Εφαρμογή ολιστικών αναλυτικών τεχνολογιών για τη μελέτη της επίδρασης της άσκησης στο μεταβολικό προφίλ

Abstract

Exercise and regular physical activity, being powerful modifiers of animal metabolism, exertnumerous beneficial effects, including reduced risk of diseases, beneficial effects on already existingdiseases and even enhanced longevity. So far, biochemichal studies on the effects of exercisefocused mainly on a small number of molecules in specific, well-characterized biochemical pathways.In this study we tried to incorporate the non-targeted holistic approach of metabonomics in the fieldof exercise biochemistry. Metabonomics, having matured over the last few years are now apowerfull tool in the investigation of the effects of a series of perturbations in a number of biologicalsystems. NMR, LC-MS and GC-MS, along with developments in data extraction and handling, are thehorsepower in the quest for a better view of systems biology through metabonomics.In the first study of the present thesis, the effects of exhaustive exercise, in our case swimming, onrat metabolism were investigated. Blood plasma and liver samples were analyzed through LC-MS,and blood and muscle samples were analyzed through biochemical assays. Nonetheless, due to poorexperimental design, the study was abandoned since the results of the preliminary phases werediscouraging.In the second study, 1H NMR-based metabonomics provided useful information for theunderstanding of metabolic changes induced by specific training schedules. Urine samples obtainedafter exercise protocols differing in as little as the rest interval between repeated sprints, not thesprints themselves, can be classified and safely predicted even when applying unsupervisedstatistical methods of analysis. In this way, important biomolecules involved in exercise biochemistrycan be identified and further studied. Separation of pre- from post-exercise samples was attributedmainly to lactate, pyruvate, hypoxanthine, compounds of the Krebs cycle, amino acids, products ofBCAA catabolism, 2-hydroxybutyrate, and hippurate. Most of these metabolites increased in urinewith exercise and have been described to also increase in muscle with exercise. Separation of the 10s from the 1 min rest interval was attributed mainly to lactate, pyruvate, alanine, compounds of theKrebs cycle, 2-oxoacids of BCAA, and 2-hydroxybutyrate. All of these metabolites increased morewith the short compared to the long interval, thus supporting the hypothesis that the former elicitedgreater metabolic disturbances than the latter as a result of the very limited time available forrecovery. The facts that such methodology can be applied to urine, a biological material readilyavailable and noninvasively obtained, and that urine reflects many of the exercise-induced changesoccurring in muscle are additional advantages. The third study, being an extension of the previous work on urine, reaffirms the power of themetabonomic approach in the study of exercise biochemistry. Serum samples acquired pre- andpost-exercise, as well as pre- and post-training, enabled us to classify and predict the effects ofexercise and training by using multivariate chemometric techniques. Separation of serum samplesaccording to the rest interval employed between repeated sprints was not possible (contrary to ourfindings in urine), possibly because serum is an equilibrium-regulated biofluid where differences inmetabolite concentrations can be mitigated by metabolite transitions between different tissues. Themain metabolites increasing after exercise were lactate, pyruvate, alanine, and an unidentifiedmetabolite resonating at 8.17 ppm. Lower concentrations after exercise were exhibited by leucine,valine, isoleucine, arginine/lysine and glycoprotein acetyls. Training induced increases inmethylguanidine, citrate, glucose, valine, taurine, trimethylamine N-oxide, choline-containingcompounds, histidines (including histidine, 1-methylhistidine, and 3-methylhistidine), andacetoacetate/acetone. Lactate, pyruvate, glycoprotein acetyls, and lipids (in general) decreased aftertraining. These findings provide further evidence for the utility of metabonomics in providing aholistic view of the short- and long-term impact of exercise on human metabolism and in identifyingnovel biomarkers of the body’s response to exercise.The fourth study describes the development of a GC-MS protocol for the untargeted analysis ofblood plasma to serve in the metabolomic study of the effects of exercise and allopurinoladministration. We describe a strategy for data treatment that employs advanced software utilities(GAVIN, Fiehn spectral library), followed by thorough statistical analysis to ensure the validity of thefindings. The current state of the art in untargeted analysis and especially in GC-MS, which employssample derivatisation, necessitates meticulous data treatment and utilization of different validationtools such as the use of QC samples. Following this approach, a number of metabolites were foundto be major differentiators between the different sample groups. Lactic and pyruvic acids, indicatingincreased carbohydrate breakdown, as well as 2-hydroxybutyric and pyroglutamic acids, indicatingincreased glutathione synthesis in response to oxidative stress, were among the majordifferentiators of exercise from the resting state. Inosine, hypoxanthine, xanthine, xanthosine anduric acid, indicating xanthine oxidase inhibition, as well as methionine, proline and leucine, indicatingincreased protein synthesis, were among the major differentiators of allopurinol administration fromplacebo. Finally, the metabolic responses to exercise were not affected by allopurinoladministration.By using holistic metabonomic analysis with LC-MS, we were able to demonstrate the repeatabilityof the metabolic fingerprint of urine before and after two swimming tests at maximum intensity. Our findings are in partial agreement with the findings of a previous work with the same samples, wherebiochemical assays for lactic acid in both the urine and the blood of athletes showed significantrepeatability in the 3 x 100 m test but not the 6 x 50 m test. Another characteristic was the inabilityto separate samples from the two different exercise protocols. This can lead to the conclusion thatthe two protocols did not cause different changes in the metabolic profile. The energy sources usedfor the two protocols are very similar, mainly from the lactic acid energy system. Also the restintervals between successive swims (5 and 10 minutes) is deemed to be sufficient to replenish theenergy systems and so each new effort starts from the same energy level.In the last study we utilized RP-UPLC-MS and NMR to study the effect of acute exercise on the urinemetabolome of young untrained men. The exercise was performed in two identical test sessions,spaced at least 3 days apart. We were able to confirm the reproducibility of these tests and todetect, identify and monitor the concentrations of several metabolites for a period of two hoursafter exercise. With both platforms we were able to identify compounds of the Krebs cycle, BCAAcatabolism, aminoacids and purine metabolism as major discriminators of exercise from rest.Interestingly, hypoxanthine was the only compound showing elevated values two hours after theend of the exercise.