Physical fitness status modulates the inflammatory proteins in peripheral blood and circulating monocytes: role of PPAR-gamma

The aim of this study was to analyze the metabolic and molecular profile according to physical fitness status (Low or High VO2max) and its impacts on peripheral and cellular inflammatory responses in healthy men. First (Phase I), inflammatory profile (TNF-α, IL-6, IL-10) was analyzed at baseline and post-acute exercise sessions performed at low (< 60% VO2max) and high (> 90% VO2max) intensities considering the individual endotoxin concentrations. Next (Phase II), monocyte cell cultures were treated with LPS alone or associated with Rosiglitazone (PPAR-γ agonist drug) to analyze cytokine production and gene expression. Monocyte subsets were also evaluated by flow cytometry. A positive relationship was observed between LPS concentrations and oxygen uptake (VO2max) (r = 0.368; p = 0.007); however, in the post-exercise an inverse correlation was found between LPS variation (Δ%) and VO2max (r = -0.385; p = 0.004). With the low-intensity exercise session, there was inverse correlation between LPS and IL-6 concentrations post-exercise (r = -0.505; p = 0.046) and a positive correlation with IL-10 in the recovery (1 h post) (r = 0.567; p = 0.011), whereas with the high-intensity exercise an inverse correlation was observed with IL-6 at pre-exercise (r = -0.621; p = 0.013) and recovery (r = -0.574; p = 0.016). When monocyte cells were treated with LPS, High VO2max individuals showed higher PPAR-γ gene expression whereas Low VO2max individuals displayed higher IL-10 production. Additionally, higher TLR-4, IKK1, and PGC-1α gene expression were observed in the High VO2max group than Low VO2max individuals. In conclusion, even with elevated endotoxemia, individuals with High VO2max exhibited higher IL-6 concentration in peripheral blood post-acute aerobic exercise and lower IL-10 concentration during recovery (1 h post-exercise). The anti-inflammatory effects linked with exercise training and physical fitness status may be explained by a greater gene expression of IKK1, TLR-4, and PGC-1α, displaying an extremely efficient cellular framework for the PPAR-γ responses.


Material and methods
Participant recruitment. The present study comprised two phases (Phase I and Phase II). In phase I, 28 healthy male individuals (age: 28.8 ± 5.6 years; body mass: 75.8 ± 9.9 kg; VO 2mean : 50.5 ± 8.8 mL kg −1 min −1 ) were recruited to participate in the study in order to analyze the peripheral inflammatory profile at baseline and in response to two acute-exercise sessions according to endotoxemia. In Phase II, another 22 healthy male individuals (age: 25.8 ± 5.7 years, body mass: 76.5 ± 14.4 kg, VO 2mean : 47.8 ± 12.3 mL kg −1 min −1 ) were recruited in order to conduct molecular analysis of monocyte cell cultures treated for 24 h.
Initially, all participants were classified as physically inactive, physically active or well-trained using the International Physical Activity Questionnaire (IPAQ). All participants were required to complete all exercise sessions. This study was approved by the local research ethics committee of the Sao Paulo State University "Júlio de Mesquita Filho" and duly registered in Brazil Platform (national electronic system created by the Federal Government to systematize the receipt of research projects involving human beings in Ethics Committees throughout the country) (CAAE: 31168714.6.0000.5402) and the research was conducted according to the 2013 Revision of the Declaration of Helsinki. Healthy men were included, without any health disorders, such as inflammatory, cardiorespiratory and osteoarticular diseases, and who had not used any ergogenic substances or medicines for at least six months prior to the study. Written informed consent was obtained from all participants prior to participation.

Maximal incremental test and aerobic exercise bouts.
