Targeted Metabolomics in High Performance Sports: Differences between the Resting Metabolic Profile of Endurance- and Strength-Trained Athletes in Comparison with Sedentary Subjects over the Course of a Training Year

tle is known about the metabolic differences between endurance and strength ath- letes in comparison with sedentary subjects under controlled conditions and about variation of the metabolome throughout one year. We hypothesized that (1) the resting metabolic pro le dif- fers between sedentary subjects and athletes and between perennially endurance- and strength- trained athletes and (2) varies throughout one year of training. We performed quantitative, targeted metabolomics (Biocrates MxP ® Quant 500, Biocrates Life Sciences AG, Innsbruck, Austria) in plasma samples at rest in three groups of male adults, 12 strength-trained (weightlifters, 20 3 years), 10 endurance-trained athletes (runners, 24 3 years), and 12 sedentary subjects (25 4 years) at the end of three training phases (regeneration, preparation, and competition) within one training year. Performance and anthropometric data showed signi cant(p< 0.05) differences between the groups. Metabolomic analysis revealed different resting metabolic pro les between the groups with acetylcarnitines, di-

strength-trained (weightlifters, 20 3 years), 10 endurance-trained athletes (runners, 24 3 years), and 12 sedentary subjects (25 4 years) at the end of three training phases (regeneration, preparation, and competition) within one training year. Performance and anthropometric data showed signi cant(p< 0.05) differences between the groups. Metabolomic analysis revealed different resting metabolic pro les between the groups with acetylcarnitines, di- and triacylglycerols, and glycerophospho- and sphingolipids, as well as several amino acids as the most robust metabolites. Furthermore, we observed changes in free carnitine and 3-methylhistidine in strength-trained athletes throughout the training year. Regular endurance or strength training induces changes in the concentration of several metabolites associated with adaptations of the mitochondrial energy and glycolytic metabolism with concomitant changes in amino acid metabolism and cell signaling. Keywords:athletes; phenotype; athlete metabolome; basal state; chronic adaptation; plasma 1. Introduction Regular training performed for many years leads to the development of characteristic athletic phenotypes depending on the training content and the combination of intensity and volume, provided that the genetic pro le is appropriate [1–3]. On one side of the exercise continuum, endurance training induces an increase in oxygen uptake, oxygen-transport capacity, and in oxidative metabolism with enhanced mitochondrial density in rather small muscles containing predominantly oxidative type I myo bers. These adaptations result in a greater resistance to fatigue and an enhanced endurance performance [4–7]. On the other side of the spectrum, resistance training leads to an increase in muscle mass with predominantly fast type II myo bers, favoring anaerobic and anabolic metabolism, and to an optimization of neuromuscular function [8,9], adaptations which enable maximal and Metabolites2023,13, 833.

Metabolites2023,13, 833 2 of 16 fast strength development. The results of a few recent studies suggest that the adaptation to extreme forms of endurance or strength training leads to characteristic metabolic pro les which can be elucidated by applying metabolomics, allowing for the simultaneous analysis of low-molecular metabolic compounds [10–13]. Little information is available regarding long-term adaptation of the basal metabolome in chronically trained individuals, especially in basal state [11,14]. To the best of our knowl- edge, there is only one landmark study which investigated metabolic differences between athletes from different sport disciplines [12]. At the time when 191 athletes reported to anti-doping controls, blood samples were obtained for the measurement of 743 metabolites. The samples were collected as part of anti-doping controls in or out of competition with a lack of information about the athletes' age, ethnicity, body mass, or their current training status, all factors with probably considerable effects on the individual metabolome. During the past years, research focused on changes of the basal metabolome of endurance-trained athletes and athletes of different sports in response to a single sport-speci c session or multiple training weeks [15–20]. However, none of these studies investigated the effects of perennial training on the resting metabolome. The aim of the present study was to investigate the differences in the resting plasma metabolome between endurance-trained athletes, strength-trained athletes, and sedentary subjects. A secondary objective was the assessment of changes in the metabolome over a one-year period involving different training and rest periods. We hypothesized that the resting plasma metabolome would exhibit distinct and signi cant characteristics between endurance-trained athletes, strength-trained athletes, and sedentary subjects. Furthermore, we expected signi cant changes in the metabolome to occur throughout the one-year study period, re ecting the in uence of different training periods and regeneration phases. Therefore, the study includes several novel elements: (i) a comparison of the plasma metabolome of the physiological extremes under controlled conditions in (ii) a longitudinal study design throughout different training and rest periods in high-performance sports applying (iii) targeted, quantitative ultra-performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) using the MxP Quant 500

