Aerobic exercise improves LPS-induced sepsis via regulating the Warburg effect in mice

We investigated the impact of aerobic exercise (AE) on multiple organ dysfunction syndrome (MODS), aortic injury, pathoglycemia, and death during sepsis. ICR mice were randomized into four groups: Control (Con), Lipopolysaccharide (LPS), Exercise (Ex), and Exercise + LPS (Ex + LPS) groups. Mice were trained with low-intensity for 4 weeks. LPS and Ex + LPS mice received 5 mg/kg LPS intraperitoneally for induction of sepsis. Histopathological micrographs showed the organ morphology and damage. This study examined the effects of AE on LPS-induced changes in systemic inflammation, pulmonary inflammation, lung permeability, and bronchoalveolar lavage fluid (BALF) cell count, oxidative stress-related indicators in the lung, blood glucose levels, plasma lactate levels, serum insulin levels, plasma high-mobility group box 1 (HMGB1) levels, glucose transporter 1 (Glut1) and HMGB1, silent information regulator 1 (Sirt-1), and nuclear factor erythroid 2-related factor 2 (Nrf-2) mRNA expression levels in lung tissue. AE improved sepsis-associated multiple organ dysfunction syndrome (MODS), aortic injury, hypoglycemia, and death. AE prominently decreased pulmonary inflammation, pulmonary edema, and modulated redox balance during sepsis. AE prominently decreased neutrophil content in organ. AE prominently downregulated CXCL-1, CXCL-8, IL-6, TNF-α, Glu1, and HMGB1 mRNA expression but activated IL-1RN, IL-10, Sirt-1, and Nrf-2 mRNA expression in the lung during sepsis. AE decreased the serum levels of lactate and HMGB1 but increased blood glucose levels and serum insulin levels during sepsis. A 4-week AE improves sepsis-associated MODS, aortic injury, pathoglycemia, and death. AE impairs LPS-induced lactate and HMGB1 release partly because AE increases serum insulin levels and decreases the levels of Glut1. AE is a novel therapeutic strategy for sepsis targeting aerobic glycolysis.

www.nature.com/scientificreports/ Detection of pulmonary permeability. To quantify the magnitude of pulmonary permeability, Wet dry weight ratio (W/D) was detected. Blotting papers were utilized to absorb liquid and blood on the surface of the lung tissues, and then the wet weight of the lung tissue was determined. The lung tissues were dried in a drying case until a stable dry weight was obtained. W/D = wet weight of the lung tissues/dry weight of the lung tissues.
Detection of the Warburg effect. The blood glucose levels were detected using a blood glucose meter (Johnson Company). Lactate in serum was detected with a colorimetric l-lactate assay kit (Abcam, Cambridge, MA, USA). The serum insulin and HMGB1 levels were detected using Commercial kits (Shino Test Corporation, Tokyo, Japan). Besides, Glu1 and HMGB1 mRNA expression levels in the lung were detected.
qRT-PCR. Total RNA which was extracted from lung tissue was used as a template for cDNA synthesis. We performed qRT-PCR as previously described 32 . GAPDH was served as the housekeeping gene. Supplementary Material 2 shows the murine PCR primer sequence information.
Assessment of survival rates. For survival analyses, 60 male mice (6 weeks old, 20-22 g) were randomly divided into three groups: (1) Con group, the mice received equal volumes of saline via intraperitoneal injection; (2) LPS group, the mice received 12 mg/kg LPS via intraperitoneal injection; (3) Ex + LPS group, the mice were trained for 4 weeks as previously described 30 . After the last training of 48 h, the mice received 12 mg/kg LPS via intraperitoneal injection. The number of dead mice was recorded every 6 h for 48 h. Each group included 20 mice.
Data processing. We performed one-way ANOVA and graphed the figures using GraphPad Prism 9 software. The significance threshold was set to P < 0.05. All data are expressed as the mean ± SD (x ± s).

AE prevented acute lung injury.
Histologic assessment showed evidence of the degree of lung injury. In the Con group ( Fig. 2A) and the Ex group (Fig. 2C), the lung tissues were intact and clear, the cells were neatly arranged, the intercellular substance was free of edema, and there were no symptoms of injury. LPS administration significantly increased lung injury, inflammatory infiltrates, and interstitial edema ( Fig. 2B) compared to Con. AE significantly reduced the degree of lung injury, inflammatory cell infiltration, and interstitial edema (Fig. 2D). AE significantly decreased lung injury score compared to LPS (P < 0.001) (Supplementary Material 1).
