Endocrinology Vol. 138, No. 2 751-755
Copyright © 1997 by The Endocrine Society
Two Signaling Pathways, from the Upper Glycolytic Flux and from the Mitochondria, Converge to Potentiate Insulin Release1
Nahoko Asanuma,
Toru Aizawa,
Yoshihiko Sato,
Thomas Schermerhorn,
Mitsuhisa Komatsu,
Geoffrey W. G. Sharp and
Kiyoshi Hashizume
Department of Geriatrics, Endocrinology and Metabolism (N.A., T.A.,
Y.S., K.H.), Shinshu University School of Medicine, Matsumoto,
Nagano-ken, Japan, and Department of Pharmacology (T.S., M.K.,
G.W.G.S.), College of Veterinary Medicine, Cornell University, Ithaca,
New York 14853-6401
Address all correspondence and requests for reprints to: Toru Aizawa, M.D., Department of Geriatrics, Endocrinology and Metabolism, Shinshu University School of Medicine, 31-1 Asahi, Matsumoto, Nagano-ken, Japan 390.
 |
Abstract
|
|---|
In the rat pancreatic ß cell, low concentrations of glucose
potentiate D-glyceraldehyde (GA)-induced insulin release
without any potentiation of the triose-induced elevation of cytosolic
free Ca2+ concentration. Namely, 23 mM
glucose strongly potentiates 5 mM GA-induced insulin
release, and the combination of stimulatory concentration of glucose
(10 mM) and 5 mM GA elicits far more than
additive insulin release: this glucose action is independent of
ATP-sensitive K+ channel closure because it can be seen in
the presence of diazoxide, an opener of the K+ channel. The
triose-induced elevation of cytosolic free Ca2+
concentration was not potentiated by the presence of 3 mM
glucose, and oxidation of labeled GA by the islet cells was not
enhanced by the presence of glucose. The glucose action can be mimicked
by mannose, but not by galactose, and was suppressed by inhibition of
glucose phosphorylation with mannoheptulose or 2-deoxyglucose. Glucose
also potentiates 2-ketoisocaproate-induced insulin release. In
contrast, a combination of GA and 2-ketoisocaproate elicits only
additive insulin release. Strikingly, 3 mM glucose does not
potentiate insulin release in response to a depolarizing concentration
of K+. Therefore, at least two signal pathways, one from
upper glycolytic flux and one from mitochondrial metabolism, must
converge to provide the potentiation of insulin release. We conclude
that the upper glycolytic flux, acting at a site unrelated to the
elevation of cytosolic free Ca2+, potentiates insulin
release triggered by triose and mitochondrial fuels.
 |
Introduction
|
|---|
AN INCREASE in ATP or the ATP/ADP
ratio caused by glucose metabolism is regarded as a determinant for
glucose stimulation of insulin exocytosis in the ß cell (1). The idea
is supported, in part, by the fact that mitochondrial fuels, such as
2-ketoisocaproate (KIC), which bypass the glycolytic pathway, stimulate
insulin release in the absence of glucose (2, 3, 4, 5, 6). Accordingly,
glycolysis is considered to be funneling sugars to the lower metabolic
pathway and not generating metabolic signals for the stimulation of
insulin release. However, in an earlier study, the islet ATP content
and insulin release did not correlate when mouse islets were stimulated
with a combination of glucose and D-glyceraldehyde (GA) or
of glucose and the mitochondrial fuel, leucine (7), indicating that the
rate of insulin release is not determined simply by that of ATP
generation. We report now that glucose amplifies insulin release
triggered by GA without any potentiation of GA-induced elevation of
Ca2+. The finding provides evidence for a unique functional
role of the upper glycolytic flux in the pancreatic ß cell.
 |
Materials and Methods
|
|---|
Pancreatic islets were obtained from male Wistar rats by
collagenase dispersion (8), and insulin release was measured in static
incubation and perifusion experiments using Krebs-Ringer bicarbonate
buffer (KRBB) containing 0.5% BSA (equilibrated with 5%
CO2-95% O2, pH 7.4), as described (9, 10, 11, 12, 13). In
static incubation experiments, 5 size-matched islets/tube were
incubated with various secretagogues for 30 min: preincubation was with
3 mM glucose for 30 min. When inhibitors were used, they
were present during the preincubation and the experimental incubation.
