Effects of acute endurance exercise and exhaustive exercise on innate immune signals induced by mtDNA

Abstract

Objective: Numerous studies have shown that mitochondrial DNA (mtDNA) can trigger innate immune signaling, and exercise can induce mitochondrial stress. Therefore, this study is aimed at investigating the influence of different types of acute exercise on the innate immune signaling triggered by mtDNA.

Methods: Male C57BL/6 mice (n = 18) were randomly and equally divided into three groups. They were control group, acute moderate-intensity endurance exercise group (AMIE), and 3-day exhaustive exercise group (EE) respectively. Mice were sacrificed immediately after exercise. The spleen, liver, and blood were taken for analysis.

Results: The amount of mtDNA in the liver cytoplasm and plasma was significantly decreased after AMIE (p < .05). However, the amount of mtDNA in plasma was increased after EE (p < .05). The mRNA expression of TFAM, and most TLR9 and cGAS/STING signaling pathway-related genes in the liver and spleen was markedly elevated, whereas the expression of those genes in leukocytes was reduced after AMIE. Furthermore, AMIE significantly decreased the protein expression of NLRP3 inflammasome in the liver (p < .05) and STING in spleen (p < .01). Also, AMIE and EE caused a drop in circulating IFN-β levels (p < .05).

Conclusion: A single bout of moderate-intensity exercise reduces mtDNA-induced innate immune signaling and suppresses inflammatory responses by decreasing hepatic cytoplasmic and circulating mtDNA. However, repeated bouts of exhaustive exercise stimulate innate immune signaling by increasing levels of circulating mtDNA.

Introduction

Mitochondria are double-membrane organelles that regulate the metabolism of various life activities in cells, such as energy metabolism, ROS production, programmed cell death, calcium ion metabolism, and mitochondrial protein input. ¹ Mitochondria and cell pattern recognition receptors (PRRs) correlate to innate immunity. PRRs recognize pathogen-associated molecular patterns (PAMPs) and activate the relevant signaling pathways to produce type I interferons (IFN-I), inflammatory cytokines, and chemokines to eliminate external pathogenic microorganisms. ¹ The major PAMPs, the conservative molecules of pathogenic microorganisms, are nucleic acids, including DNA (unmethylated CpG sequence), double-stranded RNA, single-stranded RNA, 5’-triphosphate RNA, as well as lipoproteins and cell surface glycoproteins. ² MtDNA, like bacterial DNA, contains many unmethylated CpG sequences, so PRRs can recognize mtDNA to trigger the innate immune response, such as Toll-like receptors (TLRs) and nod-like receptors (NLRs). MtDNA is released from mitochondria into the cytoplasm to trigger an innate immune response upon opening of the mitochondrial permeability transition pore or mitochondrial damage.^3,4 For example, the interaction between mtDNA and TLR9 can activate NF-κB and IFN regulatory factor 7 (IRF7) signaling by myeloid differentiation factor 88 (MyD88), increasing the production of pro-inflammatory cytokines and IFN-I. ³ cGAS can capture mtDNA to generate the 2’3’-cGAMP to activate interferon gene stimulating protein (STING). Then, the activated STING engages Tank-binding kinase 1 (TBK1) to phosphorylate IRF3, enhancing the IFN-I responses. Furthermore, STING can activate the NF-κB signaling to enhance inflammatory response.^5,6 Moreover, mtDNA enables NLRP3 inflammasome activation, which promotes the production of IL-1β and IL-18 by mediating downstream signals and enhances antibacterial immunity and antiviral immunity. ⁷

