The research was carried out within a group of 19 elite soccer players with more than 3 years of training experience in elite competition (74%) (height: 183.74 ± 5.51 cm, weight: 78.82 ± 7.29 kg, age: 22 ± 2.3 years, VO_2max: 59.69 ± 4.56 ml/kg/min).

The data on coronary risk factors, symptoms, medications, and prior cardiac events were gathered before an exercise testing, and no contraindications for the stress testing were found.

The study was done in pre-seasonal period in aerobic training regime. Daily data were recorded on the duration and intensities of training activities performed during the study period (90 minutes of training duration once a day with effort corresponding to a zone of 70-80% of maximum predicted heart rate).

All the participants gave a written informed consent.

The study was approved by Ethical committee, School of Medicine, University of Belgrade.

The following method for a maximal exercise testing was used in order to investigate a biochemical endothelial response in exercise. Maximal progressive exercise test was performed according to the modified Ellestad Memorial Hospital B protocol [21] on treadmill (T 200, Cosmed, Italy). Modified Ellestad B protocol is presented in Table 1.

The experiment started in the morning (between 8 and 9 AM). A small cathether was inserted into an antecubital vein for venous blood sampling.

Venous blood was drown with a 5-ml syringe and mixed with 100μL of 3M perchloric acid, 400μL of 20mM EDTA in a 10 ml test tube and after incubation in ice for 20 minutes, centrifuged at 15000 rpm for 5 minutes. The samples were than separated and kept at -20°C until measurement. Blood was drawn: 1) before the test, 2) in last 10 sec of each level of the exercise testing (I-V) (without test interrupting), as well as 3) in the 90th and 180th second of the recovery. O₂^- production was measured by SOD inhabitable conversion of nitro blue tetrazolium (NBT) to formazan [22]. Nitrite (NO₂^-) level, as a marker of NO metabolism, was measured by Griess reaction [23].

The pulmonary VO₂, CO₂ production (VCO₂), and the expired minute ventilation (V_E) were measured continuously through usage of an automated metabolic cart (Quark b2, Cosmed, Italy). Before each test the ambient conditions were measured, and then the gas analyzer and the flowmeter were calibrated with high precision gases. During the lower and exhaustion load levels the VO₂ values were recorded as averages of 15 s. The participants stated their subjective feeling of exhaustion by using Borg’s CR10 exhaustion scale of at least 8 [24]. The heart rate was monitored continuously and recorded as averages of 15 s using a Polar Electro Oy, Finland.

We hypothesized that the maximal oxygen consumption VO_2max was reached when the oxygen consumption got its plateau (the time when increasing of workload cannot affect an increase in oxygen consumption) [25].

Results are expressed as means ± standard error of the mean. Two-way analysis of variance (ANOVA) was used to assess variations in NO₂^- and O₂^- production depending on 1) their basal production and levels of effort, 2) achieved VO2max and levels of effort. Test of signs was used to evaluate the difference of NO₂^- and O₂^- concentration in different load levels. Linear regression was applied to identify connectivity between NO₂^- and O₂^- concentration during load progression. The significance level was set at P<0.05.

Mean basal NO₂^- and O₂^- values of 19 subjects are shown in Table 2.

Basal values of NO₂^- showed great variation between subjects (mean value 20.34 ± 14.01 nmol/ml), so we tried to assess whether there is a correlation between basal production of NO₂^- and subjects’ training history. It was shown that the relation exists, i. e. that athletes who have trained professionally more than 3 years without longer cessations in training process (more than 6 weeks) had greater basal production of NO₂^- than subjects who have trained less or had training cessations. So, we confirmed the assumption that long-term training induces increase in NO bioavailability, whilst training cessations, i. e. physical inactivity, have negative effect – it leads to reversible changes and preclude the increase in NO bioavailability. That is in accordance with studies that showed that physically active people have greater basal NO production than sedentary population [27-29], as well as that exercise induces increase in basal NO production [30-32], especially in sedentary [33] and people with some kind of endothelial dysfunction [34-37]. However, it should be emphasized that these exercise-induced changes in NO bioavailabilty are transitory and can be lost during inactivity period [32].

As original values of NO₂^- and O₂^-had high coefficient of variation, values were exponentially transformed, and transformed values are shown in Table 3. Time-course of NO₂^- and O₂^-production during exercise is shown in Fig. (1).