In perifusion, 50 islets from a single batch were placed into several
columns and all columns were perifused in parallel simultaneously (10).
After a 30-min perifusion with 3 mM glucose, test
substances were introduced. Insulin was determined by RIA, as described
(9, 10, 11, 12, 13).
Because measurement of cytosolic free Ca2+ concentration
([Ca2+]i) was performed (see below) in
modified KRBB with HEPES (14) using isolated ß cells, insulin release
also was determined under the same condition. For this purpose,
approximately 600 freshly isolated islets were dissociated in 200 µl
glucose free, Ca2+-omitted KRBB: the islets were gently
aspirated and expelled using a 100 µl pipettor for 3040 times.
Immediately thereafter, the cells were resuspended in 500 µl RPMI
1640 (Sigma, St. Louis, MO) supplemented with 10% FBS (Flow
Laboratories, North Ryde, New South Wales, Australia) and aliquotted
into 1416 tubes. By this procedure, approximately 23 x
104 cells/tube were obtained. Then, the cells were
incubated in the RPMI media (1 ml) at 37 C with 5% CO2 for
24 h. For the measurement of insulin release, both preincubation
and experimental incubation were carried out, as in the experiments
with the islets (see above) using modified KRBB containing 10
mM HEPES and 5 mM NaHCO3.
Oxidation of D-[U-14C]glyceraldehyde
(specific activity 55 mCi/mmol, American Radiolabeled Chemicals, St.
Louis, MO) was determined, as described (15) using 0.8 µCi/tube:
concentration of total (radioactive and nonradioactive, combined) GA
was 0.5 mM. Because of the relatively low specific activity
of the labeled GA, it was necessary to use a low concentration of total
GA in this experiment.
[Ca2+]i was measured using isolated ß
cells, as described above (14) using Indo-1 (Molecular Probes, Eugene,
OR). The cells were preincubated with 3 mM glucose for 30
min, washed, placed in zero glucose, and then stimulated with 5
mM GA alone or 3 mM glucose and 5
mM GA: determination of [Ca2+]i was
carried out for 30 min to critically compare the temporal profiles of
insulin secretion and [Ca2+]i.
[Ca2+]i was calculated from a standard curve
generated from known concentrations of free calcium in KRBB at 37 C
using a commercial calibration kit (Molecular Probes).
GA was obtained from Aldrich (Milwaukee, WI); KIC, diazoxide,
mannoheptulose, and 2-deoxyglucose from Sigma; and dihydroxyacetone
(DHA) from Nacalai Tesque (Kyoto, Japan). Purity of GA, KIC, and DHA
was higher than 98%. Statistical analysis was performed by one-way
ANOVA with Fishers protected least-significant difference test and
Wilcoxons ranked-sum test. P < 0.05 was considered
significant. Data are expressed as means ± SE.
 |
Results
|
|---|
Insulin release induced by 5 mM GA or 12
mM KIC was potentiated by the presence of 3 mM
glucose, and the glucose potentiation was selective for the late phase
for both GA and KIC (Fig. 1
). A marked potentiation of
the late phase of GA-induced insulin release by glucose was seen also
with 3.7 mM GA plus 3 mM glucose, despite the
fact that 3.7 mM GA alone caused only marginal insulin
release (Fig. 2
). Glucose potentiation of GA- and
KIC-induced insulin release was unequivocally demonstrated with various
combinations of glucose and GA or KIC in incubation experiments, as
well, and such glucose action was seen at a concentration as low as 2
mM (Table 1
). In stark contrast, a
combination of KIC and GA caused only additive insulin release (Table 1
). A combination of 5 mM GA and 3 mM glucose
elicited significantly greater insulin release than either of the two
alone in dispersed islet cells (Table 2
). However, in
the isolated islet cells, the response to 5 mM GA alone was
not demonstrable, and the release caused by a combination of the fuels
was not significantly greater than the sum of the release by the two
agents alone (Table 2
).
