Author response: Hepatic AMPK signaling dynamic activation in response to REDOX balance are sentinel biomarkers of exercise and antioxidant intervention to improve blood glucose control

Abstract

Article Figures and data Abstract Editor's evaluation eLife digest Introduction Results Discussion Materials and methods Data availability References Decision letter Author response Article and author information Metrics Abstract Antioxidant intervention is considered to inhibit reactive oxygen species (ROS) and alleviate hyperglycemia. Paradoxically, moderate exercise can produce ROS to improve diabetes. The exact redox mechanism of these two different approaches remains largely unclear. Here, by comparing exercise and antioxidant intervention on type 2 diabetic rats, we found moderate exercise upregulated compensatory antioxidant capability and reached a higher level of redox balance in the liver. In contrast, antioxidant intervention achieved a low-level redox balance by inhibiting oxidative stress. Both of these two interventions could promote glucose catabolism and inhibit gluconeogenesis through activation of hepatic AMP-activated protein kinase (AMPK) signaling; therefore, ameliorating diabetes. During exercise, different levels of ROS generated by exercise have differential regulations on the activity and expression of hepatic AMPK. Moderate exercise-derived ROS promoted hepatic AMPK glutathionylation activation. However, excessive exercise increased oxidative damage and inhibited the activity and expression of AMPK. Overall, our results illustrate that both exercise and antioxidant intervention improve blood glucose control in diabetes by promoting redox balance, despite different levels of redox state(s). These results indicate that the AMPK signaling activation, combined with oxidative damage markers, could act as sentinel biomarkers, reflecting the threshold of redox balance that is linked to effective glucose control in diabetes. These findings provide theoretical evidence for the precise management of diabetes by antioxidants and exercise. Editor's evaluation Redox signaling is a dynamic and concerted orchestra of interconnected cellular pathways. There is always a debate whether ROS (reactive oxygen species) could be a friend or foe. There are several paradoxical studies (both animal and human) wherein exercise health benefits were reported to be accompanied by increases in ROS generation. Utilizing the in-vitro studies as well as rats models, this manuscript illustrates the different regulatory mechanisms of exercise and antioxidant intervention on redox balance/redox state(s) that are linked to improved glucose control and thereby effective management of diabetes. https://doi.org/10.7554/eLife.79939.sa0 Decision letter Reviews on Sciety eLife's review process eLife digest Molecules known as reactive oxygen species or ROS play vital roles in healthy cells. However, ROS can act as a double-edged sword: if their levels become too high, they can be harmful and interfere with many physiological processes. Indeed, diabetes, high blood pressure and many other chronic diseases are associated with imbalances in the levels of ROS in the body. To counter high ROS levels, cells have antioxidant mechanisms that reduce the excess ROS in the cell and keep the ‘redox’ (from reduction and oxidation) balance of the cell. Exercise and antioxidant nutritional supplements have attracted much attention as drug-free interventions for diabetes. Both strategies alter the levels of ROS in the body, with exercise increasing the levels of ROS, and antioxidant supplements reducing them. Individuals with diabetes and other metabolic health issues have different ROS levels depending on the severity of the disease, age, genetics and other factors, leading to different redox states in their cells. Thus, approaches that can accurately evaluate the redox balance status of individuals are necessary for clinicians to identify what types of exercise and antioxidant supplements are beneficial and which treatments are most appropriate for each patient. Wu, Zhao, Yan, Gao et al. examined the effects of exercise and antioxidant supplements on rats with diabetes, with the aim of identifying molecules – also known as biomarkers – that