0013-7227/90/1264-1921$02.00/0 Endocrinology Copyright © 1990 by The Endocrine Society
Vol. 126, No. 4 Printed in U.S.A.
Effects of Streptozotocin-Induced Diabetes on Glucose Transport in Skeletal Muscle* R. JAMES BARNARD, JACK F. YOUNGREN, DOUGLAS S. KARTEL, AND DEBORAH A. MARTIN Department of Kinesiology, University of California, Los Angeles, California 90024-1568
ABSTRACT. Female Sprague-Dawley rats were injected with streptozotocin (45 mg/kg) to induce mild diabetes (glucose, >13 mM). Half of the animals received daily insulin injections to reduce hyperglycemia. After 10 weeks, sarcolemmal membranes were isolated from hindlimb muscles to study glucose transport, and the number of glucose transporters was assessed by cytochalasin-j8 binding. Both glucose transport (19.2 ± 1.6 us. 31.93 ±3.29 pmol/mg protein • 15 sec) and cytochalasin-/? binding (3.06 ± 0.28 vs. 6.14 ± 0.59 pmol/mg protein) were significantly (P < 0.05) reduced in the diabetic untreated rats compared to control values. Daily insulin injections restored both (P < 0.05) basal transport (33.22 ± 3.62 pmol/mg protein • 15 sec) and cytochalasin-/3 binding (5.52 ± 0.66 pmol/mg protein) to control levels. Maximum insulin stimulation (1 U/kg, iv) significantly increased (P < 0.05) both glucose transport (30.18 ± 3.76 vs. 96.48 ± 4.21 pmol/mg protein • 15 sec) and cytochalasin-/? binding (4.38 ± 0.29 vs. 9.40 ± 0.42 pmol/mg protein) in the untreated diabetic
I
N THE past 10 yr, significant progress has been made in understanding the mechanism of insulin action and some of the possible defects associated with insulinresistant states. Most of this research has been performed on fat cells or nontarget cells for insulin action. Insulin clamp studies have demonstrated that skeletal muscle is the most important target tissue for insulin action. De Fronzo (1) has estimated that 70-75% of the blood glucose removed during clamp studies is removed by skeletal muscle. Studies have also documented that the major defect, in insulin action, associated with noninsulin-independent diabetes mellitus is located in skeletal muscle (1). Insulin resistance is also known to develop in insulin-dependent diabetes mellitus. Whether the mechanisms involved in the development of insulin resistance in the two major forms of diabetes mellitus are identical is unknown, but some similarities have been found. Unfortunately, few studies have focused on cellular Received October 2, 1989. Address all correspondence and requests for reprints to: R. James Barnard, Ph.D., Department of Kinesiology, University of California, 405 Hilgard Avenue, Los Angeles, California 90024-1568. * This work was supported by NIH Grant DK-32326 and a grant from the Nathan Pritikin Research Foundation.
and control rats. However, the stimulation in the untreated diabetic rats only reached basal control levels, which was significantly (P < 0.05) below the insulin-stimulated value for the controls. In the rats receiving daily insulin injections, maximum insulin stimulation increased (P < 0.05) both glucose transport (58.67 ± 15.24 pmol/mg protein 15 sec) and cytochalasin-/? binding (6.4 ± 0 . 7 pmol/mg protein), but both transport and binding were significantly (P < 0.05) below insulin-stimulated values for the control rats. These data show that insulin deficiency adversely affected the glucose transport system in skeletal muscle. Both basal and maximum insulin-stimulated transport and the number of transport molecules were reduced. Daily insulin treatment corrected some of the defects, but maximum insulin stimulation was still significantly below values for control animals. (Endocrinology 126: 1921-1926, 1990)
aspects of skeletal muscle glucose transport in insulinresistant states, especially long term resistance. According to Ziel et al. (2), glucose transport is rate limiting for glucose metabolism in both normal and streptozotocindiabetic rats. We have recently developed an isolated skeletal muscle sarcolemmal membrane preparation to specifically study the glucose transport system (3, 4). In the present study we examined the effects of 10 weeks of treated and untreated streptozotocin-induced diabetes on glucose transport using our isolated sarcolemmal vesicle preparation.
Materials and Methods Animals Female Sprague-Dawley rats were purchased from Charles River Laboratories (Wilmington, MA) at about 5 weeks of age. The animals were divided into control, untreated diabetic, and insulin-treated diabetic groups. The diabetic groups were injected with streptozotocin (45 mg/kg, iv) and placed on a dextrose drinking solution for 3 days. On the morning of the fifth day, blood glucose was measured in all rats (One Touch, Lifescan Co., Mountainview, CA). If the morning glucose measurement was not above 13 mM in the diabetic rats, they were removed from the study. All control animals had a morning 1921
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 August 2016. at 02:41 For personal use only. No other uses without permission. . All rights reserved.
