Acute Exercise Improves Insulin Clearance and Increases the Expression of Insulin-Degrading Enzyme in the Liver and Skeletal Muscle of Swiss Mice

The effects of exercise on insulin clearance and IDE expression are not yet fully elucidated. Here, we have explored the effect of acute exercise on insulin clearance and IDE expression in lean mice. Male Swiss mice were subjected to a single bout of exercise on a speed/angle controlled treadmill for 3-h at approximately 60–70% of maximum oxygen consumption. As expected, acute exercise reduced glycemia and insulinemia, and increased insulin tolerance. The activity of AMPK-ACC, but not of IR-Akt, pathway was increased in the liver and skeletal muscle of trained mice. In an apparent contrast to the reduced insulinemia, glucose-stimulated insulin secretion was increased in isolated islets of these mice. However, insulin clearance was increased after acute exercise and was accompanied by increased expression of the insulin-degrading enzyme (IDE), in the liver and skeletal muscle. Finally, C2C12, but not HEPG2 cells, incubated at different concentrations of 5-aminoimidazole-4-carboxamide-1-β-d-ribofuranoside (AICAR) for 3-h, showed increased expression of IDE. In conclusion, acute exercise increases insulin clearance, probably due to an augmentation of IDE expression in the liver and skeletal muscle. The elevated IDE expression, in the skeletal muscle, seems to be mediated by activation of AMPK-ACC pathway, in response to exercise. We believe that the increase in the IDE expression, comprise a safety measure to maintain glycemia at or close to physiological levels, turning physical exercise more effective and safe.


Introduction
Insulin action depends on three major physiological processes: insulin sensitivity [1], insulin secretion [2], and insulin clearance [3], and each one of these processes may be influenced by several pathophysiological conditions, such as obesity and diabetes.
Insulin clearance occurs mainly in the liver due to insulin degradation mediated, primarily, by insulin-degrading enzyme (IDE) [12]. In humans [13][14][15] and animal models [16][17][18], obesity reduces insulin clearance, probably due to lower IDE expression and activity in the liver [17,18]. However, some controversies still remain because higher IDE expression and activity were reported in the liver of obese mice [19,20]. Despite these discrepancies, several studies have demonstrated that impairment on IDE expression and/or activity is closely related to the onset and development of type 2 diabetes [21][22][23][24][25][26].
Physical exercise, recommended to obese and diabetic patients, has been show to increase insulin clearance [27,28] and IDE expression [18,29] in these patients. These effects contribute to reduce the hyperinsulinemia, often associated with obesity, insulin resistance, and diabetes [15,30]. Therefore, increased insulin clearance and IDE expression could be another beneficial effect of exercise on the treatment and/or prevention of diseases related to insulin resistance.
In normoinsulinemic lean humans [27,31,32] and rats [33], physical exercise also increases insulin clearance, but none of these studies have explored the IDE expression. Here, we found that acute exercise increase insulin clearance probably due to an augmented IDE expression in the liver and skeletal muscle. We also demonstrated that activation of AMP-activated protein kinase (AMPK) might be the mechanism whereby exercise increases IDE expression, in the skeletal muscle, but not in the liver. We hypothesize that the increase of the IDE expression and insulin clearance could be a safety measure to maintain glycemia at or close to physiological levels, turning physical exercise more effective and safe.

Materials and Methods Animals
The 4-week-old male Swiss mice, acquired from the State University of Campinas Facilities, were maintained on a 12 h light-dark cycle at 20-21°C with controlled humidity during the entire experiment. The mice were allowed to feed and drink tap water ad libitum, for 8 weeks. All of the experiments were approved by the State University of Campinas Ethics Committee (approval number 1984-1), prior to starting. of exercise (EXE) or non-exercise (CTL), we measured the VO 2 of both groups ( Fig 1E). All of the experiments described below were performed immediately after the 3-h exercise.