reflect the bodies’ redox balance. They found that moderate exercise increased the levels of ROS in the liver, which, in turn, compensated by increasing the production of antioxidants to protect against the higher levels of ROS. This resulted in a healthy ‘high-level’ redox balance, in which both ROS and antioxidants levels were high in the rats. On the other hand, giving the rats antioxidant supplements decreased their levels of ROS, leading to a healthy low-level redox balance with low levels of ROS. These findings indicate that regular moderate exercise may be appropriate for people with pre-diabetes symptoms to restore a healthy redox balance. This is because the compensatory antioxidant mechanisms that kick in during exercise may be enough to counteract the excessive levels of ROS in these people. For patients with mild diabetes, exercise, antioxidant supplements, or a combination of both may be appropriate treatment, depending on their levels of ROS. Finally, patients with severe diabetes, who already have high levels of ROS, may benefit from antioxidant supplements to help reduce their excessive levels of ROS. In the future, the biomarkers identified by Wu, Zhao, Yan, Gao et al. may be used to monitor and assess the change in the redox balance status of various populations and guide personalized interventions to maintain health. Additionally, these findings provide a new strategy for precision prevention and treatment of diabetes and other metabolic diseases. Introduction Diabetes mellitus is a chronic metabolic disease, that has emerged as a global public health problem. According to the latest epidemiological data from the International Diabetes Federation, the global diabetes prevalence in 20–79 year-old was estimated to be 10.5% (536.6 million people) in 2021, and is expected to rise to 12.2% (783.2 million) in 2045 (Sun et al., 2022). With the development of genomics, proteomics, and metabolomics, it has been discovered by many studies that type 2 diabetes is associated with irreversible risk factors such as age, genetics, race, and ethnicity and reversible factors such as diet, physical activity, and lifestyle (Heald et al., 2020; Chan et al., 2021). Aerobic metabolism in glucose oxidation, mitochondrial damage, and oxidative stress have been considered to play a critical role in the occurrence and development of diabetes (Iacobini et al., 2021). Exercise and antioxidant supplements are often suggested as essential therapeutic strategies in the early stages of type 2 diabetes (Kirwan et al., 2017; Bhatti et al., 2022), with different mechanisms. It has been reported that chronic exercise training can alleviate oxidative stress and diabetic symptoms by improving cellular mitochondrial function and biogenesis in the diabetic state (Stanford and Goodyear, 2014). Contradictorily, exercise also increases reactive oxygen species (ROS) production, while prolonged or high-intensity exercise could result in mitochondrial functional impairment to aggravate complications of diabetes (Flockhart et al., 2021). Since 1970s, studies have demonstrated that 1 hr of moderate endurance exercise can increase lipid peroxidation in humans (Brady et al., 1979; Dillard et al., 1978). In 1998, Ashton directly detected increasing free radical levels in exercising humans using electron paramagnetic resonance spectroscopy (EPR) and spin capture (Ashton et al., 1998). These results led to a great deal of interest in the role of ROS in physical exercise (Powers and Jackson, 2008; Powers et al., 2011; Traverse et al., 2006). Regarding the contradiction of exercise on ROS scavenging or production, James D Watson also hypothesized that type 2 diabetes is accelerated by insufficient oxidative stress rather than oxidative stress (Watson, 2014), based on the effect of exercise on diabetes management. Although Watson’s opinions supported that exercise could treat diabetes by producing ROS, whether exercise-induced ROS production is beneficial or detrimental to diabetes is still being debated. The specific regulation of ROS produced by exercise on diabetic blood glucose in vivo is unclear. In contrast, the general view of the antioxidant treatment for diabetes is that antioxidants reduce cytotoxic ROS and oxidative products, thus