View this table:
[in this window]
[in a new window]
|
Table 1. Insulin release by the islets in response to various
combinations of fuels and effects of metabolic inhibition
|
|
On the other hand, insulin release in response to nonnutrient stimuli,
e.g. depolarization by a high concentration of
K+, was the same in the presence and absence of 3
mM glucose (Table 3
). Potentiation of
GA-induced insulin release by 3 mM glucose was completely
suppressed by inhibition of glucokinase and hexokinase with
mannoheptulose (16) and 2-deoxyglucose (17), respectively (Table 1
). On
the other hand, such glucose action was resistant to treatment with 150
µM diazoxide (Table 3
), an ATP-sensitive K+
(K+ATP) channel activator (18). Oxidation of
[U-14C]GA by the islet cells was not significantly
different in the presence and absence of 3 mM glucose: GA
converted to CO2, 77.8 ± 11.5 and 77.6 ± 4.8
pmoles/20 islets for 60 min, in the presence and absence of glucose,
respectively (n = 5 for each condition).
View this table:
[in this window]
[in a new window]
|
Table 3. Insulin release by the islets in response to various
combinations of fuels in the presence of depolarizing concentration of
K+
|
|
Mannose (10 mM) and DHA (up to 20 mM) did not
elicit insulin release by themselves. The inability of mannose to
significantly stimulate insulin release at this concentration (3) is
explained by the fact that mannose is inefficiently metabolized
compared with glucose: the apparent Km of the
islet glucokinase is 23.7 and 11.3 mM for mannose and
glucose, respectively (3). Nevertheless, GA- and KIC-induced insulin
release was potentiated by mannose and DHA (Table 1
). In contrast,
galactose, a nonmetabolizable sugar in the ß cell (2, 3), failed to
potentiate GA-induced insulin release (data not shown).
GA-induced elevation of [Ca2+]i was both
qualitatively (Fig. 3
) and quantitatively (Table 4
) unaffected by the presence of 3 mM
glucose.

View larger version (22K):
[in this window]
[in a new window]
|
Figure 3. Change in [Ca2+]i upon
stimulation with 5 mM GA alone (A), and 5 mM GA
and 3 mM glucose (B). After measurement of basal
[Ca2+]i for 600 sec, the cells were exposed
to the test substances. A representative of five tracings for each
condition is shown. Similar results were obtained in cells isolated
from four different batches of islets.
|
|
 |
Discussion
|
|---|
In the present study, we showed that glucose potentiates insulin
release triggered by nonhexose nutrients, such as GA and KIC. The data
obtained with gluco- and hexokinase inhibitors, depolarizing
concentrations of K+, and metabolizable and
nonmetabolizable sugars lead us to conclude that metabolic signal(s)
generated in the upper glycolytic flux, i.e. before the
trioses, are potentiating insulin release triggered by the lower
metabolic signal, most likely ATP. Potentiation of GA-induced insulin
release by glucose cannot be caused simply by generation of ATP from
glucose metabolism because addition of a suprathreshold concentration
(12 mM) of KIC with GA does not potentiate GA-induced
insulin release. Instead, a combination of KIC and GA elicits only
additive insulin release. In an earlier study with mouse islets,
addition of 5 mM glucose with GA or leucine, a precursor of
KIC, did not further increase islet ATP content (7). Nevertheless,
insulin release was potentiated dramatically by the addition of glucose
(7), a finding that is in good agreement with the current data. It has
been reported that addition of GA does not accelerate glucose
metabolism by the islet cells (19).
Two molecules of trioses are generated from one glucose molecule;
therefore, 3 mM glucose is metabolically equipotent to 6
mM GA. However, the insulinotropic action of GA is much
greater than glucose when the two are compared at the metabolically
equipotent concentrations (3). For example, 6 mM GA, but
not 3 mM glucose, significantly stimulate insulin release
by the islets. Accordingly, we did not perform a simple comparison of
metabolically equivalent GA and glucose. Instead, we tested a
combination of various concentrations of glucose and GA, as shown, and
found that glucose at concentrations higher than 2 mM
strongly potentiates GA-induced insulin release at all concentrations
of GA tested.
The purity of GA used here was confirmed to be higher than 98%, which
is most likely purer than that used in the earlier studies. With this
level of purity, effects of contaminants of GA, if any, would be
practically negligible. Using this preparation, we observed no
suppression of insulin release up to 15 mM, and 5
mM GA alone always produces only a small increment in
insulin release, as shown in Fig. 1
and Table 1
. Because, in previous
studies, the effect of GA was usually evaluated in the presence of 3 or
5 mM glucose, it is likely that the insulinotropic effect
of GA was already potentiated by glucose, and the potentiation effect
unnoticed.
