Muscle insulin resistance in uremic humans: glucose transport, glucose transporters, and insulin receptors JACOB E. FRIEDMAN, G. LYNIS DOHM, CHARLES W. JOHN J. CHEN, NANCY LEGGETT-FRAZIER, SAMUEL FRANCIS T. THOMAS, STUART D. LONG, AND JOSE Departments of Medicine, Biochemistry, Surgery, and Obstetrics East Carolina University, Greenville, North Carolina 27858-4354

FRIEDMAN, JACOB E., G. LYNIS DOHM, CHARLES W. ELTON, ADELA ROVIRA, JOHN J. CHEN, NANCY LEGGETT-FRAZIER, SAMUEL M. ATKINSON, JR., FRANCIS T. THOMAS, STUART D. LONG, AND JOSE F. CARO. Muscle insulin resistance in uremic

humans: glucose transport, glucose transporters, and insulin receptors. Am. J. Physiol. 261 (Endocrinol. Metab. 24): E87-

E94, 1991.-To determine the cellular basisfor insulin resistance observedin patients with uremia, we investigated insulin action in vivo and in vitro using skeletal muscleobtained from patients with chronic renal failure. Uremic subjectshad significantly reducedrates of insulin-stimulated glucosedisposal,as determined by a 3-h intravenous glucose tolerance test and using the hyperinsulinemic euglycemic clamp technique. Hepatic glucoseproduction wassimilar before (control, 76.2 t 6.3 vs. uremic, 74.2 t 6.9 mg*kg-’ l rein-‘) and during insulin infusion at 40 mUom-2~min-1(control, -60.9 t 6.6 vs. uremic, -53.9 t 6.3 mg*kg-’ .min-‘). In incubated human skeletal muscle fiber strips, basal 2-deoxy-D-glucose transport was unchanged in uremic subjectscompared with controls. However, the increasein insulin-stimulated glucosetransport was significantly reducedby 50% in musclesfrom uremic patients (P = 0.012). In partially purified insulin receptors prepared from skeletal muscle, 1251-labeled insulin binding, ,&subunit receptor autophosphorylation, and tyrosine kinaseactivity were all unchanged in uremic subjects. The abundance of insulinsensitive (muscle/fat, GLUT-4) glucose transporter protein measuredby Western blot using Mab lF8 or polyclonal antisera was similar in musclesof control and uremic patients. These findings suggestthat the insulin resistanceobservedin skeletal muscle of uremic patients cannot be attributed to defects in insulin receptor function or depletion of the GLUT-4 glucose transporter protein. An alternative step in insulin-dependent activation of the glucosetransport processmay be involved. renal failure; insulin binding; skeletal muscle;GLUT-4

INSULIN RESISTANCE associated with chronic renal failure is a widely recognized clinical abnormality. Uremic patients are characterized by elevated serum insulin levels with normal fasting glucose and by a delayed and diminished hypoglycemic response following insulin administration (20). The presence of insulin resistance in skeletal muscle of uremic patients has been documented by DeFronzo and co-workers (1, II), using the euglycemic insulin clamp combined with femoral venous catheterization, and by Westervelt (42), who reported that uremia blunted the effect of insulin on glucose uptake in the perfused forearm. 0193-1849/91

$1.50

Copyright

ELTON, ADELA ROVIRA, M. ATKINSON, JR., F. CAR0 and Gynecology, School of Medicine,

Previous studies have shown that incubation of normal rat adipose tissue with serum from uremic patients results in an in vitro insulin-resistant state (31, 32). Likewise, adipocytes from uremic rats exhibit normal glucose transport after prolonged incubation in nonuremic media (31), thus supporting the belief that a factor circulating in the serum of uremic patients is responsible for their insulin resistance. The cellular mechanism by which uremia interferes with glucose transport in insulin target tissues is, however, unknown. In chronically uremic rats, insulin binding to its receptor is unaltered in skeletal muscle (8), liver (27), and adipocytes (40). Furthermore, the tyrosine kinase activity of the insulin receptor isolated from adipocytes (40), muscle (8), and liver (27) appears to be unchanged in uremic rats. These results suggest that the defect(s) responsible for the insulinresistant state induced by uremia involves sites distal to the insulin receptor. In an effort to confirm these findings in humans, we characterized the structure and function of the insulin receptor in muscles of uremic patients, including insulin receptor binding, tyrosine kinase activity, and receptor autophosphorylation. It is now recognized that glucose transport in skeletal muscle and other tissues is mediated by a family of glucose transporter isoforms with distinct structure, function, and tissue distribution (15, 17, 22, 23, 33). Multiple mechanisms appear to be involved in the regulation of glucose transporters, including rapid changes in intrinsic activity (18, 26,37) and movement of the transporters from intracellular sites to the cell surface (10, 28, 39, 41). Decreased number and/or translocation of glucose transporters have been demonstrated in various experimental and human models of insulin resistance (3, 18, 24, 26). In the present study, we utilized a specific antibody to the insulin-sensitive glucose transporter (muscle/fat transporter, GLUT-4) (23) to determine whether decreased expression of this transporter protein can explain the insulin resistance in patients with uremia. In addition, we characterized our patient population using in vitro methods (14) to fully document the degree of insulin resistance in skeletal muscle from uremic patients. METHODS

Materials.

(30.2 Ci/mmol),

0 1991 the American

[3-3H]glucose, carrier-free

Physiological

Society

2-deoxy-D-[1,2-3H]glucose Nalz51 (-102 Ci/mmol), and E87

Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 20, 2018. Copyright © 1991 American Physiological Society. All rights reserved.

