INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

IN NORMALLY NOURISHED ADULT mammals, very few physiological manipulations result in elevations in rates of protein synthesis in skeletal muscle above rates found in the nonstressed/fed state. Resistance exercise is one such manipulation. For this reason, an extensive literature has documented the effects of this stress on skeletal muscle protein synthesis, protein degradation, and changes in the balance between these processes.

In addition to insulin's primary role in glucoregulation, this pancreatic hormone has a permissive role in regulating muscle mass stability (balance between synthesis and degradation). In nondiabetic and moderately diabetic rats, acute moderate-intensity resistance exercise causes increased rates of protein synthesis when the phenylalanine flooding dose is used to assess such rates (13). Severe insulin deficiency retards protein synthesis, and a previous study has shown that severely diabetic rats cannot increase rates of protein synthesis after moderate-intensity resistance exercise (15).

Under some circumstances, perhaps including exercise (2, 13), the initial steps in mRNA translation can be rate limiting for overall rates of protein synthesis. Insulin has an important role in regulating factors known to influence translation initiation. One such factor is eukaryotic initiation factor (eIF) 2B (eIF2B). This factor stimulates the exchange of GTP for GDP that is complexed to eIF2. This guanine nucleotide exchange reaction is a prerequisite for the formation of the ternary complex, eIF2-GTP-Met-tRNA_i, because eIF2-GDP does not bind efficiently to Met-tRNA_i. Severe insulin deficiency results in reduced activity of eIF2B in skeletal muscle (mostly type II fibers), and such reductions are concomitant with markedly reduced rates of protein synthesis (23). However, the activity of eIF2B is elevated in nondiabetic and moderately diabetic rats after resistance exercise. Thus, among other regulators, the inability of severely diabetic rats to increase rates of protein synthesis after moderate-intensity exercise could be due to an inability to increase the activity of eIF2B. No studies have reported the effects of resistance exercise on the activity of eIF2B in severely diabetic rats. This was the major goal for this study.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All experimental procedures were approved by the Institutional Animal Care and Use Committee of Pennsylvania State University. Male Sprague-Dawley rats were used in all experiments. They were housed in temperature- and humidity-controlled holding facilities with lights on at 0700 and off at 1900 and fed ad libitum a standard rodent diet (diet 5001, PMI Feeds) that contained 24% protein, 12% fat, 50% carbohydrate, 7% ash, 6% fiber, and vitamins.

Partial pancreatectomy. On the basis of previous work (10), a partial pancreatectomy procedure was used, and the procedure was modified to include rats that weighed 110-140 g as opposed to the weights (90-110 g) suggested by Foglia (18). We found that a larger percentage (83%) of the animals become diabetic when heavier rats are pancreatectomized. We also used a microcauterizer to eliminate small pancreatic blood vessels and to reduce bleeding during surgery. Sterile conditions were maintained throughout the surgery. Rats were anesthetized using isoflurane and were kept on a heated surgical pad. The procedure requires the physical removal of pancreatic tissue from the splenic, duodenal, and pyloric regions while major blood vessels are left intact. This is accomplished using sterile cotton Q-tips. Pancreatic tissue between the bile duct and the duodenum is not removed, since this approximates 10% of the original total pancreatic tissue. The resultant degree of diabetes depends on the ability of the surgeon to remove the majority of pancreatic tissue. In an effort to produce severely diabetic rats for this study, we were fastidious in removing as much pancreatic tissue as possible. At the conclusion of surgery, rats were given ampicillin (5 mg/100 g body wt sc) as an antimicrobial agent. Two weeks after partial pancreatectomy, a tail vein blood sample was obtained in the fed state to determine plasma glucose concentrations, and rats that were not severely diabetic (<475 mg/dl) were eliminated from the study. For experiments on eIF2B activity, 19 diabetic rats were randomly assigned to exercise (n = 10) or sedentary (n = 9) groups. Age-matched nondiabetic rats were randomly assigned to exercise (n = 7) or remained sedentary (n = 8) and were housed and handled in a manner identical to diabetic rats, with the exception of surgery and tail vein sampling to verify diabetic status. Using rats of a similar age, we previously showed that severely diabetic rats (5-h-fasted circulating insulin concentrations <80 pM) can perform moderate resistance exercise; however, rates of protein synthesis are not elevated in response to exercise. Previous work (12) also demonstrated that sham-operated rats had identical rates of protein synthesis during resting conditions and maintained an ability to elevate those rates after resistance exercise. Therefore, sham-operated rats were not used as controls in this study.

