Increased level of circulating cell-free mitochondrial DNA due to a single bout of strenuous physical exercise

The effect of strenuous physical exercise on the immune system has been studied during the last years. Both acute and chronic training have been investigated, often in connection with various disease syndromes, such as depression or multiple sclerosis (Hallberg et al. 2010; Mikkelsen et al. 2013; Suchanek et al. 2010; Zwetsloot et al. 2014). A single bout of physical activity leads to both cellular and humoral immune responses similar to those caused by infection, sepsis and trauma (Steensberg et al. 2000). The plasma concentration of the pro-inflammatory cytokines (e.g. TNF-α, IL-1β, IL-6 and IFN-γ) as well as acute-phase proteins increase (Boettger et al. 2010; Castellano et al. 2008). However, the degree of immune activation is reported to decrease in persons that perform regular training (Simpson et al. 2015).

Mitochondria are organelles responsible for the synthesis of ATP (Johannsen and Ravussin 2009; Scheffler 2001) and are also responsible for the generation of signaling molecules involved in apoptosis and necrosis (Johannsen and Ravussin 2009). The number of mitochondria per cell depends on its energy requirement, and e.g. muscle cells are equipped with many mitochondria to perform their regular work (Wust et al. 2018). Regular exercise has been shown to increase the number of mitochondria in muscle cells, resulting in an increased ability to synthesize ATP. In case of cellular damage, endogenous cellular substances normally not exposed to the immune system, act as danger-associated molecular patterns (DAMPs), as they are released into the circulation. These molecules are then recognized by different pattern recognition receptors, such as Toll-like receptors (TLRs), on immune cells causing a signaling cascade leading to excretion of antimicrobial substances and pro-inflammatory substances (Zhang et al. 2010a, b). Mitochondrial DNA (mtDNA), released into circulation from mitochondria under stress, has been reported to act as a DAMP, mimicking an ongoing immune-activation, mediating a pro-inflammatory response (Bhagirath et al. 2015; Chiu et al. 2003; Yu 2012). Systemic inflammatory response syndrome (SIRS) is a condition when several pro-inflammatory mediators, due to cellular damage caused by tissue-damage, trauma or hypoxia are released into the blood. Since heavy physical exercise, due to muscle damage, necrosis and oxygen deprivation has been shown to result in immune-activation resembling trauma and sepsis, the contribution of mitochondrial function has to be taken into account. Studies of trauma have revealed that inflammation, neutrophil activation and organ injury can be mediated by cell free mtDNA (cf-mtDNA) activating neutrophils via TLR-9, normally binding bacterial DNA (Zhang et al. 2010a, b).

Another useful biomarker for immune activation is the soluble urokinase plasminogen activating receptor (suPAR) (Thuno et al. 2009) however, only a few studies can be found describing the effect from intense physical activity on suPAR (Gustafsson et al. 2017; Sanchis-Gomar et al. 2013). suPAR is the soluble fragment of the multi-ligand receptor urokinase plasminogen activating receptor (uPAR or CD87). uPAR, expressed on several cell types, including neutrophils, monocytes, and endothelia cells, is involved in both proteolysis, cell-migration, angiogenesis, and inflammation. During the last decades, several studies have been investigating the possible use of suPAR as a biological marker of inflammation, organ damage, and severity of the disease, showing positive correlations between increased suPAR concentrations in both blood, urine and cerebrospinal fluid (Enocsson et al. 2015; Thuno et al. 2009).

During infection and inflammation, as well as during physical exercise, an increased number of circulating leukocytes is observed (Pedersen and Hoffman-Goetz 2000). Leukocytosis during physical exercise is caused by de-margination of leukocytes from the marginal pool caused by increased heart rate and blood flow, as well as by increased concentrations of epinephrine and cortisol (Sanchis-Gomar and Lippi 2014). Neutrophils, dendritic cells and lymphocytes have shown to increase during exercise (Pedersen and Hoffman-Goetz 2000; Suchanek et al. 2010). After exercise, the lymphocytes decrease below base-line but the levels of neutrophils remain increased (Pedersen and Hoffman-Goetz, 2000; Simpson et al. 2015).