IFN-I is primarily produced by innate immune and partial non-immune cells, such as fibroblasts and epithelial cells. ⁸ The circulating immune cells in peripheral blood perform essential immune surveillance functions. Furthermore, the spleen is the largest lymphoid organ in the body. It contains a large number of mature immune cells, including T cells and B cells, dendritic cells (DC), different subsets of macrophages, and neutrophils.^9,10 Moreover, Kupffer cells, the resident liver macrophages, which constitute 80–90% of tissue macrophages and represent about 35% of non-parenchymal hepatocytes. ¹¹ Also, the liver contains a high frequency of NK cells relative to the spleen and peripheral blood. ¹² Recently, it has been reported that the increased amount of mtDNA in cytosol engages the cGAS and promotes STING/IRF3-dependent signaling, potentiating IFN-I responses and conferring broad viral resistance. ⁵ More recently, voluntary running has been shown to increase common lymphoid progenitors (CLP) frequency in load-bearing long bones. ¹³ Endurance exercise can reduce the rate of upper respiratory tract infection (URTI) by enhancing the activity of immune cells in the circulation.^14,15 However, a J-curve relationship exists between URTI risk and increasing exercise workloads. ¹⁶ Thus, the exercise of different intensities may affect innate immunity differently. Mitochondrial stress can trigger the release of mtDNA into the cytosol. ⁵ The gut microbiota can activate the cGAS-STING-IFN-I axis and antiviral immunity by releasing bacterial DNA-containing membrane vesicles (MVs) through circulation to distal host cells. ¹⁷ Notably, IFN-I can enhance the functional activity of immune cells.^18,19 It was found that a single bout of moderate-intensity exercise tends to reduce skeletal muscle mtDNA content. Moreover, short repeated high-intensity exercise can decrease skeletal muscle mtDNA content. ²⁰ These changes may be related to oxidative stress in skeletal muscle mitochondria. ²¹ Notably, circulating mtDNA levels are elevated after a single bout of strenuous physical exercise, most likely due to increased release of skeletal muscle mtDNA.^22,23 Thus, exercise likely affects innate immunity by modulating mtDNA release. This study aims to understand the critical role of acute exercise of different intensities in the innate immune response triggered by mtDNA.

Materials and methods

Animal and exercise programs

Eighteen male C57BL/6J mice (4 weeks old) were obtained from the SLAC Laboratory Animal Research Center (Shanghai, China) and housed in a temperature-controlled room at 22 ± 2°C, 12 h light-back cycle. Mice were fed with feed, free diet, and water. The pad material was changed once every 1–2 days. All mice underwent formal training at 6 weeks of age after 2 weeks of acclimatization. The mice were randomly assigned into control, AMIE, and exhaustive exercise (EE) groups. The control mice (n = 6) did not participate in any exercise. One week before exercise, the AMIE mice (n = 6) were given one adaptive exercise with an exercise intensity of 13 m/min for 20 min to acclimate to the treadmill (BLD-PT6, BileadBiosci). After that, those mice were exercised for 40 min at 15 m/min (80%VO2max). ²⁴ To induce mitochondrial stress, the EE mice (n = 6) were forced exhaustive treadmill running for three consecutive days. ²⁵ The EE mice received 3 days of adaptive treadmill training. On day 1, mice were given 5 min to acclimate treadmill then exercised at 13 m/min, 15 m/min, and 17 m/min for 5 min for each, all at an incline of 10%. On day 2, the mice immediately exercised at 13 m/min, 15 m/min, 17 m/min, 18.5 m/min for 5 min each. On day 3, mice exercised at 15 m/min, 17 m/min, 18.5 m/min, 20 m/min for 5 min each. After familiarization, mice were allowed 3 days to recover before beginning the EE protocol. The EE protocol was as follows: At a 10% incline, the initial exercise speed was set at 13 m/min for 8 min, then increased by 2 m/minute every 3 min until reaching a speed of 20 m/minute for 3 min each. Mice were exercised for 10 min at 20 m/min, then increased by 2 m/minute every 10 min until at 30 m/min, when the speed was no longer increased until exhaustion. ²⁵ Exhaustion was determined when the mice were unwilling to run as indicated despite the negative stimulus.

Sample preparation

Immediately after the exercise, blood was collected from the heart of mice anesthetized with inhaling isoflurane with a concentration of 1–3%. Separation of plasma and leukocytes: The blood was added to the anticoagulant tube containing EDTA-2K and centrifuged at 1200g for 10 min at room temperature. The supernatant was collected and centrifuged at 1300g for 2 min at room temperature. Then, the supernatant was aspirated and subsequently stored at −80°C for later analysis. The resulting pellet was suspended erythrocyte lysate (Sangon, China) for 5 min and then centrifuged at 4000 rpm for 3 min, after which the supernatant was carefully discarded. The pellet was resuspended in PBS and centrifuged at 4000 rpm for 3 min. The supernatant is removed and the cell pellets are stored at −80°C. Plasma cell-free mitochondrial DNA was extracted using genomic DNA extraction kit (Beyotime, China).

