to the Declaration of abnormalities are reversible by exercising Helsinki. All participants remained on an and/or occur independently of effects on isocaloric diet, refrained from any physical insulin action. Long-term endurance exercise exercise for three days and fasted for 12 h training increases insulin sensitivity in before the start of the studies. On day 1, sedentary young and elderly (15), REL (16), participants underwent a frequent sampling glucose intolerant, obese or T2DM (17). 75-g OGTT and MRS. On day 2, they Exercise training for at least 4 weeks performed exercise testing. On days 3 and 5, enhances fat oxidation along with increased they exercised on a cycling ergometer. On mitochondrial mass and enzyme activities

Exercise training and ATP production day 7, measurements of day 1 were repeated collected before and in 30-min intervals for in identical fashion. 150 min for measurements of glucose, insulin Dietary assessment. Throughout the study, all and C-peptide. Dynamic insulin sensitivity volunteers were on a dietary plan reflecting was assessed by OGIS using the 120-min isocaloric diet in line with the American formula, which yields a measure of glucose Diabetes Association recommendations. clearance that has been widely exploited and Dietary intake over the last year and dietary validated against whole body insulin compliance were assessed with a modified sensitivity obtained from the euglycemic- interviewer-administered 107-item food hyperinsulinemic clamp (20). Beta cell frequency questionnaire adjusted for local function was assessed from the insulinogenic dietary habits index (21). (http://www.unihohenheim.de/wwwin140/inf Magnetic resonance spectroscopy (MRS). o/interaktives/foodfreq.htm). During the one- Participants were studied in a 3-T MR week intervention, volunteers counted all spectrometer (Bruker, Germany). A 10-cm food and beverages consumed employing circular double resonant 1H/31P surface coil common household measures to obtain six- was used for quantifying HCL and days dietary records. Nutrient/fluid intake on phosphorus metabolites. A 28-cm birdcage the days before the studies were analyzed coil was positioned over the right lower leg using 24-hour recalls. In the evenings before for measuring IMCL in soleus and tibialis the studies, participants consumed identically anterior (ant.) muscles. For non-localized 31P composed carbohydrate-enriched dinners at MRS, the right calf was positioned on the identical times. On study days, participants surface coil with the medial head of the right did not receive any calories except for the gastrocnemius muscle in the coil center. The OGTT until completion of the MRS integral of the region of phosphomonoesters measurements. (PME) covering G6P (7.1-7.4 ppm), Genotyping. Genomic DNA was extracted phosphodiesters (PDE), Pi and from blood of all participants by QIAamp phosphocreatine (PCr) were measured from DNA Blood Mini kit (Cat. No 51106, the ratio of integrated peak intensities and ß- Qiagen). Single nucleotide polymorphisms ATP resonance intensity in spectra without (SNPs), rs540467 of NDUFB6 and rs8192678 inversion and saturation assuming an ATP (Gly482Ser) of PPARGC1A, were genotyped concentration of 5.5 mmol/l (Figure 1). The using allelic discrimination assays performed assumption of constant ATP before and after with an ABI 7900 system (Applied exercise training was supported by unchanged Biosystems Inc.). The following assays were ATP/PDE ratios (data not shown), because used for rs540467 (Assay on demand, PDE levels remain constant under similar C_2334430, Applied Biosystems Inc.) and for conditions (22). Absolute quantification rs8192678, forward primer would be required to detect subtle changes of TGGAGAATTGTTCATTACTGAAATCAC myocellular ATP concentrations, which do TGT, reverse primer not necessarily reflect actual flux through GGTCATCCCAGTCAAGCTGTTTT ATP synthase. The saturation transfer together with two different probes Vic- experiment (selective irradiation of γ-ATP) CAAGACCGGTGAACTG and Fam- was employed to measure the exchange rate ACAAGACCAGTGAACTG, respectively. (k1) between Pi and ATP and to calculate Oral glucose tolerance test (OGTT). fATPase from k1x[Pi] (5, 8) (Figure 1). Participants drank a solution containing 75 g IMCL and HCL were determined within of glucose and venous blood samples were

Exercise training and ATP production volumes of interest of 1.73 cm3 (23) and 27 to exercise training among REL compared cm3 (5), respectively. with CON. REL were therefore divided into Habitual physical activity, exercise capacity two subgroups based on the difference of and training. Physical activity was assessed fATPase between baseline and after exercise by interviewer-administered questionnaire on training: responder (RESP, n=10) with a scale from one to five, low to high degree of (fATPase after – fATPase before) >0 and non- activity (24). Each participant performed an responder (NRES, n=14) with (fATPase after – incremental exhaustive exercise test on an fATPase before ) ≤0. electronically braked cycle ergometer (Lode Analytical procedures. Plasma glucose was Excalibur Sport, Groningen, Netherlands) at measured on a Glucose analyzer II (Beckman 70 revolutions/min. Respiratory gas exchange Coulter, http://www.beckmancoulter.com). measures were determined by open-air For FFA measurement (Wako Chem USA spirometry (Jäger/Viasys MasterScreen CPX, Inc., http://www.wakousa.com/), blood was Würzburg, Germany). The breath-by-breath collected into vials containing orlistat to measures were recorded and averaged over an prevent in vitro lipolysis. Lactate was interval of eight breath cycles. Heart rate was determined enzymatically (Roche, measured every 5 s (Polar Vantage NV http://www.roche.com/home.html), insulin telemetry, Polar Electro, Kempele, Finland). and C-peptide by double antibody RIA and, To regulate load intensities of training RBP-4 and visfatin in 7 controls and 8 sessions according to the individual aerobic relatives by ELISA (Phoenix Peptides, capacity, a ventilatory threshold (RCP) was Karlsruhe, Germany, determined (25), which marks the onset of www.phoenixpeptide.com/), with hyperventilation during incremental exercise inter/intraassay coefficients of variation of training mainly driven by the onset of lactic <6% (5). acidosis. RCP was determined independently Statistics. OGIS and fATPase were defined as by two investigators (intraclass correlation, primary endpoints. Sample size calculation r=0.9721) from: (i) the second upward was based on previous studies on fATPase inflection in VE (minute ventilation) curve using false positive (Zα=1.96, 2-tailed) and following the first break point at the anaerobic false negative error rates (Zß=0.84, 1-tailed), threshold, (ii) an upward inflection in the respectively. Data are presented as means±SD curve of VE/VCO2 (ventilatory equivalent for and 95% confidence intervals (text, tables) or CO2) and (iii) a downward inflection in the means±standard errors (figures). curve of PETCO2 (end-tidal volume for CO2) Unpaired/paired two-tailed t tests were used during the time course of respiratory gas for between- and within-group comparisons exchange measure variables. The two training as appropriate. Bivariate correlations were sessions (day 2, 5) consisted of 30 min in assessed with Pearson´s correlations three 10-min bouts separated by two 5-min coefficient. Multiple step-wise linear breaks with the training intensity set at 90-% regression analysis was performed for the work load determined at RCP. Thus, the dependent variables, fATPase and OGIS, three bouts of exercise training comprised of including BMI, WHR, age, triglycerides, the exercise testing and the two subsequent FFA, physical activity, VO2max, HCL, training sessions. All participants completed IMCL, caloric intake, and also for changes in exercise testing and all training sessions the above parameters during the study. All which were performed under continuous calculations were done using SPSS for supervision by a sports physiologist. Data Windows (version 8; SPSS Inc., Chicago, IL analysis revealed variable fATPase responses http://www.spss.com). Associations between

