Asp, Sven, Allan Watkinson, Nicholas D. Oakes, and Edward W. Kraegen. Prior eccentric contractions impair maximal insulin action on muscle glucose uptake in the conscious rat.J. Appl. Physiol. 82(4): 1327–1332, 1997.—Our aim was to examine the effect of prior eccentric contractions on insulin action locally in muscle in the intact conscious rat. Anesthetized rats performed one-leg eccentric contractions through the use of calf muscle electrical stimulation followed by stretch of the active muscles. Two days later, basal and euglycemic clamp studies were conducted with the rats in the awake fasted state. Muscle glucose metabolism was estimated from 2-[14C(U)]deoxy-d-glucose andd-[3-3H] glucose administration, and comparisons were made between the eccentrically stimulated and nonstimulated (control) calf muscles. At midphysiological insulin levels, effects of prior eccentric exercise on muscle glucose uptake were not statistically significant. Maximal insulin stimulation revealed reduced incremental glucose uptake above basal (P < 0.05 in the red gastrocnemius;P < 0.1 in the white gastrocnemius and soleus) and impaired net glycogen synthesis in all eccentrically stimulated muscles (P < 0.05). We conclude that prior eccentric contractions impair maximal insulin action (responsiveness) on local muscle glucose uptake and glycogen synthesis in the conscious rat.
- eccentric muscle contractions
- euglycemic clamp
exercise is generally considered to have beneficial effects on glucose homeostasis, and a single bout of concentric exercise (which involves shortening of active muscle) is a recognized enhancer of insulin action systemically and in muscle in rats and humans (6, 22, 24, 25). However, a recent study has indicated that a bout of eccentric exercise (which involves forced lengthening of active muscle) transiently impairs whole body midphysiological insulin action 2 days after the bout (17). The underlying mechanism(s) for this apparent insulin-resistant state remains obscure, but local changes in the individual muscles and/or systemic changes are possible, and in particular the former might involve a decrease in GLUT-4 content in the damaged muscles (2, 3). In vitro data obtained by using the perfused rat hindquarter technique revealed impaired insulin-stimulated glucose transport in the muscles exposed to prior eccentric contractions (4), suggesting the involvement of local changes. However, whether eccentric contractions have similar effects on local insulin action in the intact conscious rat is unknown. In the present study, we have combined the one-leg stimulation model previously described by our laboratory (3) and tracer methods for determining in vivo tissue specific glucose metabolism (14) to further elucidate mechanisms of insulin resistance induced by prior eccentric exercise. Basal, midphysiological, and maximal insulin action was measured locally in muscles in conscious rats 2 days after one-leg eccentric contractions of the calf muscles.
MATERIALS AND METHODS
Animals and diets.
All experiments were approved by the Garvan Institute Animal House Ethics Committee and complied with the National Health and Medical Research Council of Australia Guidelines for the Care and Use of Animals for Research Purposes. Adult male Wistar rats (bred in our facility) were housed individually in wire cages in a temperature-controlled room (23°C) on a 12:12-h light-dark cycle (lights on at 0600). All rats were fed ad libitum on a standard chow diet (Bavestock Feeds, Melbourne, Australia; 65% energy intake as carbohydrate).
Animal surgery and eccentric stimulation.
Rats weighing ∼300 g were anesthetized with ketamine (60 mg/kg ip; Parke Davis, Sydney, Australia) and xylazine (0.3 mg/kg im; Astra, Sydney, Australia) and fitted with jugular and carotid cannulas as previously described (14). After recovery from surgery, all rats were maintained in a sedentary state in individual cages, and feeding was continued as for preoperative schedules. Seven days after surgery, the rats were anesthetized with ketamine and xylazine, and the calf muscles on one side were electrically stimulated for eccentric contractions as described earlier (3). To facilitate in vivo metabolic studies, heavier Wistar rats (∼300 vs. 230 g) were used in this study than in the previous one (3), and, therefore, a higher pressure was applied (4 vs. 3 bar) to the piston chamber used for stretching the active muscles. The unstimulated contralateral calf muscles were used as controls. Stimulation sessions were carried out between 0800 and 1200, and after the rats recovered from anesthesia, their gait appeared completely normal. The clamp studies were conducted with the rats in the conscious state 2 days after the electrical stimulation late in the morning after a 5- to 6-h fast.