E88

GLUCOSE

TRANSPORT

[y-“‘P]ATP (tetratriethyl ammonium salt; l,OOO-3,000 Ci/mmol) were all purchased from New England Nuclear, Boston, MA. lz51-labeled goat anti-mouse immunoglobulin G (IgG) and [U-14C]sorbitol (150-250 mCi/ mmol) were obtained from ICN Radiochemicals, Irvine, CA. Sepharose (wheat germ agglutinin-Sepharose) was purchased from Miles Laboratories (Elkhart, IN), and the reagents for polyacrylamide gel electrophoresis (PAGE) were purchased from Bio-Rad Laboratories, Richmond, CA. Human insulin was supplied by Dr. Ronald Chance of Eli Lilly Research Laboratories, Indianapolis, IN. Unless otherwise stated, all other reagents and chemicals were obtained from Sigma Chemical, St. Louis, MO. Monoclonal antibody to the GLUT-4 transporter (lF8) was kindly provided by Dr. P. F. Pilch. An affinity-purified polyclonal antibody (ECU-4) specific for a carboxy-terminal synthetic peptide (12 amino acids) for GLUT-4 (23) was generated in the laboratory of Dr. Lynis Dohm. Subjects. A total of 34 patients with end-stage renal disease of diverse etiology were studied (Table 1). All patients were volunteers undergoing regular hemodialysis or peritoneal dialysis for >3 yr at the time of the study. Nondiabetic uremic patients were studied 24-48 h after the last period of dialysis unless otherwise indicated. Human muscle biopsies of rectus abdominus muscle were obtained from 21 uremic patients undergoing kidney transplant, 17 nonobese nondiabetic patients undergoing elective abdominal hysterectomy, and from 7 nondiabetic nonobese organ donors to serve as controls. Written informed consent was obtained from each patient after they were informed about the nature and potential risks of the study. General anesthesia was induced with a short-acting barbituate and was maintained by fentanyl and a nitrous oxide-oxygen mixture. Only saline was given intravenously before the biopsy. Immediately after exposing the rectus abdominus muscle, a 3 x 2 x l-cm biopsy was obtained. A specially constructed clamp was placed on muscle tissue before it was excised, and then the clamp and muscle biopsy were either frozen immediately in liquid nitrogen or rapidly transported back to the laboratory in oxygenated KrebsHenseleit buffer for incubation as previously reported (14) ktravenous glucose tolerance test. In 20 uremic TABLE

1. Patient characteristics Parameter

Age,

Y*

Sex, M/F BMI, kg/m’ Glucose, mg/dl Insulin, pU/ml 3-h IVGTT, insulin sensitivity Urea nitrogen, mg/dl Creatinine, mg/dl COz, mmol/l K’, meq/l

index

Uremic

Control

39.8t13.2 18116 23.9t3.6 91.4t2.2 16.4k1.7* 2.lt0.3*

35.2t9.6 12/22

84.7t7.4

23.6t3.6 86.9t1.3 lO.lkO.9

3.4kO.4

13.2kl.O

23.3t0.8 4.4k0.2

IN

UREMIA

patients and 23 control subjects, the minimal model for measuring insulin sensitivity was performed using the intravenous glucose tolerance test (IVGTT) described by Bergman (4). All subjects were asked to follow a diet containing 250 g carbohydrate for 3 days before the study. In 11 of the uremic patients and 10 of the control subjects, regular insulin (0.025 U/kg iv) was given at 10 min as recently suggested by Bergman (5). Plasma insulin and glucose data were analyzed using a computer program provided by Dr. Richard Bergman (University of Southern California) to obtain the insulin sensitivity index. Euglycemic insulin clamp. To further characterize the insulin resistance of our population, four uremic patients and six control patients underwent a euglycemic clamp at an insulin infusion rate of 40 mU rnD2. min-l (12). The subjects were admitted to the Outpatient Clinical Research Unit of the East Carolina University School of Medicine at 8:00 A.M. after an overnight fast. Four plasma samples were obtained during a 20-min period for determining the initial blood glucose level. After basal glucose measurements, an intravenous bolus of [3-3H] glucose (30 &i) was administered, followed by continuous infusion of 0.25 &i/min [3-3H]glucose. After a 3-h time period for isotope equilibration, a primed continuous infusion of insulin was started at 40 mu. m-2.min-1 for 2 h. Five minutes after the start of the insulin, a variable 20% glucose infusion was begun to maintain the plasma glucose concentration at approximately the basal glucose level for the next 100 min of hyperinsulinemia. During the insulin clamp the glucose infusion rate was determined by calculating the mean value during the final 30 min. The total amount of glucose metabolized by the whole body (mg rnB2. min-‘) was calculated by adding the rate of endogenous glucose production to the exogenous glucose infusion rate required to maintain euglycemia. Steady-state plasma insulin and glucose levels were calculated from the mean values during the final 30 min. The metabolic clearance rate of insulin was calculated by dividing the continuous insulin infusion rate by the mean increment in plasma insulin concentration above baseline. Hepatic glucose production in the basal state was determined by dividing the tritiated glucose infusion rate (counts/min) by the steady-state plateau of tritiated glucose specific activity achieved during the last hour of the preinsulin infusion control period. After the insulinglucose administration (euglycemic insulin clamp), a non-steady-state condition in glucose specific activity exists. The rate of glucose appearance and rate of glucose disposal were determined using the non-steady-state equation of Steele (38) as modified by Cowan and Hetenyi (9). The rate of hepatic glucose output (mg rnv2. min-‘) was calculated by subtracting the rate of exogenous glucose infusion from the rate of glucose appearance and was determined during the last 40 min of the insulin infusion. Finally, the metabolic clearance rate of glucose (mg ml-l . mD2) was calculated by dividing the rate of glucose disappearance by the mean fasting plasma glucose concentration during the last 30 min of tracer infusion. Partially purified insulin receptor preparation. Partial l

l

l

l

Values are means t SE; n = 34 subjects, except for insulin sensitivity index (Si) = ~10" min-l . (&J ml-‘), parameter derived from minimal model (5), where n = 24 uremic and n = 22 control. M, male; F, female; BMI, body mass index; IVGTT, intravenous glucose tolerance test. * Significantly different from controls, P < 0.05. l

Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 20, 2018. Copyright © 1991 American Physiological Society. All rights reserved.