Exercise training and ATP production

genotypes and exercise responses for fATPase before and after training in both groups and OGIS were analyzed with the χ2-test (Table 3). (NCSS Statistical Software, Kaysville, UT, In subgroups, responding (RESP) or USA). not responding (NRES) to exercise with stimulation of fATPase, fasting glucose, RESULTS lactate, insulin, insulinogenic index C-peptide Clinical characteristics, exercise testing and (Table 3) and 2-h post-load plasma glucose dietary intake (data not shown) were comparable before and The groups showed similar sex distribution, after training. Plasma FFA were similar at age, fat mass and distribution (Table 1). Lab baseline and increased after exercise training tests, exercise performance and physical by 39% in NRES (p=0.008). OGIS was activity including its individual components comparable at baseline, but increased only in (data not shown) were not different between RESP upon exercise training (P=0.009) groups and subgroups. Energy and nutrient (Figure 2). intake were not different between groups Intramyocellular phosphorus metabolism and before and during the 24-hours prior to ectopic lipids. Baseline fATPase did not differ studies (Table 2). The six-days dietary between groups (CON: 12.0±2.2, REL: records confirmed similar nutrient 11.1±2.9 µmol.ml muscle-1.min-1) (Figure 3). composition (Supplementary Table 1). Body After training, fATPase increased by ~18% weights neither changed in CON (before: (P=0.0001) only in CON (14.2 ±2.5 µmol.ml 71.50±12.82 vs. after: 71.35±12.75 kg) nor muscle-1.min-1) and was ~23% (P=0.002) REL (73.46±13.54 vs. 73.06±13.22 kg). higher than in REL (10.9±3.0 µmol.ml Metabolites, hormones, insulin sensitivity and muscle-1.min-1) in whom it did not change secretion. Fasting plasma glucose did not upon exercise training (Figure 3). Similarly, differ between groups (Table 3). The 2-hours k1 rose in CON (p=0.001) but not in REL plasma glucose after oral glucose loading was (p=0.7). HCL tended to be higher in REL within the normal range, slightly higher in before training (CON: 1.5±1.0%, REL: REL before (6.3±1.4 vs. CON: 5.0±1.1 5.8±7.5%, p=0.08) (Figure 4), but were not mmol/l, p=0.009) but not after exercise different after training (CON: 1.8±1.0%, (4.8±1.2 vs. 5.5±1.2 mmol/l). Plasma FFA REL: 4.4±5.4%). IMCL (Figure 4) and G6P were comparable at baseline, but higher in (data not shown) were neither different REL after training (p=0.019). C-peptide was between groups nor affected by training. higher in REL before (P=0.006) and after Interestingly, baseline fATPase was exercise training (p=0.031). Fasting plasma higher (p=0.025) in NRES than RESP before insulin tended to be increased in REL. In exercise training, but by definition only subgroups, plasma RBP-4 (before: CON: increased in RESP by ~24% (p=0.009) and 0.36±0.12, REL: 0.41±0.07, after training decreased in NRES by ~16% (p<0.001). 0.36±0.08, 0.43±0.09 ng/ml) and visfatin Similarly, k1 rose in RES (p=0.01) and (before: CON: 39±26, REL: 29±9, after decreased in NRES (p=0.02). HCL and training 35±13, 26±6 ng/ml) did not differ. At IMCL in soleus muscle were comparable baseline, OGIS was 13% lower (p=0.012) in before and after the two training sessions. REL than in CON and slightly rose in both Baseline IMCL in tibialis anterior muscle was CON by 7% (p=0.05) and REL by 12% similar, but decreased by 35% (p=0.006) in (p=0.012) after training (Figure 2). QUICKI NRES after training (data not shown). and insulinogenic index were comparable Baseline fATPase related positively to VO2 (r=0.300, p=0.046) and power output