1922
DIABETES AND GLUCOSE TRANSPORT
Endo • 1990 Vol 126 • No 4
ton, MA) were added, and the vials were counted in a Beckman LS-7500 liquid scintillation counter (Palo Alto, CA). Filter retention of counts and nonspecific uptake were measured with L-[14C]glucose and subtracted from D-[3H]glucose values to obtain specific D-glucose transport. All measurements were made in triplicate and averaged to obtain a single specific transport value. The number of glucose transport molecules in the sarcolemmal membrane was assessed by the cytochalasin-/3-binding Sarcolemmal glucose transport and cytochalasin-(i binding stud- procedure of Cushman and Wardzala (6), as described in detail previously (4). Sarcolemmal membranes were placed in glass ies culture tubes with cytochalasin-E for 10 min. D- or L-glucose The tracers D-[3H]glucose, L-[14C]glucose, D-[14C]mannitol, was then added and allowed to bind for an additional 10 min. and [3H]cytochalasin-/3 with specific activities of 331. Ci/mol, To each of the five pairs of tubes containing D- or L-glucose, cytochalasin-/? was added in concentrations ranging from 1-24 47.0 mCi/mmol, 48.7 mCi/mmol, and 18.5 Ci/mmol, respecpM and mannitol in concentrations ranging from 0.1-0.4 nM. tively, were purchased from New England Nuclear (Boston, After binding, the membranes were pelleted by centrifugation MA). Cytochalasin-jS was purchased from Aldrich Chemical at 29,000 X g for 30 min. An aliquot of the supernatant was (Milwaukee, WI), and all other chemicals from Sigma Chemical placed in one scintillation vial, and the pellets were dissolved (St. Louis, MO). in NaOH and added to a second vial. Both were then counted The procedure for sarcolemmal isolation was started between in a liquid scintillation counter. All samples were made in 0800-0900 h; the animals were not fasted, and the insulinduplicate, and specific binding was determined by subtracting treated animals had received their regular insulin injection the the percentage of [14C] mannitol bound from the percentage of previous evening. For the maximum insulin stimulation studies, 3 [ H]cytochalasin-j8 bound. The total binding for cytochalasinthe rats were given an iv injection of purified porcine insulin j8 was calculated according to Scatchard (7). (1 U/kg) and killed 10 min later. Sarcolemmal vesicles were All data were placed on a computer for statistical comparison isolated using the methods described in detail previously (3). using SAS (Statistical Analysis Systems, Cary, NC) programs. The quadriceps, gastrocnemius, and plantaris muscles (all fastGroup means were compared using a two-way analysis of varitwitch) from each rat were trimmed, minced, and then homogance, followed by Tukey's test for significant differences. P < enized using a Polytron (Brinkmann Instruments, Westbury, 0.05 was considered statistically significant. Data are expressed NY). The muscles from two rats were combined for each as the mean ± SE. membrane preparation. After homogenization, the homogenate was filtered through wire mesh to remove connective tissues and mixed with KC1 Na pyrophosphate to dissolve contractile Results proteins. After vigorous mixing, the solution was centrifuged at 184,000 X g, and the pellet was resuspended in homogenizing Table 1 presents the clinical data obtained from the medium and incubated for 45 min at 30 C with 20 mg DNAase. rats during the final week of the experiment. The unAfter a low speed spin at 750 X g for 15 min, the supernatant treated diabetic rats had a lower body weight and higher was spun at 184,000 X g for 15 min. The pellet was resuspended morning glucose value than either the control or insulinin 45% sucrose and placed on the bottom of a sucrose gradient. treated group. While the daily insulin injections normalAfter 16 h of centrifugation at 64,000 X g, the sarcolemmal ized body weight in the treated group, blood glucose was vesicles were harvested from the F2 fraction, which had a still significantly higher than control values, but was density of about 1.1 g/ml. The activity of the enzyme potassignificantly lower than that in the untreated diabetic sium-stimulated p-nitrophenylphosphatase (KpNPPase) was group. The daily insulin injections in the treated group used as a sarcolemmal membrane marker; protein was measranged from 4-7 U/day. ured using the Bradford method (3). The percentage of vesicles that formed rightside out was assessed by measuring the Table 2 shows the sarcolemmal isolation data. There KpNPPase activity before and after deoxycholate treatment of was no significant difference in yield of sarcolemmal the vesicles, as described previously (3). After isolation, the protein per g wet wt of muscle among any of the groups. membranes were frozen and stored in liquid nitrogen. In addition, there was no significant difference in the Specific D-glucose transport was measured using the equilibrium exchange method of Ludvigsen and Jarett (5), as described TABLE 1. Final body weight and blood glucose data in detail previously (3, 4). Transport was measured for 15 sec Blood glucose at 37 C in a solution containing 180 nM D- and L-glucose. The Group BW(g) (mM) reaction was stopped by the addition of 3 mM HgCl2, and transported and untransported glucose were separated by ultraControl 290 ± 6 4.0 ± 0.2 filtration using Gelman 0.45-jum filters (product 63068, Ann Untreated diabetic 232 ± 5 a 19.9 ±0.7° Insulin-treated diabetic 296 ± 186 6.8 ± 1.2a>6 Arbor, MI). The filters were transferred to scintillation vials 0 and dissolved in 600 n\ ethylene glycol monomethyl ether. Ten Significantly different from control, P < 0.05. 6 millimeters of Aquasol-2 (DuPont-New England Nuclear, BosSignificantly different from untreated diabetic, P < 0.05.
glucose reading below 5 mM. The insulin-treated group received daily (late afternoon) sc injections of Iletin NPH I insulin (Eli Lilly Co., Indianapolis, IN) in quantities sufficient to reduce morning glucose levels to below 8 mM and as close to 6 mM as possible. The morning glucose level was measured weekly, and the insulin dosage adjusted as needed. Standard rat chow and water were provided ad libitum with the animals on a 12-h light, 12-h dark cycle, starting at 0700 h.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 August 2016. at 02:41 For personal use only. No other uses without permission. . All rights reserved.