Before initiation of glucose clamp studies, a basal blood sample (0.6 ml) was taken for measurement of plasma glucose and insulin, and red blood cells were resuspended in saline and returned. Insulin infusion was then commenced at a rate of 0.25 (midphysiological insulin concentrations) or 2.0 U ⋅ kg−1 ⋅ h−1(maximal insulin levels) while the blood glucose concentration was clamped at 4.5 mM. Potassium (KCl) was infused at 0.3 and 1.2 mmol ⋅ kg−1 ⋅ h−1, respectively. After ∼75 min, the glucose infusion rate (GIR) necessary to keep a constant plasma glucose level was stable. A bolus injection of 60 μCid-[3-3H]glucose (Amersham) and 30 μCi 2-[14C(U)]deoxy-d-glucose (2-DG; Du Pont-New England Nuclear, Boston, MA) was then administered via the jugular line, the line was carefully washed with physiological saline, and blood samples (0.2 ml) were taken at 3, 6, 10, 15, 20, 30, and 45 min for determination of plasma glucose and tracers levels. At the completion of the studies (45 min after tracer administration), a final blood sample (0.6 ml) was taken. Rats were quickly anesthetized (110 mg/kg iv pentobarbital sodium) and the following hindquarter muscles were rapidly removed and freeze-clamped with tongs precooled in liquid N2(all within 90 s) for subsequent analysis: white part of the gastrocnemius (WG; consisting mainly of fast-twitch white fibers), soleus (S; consisting mainly of slow-twitch oxidative fibers), and red part of the gastrocnemius (RG; consisting mainly of fast-twitch red fibers).
Blood and plasma glucose concentrations were determined by using a glucose analyzer (model 23AM, Yellow Springs Instruments, Yellow Springs, OH). Plasma samples for determination of insulin were kept at −20°C until assayed by a double-antibody radioimmunoassay as previously described (18). Plasma samples for determination of tracer concentration were deproteinized immediately after collection in 5.5% ZnSO4 and saturated Ba(OH)2. An aliquot of supernatant was dried down to remove tritiated water and was redissolved in an aqueous scintillant (Picofluor 40; Packard Instruments, Rockville, MD). Radioactivity counting was performed in a liquid scintillation spectrophotometer (Beckman Instruments, Fullerton, CA). Muscle glycogen concentration was assayed as previously described (12) in the basal and 2 U ⋅ kg−1 ⋅ h−1clamp groups (tissue requirements for GLUT-4 assay precluded measurement of glycogen in the 0.25 U ⋅ kg−1 ⋅ h−1 group). GLUT-4 content was assayed in the muscles of rats exposed to insulin at 0.25 U ⋅ kg−1 ⋅ h−1by a modification of a previously described procedure (2). In brief, the GLUT-4 protein content was quantified by Western blot using a mouse monoclonal primary antibody directed against the13COOH-terminal amino acids of GLUT-4 and a horseradish peroxidase-labeled goat anti-mouse antibody as described previously (2). GLUT-4 values are given in arbitrary units, expressed as percentage of a rat heart standard per milligram sample protein. For histological determination of leucocyte/phagocyte infiltration and focal muscle necrosis, embedded frozen muscle samples were cut in a microtome at a thickness of 10 μm and were stained with hematoxylin and eosin.
Muscle glucose uptake [estimated as the glucose metabolic rate index (Rg′) of phosphorylated 2-DG] and [3H]glucose incorporation into glycogen (net glycogen synthesis rate) were calculated as previously described (15, 18). In brief, estimates of whole body rates of insulin-stimulated glucose disposal and hepatic glucose output were obtained from the plasma disappearance of [3H]2-DG by using methods described previously (15). Estimates of the rate of glucose uptake by individual tissues (Rg′) were based on tissue accumulation of phosphorylated 2-DG (18). The insulin-stimulated glucose uptake above basal was calculated by subtracting the average basal Rg′ from the midphysiological or maximal Rg′ in each muscle type and leg [control and eccentrically stimulated (Ecc)].
All experimental data were expressed as means ± SE. A two-way repeated-measures analysis of variance on one factor was used, and Student’s paired t-test with the Bonferroni correction was used as post hoc test.
Histological analysis revealed that inflammatory cells had accumulated in all Ecc-calf muscles. This was most pronounced in the RG (Fig.1).
The basal Rg′ was marginally higher in Ecc-WG compared with control (P < 0.1; Table1), whereas there were no significant differences between Ecc- and control muscles in WG at midphysiological and maximal insulin concentrations. In contrast, in the RG and S, Rg′ was unaffected by prior eccentric contractions at basal and midphysiological insulin concentrations and was lower at maximal insulin concentrations compared with control (P < 0.05). The maximal incremental insulin-stimulated glucose uptake above basal was reduced in all Ecc-muscles (P < 0.05 in RG andP < 0.1 in WG and S; Fig.2), whereas no differences were found in these muscles exposed to midphysiological insulin concentrations. The ratio of phosphorylated to free 2-DG was not different between Ecc- and control muscles at any insulin concentration (data not shown).