uence of different training periods and regeneration phases. Therefore, the study includes several novel elements: (i) a comparison of the plasma metabolome of the physiological extremes under controlled conditions in (ii) a longitudinal study design throughout different training and rest periods in high-performance sports applying (iii) targeted, quantitative ultra-performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) using the MxP Quant 500 kit (Biocrates Life Sciences AG, Innsbruck, Austria). 2. Materials and Methods 2.1. Participants Three groups of male adults, strength-trained athletes (ST;n= 12, age:20.2 2.6 years, mass: 80.3 13.0 kg, height: 175.0 8.7 cm, BMI: 26.0 2.5 kg/m 2 , body fat:10.6 4.3%, VO2max: 42.5 4.7 mL min 1 kg 1 ), endurance-trained athletes (ET;n= 10,age: 24.0 2.9 years, mass: 66.5 9.2 kg, height: 176.8 6.8 cm, BMI: 21.2 2.1 kg m 2 , body fat: 8.3 1.4%, VO2max: 65.1 4.8 mL min 1 kg 1 ), and a group of sedentary sub- jects (CG;n= 12, aged 24.8 4.2 years, mass: 81.9 18.4 kg, height: 180.8 8.8 cm, BMI: 25.2 6.1 kg m 2 , body fat: 14.9 6.5%, VO2max: 41.2 6.5 mL min 1 kg 1 ) were inves- tigated. The strength-trained group included weightlifters of the German Junior National Team (n= 8) and regional weightlifters (n= 4), practicing competitive sports for8 3 years at national and international levels. The endurance-trained group consisted of track and eld runners with a specialization in middle- and long-distance running (800–10,000 m) as well as in marathon running, practicing competitive sports for9 4 yearsat regional and national levels. The control group included non-active, healthy participants who never underwent periodized training and who were unexperienced in endurance and resistance training. 2.2. Experimental Design Participants attended the laboratory on four separate occasions, the preliminary testing and three laboratory visits. Figure mental design. All participants received a medical examination before the preliminary testing. Participants were excluded if they presented evidence of musculoskeletal disor-

Metabolites2023,13, 833 3 of 16 ders (e.g., arthrosis, spondylarthrosis, spinal deformities), cardiovascular diseases (e.g., coronary heart disease, hypertension, cardiac arrythmias), acute or chronic diseases (e.g., infectious diseases, muscle diseases), or diseases related to obesity (e.g., type 2 diabetes mellitus, metabolic syndrome), as well as coagulation activity disorders and if they took regular medication.Metabolites 2023, 13, x FOR PEER REVIEW 3 of 17 testing. Participants were excluded if they presented evidence of musculoskeletal disor- ders (e.g., arthrosis, spondylarthrosis, spinal deformities), cardiovascular diseases (e.g., coronary heart disease, hypertension, cardiac arrythmias), acute or chronic diseases (e.g., infectious diseases, muscle diseases), or diseases related to obesity (e.g., type 2 diabetes mellitus, metabolic syndrome), as well as coagulation activity disorders and if they took regular medication. Figure 1. Schematic overview of the study design. (a) Chronically strength-trained (weightlifters, ST, 12 males) and endurance-trained athletes (runners, ET, 10 males) as well as healthy sedentary subjects (control group, CG, 12 males) were recruited. (b) Dietary control 24 h pre-laboratory visits. Subjects had to refrain from drugs, alcohol, nicotine, caffeine, and supplements 48 h before each visit. They had to avoid intense physical activity or training 24 h before each visit. All participants consumed a standardized dinner the day before each visit and a standardized breakfast on each visit. All participants performed three laboratory visits within one training year each during the periods of regeneration, preparation, and competition. Venous blood samples were taken from the forearm vein 60 min postprandial. (c) Sample treatment and analysis were performed by UPLC- MS/MS and MetIDQ Software according to Biocrates MxP Quant 500 kit workflow. Data treatment and statistical analysis were performed in the software R (version 4.0.3, R Core Team, Vienna, Aus- tria) and MetaboAnalyst 4.0 (https://www.metaboanalyst.ca). The figure was created with BioRen- der.com. 2.3. Preliminary Testing Body height and mass were measured using a standard stadiometer and a calibrated scale (Seca, Hamburg, Germany), respectively. The percentage of body fat was calculated from skinfold thickness measurements (Holtain, Crymych, UK) at 3 sites [21]. All partici- pants performed an incremental exercise test to exhaustion either on a treadmill (ET and Figure 1.