DIABETES AND GLUCOSE TRANSPORT
1923
TABLE 2. Effect of untreated and insulin-treated diabetes on sarcolemma membranes Sarcolemma yield (mg protein/g muscle)
Group
KpNPPase h)
Control Control, insulin-stim Untreated diabetic Untreated diabetic, insulin-stim Insulin-treated diabetic Insulin-treated, insulin-stim
0.68 ± 0.08 8.05 ± 2.38 0.72 ± 0.12 6.82 ± 0.84 0.76 ± 0.15 7.74 ± 2.58 0.91 ± 0.15 5.91 ± 1.23 0.77 ± 0.08 7.34 ± 0.46 0.62 ± 0.04 6.25 ± 0.57 Insulin stimulated (insulin-stim) were given 1 U/kg regular porcine insulin 10 min before isolation. Values are the mean ± SE. Each group contained five membrane preparations. No significant differences were found for either the yield (milligrams of protein) or purity (KpNPPase activity).
activity of KpNPPase, the sarcolemma marker enzyme, indicating no significant differences in the purity of the sarcolemma membrane preparation among any of the groups. The percentage of vesicles that formed rightside out was measured in control, untreated diabetic and insulin-treated groups, and values were 56 ± 2, 60 ± 8, and 60 ± 1, respectively (not significantly different). Figure 1 shows the 15-sec specific D-glucose transport data. Under basal conditions, glucose transport was significantly (P < 0.05) decreased in the untreated diabetic group compared to that in controls (19.2 ± 1.6 vs. 31.9 ± 3.3 pmol/mg protein-15 sec). Insulin treatment restored basal transport to control levels (30.2 ± 3.8 pmol/mg protein-15 sec; P < 0.05). Maximum insulin stimulation increased transport 3-fold in the control animals to 96.5 ± 4.2 pmol/mg protein -15 sec (P < 0.01). In the untreated diabetic group, insulin stimulation resulted in only a 1.6-fold increase in glucose transport to 30.2 ± 3.8 EFFECT OF STREP DIABETES ON SARCOLEMMAL VESICLE GLUCOSE TRANSPORT
pmol/mg protein-15 sec (P < 0.05), which was significantly (P < 0.05) lower than the maximum stimulation achieved in the control group. In the insulin-treated group, insulin stimulation resulted in a 1.8-fold increase in glucose transport to 58.8 ± 15.2 pmol/mg protein • 15 sec (P < 0.05), which was still significantly (P < 0.05) below the maximum transport achieved in the control group. Figure 2 shows the cytochalasin-/?-binding sites for both basal and insulin-stimulated conditions. Under basal conditions cytochalasin-/? binding was significantly (P < 0.05) reduced in the untreated diabetic group compared to that in controls (3.06 ± 0.28 us. 6.14 ± 0.53 pmol/mg protein). With maximum insulin stimulation, cytochalasin-/? binding was significantly (P < 0.05) increased in the control group to 9.4 ± 0.72 pmol/mg protein. In the untreated diabetic group maximum insulin stimulation increased cytochalasin-/? binding to 4.38 ± 0.29 pmol/mg protein (P < 0.05), which was signifiEFFECT OF STREP DIABETES ON SARCOLEMMAL VESICLE CYTOCHALASIN-13 BINDING 12 -i
D
•
Basal N = 5
10 -
Insulin Stimulated N = 5
Basal N = 5 Insulin Stimulated N = 5
I
O)
o a.
Q.
O
o. CONTROL
UNTREATED DIABETIC
INSULIN TREATED DIABETIC
FIG. 1. Effect of streptozotocin diabetes on sarcolemmal vesicle glucose transport. Basal transport was reduced in the untreated diabetic group (P < 0.05). Insulin stimulation increased transport in all three groups (P < 0.05); however, the maximum response was significantly different in all three groups (P < 0.05). n = 5 refers to the number of sarcolemmal vesicle preparations studied for each group.
CONTROL
UNTREATED DIABETIC
INSULIN TREATED DIABETIC
FIG. 2. Effect of streptozotocin diabetes on cytochalasin-/3 binding in sarcolemmal vesicles. Basal binding was significantly reduced in the untreated diabetic group (P < 0.05), and all three maximum values were significantly different (P < 0.05). n = 5 refers to the number of sarcolemmal vesicle preparations studied for each group.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 August 2016. at 02:41 For personal use only. No other uses without permission. . All rights reserved.
1924
DIABETES AND GLUCOSE TRANSPORT
cantly (P < 0.05) below the insulin-stimulated value for the controls. In the insulin-treated group, basal cytochalasin-/? binding was 5.52 ± 0.66 pmol/mg protein, which was not significantly different from that in the control group. Maximum insulin stimulation increased (P < 0.05) cytochalasin-/? binding to 6.46 ± 0.76 pmol/ mg protein, which was significantly (P < 0.05) below the maximum stimulation for the control group.