Exercise training and ATP production

(r=0.361, p=0.030) each at maximum and difference in OGIS is in line with the RCP. This also held true for fATPase after variability of insulin sensitivity in overweight exercise training (data not shown). Baseline (26), healthy humans and T2DM relatives fATPase correlated positively with OGIS (16) obtained with clamp tests. Severely (r=0.360, p=0.031). HCL related negatively to insulin-resistant offspring of T2DM exhibit OGIS (r=-0.675, p<0.001) and positively to increases in plasma FFA, body fat mass, plasma insulin (r=0.761, p<0.001). Multiple IMCL and HCL along with impaired insulin- regression analysis revealed that HCL and stimulated glucose transport/phosphorylation baseline caloric intake explained ~49% of the (11). Such individuals have ~30% lower variation of baseline OGIS (p=10-5). Changes fATPase along with ~40% lower muscle in fATPase (before vs. after exercise sessions) mitochondrial density than healthy humans. were neither related to changes in OGIS In our study, fATPase was similar to that of (r=0.001, p=0.993) nor to changes in caloric the present and previous controls (4, 5, 8) intake (before vs. mean caloric intake during suggesting that their ATP production suffices the intervention week; r=0.078, p=0.665) for non-exercising conditions. This is in which did not relate to each other, agreement with unchanged mRNA and respectively (r=-0.201, p=0.261). Multiple protein expression of PGC-1α, regulators of regression analysis did not identify any mitochondrial biogenesis and their anthropometric, dietary or lab parameters as downstream effectors in insulin-resistant independent predictors of the post-exercise relatives (11). However, comparable baseline changes in fATPase. fATPase between REL and CON does not Associations between SNPs in the NDUFB6 allow conclusions on differences in and PPARGC1A genes and exercise mitochondrial number/function, but might responses of fATPase and insulin sensitivity. simply reflect normal basal metabolic rate in The NDUFB6 SNP, rs540467 G/A, was both groups. Although overall mitochondrial associated with resistance to stimulation of function can be reduced in REL (11), it was fATPase and OGIS upon exercise training found to be normal in patients with T2DM at (Table 4). Exercise training increased baseline (7, 8, 27, 28) and only reduced upon fATPase in 74% of the G/G carriers, but only insulin stimulation (7, 8). Several groups have in 33% carrying the A-allele (p=0.02 for a further shown dissociation between the level dominant model). More G/G genotype of markers for mitochondrial content and in carriers (84%) increased OGIS after exercise vitro oxidative capacity. Only some (29, 30), training compared with A-allele carriers but not others (6, 9) found differences in (53%, p=0.05 for a dominant model). The mitochondrial function adjusted for PPARGC1A SNP (rs8192678, Gly482Ser) mitochondrial mass. neither related to fATPase nor OGIS This study demonstrates that short- responses to exercise training (Table 4). term exercise training (i) uniformly raises myocellular basal metabolic rate as assessed DISCUSSION with fATPase independently of insulin At baseline, REL and CON did not differ in sensitivity in healthy humans, (ii) does not parameters commonly interfering with insulin affect mean fATPase in REL, but (iii) sensitivity and fATPase including age, fat identifies subgroups of REL with different mass, physical activity and aerobic capacity. exercise responses. In some REL, these Nevertheless, REL were slightly less insulin conditions unmask a variation to increase sensitive based on OGIS which corresponds ATP production reflecting altered adaptation to M-values of ~10 mg.kg-1.min-1. The small of basal metabolic rate to short-term exercise

Exercise training and ATP production training. In the face of normal fasting glucose intensity exercise, whole body oxygen and FFA, hereditary factors seem to play an consumption returns to pre-exercise levels important role, as this variation in response to (38, 39) whereas both IMCL re-synthesis and exercise training was also seen in carriers of lipid oxidation continued to rise up to more the A-allele of the NDUFB6 SNP, rs540467, than 40 hours (38, 39). The post-exercise rise which has been associated with impaired in fATPase would thereby result from insulin action and T2DM risk (12). increased ATP demand for augmented lipid Several mechanisms could explain the turnover. Of note, a portion of the Pi/ATP rise in fTAPase in CON and in RESP. While exchange may be catalyzed by long-term exercise training possibly glyceraldehyde-3-phosphate dehydrogenase stimulates fat oxidation (18, 19) and electron and 3-phosphoglycerate kinase. Although the transport chain activity (31) through increased contribution of anaerobic glycolysis to mitochondrial size/number (19, 32), it is muscular ATP production is relevant during unlikely that mitochondrial biogenesis initial states of exercise and short-time high- contributed to the rise in fATPase after short intensity exercise, it ceases in the absence of duration of training. Of note, it is conceivable muscle contraction and during aerobic resting that even if mitochondrial capacity is conditions (40, 41). Of note, despite the impaired, putative regulators of oxidative known negative relationship between HCL phosphorylation could increase and maintain and baseline insulin sensitivity at baseline the balance between ATP demand and supply (14), the insulin sensitivity improved after and thereby baseline fATPase. We detected a exercise training despite no reductions in variation occurring despite normal basal IMCL and HCL. This is an important finding fATPase in insulin sensitive relatives. In of the study pointing to independent NRES, the decreased fATPase affected regulation of insulin sensitivity and ectopic adaptation of the myocellular basal metabolic lipids by short term exercise training. rate and could therefore reflect reduced need The absent fATPase response in or sufficient capability to supply ATP to meet NRES was recorded at 48 h after the last bout energy demands or result from impaired of exercise, when insulin secretion, G6P, pH oxidative ATP production (33). and IMCL were not altered in all groups. This study also shows that IMCLs Thus, acute effects occurring within the first were not different from baseline at 48 hours 24 h after exercise such as stimulation of after one week of three bouts of exercise. intramyocellular glucose Interestingly, NRES had higher plasma FFA transport/phosphorylation (42) and IMCL suggesting that increased lipid availability depletion (26, 37) unlikely contributed to could contribute to impaired fATPase (5), fATPase alterations. Of note, exercise training although this might alternatively lead to slightly but significantly stimulated insulin increased mitochondrial function and sensitivity in both groups in agreement with biogenesis as demonstrated in rodent models previous studied T2DM relatives (16). Insulin (34, 35). Fat oxidation may rise regardless of resistant relatives respond to 6-weeks insulin sensitivity even after one week of endurance training with increased insulin exercise training due to increased expression sensitivity and substrate oxidation (10, 16). of enzymes involved in lipid metabolism (36). On the other hand, combined diet and training At 48 hours after one bout of exercise, for 2 weeks was required to improve insulin insulin-stimulated glucose disposal (16, 26) sensitivity and IMCL in T2DM (43). and IMCL can be increased or unchanged (17, However, in the present study, insulin 37). Within 12 hours after moderate-to-high sensitivity failed to explain the fATPase