For measurement of mtDNA in the cytosol, tissues (liver and spleen) with the same mass of each sample were homogenized with a glass homogenizer in 100 mm Tricine-NaOH solution, pH 7.4, containing 0.25 m sucrose, 1 mm EDTA and protease inhibitor, then were centrifuged at 700 g for 10 min at 4°C. The volume of the supernatant was normalized, followed by centrifugation at 10,000 g for 30 min at 4°C to produce a supernatant corresponding to the cytosolic fraction without mitochondria. ²⁶ Genomic DNA extraction kit (Beyotime, China) was then used for DNA isolation in the cytosolic fraction. The copy number of DNA encoding cytochrome c oxidase 1 (COX1) was measured by quantitative real-time PCR with the same volume of the DNA solution. The remaining tissues were snap-frozen in liquid nitrogen and further stored at −80°C for further use.

RNA extraction and quantitative real-time PCR

Total RNA was extracted from frozen tissue with the use of TRIZOL (Invitrogen) according to the manufacturer’s specifications. The total RNA extracted was used 1 μg for reverse transcription synthesis of cDNA. ReverTra Ace qPCR RT Kit (TOYOBO, Osaka, Japan) was used for reverse transcription. Fluorescence quantitative PCR was performed using ABI QuantStudio 3 real-time fluorescence quantitative PCR instrument and software (ABI, California, USA) with 2 μl of cDNA, 10 μl Hieff® qPCR SYBR Green Master Mix (Low Rox Plus) (yeasen, China), 0.4 μl forward primer (10 μm), 0.4 μl reverse primer (10 μm) and 7.2 μl H2O. The thermocycling conditions for PCR were as follows: 95°C for 5 min, followed by 40 cycles of 95°C for 10 s, 60°C for 20 s, and 72°C for 20 s. Copy number of cytoplasmic free mtDNA was standardized by COX I fragment compared with nuclear DNA coding GAPDH. ²⁷ The remaining target genes were standardized by 18S rRNA. Mice PCR target gene primer sequences are as follows:

TLR9: F:CTCCCAACATGGTTCTCCGTC

R:AGGCTTCAGCTCACAGGGTA

TOM70: F:TTATCACCACCGAGGACAGC

R:GCCTGGCGATACAATGCAAAA

TBK1: F:AGAGTACCTGCATCCGGACA

R:CGGTAGCCCCGTACTTCTTC

IRF3: F:CGGAGGCTTAGCTGACAAAGA

R:ATGCTCTAGCCAGGGGAGGA

TFAM:F:CCGGGCCATCATTCGTCG R:CAGACAAGACTGATAGACGAGGG

IRF7:F:AGCTTGGATCTACTGTGGGC R:GGGTTCCTCGTAAACACGGT

IFN-β:F:CGTGGGAGATGTCCTCAACT R:CTGAAGATCTCTGCTCGGACC

STING: F:GCCCTGTCACTTTTGGTCCT

R:TGGAGTATGGCATCAGCAGC

IL-6:F:CCCCAATTTCCAATGCTCTCC R:CGCACTAGGTTTGCCGAGTA

cGAS:F:GAACATGTGAAGATTTCTGCTCC R:GACTCAGCGGATTTCCTCGT

MAVS F:CGCCCTGTGTTGGACATTCT

R:AAGCCCGCAGTCGATCAAGA

Myd88:F: ATGACCCCCTAGGACAAACG

R:AGGCTGAGTGCAAACTTGGT

COXI F:TCCAGCTATACTATGAGCCTT

R:AAACAATTCCGGTTAGACCA

GAPDH:F:TCCCAGCTTAGGTTCATCAGG R:AGATTGCTACGCCATAGGTC

18SrRNA

F:AGCTTGCGTTGATTAAGTCCCT R:GCCTCACTAAACCATCCAATCGG

All primers were synthesized by Sangon (Shanghai, China). All the samples were analyzed in duplicate and data were analyzed by 2 ^−ΔΔCT method.

ELISA

To investigate the changes in inflammatory levels and immune factors in the circulation immediately after exercise, plasma IL-6 (Shanghai mlbio) and IFN-β (Shanghai mlbio) levels were analyzed according to the manufacturers’ instructions.

