Intravascular ADP and soluble nucleotidases contribute to acute prothrombotic state during vigorous exercise in humans

Abstract

Extracellular ATP and ADP trigger vasodilatatory and prothrombotic signalling events in the vasculature. Here, we tested the hypothesis that nucleotide turnover is activated in the bloodstream of exercising humans thus contributing to the enhanced platelet reactivity and haemostasis. Right atrial, arterial and venous blood samples were collected from endurance-trained athletes at rest, during submaximal and maximal cycle ergometer exercise, and after early recovery. ATP-specific bioluminescent assay, together with high-performance liquid chromatographic analysis, revealed that plasma ATP and ADP concentrations increased up to 2.5-fold during maximal exercise. Subsequent flow cytometric analysis showed that plasma from exercising subjects significantly up-regulated the surface expression of P-selectin in human platelets and these prothrombotic effects were diminished after scavenging plasma nucleotides with exogenous apyrase. Next, using thin layer chromatographic assays with [gamma-(32)P]ATP and (3)H/(14)C-labelled nucleotides, we showed that two soluble nucleotide-inactivating enzymes, nucleotide pyrophosphatase/phosphodiesterase and nucleoside triphosphate diphosphohydrolase, constitutively circulate in human bloodstream. Strikingly, serum nucleotide pyrophosphatase and hydrolase activities rose during maximal exercise by 20-25 and 80-100%, respectively, and then declined after 30 min recovery. Likewise, soluble nucleotidases were transiently up-regulated in the venous blood of sedentary subjects during exhaustive exercise. Human serum also contains 5'-nucleotidase, adenylate kinase and nucleoside diphosphate (NDP) kinase; however, these activities remain unchanged during exercise. In conclusion, intravascular ADP significantly augments platelet activity during strenuous exercise and these prothrombotic responses are counteracted by concurrent release of soluble nucleotide-inactivating enzymes. These findings provide a novel insight into the mechanisms underlying the enhanced risk of occlusive thrombus formation under exercising conditions.

Figure 1. Plasma from exercising humans stimulates P-selectin expression on platelet surface

Human platelets from healthy volunteers were pretreated for 20 min with plasma samples collected from the right atrium of trained athletes at rest, after submaximal (sub) and maximal (max) exercise and after 10 min recovery (rec). A and B, the platelets were then stained with anti-P-selectin mAb WAPS-12.2 (histograms 1, 2 and 3) or with negative control mAb 3G6 (Neg. control) and analysed by flow-cytometry. The abscissa represents the logarithmic scale of fluorescence intensity and the ordinate shows the relative cell number. C, the percentage of P-selectin expression after platelet-plasma co-incubation was determined from the above histograms (mean +s.e.m.; *P < 0.05 as compared with rest; n= 8). Plasma from exercising subjects was pretreated with apyrase (1 unit ml⁻¹) prior to addition to platelets (**P < 0.05 as compared with non-treated plasma; n= 4). Platelets were also incubated in RPMI-1640 with and without 1 μmol l⁻¹ ADP (***P < 0.05 as compared with medium alone; n= 3).

Figure 2. Effect of exercise on adenine nucleotide levels in blood plasma

Human blood was collected from the trained athletes at rest, after submaximal (sub) and maximal (max) exercise and after 10 min recovery (rec). A, plasma ATP was assayed using luciferin–luciferase luminometry. B, HPLC analysis of ADP and AMP concentrations in plasma prepared from the brachial artery blood of exercising humans. Data are mean +s.e.m. (n= 8–10). *P < 0.05 as compared with rest.

Figure 3. Pattern of [γ-³²P]ATP metabolism in human plasma and serum

A, serum and EDTA-containing or heparinized plasma from venous blood were incubated with 10 μmol l⁻¹[γ-³²P]ATP, separated using TLC and developed by autoradiography. Serum was additionally ultracentrifuged or filtered through 0.22 μm filters prior to adding [γ-³²P]ATP. B and D, serum from right atrial blood of exercising trained subjects was incubated with [γ-³²P]ATP for 10–35 min. Arrows indicate the positions of ATP, inorganic phosphorus (P_i) and pyrophosphate (PP_i). Blanks show the radiochemical purity of [γ-³²P]ATP. C, serum NPPs were quantified by acquiring the images shown in B and measuring the band intensities corresponding to [γ-³²P]ATP-derived ³²PP_i. E, likewise, NTPDase activities were determined from the autoradiography in D by the rate of ³²P_i cleavage from [γ-³²P]ATP. Data are presented as box-and-whiskers plots (n= 10). *P < 0.05 as compared with rest.

Figure 4. Effect of exercise on nucleotide-converting activities in human serum

Serum was prepared at rest, during submaximal (sub) and maximal (max) exercise, and after ensuing 10 min recovery. ATPase (A), ADPase (B), and 5′-nucleotidase (C) were assayed by TLC with 100 μmol l⁻¹[¹⁴C]ATP, 50 μmol l⁻¹[³H]ADP and 300 μmol l⁻¹[³H]AMP as respective substrates. Nucleotide-phosphorylating enzymes, adenylate kinase (D) and NDP kinase (E), were assayed using γ-phosphate donating ATP (800 μmol l⁻¹) and 500 μmol l⁻¹[³H]AMP or [³H]ADP as respective phosphate acceptors. Enzymatic activities are presented as box-and-whiskers plots expressed as nanomoles of nucleotide substrates hydrolysed (A–C) or phosphorylated (D and E) by 1 ml serum per hour (n= 10).

Figure 5. Hydrolysis of [³H]ADP by human serum during exercise and recovery

Blood was prepared from the femoral artery (A), right atrium (B) and femoral vein (C) of the trained athletes at rest, during incremental exercise, and ensuing 30 min recovery. Serum NTPDase was assayed by TLC using 50 μmol l⁻¹[³H]ADP as substrate. Data are mean ±s.e.m. (n= 6). *P < 0.05 as compared with rest.

Figure 6. ATP level and nucleotidase activities in the blood of sedentary subjects

Venous blood was collected at rest, during maximal exercise (max) and after 30 min recovery (rec). A, plasma ATP was assayed using luciferin–luciferase luminometry. B and C, serum was incubated with 10 μmol l⁻¹[γ-³²P]ATP for 10–35 min and separated by TLC. NPP and NTPDase activities were quantified from the autoradiographic images by measuring the band intensities corresponding to [γ-³²P]ATP-derived ³²PP_i and ³²P_i, as shown in Fig. 3. D, NTPDase was assayed by TLC using 50 μmol l⁻¹[³H]ADP as substrate. Data are presented as box-and-whiskers plots (n= 7). *P < 0.05 as compared with rest.