DIABETES-INSULIN-GLUCAGON-GASTROINTESTINAL

Partial Blockade of Kv2.1 Channel Potentiates GLP-1’s

Insulinotropic Effects in Islets and Reduces Its Dose
Required for Improving Glucose Tolerance in Type 2
Diabetic Male Mice

Downloaded from https://academic.oup.com/endo/article-abstract/156/1/114/2800622 by guest on 30 January 2020

Rauza Sukma Rita, Katsuya Dezaki, Tomoyuki Kurashina, Masafumi Kakei,
and Toshihiko Yada
Division of Integrative Physiology (R.S.R., K.D., T.K., T.Y.), Department of Physiology, Jichi Medical
University School of Medicine, Shimotsuke, Tochigi 329-0498, Japan; Department of Internal Medicine
(M.K.), Saitama Medical Center, Jichi Medical University School of Medicine, Saitama 337-8503, Japan;
and Department of Development Physiology (T.Y.), Division of Adaptation Development, National
Institute for Physiological Sciences, Okazaki, Aichi 444-8585, Japan

Glucagon-like peptide-1 (GLP-1)-based medicines have recently been widely used to treat type 2
diabetic patients, whereas adverse effects of nausea and vomiting have been documented. Inhi-
bition of voltage-gated K⫹ channel subtype Kv2.1 in pancreatic ␤-cells has been suggested to
contribute to mild depolarization and promotion of insulin release. This study aimed to determine
whether the blockade of Kv2.1 channels potentiates the insulinotropic effect of GLP-1 agonists.
Kv2.1 channel blocker guangxitoxin-1E (GxTx) and GLP-1 agonist exendin-4 at subthreshold con-
centrations, when combined, markedly increased the insulin release and cytosolic Ca2⫹ concen-
tration ([Ca2⫹]i) in a glucose-dependent manner in mouse islets and ␤-cells. Exendin-4 at sub-
threshold concentration alone increased islet insulin release and ␤-cell [Ca2⫹]i in Kv2.1⫹/⫺ mice. The
[Ca2⫹]i response to subthreshold exendin-4 and GxTx in combination was attenuated by pretreat-
ment with protein kinase A inhibitor H-89, indicating the protein kinase A dependency of the
cooperative effect. Furthermore, subthreshold doses of GxTx and GLP-1 agonist liraglutide in
combination markedly increased plasma insulin and improved glucose tolerance in diabetic db/db
mice and NSY mice. These results demonstrate that a modest suppression of Kv2.1 channels dra-
matically raises insulinotropic potency of GLP-1-based drugs, which opens a new avenue to reduce
their doses and associated adverse effects while achieving the same glycemic control in type 2
diabetes. (Endocrinology 156: 114 –123, 2015)

lucose metabolism-induced closure of ATP-sensitive
G K⫹ (KATP) channels and membrane depolarization
open the voltage-dependent Ca2⫹ channels, triggering
Ca2⫹ influx and insulin secretion. Glucose-induced depo-
larization activates voltage-gated potassium channels
(Kv), which repolarizes ␤-cells (1–3). Kv2.1 encoded by
KCNB1 is expressed in islet ␤-cells in rodents and humans
(4 – 6). Studies with the pharmacological Kv2.1 blockers
(4, 7, 8) and Kv2.1 knockout (KO) mice (9) have suggested
that this channel physiologically limits glucose-induced
plasma membrane excitability and insulin secretion in
␤-cells, although most of the pharmacological Kv2.1
blockers can also inhibit Kv2.2 (7, 10), which is reported
to be expressed in ␤-cells (10). It was reported that the
KCNB1 single-nucleotide polymorphism genotype is as-
sociated with an increased risk of type 2 diabetes in the
Chinese Han population (11). The ␤-cell Kv channels are
open when the ␤-cells are depolarized by glucose. Hence,

ISSN Print 0013-7227 ISSN Online 1945-7170
Printed in U.S.A.
Copyright © 2015 by the Endocrine Society
Received September 1, 2014. Accepted October 14, 2014.
First Published Online October 22, 2014
Abbreviations: AUC, area under the curve; [Ca2⫹]i, cytosolic Ca2⫹ concentration; CPT-
cAMP, 8-pCPT-2⬘-O-Me-cAMP-AM; DPP-4, dipeptidyl peptidase-4; Epac, exchange pro-
teins directly activated by cAMP; Ex-4, exendin-4; GLP-1, glucagon-like peptide-1; GxTx,
guangxitoxin-1E; HKRB, HEPES-added Krebs-Ringer bicarbonate buffer; IPGTT, ip glucose
tolerance tests; KATP, ATP-sensitive K⫹; KO, knockout; Kv, voltage-gated potassium chan-
nel; OGTT, oral glucose tolerance tests; Phe-cAMP, 6-Phe-cAMP; PKA, protein kinase A;
TRPM2, transient receptor potential melastatin 2.