alleviating diabetes and achieving glycemic control (Rahimi et al., 2005). Our previous study also found that hepatic mitochondrial ROS scavengers and antioxidant substances inhibited the oxidative products such as Malondialdehyde (MDA) and 4-HNE in diabetic animals and favored glycemic control (Wu et al., 2019; Wu et al., 2021). Exercise-induced oxidation and antioxidant administration, as two opposite approaches, could achieve the regulation of diabetes, respectively. However, the differences in redox mechanisms between these two approaches to diabetes treatment have not been fully understood. It is well established that the increase of skeletal muscle glucose uptake during exercise is crucial in glycemic control (Holloszy, 2005; Holloszy et al., 1986; Greiwe et al., 1999). Considering that the liver is another vital organ for maintaining blood glucose homeostasis, including storing, utilizing, and producing glucose, exercise-induced hepatic redox metabolism is also significant. The activation of hepatic AMP-activated protein kinase (AMPK), which acts as a ‘metabolic master switch’, alleviates diabetes symptoms by reducing glycogen synthesis, increasing glycolysis, and promoting glucose absorption in surrounding tissues (Viollet et al., 2006). Therefore, the activation of AMPK in the liver is significant for regulating glucose and lipid metabolism in the blood. Zmijewski et al. found that AMPK could be activated by hydrogen peroxide stimulation through direct oxidative modification (Zmijewski et al., 2010). In contrast, other studies suggested that oxidative stress could disrupt the activation of the AMPK signaling pathway (Ren et al., 2021; Hawley et al., 2010). Our previous study explored the mechanism by which redox status contributes to hepatic AMPK dynamic activation. Under a low ROS microenvironment, GRXs mediated S-glutathione modification activates AMPK to improve glucose utilization. In contrast, under an excessive ROS microenvironment, sustained high level ROS might cause loss of AMPK protein (Dong et al., 2016b). These studies indicate that oxidative modification can directly regulate AMPK activity in liver cells, thus activating downstream signaling pathways to regulate glucose and lipid metabolism. However, it is unclear why two seemingly contradictory phenomenon of antioxidant intervention and exercise-induce ROS can promote AMPK activation. Moderate exercise has been proved significantly elevate systemic ROS. At the same time, endogenous antioxidant defences also increased to counteract increased levels of ROS induced by exercise (Parker et al., 2014). Thus, we hypothesized that both antioxidants and exercise could reach either high- or low-level redox balance in diabetic individuals. Moreover, the activity and expression of AMPK might be a biomarker of redox balance in vivo. Hence, the present study was designed to understand the different mechanisms of exercise and antioxidant intervention in diabetes and verify the activation of hepatic AMPK as a hallmark of dynamic redox balance. First, we utilized the streptozotocin-high fat diet (STZ-HFD) induced type 2 diabetic model (T2DM) in rats to clarify the hepatic redox status in T2DM rats after the exercise or antioxidant intervention. Then, according to the exercise intensity and mode, we divided the exercise groups into three modes and found that AMPK activation could serve as a sentinel biomarker of redox balance and moderate exercise in diabetic management. In this study, we found that AMPK activation and its downstream pathways could reflect the threshold of exercise or antioxidant administration for diabetes management (Figure 1). This study provides clues for the personalized management of diabetes by antioxidants and exercise. Figure 1 Download asset Open asset A model of the different mechanisms of exercise and antioxidant intervention in diabetes. A graphical abstract of this study. Moderate exercise upregulated compensatory antioxidant capability and reached a high-level redox balance, whereas antioxidant intervention achieved a low-level redox balance by inhibiting oxidative stress for treating diabetes. ROS: reactive oxygen species; AMPK: AMP-activated protein kinase. Results Exercise promotes antioxidant levels through producing