Exercise training and ATP production response suggesting that effects of short-term adaptations were related to changes in exercise training on mitochondrial function mitochondrial number, localization or and insulin sensitivity can be dissociated in function. In this context it is of note that skeletal muscle. Although one cannot exclude activation of PPARs might improve insulin that mitochondrial oxidation rises before sensitivity by transcriptional control of changes in insulin sensitivity in healthy mitochondrial function due to enhanced fatty humans, previous studies measuring muscular acid oxidation and regulation of aerobic metabolites and enzyme activities reported capacity. The Botnia study identified a that lactate, phosphocreatine hydrolysis and common variant in the PPARγ gene (P12A) glycogen depletion are reduced after 5 days of as one of the best predictors of future training whereas maximal succinate development of T2DM (48). Recent data dehydrogenase activity as a surrogate of provide evidence that even within 300 min mitochondrial function increased only after 31 after one bout of exercise expression and days (44). As insulin action was impaired in activation of PGC-1α, AMP-dependent endurance trained individuals who stop protein kinase phosphorylation, nuclear exercise training, despite unchanged oxidative respiratory factor-1 and cytochrome c oxidase capacity (45), other mechanisms than are increased in lean but not in insulin- oxidative capacity such as muscle glycogen resistant obese humans (49). Although, synthesis likely contribute to changes in previous studies suggest roles for PPARGC1A OGIS. At 48 hours of one bout of exercising, in oxidative phosphorylation, insulin action, overweight insulin resistant relatives VO2max and exercise response (12, 13, 47, improved their insulin-stimulated glucose 49), the present study did not find such disposal associated with increased glycogen associations for the PPARGC1A Gly482Ser synthesis (16). This suggests that increased gene variant probably due to lack of statistical glycogen storage mainly accounts for power. Also, longer duration of training might improved insulin action after exercise (46). be needed to reveal effects of this gene As our patients were not on hypercaloric diet, variant, as PPARGC1A Gly482Ser predicts their glycogen stores were unlikely endurance capacity (50). replenished to pre-exercise levels which is The present study has some known to be an important modulator of limitations. First, quantitative differences insulin sensitivity for at least 72 h after the from previous studies might be due to the last exercise session (46). exercise protocols, degree of insulin The A-allele carriers of the NDUFB6 resistance or the individual exercise capacity. SNP, rs540467, showed a variation in the To control for variations in exercise capacity, response to exercise in line with an inherited participants were exercising at 90% of their difference. NDUFB6 is among the genes RCP, the highest work load at which encoding the respiratory chain, which show oxidative phosphorylation is adequate for lower muscular expression in T2DM patients energy demand (51). Second, minor changes than controls (47). This SNP was also in amount and timing of caloric intake relative associated with resistance to an increase in to exercise could have modulated OGIS and OGIS after exercise training in line with a fATPase (52). But unchanged caloric intake reported relationship with glucose disposal and identical timing of meals render this and risk of T2DM (12). The present study possibility unlikely. Third, glucose ingestion further supports the important role for this during the OGTT before MRS measurements gene in ATP production in humans in vivo, could have differently affected fATPase. but cannot determine whether these However, the absence of differences in

Exercise training and ATP production

plasma glucose between groups at 150 min of short-term exercise training, nor do they have the OGTT and the start of MRS after further improved insulin sensitivity, and these same two hours and fATPase does not support such individuals carry a risk polymorphism in the effect. Fourth, fATPase could have been NDUFB6 gene. affected by alterations of ATP concentrations which can increase up to 18 h after exercise ACKNOWLEDGEMENTS (53). Such rise of ATP would have led to We would like to acknowledge the lower fATPase values thereby cooperation of Drs.Attila Brehm, Hannes underestimating the actual ATP synthase flux. Wondratsch, Herbert Dworak as well as Peter However, the observation of unchanged Nowotny. We are also grateful to Professor ATP/PDE ratios and the timing of Wilfried Grossmann, Faculty of Computer measurement at 48 h after the last bout of Science, University Vienna, Austria, for his exercise do not support alterations of ATP expert advice and critical analysis of the levels. Moreover, exchange rates, k1, which statistical evaluation of the data and to do not depend on substrate concentrations Margareta Svensson for extracting DNA. This behaved similarly as fATPase between and study was supported by grants from the within groups before and after training European Foundation for the Study of underlining that the observed differences Diabetes (EFSD, Novo Nordisk and GSK reflect variation in ATP synthase flux. Grants), Austrian Science Foundation Finally, no muscle biopsies were taken so that (P15656), Austrian National Bank (OENB neither mass-adjusted mitochondrial function 11459), Hochschuljubiläumsstiftung Vienna, nor possible changes in the expression of and unrestricted grants by Novo Nordisk and transcriptional modulators could be assessed. Baxter to MR, by grants from the Swedish Unchanged plasma RBP-4 and visfatin do not Research Council and the Wallenberg exclude altered expression/activity of these Foundation to LG, a Research Grant Award factors. by the Austrian Diabetes Association (ÖDG) In conclusion, training-induced to GKB and a grant from Regione Veneto increases of insulin sensitivity and ATP (Biotech DGR 2702/10-09-04) to GP. The synthesis indicate an important role of early funding bodies had no role in study design, adaptation of basal metabolic rate in healthy data collection and analysis, decision to humans. Conversely, some REL do not publish or preparation of the manuscript. stimulate ATP production in response to

REFERENCES

1. Eriksson, J., Franssila-Kallunki, A., Ekstrand, A., Saloranta, C., Widen, E., Schalin, C., and Groop, L. 1989. Early metabolic defects in persons at increased risk for non-insulin- dependent diabetes mellitus. N Engl J Med 321:337-343. 2. Roden, M., and Shulman, G.I. 1999. Applications of NMR spectroscopy to study muscle glycogen metabolism in man. Annu Rev Med 50:277-290. 3. Szendroedi, J., and Roden, M. 2008. Mitochondrial fitness and insulin sensitivity in humans. Diabetologia 51:2155-2167. 4. Petersen, K.F., Befroy, D., Dufour, S., Dziura, J., Ariyan, C., Rothman, D.L., DiPietro, L., Cline, G.W., and Shulman, G.I. 2003. Mitochondrial dysfunction in the elderly: possible role in insulin resistance. Science 300:1140-1142.