The aim of this study was to investigate if a bout of strenuous physical exercise leads to changed levels of circulating cf-mtDNA in plasma and to examine if a correlation between cf-mtDNA levels, suPAR, and total leukocytes could be established. Moreover, a possible correlation of membrane-bound uPAR (CD87) and free suPAR was investigated.

During this study, the effect of strenuous physical exercise on circulation of various leukocytes and inflammatory markers, such as circulating cf-mtDNA and suPAR were measured. Maximal heart rate of the participants was reached at individual time points ranging from 16.5 to 25.5 min, with a median value of 20.5 min (Table 1).

This study was performed to investigate if a single bout of strenuous physical exercise affects the concentration of cf-mtDNA and suPAR in plasma. Furthermore, the concentration of total leukocytes was measured during the study, together with the presence of membrane-bound uPAR on the different leukocytes.

The results clearly show that the concentration of cf-mtDNA was dynamically affected during the measurements. At sub-maximal and maximal work-load the concentration was mildly increased but a significant increase of concentrations could be seen after both 30 and 90 min of rest, compared to baseline levels. These results indicate that strenuous physical exercise induces some kind of cellular damage, leading to necrosis and the liberation of mitochondrial DNA. The results from the present study regarding cf-mtDNA differ from other studies showing decreased values of cf-mtDNA (Beiter et al. 2011; Helmig et al. 2015). In one extensive study, repeated bouts of exhaustive exercise were carried out, giving rise to large increases of both cf-mtDNA but especially cf-nDNA (Stawski et al. 2017). A possible reason for the increased levels of cf-DNA, both nuclear- and mitochondrial origin, could be DNA liberation from neutrophils. Liberation of DNA from neutrophils, forming neutrophil-extracellular traps (NET) have been shown to take place rapidly after stimulation (Yousefi et al. 2009). Possible reasons for the contradicting results could be technical differences during sample treatment and DNA-isolation. Slightly different protocols are in use, with differences of centrifugation speed and time, among other things (Dantham et al. 2016; Kumar et al. 2017). A reason for the raised values of the ratios between cf-mtDNA and n-DNA could be the fact that the primers targeting b2M do not bind to shorter fragments of damaged nuclear DNA (Jiang and Lo 2016).

Previous studies have shown that tissue damage occurs during physical exercise (Zwetsloot et al. 2014). Muscle cells, heavily loaded during physical exercise, are richly equipped with mitochondria. Hypoxia and tissue damage, due to over-load, are likely causes for the liberation of cf-mtDNA from these cells (Zhang et al. 2010a, b).

In a recent study also investigating the release of cf-mtDNA, another more extended exercise regime was used. The time of a tread-mill run was set to 90 min but only 60% of VO₂ was used. During this weaker bout of physical work, a decreased level of cf-mtDNA was registered (Shockett et al. 2016), indicating that different grades of physical labour have different effects. Short term hard labour probably causes tissue damages, reminding of trauma, while prolonged medium exercise causes an increase of circulating neutrophils, absorbing liberated cf-mtDNA, resulting in the observed decrease. The samples from the different participants of the present study showed a substantial individual variation. According to previous studies, the concentration of cf-mtDNA is influenced by several factors, such as age and health status (Pinti et al. 2014). The amount of cf-mtDNA has been reported to increase significantly after 50 years of age, however, with large individual differences. Age is an important parameter, probably due to a gradual decrease of muscle mass, leading to increased necrosis and liberation of endogenous substances, including mtDNA (Mikkelsen et al. 2013). Increased levels of cf-mtDNA in plasma could also be a result of decreased ability to clear it from the circulation (Kung et al. 2012). The mechanisms of clearing of cf-mtDNA have been discussed and studied by others and circulating neutrophils have been suggested to bind cf-mtDNA to TLR's acting as a sink (Shockett et al. 2016). Another theory is that increase of plasma DNase I activity, measured directly after exhaustive short-term exercise, could help restoring levels to base-line values (Beiter et al. 2011). Recently, it was reported that a sequence of DNA, targeted by the primers for the ND2-gene in the mitochondrial genome, is present on chromosome 1 in the human nuclear genome (NCBI Gene ID: 100652939). However, during our previous study of cf mtDNA, another primer-pair, targeting the ND1 gene in the mitochondrial genome, both pairs of primers were found to correlate to a high degree (Lindqvist et al. 2018).

