Hypoxia Stimulates Glucose Transport in
Insulin-Resistant Human Skeletal Muscle
John L. Azevedo, Jr., Julie O. Carey, Walter J. Pories, Patricia G. Morris, and G. Lynis Dohm
causes translocation of GLUT4 to the plasma membrane, as
does insulin, although no decrease in GLUT4 was observed
from the intracellular membrane fraction when muscle was
stimulated by contraction (8).
It has been demonstrated that hypoxia stimulates glucose
transport in muscle, as does exercise or contraction (5).
There is indirect evidence that hypoxia acts through the
contraction signaling pathway to stimulate glucose transport. Hypoxia and contraction stimulation are additive with
maximal insulin stimulation; however, hypoxia and contraction are not additive with one another. Contraction-stimulated glucose transport is hypothesized to be mediated by
Ca2+. Calcium is released from the sarcoplasmic reticulum
upon excitation-contraction coupling, which stimulates
translocation of GLUT4 to the plasma membrane, thus
increasing glucose transport. It is also thought that hypoxia
causes the release of Ca2+ from the sarcoplasmic reticulum
(5). Dantrolene, a blocker of Ca2+ release from the sarcoplasmic reticulum, also blocks hypoxia-stimulated glucose
transport in incubated skeletal muscle.
Because contraction and hypoxia may stimulate glucose
transport through a signaling pathway distinct from insulin,
the purpose of the present study was to assess the function
of the glucose transport effector system by stimulating
glucose transport with hypoxia in the human muscle incubation system previously described (9). If a defect in the
insulin signaling pathway causes the decreased glucose
transport observed in muscle of obese and NIDDM patients,
uscle is responsible for the majority of insulin- we reasoned that the contraction/hypoxia signaling pathway
stimulated whole-body glucose disposal in hu- should produce a normal response if the glucose transport
mans (1). Decreased glucose transport in effector system (translocation and activation of GLUT4) is
muscle from obese and obese non-insulin-de- normal.
pendent diabetes mellitus (NIDDM) patients is well documented; however, the underlying defect for this difference
has not been elucidated. The lack of transduction of the RESEARCH DESIGN AND METHODS
insulin signal is a possible locus for a defect in glucose The experimental protocol was approved by the East Carolina Univertransport associated with obesity and NIDDM (2). Caro et al. sity Policy and Review Committee on Human Research, and informed
was obtained from all patients. Patients were categorized as
(3) reported that in obese patients with and without NIDDM, consent
lean, obese, or obese NIDDM according to body mass index (BMI)
insulin-receptor tyrosine kinase activity in muscle was de- (weight in kg/[height in m]2) and the National Diabetes Data Group
criteria (10). Muscle biopsies were obtained from obese and obese
creased.
In addition to stimulation by insulin, muscle has the ability NIDDM patients undergoing gastric bypass surgery, while biopsies from
and some obese patients were obtained from elective abdominal
to increase glucose transport in response to contraction lean
surgical procedures, primarily hysterectomies. All patients were women.
independent of insulin (4-7). Furthermore, contraction Blood sampling. Blood samples for glucose and insulin analysis were
Insulin and muscle contraction stimulate glucose transport into muscle cells by separate signaling pathways, and
hypoxia has been shown to operate via the contraction
signaling pathway. To elucidate the mechanism of insulin
resistance in human skeletal muscle, strips of rectus
abdominis muscle from lean (body mass index [BMI]
<25), obese (BMI >30), and obese non-insulin-dependent diabetes mellitus (NIDDM) (BMI >30) patients
were incubated under basal and insulin-, hypoxia-, and
hypoxia + insulin-stimulated conditions. Insulin significantly stimulated 2-deoxyglucose transport approximately twofold in muscle from lean (P < 0.05) patients,
but not in muscle from obese or obese NIDDM patients.
Furthermore, maximally insulin-stimulated transport
rates in muscle from obese and diabetic patients were
significantly lower than rates in muscle from lean patients (P < 0.05). Hypoxia significantly stimulated glucose transport in muscle from lean and obese patients.