with BioRen- der.com. 2.3. Preliminary Testing Body height and mass were measured using a standard stadiometer and a calibrated scale (Seca, Hamburg, Germany), respectively. The percentage of body fat was calculated from skinfold thickness measurements (Holtain, Crymych, UK) at 3 sites [21]. All partici- pants performed an incremental exercise test to exhaustion either on a treadmill (ET and Figure 1. Schematic overview of the study design. (a) Chronically strength-trained (weightlifters, ST, 12 males) and endurance-trained athletes (runners, ET, 10 males) as well as healthy sedentary subjects (control group, CG, 12 males) were recruited. (b) Dietary control 24 h pre-laboratory visits. Subjects had to refrain from drugs, alcohol, nicotine, caffeine, and supplements 48 h before each visit. They had to avoid intense physical activity or training 24 h before each visit. All participants consumed a standardized dinner the day before each visit and a standardized breakfast on each visit. All participants performed three laboratory visits within one training year each during the periods of regeneration, preparation, and competition. Venous blood samples were taken from the forearm vein 60 min postprandial. (c) Sample treatment and analysis were performed by UPLC-MS/MS and MetIDQ Software according to Biocrates MxP Quant 500 kit work ow. Data treatment and statistical analysis were performed in the software R (version 4.0.3, R Core Team, Vienna, Austria) and MetaboAnalyst 4.0 (https://www.metaboanalyst.ca). The gure was created with. 2.3. Preliminary Testing Body height and mass were measured using a standard stadiometer and a calibrated scale (Seca, Hamburg, Germany), respectively. The percentage of body fat was calculated from skinfold thickness measurements (Holtain, Crymych, UK) at 3 sites [21]. All partici- pants performed an incremental exercise test to exhaustion either on a treadmill (ET and CG, ELG70, Woodway USA Inc., Waukesha, WI, USA) or on a cycle ergometer (ST, Excalibur

Metabolites2023,13, 833 4 of 16 Sport, Lode BV Medical Technology, Groningen, The Netherlands) to assess cardiorespira- tory performance. Treadmill spiroergometry started with a 1-min warm-up at 4 km h 1 . Then, running velocity (start at 6 km h 1 , incline 1.5%) was increased by2 km h 1 every 3 min until volitional exhaustion, followed by 5 min of recovery at 4 km h 1 and 5 min of passive rest. Cycling spiroergometry started at 50 W. The load was increased by 50 W every 3 min until volitional exhaustion, followed by 5 min of recovery with 50 W and 5 min of passive rest. Exhaustion was considered if at least two of the following criteria were met: high levels of blood lactate concentration (BLa; 8–10 mmol L 1 ); a plateau in oxygen uptake (VO2) despite increasing work rate; a respiratory exchange ratio (RER) above 1.1. During each test, oxygen uptake (VO2), carbon dioxide release (VCO2), and ventilation (VE) were recorded with a breath-by-breath spirometry system (Geratherm Respiratory GmbH, Bad Kissingen, Germany) and the corresponding software Blue Cherry (version 1.3.0.5, Geratherm Respiratory GmbH, Bad Kissingen, Germany) using an individual adjusted face mask. Before each test, both sensors were calibrated with known gas concen- trations and the owmeter with a 3L-syringe according to the manufacturer's instructions. Heart rate (HR) was recorded continuously with a 12-lead ECG using the Amedtec ECGpro Software (version 4.21.0, AMEDTEC Medizintechnik Aue GmbH, Aue, Germany) and self- adhesive electrodes in all three groups. Maximum oxygen uptake (VO2max) and RERmax were detected as the highest 30 s average values at the time of volitional exhaustion. If participants did not nish the entire increment, their maximal running velocity or cycling performance was linearly interpolated. 2.4. Standardization Participants were asked to refrain from any drugs (e.g., nicotine, alcohol, medication, caffeine) and soft drinks or special teas (green or black tea) 48 h before each laboratory visit. Nutritional supplementation (e.g., creatine, beetroot) was not allowed 48 h before each visit. All participants were asked to avoid intense physical activity and training 24 h before each visit. Participants were asked