Discussion Skeletal muscle, the most important target tissue for insulin, is adversely affected by 10 weeks of mild streptozotocin diabetes. Our data show that glucose transport, under both basal and insulin-stimulated conditions, is significantly reduced in this untreated diabetic model. Daily insulin injections restored basal glucose transport to control levels; however, maximum insulin stimulation was still significantly reduced compared to that in controls. The reduction in glucose transport in the untreated diabetic rat was due primarily to a reduction in the number of glucose transport molecules in the sarcolemmal membrane, as assessed by cytochalasin-/? binding. The diabetic state, either treated or untreated, did not affect the yield or purity of the isolated sarcolemmal membranes or the percentage of vesicles that formed rightside out. According to De Fronzo (1), skeletal muscle is the major site of insulin resistance associated with diabetes mellitus, especially noninsulin-dependent diabetes mellitus. Insulin resistance also develops with insulin-dependent diabetes mellitus. Whether the mechanisms responsible for insulin resistance in the two major forms of diabetes mellitus are similar remains to be determined. It has been reported that insulin binding is increased in the streptozotocin-treated insulin-deficient rat (8-14), but is decreased in the obese rat (15, 16), both of which develop insulin resistance. This suggests that different mechanisms may be involved. While an extensive body of research has been published using the streptozotocin diabetic rat model, almost all of the studies have been performed on fat cells, with only limited information obtained from skeletal muscle. The use of an isolated sarcolemmal vesicle preparation obtained from skeletal muscle has enabled our laboratory to focus directly on the glucose transport system and the mechanisms of insulin action (3, 4). In the present study basal glucose transport was reduced by 40% after 10 weeks of streptozotocin-induced untreated diabetes. These results are similar to those reported for isolated fat cells (8-10,12, 17). Studies on skeletal muscle, either hindlimb perfusion (18) or in vitro incubated muscle (11, 19), have reported decreases in 3-methylglucose or 2deoxyglucose uptake of 40-70% under basal conditions
Endo • 1990 Vol 126 • No 4
for mild and severe diabetes, respectively. Since insulin is significantly reduced in this untreated diabetic model, it is not surprising to see a decrease in basal glucose transport. Our previous studies (4) have indicated that insulin increases glucose transport in skeletal muscle by both translocation of additional glucose transport molecules into the sarcolemma membrane as well as activation of transporters normally present in the membrane. The data for insulin stimulation in control animals in the present study also support this concept, i.e. transport was increased by 302% while cytochalasin-/? binding was increased by only 154% after insulin stimulation. Another interpretation may be made, however, on the basis of recent data showing different basal vs. insulin-responsive isoforms of glucose transporters (20). The insulinresponsive isoform of glucose transporter may simply transport at a faster rate than the basal form, and thus, translocation could be the sole mechanism involved in insulin action. The data from the untreated diabetic and insulin-treated groups in the present study, however, do not support this concept, in that the insulin-stimulating effect was similar for both transport and cytochalasin binding; i.e. in the untreated diabetic animals transport was increased by 157%, while cytochalasin-/3 binding was increased by 143%, and in the insulin-treated group the values were increased by 195% and 117% for transport and cytochalasin-/? binding, respectively. In the present report cytochalasin-/? binding studies showed a 50% decrease in the number of glucose transport molecules in the sarcolemma membranes of untreated diabetic rats. The difference between the 40% decrease in glucose transport and the 50% decrease in the number of glucose transport molecules is within the error of measurement, but shows that the decreased glucose transport is due mainly to a decrease in available transport molecules. Ramlal et al. (21) also found the number of cytochalasiniS-binding sites to be reduced by approximately 50% in plasma membranes from diabetic rat muscle under basal conditions. These results, collectively, suggest that the main effect of diabetes in the streptozotocin diabetic model is to reduce the number of basal transporters as well as the number of transporters translocated into the plasma membrane after insulin stimulation. In previous studies (3,4) we demonstrated that a single insulin injection of 1 U/kg, iv, resulted in maximum stimulation of the glucose transport system. In the present study maximum insulin-stimulated transport achieved in the untreated diabetic sarcolemma preparations only reached the basal transport level in the control sarcolemma preparations. These results are similar to those reported for fat cells (8-10, 12, 17) and indicate postreceptor binding defects. Decreases in autophosphorylation and the tyrosine kinase activity of the insulin
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 August 2016. at 02:41 For personal use only. No other uses without permission. . All rights reserved.
DIABETES AND GLUCOSE TRANSPORT receptor have been reported (22-28) for liver and muscle insulin receptors, as has a decrease in the fat cell intracellular pool of transporters available for translocation (29-31). More recently, Garvey et al. (32) reported a decrease in mRNA for the insulin-sensitive isoform of glucose transporter in muscle from streptozotocin-induced diabetic rats, and Nishimura et al. (33) reported a decrease in the microsomal pool of glucose transporters. Whether these cellular changes in the untreated streptozotocin-induced diabetic rat are the result of hyperglycemia or hypoinsulinemia is not known, especially for long term diabetes as used in the present study. Walker et al. (34) have recently reported short term studies using L6 muscle cells in which they found that both glucose and insulin can regulate the cellular production of glucose transporters. Daily insulin injections in the diabetic animals reduced blood glucose close to the normal range measured in the morning; however, the injections did not completely correct the defects in the glucose transport system induced by streptozotocin. Under basal conditions, glucose transport and cytochalasin-j8 binding were normal in the insulin-treated animals. However, maximum insulin stimulation results showed significant reductions in both glucose transport and cytochalasin-/? binding compared to control values. Thus, the daily insulin injections only partially corrected the diabetic state. Further improvements may have been achieved with multiple insulin injections, as used by most diabetic patients. The increases in transport and cytochalasin-/3 binding after insulin treatment are consistent with the data reported by Walker et al. (34) showing that insulin treatment induces an increase in glucose transporter production in L6 cells. Short term studies with streptozotocin diabetic rats have shown that daily insulin injections for 1 week resulted in a hyperresponsive state in fat cells (30, 31, 35). Maximum glucose transport was higher than the control value, but the numbers of glucose transporters in the plasma membrane and microsomal fraction was similar to those in controls. These results indicate that insulin treatment restores the intracellular pool and increases the intrinsic activity of the glucose transporters. Our results with long term (10 weeks) diabetes with insulin treatment showed a reduction in both maximum glucose transport as well as the number of glucose transporters translocated into the sarcolemmal membrane. Whether the difference between our data on skeletal muscle and the data on fat cells is due to a difference in tissues, the duration of diabetes, or some other aspect of experimental design remains to be determined. In summary, the results of our studies show that mild untreated streptozotocin diabetes (45 mg/kg, iv) maintained for 10 weeks resulted in hyperglycemia (>13 mM)
1925
and a reduction in both basal and maximum insulinstimulated glucose transport in skeletal muscle. The reduction in glucose transport was associated with a reduction in the number of glucose transport molecules in the sarcolemmal membrane. Daily insulin treatment reduced morning glucose values (
Vol. 126, No. 4 Printed in U.S.A.