Glucose potentiation of GA-induced insulin release is demonstrable at a
glucose concentration as low as 2 mM. Furthermore, it is
independent of membrane potential changes because it can be seen in ß
cells clamped with high concentrations of K+ and diazoxide,
a K+ATP channel opener (18). From such observations, it can
be deduced that this is not a potentiation of Ca2+-induced
exocytosis or of the [Ca2+]i response itself.
To directly verify the idea, [Ca2+]i was
measured using isolated islet cells. As we expected, GA-induced
elevation of [Ca2+]i was not potentiated by
the presence of 3 mM glucose. However, insulin release
induced by a combination of 3 mM glucose and 5
mM GA was not significantly greater than the sum of the
release by the two agents alone in the isolated islet cells. Therefore,
the data on [Ca2+]i should be interpreted
with reserve. Direct enhancement of GA metabolism by glucose was ruled
out because addition of 3 mM glucose does not significantly
increase oxidation of GA.
To explore the mechanism of this newly found glucose action, effects of
DHA, which is phosphorylated and enters the glycolytic pathway as DHA
phosphate (DHAP), were evaluated. DHA also potentiated GA- and
KIC-induced insulin release: the weak effect of DHA on GA-induced
insulin release could be caused by the fact that DHA, having a higher
affinity for triokinase than GA, may interfere with GA metabolism in
the ß cell. Although activation of the glycerol-phosphate shuttle (2, 3) might be responsible for the glucose action described here,
replenishment of cytosolic NAD through the glycerol-phosphate shuttle
is not likely to be responsible for glucose potentiation of GA-induced
insulin release because addition of nicotinamide, which increases
cellular NAD (20), caused minimal effect on GA-induced insulin release
(data not shown). The other signal arising from DHAP, activation of
protein kinase C via conversion of DHAP to diacylglycerol
(21), may be a mediator of the glucose action described here because
glutamine, which provides the glycerol-3-phosphate backbone for DAG
production, strongly potentiates leucine-induced insulin release, as
does glucose (22). Phospholipase C is not involved in the glucose
action described here because the addition of a low concentration of
glucose with a stimulatory concentration of KIC does not further
increase phosphoinositide hydrolysis by the islets (23).
Although the glucose-induced rise in ß cell
[Ca2+]i is important for its insulinotropic
action (24), glucose augmentation of insulin release occurs without any
additional increase in [Ca2+]i (5, 25, 26).
Augmentation can occur even without any increase in
[Ca2+]i (14). Glucose potentiation of insulin
release without an additional elevation in
[Ca2+]i is unequivocally demonstrated in the
present study, as well. The action of glucose described here is
distinct from the previously described K+ATP
channel-independent glucose actions (5, 10, 12, 13) in the following
aspects: 1) this is not a direct potentiation of
Ca2+-evoked insulin exocytosis; 2) the potentiation is
demonstrable at a glucose concentration as low as 2 mM; and
3) it is dependent upon the nutrients, GA and KIC.
In summary, at least two signals derived from the upper glycolytic
pathway of glucose metabolism and from the mitochondria must converge
to cause the potentiation of insulin release. They do this by an effect
on stimulation-secretion coupling or by supplying substrates in the
correct balance to the lower glycolytic pathway, so as to fully
activate fuel-responsive effectors for insulin exocytosis. When the ß
cell is stimulated with graded concentrations of glucose, the increase
in insulin release is greater than the increase in glucose metabolism,
as evidenced by a much greater Hill constant for the former than the
latter (3). The reason for this apparent discrepancy has been unknown.
Now, however, it can, at least in part, be attributed to the
potentiation of insulin release by glucose, as shown here.
 |
Acknowledgments
|
|---|
The authors thank Dr. H. Ishida for constructive discussion.
 |
Footnotes
|
|---|
1 This work was supported by a grant-in-aid for scientific research
from the Ministry of Education Science and Culture, Japan (to T.A.),
and by NIH Grant RO1-DK-42063 (to G.W.G.S). 