Fig. (1). Time-course of NO2- and O2- production during the exercise.

Statistically significant (p<0.05) difference in NO₂^-production was found between resting state and grade I (increase in NO₂^-production), and between stage I and stage II (decrease in NO₂^- production). NO₂^-stayed decreased (lower than basal) until the end of testing. Increased NO₂^- production at the beginning of the exercise test can be explained by increased blood flow and shear stress [38]. The subsequent drop in NO₂^- production may be explained by effects of ROS (O₂^-) on NO biovailability, since O₂^- acts as NO scavenger, decreasing his bioavailability [15]. This reaction is especially evident when the levels of NO increase into the elevated nanomolar range and approach the local concentrations of superoxide dismutase (SOD) [39], because the reaction between NO and O₂^- is 3 times faster than SOD-mediated dismutation of O₂^-[11, 40]. Later, during the test, the NO level declines, probably as a consequence of a structural adaptation of blood vessel smooth muscle cells in trained subjects [41], the adaptation that subjects in our study gained during long-term training history. Tinken et al. [42] recently confirmed this mechanism of blood vesells adaptation to repetitive exposure to shear stress, emphasizing the time-course of endothelial adaptation - functional changes in the conduit arteries precede structural adaptations in vivo, i.e. NO may play a role in the short-term adaptations to exercise (over a few weeks), but long-term exercise (over a period of years) is associated with another mechanism of endothelial adaptation to increased blood flow [43]. It is possible that adaptations to meet metabolic demands with training evolve from, at least in part, NO-mediated vasodilation in the short-term, to longer term adaptations such as metabolic enzyme changes and vascular restructuring [41-43]. Structural changes (including increased vessel diameter) minimize or even eliminate the need for enhanced release of NO, so it returns to the baseline [42, 43].

Generally, it is thought that exercise increases ROS production, especially intense aerobic exercise during which athletes consume large amounts of oxygen (2-5% of that oxygen, due to electron laekage in mitochondria chain, converts to O₂^-) [15]. During our testing, statistically significant (p<0.05) increase in O₂^- production was also found between the resting state and grade I, and also, between the grade IV and V and the first phase of the recovery (R1). The increase of O₂^- levels with the beginning of the test and latter O₂^- decrease could be explained by mitochondrial properties to produce free radicals – mitochondrial free radical production increases during state 4 respiration (low VO₂; high membrane potential; low ATP production), while during state 3 respiration (high VO₂; lower membrane potential; high ATP production) it decreases [44]. That is in accordance with results of Herrero and Barja, who confirmed the decrease of radical production as VO₂ and ATP synthesis increase, and observed the lowest O₂^- concentration at the point of anaerobic threshold [45] (our subjects achieved their anaerobic threshold somewhere near the end of the stage IV). After the point of respiratory compensation (anaerobic threshold), O₂^- concentration started to increase again, which partly can be explained by U-shaped radical formation related to the mitochondrial pO₂ and/or an alternative mechanism – mechanism of xanthine oxidase O₂^- production during ischaemia and reperfusion [8, 46]. Also, high intensity exercise is often accompanied by increased secretion of the catecholamines which can undergo autooxidation, with and without oxygen, and produce O₂^- [47].

As basal NO₂^-values of athletes varied to the great extent, we divided them into three groups (depending on their basal NO₂^-value - B1: NO₂^-> 20 nmol/ml, B2: NO₂^-10-20 nmol/ml, B3: NO₂^-< 10 nmol/ml), in order to detect differencies in time-course of NO₂^-production between groups, i.e. to see whether basal value of NO₂^-affects time-course of NO₂^-production during the test. Analysing the results of these three groups (The two-way ANOVA), we concluded that there were differences in NO₂^-production between groups (B1-B3 and B2-B3), but that the differencies were not related to the level of effort. The groups producted higher or lower concentrations of NO₂^-, but the pattern of NO₂^-production during the test was similar (Fig. 2). All three groups showed initial increase in NO₂^-production, and the increase was the lowest in the group with the lowest basal NO₂^-production. After the first level of effort, NO₂^- production dropped in all three groups, and stayed decreased and relatively stable to the end of the test. The group with the lowest basal NO₂^-production showed the least oscilations in NO₂^-production during the test. That is in accordance with the results of the study that compared time-course of NO₂^-production between athletes from different sports [48]. That study showed statistically significant variations in NO₂^-production between athletes of different sport, both in basal and active conditions, and also that athletes with the lowest basal NO₂^-production had the least oscilations in NO₂^-production during the test.