GLUCOSE

TRANSPORT

purification of the insulin receptor from human muscle was performed according to the procedure described by Burant et al. (6) with minor modifications (7). Recovery of ‘““I-labeled insulin binding activity in partially purified receptor preparations from crude extracts was >85% in both experimental groups. “51-insulin binding. Partially purified insulin receptor from muscle (15 pugprotein) was incubated at 4°C for 16 h in the presence of lo-" M 1251-insulin, with or without increasing concentrations of unlabeled insulin as previously described (6). The specific binding of 1251-insulin was determined at each insulin concentration and used for Scatchard analysis. The nonspecific insulin binding was ~10% of total binding. The ‘““I-insulin binding data were used to equalize the receptor number in all further studies. Cross-linkage of solubilized insulin receptors with 1251insulin. After optimal 12”1-insulin binding in the presence and absence of unlabeled insulin ( low7 M), equal amounts of the ‘“‘I-insulin receptor complex were covalently cross-linked with 0.5 mM disuccinimidyl suberate for 15 min at 4°C by the method of Pilch and Czech (35). PAGE was performed according to the method of Laemmli (30) in a 7.5% gel. The gels were fixed, stained with 0.25% Coomassie blue, dried, and autoradiographed for 24 h at -70°C with Kodak X-Omat film. Tyrosine-specific protein kinase activity and autophosphorylation of insulin receptors. Equal amounts of human muscle insulin receptors from the two groups were incubated in the absence and presence of different concentrations of insulin at 4°C for 16 h. After this incubation, the tyrosine-specific protein kinase activity was determined at room temperature by the method of Grunberger et al. (19), with minor modifications as described previously (7). Autophosphorylation of the insulin receptors was initiated by adding [T-~~P]ATP (20 &i, 5 PM) in the presence of (in mM) 5 MgC12 and 10 MnC12 in 50 N-2-hydroxyethylpiperazine-N’-2-ethanesulfonic acid (HEPES) buffer, pH 7.4. The reaction was stopped by adding an equal volume of 50 mM HEPES buffer containing (in mM) 10 EDTA, 100 NaF, 20 pyrophosphate, and 4 ATP. This mixture was incubated with l:200 dilutions of polyclonal insulin receptor antibody (IRA no. 16) raised in rabbits against rat liver insulin receptors (7). The immunoprecipitate was then suspended in Laemmli buffer containing 50 mM dithiothreitol (DTT), boiled for 5 min, and was subjected to sodium dodecyl sulfate (SDS)-PAGE (slab gel, 7.5%). The ,&subunit of the insulin receptor (-95-kDa band) was localized in the autoradiography and was measured by densitometry. 2-Deoxy-D-glucose transport by muscle fiber strips. Muscle fiber strips weighing 50-80 mg were “teased” from fresh muscle biopsies obtained at the time of surgery and were incubated in 4.0 ml of medium consisting of Krebs-Henseleit buffer, 1% bovine serum albumin, 1.0 mM pyruvate, and with or without 10s7M insulin. Transport was measured in the presence of 5 mM 2-deoxy-Dglucose for 60 min as described previously (14). Analysis of glucose transporter protein. Approximately 200 mg of muscle were thawed and homogenized (1:6 w-t/ vol) using a Polytron homogenizer at a setting of 7 for

IN

UREMIA

E89

15 s in a buffer containing (in mM) 25 HEPES, 4 EDTA, and 25 benzamidine, as well as 1 PM each of leupeptin, pepstatin, and aprotinin. The samples were spun for 1 h at 100,000 g in a Beckman 42.2 Ti rotor at 4”C, and the supernatant was removed. The pellet was resuspended in Triton X-100 to a final concentration of 1%; the samples were incubated for 1.5 h at 4°C with occasional mixing then centrifuged for 35 min in a Beckman 42.2 Ti rotor at 40,000 revolutions/min. For Western analysis, the supernatant containing crude membranes (50 pg protein) was mixed overnight with 50 ~1 Laemmli sample buffer containing 5% DTT and was brought to 100 ,~l volume with buffer containing 25 mM tris( hydroxymethyl)aminomethane (Tris), 0.19 M glycine, and 1% SDS (pH 8.3). Proteins were separated by SDS-PAGE on an 8% resolving gel using Laemmli’s method (30) and were transferred from the gel to an Immobilon membrane by electrotransfer. The membrane was blocked for 2 h with 5% Carnation low-fat instant milk in Tris-buffered saline (TBS) and was followed by incubation in primary antibody (Mab lF8 or polyclonal antisera ECU-4). After 16 h incubation, the membranes were washed alternatively in TBS and TBS-0.05% Tween and were probed overnight with 12”1-goatanti-mouse IgG, or 1251-goatantirabbit IgG. Autoradiography was carried out for 48 h at -7O”C, and’ the resulting autoradiograph was analyzed by densitometry. The results were expressed relative to an internal standard control (rat heart membrane) run on each gel. Statistical analysis. Differences between uremic and control patients were analyzed using one-way analysis of variance with unequal subclass numbers in an SAS program (Statistical Analyses System, Cary, NC). The 0.05 level of probability was accepted as the level of statistical significance. All data are presented as means t SE. RESULTS