ROS, leading to a high level of REDOX balance in the liver To investigate the hepatic redox regulation in diabetes after exercise intervention, we established the T2DM rat model by feeding HFD followed by a low dose of STZ injection (35 mg/kg). The exercise intervention was started from day-0 to day-28 (Figure 2A). According to previous studies, the initial speed of exercise was 15 m/min, and the speed was increased by 3 m/min every 5 min. After the speed reached 20 m/min, the speed was maintained for another 60 min with slope of 5%. The exercise intensity was 64–76% VO2max (Qin et al., 2020). The low-intensity continuous exercise (CE) can be regarded as aerobic exercise. Figure 2 Download asset Open asset Moderate exercise induced reactive oxygen species (ROS) production in exercise group and increased the antioxidant status. (A) Experimental design. Type 2 diabetic model (T2DM) rats model was fed by high-fat diet plus a low dose of streptozotocin (STZ) injection (35 mg/kg). The high-fat diet (HFD, 60% calories from fat) was started from the 1st week to the 8th week. The exercise intervention was started from 1st week to 4th week. (B–D). Representative protein level and quantitative analysis of NADPH oxidase 4 (NOX4) (67 kDa), cyclooxygenase 2 (COX2) (17 kDa) and Actin (45 kDa) in the rats in the control (Ctl), T2D, and T2D+continuous exercise (CE) groups.(E–G). Representative protein level and quantitative analysis of Nrf2(97 kDa), Sestrin2 (56 kDa) and Actin (45 kDa) in the rats in the Ctl, T2D, and T2D+CE groups.(H–K). Representative protein level and quantitative analysis of PRX1 (27 kDa), Grx1 (17 kDa), Trx1 (12 kDa), and Actin (45 kDa) in the rats in the Ctl, T2D and T2D+CE groups. The rat livers were homogenized by 1% SDS and analyzed by Western blots with the appropriate antibodies. (L–M). Representative protein level and quantitative analysis of 3-NT and Actin (45 kDa) in the rat in the Ctl, T2D and T2D+CE groups. (N–O). Liver protein carbonylation (N) and MDA content (O) level was detected in the rats of Ctl, T2D, T2D+CE groups. (ns: not significant; *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 compared with all groups by one-way ANOVA and Tukey’s post hoc test; data are expressed as the mean ± SEM; n=4–8 per group). Figure 2—source data 1 Full western blot images. https://cdn.elifesciences.org/articles/79939/elife-79939-fig2-data1-v2.pdf Download elife-79939-fig2-data1-v2.pdf Figure 2—source data 2 Normalized grey value of western blot data. https://cdn.elifesciences.org/articles/79939/elife-79939-fig2-data2-v2.xlsx Download elife-79939-fig2-data2-v2.xlsx The ROS-generating nicotinamide adenine dinucleotide phosphate oxidases (NOXs) have been recognized as one of the main sources of ROS production in cells (Panday et al., 2015). Cyclooxygenase 2 (COX2) activity could also act as a stimulus for ROS production (Burdon et al., 2007). The expressions of NADPH oxidase 4 (NOX4) and COX2 in the liver were increased in the diabetic group. After exercise intervention, NOX4 and COX2 level were further up-regulated compared with the diabetic group (Figure 2B–D). These results indicate the exercise intervention up-regulated ROS production. Next, we detected the expression of antioxidant enzymes in liver tissue. Nuclear factor erythroid 2–related factor 2 (Nrf2) is the central regulator of the threshold mechanisms of oxidative stress and ROS generation (McMahon et al., 2003). With the increase of ROS level in the development of diabetes, Nrf2 was activated to induce the transcription of several antioxidant enzymes (Bitar and Al-Mulla, 2011; Jiang et al., 2010). We found an increase in Nrf2 expression in diabetic rats (Figure 2E–F). After CE intervention, the level of Nrf2 levels further increased, indicating that exercise intervention could activate antioxidant system (Figure 2E–F). Under stress conditions, Nrf2 translocates to the nucleus and binds to antioxidant response elements (AREs), which results in the expression of diverse antioxidant and metabolic genes, such as thioredoxin (Trx), to relieve oxidative damage (Krajka-Kuźniak et al., 2017). Thioredoxin-1 (Trx-1), a type of cytosolic isoform of Trx, has been widely accepted to regulate glutathione metabolism with GRX and PRX. After CE intervention, we found the protein expression of GRX1 and TRX1 were up-regulated (Figure 2H J–K). Notably, the PRX expression also showed a trend of increase (Figure 2). Sestrin2 is a cysteine sulfinyl reductase that plays crucial roles in regulation of antioxidant actions (Budanov et al., 2010). As an endogenous antioxidant, the hepatic sestrin2 level also showed a significant increase in the CE group (Figure 2E G). Since excess ROS can cause the increase of oxidative damage (Murphy et al., 2022), we further detected the protein damage and lipid peroxidation to determine the redox status. 