114 endo.endojournals.org Endocrinology, January 2015, 156(1):114 –123 doi: 10.1210/en.2014-1728

doi: 10.1210/en.2014-1728
the blockade of Kv2.1 channels in ␤-cells could promote
insulin release with a lower risk of hypoglycemic events
compared with the KATP channel blockers.
Glucagon-like peptide-1 (GLP-1), a physiological in-
cretin hormone, is secreted from the intestinal L cells in
response to meals and enhances insulin release to maintain
normoglycemia (12, 13). Circulating GLP-1 is rapidly de-
graded by dipeptidyl peptidase-4 (DPP-4) with a half-life
of approximately 1–2 minutes. Currently chemically mod-
ified DPP-4-resistant GLP-1 agonists and DPP-4 inhibi-
tors have been clinically used for treating type 2 diabetic
patients and shown to confer a much lesser risk of hypo-
glycemia than sulfonylureas. However, nausea and vom-
iting have been documented as the common adverse events
of GLP-1 agonists (14, 15). A practical way to reduce these
adverse events is to lower their doses. This could be
achieved by combining another substance or mechanism
that can collaborate with GLP-1 agonists on insulin
release.
GLP-1 potentiates glucose-induced insulin release from
pancreatic ␤-cells by increasing the intracellular cAMP,
whose major actions are the promotion of distal secretory
processes associated with exocytosis (16) and facilitation
of Ca2⫹ influx through voltage-dependent L-type chan-
nels (17). In addition, GLP-1 reportedly attenuates ␤-cell
Kv channels, possibly via cAMP signaling (18). This mech-
anism is suggested to partly contribute to the prolonged
depolarizing effect of GLP-1 (18). A method to more ef-
fectively inhibit Kv channels is expected to enhance the
GLP-1 action on islet ␤-cells and facilitate GLP-1-based
therapy.
In this study, we aimed to determine whether pharma-
cological (guangxitoxin-1E) or genetic blockade of Kv2.1
channels (Kv2.1 hetero-KO) could enhance the action of
GLP-1 agonists to potentiate the glucose-induced insulin
release and Ca2⫹ concentration ([Ca2⫹]i) increases in
mouse islet ␤-cells. The present results show that partial
blockade of Kv2.1 channels markedly enhances the GLP-1
action to promote glucose-induced insulin release and
[Ca2⫹]i increases in mouse islet ␤-cells. When GLP-1 ag-
onist and Kv2.1 channel blocker at subthreshold doses
were combined, they acted synergistically to induce insulin
secretion in normal mice and to correct hyperglycemia
during a glucose tolerance tests in type 2 diabetic db/db
mice and NSY mice.
Materials and Methods
Animals
Wild-type C57BL/6J mice (Japan SLC), Kv2.1-KO mice (9)
(Deltagen, Inc), db/db mice (Japan Clea), and NSY mice (Japan
endo.endojournals.org 115
SLC) were purchased and housed on a 12-hour light, 12-hour
dark cycle in accordance with our institutional guidelines and
with the Japanese Physiological Society’s guidelines for animal
care. The Kv2.1-KO mice were backcrossed onto a C57BL/6J
mice at least for nine generations. The disrupted sequence in
genomic DNA from the KO mice was detected using PCR to
produce an amplicon with one primer inside the targeting se-
quence in combination with a Kv2.1-specific primer. Male age-
matched (10 wk old) Kv2.1 hetero-KO (Kv2.1⫹/⫺) mice and
wild-type littermates as controls were used. All mice were given
free access to rodent normal chow and water. Experimental pro-
tocols for animal studies were approved by our institutional
Downloaded from https://academic.oup.com/endo/article-abstract/156/1/114/2800622 by guest on 30 January 2020
committee on animal care.
Preparation of pancreatic islets and single ␤-cells
Islets of Langerhans were isolated by collagenase digestion, as
reported elsewhere (19) with slight modification. Animals were
anesthetized by an ip injection of pentobarbitone at 80 mg/kg,
followed by an injection of collagenase at 1.05 mg/mL (Sigma-
Aldrich) into the common bile duct. Collagenase was dissolved
in HEPES-added Krebs-Ringer bicarbonate buffer (HKRB) so-
lution (in millimoles): NaCl 129, NaHCO3 5.0, KCl 4.7,
KH2PO4 1.2, CaCl2 2.0, MgSO4 1.2, and HEPES 10 (pH 7.4)
with NaOH, supplemented with 5.6 mM glucose and 0.1% BSA.
The HKRB solution containing 0.1% BSA was used for the mea-
surements of cytosolic Ca2⫹ concentrations ([Ca2⫹]i) and insulin
release but not for the patch-clamp study. The pancreas was
dissected out and incubated at 37°C for 16 minutes. Islets were
hand collected under a microscope and were immediately used
for the measurement of insulin secretion. For ␤-cell experiments,
islets were dispersed into single cells in Ca2⫹-free HKRB, and the