As shown in Table 1, the uremic patients were hyperinsulinemic, with fasting glucose levels similar to controls, which is indicative of overall insulin resistance. Similarly, the 3-h IVGTT yielded a significantly reduced insulin sensitivity index (P < 0.05) according to the minimal model in the uremic patients. To establish the site of insulin resistance, a hyperinsulinemic euglycemic clamp was performed in a sample of our population (Table 2). Hepatic glucose output during euglycemia was similar in both groups under basal conditions and during the insulin infusion. The insulin infusion rate of 40 mU~m-2~min-1 resulted in a higher steady-state insulin level in the uremic patients than in control patients (P c 0.05). Despite higher steady-state insulin levels, however, the rate of whole body glucose metabolism was significantly less in the uremic patients (P < 0.05), indicating peripheral insulin resistance. Because skeletal muscle is the major site of insulinstimulated glucose disposal, the influence of the uremic condition on skeletal muscle glucose transport was determined in vitro using muscle biopsies obtained from patients at the time of surgery. The rate of 2-deoxy-Dglucose uptake in muscle fiber strips incubated under

Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 20, 2018. Copyright © 1991 American Physiological Society. All rights reserved.

E90

GLUCOSE

TRANSPORT

IN UREMIA

2. Summary of glucose metabolism, h.epatic glucose production, and plasma insulin response during steady-state glucose clamp

TABLE

Group

n

Amount of Glucose Metabolized, mg mm’.min-’

Hepatic Glucose Production, mg. kg-‘. mm’ Basal

Uremic 4 245.2f16.6* 74.2k6.9 76.2zk6.3 Control 6 347.6k16.0 Values are means t SE. * Significantly different from controls, P < 0.05.

basal conditions (no insulin) was similar in uremic patients and control subjects (control, 26.0 + 3.1; uremic, 27.7 + 4.9 nm +g dry muscle-l. min-‘). However, maximal insulin-stimulated glucose transport above basal was significantly reduced by 50% in muscle from uremic subjects (P = 0.012; Fig. 1). Among the uremic patients, timing of hemodialysis was found to affect the degree of insulin resistance. Insulin stimulation over basal was highest in two patients who underwent dialysis only 12 h before surgery (26 + 2 nmol. g dry muscle-l +min-l) and nearly absent in one patient dialyzed 6 days before surgery (2 nmol . g dry muscle-l. min-‘) . To investigate whether the insulin resistance observed in uremic patients was due to a reduction in glucose transporters, we determined the abundance of GLUT-4 glucose transporter protein in muscles of uremic patients by Western blot analysis (Fig. 2). With the use of identical protein concentrations, the abundance of transporter protein (45kDa band) was quantitated by densitometry and revealed no significant differences between uremic patients and control subjects. This experiment was repeated using antiserum lF8 with similar results. The amount of protein recovered per gram of starting tissue was the same in control and uremic muscle (control, 987 f 150; uremic 1,000 rfr 198 pg protein/g muscle). To determine whether alterations in insulin-signaling mechanisms can explain the decreased insulin-stimulated glucose transport in the presence of normal levels of glucose transporters, we studied insulin receptor struc0

CONTROL

m

UREMIC

-L x

(N = 8)

iI (N = 11)

FIG. 1. Effect of insulin (lo-? M) on 2-deoxy-D-glucose transport (stimulated basal) in incubated human muscle from control (n = 8) and uremic (n = 11) patients. *Significantly less than controls, P = 0.012.

Postinsulin

Steady-State Plasma Insulin Concentration, a/m1

-60.9k6.6 -53.9k6.3

114.8+7.8* 88.6k5.8

0

CONTROL

m

UREMIC

Steady-State Plasma Glucose Concentration, w/d1 91.2+1.5 93.6k1.6

T

FIG. 2. Quantitation of glucose transporter protein GLUT-4 in skeletal muscle from control and uremic patients. Protein (50 fig) from muscle of control, uremic, and 50 pg rat heart membrane (standard) was electrophoresed using SDS-polyacrylamide gel electrophoresis (PAGE). Proteins were transferred to Immobilon membrane and immunoblotted with polyclonal antiserum ECU-4 as described in METHODS. The 45-kDa band from autoradiogram was scanned, and results were expressed relative to rat heart membrane standard applied to gel. Results are mean + SE from 7 control and 7 uremic patients run on 2 separate Western blots.

ture and function in these patients. Because of the limitation of sample size, the receptor studies and the glucose transporter studies were carried out on muscle from different patients but with identical degrees of uremia. Scatchard analysis of 1251-insulin binding from partially purified muscle insulin receptors (Fig. 3) indicates that the specific 1251-insulin binding at tracer concentration (10-l’ M) is lower but not significantly different in uremic patients compared with control subjects (9.92 k 1.06 and 12.01 + 1.99 ng.ml-‘. 15 pg-‘, respectively). There was no difference in the lz51-insulin binding between groups at any insulin concentration tested. To further characterize the insulin receptor, we determined the electrophoretic mobility of the (Y- and /3subunits with the use of affinity labeling and autophosphorylation techniques (Fig. 4). Partially purified insulin receptors were incubated with 1251-insulin and disuccinimidyl suberate under reducing conditions. A protein with a molecular mass of -125 kDa is labeled (lanes 4 and 9), corresponding to the molecular mass of the CYsubunit of the insulin receptor, and its radioactivity was totally displaced by an excess of unlabeled insulin (lanes 5 and lo), supporting the specificity of insulin binding to this band. The electrophoretic migration of the crosslinked a-subunit of the insulin receptor was similar in uremic patients and control subjects. In these cross-

Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 20, 2018. Copyright © 1991 American Physiological Society. All rights reserved.