Acute resistance exercise. Details of the exercise protocol have been described previously (16). Briefly, rats were operantly conditioned to touch an illuminated bar low on a Plexiglas exercise cage and then were taught to stand and touch an illuminated bar that was located high on the opposite wall of the cage. Electrical foot shock (<2 mA, 60 Hz) was used to reinforce these movements. Once the learning process was completed (3-4 sessions), weighted vests were strapped over the scapulae and the rats were required to touch the high bar 50 times during one acute exercise session. We defined "acute" resistance exercise as four separate sessions with 1 day of rest between sessions. The rats performed 50 repetitions each day with 0.2 (day 1), 0.4 (days 2 and 3), and 0.6 (day 4) g weighted vest/g body wt. Previous work had shown that a rat that was naive to the lifting procedure would not lift the 0.6 g/g body wt on the 1st day weights were applied to the vest. This protocol can be considered acute, because it does not result in changes in muscle weight (17); however, it is probable that some adaptations that are characteristic of a trained state occur within the first few exercise sessions (4). Exercise sessions occurred in the dark (red light) in the late afternoon. Rats that did not perform exercise (sedentary) were placed in the lifting cages at least three times during the week of acute exercise and were given five electric shocks to simulate some of the stress experienced by the exercised groups. One of these shock control sessions occurred 16 h before the determination of rates of protein synthesis. All rats were anesthetized with isoflurane and killed 16 h after the last bout of acute exercise, and food was withdrawn during the last 5 h of this period.

Activity of eIF2B. Immediately after excision, gastrocnemius muscle was homogenized in 4 vol of a buffer specifically designed for measuring muscle eIF2B activity (buffer A). Buffer A (23) was made the day before use and contained (in mM) 20 triethanolamine, 2 magnesium acetate, 150 KCl, 0.5 dithiothreitol, 0.1 EDTA (Na₂), 25 sucrose, 5 EGTA, and 50 -glycerophosphate. The homogenate was centrifuged for 20 min at 10,000 g and 4°C, and the postmitochondrial supernatant was used immediately for the assay. eIF2 was complexed to [³H]GDP in buffer B, which contained 62.5 mM MOPS, 125 mM KCl, 1.25 mM dithiothreitol, and 0.25 mg/ml BSA. Buffer B (104 µl) and 42.8 µl of H₂O, 10.5 µl of eIF2, and 2.3 µl of [³H]GDP were mixed by tube inversion at 30°C for 10 min. A nonradioactive GDP mix was made by combining 1.2 mg of GDP, 10 ml of buffer C (buffer B with the addition of 2.5 mM magnesium acetate), and 2 ml of H₂O. The assay was started with the addition of 40 µl of postmitochondrial supernatant to 161 µl of assay buffer C, 140 µl of H₂O, and 40 µl of eIF2-[³H]GDP. Aliquots (60 µl) were taken at 8, 30, 60, 180, 360, and 540 s. The eIF2-[³H]GDP complex was vacuum captured on nitrocellulose filters, which were then dissolved by vortexing in Filtron-X scintillation fluid. Beta radiation was quantified using liquid scintillation counting, with appropriate correction for quench due to the dissolved filters. Samples from individual muscles were assayed in duplicate.