For Phases I and II, all participants completed a maximal incremental test on a cycle ergometer (Inbrasport CG-04, Embramed, Porto Alegre, Brazil) to determine maximal oxygen uptake. The initial workload was 35 watts for sedentary individuals, 70 watts for the physically active, and 105 watts for well-trained individuals, with an increase of 25 watts every 3-min, and under constant speed (70-90 rpm) until exhaustion 27 . Voluntary exhaustion criteria were gas exchange ratio > 1.1, HR max > 90% of the maximum expected for age, and rating of perceived exertion (RPE) > 18. The maximum workload (W max ) and maximal oxygen uptake (VO 2max ) were assessed by a breath-by-breath gas analyzer (Quark PFT, Cosmed, Rome, Italy). In Phase I, ventilatory thresholds (aerobic and anaerobic thresholds) were determined by the VE•VO 2 -1 vs. workload and VE•VCO 2 -1 vs. workload, as suggested by Binder and colleagues 28 , to prescribe acute-exercise sessions at low and high-intensities.
In Phase I, aerobic acute-exercise sessions started with 5 min warm-up on a cycle ergometer at 30% of W max for all intensities. Randomly, two sessions (with at least 48 h interval between the sessions) were performed at low (< 60% VO 2max -90% of aerobic threshold) and high (> 90% VO 2max -midpoint between anaerobic threshold and W max ) intensities until exhaustion or up to 60 min 28  www.nature.com/scientificreports/ Blood collection and Isolation of human peripheral mononuclear cells. In both Phases, blood samples were collected by peripheral puncture of a forearm vein. In Phase I, at baseline (1.5 h after breakfast and immediately before exercise sessions), immediately post-session (post-exercise), and recovery (60 min after the end-session). Blood sample treatment were processed as previously described by the authors 22 . In Phase II, the peripheral mononuclear cell isolation, mediated by adherence protocol in order to obtain the monocytes, was carried out as previously conducted by the authors 29 . A standard breakfast was offered respecting the 25% of total energy value and the recommended macronutrients proportion as previous mentioned 29 .
Flow cytometry. For  www.nature.com/scientificreports/ Statistical analysis. Data normality was verified using the Shapiro-Wilk test and a non-parametric analysis was adopted for peripheral variables of Phase I, as one or more parameters followed a non-parametric distribution, and for the Phase II was adopted parametric analyses given the normal distribution of data. In Phase I, to compare the LPS concentration according to physical fitness, all subjects were classified according to LPS mean values (low or high endotoxemia) and the comparison of the physical fitness was conducted by the Mann-Whitney test. Additionally, the relationship between physical fitness and LPS concentration (at baseline and post-exercise sessions) as well as the relationship between LPS release and cytokine production at low and high intensity exercise sessions was verified through Spearman correlation. In Phase II, to compare the groups according to physical fitness (low and high VO 2max ), the Student's t-test for independent samples was used. Differences in cytokine concentration, in the monocyte culture supernatant, as well as the gene expression, between the physical fitness groups exposed to different stimulus (LPS, Rosiglitazone, and LPS + Rosiglitazone) were analyzed by ANOVA two-way (group X treatment), and when an interaction (treatment x group) was observed, a Bonferroni post hoc test was conducted and the partial eta-square for ANOVA was presented for main effect of time (η 2 ). Statistical significance was set at P < 0.05 and the data were analyzed using the Statistical Package for Social Sciences 22.0 (SPSS Inc. Chicago. IL. USA).
Correlation between endotoxemia mediated by exercise sessions and peripheral cytokine concentrations. The relationship between LPS concentration post-exercise and cytokine productions after low and high intensity exercises is shown in Table 2. At low-intensity exercise there was an inverse correlation between LPS release and IL-6 immediately post-exercise (p = 0.046) and a positive correlation with IL-10 concentration in the recovery (p = 0.011), whereas, at high-intensity exercise, an inverse correlation was observed with IL-6 concentration at pre-exercise (p = 0.013) and recovery period (p = 0.016).

Influence of LPS stimulation on activation or inhibition of PPAR-γ according to physical fitness status.