Intraperitoneal insulin tolerance test (ipITT)
Non-fasted mice received an intraperitoneal bolus of insulin (1 U kg -1 ). The blood glucose was measured using test strips (Accu-Chek Performa II) at baseline (0 min, before receiving insulin) and at 5, 10, 15, 20, 25 and 30 min after the administration of the insulin bolus. Glucose values were converted to natural logarithmic (Ln). The slope was calculated using linear regression (time × Ln [glucose]) and multiplied by 100 to obtain the glucose decay rate constant during the insulin tolerance test (k ITT , % min -1 ).

Insulin clearance
The plasma insulin concentrations were also evaluated during the insulin tolerance test, and the insulin clearance was calculated as previously described [35]. The rate constant for insulin loss (insulin decay) was assessed by converting the insulin values to natural logarithmic (Ln); the slope was calculated using linear regression (time × Ln [insulin]); and the results were multiplied by 100 to obtain the insulin decay rate constant (% min -1 ).

Tissue samples
The mice were killed in a CO 2 -saturated atmosphere, immediately followed by decapitation. We extracted liver and muscle samples from the mice 5 min after an intraperitoneal bolus of 1 U kg -1 insulin (Humulin1R, Eli Lilly, São Paulo, Brazil) or saline solution (0.9% NaCl wt/vol). The liver and muscle samples were snap frozen in liquid nitrogen and stored for subsequent protein. After, these tissues were homogenized using a lysate buffer (10 mmol L -1 EDTA, 100 mmol L -1 Tris base, 100 mmol L -1 sodium pyrophosphate, 100 mmol L -1 sodium fluoride, 10 mmol L -1 sodium orthovanadate, 2 mmol L -1 PMSF, 1% Triton X-100 and 1 μg mL -1 aprotinin). Pancreatic islets were isolated from the mice immediately after the exercise, as previously described [36].

Glucose-stimulated insulin secretion (GSIS)
Batches of 10 islets were pre-incubated for 1 h in Krebs-Henseleit buffer solution (KHBS) containing 0.5 g l -1 of BSA and 5.6 mmol l -1 glucose and equilibrated at 95% O 2 and 5% CO 2 at 37°C. The medium was discarded, and the islets were incubated for an additional hour in 1 ml of KHBS containing 2.8 mmol l -1 or 16.7 mmol l -1 of glucose. Subsequently, the supernatant fraction was collected to evaluate insulin secretion, and the remaining islets were homogenized in an alcohol and acid solution to measure the total insulin content by radioimmunoassay [37]. supplemented with 10% vol./vol. FBS and 1% vol./vol. penicillin-streptomycin, under a humidified condition with 5% CO 2 at 37°C. After obtain total confluence, the cells were differentiated using DMEM high glucose containing 2% vol./vol. horse serum for 5 days. After that, HEPG2 and differentiated C2C12 cells were incubate at 250, 500 and 750 μmol l -1 of 5-aminoimidazole-4-carboxamide-1-β-d-ribofuranoside (AICAR) (TOCRIS Bioscience, Bristol, England, UK), for 3-h. After, the cells were collected in trypsin/EDTA, washed with phosphate-buffered saline (PBS), and homogenized in urea anti-protease/anti-phosphatase buffer for subsequent western blot analysis.

Western blotting
Western blotting for protein expression and phosphorylation were performed as previously described [38]. The primary antibodies used for Western blotting were as follows: anti-phospho-AMPK and anti-phospho-ACC (Cell Signaling Technology, Boston, MA, USA); anti-IDE, anti-GAPDH, anti-phospho-IR and anti-phospho-Akt (Santa Cruz Biotechnology, Dallas, TX, EUA).

Statistical analyses
Point-to-point and groups of mice were compared by Student's t-test. Groups from the in vitro experiments (HEPG2 and C2C12 cell culture) were compared by one-way ANOVA using the unpaired Tukey's post hoc test (GraphPad Prism 5, La Jolla, CA, USA). Data are presented as the mean ± standard error means (SEM). A value of p<0.05 was considered to be statistically significant.

Exercise altered metabolic parameters
There was no change in body weight (Fig 1A), liver (Fig 1B), perigonadal fat (Fig 1C), or gastrocnemius weight ( Fig 1D) after a single bout of exercise; however, the VO 2 increased during the experiment (Fig 1E).