Exercise training and ATP production

5. Brehm, A., Krssak, M., Schmid, A.I., Nowotny, P., Waldhausl, W., and Roden, M. 2006. Increased lipid availability impairs insulin-stimulated ATP synthesis in human skeletal muscle. Diabetes 55:136-140. 6. Kelley, D.E., He, J., Menshikova, E.V., and Ritov, V.B. 2002. Dysfunction of mitochondria in human skeletal muscle in type 2 diabetes. Diabetes 51:2944-2950. 7. Stump, C.S., Short, K.R., Bigelow, M.L., Schimke, J.M., and Nair, K.S. 2003. Effect of insulin on human skeletal muscle mitochondrial ATP production, protein synthesis, and mRNA transcripts. Proc Natl Acad Sci U S A 100:7996-8001. 8. Szendroedi, J., Schmid, A.I., Chmelik, M., Toth, C., Brehm, A., Krssak, M., Nowotny, P., Wolzt, M., Waldhausl, W., and Roden, M. 2007. Muscle mitochondrial ATP synthesis and glucose transport/phosphorylation in type 2 diabetes. PLoS Med 4:e154. 9. Boushel, R., Gnaiger, E., Schjerling, P., Skovbro, M., Kraunsoe, R., and Dela, F. 2007. Patients with type 2 diabetes have normal mitochondrial function in skeletal muscle. Diabetologia 50:790-796. 10. Ostergard, T., Andersen, J.L., Nyholm, B., Lund, S., Nair, K.S., Saltin, B., and Schmitz, O. 2006. Impact of exercise training on insulin sensitivity, physical fitness, and muscle oxidative capacity in first-degree relatives of type 2 diabetic patients. Am J Physiol Endocrinol Metab 290:E998-1005. 11. Morino, K., Petersen, K.F., Dufour, S., Befroy, D., Frattini, J., Shatzkes, N., Neschen, S., White, M.F., Bilz, S., Sono, S., et al. 2005. Reduced mitochondrial density and increased IRS-1 serine phosphorylation in muscle of insulin-resistant offspring of type 2 diabetic parents. J Clin Invest 115:3587-3593. 12. Ling, C., Poulsen, P., Simonsson, S., Ronn, T., Holmkvist, J., Almgren, P., Hagert, P., Nilsson, E., Mabey, A.G., Nilsson, P., et al. 2007. Genetic and epigenetic factors are associated with expression of respiratory chain component NDUFB6 in human skeletal muscle. J Clin Invest 117:3427-3435. 13. Ling, C., Wegner, L., Andersen, G., Almgren, P., Hansen, T., Pedersen, O., Groop, L., Vaag, A., and Poulsen, P. 2007. Impact of the peroxisome proliferator activated receptor- gamma coactivator-1beta (PGC-1beta) Ala203Pro polymorphism on in vivo metabolism, PGC-1beta expression and fibre type composition in human skeletal muscle. Diabetologia 50:1615-1620. 14. Roden, M. 2006. Mechanisms of Disease: hepatic steatosis in type 2 diabetes-- pathogenesis and clinical relevance. Nat Clin Pract Endocrinol Metab 2:335-348. 15. Coggan, A.R., Spina, R.J., King, D.S., Rogers, M.A., Brown, M., Nemeth, P.M., and Holloszy, J.O. 1992. Skeletal muscle adaptations to endurance training in 60- to 70-yr-old men and women. J Appl Physiol 72:1780-1786. 16. Perseghin, G., Price, T.B., Petersen, K.F., Roden, M., Cline, G.W., Gerow, K., Rothman, D.L., and Shulman, G.I. 1996. Increased glucose transport-phosphorylation and muscle glycogen synthesis after exercise training in insulin-resistant subjects. N Engl J Med 335:1357-1362. 17. Kiens, B. 2006. Skeletal muscle lipid metabolism in exercise and insulin resistance. Physiol Rev 86:205-243. 18. Holloszy, J.O., and Coyle, E.F. 1984. Adaptations of skeletal muscle to endurance exercise and their metabolic consequences. J Appl Physiol 56:831-838.

Exercise training and ATP production

19. Menshikova, E.V., Ritov, V.B., Toledo, F.G., Ferrell, R.E., Goodpaster, B.H., and Kelley, D.E. 2005. Effects of weight loss and physical activity on skeletal muscle mitochondrial function in obesity. Am J Physiol Endocrinol Metab 288:E818-825. 20. Mari, A., Pacini, G., Murphy, E., Ludvik, B., and Nolan, J.J. 2001. A model-based method for assessing insulin sensitivity from the oral glucose tolerance test. Diabetes Care 24:539-548. 21. Seltzer, H.S., Allen, E.W., Herron, A.L., Jr., and Brennan, M.T. 1967. Insulin secretion in response to glycemic stimulus: relation of delayed initial release to carbohydrate intolerance in mild diabetes mellitus. J Clin Invest 46:323-335. 22. Newcomer, B.R., and Boska, M.D. 1997. Adenosine triphosphate production rates, metabolic economy calculations, pH, phosphomonoesters, phosphodiesters, and force output during short-duration maximal isometric plantar flexion exercises and repeated maximal isometric plantar flexion exercises. Muscle Nerve 20:336-346. 23. Krssak, M., Falk Petersen, K., Dresner, A., DiPietro, L., Vogel, S.M., Rothman, D.L., Roden, M., and Shulman, G.I. 1999. Intramyocellular lipid concentrations are correlated with insulin sensitivity in humans: a 1H NMR spectroscopy study. Diabetologia 42:113- 116. 24. Baecke, J.A., Burema, J., and Frijters, J.E. 1982. A short questionnaire for the measurement of habitual physical activity in epidemiological studies. Am J Clin Nutr 36:936-942. 25. Beaver, W.L., Wasserman, K., and Whipp, B.J. 1986. A new method for detecting anaerobic threshold by gas exchange. J Appl Physiol 60:2020-2027. 26. Jamurtas, A.Z., Theocharis, V., Koukoulis, G., Stakias, N., Fatouros, I.G., Kouretas, D., and Koutedakis, Y. 2006. The effects of acute exercise on serum adiponectin and resistin levels and their relation to insulin sensitivity in overweight males. Eur J Appl Physiol 97:122-126. 27. Schrauwen-Hinderling, V.B., Kooi, M.E., Hesselink, M.K., Jeneson, J.A., Backes, W.H., van Echteld, C.J., van Engelshoven, J.M., Mensink, M., and Schrauwen, P. 2007. Impaired in vivo mitochondrial function but similar intramyocellular lipid content in patients with type 2 diabetes mellitus and BMI-matched control subjects. Diabetologia 50:113-120. 28. Trenell, M.I., Hollingsworth, K.G., Lim, E.L., and Taylor, R. 2008. Increased daily walking improves lipid oxidation without changes in mitochondrial function in Type 2 diabetes. Diabetes Care. 29. Mogensen, M., Sahlin, K., Fernstrom, M., Glintborg, D., Vind, B.F., Beck-Nielsen, H., and Hojlund, K. 2007. Mitochondrial respiration is decreased in skeletal muscle of patients with type 2 diabetes. Diabetes 56:1592-1599. 30. Ritov, V.B., Menshikova, E.V., He, J., Ferrell, R.E., Goodpaster, B.H., and Kelley, D.E. 2005. Deficiency of subsarcolemmal mitochondria in obesity and type 2 diabetes. Diabetes 54:8-14. 31. Holloszy, J.O., Oscai, L.B., Don, I.J., and Mole, P.A. 1970. Mitochondrial citric acid cycle and related enzymes: adaptive response to exercise. Biochem Biophys Res Commun 40:1368-1373. 32. Toledo, F.G., Watkins, S., and Kelley, D.E. 2006. Changes induced by physical activity and weight loss in the morphology of intermyofibrillar mitochondria in obese men and women. J Clin Endocrinol Metab 91:3224-3227.