were asked to refrain from any drugs (e.g., nicotine, alcohol, medication, caffeine) and soft drinks or special teas (green or black tea) 48 h before each laboratory visit. Nutritional supplementation (e.g., creatine, beetroot) was not allowed 48 h before each visit. All participants were asked to avoid intense physical activity and training 24 h before each visit. Participants were asked to not change their style of living (including mode of transportation) during the study. The basal metabolome was measured at the end of three characteristic training phases (for details see Tables S1 and S2): preparation (high volume and low intensity training), competition (regular competitions as well as low-volume, high-intensity training), and regeneration (no training at all or low-volume, low-intensity, unspeci c training). All basal measurements were used for further analysis and declared together as the resting metabolic pro le, independent of training phases. The results of the resting metabolic pro le were part of a larger study where standardized endurance and strength tests were performed to investigate exercise-induced metabolic pro le differences. The subjects of CG performed both the endurance and strength tests on different occasions. Therefore, their basal metabolome was determined twice. Due to the testing procedure and duration of the exercise tests and because of test- ing athletes in their competitive phase, it was not possible to test them in a fasted state. Therefore, all participants received nutrition counselling at the beginning of the study and a nutrition plan for a standardized dinner in the evening before each visit. They were served a standardized breakfast in the morning of each visit. All participants had to choose between three different meals in the evenings and in the mornings and stay with the same self-selected meal and size as well as the same mealtime during the study. Each meal consisted of an equal nutrient distribution (55–60% carbohydrates, 25–30% proteins, 13–17% fat). Water was consumed ad libitum. Compliance with the diet was con rmed with food diaries the day before the trials and a food photography method for lunch be- tween 6 and8 p.m.the evening before the trials [22]. Breakfast time

as the same mealtime during the study. Each meal consisted of an equal nutrient distribution (55–60% carbohydrates, 25–30% proteins, 13–17% fat). Water was consumed ad libitum. Compliance with the diet was con rmed with food diaries the day before the trials and a food photography method for lunch be- tween 6 and8 p.m.the evening before the trials [22]. Breakfast time was between 7 and 10 a.m.according to the participants' schedule and consumed in a maximum of 20 min. To prevent possible bias related to circadian rhythms, all laboratory visits were performed at the same time of the day and sequence throughout the study [23].