In order to investigate the influence of LPS on activation or inhibition of PPAR-γ the monocyte culture was performed in the presence and absence of LPS for 24 h. Figure 3 shows the gene expression of PPAR-γ with or without LPS stimulation according to physical fitness status (Low or High VO 2max ) with no statistical differences found in the control treatment when compared the fitness groups; however, when compared the fitness groups in response to LPS treatment, the High VO 2max group demonstrated a significant increase (p = 0.009) in PPAR-γ mRNA.
For AMPK, there was a main effect of treatment (F = 33.057, p < 0.001, η 2 : 0.768) and significant differences between groups (F = 5.930, p = 0.035, η 2 : 0.372), with higher expression in the Low VO 2max group compared with the High VO 2max group, independently of treatment condition. For PGC-1α, there was a main effect of treatment (F = 45.989, p < 0.001, η 2 : 0.821), significant differences between groups (F = 10.532, p = 0.009, η 2 : 0.513), and statistically significant interactions (F = 6.076, p = 0.009, η 2 : 0.378). PGC-1α was more highly expressed in the High VO 2max group compared with the Low VO 2max group in all treatment conditions; in addition, LPS + Rosiglitazone seemed to generate a higher expression when compared with the other treatments in the group with better physical fitness status. For PPAR-γ treated with other stimulus, there was a main effect of treatment (F = 27.859, p < 0.001, η 2 : 0.736), verifying higher expression under isolated Rosiglitazone and LPS + Rosiglitazone stimulations, when compared with LPS treatment, with isolated Rosiglitazone showing higher values when compared with LPS + Rosiglitazone.

Discussion
Here, we showed for the first time that physical fitness status modulates the peripheral and cellular inflammatory response. We identified a positive relationship between LPS concentration and VO 2max , independently of intensities of exercise sessions, as well as an association between LPS and cytokine concentrations, mainly IL-6 and IL-10 post-exercise. In addition, High VO 2max group exhibited higher PPAR-γ, TLR-4, IKK1, and PGC-1α gene expressions and lower IL-10 production in monocytes cell culture after LPS treatment. These results suggest that PPAR-γ pathway is highly expressed in trained individuals and this positive profile may act directly on the exercise-mediated peripheral response. www.nature.com/scientificreports/ Middle or higher endotoxin concentrations at rest are related with unhealthy routine, such as sedentary lifestyle and poor nutrition, as well as with stress responses, such as after exercise training performed at higher volume, duration, and/or intensity. In the present study were evaluated LPS concentrations 1.5 h after breakfast, and found an augmented in endotoxin in higher VO 2max compared to low VO 2max individuals. These results may be explained, at least in part, by previous exercise-induced gastrointestinal symptoms 30 and total energy or nutrient intake 31 , which may lead to higher LPS activity, suggesting that future studies should better control these parameters at rest. On the other hand, our findings are in agreement with previous studies conducted with exercised-animals (exercise training for 4 weeks with gradual increase in exercise training per week (12 m·min −1 -21 m·min −1 for 60 min ) and athletes (triathletes) that observed smaller increases in LPS before an exercise session, when compared with untrained groups, suggesting a possible LPS tolerance in well-trained subjects 9,13 .
It is noteworthy to emphasize that LPS is a gut-derived bacterial endotoxin resulting from dysbiosis stimulated by hypercaloric/hyperlipidic diets, as well as sedentary behavior, favoring the activation of pro-inflammatory signaling pathways 32,33 . It is important to highlight that gut-microbiota act as an endocrine organ fully integrated in the host metabolism given its ability to produce and release biomarkers, as hormones and short chain fatty acids (SCFAs). Besides, its bacterial diversity change in response to several internal and external stimuli, as nutrition habits and physical exercise practice 34 .
Considering the gut-microbiota, it is well established in the literature that physical fitness status may directly influence the diversity/type and abundance of several intestinal bacteria and, in this context, it is suggested that regular physical exercise practitioners individuals have greater diversity and abundance of bacteria beneficial to health 35,36 . In this sense, according with gut-microbiome composition, the production and release of metabolites derived from microbial fermentation (i.e. butyrate, propionate and acetate) may be affected and these metabolites are able to regulate several cell and tissue functions given that these metabolites have pro and/ or anti-inflammatory properties 37,38 .
