Physiological extremes of the human blood metabolome: A metabolomics analysis of highly glycolytic, oxidative, and anabolic athletes

Abstract

Human metabolism is highly variable. At one end of the spectrum, defects of enzymes, transporters, and metabolic regulation result in metabolic diseases such as diabetes mellitus or inborn errors of metabolism. At the other end of the spectrum, favorable genetics and years of training combine to result in physiologically extreme forms of metabolism in athletes. Here, we investigated how the highly glycolytic metabolism of sprinters, highly oxidative metabolism of endurance athletes, and highly anabolic metabolism of natural bodybuilders affect their serum metabolome at rest and after a bout of exercise to exhaustion. We used targeted mass spectrometry-based metabolomics to measure the serum concentrations of 151 metabolites and 43 metabolite ratios or sums in 15 competitive male athletes (6 endurance athletes, 5 sprinters, and 4 natural bodybuilders) and 4 untrained control subjects at fasted rest and 5 minutes after a maximum graded bicycle test to exhaustion. The analysis of all 194 metabolite concentrations, ratios and sums revealed that natural bodybuilders and endurance athletes had overall different metabolite profiles, whereas sprinters and untrained controls were more similar. Specifically, natural bodybuilders had 1.5 to 1.8-fold higher concentrations of specific phosphatidylcholines and lower levels of branched chain amino acids than all other subjects. Endurance athletes had 1.4-fold higher levels of a metabolite ratio showing the activity of carnitine-palmitoyl-transferase I and 1.4-fold lower levels of various alkyl-acyl-phosphatidylcholines. When we compared the effect of exercise between groups, endurance athletes showed 1.3-fold higher increases of hexose and of tetradecenoylcarnitine (C14:1). In summary, physiologically extreme metabolic capacities of endurance athletes and natural bodybuilders are associated with unique blood metabolite concentrations, ratios, and sums at rest and after exercise. Our results suggest that long-term specific training, along with genetics and other athlete-specific factors systematically change metabolite concentrations at rest and after exercise.

Keywords: athlete; energy metabolism; exercise biomarker; exercise phenotype.

FIGURE 1

Overview of the study design where a standardized bicycle ramp test was performed with a continuously increasing load of 30 watts per minute until voluntary exhaustion

FIGURE 2

PLS‐DA score plot showing baseline (●) and post‐exercise (■) serum metabolite profiles within 75% confidence intervals (shading) of natural bodybuilders (B1‐B4), endurance athletes (E1‐E6), sprinters (S1–S5) and untrained controls (C1–C4)

FIGURE 3

Concentration changes for every participant between baseline (○) and post‐exercise (□) for isoleucine (a) and PC aa C36:6 (b), contributing most to the separation of natural bodybuilders and the CPT1‐ratio (c) and PC ae C38:6 (d) contributing most to the separation of endurance athletes

FIGURE 4

Volcano plot (a) showing significant metabolite changes (in black; α < 2.58*10⁻⁴) after graded cycle exercise in all participants and metabolites with the highest concentration changes from baseline (○) to post‐exercise (□) including (b) alanine, (c) spermidine/putrescine and (d) serotonin/tryptophan

FIGURE 5

Among all 194 metabolite measures, hexose (a), taurine (b) and tetradecenoylcarnitine (c) showed suggestive group‐specific responses between baseline (○) and post‐exercise (□)