In vivo rates of protein synthesis. To date, our report (15) has been the only one to make the observation that severely diabetic rats cannot increase rates of protein synthesis after resistance exercise. Therefore, we repeated that study using five or six additional rats per group. All exercise, handling, and fasting conditions were identical to those described above for rats used in the eIF2B activity experiments. Rats were anesthetized with isoflurane, and the left carotid artery and right jugular vein were cannulated. Total time between the onset of anesthesia and completion of surgery was 10-16 min. Rats remained unconscious on a heated surgical pad during the flooding dose protocol, which immediately followed the insertion of catheters. Arterial blood (1 ml) was taken for the determination of insulin and glucose. After cannulation, a flooding dose (19) of L-[2,3,4,5,6-³H]phenylalanine (1 mCi/rat; Amersham Life Science, Arlington Heights, IL) in nonradioactive phenylalanine (150 mM; l ml/100 g body wt total volume) was injected into the venous catheter over a 20-s period. Arterial blood (1 ml) was taken at 6 and 10 min, and then the gastrocnemius muscle was excised. Muscles were dropped into liquid nitrogen. Rates of protein synthesis were measured separately for each animal. Frozen muscles were stored at 70°C until radiolabeled phenylalanine incorporation into trichloroacetic acid-precipitable protein was measured on pulverized tissue. Beta radiation in the solubilized protein precipitate was measured using liquid scintillation counting with appropriate correction for volume, color, and particle quenching. Plasma specific radioactivity of phenylalanine was analyzed after dabsylation (8) of the amino acid and quantification of the amount of phenylalanine using HPLC. Radioactivity in the phenylalanine chromatographic peak was measured by liquid scintillation counting with appropriate correction for quench. The specific radioactivity of plasma phenylalanine was used as the estimate of the precursor pool for phenylalanine (7, 19, 28). Protein determinations were made using the biuret method, and those values were used in the calculation of rates of protein synthesis. The amount of radiolabeled phenylalanine incorporated into muscle protein was calculated using the methods of Garlick et al. (19).

Statistical analysis. Data were analyzed using the PROC Mixed General Linear Model program of SPSS. The model was a 2 × 2 groups-by-treatment design with diabetes status (nondiabetic × diabetic) and exercise (sedentary × exercised) as the groups and treatments, respectively. When significant F ratios were found, post hoc Student-Newman-Keuls tests were applied to locate means that differed significantly at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Physical and physiological characteristics of the groups of rats are provided in Table 1. Because there was no statistical difference in physical and physiological characteristics between the groups of rats used for the eIF2B and rates of protein synthesis experiments, those data were combined and are presented in Table 1. As expected, diabetic rats weighed less and had lower arterial insulin concentrations and higher plasma glucose than nondiabetic rats. Hematocrit and hemoglobin were not different (data not shown) between groups, although a tendency toward higher values in diabetic groups was noted.

Rates of protein synthesis for gastrocnemius muscle in response to exercise are provided in Fig. 1. These responses are similar (not significantly different from data in Fig. 1 of Ref. 15) to those previously reported, except the basal rates of protein synthesis are somewhat lower than those previously reported. Rates of protein synthesis were higher in the exercised nondiabetic group, but no elevation was apparent in the severely diabetic group.

View larger version (31K): [in this window] [in a new window]   Fig. 1.   In vivo rates of protein synthesis for nondiabetic (n = 6/group) and severely diabetic (n = 5/group) rats that exercised (Ex) or remained sedentary (Sed). *Rates were significantly (P < 0.05) higher for nondiabetic than for severely diabetic rats and were increased by exercise. **Rates were lower in severely diabetic rats but were not affected by exercise.

The activity of eIF2B was higher in exercised nondiabetic than in sedentary nondiabetic rats, but the same relative intensity of exercise did not alter eIF2B activity in severely diabetic rats (Fig. 2). Severely diabetic rats had lower eIF2B activity than nondiabetic rats, a finding previously reported (23).