An increased level of suPAR could be seen, during the exercise and the following rest, however, suPAR levels were only significantly raised 90 min after maximal work-load (Fig. 3a). Plasma concentrations of less than 4 ng/mL are considered as normal, while levels above 6 ng/mL may indicate illnesses such as cancer, diabetes or cardiovascular disease (Eugen-Olsen et al. 2010; Gustafsson et al. 2017; Kofoed et al. 2008). The cell-surface expression of the membrane-bound variant, uPAR, on the neutrophils was decreased during the same time, giving an almost inverted expression pattern, compared to suPAR (Fig. 3b). The negative correlation, although not significant, between suPAR and cell-surface expression of uPAR, measured as mean-fluorescent intensity (MFI) on granulocytes, supports the results of the study. The changes of suPAR-expression seen in this study are similar as during a study performed on patients suffering from major depressive disorder (MDD) (Gustafsson et al. 2017). An almost identical pattern of plasma-suPAR was found during that study, indicating that a single bout of strenuous physical exercise has a substantial effect on suPAR levels. The rather immediate increase of suPAR and decrease of uPAR fits well with the theory that uPAR is cleaved by a mixture of activated proteases and acute-phase proteins (Thuno et al. 2009). The concentration of suPAR measured by Sanchis-Gomar et al. (2014) before and after a football match was not reported to change, however, the subjects were all athletes, probably accustomed to physical exercise. The blood samples after the match were, however, taken 12 h after the match was finished. In a recent study, including 80 participants and 53 controls, it was reported that mild physical training, during a five-month period of time, did not have any effect on suPAR-levels (Rohde et al. 2018). On the other hand, a healthy lifestyle has been shown to significantly affect the levels of suPAR (Haupt et al. 2019) with higher concentrations measured in smokers and physically inactive participants. One interpretation of the different results is that suPAR levels indicate an acute inflammatory response, as mimicked during heavy physical exercise, but also a low-grade on-going inflammation, manifested as a result of unhealthy and sedentary lifestyle.

The expression of uPAR has been found on several blood leukocytes, including granulocytes, monocytes and activated T-lymphocytes (Koch et al. 2014; Thuno et al. 2009). The leukocytosis, resulting from de-margination of cells during heavy exercise, probably contributes to the increase of suPAR, since the levels of suPAR and total leukocytes have a similar concentration profile. Another explanation to the increase of suPAR could be a possible microbial translocation across the gut, resulting in the occurence of lipopolysaccharide (LPS) (Brenchley et al. 2006; Ng et al. 2008) in connection with physical exercise.

Several different proteases have been demonstrated to cleave uPAR in the linker region, including the urokinase plasminogen activator (uPA) itself, plasmin, chymotrypsin, and various MMPs (Andolfo et al. 2002). Increased plasma levels of uPA have previously been shown to result from physical exercise (Dooijewaard et al. 1991). When uPA is bound to uPAR, the receptor is cleaved, and suPAR, the soluble form of uPAR, is released. Since it has also been shown that heavy exercise can contribute to elevated levels of metalloprotease-9 (MMP-9) (Rullman et al. 2013), this protease could also be involved in the cleavage of uPAR leading to suPAR production.

One limitation of the study is the lack of blood-samples at later time-points. It would be interesting to measure the levels of both cf-mtDNA and suPAR during the following day and compare that to the leukocyte concentration.

A further limitation of the study is the low number of participants, making excessive conclusions from the results difficult. However, the aim of the study was reached, and was in line with the study by Stawski et al. (2017), encouraging further examinations of the effect from exercise on the immune response and the use of cf-mtDNA and suPAR in measurements. The increased levels of plasma IL-6, measured in connection to extensive exercise (Ullum et al. 1994), further strengthens the results indicating that exercise induces an inflammatory reaction manifested in several ways. Further studies will have to be performed to clarify the mechanisms involved.