Exercise training and ATP production

33. Weiss, R.G., Gerstenblith, G., and Bottomley, P.A. 2005. ATP flux through creatine kinase in the normal, stressed, and failing human heart. Proc Natl Acad Sci U S A 102:808-813. 34. Hancock, C.R., Han, D.H., Chen, M., Terada, S., Yasuda, T., Wright, D.C., and Holloszy, J.O. 2008. High-fat diets cause insulin resistance despite an increase in muscle mitochondria. Proc Natl Acad Sci U S A 105:7815-7820. 35. Turner, N., Bruce, C.R., Beale, S.M., Hoehn, K.L., So, T., Rolph, M.S., and Cooney, G.J. 2007. Excess lipid availability increases mitochondrial fatty acid oxidative capacity in muscle: evidence against a role for reduced fatty acid oxidation in lipid-induced insulin resistance in rodents. Diabetes 56:2085-2092. 36. Putman, C.T., Jones, N.L., Hultman, E., Hollidge-Horvat, M.G., Bonen, A., McConachie, D.R., and Heigenhauser, G.J. 1998. Effects of short-term submaximal training in humans on muscle metabolism in exercise. Am J Physiol 275:E132-139. 37. Krssak, M., Petersen, K.F., Bergeron, R., Price, T., Laurent, D., Rothman, D.L., Roden, M., and Shulman, G.I. 2000. Intramuscular glycogen and intramyocellular lipid utilization during prolonged exercise and recovery in man: a 13C and 1H nuclear magnetic resonance spectroscopy study. J Clin Endocrinol Metab 85:748-754. 38. Kiens, B., and Richter, E.A. 1998. Utilization of skeletal muscle triacylglycerol during postexercise recovery in humans. Am J Physiol 275:E332-337. 39. Schenk, S., and Horowitz, J.F. 2007. Acute exercise increases triglyceride synthesis in skeletal muscle and prevents fatty acid-induced insulin resistance. J Clin Invest 117:1690-1698. 40. Quistorff, B., Johansen, L., and Sahlin, K. 1993. Absence of phosphocreatine resynthesis in human calf muscle during ischaemic recovery. Biochem J 291 (Pt 3):681-686. 41. Lanza, I.R., Wigmore, D.M., Befroy, D.E., and Kent-Braun, J.A. 2006. In vivo ATP production during free-flow and ischaemic muscle contractions in humans. J Physiol 577:353-367. 42. Price, T.B., Perseghin, G., Duleba, A., Chen, W., Chase, J., Rothman, D.L., Shulman, R.G., and Shulman, G.I. 1996. NMR studies of muscle glycogen synthesis in insulin- resistant offspring of parents with non-insulin-dependent diabetes mellitus immediately after glycogen-depleting exercise. Proc Natl Acad Sci U S A 93:5329-5334. 43. Tamura, Y., Tanaka, Y., Sato, F., Choi, J.B., Watada, H., Niwa, M., Kinoshita, J., Ooka, A., Kumashiro, N., Igarashi, Y., et al. 2005. Effects of diet and exercise on muscle and liver intracellular lipid contents and insulin sensitivity in type 2 diabetic patients. J Clin Endocrinol Metab 90:3191-3196. 44. Phillips, S.M., Green, H.J., Tarnopolsky, M.A., Heigenhauser, G.J., and Grant, S.M. 1996. Progressive effect of endurance training on metabolic adaptations in working skeletal muscle. Am J Physiol 270:E265-272. 45. Vukovich, M.D., Arciero, P.J., Kohrt, W.M., Racette, S.B., Hansen, P.A., and Holloszy, J.O. 1996. Changes in insulin action and GLUT-4 with 6 days of inactivity in endurance runners. J Appl Physiol 80:240-244. 46. Cartee, G.D., Young, D.A., Sleeper, M.D., Zierath, J., Wallberg-Henriksson, H., and Holloszy, J.O. 1989. Prolonged increase in insulin-stimulated glucose transport in muscle after exercise. Am J Physiol 256:E494-499. 47. Mootha, V.K., Lindgren, C.M., Eriksson, K.F., Subramanian, A., Sihag, S., Lehar, J., Puigserver, P., Carlsson, E., Ridderstrale, M., Laurila, E., et al. 2003. PGC-1alpha-

Exercise training and ATP production

responsive genes involved in oxidative phosphorylation are coordinately downregulated in human diabetes. Nat Genet 34:267-273. 48. Lyssenko, V., Almgren, P., Anevski, D., Orho-Melander, M., Sjogren, M., Saloranta, C., Tuomi, T., and Groop, L. 2005. Genetic prediction of future type 2 diabetes. PLoS Med 2:e345. 49. De Filippis, E., Alvarez, G., Berria, R., Cusi, K., Everman, S., Meyer, C., and Mandarino, L.J. 2008. Insulin-resistant muscle is exercise resistant: evidence for reduced response of nuclear-encoded mitochondrial genes to exercise. Am J Physiol Endocrinol Metab 294:E607-614. 50. Lucia, A., Gomez-Gallego, F., Barroso, I., Rabadan, M., Bandres, F., San Juan, A.F., Chicharro, J.L., Ekelund, U., Brage, S., Earnest, C.P., et al. 2005. PPARGC1A genotype (Gly482Ser) predicts exceptional endurance capacity in European men. J Appl Physiol 99:344-348. 51. Billat, V.L., Sirvent, P., Py, G., Koralsztein, J.P., and Mercier, J. 2003. The concept of maximal lactate steady state: a bridge between biochemistry, physiology and sport science. Sports Med 33:407-426. 52. Stephens, B.R., and Braun, B. 2008. Impact of nutrient intake timing on the metabolic response to exercise. Nutr Rev 66:473-476. 53. Kimber, N.E., Heigenhauser, G.J., Spriet, L.L., and Dyck, D.J. 2003. Skeletal muscle fat and carbohydrate metabolism during recovery from glycogen-depleting exercise in humans. J Physiol 548:919-927.