View larger version (26K): [in this window] [in a new window]   Fig. 2.   Activity of eukaryotic initiation factor 2B (eIF2B). Values are means ± SE for 7-10 rats/group. *Activities were higher in exercised nondiabetic than in sedentary nondiabetic rats. **Activity of eIF2B was lower in exercised and sedentary diabetic than in nondiabetic rats and was not affected by exercise in severely diabetic rats.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

These data increase our knowledge about the effects of resistance exercise on protein metabolism, in that severely diabetic rats cannot increase the activity of eIF2B in skeletal muscle after exercise, whereas nondiabetic rats can. These findings are consistent with previous work (15) as well as the verification study reported here demonstrating that severely diabetic rats (arbitrarily defined as having arterial glucose concentration in the 5-h-fasted state >475 mg/dl and plasma insulin concentrations below 100 pM) cannot increase rates of protein synthesis in mixed muscle proteins in response to exercise. This is in contrast to moderately diabetic rats (arbitrarily defined as having an arterial glucose of ~300-450 mg/dl and plasma insulin concentrations 180 pM) that can increase rates of protein synthesis in mixed muscle proteins after moderate-intensity resistance exercise. Plasma insulin concentrations in Table 1 for severely diabetic rats are somewhat higher than the ~80 pM critical concentration, below which we previously proposed to prohibit elevations in rates of protein synthesis due to exercise. This suggests the need to use caution in placing too much emphasis on any one absolute concentration of any one anabolic factor as having a determining role for protein synthesis. However, severe hypoinsulinemia clearly prohibits anabolic responses to exercise.

Many factors control the rate of mRNA translation, and some of these are regulated by insulin. Two factors that are particularly important are eIF2B and eIF4E (35). Briefly, eIF2B regulates guanine nucleotide exchange on eIF2, whereas eIF4E is important for the association of the 40S ribosomal subunit to mRNAs that have the 5'-terminal m⁷GTP cap. With use of an exercise protocol identical to that used in this study, initial reports show that eIF2B activity is increased in nondiabetic and moderately diabetic rats (13), while no change in eIF4E is apparent (14), although a trend toward increased formation of the eIF4E · eIF4G complex was noted. Elevations in the amount of this complex are consistent with elevations in rates of protein synthesis. As has been demonstrated many times (23, 26), the activity of eIF2B is reduced in severely diabetic rats, and data in Fig. 2 show that this activity is not elevated after exercise. This observation suggests but does not prove that eIF2B is causally linked to the lack of an anabolic response.

Regulation of eIF2B activity is complex (24, 26). The eIF2B protein is a heteropentamer consisting of -, -, -, -, and -subunits, with the -subunit being the only one that is phosphorylated under conditions that change the anabolic status of the cell. A full review of the conditions and factors that regulate eIF2B activity is not appropriate here, yet many possibilities exist for the lack of an increase in eIF2B activity in severely diabetic rats. Briefly, guanine nucleotide exchange on eIF2 depends on the availability of substrate (eIF2-GDP) cofactors such as Mg²⁺, cellular energy status (NADPH/NADP, ATP/ADP), and the activity of several protein kinases that phosphorylate other eIFs (20, 38). Of particular importance are the states of phosphorylation of eIF2 and eIF2B. Phosphorylation of eIF2 on Ser-51 clearly leads to reduced eIF2B activity (6), whereas phosphorylation of eIF2B has been shown to increase (1), decrease (42), or have no effect (33) on eIF2B activity. Decreased phosphorylation of eIF2 on Ser-51 may be related to contraction, but not insulin-induced changes in protein synthesis, because hypoinsulinemia does not alter the phosphorylation of eIF2 (22, 23). It must be appreciated that, with only a small number of exceptions, the literature has relied on cell culture or reticulocyte lysate preparations for our detailed understanding of eIF2B function, and it is clear that regulation of this important factor is more complicated when studied in vivo.

Although the eIF4E system does not seem to be involved in elevations of protein synthesis after exercise in nondiabetic rats (14), it remains possible that the eIF4E system is involved in the lack of an anabolic response observed in this study. Severely diabetic rats have deficits in eIF4E function, such that the amount of eIF4E complexed to its binding protein (binding protein-1), i.e., eIF4E · 4E-BP1, increases with diabetes (25) and thus reduces the availability of eIF4E for stimulation of translation initiation. In this initial study, we chose to focus on eIF2B, rather than eIF4E, because exercise does not stimulate all components of the eIF4E system, and the phenomenon we were trying to explain seemed more related to eIF2B. Future studies could be expanded to include more factors and/or more specific regulators of eIF2B.