Effects of Streptozotocin-Induced Diabetes on Glucose Transport in Skeletal Muscle* R. JAMES BARNARD, JACK F. YOUNGREN, DOUGLAS S. KARTEL, AND DEBORAH A. MARTIN Department of Kinesiology, University of California, Los Angeles, California 90024-1568
ABSTRACT. Female Sprague-Dawley rats were injected with streptozotocin (45 mg/kg) to induce mild diabetes (glucose, >13 mM). Half of the animals received daily insulin injections to reduce hyperglycemia. After 10 weeks, sarcolemmal membranes were isolated from hindlimb muscles to study glucose transport, and the number of glucose transporters was assessed by cytochalasin-j8 binding. Both glucose transport (19.2 ± 1.6 us. 31.93 ±3.29 pmol/mg protein • 15 sec) and cytochalasin-/? binding (3.06 ± 0.28 vs. 6.14 ± 0.59 pmol/mg protein) were significantly (P < 0.05) reduced in the diabetic untreated rats compared to control values. Daily insulin injections restored both (P < 0.05) basal transport (33.22 ± 3.62 pmol/mg protein • 15 sec) and cytochalasin-/3 binding (5.52 ± 0.66 pmol/mg protein) to control levels. Maximum insulin stimulation (1 U/kg, iv) significantly increased (P < 0.05) both glucose transport (30.18 ± 3.76 vs. 96.48 ± 4.21 pmol/mg protein • 15 sec) and cytochalasin-/? binding (4.38 ± 0.29 vs. 9.40 ± 0.42 pmol/mg protein) in the untreated diabetic
I
N THE past 10 yr, significant progress has been made in understanding the mechanism of insulin action and some of the possible defects associated with insulinresistant states. Most of this research has been performed on fat cells or nontarget cells for insulin action. Insulin clamp studies have demonstrated that skeletal muscle is the most important target tissue for insulin action. De Fronzo (1) has estimated that 70-75% of the blood glucose removed during clamp studies is removed by skeletal muscle. Studies have also documented that the major defect, in insulin action, associated with noninsulin-independent diabetes mellitus is located in skeletal muscle (1). Insulin resistance is also known to develop in insulin-dependent diabetes mellitus. Whether the mechanisms involved in the development of insulin resistance in the two major forms of diabetes mellitus are identical is unknown, but some similarities have been found. Unfortunately, few studies have focused on cellular Received October 2, 1989. Address all correspondence and requests for reprints to: R. James Barnard, Ph.D., Department of Kinesiology, University of California, 405 Hilgard Avenue, Los Angeles, California 90024-1568. * This work was supported by NIH Grant DK-32326 and a grant from the Nathan Pritikin Research Foundation.
and control rats. However, the stimulation in the untreated diabetic rats only reached basal control levels, which was significantly (P < 0.05) below the insulin-stimulated value for the controls. In the rats receiving daily insulin injections, maximum insulin stimulation increased (P < 0.05) both glucose transport (58.67 ± 15.24 pmol/mg protein 15 sec) and cytochalasin-/? binding (6.4 ± 0 . 7 pmol/mg protein), but both transport and binding were significantly (P < 0.05) below insulin-stimulated values for the control rats. These data show that insulin deficiency adversely affected the glucose transport system in skeletal muscle. Both basal and maximum insulin-stimulated transport and the number of transport molecules were reduced. Daily insulin treatment corrected some of the defects, but maximum insulin stimulation was still significantly below values for control animals. (Endocrinology 126: 1921-1926, 1990)
aspects of skeletal muscle glucose transport in insulinresistant states, especially long term resistance. According to Ziel et al. (2), glucose transport is rate limiting for glucose metabolism in both normal and streptozotocindiabetic rats. We have recently developed an isolated skeletal muscle sarcolemmal membrane preparation to specifically study the glucose transport system (3, 4). In the present study we examined the effects of 10 weeks of treated and untreated streptozotocin-induced diabetes on glucose transport using our isolated sarcolemmal vesicle preparation.
Materials and Methods Animals Female Sprague-Dawley rats were purchased from Charles River Laboratories (Wilmington, MA) at about 5 weeks of age. The animals were divided into control, untreated diabetic, and insulin-treated diabetic groups. The diabetic groups were injected with streptozotocin (45 mg/kg, iv) and placed on a dextrose drinking solution for 3 days. On the morning of the fifth day, blood glucose was measured in all rats (One Touch, Lifescan Co., Mountainview, CA). If the morning glucose measurement was not above 13 mM in the diabetic rats, they were removed from the study. All control animals had a morning 1921
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 August 2016. at 02:41 For personal use only. No other uses without permission. . All rights reserved.