Received June 28, 1996.
 |
References
|
|---|
-
Henquin J-C 1994 Cell biology of insulin
secretion. In: Kahn CR, Weir CG (eds) Joslins Diabetes Mellitus, 13th
ed. Lea & Febiger, Philadelphia, pp 5680
-
Sener A, Malaisse WJ 1984 Nutrient metabolism in
islet cells. Experientia 40:10261035[CrossRef][Medline]
-
Meglasson MD, Matschinsky FM 1986 Pancreatic islet
glucose metabolism and regulation of insulin secretion. Diabetes Metab
Rev 2:163214[Medline]
-
Hedeskov CJ 1980 Mechanism of glucose-induced
insulin secretion. Physiol Rev 60:442509[Free Full Text]
-
Gembal M, Gilon P, Henquin J-C 1992 Evidence that
glucose can control insulin release independently from its action on
ATP-sensitive K+ channels in mouse B cells. J Clin
Invest 89:12881295
-
Kelley GG, Zawalich KC, Zawalich WS 1994 Calcium
and a mitochondrial signal interact to stimulate phosphoinositide
hydrolysis and insulin secretion in rat islets. Endocrinology 134:16481654[Abstract/Free Full Text]
-
Ashcroft SJH, Weersinghe LCC, Randle PJ 1973 Interrelationship of islet metabolism, adenosine triphosphate content
and insulin release. Biochem J 132:223231[Medline]
-
Lacy PE, Kostianovsky M 1967 Method for the
isolation of intact islets of Langerhans from the rat pancreas.
Diabetes 16:3539[Medline]
-
Komatsu M, Yokokawa N, Takeda T, Nagasawa Y, Aizawa T,
Yamada T 1989 Pharmacological characterization of the
voltage-dependent calcium channel of pancreatic B-cell. Endocrinology 125:20082014[Abstract/Free Full Text]
-
Sato Y, Aizawa T, Komatsu M, Okada N, Yamada T 1992 Dual functional role of membrane depolarization/Ca2+ influx
in rat pancreatic B-cell. Diabetes 41:438443[Abstract]
-
Komatsu M, Aizawa T, Yokokawa N, Sato Y, Okada N, Takasu
N, Yamada T 1992 Mastoparan-induced hormone release from the rat
pancreatic islets. Endocrinology 130:221228[Abstract/Free Full Text]
-
Aizawa T, Sato Y, Ishihara F, Taguchi N, Komatsu M,
Suzuki N, Hashizume K, Yamada T 1994 ATP-sensitive K+
channel-independent glucose action in rat pancreatic ß-cell. Am
J Physiol 266:C622C627
-
Taguchi N, Aizawa T, Sato Y, Ishihara F, Hashizume
K 1995 Mechanism of glucose-induced biphasic insulin release:
physiological role of adenosine triphosphate-sensitive K+
channel-independent glucose action. Endocrinology 136:39423948[Abstract]
-
Komatsu M, Schermerhorn T, Aizawa T, Sharp GWG 1995 Glucose-induced insulin release in the absence of extracellular calcium
and absence of any increase in intracellular calcium. Proc Natl Acad
Sci USA 92:1072810732[Abstract/Free Full Text]
-
Ishihara F, Aizawa T, Taguchi N, Sato Y, Hashizume
K 1994 Differential metabolic requirement for initiation and
augmentation of insulin release by glucose: a study with rat pancreatic
islets. J Endocrinol 143:497503[Abstract/Free Full Text]
-
Malaisse WJ, Lea MA, Malaisse-Lagae F 1968 The
effect of mannoheptulose on the phosphorylation of glucose and the
secretion of insulin by islets of Langerhans. Metabolism 17:126132[CrossRef][Medline]
-
Manuel Y, Keenoy B, Zahner D, Malaisse WJ 1992 Dissociated effects of 2-deoxy-D-glucose on
D-[2-3H]glucose and
D-[5-3H]glucose conversion into
3HOH in rat erythrocytes. Biochem J 288:433438
-
Trube G, Rorsman P, Ohno-Shousaku T 1986 Opposite
effects of tolbutamide and diazoxide on the ATP-sensitive
K+ channel in mouse pancreatic ß-cells. Pflugers Arch 407:493499[CrossRef][Medline]
-
Hellman B, Idahl L-Å, Lernmark Å, Sehlin J,
Täljedal I-B 1974 The pancreatic ß-cell recognition of
insulin secretagogues: comparisons of glucose with glyceraldehyde
isomers and dihydroxyacetone. Arch Biochem Biophys 162:448457[CrossRef][Medline]
-
Deery DJ, Taylor KW 1973 Effects of azaserine and
nicotinamide on insulin release and nicotinamide-adenine dinucleotide
metabolism in isolated rat islets of Langerhans. Biochem J 134:557563[Medline]
-
Dunlop ME, Larkins RG 1985 Pancreatic islets
synthesize phospholipids de novo from glucose via
acyl-dihydroxyacetone phosphate. Biochem Biophys Res Commun 132:467473[CrossRef][Medline]
-
Malaisse WJ, Sener A, Malaisse-Lagae F, Welsh M,
Matthews DE, Bier DM, Hellerström C 1982 The
stimulus-secretion coupling of amino acid-induced insulin release.