AE attenuated neutrophil content in lung tissue. Compared with the Con group ( Fig. 3A) and the Ex group (Fig. 3C), there was a large amount of neutrophil infiltration after LPS administration (Fig. 3B). Compared with the LPS group, a 4-week exercise pretreatment prevented the upregulation of the neutrophil infiltration (Fig. 3D). The number of neutrophils increased significantly after the administration of LPS (P < 0.001). A 4-week exercise pretreatment prevented the upregulation of the number of neutrophils (P < 0.001) in mice with sepsis ( Fig. 3E).
Cell counts in BALF. LPS injection prominently increased the number of macrophages (P < 0.001) and neutrophils (P < 0.001) in BALF. A 4-week exercise pretreatment prevented the increase in the number of macrophages (P < 0.01) and neutrophils (P < 0.01) in mice with sepsis. LPS injection or AE administration did not change the number of lymphocytes and eosinophils in BALF (Table 1). LPS injection prominently increased neutrophil content in the liver (P < 0.001), kidney (P < 0.001), and heart tissues (P < 0.001), while exercise prominently attenuated neutrophil content in the liver (P < 0.01), kidney (P < 0.01), and heart tissues (P < 0.01) during sepsis (Supplementary Material 1).

Con
AE relieved liver injury. The liver lobules of the mice in the Con group (Fig. 4A,E) and the Ex group ( Fig. 4B,F) were intact and clear, the cells were neatly arranged, the intercellular substance was free of edema, the liver stripes were clear and regular, and there were no symptoms of injury. The liver lobules of the mice in the LPS group were severely damaged, the liver cells swelled, the intercellular substance disappeared, and there was a large amount of neutrophil infiltration (Fig. 4C,G). The liver lobules of the mice in the Ex + LPS group had significantly less liver tissue structural damage, with clearer liver lobules and a small amount of neutrophil infiltration (Fig. 4D,H).
LPS administration notably upregulated the markers of liver disease ALT (P < 0.001) and AST (P < 0.001) levels compared to Con. AE notably downregulated the ALT (P < 0.05) and AST (P < 0.05) levels compared to LPS. AE notably reduced the lung injury score compared to LPS (P < 0.001) ( Table 4).
AE relieved kidney injury. The kidney tissues of the mice in the Con group (Fig. 5A,E) and in the Ex group (Fig. 5B,F) were intact and clear, the cells were neatly arranged, the intercellular substance was free of edema, and there were no symptoms of injury. Cortical tubular epithelial cells were well-shaped, and almost every epithelial cell owned intact nuclei. The kidneys of the mice in the LPS group were severely damaged, the cells swelled, and the intercellular substance disappeared, accompanied by a large amount of neutrophil (black arrows) and hemocyte infiltration, severe epithelial vacuolization (yellow arrows), flattening of the tubular epithelium (blue arrows), and the appearance of an atypical shape with almost no nuclei (green arrow) (Fig. 5C,G). The kidneys of the mice in the Ex + LPS group had significantly less kidney tissue structural damage than those of the LPS group, with clearer nephrons and a small amount of inflammatory cell infiltration, less flattening of tubular epithelium, and less vacuolization than those of septic mice, and the degree of damage was significantly reduced (Fig. 5D,H).
Compared with the Con group, LPS administration notably increased the levels of markers of kidney injury Cre (P < 0.001) and BUN (P < 0.001), while Cre (P < 0.05) and BUN (P < 0.05) levels were notably decreased in the Ex + LPS group compared with those in the LPS group. Compared with the LPS group, AE notably reduced the liver injury score (P < 0.001) ( Table 4).  Table 4. Effects of AE on Gre, BUN, ALT and AST levels in serum. The markers of liver damage (ALT and AST) and markers of kidney damage (Cre and BUN) in serum were detected. @@@ P < 0.001, Con versus LPS groups. *P < 0.05, LPS versus Ex + LPS groups. **P < 0.01, LPS verus Ex + LPS groups. www.nature.com/scientificreports/ AE prevented septicemic cardiomyopathy. In the Con group (Fig. 6A) and the Ex group (Fig. 6C), the myocardial tissue was uniformly stained, the myocardial fibers were arranged regularly and the interstitial spaces were normal. In the LPS group, myocardial tissues were disordered, myocardial degeneration occurred, dissolution occurred, and a large number of inflammatory cells infiltrated the muscle space (Fig. 6B). The inflammatory cells in the Ex + LPS group were less infiltrated, and the myocardial fiber tissue structure was normal. The distribution of muscle fibers was improved, but it did not completely return to the normal form (Fig. 6D). AE notably reduced the myocardial injury score (P < 0.001) compared to LPS (Supplementary Material 1).