In general, regular physical exercise practice acts as a protective behavior, modulating intestinal bacteria composition/diversity directly related to modulations in metabolism and immunity, especially to improve the SCFAs production and restore the intestinal epithelial barrier 39 . Indeed, other studies suggested that exercise training modulate Kupffer cells function enhancing the in vivo endotoxin clearance in animals, impacting directly the inflammatory response 40,41 . Thus, gut-microbiota and nutritional habits, directly associated with gut modulation, are determining factors for the regulation of the cell functionality and both parameters must be considered. www.nature.com/scientificreports/ Changes in LPS concentration after an exercise session may directly, or indirectly, influence inflammatory cytokine release; our results showed that individuals with High VO 2max exhibited small changes in LPS concentrations post-exercise, however, we found higher IL-6 concentrations immediately after exercise performed at low and high intensities and lower IL-10 concentrations 60-min after low-intensity exercise in this group, suggesting that, inflammatory responses may be directly associated with physical fitness status, and that cytokine modulation in trained individuals occurs, at least in part, through LPS-independent mechanisms.
In this way, Ortega 42 proposes that there are bioregulatory effects arising from exercise practice dependent on individual metabolic profile, given that in healthy individuals exercise training should stimulate an inflammatory response whereas in individuals with an installed inflammatory profile the same effort should mediate an anti-inflammatory response. According to some studies, physically active men and athletes have higher IL-6 mRNA expression in skeletal muscle after an exercise training session 43,44 , whereas higher IL-10 expression and release may be observed in individuals with an established inflammatory profile, such as sedentary and overweight individuals, as an attempt to inhibit an inflammatory condition represented by the high expression and release of inflammatory cytokines 45 .
It is important to highlight that IL-6 acts as an "energy sensor in response to decreased glycogen stores observed across effort sessions, increasing the bioavailability of energy substrate by glucose metabolism regulation via PI3-K/AKT, GLUT-4 activity and lipolysis 20,21 . On the other hand, as previously mentioned, IL-6 has a pleiotropic function, having anti and/or pro inflammatory properties. Considering its pro-inflammatory role, Peng and colleagues 46 investigated the relationship between homocysteine, PPAR-γ and IL-6 gene expression in peripheral blood mononuclear cell cultures treated with PPAR-γ activators (troglitazone and rosiglitazone), and observed that homocysteine can promote the expression of inflammatory factors, such as IL-6. However, IL-6 concentrations were significantly reduced in PBMC treated with PPAR-γ activators culture supernatants, suggesting that the PPAR-γ activation can inhibit the pro-inflammatory factors production, mainly in population with inflammatory diseases. All these reports explain, at least in part, our findings regarding the cytokine concentrations in peripheral blood and monocyte culture supernatant, evidencing an intimate and dependent relationship between inflammatory responses, molecular activations and physical fitness status.
Our results seem to support the initial hypothesis that some molecular adaptations and mechanisms, mainly linked with PPAR-γ activation/signaling, may be imposed by long-term and regular exercise practice parallel with physical fitness status improvements which, at least in part, were proven when higher gene expression of TLR-4, IKK1, PGC-1α, and PPAR-γ was found in individuals with High VO 2max , independently of monocyte phenotypic profile. Regarding TLR-4, differences were observed between physical fitness status and monocyte cell culture treatments, with higher expression in the High VO 2max group, especially under LPS + Rosiglitazone stimulation, suggesting an increase and/or immune response improvement against stressor stimuli which may be associated with adaptations imposed by exercise training. In this line, Nickel and colleagues 47 showed that aerobic training (≤ 40 km/week for obese and ≥ 55 km/week for lean athletes) during 10-weeks induced an increase in the expression of TLR receptors, mainly TLR-4 and TLR-7, supporting our hypothesis that chronic exercise training may be able to modulate the TLR-activated axis.