Exercise training and ATP production

Table 1. Baseline clinical and laboratory characteristics (means±SD, [95% CI]) of controls (CON) and first-degree relatives of type 2 diabetic patients (REL) and REL subgroups: responders (RESP) and non-responders (NRES) of ATP-synthesis to exercise training.

CON REL RESP NRES N (female/male) 12 (6/6) 24 (13/11) 10 (4/6) 14 (9/5) Age (years) 37±11, [30-44] 40±12, [35-45] 41±15, [31-52] 39±11, [33-46] Body weight (kg) 71±13, [63-79] 73±14, [68-79] 76±14, [65-86] 72±13, [64-79] Body mass index (kg/m2) 23±2, [22-25] 25±4, [24-27] 26±4, [23-29] 25±4, [22-27] Waist-to-hip ratio 0.83±0.08, [0.78-0.88] 0.85±0.05, [0.83-0.87] 0.87±0.04, [0.84-0.90] 0.84±0.05, [0.81-0.87] Systolic blood pressure (mmHg) 124±19, [112-135] 126±11, [121-130] 127±11, [119-135] 125±12, [118-132] Diastolic blood pressure (mmHg) 82±10, [76-88] 83±8, [79-86] 82±8, [76-88] 84±8, [79-88] Hemoglobin A1c (%) 5.3±0.3, [5.1-5.5] 5.4±0.3, [5.3-5.6] 5.5±0.4, [5.2-5.7] 5.4±0.3, [5.3-5.6] Triglycerides (mmol/l) 0.87±0.50, [0.53-1.20] 1.23±0.70, [0.94-1.53] 1.32±0.57, [0.91-1.73] 1.18±0.80, [0.72-1.64] HDL-cholesterol (mmol/l) 1.7±0.45, [1.4-2.0] 1.5±0.4, [1.3-1.6] 1.4±0.4, [1.1-1.7] 1.5±0.4, [1.3-1.7] LDL-cholesterol (mmol/l) 3.0±0.7, [2.5-3.4] 3.5±0.9, [3.1-3.9] 3.8±0.9, [3.2-4.4] 3.3±1.0, [2.8-3.9] Physical activity (scale 1 to 5) 2.7±0.3, [2.5-2.9] 2.9±0.3, [2.7-3.0] 2.9±0.3, [2.6-3.1] 2.9±0.3, [2.7-3.1] -1 -1 VO2max (ml.kg .min ) 33.6±6.0, [29.8-37.5] 30.6±6.4, [27.8-33.3] 29.6±5.5, [25.7-33.6] 31.2±7.1, [27.1-35.3] -1 -1 VO2RCP (ml.kg .min ) 26.2±4.8, [23.1-29.2] 22.8±5.7, [20.4-25.2] 22.1±4.7, [18.7-25.4] 23.3±6.4, [19.6-27.0]

Exercise training and ATP production

Table 2. Nutrient intake obtained from 24-hours dietary recalls (means±SD, [95% CI]) in controls (CON) and first-degree relatives of type 2 diabetic patients (REL). Dietary data after training were not available from one relative who lost the dietary record. There were no statistical differences within and between groups. CON REL Before After training Before After training Caloric intake (kcal/day) 2274±843, 2045±904, 1937±614, 1820±557, [1738-2809] [1470-2619] [1671-2203] [1579-2061] % of daily energy intake: Carbohydrate (%) 39±11, [32-46] 43±10, [37-50] 44±12, [39-49] 44±9, [40-48] Fat (%) 40±10, [34-47] 37±8, [32-42] 37±10, [33-41] 35±8, [31-38] Protein (%) 17±6, [14-21] 17±6, [13-20] 17±6, [15-20] 19±5, [16-21] Saturated fat (%) 18±6, [14-21] 17±4, [15-19] 16±5, [14-18] 15±5, [13-17] n-3 fatty acids (%) 1±0, [1-2] 1±1, [1-2] 1±1, [1-1] 1±0, [1-1] n-6 fatty acids (%) 5±3, [3-7] 4±2, [3-5] 5±2, [4-6] 6±3, [5-7] Cholesterol (g/day) 0.38±0.17,[0.27-0.50] 0.38±0.20,[0.26-0.51] 0.34±0.22, [0.25-0.44] 0.30±0.17, [0.22-0.37]