AE prevented aortic injury. In the Con group (Fig. 7A) and the Ex group (Fig. 7C), the aorta was uniformly stained and arranged regularly, the endothelium was smooth and orderly, and the elastic fibers had a regular wavy-like shape. After LPS administration, the endothelium was not smooth and regular, the elastic fib-  www.nature.com/scientificreports/ ers of the media became sparse, elastic fibers lost a regular wavy-like shape, and LPS administration significantly increased the aortic media thickness and decreased the area ratio of elastic fibers (Fig. 7B). AE notably increased the area ratio of elastic fibers and increased the aorta media thickness during sepsis (Fig. 7D). LPS prominently increased the medial thickness of the aorta (P < 0.001) and prominently decreased the medium membrane elastic fiber area ratio (P < 0.001), which was reserved by a 4-week AE (Supplementary Material 1). AE prevented oxidative stress injury in lung tissue. LPS injection prominently upregulated MDA ( Fig. 8A; P < 0.001) and MPO ( Fig. 8B; P < 0.001) expression but prominently downregulated GSH ( Fig. 8C; P < 0.001) and SOD ( Fig. 8D; P < 0.001) expression compared to Con. AE prominently downregulated MDA (P < 0.001) and MPO (P < 0.001) expression but prominently upregulated SOD (P < 0.05) and GSH (P < 0.05) expression in lung tissue of septic mice.
AE relieved pulmonary edema. Compared with the Con group, the W/D of lung tissues increased significantly after LPS injection (P < 0.001). AE prevented the upregulation of W/D in lung tissues (P < 0.01; Fig. 10B).

Discussion
Previous studies have demonstrated that the Warburg effect was involved in innate and adaptive immunity 21 . However, it was not clear whether AE can regulate the Warburg effect during sepsis. There was convincing evidence that increased serum lactate levels were a biomarker of mortality and MODS during sepsis and lactate clearance was a novel therapeutic strategy for sepsis [23][24][25] . It was well documented that HMGB1 was involved in the development of sepsis and was a promising therapeutic target for sepsis treatment 26,27 . Increased aerobic glycolysis released a lot of lactate which stimulated macrophages to release HMGB1 28 . Our data demonstrated that LPS injection prominently upregulated the levels of lactate and HMGB1 in serum, while AE prominently www.nature.com/scientificreports/ downregulated the levels of lactate and HMGB1 in serum during sepsis. It has been suggested that insulin can decrease serum HMGB1 levels in septic animals 1 , implying a possible role of glucose metabolism in the regulation of HMGB1 release. It's well documented that exercise can improve insulin resistance. We came up with a hypothesis that exercise may serum insulin concentrations during sepsis. Our data demonstrated that exercise prominently increased serum insulin concentrations during sepsis. Based on these results, AE improved sepsis via impairing LPS-induced HMGB1 release. AE impaired HMGB1 release via decreasing lactate and Glut1 expression and increasing insulin expression during sepsis. Our research found a new mechanism through which AE improved sepsis. Increased aerobic glycolysis consumed a large amount of glucose. Strikingly, our data demonstrated that LPS administration markedly decreased blood glucose levels. However, a 4-week exercise pretreatment markedly  www.nature.com/scientificreports/ increased blood glucose levels. Some studies found that LPS administration resulted in hyperglycemia 37,38 . However, other studies found that LPS administration resulted in hypoglycemia [39][40][41] . Our data demonstrated that LPS administration resulted in hypoglycemia, in part because increased aerobic glycolysis consumed a large amount of blood glucose during sepsis. Hypoglycemia was a common complication of sepsis. If sepsis-induced hypoglycemia was not treated promptly, patients can deteriorate further and fall into a coma, which was easily confused with coma caused by infection, leading to delayed diagnosis and treatment or even death. Hence, septicemia patients should be monitored, and the blood glucose should be supplemented in a timely manner. Our data showed that AE prevented the decrease in blood glucose levels. Our data demonstrated that AE can be used as a preventive tool for LPS-induced hypoglycemia.