GLUCOSE

TRANSPORT

g 0.10 L s c 2

al 0.05

0.00 0

2

1

Bound Insulin ( ng/ 15pg Protein) FIG. 3. Insulin binding to solubilized insulin receptors from skeletal muscle. Wheat germ agglutinin (WGA)-purified solubilized insulin receptors (-10 pg protein) were incubated with ‘251-labeled insulin (1 x 10-i’ M) with increasing concentrations of unlabeled insulin (lo-’ to ’ 10m6M) in a total volumeof 250 ~1 of 50 mM HEPES buffer, pH 7.4, containing 0.1% Triton X-100, and incubation with polyethylene glycol (12.5% final concentration) using bovine y-globulin as carrier protein. Scatchard plots are of specific insulin binding data (means f SE) of 6 separate experiments from uremic and control patients.

linking experiments, we used the same conditions as those for ‘251-insulin binding (Fig. 3) with regard to incubation time, temperature, and amount of protein (15 pg) chromatographed. In uremic patients and control subjects, incubation of receptors with insulin increased the incorporation of 32P from [T-~*P]ATP into a protein of molecular mass between 90 and 95 kDa (Fig. 4, lanes 2,3, 7, B), which specifically was immunoprecipitated by insulin receptor antibodies, thus corresponding to the /3-

Control

E91

IN UREMIA

subunit of the insulin receptor. No change in the electrophoretical mobility of this protein was seen in uremic muscles (n = 7) compared with control muscles. Densitometry of the autoradiograms showed no differences in the radiolabeled bands for the nonstimulated and insulin-stimulated receptor in control subjects (lanes 13) vs. uremic patients (lanes 6-8). Because it has been demonstrated that insulin receptor kinase activity toward exogenous substrates may be different from that observed with autophosphorylation (16), we also measured receptor kinase activity using the exogenous substrate polyglutamyltyrosine (4:l) [poly(Glu4Tyr’)] as the phosphoacceptor (Fig. 5). In the absence of insulin, equal amounts of receptors from the two groups phosphorylated poly(Glu4Tyr1) to a similar degree (912 f 102 vs. 999 + 67 counts/min for uremic patients and control subjects, respectively). Exogenous substrate phosphorylation was similar between uremic patients and control subjects at submaximal (lo-’ M) and maximal (10m7 M) insulin concentrations. DISCUSSION

Numerous studies both in humans and animals have demonstrated that uremia produces an insulin-resistant state (reviewed in Ref. 11). With the use of the minimalmodel method to determine insulin sensitivity (5), our data confirm that uremic patients are severely resistant to insulin. To determine the site of the insulin resistance, we performed a euglycemic clamp in our patients. From our own in vivo data and those of others (1,ll) it is clear that the resistance to insulin is not associated with excess

Uremic

FIG. 4. Representative autoradiograph of SDS-PAGE of solubilized insulin receptors from skeletal muscle. ‘251-insulin binding was performed in WGApurified solubilized receptor, as described in Fig. 3, in presence and absence of 1 x 10m6 M unlabeled insulin. ‘?-insulin receptor complex was cross-linked with 0.5m M dissuccinimidyl suberate, reduced with 50 mM dithiothreitol, and subjected to SDS-PAGE in 7.5% acrylamide resolving gel. Fixed, stained, and dried gels were autoradiographed on Kodak X-OMAT film with intensifying screen at -70°C for 24 h. Autoradiography was quantitated by densitometry. Autophosphorylation of the &subunit in receptors from uremic patients was 124 + 25% of control at 10e9 M insulin, and 107 * 7.7% of control at lo-’ M insulin

a Subunit P Subunit

(n = 5).

Insulin (M) 1251nsulin (MI Lane

0

-9

10

lo7

0 -10

10 M6 0

IO9

,07

0

lo6

-10

0

0

0

10

10

0

0

0

IO’”

1

2

3

4

5

6

7

8

9

IO’O 10

Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 20, 2018. Copyright © 1991 American Physiological Society. All rights reserved.

E92

GLUCOSE

0

TRANSPORT

0 Controls 0 Uremics

kg Insulin

&A)

5. Polyglutamyltyrosine(4:l) [poly(Glu4Tyr’)] phosphorylation by skeletal muscle-derived insulin receptors. WGA-purified solubilized insulin receptors from uremic patients and control subjects were incubated in presence and absence of different concentrations of insulin, as described in Fig. 4. ATP (100 pmol, 20 &i/vial) was then added in presence of 2.5 mg/ml poly(Glu4Tyr’) and 10 mM MgC12 in 50 mM HEPES. After 30 min at room temperature, reaction was stopped with 10% trichloroacetic acid containing 10 mM pyrophosphate and 3% bovine serum albumin. Data are means k SE from 6 separate experiments. FIG.