Western blotting

Total proteins from tissues were extracted using the pre-cooled RIPA lysis buffer (Thermo, USA) with protease inhibitor cocktail (Beyotime, China). BCA protein assay was used for protein quantification (Beyotime, China), and the protein (25 μg) was separated by SDS-PAGE. PVDF membrane (Millipore) was used and was blocked by 5% non-fat milk at room temperature for 1h, and the primary antibody was incubated overnight at 4°C. The following antibodies were used: anti-cGAS (31,659S, CST, 1:1000), anti-STING (ab92605, Abcam, 1:1000), anti-IRF3(ab68481, Abcam, 1:1000), anti-IFN-β(DF6471, Affinity, 1:500), anti-TOM70 (ab83841, Abcam, 1:1000),anti-TBK1(ab40676, Abcam, 1:1000),anti-β-actin (T0022, Affinity, 1:3000), anti-TFAM(ab47517, Abcam, 1:1000), anti-NLRP3 (ab210491, Abcam, 1:1000), anti-pro-IL-1β/IL-1β(AF5103, Affinity, 1:500). After overnight incubation with primary antibody, the PVDF membrane was incubated with secondary antibody (1:10,000 Jackson) for 1 h at room temperature and ECL development (Biorad,USA). Densitometry was semi-quantified with ImageJ 1.8.0 software, and results were normalized with β-actin.

Statistical analysis

The data were represented by mean ± SEM, and GraphPad Prism 7 was used for one-way ANOVA followed by Dunnett’s many-to-one test. Compared with control group, * represents p < .05, ** represents p < .01, *** representsp < .001.

Results

AMIE reduces the amount of mtDNA in cytoplasm of liver and plasma

To assess different types of acute exercise on the content of free mtDNA in the cytoplasm of different immune organs and plasma, the content of mtDNA at these sites was evaluated immediately after exercise. Interestingly, as illustrated in Figure 1(a), (c), (g) and (i), AMIE decreased the content of free mtDNA in liver cytoplasm and plasma (p < 0.05), whereas EE increased the amount of free mtDNA in plasma compared to control (p < .05), suggesting that different intensity of acute exercise regulates the amount of free mtDNA in blood. Moreover, EE increased the number of nuclear DNA fragments GAPDH and 18S in the cytoplasm of the spleen(p < .05) (Figure 1(e) and (f)), which also suggests EE is probably to cause damage to splenocyte that sheds more nuclear DNA fragments into the cytoplasm.

Figure 1.

The effect of different intensity of acute exercise on free mtDNA content from the cytoplasm in liver, spleen, and the blood circulation. (a, b) Free mtDNA (COX I, (a)) and nuclear DNA (GAPDH, (b)) in the cytoplasm of liver in mice after different intensity of acute exercise (n = 6). (c) Ratio of mtDNA to nuclear DNA in liver (n = 6). (d, e, f) Free mtDNA (COX I, (d)) and nuclear DNA (GAPDH and 18S, (e), (f)) in the cytoplasm of spleen in mice after different intensity of acute exercise (n = 6). (g, h) Cell-free mtDNA (COX I, (g)) and nuclear DNA (GAPDH, (h)) in the plasma of mice after different intensity of acute exercise (n = 6). (i) Ratio of mtDNA to nuclear DNA in plasma (n = 6). *, and ** indicates p < .05, and p < .01 versus Control.

Exercise does not alter the TFAM expression in liver

Since mtDNA is released to cytosol by TFAM deficiency, ⁵ we measured the TFAM mRNA levels and protein expression in liver, spleen, and leukocytes. The mRNA level of TFAM in the liver was significantly increased in EE compared to control (p < .01). However, AMIE considerably decreased TFAM mRNA levels in leukocytes compared to control (p < .01) (Figure 2(a)). In addition, neither AMIE nor EE significantly altered the TFAM protein level in liver (Figure 2(b)), suggesting that exercise-induced changes in content of free mtDNA in liver cytoplasm are not related to the expression of hepatic TFAM. Unfortunately, we did not detect the expression of TFAM protein in the spleen and leukocytes.

Figure 2.

Effects of acute exercise of different intensity on expression of TFAM in liver, spleen, and leukocytes. (a) Gene expression levels of TFAM in liver, spleen, and leukocytes after acute exercise of different intensity (n = 6). Expression was normalized to 18S RNA. (b) Expression of TFAM in liver after acute exercise of different intensity (n = 6). The density of each band was normalized to β-actin. ** indicates p < .01 versus Control.