3-Nitrotyrosine (3-NT) and protein carbonylation are biomarkers of reactive nitrogen species (RNS) and ROS modified proteins (Ahsan, 2013; Wong et al., 2010). We found that the CE intervention reduced the 3-NT level and did not further decrease the protein carbonylation level (Figure 2L–N). MDA, a biomarker of lipid peroxidation, was also significantly up-regulated in the diabetic group but decreased in exercise group (Figure 2O). These results indicate that the high ROS production in the CE group could, instead, increase the antioxidant status to avoid oxidative damage. It suggests that CE can promote redox to reach a high level of balance. Therefore, even if exercise increases the ROS-generating enzymes NOX4 and COX2, the increase in ROS production does not lead to oxidative damage. Antioxidant intervention alleviates blood glucose through reducing oxidative stress, leading to a low level of REDOX balance in the liver Recent studies have suggested that NADPH oxidase is one of the primary sources of ROS (Panday et al., 2015; López-Acosta et al., 2018). Apocynin has already been characterized as an NADPH oxidase inhibitor in the early 1980s, and it can also act as an antioxidant (Heumüller et al., 2008). Our previous study showed that apocynin intervention alleviated blood glucose by inhibiting oxidative products. In this study, the antioxidant supplement was also started from day-0 to day-28 in this study (Figure 3A). We found that apocynin supplement decreased the protein carbonylation level and MDA level in the liver (Figure 3B–C). Also, the total antioxidant capacity (TAOC) level was increased after apocynin treatment, indicating the decrease of oxidation level (Figure 3D). Moreover, as endogenous antioxidant, the Sestrin2 and Nrf2 expression decreased after apocynin intervention (Figure 3E–G). These results indicate that the antioxidant intervention reduced the ROS in diabetic hepatocytes, thereby decreasing the ROS-induced compensatory upregulation of Sestrin2 and Nrf2. Figure 3 Download asset Open asset Antioxidant intervention alleviates blood glucose through promoting the upregulation of reducing levels. (A) Experimental design. Type 2 diabetic model (T2DM) rats model was fed by high-fat diet plus a low dose of streptozotocin (STZ) injection (35 mg/kg). The apocynin intervention was started from 1st week to 4th week. (B–D) Liver protein carbonylation (B), MDA content (C) and TAOC (D) level were detected in the rats of control (Ctl), T2D and T2D+APO groups. (E–H) Representative protein level and quantitative analysis of Nrf2 (97 kDa), Sestrin2 (57 kDa), Glut2 (60–70 kDa), and HSP90 (90 kDa) in the rats in the Ctl, T2D and T2D+APO groups. (I) Postprandial blood glucose levels of Ctl, T2D, T2D+continuous exercise (CE) and T2D+APO groups at the end of 8th week. (J) Fasting blood glucose levels of Ctl, T2D, T2D+CE and T2D+APO groups at the end of 8th week. (K) Blood glucose level after oral glucose administration (0 min, 60 min, and 120 min) in Ctl, T2D, T2D+CE and T2D+APO groups at the end of 8th week (*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 compared with all groups by one-way ANOVA and Tukey’s post hoc test; data are expressed as the mean ± SEM; n=4–8 per group). Figure 3—source data 1 Full western blot images. https://cdn.elifesciences.org/articles/79939/elife-79939-fig3-data1-v2.pdf Download elife-79939-fig3-data1-v2.pdf Figure 3—source data 2 Normalized protein carbonylation, MDA content, TAOC, grey value of western blot data and blood glucose level. https://cdn.elifesciences.org/articles/79939/elife-79939-fig3-data2-v2.xlsx Download elife-79939-fig3-data2-v2.xlsx Consistently, Glut2, a glucose