hepatic glucose production or with a defect in insulin secretion. Rather, the insulin resistance is due to a defect in peripheral glucose disposal. It should be noted that, at lower insulin levels, recent studies have found impaired inhibitory effects of insulin on hepatic glucose production (36) in uremia. Nevertheless, Alvestrand et al. and DeFronzo et al. (1,ll .) have shown that, in uremia, insulin-stimulated glucose uptake across the leg is reduced and corresponds to the decrease observed in whole body glucose disposal. From these data it appears likely that skeletal muscle is a major site of impaired insulin action in patients with uremia. Although in vivo techniques, such as the euglycemic clamp, have been extremely important in defining the changes that occur in insulin-resistant states, further study of the cellular mechanisms responsible for insulin resistance in patients with uremia has been very limited. Recently, we developed a novel in vitro human muscle preparation well suited for investigating the mechanism of insulin resistance ( 14). We utilized our incubated human muscle fiber strip preparati .on to tes t whether, in the absence of circulating factors, skele tal muscle from uremic patients is insulin resistant. We found no change in basal glucose transport rates between control subjects and uremic patients, but we found a significantly decreased rate of maximal insulin-stimulated glucose transport in muscles of uremic patients compared with control subjects. These data provide direct evidence that insulin responsiveness is impaired in the skeletal muscle of uremic patients and suggests that uremia interferes with insulin action, perhaps at the level of the glucose transport system. It is now well established that transport of glucose by insulin-sensitive tissues is mediated, at least partially, by recruitment of glucose transporters from intracellular sites to the cell surface, (10,28,39,41) and by alterations in intrinsic activity of the transporter (18, 22, 37). More recently, a novel insulin-sensitive glucose transporter

IN

UREMIA

isoform (GLUT-4), unique to muscle and adipose tissue, has been identified using a monoclonal antibody, IF8 (23). In an attempt to identify the primary cellular defect in uremia, we examined the abundance of GLUT-4 transporter protein in muscles of uremic patients. We found no significant difference in the abundance of GLUT-4 in muscles of patients with uremia (Fig. 3). These results differ from those of Jacobs et al. (21), who used the cytochalasin B assay to study glucose transporters in adipocytes from uremic rats. They found reduced translocation of fat cell glucose transporters in vitro, due in part to a 17% decrease in total transporter concentration in membranes isolated from uremic rats. The reasons for this difference may be due to the experimental model used as well as to methodological differences. The membrane isolation technique recovers only a fraction of the glucose transporters based on marker enzymes, and, because cytochalasin-B does not distinguish between the transporter isoforms, it cannot be determined for certain which isoform is affected. However, regardless of these differences, because the decrease in insulinstimulated glucose transport is greater than could be explained by any difference in glucose transporters, it suggests that there may be a functional defect, i.e., decreased intrinsic activity or reduced translocation of glucose transporters in uremic tissue. Reduced glucose transport activity in the absence of changes in glucose transporter number has been reported previously in several studies of insulin resistance. For example, Karnieli et al. (26) and Kahn et al. (25) reported that reversal of diabetes in the insulin-treated diabetic rat is associated with a threefold increase in glucose transport, accompanied by just a minor increase in total transporter levels. Likewise, it was recently reported that, despite significant insulin resistance in db/db obese mice (29) and in type II diabetics (34), there was little change in the level of GLUT-4 glucose transporter protein in skeletal muscle. Thus it is becoming increasingly accepted that the mechanism of glucose transport is not exclusively regulated by the number of glucose transporters. Other factors such as reduced translocation, changes in intrinsic activity, or altered signal transduction may be important in inducing insulin resistance in uremia. Recent reports suggest that phosphorylation of the glucose transporter may modulate changes in translocation and intrinsic activity (22). Uremic muscle is characterized by a number of abnormalities in energy metabolism, particularly a decrease in both ATP and phosphocreatine content (13), which could theoretically interfere with the activation of ATP-consuming processes. A decrease in insulin-stimulated glucose transport could be associated with a defect in the insulin-signaling pathway. However, we found no differences in insulin binding to solubilized insulin receptors from muscle of uremic patients. Furthermore, no alteration in the asubunit of the insulin receptor was seen with regard to its ability to bind insulin and in its molecular mass. The interaction of insulin with its receptor in the plasma membrane is known to trigger both phosphorylation and dephosphorylation of several cellular proteins. Because insulin binding induces ,&subunit phosphorylation of the insulin receptor and because the insulin receptor itself is

Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 20, 2018. Copyright © 1991 American Physiological Society. All rights reserved.

GLUCOSE

TRANSPORT

a tyrosine-specific protein kinase, it has been hypothesized that abnormalities of this function could explain insulin resistance. We found that tyrosine kinase activity of the insulin receptor toward the receptor itself and an artific ial exogenous substrate were both normal i n muscle from uremic patients. These findings must be i.nterpreted with caution, however, since there may be fundamental differences in these processes if they were examined in vitro and in vivo (43). Our findings are in agreement with our previous results in muscle from chronic uremic rats (8) and with a recent report by Bak et al. (Z), which appeared while this manuscript was in preparation, in human muscle from uremic subjects. Thus, since insulin binding and kinase activity are not decreased in uremia, it appears that defects following the interaction of insulin with its receptor are responsible for the insulin resistance. In summary, our data indicate that reduced insulinstimulated glucose transport in skeletal muscle of uremic patients is accompanied by normal insulin receptor binding and kinase activation. Furthermore, our results indicate that the mechanism responsible for insulin resistance in uremia does not involve a reduction in GLUT-4 glucose transporter protein. Further studies are needed to determine whether defects in transporter activation or translocation can explain the insulin resistance in uremia.

IN

8.

9.

10.

11.

12.

13.

14.

15.

16. We sincerely thank the volunteers who agreed to participate in this study. We thank Karen Parker of the ECU Transplant Center for help in compiling the patient data and Chenoa Jones for performing the insulin assays. This work was supported by National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) Grants 5ROl-DK-32585-05 and POl-DK-3629601. J. E. Friedman is supported by NIDDK National Research Service Award 1 F32 DK-08477-01. Address for reprint requests: J. F. Caro, Section of Endocrinology, Dept. of Medicine, School of Medicine, East Carolina Univ., Greenville, NC 27858-4354. Received

6 September

1990; accepted

in final

form

24 January

1991.

17.

18.

19.