Metabolites2023,13, 833 5 of 16 2.5. Sample Handling Venous blood samples (single 4.9 mL tube, EDTA, S-monovette, Sarstedt, Nümbrecht, Germany) were taken from the forearm vein of all participants in a seated position 60 min postprandial. After collection, the tube was immediately stored in crushed ice (4 C) for no longer than two hours. Tubes were then centrifuged at 4000 gat 4 C for 10 min. After separation, supernatant plasma was instantly aliquoted, snap frozen in liquid nitrogen, and stored at 80 C until analysis. 2.6. Targeted Metabolomics Analysis The Biocrates MxP ® Quant 500 kit (Biocrates Life Sciences AG, Innsbruck, Austria) can be used for analysis of up to 630 metabolites from 26 compound classes of widely different structures and polarities. Compound classes include lipids like acylcarnitines (Cx:y), hy- droxylacylcarnitines [C(OH)x:y] and dicarboxylacylcarnitines (Cx:y-DC), lysophopatidyl- cholines, phosphatidylcholines, sphingomyelins (SMx:y) and sphingomyelin derivatives [SM(OH)x:y], ceramides and derivatives (cer-, hexcer-, hex2cer- and hex3cer-), cholesteryl esters, and diglycerides and triglycerides (the rst fatty acid is counted individually, in the case of three fatty acids, the last two fatty acids are summed), which are all measured by FIA-MS/MS, as well as amino acids, amino acid-related compounds, bile acids, bio- genic acids, biogenic amines, the sum of hexoses (H1), p-cresol sulfate, carboxylic acids, fatty acids, hormones and related metabolites (abscisic acid, cortisol, cortisone, dehy- droepiandrosterone sulfate; DHEAS), indoles and derivatives (indole, 3-indoleacetic acid, 3-indolepropionic acid, indoxyl sulfate), xanthine and hypoxanthine, choline, trigonelline, and trimethylamine N-oxide (TMAO), which are determined by UPLC-MS/MS. In brief, 10 L of human plasma were pipetted on a 96 well-plate containing internal standards and dried under a nitrogen stream using a positive pressure manifold (Waters). A total of 50 L of a 5% phenyl isothiocyanate (PITC) solution was added to each well to derivatize amino acids and biogenic amines. After 1 h incubation at room temperature, the plate was dried again. To extract the metabolites, 300 L 5 mM ammonium acetate in methanol was pipetted to each lter and incubated for 30 min. The extract was eluted into a new 96-well plate using positive pressure. For further LC-MS/MS analyses, 150

each well to derivatize amino acids and biogenic amines. After 1 h incubation at room temperature, the plate was dried again. To extract the metabolites, 300 L 5 mM ammonium acetate in methanol was pipetted to each lter and incubated for 30 min. The extract was eluted into a new 96-well plate using positive pressure. For further LC-MS/MS analyses, 150 L of the extract was diluted with an equal volume of water. For FIA-MS/MS analyses, 10 L extract was diluted with 490 L of FIA solvent (provided by Biocrates). After dilution, LC-MS/MS and FIA-MS/MS measurements were performed. For chromatographical separation, an UPLC I-class PLUS (Waters) system was used coupled to a SCIEX QTRAP 6500+ mass spectrometry system in electrospray ionization (ESI) mode. Data was generated using the Analyst (Sciex) software suite and transferred to the MetIDQ software (Biocrates Life Sciences AG), which was used for further data processing and analysis. All metabolites were identi ed using isotopically labeled internal standards and multiple reaction monitoring (MRM) using optimized MS conditions as provided by Biocrates. For quanti cation, either a seven-point calibration curve or one-point calibration was used depending on the metabolite class. Sample orders were randomized to ensure that the results obtained are not in uenced by the order of analysis. 2.7. Statistical Analysis All data preprocessing and analysis steps were performed using R (version 4.0.3, R Core Team, Vienna, Austria) [24]. Where appropriate, the web-based tool MetaboAnalyst 4.0 (https://www.metaboanalyst.ca) [ 25] was used. We used a multiple step procedure to ensure data quality. First, metabolites with more than 20% missing values (i.e., values lower than level of detection, <LOD) were removed from the data set [26]. Second, missing values were imputed using the k-nearest (k = 3) neighbor method [27,28]. In a third step, potential outliers were visually detected via principal component analysis (PCA). Samples far outside the 95% con dence interval were regarded as strong outliers. The nal data matrix (367 metabolites and 34 samples) was used for further analysis in regard to group (ST, ET, CG) and training phase (PP, CP, RP). The control group was matched to

In a third step, potential outliers were visually detected via principal component analysis (PCA). Samples far outside the 95% con dence interval were regarded as strong outliers. The nal data matrix (367 metabolites and 34 samples) was used for further analysis in regard to group (ST, ET, CG) and training phase (PP, CP, RP). The control group was matched to both

This study investigates metabolic profiles of athletes and sedentary subjects over a training year.