There were no significant differences in hypoxia-stimulated glucose transport rates among lean, obese, and
obese NIDDM groups. Hypoxia + insulin significantly
stimulated glucose transport in lean, obese, and diabetic
muscle. The results of the present study suggest that the
glucose transport effector system is intact in diabetic
human muscle when stimulated by hypoxia. Diabetes
44:695-698, 1995
M
From the Departments of Biochemistry (J.L.A., J.O.C., G.L.D.) and Surgery (W.J.P.,
P.G.M.), School of Medicine, East Carolina University, Greenville, North Carolina
Address correspondence and reprint requests to Dr. John L. Azevedo, Jr.,
Department of Biochemistry, School of Medicine, East Carolina University,
Greenville, NC 27858.
Received for publication 20 October 1994 and accepted in revised form 1 February
1995.
ANOVA, analysis of variance; BMI, body mass index; 2-DOG, 2-deoxyglucose;
KHB, Krebs-Henseleit buffer; NIDDM, non-insulin-dependent diabetes mellitus.
DIABETES, VOL. 44, JUNE 1995
taken intraoperatively.
Muscle biopsy. Surgery was performed on the patients after an
overnight fast. General anesthesia was initiated with a short-acting
barbiturate and maintained with a fentanyl and nitrous oxide-oxygen
mixture. A 3 x 2 X 0.5 cm muscle biopsy was obtained from the rectus
abdominis muscle after entry into the abdominal cavity.
Muscle strip incubation. Immediately after removal, a given muscle
sample was placed in an airtight container with oxygenated KrebsHenseleit buffer (KHB) for transport to the laboratory. Muscle strips
were incubated in a modified incubation system described earlier (9).
695
HYPOXIA STIMULATES GLUCOSE TRANSPORT
TABLE 1
Patient profile
Lean
Age (years)
BMI (kg/m2)
Glucose (mmol/1)
Insulin (pmol/1)
37.0 ±
22.5 ±
4.7 ±
16.2 ±
1.4 (11)
0.7 (11)
0.2 (9)
4.2 (9)
Obese
39.8 ±
43.4 ±
5.9 ±
36.6 ±
Obese NIDDM
2.8 (10)
44.8 ±
3.0 (10)* 59.3 ±
0.4 (8)
9.8 ±
6.0 (8)* 103.8 ±
2.5 (5)
10.1 (5)*
2.0 (5)*
13.2 (5)*t
Data are means ± SE (n). Glucose and insulin data are intraoperative values. *P< 0.05 vs. lean control subjects. | P < 0.05 vs. obese
subjects.
Briefly, muscle strips weighing ~25 mg were teased from the muscle
biopsy sample and placed in Lucite clips to maintain them at resting
length. Muscle strips were first preincubated for 60 or 90 min under
normoxic or hypoxic conditions. Insulin was added 10 min before the
end of the preincubation period, thus yielding four groups: basal and
insulin- (10~7 mol/1), hypoxia-, or hypoxia + insulin (10~7 mol/1)stimulated. Initially, a 90-min preincubation period was used; however,
it was found that hypoxia stimulated glucose transport maximally using
a 60-min preincubation period; therefore, the data sets were combined.
The strips were then transferred to incubation chambers for 60 min
under identical conditions as preincubation, with the exception that the
incubation media contained 0.5 mmol/1 2-deoxyglucose (2-DOG), 20
mmol/1 sorbitol, 2 |xCi 2-[l,2-3H(N)]2-DOG to quantify glucose transport,
and 0.1 |xCi [U-14C]-D-sorbitol to quantify extracellular space. Preincubation and incubation volumes were 4 ml. Normoxic samples were
continuously gassed with 95% 0^5% CO2, while hypoxic samples were
continuously gassed with 95% N2/5% CO2. Preincubation and incubation
chambers were in a shaking water bath at 29°C. After incubation, muscle
strips were transferred to ice-cold KHB to wash excess 2-DOG and
sorbitol from the samples for two 5-min periods. After washing, muscle
strips were blotted, weighed, and solubilized in 0.5 ml 0.32 mol/1
hexadecyltrimethyl ammonium bromide and 0.29 mol/1 potassium hydroxide in a 1:1 mixture of MeOH and H2O. Solubilized muscle strips and
incubation media samples were counted in a Beckman LS 5000 TD liquid
scintillation counter preset to count 14C and 3H channels simultaneously.