Table 3. First-degree relatives of type 2 diabetic patients (REL) and of responder (RESP) and non-respondern (NRES) of ATP synthesis to exercise training.*P<0.05 before vs. after exercise, ** P<0.05 CON vs. REL at baseline, §P<0.05 CON vs. REL after exercise training. Before exercise After Exercise Before exercise After Exercise Glucose (mmol/l) CON 4.9±0.3, [4.7-5.1] 4.9±0.3, [4.7-5.1] RESP 5.0±0.4, [4.8-5.3] 5.0±0.4, [4.7-5.3] REL 5.0±0.4, [4.8-5.1] 5.0±0.4, [4.8-5.1] NRES 4.9±0.4, [4.7-5.2] 5.0±0.5, [4.7-5.2] Lactate (mmol/l) CON 1.1±0.3, [0.9-1.3] 1.0±0.2*, [0.8-1.1] RESP 1.3±0.5,[1.0-1.7] 1.2±0.3, [0.9-1.4] REL 1.5±0.5**, [1.3-1.7] 1.3±0.5*, [1.0-1.5] NRES 1.6±0.4, [1.3-1.8] 1.3±0.6, [1.0-1.6] Free fatty acids (µmol/l) CON 411±190, [290-531] 502±154, [404-600] RESP 488±195, [348-628] 510±204, [364-655] REL 443±160, [376-511] 540±204*, [450-631] NRES 408±128, [338-486] 566±209*, [434-699] Insulin (pmol/l) CON 27±8, [22-32] 23±7, [19-27] RESP 36±22, [21-52] 36±20, [22-50] REL 41±34, [27-55] 39±31, [26-52] NRES 45±41, [22-69] 40±37, [19-62] C-peptide (nmol/l) CON 1.6±0.4, [1.3-1.8] 1.5±0.4, [1.3-1.8] RESP 2.4±0.8, [1.9-3.0] 2.3±1.0, [1.6-3.0] REL 2.3±0.8**, [1.9-2.6] 2.2±1.0§, [1.8-2.6] NRES 2.2±0.8, [1.7-2.7] 2.1±1.0, [1.6-2.7] QUICKI CON 0.51±0.05, [0.48-0.54] 0.52±0.04, [0.50-0.54] RESP 0.48±0.06, [0.44-0.52] 0.48±0.05, [0.44-0.51] REL 0.47±0.07, [0.45-0.50] 0.48±0.07, [0.46-0.51] NRES 0.47±0.07, [0.43-0.51] 0.49±0.08, [0.45-0.53] Insulinogenic index CON 4.3±1.5, [3.3-5.2] 6.1±4.2, [3.4-8.8] RESP 4.3±2.7, [2.4-6.2] 5.0±3.6, [2.4-7.5] REL 4.8±2.4, [3.8-5.9] 6.0±4.7, [4.0-8.0] NRES 5.2±2.2, [4.0-6.5] 6.8±5.4, [3.7-9.9]

Exercise training and ATP production

Table 4. Association between SNPs in the NDUFB6 (rs540467) and PPARGC1A (rs8192678, Gly482Ser) genes and response to exercise training regarding stimulation of ATP synthesis (fATPase) and dynamic insulin sensitivity (OGIS) in responders and non-responders. Chi-square tests were performed to analyze associations between genotypes and response to exercise for fATPase and OGIS.

NDUFB6 rs540467 G/G G/A A/A P n n n fATPase Responders 14 4 1 0.02* Nonresponders 5 9 1 OGIS Responders 16 7 1 0.05* Nonresponders 3 6 1

PPARGC1A rs8192678 Gly/Gly Gly/Ser Ser/Ser P n n N fATPase Responders 9 10 0 0.24 Nonresponders 7 6 2 OGIS Responders 11 11 2 0.64 Nonresponders 5 5 0 * p value indicating significance for a dominant model.

Exercise training and ATP production

Figure Legends

Figure 1. 31P magnetic resonance spectrum from acquired at 3 Tesla employing a surface coil (TR = 15 s, NS = 16) positioned under the calf muscle of one participant. The spectrum shows intramyocellular phosphomonoesters (PME) including glucose-6-phosphate (G6P), inorganic phosphate (Pi), phosphodiesters (PDE), phosphocreatine (PCr) and adenosine-triphosphate (ATP). Insert: 31P spectra with saturation of γ-ATP (bottom) and with saturation mirrored around Pi (top), which was always used to account and correct for direct saturation of the resonance frequency pulse.

Figure 2. Dynamic insulin sensitivity as assessed from the oral glucose tolerance test (OGIS) in individuals without (CON, n=12) or with (REL, n=24) first-degree relatives with type 2 diabetes and in REL-subgroups responding (RESP, n=10) or not responding (NRES, n=14) with increased ATP-synthesis after exercise training sessions. Black horizontal bars indicate mean values of the respective groups.

* p=0.049 CON before vs. after, ** p=0.012 REL before vs. after, † p=0.009 RESP before vs. after, ‡ p=0.012 CON vs. REL before, § p=0.003, CON vs. RESP before, $ p=0.031 CON vs. RESP after exercise

Figure 3. Flux through skeletal muscle ATP synthase (fATPase) in individuals without (CON, n=12) or with (REL, n=24) first-degree relatives with type 2 diabetes and in REL-subgroups responding (RESP, n=10) or not responding (NRES, n=14) with increased ATP-synthesis after exercise training sessions. Black horizontal bars indicate mean values of the respective groups.

* p<0.001 CON and NRES before vs. after, ** p=0.002, CON vs. REL after, † p=0.010 CON vs. RESP before, $ p=0.009 RESP before vs. after, § p=0.024 RESP vs. NRES before, ‡ p<0.001 CON vs. NRES after exercise

Figure 4. Absolute changes (delta; means±SEM) in dynamic insulin sensitivity (OGIS), flux through ATP synthase (fATPase) and lipid concentrations in liver (HCL) and soleus muscle (IMCL) in individuals without (CON, n=12) or with (REL, n=24) first-degree relatives with type 2 diabetes and in REL subgroups responding (RESP, n=10) or not responding (NRES, n=14) with increased ATP synthesis after exercise training sessions.

* p=0.005 CON vs. REL, ** p<0.001 RESP vs. NRES, § p<0.001 CON vs. NRES, † p=0.014 REL vs. RESP, $ p=0.024 REL vs. NRES

Exercise training and ATP production

Figure 1. PCr

Pi PCr

PDE

PME Saturation γ−ΑΤΠ α-ATP β−ΑΤΠ Pi PME PDE

Saturation ∆Pi

γ−ΑΤΠ α -ATP β−ΑΤΠ

ppm 10 5 0 -5 -10 -15

Figure 2.

Dynamic insulin sensitivity (OGIS)

800 ** * ‡

-2 600 †$ .m -1 400 ml.min

200

0 CON REL RES NRES before after before afterbefore after before after exercise

Exercise training and ATP production

Figure 3.

ATP synthetic flux (fATPase) 20 ***† $ * ‡

16 -1

.min 12 -1

8 µmol.ml 4

0 CON REL RES NRES before after before afterbefore after before after exercise

Figure 4.

80 Δ OGIS Δ fATPase

60 -2 4

.m † 1 $ - 40 *** -1 2 .min

ml.min 20 -1

0 0 µmol.ml

-2 Δ HCL Δ IMCL

0.0 0.4

-1.0 0.2

-2.0 0.0 % water signal % water % water signal % water

-3.0 -0.2 CON REL RES NRES CON REL RES NRES