1922
DIABETES AND GLUCOSE TRANSPORT
Endo • 1990 Vol 126 • No 4
ton, MA) were added, and the vials were counted in a Beckman LS-7500 liquid scintillation counter (Palo Alto, CA). Filter retention of counts and nonspecific uptake were measured with L-[14C]glucose and subtracted from D-[3H]glucose values to obtain specific D-glucose transport. All measurements were made in triplicate and averaged to obtain a single specific transport value. The number of glucose transport molecules in the sarcolemmal membrane was assessed by the cytochalasin-/3-binding Sarcolemmal glucose transport and cytochalasin-(i binding stud- procedure of Cushman and Wardzala (6), as described in detail previously (4). Sarcolemmal membranes were placed in glass ies culture tubes with cytochalasin-E for 10 min. D- or L-glucose The tracers D-[3H]glucose, L-[14C]glucose, D-[14C]mannitol, was then added and allowed to bind for an additional 10 min. and [3H]cytochalasin-/3 with specific activities of 331. Ci/mol, To each of the five pairs of tubes containing D- or L-glucose, cytochalasin-/? was added in concentrations ranging from 1-24 47.0 mCi/mmol, 48.7 mCi/mmol, and 18.5 Ci/mmol, respecpM and mannitol in concentrations ranging from 0.1-0.4 nM. tively, were purchased from New England Nuclear (Boston, After binding, the membranes were pelleted by centrifugation MA). Cytochalasin-jS was purchased from Aldrich Chemical at 29,000 X g for 30 min. An aliquot of the supernatant was (Milwaukee, WI), and all other chemicals from Sigma Chemical placed in one scintillation vial, and the pellets were dissolved (St. Louis, MO). in NaOH and added to a second vial. Both were then counted The procedure for sarcolemmal isolation was started between in a liquid scintillation counter. All samples were made in 0800-0900 h; the animals were not fasted, and the insulinduplicate, and specific binding was determined by subtracting treated animals had received their regular insulin injection the the percentage of [14C] mannitol bound from the percentage of previous evening. For the maximum insulin stimulation studies, 3 [ H]cytochalasin-j8 bound. The total binding for cytochalasinthe rats were given an iv injection of purified porcine insulin j8 was calculated according to Scatchard (7). (1 U/kg) and killed 10 min later. Sarcolemmal vesicles were All data were placed on a computer for statistical comparison isolated using the methods described in detail previously (3). using SAS (Statistical Analysis Systems, Cary, NC) programs. The quadriceps, gastrocnemius, and plantaris muscles (all fastGroup means were compared using a two-way analysis of varitwitch) from each rat were trimmed, minced, and then homogance, followed by Tukey's test for significant differences. P < enized using a Polytron (Brinkmann Instruments, Westbury, 0.05 was considered statistically significant. Data are expressed NY). The muscles from two rats were combined for each as the mean ± SE. membrane preparation. After homogenization, the homogenate was filtered through wire mesh to remove connective tissues and mixed with KC1 Na pyrophosphate to dissolve contractile Results proteins. After vigorous mixing, the solution was centrifuged at 184,000 X g, and the pellet was resuspended in homogenizing Table 1 presents the clinical data obtained from the medium and incubated for 45 min at 30 C with 20 mg DNAase. rats during the final week of the experiment. The unAfter a low speed spin at 750 X g for 15 min, the supernatant treated diabetic rats had a lower body weight and higher was spun at 184,000 X g for 15 min. The pellet was resuspended morning glucose value than either the control or insulinin 45% sucrose and placed on the bottom of a sucrose gradient. treated group. While the daily insulin injections normalAfter 16 h of centrifugation at 64,000 X g, the sarcolemmal ized body weight in the treated group, blood glucose was vesicles were harvested from the F2 fraction, which had a still significantly higher than control values, but was density of about 1.1 g/ml. The activity of the enzyme potassignificantly lower than that in the untreated diabetic sium-stimulated p-nitrophenylphosphatase (KpNPPase) was group. The daily insulin injections in the treated group used as a sarcolemmal membrane marker; protein was measranged from 4-7 U/day. ured using the Bradford method (3). The percentage of vesicles that formed rightside out was assessed by measuring the Table 2 shows the sarcolemmal isolation data. There KpNPPase activity before and after deoxycholate treatment of was no significant difference in yield of sarcolemmal the vesicles, as described previously (3). After isolation, the protein per g wet wt of muscle among any of the groups. membranes were frozen and stored in liquid nitrogen. In addition, there was no significant difference in the Specific D-glucose transport was measured using the equilibrium exchange method of Ludvigsen and Jarett (5), as described TABLE 1. Final body weight and blood glucose data in detail previously (3, 4). Transport was measured for 15 sec Blood glucose at 37 C in a solution containing 180 nM D- and L-glucose. The Group BW(g) (mM) reaction was stopped by the addition of 3 mM HgCl2, and transported and untransported glucose were separated by ultraControl 290 ± 6 4.0 ± 0.2 filtration using Gelman 0.45-jum filters (product 63068, Ann Untreated diabetic 232 ± 5 a 19.9 ±0.7° Insulin-treated diabetic 296 ± 186 6.8 ± 1.2a>6 Arbor, MI). The filters were transferred to scintillation vials 0 and dissolved in 600 n\ ethylene glycol monomethyl ether. Ten Significantly different from control, P < 0.05. 6 millimeters of Aquasol-2 (DuPont-New England Nuclear, BosSignificantly different from untreated diabetic, P < 0.05.
glucose reading below 5 mM. The insulin-treated group received daily (late afternoon) sc injections of Iletin NPH I insulin (Eli Lilly Co., Indianapolis, IN) in quantities sufficient to reduce morning glucose levels to below 8 mM and as close to 6 mM as possible. The morning glucose level was measured weekly, and the insulin dosage adjusted as needed. Standard rat chow and water were provided ad libitum with the animals on a 12-h light, 12-h dark cycle, starting at 0700 h.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 August 2016. at 02:41 For personal use only. No other uses without permission. . All rights reserved.