Statistical analysis. One-way factorial analysis of variance (ANOVA)
was used to determine differences between groups, and one-way repeated-measures ANOVA was used to determine differences due to treatments within a group. When differences were indicated, a NewmanKeuls post hoc test was used to locate the difference. If data did not
fulfill the equal variance or normality criteria of ANOVA, then a
Kruskal-Wallis ANOVA test on ranks was run, and differences between
groups were determined using chi-squared analysis. The a was set at
0.05.
RESULTS
Patient profile. There were no significant differences in
ages between the three groups of patients studied (Table 1).
BMI was significantly (P < 0.05) greater in obese and NIDDM
patients compared with lean control subjects, but obese and
NIDDM groups were not different from one another. There
were no significant differences between glucose concentrations in lean and obese patients, although insulin concentrations in obese patients were significantly greater than those
in lean patients. However, glucose and insulin concentrations were significantly higher in NIDDM patients compared
with lean patients; in addition, insulin concentrations in
NIDDM patients were significantly greater than in obese
patients (Table 1).
Glucose transport. Glucose transport was measured after
120- and 150-min incubations. There were no differences in
2-DOG transport rates as a function of incubation duration;
therefore, data from these two groups were pooled (data not
shown). Basal 2-DOG transport rates in the three groups
were not different from one another (Fig. 1). Insulin stimulated 2-DOG transport significantly (P < 0.05) in muscle from
lean patients, only slightly in muscle from obese patients,
696
and not at all in muscle from NIDDM patients. Furthermore,
insulin-stimulated transport rates in obese and NIDDM patients were significantly less than those in lean patients.
Hypoxia stimulated 2-DOG transport significantly, compared
with basal, in muscle from lean patients. In muscle from
obese patients, hypoxia significantly stimulated 2-DOG transport over basal and insulin-stimulated conditions. Hypoxia
stimulated 2-DOG transport more than twofold in NIDDM
muscle, although this did not reach significance (P = 0.069).
Hypoxia + insulin stimulated 2-DOG transport in muscle
from lean patients significantly over basal and insulin- and
hypoxia-stimulated conditions. In muscle from obese patients, hypoxia + insulin stimulated 2-DOG transport significantly over basal and insulin-stimulated conditions but not
over hypoxia stimulation alone. Hypoxia + insulin stimulated 2-DOG transport significantly over basal and insulinstimulated conditions in muscle from NIDDM patients. There
were no differences between lean, obese, and NIDDM groups
under basal and hypoxia- and hypoxia + insulin-stimulated
conditions. The predicted combination of hypoxia and insulin was compared with hypoxia + insulin-stimulated glucose
transports after basal transport rates were subtracted (Fig.
2). There were no differences between the predicted and the
actual transport values in muscle from lean, obese, and
NIDDM patients, suggesting that hypoxia and insulin are
additive.
DISCUSSION
It has been hypothesized that hypoxia stimulates glucose
transport via the same pathway as contraction (5). Previous
work with rat skeletal muscle has shown that hypoxia and
contractile activity stimulated glucose transport to the same
degree (4,5) and that each yielded an additive quality when
coupled with insulin stimulation. Furthermore, when hypoxia and contraction were coupled, the magnitude of
glucose transport stimulation was no greater than that for
either treatment alone.
It has been shown repeatedly that exercise causes translocation of GLUT4 from interior pools to the surface membrane (11-15). It has also been shown that hypoxia causes
translocation of GLUT4 (5). Thus, it may be surmised that
because exercise and hypoxia stimulate glucose transport to
DC
O
a.
14
12
a,b,c
•
LEAN (n = 11)
0
OBESE (n = 10)
D NIDDM (n = 5)
5
io
•- -?
a.b
8
O §
O o 6
U'E
X
o
111
Q
•
CM
BASAL
INSULIN HYPOXIA HYP + INS
FIG. 1. Basal and insulin-, hypoxia-, and hypoxia + insulin-stimulated
2-DOG transport in human skeletal muscle from lean, obese, and obese
NIDDM patients: a, significantly different from basal; b, significantly
different from insulin-stimulated; c, significantly different from hypoxia;
and d, significantly different from lean. Values are means ± SE.
DIABETES, VOL. 44, JUNE 1995
J.L. A2EVEDO AND ASSOCIATES
10
•
o
PREDICTED
D ACTUAL
Q.