In addition to insulin, several other hormones are known to stimulate or suppress protein synthesis and thus may be involved in the failure of protein synthesis to increase after exercise in severely diabetic rats. Two notable possibilities are glucocorticoids (inhibitory) and insulin-like growth factor I (IGF-I) (stimulatory). Although we did not measure these hormones in this study, previous work from our group or collaborators allows some speculation. Acute in vivo exposure of rats to dexamethasone results in suppression of protein synthesis with regulation at the level of translation initiation. Shah et al. (37) demonstrated that such suppression is due to changes in the eIF4E, rather than the eIF2B, system. The observed reductions in protein synthesis in this study may be partially due to mechanisms involving the eIF4E system because of excess glucocorticoids found in diabetic animals (39). It is clear, however, that reductions in protein synthesis and eIF2B activity are restored to normal basal values very rapidly after treatment with insulin alone (23). Clearly, more work must focus on the potential interaction between glucocorticoids and insulinemia during anabolic states. IGF-I may facilitate an anabolic response during periods of moderate hypoinsulinemia. Farrell et al. (13) demonstrated that muscle IGF-I was markedly elevated in moderately diabetic rats after resistance exercise, but this response was not observed in nondiabetic rats. This apparent compensation was associated with the ability of moderately diabetic rats to increase rates of protein synthesis after exercise. Because severely diabetic rats in this study could not elevate rates of protein synthesis, an IGF-I compensation was either insufficient or was absent.

The data in Fig. 1 support a permissive role for insulin in a rat model of type 1 diabetes mellitus. Some studies conducted on humans with type 1 diabetes mellitus, however, suggest that acute withdrawal of insulin for ~12 h does not alter rates of protein synthesis in mixed muscle proteins (40). This study differs from our study in at least two important ways. 1) Species differences may exist, because the data from rats consistently show that severe diabetes reduces rates of protein synthesis in skeletal muscle composed primarily of type II fibers (15, 21, 23, 27, 36), whereas the data from diabetic humans (3, 5, 34, 40) show that short-term insulin deprivation does not alter rates of skeletal muscle protein synthesis. 2) The rats used in the present study had been in a severely hypoinsulinemic state for 5 wk before measurement of amino acid incorporation into mixed muscle protein. Long-term hyperglycemia/hypoinsulinemia is known to produce cellular adaptations that can adversely affect translational control (21, 27). Additionally, in contrast to studies using humans, we did not measure rates of protein degradation, which are known to be higher in diabetic humans (34) and rodents (39), and this would affect protein stability. Additional studies are needed to determine the mechanisms that clarify a role for insulin in these two species during periods of anabolism.

Any practical application of these data must be done cautiously. To our knowledge, only three studies (9, 29, 31) have appeared in which humans with well-controlled type 1 diabetes performed resistance exercise training; however, in two (29, 31) of these three studies, aerobic and resistance exercise were combined. Those studies are encouraging, in that each included a statement that no untoward effects of the weight training were noted; however, specific details of how this was determined were not provided. In each study, the participants with type 1 diabetes mellitus increased muscle strength, and in two (9, 32) of the three studies, significant reductions in glycosylated hemoglobin A1c were found. The later finding specific to resistance exercise is in contrast to the general consensus that hemoglobin A1c does not change in humans with type 1 diabetes in response to chronic endurance exercise (41, 43). The present study using rats focused on acute resistance exercise; however, moderately diabetic rats can increase muscle mass in response to 10 wk of resistance exercise (11). The present study suggests that severely diabetic rats probably would not gain muscle mass as a result of this form of exercise.

In summary, severe hypoinsulinemia is associated with reduced activity of eIF2B, which is not normalized by resistance exercise. This diabetes-related deficit in the ability to make new proteins implies that translational control is compromised at a time when the body needs to invoke anabolic mechanisms.