sensor in the liver, was increased in diabetic liver and decreased after the apocynin supplement (Figure 3E, H). The postprandial blood glucose, fasting blood glucose, and oral glucose tolerance test (2 hr after oral glucose, (oral glucose tolerance test)OGTT) were decreased in the apocynin intervention group compared with the diabetic rat group (Figure 3K). Consistent with the apocynin intervention group, the exercise group also showed lower postprandial blood glucose and fasting blood glucose levels and OGTT (Figure 3K). These studies indicate that the apocynin treatment improved the diabetes through inhibiting ROS level and protein oxidative damage to achieve a low-level redox balance. Moderate exercise-generated ROS promotes activation of AMPK by phosphorylation and reduces blood glucose level, while excessive exercise-generated oxidative stress reduces AMPK expression and exacerbates diabetes. In order to find out the biomarkers that could reflect moderate exercise to improve blood glucose control, diabetic rats were divided into short-term CE, intermittent exercise (IE), and excessive exercise (EE) according to the exercise intensity and mode (Qin et al., 2020). We found that the random blood glucose and 2 hr OGTT in CE and IE treated diabetic rats were decreased (Figure 4A–B). In contrast, EE intervention did not improve blood glucose but slightly increased random and 2 hr OGTT (Figure 4A–B). Figure 4 Download asset Open asset Moderate exercise-generated reactive oxygen species (ROS) promotes activation of AMP-activated protein kinase (AMPK) by phosphorylation and reduces blood glucose level, while excessive exercise- generated oxidative stress reduces AMPK expression and exacerbates diabetes. (A) Postprandial blood glucose levels of control (Ctl), type 2 diabetic (T2D), T2D+continuous exercise (CE), T2D+intermittent exercise (IE) and T2D+excessive exercise (EE) groups at the end of 8th week. (B) Blood glucose level after oral glucose administration in Ctl, T2D, T2D+CE, T2D+IE and T2D+EE groups at the end of 8th week. (C–E) Representative protein level and quantitative analysis of NADPH oxidase 4 (NOX4) (67 kDa), cyclooxygenase 2 (COX2) (17 kDa) and HSP90 (90 kDa) in the rats in the Ctl, T2D, T2D+CE, T2D+IE, and T2D+EE groups. (F–I) Representative protein level and quantitative analysis of Ace-SOD2 (27 kDa), SOD2 (17 kDa), Grx1 (17 kDa), Trx1 (12 kDa) and Actin (45 kDa) in the rats in the Ctl, T2D, T2D+CE, T2D+IE and T2D+EE groups. (J–L) Liver protein carbonylation content (J), liver MDA content (K) and AMP/ATP ratio (L) were detected in the rats of Ctl, T2D, T2D+CE, T2D+IE and T2D+EE groups. (M–N). Representative protein level and quantitative analysis of P-AMPK (67 kDa), AMPK (67 kDa) and Actin (45 kDa) in the rats in the Ctl, T2D, T2D+CE, T2D+IE and T2D+EE groups. (ns: not significant; *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 compared with all groups by one-way ANOVA and Tukey’s post hoc test; data are expressed as the mean ± SEM; n=4–8 per group). Figure 4—source data 1 Full western blot images. https://cdn.elifesciences.org/articles/79939/elife-79939-fig4-data1-v2.pdf Download elife-79939-fig4-data1-v2.pdf Figure 4—source data 2 Normalized grey value of western blot data. https://cdn.elifesciences.org/articles/79939/elife-79939-fig4-data2-v2.xlsx Download elife-79939-fig4-data2-v2.xlsx We detected the increase of ROS production-related enzymes, such as NOX4 and COX2 in the EE group, indicating the highest oxidation level (Figure 4C–E). Next, we detected the expression of antioxidant enzymes and oxidative damage in the liver tissue of exercise-treated type 2 diabetic (T2D) rats. As shown in Figure 4F–G, IE intervention increased the activity of MnSOD as shown by decreased level of acetylation compared with the diabetic rats. The expression of GRX and TRX were up-regulated after CE intervention (Figure 4F H–I). Furthermore, we detected the oxidative damage in these three modes. The results showed that the CE and IE group did not obviously change the protein carbonylation level. However, the EE intervention promoted the protein carbonylation in the liver, indicating this mode of action is not due to free radical scavenging but oxidative damage (Figure 4J). In addition to the