Metabolic response of pancreatic islets to L-glutamine and
L-leucine. J Biol Chem 257:87318737[Abstract/Free Full Text]
-
Rasmussen H, Zawalich KC, Zawalich WS 1995 Metabolic activation of Ca2+-independent phosphoinositide
hydrolysis in ß-cells and its role in the control of insulin
secretion. Endocrinology 136:577584[Abstract]
-
Wollheim CB, Sharp GWG 1981 Regulation of insulin
release by calcium. Physiol Rev 61:914973[Free Full Text]
-
Martin F, Reig JA, Soria B 1995 Secretagogue-induced changes in [Ca2+]i
single rat pancreatic islets and correlation with simultaneously
measured insulin release. J Mol Endocrinol 15:177185[Abstract/Free Full Text]
-
Komatsu M, Aizawa T, Takasu N, Yamada T 1989 Glucose raises cytosolic free calcium in the rat pancreatic islets.
Horm Metab Res 21:405409[Medline]
This article has been cited by other articles:

|
 |

|
 |
 
M. Fukaya, A. Mizuno, H. Arai, K. Muto, T. Uebanso, K. Matsuo, H. Yamamoto, Y. Taketani, T. Doi, and E. Takeda
Mechanism of rapid-phase insulin response to elevation of portal glucose concentration
Am J Physiol Endocrinol Metab,
August 1, 2007;
293(2):
E515 - E522.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
Q. Huang, S. Bu, Y. Yu, Z. Guo, G. Ghatnekar, M. Bu, L. Yang, B. Lu, Z. Feng, S. Liu, et al.
Diazoxide Prevents Diabetes through Inhibiting Pancreatic {beta}-Cells from Apoptosis via Bcl-2/Bax Rate and p38-{beta} Mitogen-Activated Protein Kinase
Endocrinology,
January 1, 2007;
148(1):
81 - 91.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
L. Juntti-Berggren, D.-L. Webb, P. O.G. Arkhammar, V. Schultz, E. K. H. Schweda, K. Tornheim, and P.-O. Berggren
Dihydroxyacetone-induced Oscillations in Cytoplasmic Free Ca2+ and the ATP/ADP Ratio in Pancreatic {beta}-Cells at Substimulatory Glucose
J. Biol. Chem.,
October 17, 2003;
278(42):
40710 - 40716.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
M. Chen, Z. Yang, R. Wu, and J. L. Nadler
Lisofylline, a Novel Antiinflammatory Agent, Protects Pancreatic {beta}-Cells from Proinflammatory Cytokine Damage by Promoting Mitochondrial Metabolism
Endocrinology,
June 1, 2002;
143(6):
2341 - 2348.
[Abstract]
[Full Text]
[PDF]
|
 |
|

|
 |

|
 |
 
I. Miwa, N. Ichimura, M. Sugiura, Y. Hamada, and S. Taniguchi
Inhibition of Glucose-Induced Insulin Secretion by 4-Hydroxy-2-Nonenal and Other Lipid Peroxidation Products
Endocrinology,
August 1, 2000;
141(8):
2767 - 2772.
[Abstract]
[Full Text]
[PDF]
|
 |
|