The glycogen concentration was lower in the Ecc-WG and Ecc-RG muscles compared with control in basal rats and after maximal insulin stimulation (P < 0.05). In addition, glycogen content was lower in the Ecc-S compared with control after maximal insulin stimulation (P < 0.05), whereas glycogen content was unaffected in the basal Ecc-S (Table 2). Rates of glucose incorporation into glycogen were lower in all Ecc-muscles compared with control during maximal insulin stimulation (P< 0.05), whereas no differences were found in muscles exposed to basal concentrations of insulin (Table 3).
Total GLUT-4 content was lower in the Ecc-WG and Ecc-RG muscles compared with the respective unstimulated control leg muscles [22.8 ± 4.6 vs. 32.6 ± 2.7 (P < 0.05) and 38.0 ± 6.5 vs. 63.8 ± 4.7 arbitrary units (P < 0.05), respectively], whereas the content was unaffected by prior eccentric contractions in the S (113.5 ± 12.6 vs. 110.5 ± 7.9 arbitrary units).
The principal finding in this study is that prior eccentric muscle contractions impair maximal insulin action (responsiveness) on muscle glucose uptake and net glucose incorporation into glycogen in the conscious rat.
A single bout of concentric exercise is a recognized enhancer of insulin action systemically and in muscle in rats and humans (6, 22,24, 25), whereas a previous study (17) described transient whole body midphysiological insulin resistance 2 days after a bout of unaccustomed eccentric exercise. Whole body insulin resistance can be caused by local changes in the individual exercised muscles and/or by systemic changes, and the present study supports the importance of local changes. All three calf muscles were apparently recruited extensively during the eccentric contractions because inflammatory cells accumulated in all fiber types, and incremental glucose uptake above basal was impaired or marginally impaired (P < 0.1) in all Ecc-muscles during maximal insulin stimulation (insulin concentration of 678 ± 102 μU/ml), whereas it was unchanged at midphysiological concentrations (insulin concentration of 86 ± 7 μU/ml). The changes in muscle insulin action only at maximal and not at midphysiological concentrations were unexpected and could reflect that any less-pronounced difference at the midphysiological concentration might be lost in the experimental variation. The importance of the local changes is supported by a another recent study using the rat hindquarter technique and showing insulin-resistant glucose transport in muscles 2 days after eccentric contractions (3). Insulin-stimulated glucose transport was impaired in the Ecc-WG in the presence of upper physiological (200 μU/ml) and maximal (20,000 μU/ml) insulin concentrations and in the Ecc-RG during maximal insulin stimulation (20,000 μU/ml), in the presence of inflammatory cells in both muscles, whereas glucose transport was unaffected or increased in the Ecc-S in the absence of inflammatory cells. The disparate effects on the S in the present and previous studies might be caused by different rat sizes used or pressure applied for stretching the active calf muscles. Both factors could change the strain on the individual calf muscles during the eccentric contractions, but this is hard to evaluate without individual measurements of tension in the different muscles. The mechanism(s) behind the observed local insulin resistance remains obscure but might involve the cytokine tumor necrosis factor-α (TNF-α). TNF-α is produced by inflammatory cells and possibly by muscle cells (28), and studies have shown that the cytokine induces insulin resistance (21) and that it is overexpressed in muscle tissue from insulin-resistant and diabetic subjects (26). It can be hypothesized that TNF-α produced by inflammatory cells and possibly even by damaged myocytes causes insulin resistance in the muscles subjected to prior eccentric contractions.
The GIR necessary to maintain euglycemia was not different between the rats that were used in the present study and age- and weight-matched rats with one-leg concentrically stimulated muscles (unpublished observations) at midphysiological or maximal insulin concentrations. This is in accord with human experiments by King et al. (16), who found no effect on GIR during a hyperglycemic clamp at insulin levels of ∼40 μU/ml ∼36 h after maximal eccentric contractions of both thighs, but it is in contrast to findings by Kirwan et al. (17), who described reduced submaximal (insulin at 35 μU/ml) GIR in subjects 2 days after they ran for 30 min on a treadmill with a negative incline. Neither of these human studies investigated the local effects, and the different results on whole body insulin action might be due to the nature of the eccentric contractions (in situ electrical stimulation and maximal eccentric contractions of both thighs vs. running on a treadmill with a negative incline) or the muscle mass involved in the eccentric contractions. It should be borne in mind when present and past studies are analyzed that the calf muscles only constitute a minor fraction of the total rat mass, and, therefore, a change in the eccentrically stimulated muscles would not necessarily be measurable systemically.