Generally, TLR-4 activity is associated with inflammatory signaling, however its action is directly stress or agent and/or binder-dependent given that the TLR-4/MyD88 signaling pathway results in pro-inflammatory production and release, while the signal transduction by TLR-4/TRAM/TRIF/TRAF3 results in IFN-β production (related to innate immune response) and in anti-inflammatory cytokine release 48 . Therefore, TLR-4 may be associated with a better inflammatory response and, consequently, acts in a protective way on metabolism.
In this scenario, another important protein related to inflammatory response is IKK1 (or IKKα), a catalytic subunit of the IkB kinase complex that keeps NF-kB sequestered in the cytosol. In the present study we found higher expression of IKK1 in the High VO 2max group, when compared with the Low VO 2max group, leading us to hypothesize that a positive feedback occurs in the regulation of this protein in order to stimulate an alternative NF-kB pathway (maybe linked with IKK1-RelB activation).
Concerning PPAR-γ upstream marker signaling, PGC-1α is a co-activator of all PPAR isoforms, synergistically coordinating metabolic and inflammatory effects, such as AMPK-mediated mitochondrial regulation and expression of glucose membrane transport protein (GLUT-4) 49 , and insulin sensitivity 50 , as well as repressing inflammatory cytokines production through reduction in NF-kB phosphorylation 51 . According to Krämer and colleagues 52 , higher PGC-1α and PGC-1β expression as well as PPAR-α and PPAR-δ are observed in athletes (cyclists), when compared with physically active subjects; our findings corroborate these results given that High VO 2max group exhibited greater expression of the PPAR co-activator and was stimulus-independent, leading us to hypothesize that physical fitness status enables protective metabolic adaptations, mainly through imposing an anti-inflammatory environment even when conjugated pro and anti-inflammatory stimulus.
The positive modulation of the PGC-1α/PPAR-γ axis observed in the molecular environment may orchestrate the anti-inflammatory responses observed in the peripheral/systemic blood after acute aerobic exercise explains the circulating IL-6 and IL-10 concentrations. These findings allow us to understand that beneficial peripheral responses to exercise training are modulated directly by the molecular/cellular adaptations of "individual biological machinery" and a recent study conducted by Dorneles and colleagues 53 corroborates and strengthens our hypothesis.
Considering the interesting findings of the present study, some limitations should be mentioned, such as the absence of information about relative and functional quantities of the main proteins investigated herein in order to verify if the gene expression is accompanied by a greater active and functional quantity of all proteins related to the PPAR-γ pathway. In addition, not having analyzed and explored other signaling pathways that could be in crosstalk with our experimental model, especially in cell culture with the drug combination, is an important limitation to be considered. Future studies should be conducted to answer these aforementioned gaps to better understand the metabolic mechanisms as well as the interactions between signaling pathways. Moreover, other Scientific RepoRtS | (2020) 10:14094 | https://doi.org/10.1038/s41598-020-70731-6 www.nature.com/scientificreports/ studies should consider different populations, especially patients with an inflammatory profile, to identify the impacts of exercise training in the illness scenario as well as its possible ability to mediate reprogramming at the cellular level.
In conclusion, even with elevated endotoxemia at rest 1.5 h after breakfast, individuals with greater VO 2max exhibited higher IL-6 concentration in peripheral blood post-acute aerobic exercise sessions, and lower IL-10 concentrations during recovery, evidencing that physical fitness status directly impacts the inflammatory responses through LPS-independent mechanisms. The anti-inflammatory effects associated with regular exercise training and, consequently, physical fitness status may be explained by the greater expression of proteins and cellular receptors, such as IKK1, TLR-4, and PGC-1α, in the group with High VO 2max displaying an extremely efficient cellular framework for fast and successful responses orchestrated by PPAR-γ (Fig. 7).