Exercise altered glycemia and insulinemia
After a single 3 h bout of exercise, the mice exhibited reduced glycemia (Fig 2A) and insulinemia (Fig 2B), compared with the control mice.

Exercise increased insulin tolerance
To explain the apparent paradox of lower insulinemia and glycemia in exercised mice, we evaluated insulin response using the ipITT. After exercise, the mice were more responsive to insulin, as demonstrated by a lower ipITT value ( Fig 3A) and a higher k ITT value ( Fig 3B). As shown by the area under the curve (AUC) of glucose values (Fig 3C), the increased insulin tolerance resulted in a lower glycemia during the test.

Exercise increased the phosphorylation of AMPK-ACC but not IR-Akt
The phosphorylation levels of IR ( Fig 4A) and Akt (Fig 4B) were not changed, whereas phosphorylated AMPK (Fig 4C) and phosphorylated ACC (Fig 4D) were increased in the liver, soleus and gastrocnemius of mice, after exercise.

Exercise increased GSIS from pancreatic islets
Since the increased insulin sensitivity observed does not explain the lower concentration of insulin, in the plasma of exercised mice, we analyzed the GSIS of ex vivo pancreatic islets, isolated from the control and exercised mice. Exercise increased insulin secretion at sub-and supra-stimulatory glucose conditions (Fig 5A), and this insulin increase was accompanied by an increase in beta-cell function, as demonstrated by the higher GSIS (Fig 5B). Furthermore, exercise did not alter the total insulin content (Fig 5C).

Exercise increased insulin clearance and decay
Because the increased GSIS does not explain the reduction in insulinemia, we evaluated insulin clearance in the exercised mice. We found that exercise increased insulin clearance (Fig 6A) and the insulin decay rate after insulin administration (Fig 6B), resulting in lower insulin levels during ITT, as judged by the AUC of insulin (Fig 6C). These results explain the reduced insulinemia in exercised mice.

Exercise increased IDE expression in liver and skeletal muscle
We also investigated the mechanism by which insulin clearance was reduced in response to exercise. IDE protein levels were increased in the liver, soleus and gastrocnemius tissues (Fig  7), which indicates that insulin clearance in mice is likely to be increased as a result of increased IDE expression.

AMPK activation increases IDE expression in C2C12, but not in HEPG2 cells
Finally, we evaluated whether activation of AMPK-ACC pathway could increase the IDE expression. In HEPG2 cells, 3-h of treatment with different concentrations of AICAR did not change the IDE expression ( Fig 8A). However, in C2C12 cells, 3-h incubation at 250 and 500 μmol l -1 AICAR, significantly increased IDE expression (Fig 8B). These data indicate that increased IDE expression in the skeletal muscle, in response to exercise, could be mediated by activation of AMPK-ACC pathway.