It is well established that muscle glycogen concentration remains subnormal several days after eccentric contractions despite a carbohydrate-rich diet (3, 5, 7, 9, 23, 29). Consistently, muscle glycogen concentrations in WG and RG at the end of the clamps were lower in the Ecc-leg than in the control leg, and it was reduced to the same extent as described previously (3, 4). Reduced muscle glycogen content can be caused by decreased glycogen anabolism and/or increased catabolism, and insulin action is of interest in this context because it is the most potent glycogen anabolic hormone. A previous study (4) suggested that insulin-stimulated net glycogen synthesis might be reduced in Ecc-muscle, but no previous studies have directly measured insulin stimulated glycogen synthesis after eccentric contractions. The present study demonstrating reduced net glucose incorporation into glycogen in the maximally stimulated Ecc-muscles confirms this suggestion. However, the basal and midphysiological Rg′ and basal glucose incorporation into glycogen were not significantly different from control muscles in the Ecc-RG or Ecc-S, whereas the basal Rg′ was increased and unchanged at midphysiological insulin concentrations in the Ecc-WG. The unchanged or increased glucose uptake found in the present study and the recently reported unchanged activity of glycogen synthase (4) with physiological insulin elevation make it less likely that altered insulin action alone can account for the subnormal muscle glycogen concentrations found for several days after eccentric contractions in the conscious rat. This suggests glycogen catabolism as the major determinant, but the present data do not allow for any final conclusion in that respect. Alternatively, reduced basal glucose uptake into the damaged muscles could be masked by the higher glucose utilization in inflammatory cells (7, 10, 27) that are present in the eccentrically damaged muscles; also, if the cytosolic Ca2+ concentration is increased in damaged muscle, this may cause increased glucose transport (30). Accumulation of inflammatory cells in eccentrically damaged muscles is pronounced in rats, whereas it has been found to be less pronounced in most human studies, and could indicate that the muscles in the present study were damaged to a greater extent than in previous human studies (2, 7). However, the glycogen in the present study was reduced to the same degree onday 2 after eccentric contractions in basal rats (WG 23%, RG 18%) as previously reported in a human study (18%) (2), and previous studies revealed that muscle glycogen repleted similarly after eccentric contractions in both species (2, 3). The almost identical glycogen pattern after eccentric contractions in both species despite different accumulation of inflammatory cells makes it less likely that there is a causal relation between subnormal glycogen and accumulation of inflammatory cells, as suggested by previous studies (10, 27).
Transmembrane glucose transport is thought to be the rate-limiting factor for the glucose uptake under most circumstances (11, 20). Insulin-induced increase in muscle glucose uptake has been found to be dependent on the GLUT-4 protein content (1, 8, 13), but whether this holds true in damaged muscle is uncertain. Our laboratory (3) previously reported that the GLUT-4 protein content was particularly affected by eccentric muscle contractions, whereas the concentration of another insulin-regulatable protein, glycogen synthase, was unchanged. In accord with previous rat (3, 19) and human (2) studies, GLUT-4 content was 30% lower in the Ecc-WG, 40% lower in the Ecc-RG, and unchanged in the Ecc-S compared with control (3,19). Consistent with the view that total GLUT-4 content determines the capacity for maximal glucose uptake, maximal insulin-stimulated Rg′ above basal was reduced 28% in the Ecc-WG and was reduced 24% in Ecc-RG. In contrast, the maximal insulin-stimulated glucose uptake was reduced 21% in the Ecc-S compared with control in the face of unchanged GLUT-4 content, which suggests that the rate-limiting step could move toward intracellular disposal when rat muscles are exposed to maximal insulin and glucose (11).
In conclusion, we found that prior eccentric contractions impair maximal insulin action (responsiveness) on the local muscle glucose uptake and net glucose incorporation into glycogen in the conscious rat.
Souad Camilleri and Vicki Theos provided skilled technical assistance. The Diabetes Research Group at the Garvan Institute and especially Stuart Furler are sincerely thanked by S. Asp for the hospitality, fruitful discussions, and help with the initial glucose clamps. Don Chisholm and Erik A. Richter are thanked for making the study possible. The GLUT-4 antibody was kindly donated by Dr. Per Norup Jørgensen (Novo Nordisk, Bagsværd, Denmark).
Address for reprint requests: S. Asp, Copenhagen Muscle Research Centre, August Krogh Institute, Univ. of Copenhagen, 13 Universitetsparken, DK-2100 Copenhagen, Denmark (E-mail:).
S. Asp′s stay at the Garvan Institute of Medical Research was supported by the Danish Research Council, Danish National Research Foundation Grant 504-14, and the Garvan Research Foundation. A. Watkinson, N. D. Oakes, and E. W. Kraegen were supported by the National Health and Medical Research Council of Australia.
- Copyright © 1997 the American Physiological Society