REFERENCES 1. ALVESTRAND, A., J. WAHREN, D. SMITH, AND R. A. DEFRONZO. Insulin-mediated potassium uptake is normal in uremic and healthy subjects. Am. J. Physiol. 246 (Endocrinol. Metab. 9): E174E180, 1984. 2. BAK, J. F., 0. SCHMITZ, S. S. SORENSEN, J. FROKJAER, T. KJAER, AND 0. PEDERSON. Activity of insulin receptor kinase and glycogen synthase in skeletal muscle from patients with chronic renal failure. Acta Endocrinol. 121: 744-750, 1989. 3. BERGER, J., C. BISWAS, P. P. VICARIO, H. V. STROUT, R. SAPERSTEIN, AND P. F. PILCH. Decreased expression of the insulinresponsive glucose transporter in diabetes and fasting. Nature Lond. 340: 70-74,1989. 4. BERGMAN, R. N. Toward physiological understanding of glucose tolerance. Minimal model approach. Diabetes 38: 1512-1521,1989. 5. BERGMAN, R. N., J. C. BEARD, AND M. CHEN. The minimal modeling method. Assessment of insulin sensitivity and B-cell function in vivo. In: Methods in Diabetes Research. Clinical Methods, edited by W. L. Clarke, J. Larner, and S. L. Pohl. Baltimore, MD: Wiley, 1986, vol. III, p. 15-34. 6. BURANT, C. F., M. K. TREUTELAAR, AND M. G. BUSE. In vitro and in vivo activation of the insulin receptor kinase in control and denervated skeletal muscle. Diabetes 33: 704-708, 1986. 7. CARO, J. F., 0. ITTOOP, W. J. PORIES, D. MEELHEIM, E. G. FLICKINGER, F. THOMAS, M. JENQUIN, J. F. SILVERMAN, P. G. KHAZANIE, AND M. K. SINHA. Studies on the mechanism of insulin -- - _-----

20. 21.

22.

23.

24.

25.

26.

27. - _-

UREMIA

E93

resistance in the liver from humans with non-insulin dependant diabetes: insulin action and binding in isolated hepatocytes, insulin receptor structure and kinase activity. J. CZin. Invest. 78: 249-258, 1986. CECCHIN, F., 0. ITTOOP, M. K. SINHA, AND J. F. CARO. Insulin resistance in uremia: insulin receptor kinase activity in liver and muscle from chronic uremic rats. Am. J. PhysioZ. 254 (Endocrinol. Metab. 17): E394-E401, 1988. COWAN, J. S., AND G. HETENYI. Glucoregulatory responses in normal and diabetic dogs recorded by a new tracer method. Metab. CZin. Exp. 20: 360-372, 1971. CUSHMAN, S. W., AND L. J. WARDZALA. Potential mechanism of insulin action on glucose transport in the isolated rat adipose cell. J. Biol. Chem. 259: 4758-4762,198O. DEFRONZO, R. A., R. ANDRES, P. EDGAR, AND W. G. WALKER. Carbohydrate metabolism in uremia: a review. Medicine BaZtimore 52: 469-481, 1973. DEFRONZO, R. A., J. D. TOBIN, AND R. ANDRES. Glucose clamp technique: a method for quantifyng insulin secretion and resistance. Am. J. Physiol. 237 (Endocrinol. Metab. Gastrointest. Physiol. 6): E214-E223,1979. DEL CANALE, S., E. FIACCADORI, N. RONDA, K. SODERLUND, C. ANTONUCCI, AND A. GUARIGLIA. Muscle energy metabolism in uremia. Metab. CZin. Exp. 35: 981-983, 1986. DOHM, G. L., E. B. TAPSCOTT, W. J. PORIES, D. J. DABBS, E. G. FLICKINGER, D. MEELHEIM, T. FUSHIKI, S. M. ATKINSON, C. W. ELTON, AND J. F. CARO. An in-vitro human muscle preparation suitable for metabolic studies. Decreased insulin stimulation of glucose transport in muscle from morbidly obese and diabetic subjects. J. CZin. Invest. 82: 486-494, 1988. FLIER, J. S., M. M. MUECKLER, A. L. MCCALL, AND H. F. LODISH. Distribution of glucose transporter messenger RNA transcripts in tissues of rat and man. J. CZin. Invest. 79: 657-661, 1987. FREIDENBERG, G. R., R. R. HENRY, H. H. KLEIN, D. R. REICHERT, AND J. F. OLEFSKY. Decreased kinase activity of insulin receptors from adipocytes of non-insulin dependent diabetic (NIDDM) subjects. J. CZin. Invest. 79: 240-250, 1987. FUKUMOTO, H., S. SEINO, H. IMURA, Y. SEINO, R. L. EDDY, Y. FUKUSHIMA, M. G. BYERS, T. B. SHOWS, AND G. I. BELL. Sequence, tissue distribution and chromosomal localization of mRNA encoding a human glucose transporter-like protein. Proc. NatZ. Acad. Sci. USA 85: 5434-5438, 1988. GARVEY, T., T. P. HUECKSTEADT, S. MATTHAEI, AND J. M. OLEFSKY. Role of glucose transporters in cellular insulin resistance of type II noninsulin-dependent diabetes mellitus. J. CZin. Invest. 81: 1528-1536, 1988. GRUNBERGER, G., Y. ZICK, J. ROTH, AND P. GORDON. Protein kinase activity of the insulin receptor in human circulating and cultured mononuclear cells. Biochem. Biophys. Res. Commun. 115: 560-566, 1983. HORTON, E. S., C. JOHNSON, AND H. E. LEBOVITZ. Carbohydrate metabolism in uremia. Ann. Intern. Med. 68: 63-74, 1968. JACOBS, D. B., G. R. HAYES, J. A. TRUGLIA, AND D. H. LOCKWOOD. Alterations of glucose transporter systems in insulin resistant uremic rats. Am. J. Physiol. 257 (Endocrinol. Metab. 20): E193E197,1989. JAMES, D. E., J. F. HIKEN, AND J. C. LAWRENCE. Isoproterenol stimulates phosphorylation of the insulin-regulatable glucose transporter in rat adipocytes. Proc. Natl. Acad. Sci. USA 86: 83688372, 1989. JAMES, D. E., M. STRUBE, AND M. MUECKLER. Molecular cloning and characterization of an insulin-regulatable glucose transporter. Nature Lond. 338: 83-87,1989. KAHN, B. B., M. J. CHARRON, H. F. LODISH, S. W. CUSHMAN, AND J. F. FLIER. Differential regulation of two glucose transporters in adipose cells from diabetic and insulin-treated diabetic rats. J. Clin. Invest. 84: 404-411, 1989. KAHN, B. B., AND S. W. CUSHMAN. Mechanism for markedly hyperresponsive insulin stimulated glucose transport in adipose cells from insulin treated streptozotocin diabetic rats. J. BioZ. Chem. 262: 5118-5224,1987. KARNIELI, E., M. ARMONI, P. COHEN, Y. KANTNER, AND R. RAFAELOFF. Reversal of insulin resistance in diabetic rat adipocytes by insulin therapy. Diabetes 36: 925-931, 1987. KAUFMAN, J. M., AND J. F. CARO. Insulin resistance in uremia:

Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 20, 2018. Copyright © 1991 American Physiological Society. All rights reserved.

E94

28.

29.

30. 31.

32.

33.

34.

35.

GLUCOSE

TRANSPORT

characterization of insulin action, binding, and processing in isolated hepatocytes from chronic uremic rats. J. Ctin. Invest. 71: 69%708,1983. KLIP, A., T. RAMLAL, D. YOUNG, AND J. 0. HOLLOSZY. Insulin induced translocation of glucose transporters in rat hind limb muscles. FEBS Lett. 224: 224-229, 1987. KORANYI, L., D. JAMES, M. MUECKLER, AND M. A. PERMUTT. Glucose transporter levels in spontaneously obese (db/db) insulin resistant mice. J. Clin. Inuest. 85: 962-967, 1990. LAEMMLI, U. K. Cleavage of structural protein during the assembly of the head of bacteriophage T4. Nature Lond. 277: 680-685,197O. MALOFF, B. L., M. L. MCCALEB, AND D. H. LOCKWOOD. Cellular basis of insulin resistance in chronic uremia. Am. J. Physiol. 245 (Endocrinol. Metab. 8): E178-E184, 1983. MCCALEB, M. L., R. MEVORACH, R. B. FREEMAN, M. S. Izzo, AND D. H. LOCKWOOD. Induction of insulin resistance in normal adipose tissue by uremic human serum. Kidney Int. 25: 416-421,1984. MUECKLER, M., C. CARUSO, S. A. BALDWIN, M. PANICO, I. BLENCH, H. R. MORRIS, J. W. ALLARD, G. E. LIENHARD, AND H. F. LODISH. Sequence and structure of a human glucose transporter. Science Wash. DC 229: 941-945,1985. PEDERSEN, O., J. F. BAK, P. H. ANDERSEN, S. LUND, D. E. MOLLER, J. S. FLIER, AND B. B. KAHN. Evidence against altered expression of Glut-l or Glut-4 in skeletal muscle of patients with obesity and NIDDM. Diabetes 39: 865-870, 1990. PILCH, P. F., AND M. P. CZECH. The subunit structure of the high affinity insulin receptor: evidence for a disulfide linked receptor

IN

36.

37.

38. 39.

40.

41.

42. 43.

UREMIA

complex in fat cell and liver plasma membrane. J. Biol. Chem. 255: 1722-1731,198O. SCHMITZ, 0. Peripheral and hepatic resistance to insulin and hepatic resistance of glucagon in uremic subjects: studies at physiologic and supraphysiologic hormone levels. Actu EndocrinoZ. 118: 125-134,1988. SIMPSON, I. A., AND S. W. CUSHMAN. Hormonal regulation of mammalian glucose transport. Anna Rev. Biochem. 55: 1059-1089, i986. STEELE, R. Influence of glucose loading and of injected insulin on hepatic glucose output. Ann. NY Acad. Sci. 82: 420-430, 1959. SUZUKI, K., AND T. KONO. Evidence that insulin causes translocation of glucose transport activity to the plasma membrane from an intracellular storage site. Proc. Nutl. Acud. Sci. USA 77: 25422545,198O. TRUGLIA, J. A., G. R. HAYES, AND D. H. LOCKWOOD. Intact adipocyte insulin receptor phosphorylation and in vitro tyrosine kinase activity in animal models of insulin resistance. Diabetes 37: 147-153,1988. WARDZALA, L. J., AND B. JEANRENAUD. Potential mechansim of insulin action on glucose transport in the isolated rat diaphragm. J. Biol. Chem. 256: 7090-7093,198l. WESTERVELT, F. B. Abnormal carbohydrate metabolism in uremia. Am. J. Clin. Nutr. 21: 423-425, 1968. WHITE, M. F., S. TAKAYAMA, AND C. R. KAHN. Differences in the sites of phosphorylation of the insulin receptor in vivo and in vitro J. Biol. Chem. 260: 9470-9478,1985.

Downloaded from www.physiology.org/journal/ajpendo by ${individualUser.givenNames} ${individualUser.surname} (130.070.008.131) on October 20, 2018. Copyright © 1991 American Physiological Society. All rights reserved.