V)
111
0)
f
o i
>o
o
D
_l
X
o
UJ
Q
LEAN
OBESE
DIABETIC
FIG. 2. Predicted versus actual result of the effects of insulin
stimulation + hypoxia stimulation. Predicted data were calculated by
subtracting basal transport rates from the insulin-stimulated rates and
the hypoxia-stimulated rates then adding the differences, e.g., [(insulin basal) + (hypoxia - basal)] (predicted). Actual data were calculated by
simply subtracting the basal transport rates from the hypoxia+insulinstimulated data, e.g., [(hypoxia + insulin) - basal] (actual).
and that the defect may lie upstream of the effector system,
i.e., the insulin signaling pathway. Alternatively, there may be
two different pools of transporters, one that is accessed by
insulin, and one that is accessed by contraction/hypoxia.
Insulin-resistant muscle may not respond to insulin due to a
defective insulin-sensitive pool; however, the contraction/
hypoxia pool is still intact during insulin resistance.
An alternative explanation for insulin resistance could
involve the shifting of glucose transporters from the insulinsensitive pool to the contraction-sensitive pool. However, if
this were the case, when muscle is stimulated by contraction
(or hypoxia), one would expect a response that would be
similar in magnitude to that of the combination of insulin and
hypoxia in lean control muscle. However, this did not occur.
Hypoxia-stimulated glucose transport response was similar
in muscle from lean, obese, and obese NIDDM patients,
which supports the hypothesis that the defect in insulinresistant muscle lies in the insulin signaling pathway.
The present study demonstrates that hypoxia is able to
stimulate glucose transport in insulin-resistant human skeletal muscle. This suggests that as with many other functions
in the body, there is redundancy in signaling glucose transport and that the problem lies in how to signal transport. It is
apparent from these data and previous work (9,16) that
insulin-stimulated glucose transport is defective in obese
humans with and without NIDDM; however, the present
study provides evidence that given the appropriate stimulus,
glucose transport is functional.
the same degree and cause translocation of GLUT4, hypoxia
acts through the exercise signaling pathway. The present
study used hypoxia as an alternative to contraction to test
the hypothesis that the glucose transport effector system is
intact in insulin-resistant human muscle. These data show a
twofold increase in glucose transport in muscle from lean
patients in response to insulin stimulation, which is in ACKNOWLEDGMENTS
agreement with previous work (9,16). Hypoxia stimulated
This study was supported by National Institutes of Health
glucose transport to approximately the same magnitude as
Grant DK-46121. J.L.A. is supported by National Institute of
did insulin in lean muscle. In muscle from obese and NIDDM
Diabetes and Digestive and Kidney Diseases Fellowship
groups, hypoxia stimulated glucose transport between twoF32-DK-08830.
and threefold, and hypoxia-stimulated transport was not
We thank M.E. Brinn and E.B. Tapscott, Jr., for expert
different from that in lean control subjects.
technical assistance.
Evidence to support the hypothesis that the major defect
in insulin-resistant muscle is in insulin signaling is provided
by experiments on contraction- or exercise-stimulated glu- REFERENCES
DeFronzo RA, Gunnarsson R, Bjorkman 0, Olsson M, Wahren J: Effects
cose transport in muscle from insulin-resistant rats (17,18). 1. of
insulin on peripheral and splanchnic glucose metabolism in noninsuDolan et al. (17) showed that in situ contraction stimulated
lin-dependent (type II) diabetes mellitus. J Clin Invest 76:149-155, 1985
glucose transport in muscle from insulin-resistant obese rats 2. Yki-Jarvinen H: Evidence for a primary role of insulin resistance in the
of type 2 diabetes. Ann Med 22:197-200, 1990
to a magnitude comparable with that in muscle from lean 3. pathogenesis
Caro JF, Sinha MK, Raju SM, Ittoop O, Pories WJ, Flickinger EG,
control rats. In addition, contraction appeared to restore
Meelheim D, Dohm GL: Insulin receptor kinase in human skeletal muscle
from obese subjects with and without noninsulin dependent diabetes. J
insulin responsiveness in muscle from obese rats, as contracClin Invest 79:1330-1337, 1987
tion + insulin yielded much higher glucose transports than 4. Brozinick
JT Jr, Etgen GJ Jr, Yaspelkis BB III, Ivy JL: Contractionwould have been predicted by adding the individual values of
activated glucose uptake is normal in insulin-resistant muscle of the
obese Zucker rat. JAppl Physiol 73:382-387, 1992
insulin- and contraction-stimulated glucose transport. In the
5. Cartee GD, Douen AG, Ramlal T, Klip A, Holloszy JO: Stimulation of
present study, the predicted and actual values were virtually
glucose transport in skeletal muscle by hypoxia. JAppl Physiol 70:15931600, 1991
identical in muscle from lean (8.0 vs. 8.1, respectively) and
6. Cartee GD, Holloszy JO: Exercise increases susceptibility of muscle
obese (4.9 and 5.4, respectively) patients (Fig. 2).