AMIE and EE have opposite effects on TLR9 and cGAS/STING signaling

To further clarify the changes in mtDNA downstream signaling, we measured the expression of TLR9 and cGAS/STING signaling-related molecules in the liver, spleen, and leukocytes. We observed that EE significantly increased the mRNA levels of IRF7 (p < .05), TBK1 (p < .01), IRF3 (p < .05), and IFN-β (p < .05) in the liver compared to control (Figure 3(a)). Except for IRF3 and IFN-β, EE significantly increased the mRNA levels of TLR9 and cGAS/STING signaling-related molecules and IL-6 in the spleen compared to control (p < .05) (Figure 3(b)). However, the mRNA levels of TLR9 (p < .001), MyD88 (p < .05), TBK1 (p < .001), and IRF3 (p < .01) in AMIE were significantly lower in leukocytes compared to control (Figure 3(c)). It is worth noting that the knockout of cGAS in immune cells fails to produce IFN-I and other antiviral cytokines in response to DNA virus infection, ²⁸ indicating that cGAS plays a critical role in the innate immune response. Therefore, we detected the changes in protein levels of cGAS-STING pathway. We observed that AMIE reduced the STING protein expression in the spleen (p < .05) (Figure 3(e)), which reflects the anti-inflammatory effect of AMIE. In addition, AMIE and EE significantly reduced the level of IFN-β in plasma (p < .05) (Figure 3(g)), suggesting that the body may have weakened immunity immediately after exercise, susceptible to pathogen infection. However, there was no significant change in the expression levels of cGAS/STING signaling-related proteins in liver and leukocytes (Figure 3(d) and (f)). In addition, there was no significant difference in the changes of plasma IL-6 levels among the groups (Figure 3(h)).

Figure 3.

Effects of acute exercise of different intensity on TLR9 and cGAS/STING signaling in liver, spleen, and leukocytes. (a, b, c) Effects of acute exercise of different intensity on mRNA expression of TLR9 and cGAS/STING signaling related molecules in liver (a), spleen (b), and leukocytes (c) (n = 6). Expression was normalized to 18S RNA. (d, e, f) Effects of acute exercise of different intensity on protein expression of TLR9 and cGAS/STING signaling related molecules in liver (d), spleen (e), and leukocytes (f) (n = 6). Expression was normalized to 18S RNA. (g, h) Effects of different intensity of acute exercise on plasma levels of IFN-β (g) and IL-6 (h) (n = 6). The density of each band was normalized to β-actin. *, **, and *** indicates p < .05, p < .01, and p < .001 versus Control.

AMIE represses the activation of the NLRP3 inflammasome in liver

MtDNA also engages the NLRP3 inflammasome to trigger an inflammatory response. ⁷ Therefore, we also examined the downstream signaling of NLRP3 inflammasome. As shown in Figure 4(a), AMIE significantly decreased the protein expression of NLRP3 inflammasome in the liver compared to control (p < .05). However, the biologically functional IL-1β from pro-IL-1β (IL-1β/pro-IL-1β) in the liver and spleen did not change after different intensity exercises (Figure 4(b) and (c)). Moreover, we did not detect the expression of NLRP3 inflammasomes in the spleen and leukocytes, possibly due to low expression of NLRP3 inflammasomes.

Figure 4.

AMIE reduces the expression of NLRP3 inflammasome in liver. (a, b) Effects of acute exercise of different intensity on expression of NLRP3 inflammasome (a) and IL-1β (IL-1β/pro-IL-1β) in liver (b) (n = 6). (c) The effect of different intensity of acute exercise on the expression of IL-1β (IL-1β/pro-IL-1β) in spleen (n = 6). The density of each band was normalized to β-actin. * indicatesp < .05 versus Control.

Acute exercise does not alter the protein level of TOM70 in different immune tissues

Given that the interaction between translocases of outer membrane 70 (TOM70) and MAVS and overexpression of TOM70 can enhance IRF3-mediated generation of IFN-β, ²⁹ we measured the expression level of TOM70 in different immune tissues. We observed that EE significantly increased the mRNA level of TOM70 in the liver (p < .01) and spleen (p < .05), while AMIE significantly decreased the mRNA level of TOM70 in leukocytes compared to control (p < .01) (Figure 5(a)). However, neither AMIE nor EE alters the TOM70 protein expression in the liver, spleen, and leukocytes (Figure 5(b)).