Discussion
The effects of physical exercise on insulin clearance and IDE expression, in healthy lean subjects, are not yet fully elucidated. In the literature there is evidence either in favor to an increase [27,[31][32][33], or to a decrease [39] in insulin clearance, induced by exercise. Here, we found that a single bout of exercise reduces insulinemia in lean mice, mainly by increase insulin clearance, probably due to an augmentation of the IDE expression in the liver and skeletal muscle. We suppose that this augmentation in the IDE expression is mediated by activation of AMP-K-ACC pathway in the skeletal muscle, but not in the liver. We believe that the increase in insulin clearance and IDE expression contribute to avoid a hazardous decrease in glycemia during the physical exercise, turning it safer and effective.
It is known that exercise reduces plasma insulin concentration [40,41] and this effect is explained, at least in part, by decrease in insulin secretion [42,43]. However, no alteration [44] or even increase in the insulin release, after exercise, were observed [45][46][47]. Corroborating these last studies, we found an increased GSIS in isolated pancreatic islet from exercised mice (Fig 5). We speculate that this increase in secretion is due to a cross-talk between the skeletal muscle and pancreas, mediated by myokines [48], particularly interleukin-6 (IL-6) [47,49]. Irrespective to the mechanism, increased insulin secretion does not explain the lower insulinemia found in our exercised mice. The insulin concentration in the plasma depends on the balance between insulin secretion and clearance. Therefore, the increased insulin clearance observed (Fig 6) could explain why the insulinemia is reduced in mice after a single bout of exercise, despite an increased GSIS. In agreement, some studies also demonstrated that acute [46] and chronic [31,33] exercise induce an increase of the insulin clearance. However none of these studies measured the IDE expression.
Insulin clearance is primarily dependent on degradation of insulin, which is mediated mainly by IDE in the liver [3]. In lean trained Swiss mice, rested for 24 hours, the reduction in insulin clearance was attributed to a lower IDE expression in the liver, despite a higher IDE expression in the skeletal muscle [39]. In this line, and corroborating previous work [29], we observed here that the expression of IDE was increased by two-fold in the liver of exercised mice, explaining, at least in part, the augmented insulin clearance. Interestingly, IDE was also increased in the soleus (by eight-fold) and gastrocnemius (by two-fold) skeletal muscles of exercised mice. These results were intriguing, considering that the skeletal muscle is not the primary organ responsible for insulin degradation. However, during exercise, blood flow is significantly increased in the skeletal muscle. Thus, this tissue could play an important role in the insulin removal and degradation during physical activity. In fact, in diet-induced obese mice, an expressive increase in IDE activity was observed only in acute exercised muscle [18].
In diet-induced obese mice, which display hyperinsulinemia, acute exercise also increases insulin clearance and IDE expression, normalizing the insulinemia, and this could be an important beneficial effect of acute exercise, in diseases related to insulin resistance. By contrast, lean mice do not display hyperinsulinemia, and we believe that the increase in IDE expression and insulin clearance, in these mice, are important for the effectiveness and safety of physical exercise. During and immediately after intense exercise, a substantial insulin-independent increase in glucose uptake by skeletal muscle is observed [50][51][52]. This increased uptake, associated with increased insulin secretion by the islets, would provoke a fast and hazardous decrease in glycemia. However, a significant increase in the degradation of insulin, by the active skeletal muscle, could prevent the depletion of the glucose stock from the plasma.
Although the effect of exercise on insulin clearance in lean subjects needs more investigations, it is well known that exercise improves insulin sensitivity [53] and reduces glycemia [54,55]. Here, despite an increased insulin tolerance (Fig 3), acute exercise per se did not activate the canonical insulin signaling pathway (IR-Akt) (Fig 4), in agreement with previous data [56,57]. In fact, acute exercise increased the activity of the AMPK-ACC pathway in the skeletal muscle (Fig 4), corroborating previous reports [58,59].
Finally, we investigated if activation of AMPK-ACC pathway contributes to the increase in IDE expression. Interestingly, an increased IDE expression was noticed in C2C12, but not in HEPG2 cells, after exposure to AICAR for 3-h (Fig 8). Thus, activation of AMPK, by exercise, could contribute to the increase of IDE expression in the skeletal muscle. It is of note that IDE can also degrade receptor-bound insulin [60] interrupting IR phosphorilation and activation. Thus, higher activity of the AMPK-ACC pathway could block the insulin signaling, via increase IDE expression, corroborating our hypothesis that increase in the degradation of insulin, by the active skeletal muscle, would protect against hypoglycemia during physical activities.
In conclusion, we present evidence that a single bout of exercise increases insulin clearance to maintain glycemia at or close to physiological levels. This effect may occur via an increase in the expression of IDE in the liver and skeletal muscle, which may have implications for the effectiveness and safety of physical exercise in lean mice. Our data also suggest that activation of AMPK could be the mechanism whereby acute exercise increases IDE expression in the skeletal muscle.
Supporting Information S1 Fig. VO 2 max test. The treadmill exercise for VO 2 max test included a warm-up period of 5 min at 10 cm sec -1 with subsequently increase in the treadmill speed by 5 cm sec -1 each minute, until the mice reached exhaustion (for details, see Materials and Methods). The data are presented as the mean ± S.E.M., n = 10. (TIF) S1 Table. VO 2 max test data. VO 2 max, maximum speed and run distance reached by the mice (1-10) during the VO 2 max test. The mean data are presented as the mean ± S.E.M., n = 10. (DOCX)