Our data revealed that a 4-week exercise pretreatment reduced sepsis-associated lung, liver, kidney, and heart injury. Our data were consistent with previous conclusions [15][16][17][18][19] . Strikingly, we found that AE prevented sepsisinduced aortic injury. The aortic injury occurred before the onset of inflammatory infiltration and organ injury. Previous studies have overlooked the effects of sepsis on the aorta. Aorta could regulate arterial blood pressure and hypotension was one of the most frequent complications of sepsis. Sepsis-induced hypotension led to organ dysfunction and septic shock, which were the most severe complications of sepsis and deadly disease 20 . Our data identified that LPS administration increased aortic media thickness and reduced the area ratio of elastic fibers, which was improved by a 4-week AE pretreatment. These results demonstrated that AE could be used as a preventive tool for LPS-induced aortic injury.
Previous researches demonstrated that neutrophils were involved in the development of sepsis 42 . Previous work identified that deleterious accumulation of neutrophils in organs resulted in MODS 43 . Our data showed that LPS injection resulted in severe neutrophil infiltration in lung, liver, kidney, and heart tissues. Strikingly, our data showed that AE reduced neutrophil infiltration in lung, liver, kidney, and heart tissues during sepsis. Our data identified that deleterious activation of neutrophils was a critical reason leading to host tissue injury and organ damage during sepsis and AE improved sepsis-induced MODS in part by decreasing neutrophil content. There was convincing evidence that TNF-α, CXCL-1, and CXCL-8 were chemotactic for neutrophils. Our data demonstrated that AE prominently reduced BALF levels of CXCL-1, CXCL-8, and TNF-α and CXCL-1, CXCL-8, and TNF-α mRNA expression levels in lung tissue. Hence, AE inhibited neutrophil infiltration via suppressing TNF-α, CXCL-1, and CXCL-8 expression.
Our data demonstrated that LPS injection led to an excessive inflammatory response, pulmonary edema, and the infiltration of inflammatory cells, which were the three main features of ALI 6 . A 4-week exercise pretreatment improved the degree of pulmonary edema and neutrophil infiltration. Hence, AE could be used as a preventive tool for sepsis-associated ALI.
Oxidant/antioxidant imbalance was involved in the pathogenesis of sepsis and LPS administration led to oxidative stress injury 5 . We demonstrated the antioxidant effects of exercise during sepsis. Our data identified that AE prominently increased MDA and MPO expression and prominently decreased SOD and GSH expression during sepsis. Based on these results, AE was a preventive tool for sepsis partly because AE increased antioxidant capacity in lung tissue.
Sirt-1 exerted the effects of promoting lung cell proliferation and vitality and exerted immunomodulatory effects and modulated redox balance 44,45 . Activation of Sirt-1 by agents including resveratrol improved sepsis because Sirt-1 reduced inflammatory response and modulated redox balance 46,47 . Sirt-1 was known to protect against sepsis through Sirt-1/Nrf-2 signaling 47 . Previous studies showed that exercise activated Sirt-1 in muscle tissue because exercise increased NAD/NADH ratio 48 . Sirt-1/Nrf-2 signaling was established as a crucial mechanism underlying lung protection, our data identified that AE could activate Sirt-1/Nrf-2 signaling and reduced inflammatory response and modulated redox balance. Sirt-1/Nrf-2 signaling was a novel therapeutic strategy for sepsis. In contrast to traditional therapeutic methods, AE was a comprehensive intervention treatment. AE improved sepsis through multiple mechanisms simultaneously. The protective effects of 4 weeks of exercise pretreatment on LPS-induced changes MODS, aortic injury, neutrophilic inflammation, pulmonary inflammation, and oxidative stress injury were detected. All of these factors cooperated to improve the survival rate of septic mice.