Figure 5.

Effects of acute exercise of different intensity on TOM70 expression of liver, spleen, and leukocytes. (a) Effects of different intensity of acute exercise on the level of TOM70 mRNA in liver, spleen and leukocytes (n = 6). Expression was normalized to 18S RNA. (b) Effects of different intensity of acute exercise on the protein expression of TOM70 in liver, spleen and leukocytes (n = 6). (c) Possible mechanisms of mtDNA regulation of innate immunity under exercise stress. The density of each band was normalized to β-actin. * indicates p < .05 versus Control.

Discussion

Based on the epidemiological investigation, individuals exercising moderately may lower their risk of URTI. In contrast, those undergoing strenuous exercise regimens may have a higher risk. ¹⁶ Thus, moderate exercise may strengthen the innate immune system, while strenuous exercise may impair the innate immune response. We hypothesized that moderate exercise could increase the amount of free mtDNA in the cytoplasm to stimulate the innate immune response. However, we observed that a single bout of moderate-intensity exercise significantly reduced the free mtDNA content in the cytoplasm of the liver and plasma. It was noted that cell-free plasma mtDNA declines immediately after 60% VO2max aerobic exercise for 90 min. ³⁰ Moreover, circulating mtDNA levels in professional male volleyball players are lower than those of normal people. ³¹ These studies suggest that exercise with an appropriate intensity may decrease mitochondrial damage and release of mtDNA, which may contribute to maintaining mitochondrial health. However, our data showed that repeated bouts of exhaustive exercise resulted in a substantial accumulation of mtDNA in plasma and increased cytoplasmic nuclear DNA fragments in spleen compared to control. Indeed, plasma cell-free mtDNA levels are substantially elevated, and the spleen is damaged immediately after exhaustive exercise.^32,33 Moreover, strenuous exercise can cause skeletal muscle damage. ³⁴ Therefore, the increase in circulating mtDNA levels may be related to the release of mtDNA caused by the damage to various tissues and organs. In turn, circulating mtDNA can cause damage to various tissues and organs. For example, circulating mitochondrial DNA promotes acute kidney injury and myocardial ischemia-reperfusion injury by activating innate immune signaling.^35,36 Furthermore, the increase of mtDNA in plasma may also be derived from the release of mtDNA leukocytes. For example, viable neutrophils release mtDNA after the stimulation of CSF-2 and complement C5a, in which the process is not associated with apoptosis and necrosis. ³⁷ In addition, strenuous exercise increases CSF-2 and C5a in blood circulation.^38,39 Therefore, a rise in cf mt-DNA after exhausting exercise may result from multiple reasons. Notably, these increased levels of circulating mtDNA are likely due to release from skeletal muscle.^22,23

We further examined the TFAM level that determines the release of mtDNA. ⁵ We found that no difference in TFAM protein expression in liver among all groups, which means that the reduction of cytoplasmic mtDNA is independent of the level of TFAM in the liver. However, repeated bouts of exhaustive exercise increased the mRNA level of TFAM in the liver. In contrast, a single bout of moderate-intensity exercise decreased the mRNA level of TFAM in leucocytes. Furthermore, TFAM is linked to the biogenesis of mitochondria. ⁴⁰ These results suggest that repeated bouts of exhaustive exercise may be related to mitochondrial biogenesis in the liver. However, the significance of reduced mRNA expression of TFAM in leukocytes caused by a single bout of moderate-intensity exercise remains unclear.