Fig. (2). Time-course of NO2- production during the exercise test - dependance on its basal values.

The two-way ANOVA analysis of O₂^-production during exercise between groups based on basal O₂^-production (B1: O₂^-> 10 nmol/ml, B2: O₂^-= 5-10 nmol/ml and B3: O₂^-< 5 nmol/ml) showed that there was significant variation in O₂^- production between groups, but that it wasn’t in a relation with levels of effort (Fig. 3). We concluded that production of O₂^-is pretty individual and depends on many factors. For example, basal O₂^-values depend on physical activity during preceding day(s), i. e. exercise-induced muscle damage and inflammation cause delayed free radical production by neutrophil NADPH [3]. Cyclist in study by Cubrilo [49], had higher basal values of O₂^-than rowers and taekwondo athletes, due to intensive and long trainings before they participated in the study. Some parameters of oxidative stress do not show themselves immediately after exercise, but hours or days later [50].

Fig. (3). Time-course of O2- production during the exercise test - dependance on its basal values.

In order to determine whether dynamics of NO₂^- and O₂^- production during the exercise test depends on the level of aerobic capacity of athletes, we divided athletes into three groups, based on the VO_2max achieved on the test – G1: VO_2max > 65 ml/kg, G2: VO_2max = 60-65 ml/kg, and G3: VO_2max < 60 ml/kg (Table 4).

We didn’t find significant variations in time-course of NO₂^- production between groups, i.e. there wasn’t significant correlation between VO_2max and time-course of NO₂^- production during exercise, but that there was significant correlation between VO_2max and NO₂^- basal production, since athletes who had basal NO₂^- production > 20 nmol/ml achieved higher VO_2max on the test (mean value VO_2max = 65,84 ml/kg) compared to athletes who had NO₂^- basal production < 20 nmol/ml (mean value VO_2max = 60,82 ml/kg).

Correlation between NO bioavailability and exercise capacity was also shown in many other studies. Rassaf et al. showed that the capacity of the vasculature to produce NO, and thus nitrite, predicts maximal power and duration of exercise in 55 healthy subjects [51]. Hambrecht et al. suggested that improvement in NO-mediated vascular function improves cardiac function and VO_2max in patients with chronic heart failure [36]. Allen et al. also showed significant relation between VO_2max and baseline levels of nitrite and nitrate (NOx) [52], but he also found correlation between VO_2max and the increase in NO₂^- or NOx from baseline to recovery [52-54], which is opposite to our results (both nitrite and nitrate increased during exercise in his studies, which is opposite to our results, too).

Analysis of O₂^- production during exercise showed that there were differencies in time-course of O₂^- production between groups (G1-G2 and G1-G3) – groups with higher aerobic capacity showed decrease in O₂^- production after the initial increase in the beginning of the test, while in group with lower VO_2max concentrations of O₂^- continued to rise until the stage III. Trend of decrease in O₂^- production in groups with high VO_2max may be explained by metabolic, cardiovascular, respiratory and endocrine adaptations that developed simultaneously with aerobic capacity development [55]. It was shown that antioxidants appear in blood within less than 5 minutes after occurrence of free radicals [56], i.e during the stage II of the exercise test, so decrease in O₂^- production may also be explained by activity of antioxidants.

The reaction between NO (produced as a response of endothelium to increased blood flow and shear stress) and O₂^- (produced as a result of increased physical activity) is 3 times faster the SOD-catalysed dismutation of superoxide, so when the levels of NO increase into the elevated nanomolar range and approach the local concentrations of SOD, NO is able to compete with SOD for the scavenging of O₂^-[39]. This interaction results in alleviating O₂^-toxic effects, but also decreases effects of NO, for example NO-mediated vasodilatation [15].

By following ratio of their production during the exercise test, we detected the point of their most intensive reaction, which turns to be near the grade IV, at the level of anaerobic threshold (Fig. 1). It suggests that lactate threshold could be of a crucial importance not only in anaerobic and aerobic metabolism but in mechanisms of signal transductions as well.