DIABETES AND GLUCOSE TRANSPORT
1923
TABLE 2. Effect of untreated and insulin-treated diabetes on sarcolemma membranes Sarcolemma yield (mg protein/g muscle)
Group
KpNPPase h)
Control Control, insulin-stim Untreated diabetic Untreated diabetic, insulin-stim Insulin-treated diabetic Insulin-treated, insulin-stim
0.68 ± 0.08 8.05 ± 2.38 0.72 ± 0.12 6.82 ± 0.84 0.76 ± 0.15 7.74 ± 2.58 0.91 ± 0.15 5.91 ± 1.23 0.77 ± 0.08 7.34 ± 0.46 0.62 ± 0.04 6.25 ± 0.57 Insulin stimulated (insulin-stim) were given 1 U/kg regular porcine insulin 10 min before isolation. Values are the mean ± SE. Each group contained five membrane preparations. No significant differences were found for either the yield (milligrams of protein) or purity (KpNPPase activity).
activity of KpNPPase, the sarcolemma marker enzyme, indicating no significant differences in the purity of the sarcolemma membrane preparation among any of the groups. The percentage of vesicles that formed rightside out was measured in control, untreated diabetic and insulin-treated groups, and values were 56 ± 2, 60 ± 8, and 60 ± 1, respectively (not significantly different). Figure 1 shows the 15-sec specific D-glucose transport data. Under basal conditions, glucose transport was significantly (P < 0.05) decreased in the untreated diabetic group compared to that in controls (19.2 ± 1.6 vs. 31.9 ± 3.3 pmol/mg protein-15 sec). Insulin treatment restored basal transport to control levels (30.2 ± 3.8 pmol/mg protein-15 sec; P < 0.05). Maximum insulin stimulation increased transport 3-fold in the control animals to 96.5 ± 4.2 pmol/mg protein -15 sec (P < 0.01). In the untreated diabetic group, insulin stimulation resulted in only a 1.6-fold increase in glucose transport to 30.2 ± 3.8 EFFECT OF STREP DIABETES ON SARCOLEMMAL VESICLE GLUCOSE TRANSPORT
pmol/mg protein-15 sec (P < 0.05), which was significantly (P < 0.05) lower than the maximum stimulation achieved in the control group. In the insulin-treated group, insulin stimulation resulted in a 1.8-fold increase in glucose transport to 58.8 ± 15.2 pmol/mg protein • 15 sec (P < 0.05), which was still significantly (P < 0.05) below the maximum transport achieved in the control group. Figure 2 shows the cytochalasin-/?-binding sites for both basal and insulin-stimulated conditions. Under basal conditions cytochalasin-/? binding was significantly (P < 0.05) reduced in the untreated diabetic group compared to that in controls (3.06 ± 0.28 us. 6.14 ± 0.53 pmol/mg protein). With maximum insulin stimulation, cytochalasin-/? binding was significantly (P < 0.05) increased in the control group to 9.4 ± 0.72 pmol/mg protein. In the untreated diabetic group maximum insulin stimulation increased cytochalasin-/? binding to 4.38 ± 0.29 pmol/mg protein (P < 0.05), which was signifiEFFECT OF STREP DIABETES ON SARCOLEMMAL VESICLE CYTOCHALASIN-13 BINDING 12 -i
D
•
Basal N = 5
10 -
Insulin Stimulated N = 5
Basal N = 5 Insulin Stimulated N = 5
I
O)
o a.
Q.
O
o. CONTROL
UNTREATED DIABETIC
INSULIN TREATED DIABETIC
FIG. 1. Effect of streptozotocin diabetes on sarcolemmal vesicle glucose transport. Basal transport was reduced in the untreated diabetic group (P < 0.05). Insulin stimulation increased transport in all three groups (P < 0.05); however, the maximum response was significantly different in all three groups (P < 0.05). n = 5 refers to the number of sarcolemmal vesicle preparations studied for each group.
CONTROL
UNTREATED DIABETIC
INSULIN TREATED DIABETIC
FIG. 2. Effect of streptozotocin diabetes on cytochalasin-/3 binding in sarcolemmal vesicles. Basal binding was significantly reduced in the untreated diabetic group (P < 0.05), and all three maximum values were significantly different (P < 0.05). n = 5 refers to the number of sarcolemmal vesicle preparations studied for each group.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 August 2016. at 02:41 For personal use only. No other uses without permission. . All rights reserved.
1924
DIABETES AND GLUCOSE TRANSPORT
cantly (P < 0.05) below the insulin-stimulated value for the controls. In the insulin-treated group, basal cytochalasin-/? binding was 5.52 ± 0.66 pmol/mg protein, which was not significantly different from that in the control group. Maximum insulin stimulation increased (P < 0.05) cytochalasin-/? binding to 6.46 ± 0.76 pmol/ mg protein, which was significantly (P < 0.05) below the maximum stimulation for the control group.