Moderate exercise can play an anti-inflammatory role in various ways, such as reducing macrophage infiltration and inducing the phenotypic switching from M1 to M2 polarization of macrophages in white adipose tissue, decreasing the expression of TLR4 on the monocyte surface and the numbers of pro-inflammatory monocytes in circulation. ⁴¹ Indeed, 6 months of aerobic exercise training down-regulates pro-inflammatory genes and up-regulates anti-inflammatory genes in leukocytes. ⁴² However, acute high-intensity exercise has a pro-inflammatory effect. It has been demonstrated that strenuous exercise dramatically increases the plasma levels of TNF-α and IL-1β, ⁴³ and induces mitochondrial damage and increased expression of TNF-α in skeletal muscle. ⁴⁴ Our findings are consistent with their reports. We observed that the mRNA level of TLR9, MyD88, TBK1, and IRF3 in leucocytes, the protein expression of STING in spleen, and the content of IFN-β in plasma were decreased immediately after a single bout of moderate-intensity exercise, which supports the anti-inflammatory effect of moderate exercise. Moreover, the amount of mtDNA in the liver cytoplasm and plasma was remarkably decreased immediately after a single bout of moderate-intensity exercise, which may cause the decline in plasma IFN-β. Thus, endurance exercise slows mtDNA-induced innate immune responses, which appear detrimental to anti-infection immunity. It has been proposed that the decrease in lymphocyte numbers after endurance exercise is related to the redistribution of immune cells to peripheral tissues to enhance immune surveillance, ⁴⁵ which is also consistent with our finding that the expression levels of proinflammatory cytokine genes in circulating leukocytes are decreased after a single bout of moderate-intensity exercise. Notably, 6 weeks of voluntary running increases the expression of CXCL12 in bone marrow (LepR+) stromal cells and reduces the proliferation of hematopoietic stem and progenitor cells (LSKs) by reducing leptin secretion in mouse adipose tissue, thereby reducing the production of inflammatory leukocytes. However, in this context, the reduction in the number of inflammatory leukocytes does not weaken the immunity of the mice, but enhances the anti-infection ability of the mice. ⁴⁶ This provides strong evidence that the anti-inflammatory effects of moderate exercise can enhance the ability to fight infection.

The repeated bouts of exhaustive exercise increased the mRNA expression of IRF7, TBK1, IRF3, and IFN-β in liver. They elevated the gene expression of TLR9 and cGAS/STING signaling-related genes in the spleen in addition to IRF3 and IFN-β. These may be associated with DAMPs released by skeletal muscle injury. ³⁴ However, our data also indicated no difference in expression levels of these proteins, and the reasons remain unclear. Moreover, the IFN-β level in plasma was also reduced immediately after repeated bouts of exhaustive exercise. It has been found that the mitochondrial membrane potential is reduced, and the propensity of apoptosis is increased in leukocytes after high-intensity exercise training. ⁴⁷ Thus, we concluded that this might be why the circulating cell-free mtDNA content is increased, but the level of IFN-β is decreased. However, the precise mechanism remains to be elucidated. NLRP3 inflammasome activation can lead to hepatocyte pyroptosis, liver hepatitis, and fibrosis. ⁴⁸ It is worth noting that voluntary wheel running decreases the expression of NLRP3 inflammasome in the heart of obese mice. ⁴⁹ Indeed, we found that a single bout of moderate-intensity exercise reduced the expression of NLRP3 inflammasome in the liver, suggesting that moderate exercise may help to protect the liver and maintain liver health. These results suggest the anti-inflammatory effect of a single bout of moderate-intensity exercise and the pro-inflammatory effect of repeated bouts of exhaustive exercise. Based on the above discussion, we propose a possible mechanism by which mtDNA regulates innate immunity under exercise stress (Figure 5(c)). Furthermore, we observed that the TOM70 gene expression in liver and spleen was increased after repeated bouts of exhaustive exercise. In contrast, the mRNA level of TOM70 in leukocytes was reduced after AMIE. However, we found no difference in the expression level of TOM70 protein between groups, suggesting that the effect of exercise on innate immunity may be independent of TOM70 protein expression. No report on the effects of exercise on TOM70 in the liver, spleen, and leukocytes, and the specific mechanism remains to be further studied. Notably, although we explored mtDNA-induced innate immune signaling in some immune organs, this study still has some limitations. First, skeletal muscle is an organ directly stimulated by exercise. However, we did not detect the changes in cytoplasmic mtDNA content of skeletal muscle after exercise. Thus, it cannot be determined whether the elevated levels of plasma mtDNA after exhaustive exercise are derived from skeletal muscle. Second, the sample size was not calculated in this study. Finally, repeated bouts of exhaustive exercise result in changes in the production of nuclear DNA fragments in the cytoplasm of the spleen. Therefore, the corresponding mtDNA changes could not be normalized in this study.

Conclusion

In summary, we conclude that acute moderate-intensity endurance exercise reduces mtDNA-induced innate immune signaling and suppresses inflammatory responses by decreasing hepatic cytoplasmic and circulating mtDNA. However, repeated bouts of exhaustive exercise stimulate innate immune signaling by increasing levels of circulating mtDNA.