glucose transport to activation by various stimuli. Am J Physiol 258:
Kusunoki et al. (18) were able to show that in individual
E390-E393, 1990
muscles, an acute bout of treadmill running stimulated 7. Fushiki T, Wells JA, Tapscott EB, Dohm GL: Changes in glucose
transporters in muscle in response to exercise. Am J Physiol 256:E580glucose transport to the same degree in insulin-resistant rats
E587; 1989
as in controls, whereas insulin did not stimulate glucose 8. Douen AG, Ramlal T, Hip A, Young DA, Cartee GD, Holloszy JO:
Exercise-induced increase in glucose transporters in plasma membranes
transport in any muscle type. In addition, glucose uptake was
of rat skeletal muscle. Endocrinology 124:449-454, 1989
still elevated compared with baseline after exercise, which 9. Dohm GL, Tapscott EB, Pories WJ, Dabbs DJ, Flickinger EG, Meelheim
D, Fushiki T, Atkinson SM, Elton CW, Caro JF: An in vitro human muscle
appeared to change reciprocally with glycogen concentrapreparation suitable for metabolic studies: decreased insulin stimulation
tion. The present study demonstrates that hypoxia-stimuof glucose transport in muscle from morbidly obese and diabetic
lated glucose transport in insulin-resistant human skeletal
subjects. J Clin Invest 82:486-494, 1988
muscle is equivalent to that of lean control subjects; this 10. National Diabetes Data Group: Classification and diagnosis of diabetes
mellitus and other categories of glucose intolerance. Diabetes 28:1039suggests, as do the data of Dolan et al. (17) and Kusunoki et
1057, 1979
al. (18), that the glucose transport effector system is intact 11. Douen AG, Ramlal T, Cartee GD, Klip A: Exercise modulates the
DIABETES, VOL. 44, JUNE 1995
697
HYPOXIA STIMULATES GLUCOSE TRANSPORT
insulin-induced translation of glucose transporters in rat skeletal muscle.
FEBSLett 261:256-260, 1990
12. Douen AG, Ramlal T, Rastogi S, Bilan PJ, Cartee GD, Vranic M, Holloszy
JO, Klip A: Exercise induces recruitment of the "insulin-responsive
glucose transporter": evidence for distinct intracellular insulin- and
exercise-recruitable transporter pools in skeletal muscle. J Biol Chem
265:13427-13420, 1990
13. Etgen GJ, Memon AR, Thompson GA Jr, Ivy JL: Insulin- and contractionstimulated translocation of GTP-binding proteins and GLUT4 protein in
skeletal muscle. J Biol Chem 268:20164-20169, 1993
14. Goodyear LJ, Hirshman MF, Horton ES: Exercise-induced translocation
of skeletal muscle glucose transporters. Am J Physiol 261:E795-E799,
1991
698
15. Hirshman MF, Wallberg-Henriksson H, Wardzala LJ, Horton ED, Horton
ES: Acute exercise increases the number of plasma membrane glucose
transporters in rat skeletal muscle. FEBS Lett 238:235-239, 1988
16. Friedman JE, Dohm GL, Leggett-Frazier N, Elton CW, Tapscott EB,
Pories WJ, Caro JF: Restoration in insulin responsiveness in skeletal
muscle of morbidly obese patients after weight loss. J Clin Invest
89:701-705, 1992
17. Dolan PL, Tapscott EB, Dorton PJ, Dohm GL: Contractile activity
restores insulin responsiveness in skeletal muscle of obese Zucker rats.
Biochem J 289:423-426, 1993
18. Kusunoki M, Storlien LH, MacDessi J, Oakes ND, Kennedy C, Chisolm DJ,
Kraegen EW: Muscle glucose uptake during and after exercise is normal
in insulin-resistant rats. Am J Physiol 264:E167-E172, 1993
DIABETES, VOL. 44, JUNE 1995