protein damage, hepatic MDA concentration showed significant up-regulation in the diabetic group but decreased in CE and IE group (Figure 4K), while increased MDA in the EE group indicates oxidative damage. Among these three exercise modes, the IE group showed the lowest level of oxidation (the minor increase in NOX4 and a slight decrease in carbonylation). Although the levels of antioxidant enzymes such as GRX and TRX did not increase, the activity of MnSOD also increased significantly (Figure 4F-G). The reduction of MDA level also indicates IE group did not form oxidative damage (Figure 4K), indicating the IE group could also maintain a relatively high level of redox balance. Nevertheless, the decrease of antioxidant enzymes and increase of oxidative damage in the EE group indicates that the REDOX balance was disrupted. Notably, the phosphorylation of AMPK showed different patterns in three kinds of exercise, among which both CE and IE intervention could promote the phosphorylation of AMPK compared to the diabetic rats (Figure 4M–N). EE intervention did not increase the content of AMPK phosphorylation, which might be caused by the reduction of AMPK level. Meanwhile, the ratio of AMP to ATP was detected, and exercise-activated AMPK did not exhibit AMP-dependent characteristics at this time (Figure 4L). These results suggest that moderate exercise-generated ROS may directly promote AMPK activation by phosphorylation without AMP upregulation and reduce blood and liver glucose levels. However, excessive exercise-generated oxidative stress reduces AMPK expression and exacerbates diabetes. Moderate exercise promoted glycolysis and mitochondrial tricarboxylic acid cycle and inhibited the gluconeogenesis in the liver of diabetic rats Next, we further explored the mechanism by which inhibiting blood glucose during CE and IE intervention. Fructose-2,6-diphosphate (F-2,6-P2; also known as F-2,6-BP), which is a product of the bifunctional enzyme 6-phosphofructose 2-kinase/fructose 2,6-diphosphatase 2 (PFK/FBPase 2, also known as PFKFB2), is a potent regulator of glycolytic and gluconeogenic flux. The phospho-PFKFB2 to PFKFB2 ratio represents the glycolytic rate. A high ratio of phospho-PFKFB2:PFKFB2 leads to an increase in the F-2,6-P2 level and the allosteric activation of phosphor-fructose kinase 1 (PFK1), while a low ratio leads to a decrease in F-2,6-P2 and an increase in gluconeogenesis (Okar et al., 2001). The overexpression of bifunctional enzymes in mouse liver can reduce blood glucose levels by inhibiting hepatic glucose production (Wu et al., 2001). Therefore, bifunctional enzymes are also a potential target for reducing hepatic glucose production. In our study, the p-PFK2:PFK2 ratio decreased in the diabetic rats but was enhanced by CE and IE intervention (Figure 5A–C), suggesting that CE and IE could reverse gluconeogenesis to glycolysis by enhancing PFK/FBPase. Meanwhile, the substrates of the glycolytic pathway (such as DHAP, Figure 5D) and the tricarboxylic acid cycle (such as citrate, succinate, and malate, Figure 5D) showed an upward trend. Figure 5 Download asset Open asset Moderate exercise promoted glycolysis and mitochondrial tricarboxylic acid cycle and inhibited the gluconeogenesis in the liver of diabetic rats. (A–B) Representative protein level and quantitative analysis of P-PFK2 (64 kDa), PFK2 (64 kDa) and GAPDH (37 kDa) in the rats in the control (Ctl), T2D, T2D+continuous exercise (CE) and T2D+intermitten exercise (IE) groups. (C). Liver glucose level after oral glucose administration in Ctl, T2D, T2D+CE, and T2D+IE groups at the end of 8th week. (D) Relative concentrations of substrates for glycolysis (DHAP and Lactate) and the tricarboxylic acid cycle (citrate, succinate and malate) in the rats of Ctl, T2D, T2D+CE, and T2D+IE groups. The concentration of substrates was analyzed by LC-MS/MS. (E–G) Representative protein level and quantitative analysis of FoxO1 (82 kDa), GLUT2 (60–70 kDa) and Actin (45 kDa) in the rats in the Ctl, T2D, T2D+CE, and T2D+IE groups. (H–I) Expression of hepatic Pepck and G6C mRNA in the Ctl, T2D, T2D+CE, and T2D+IE groups were evaluated by real-time PCR analysis. Values represent mean ratios of Pepck and G6pase transcripts normalized to GAPDH transcript levels.