Discussion Skeletal muscle, the most important target tissue for insulin, is adversely affected by 10 weeks of mild streptozotocin diabetes. Our data show that glucose transport, under both basal and insulin-stimulated conditions, is significantly reduced in this untreated diabetic model. Daily insulin injections restored basal glucose transport to control levels; however, maximum insulin stimulation was still significantly reduced compared to that in controls. The reduction in glucose transport in the untreated diabetic rat was due primarily to a reduction in the number of glucose transport molecules in the sarcolemmal membrane, as assessed by cytochalasin-/? binding. The diabetic state, either treated or untreated, did not affect the yield or purity of the isolated sarcolemmal membranes or the percentage of vesicles that formed rightside out. According to De Fronzo (1), skeletal muscle is the major site of insulin resistance associated with diabetes mellitus, especially noninsulin-dependent diabetes mellitus. Insulin resistance also develops with insulin-dependent diabetes mellitus. Whether the mechanisms responsible for insulin resistance in the two major forms of diabetes mellitus are similar remains to be determined. It has been reported that insulin binding is increased in the streptozotocin-treated insulin-deficient rat (8-14), but is decreased in the obese rat (15, 16), both of which develop insulin resistance. This suggests that different mechanisms may be involved. While an extensive body of research has been published using the streptozotocin diabetic rat model, almost all of the studies have been performed on fat cells, with only limited information obtained from skeletal muscle. The use of an isolated sarcolemmal vesicle preparation obtained from skeletal muscle has enabled our laboratory to focus directly on the glucose transport system and the mechanisms of insulin action (3, 4). In the present study basal glucose transport was reduced by 40% after 10 weeks of streptozotocin-induced untreated diabetes. These results are similar to those reported for isolated fat cells (8-10,12, 17). Studies on skeletal muscle, either hindlimb perfusion (18) or in vitro incubated muscle (11, 19), have reported decreases in 3-methylglucose or 2deoxyglucose uptake of 40-70% under basal conditions
Endo • 1990 Vol 126 • No 4
for mild and severe diabetes, respectively. Since insulin is significantly reduced in this untreated diabetic model, it is not surprising to see a decrease in basal glucose transport. Our previous studies (4) have indicated that insulin increases glucose transport in skeletal muscle by both translocation of additional glucose transport molecules into the sarcolemma membrane as well as activation of transporters normally present in the membrane. The data for insulin stimulation in control animals in the present study also support this concept, i.e. transport was increased by 302% while cytochalasin-/? binding was increased by only 154% after insulin stimulation. Another interpretation may be made, however, on the basis of recent data showing different basal vs. insulin-responsive isoforms of glucose transporters (20). The insulinresponsive isoform of glucose transporter may simply transport at a faster rate than the basal form, and thus, translocation could be the sole mechanism involved in insulin action. The data from the untreated diabetic and insulin-treated groups in the present study, however, do not support this concept, in that the insulin-stimulating effect was similar for both transport and cytochalasin binding; i.e. in the untreated diabetic animals transport was increased by 157%, while cytochalasin-/3 binding was increased by 143%, and in the insulin-treated group the values were increased by 195% and 117% for transport and cytochalasin-/? binding, respectively. In the present report cytochalasin-/? binding studies showed a 50% decrease in the number of glucose transport molecules in the sarcolemma membranes of untreated diabetic rats. The difference between the 40% decrease in glucose transport and the 50% decrease in the number of glucose transport molecules is within the error of measurement, but shows that the decreased glucose transport is due mainly to a decrease in available transport molecules. Ramlal et al. (21) also found the number of cytochalasiniS-binding sites to be reduced by approximately 50% in plasma membranes from diabetic rat muscle under basal conditions. These results, collectively, suggest that the main effect of diabetes in the streptozotocin diabetic model is to reduce the number of basal transporters as well as the number of transporters translocated into the plasma membrane after insulin stimulation. In previous studies (3,4) we demonstrated that a single insulin injection of 1 U/kg, iv, resulted in maximum stimulation of the glucose transport system. In the present study maximum insulin-stimulated transport achieved in the untreated diabetic sarcolemma preparations only reached the basal transport level in the control sarcolemma preparations. These results are similar to those reported for fat cells (8-10, 12, 17) and indicate postreceptor binding defects. Decreases in autophosphorylation and the tyrosine kinase activity of the insulin
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 10 August 2016. at 02:41 For personal use only. No other uses without permission. . All rights reserved.
DIABETES AND GLUCOSE TRANSPORT receptor have been reported (22-28) for liver and muscle insulin receptors, as has a decrease in the fat cell intracellular pool of transporters available for translocation (29-31). More recently, Garvey et al. (32) reported a decrease in mRNA for the insulin-sensitive isoform of glucose transporter in muscle from streptozotocin-induced diabetic rats, and Nishimura et al. (33) reported a decrease in the microsomal pool of glucose transporters. Whether these cellular changes in the untreated streptozotocin-induced diabetic rat are the result of hyperglycemia or hypoinsulinemia is not known, especially for long term diabetes as used in the present study. Walker et al. (34) have recently reported short term studies using L6 muscle cells in which they found that both glucose and insulin can regulate the cellular production of glucose transporters. Daily insulin injections in the diabetic animals reduced blood glucose close to the normal range measured in the morning; however, the injections did not completely correct the defects in the glucose transport system induced by streptozotocin. Under basal conditions, glucose transport and cytochalasin-j8 binding were normal in the insulin-treated animals. However, maximum insulin stimulation results showed significant reductions in both glucose transport and cytochalasin-/? binding compared to control values. Thus, the daily insulin injections only partially corrected the diabetic state. Further improvements may have been achieved with multiple insulin injections, as used by most diabetic patients. The increases in transport and cytochalasin-/3 binding after insulin treatment are consistent with the data reported by Walker et al. (34) showing that insulin treatment induces an increase in glucose transporter production in L6 cells. Short term studies with streptozotocin diabetic rats have shown that daily insulin injections for 1 week resulted in a hyperresponsive state in fat cells (30, 31, 35). Maximum glucose transport was higher than the control value, but the numbers of glucose transporters in the plasma membrane and microsomal fraction was similar to those in controls. These results indicate that insulin treatment restores the intracellular pool and increases the intrinsic activity of the glucose transporters. Our results with long term (10 weeks) diabetes with insulin treatment showed a reduction in both maximum glucose transport as well as the number of glucose transporters translocated into the sarcolemmal membrane. Whether the difference between our data on skeletal muscle and the data on fat cells is due to a difference in tissues, the duration of diabetes, or some other aspect of experimental design remains to be determined. In summary, the results of our studies show that mild untreated streptozotocin diabetes (45 mg/kg, iv) maintained for 10 weeks resulted in hyperglycemia (>13 mM)
1925
and a reduction in both basal and maximum insulinstimulated glucose transport in skeletal muscle. The reduction in glucose transport was associated with a reduction in the number of glucose transport molecules in the sarcolemmal membrane. Daily insulin treatment reduced morning glucose values (