single cells were plated sparsely on coverslips and maintained for
1 day at 37°C in an atmosphere of 5% CO2 and 95% air in
Eagle’s minimal essential medium containing 5.6 mM glucose
supplemented with 10% fetal bovine serum, 100 ␮g/mL strep-
tomycin, and 100 U/mL penicillin.
Measurements of insulin release in mouse islets
For the measurements of the insulin release, groups of 10 islets
were incubated for 1 hour at 37°C in HKRB with 2.8 mM glucose
for stabilization, followed by test incubation for 1 hour in HKRB
with 2.8 or 8.3 mM glucose. Guangxitoxin-1E (GxTx) (Peptide
Institute), exendin-4 (Ex-4) (Sigma-Aldrich), and GxTx with
Ex-4 were present throughout the incubation. Insulin release in
islets was determined by an ELISA kit (Morinaga Institute of
Biological Science).
Patch-clamp experiments in mouse single ␤-cells
Conventional whole-cell currents were recorded using an am-
plifier (Axopatch 200B; Molecular Devices) in a computer using
pCLAMP10.2 software, as reported (20). For the conventional
whole-cell clamp experiments, a pipette solution contained (in
millimoles): KCl 50, K2SO4 35, MgCl2 5, EGTA 11, CaCl2 1,
HEPES 11, and ATP-2Na 5 (pH 7.2) with KOH. HKRB solution
containing 5.6 mM glucose was used for external solution. Single
␤-cells were voltage clamped at a holding potential of ⫺70 mV
and then shifted to the test potentials to 0 mV with the pulses of
100 milliseconds duration for Kv channel currents at room tem-
perature (25°C).
116 Sukma Rita et al Kv2.1 Blockade Enhances Insulinotropic GLP-1 Action
A A
B evoked by a step pulse to 0 mV from a holding potential of ⫺70 mV
C Dissociated single ␤-cells on coverslips were mounted in an
Figure 1. Kv2.1 channel blockade cooperates with Ex-4 to enhance
insulin release in mouse isolated islets. A, In isolated islets from wild-type
mice, glucose (8.3 mM)-induced insulin release was significantly enhanced
by the combination of Ex-4 (30 pM) and GxTx (3 nM), either of which
individually failed to significantly increase the insulin release. High
concentration of GxTx at 1 ␮M alone enhanced the insulin release at 8.3
mM glucose, and Ex-4 (30 pM) failed to potentiate it. B, GxTx (3 nM)
leftward shifted the concentration-response curve of Ex-4 with EC50
values of 127.6 pM for control and 34.1 pM for GxTx, respectively. *, P ⬍
.05; **, P ⬍ .01 vs control (n ⫽ 8 –9 tubes of islets from six mice for batch
incubation). C, In isolated islets from Kv2.1⫹/⫺ mice, the glucose (8.3 mM)-
induced insulin release was enhanced by the subthreshold concentration of
Ex-4 (30 pM). *, P ⬍ .05 vs 8.3 mM glucose in the wild type; #, P ⬍ .05 vs 8.3
mM glucose in Kv2.1⫹/⫺ (n ⫽ 8–10 tubes of batch incubation).
Endocrinology, January 2015, 156(1):114 –123
B
Downloaded from https://academic.oup.com/endo/article-abstract/156/1/114/2800622 by guest on 30 January 2020
Figure 2. ␤-Cell Kv currents are partially blocked by a low
concentration of GxTx (3 nM) and Kv2.1 hetero-KO. Current traces
were measured in a mouse single ␤-cell under a conventional whole-
cell clamp mode. Data were recorded in the presence of 5.6 mM
glucose and 100 ␮M tolbutamide. A, Upper panel, Exposure to low
concentration of GxTx (3 nM) partially decreased the amplitudes of
delayed outward currents. Lower panel, GxTx (3 nM) significantly
decreased peak amplitude of the current density in ␤-cells (n ⫽ 9
single ␤-cells). *, P ⬍ .05. B, Upper panel, Typical traces of Kv currents
in ␤-cells of Kv2.1⫹/⫺ mice and wild-type littermates. Lower panel,
Peak amplitudes of Kv current density were diminished in ␤-cells of
Kv2.1⫹/⫺ mice compared with wild-type littermates.*, P ⬍ .05 (n ⫽ 20
single ␤-cells).
Measurements of [Ca2ⴙ]i in single ␤-cells
open chamber and superfused in HKRB. Cytosolic [Ca2⫹]i in
single ␤-cells were measured at 31°C by dual-wavelength fura-2
microfluorometry with excitation at 340/380 nm and emission at
510 nm using a cooled charge-coupled device camera (19, 21).
The ratio image was produced on an Aquacosmos system
(Hamamatsu Photonics). Data were taken exclusively from the
cells that fulfilled the reported morphological and physiological
criteria of ␤-cells including the diameter and responsiveness to
glucose (8.3 mM) and KATP channel blocker tolbutamide (100
␮M) (17). In some experiments, we measured [Ca2⫹]i in the cells
that failed to respond to 8.3 mM glucose but responded to tol-
butamide. For [Ca2⫹]i measurements, ␤-cells were prepared
from at least three mice in each experiment.
Intraperitoneal glucose tolerance tests and oral
glucose tolerance tests
An ip glucose tolerance tests (IPGTT) were performed with
male Kv2.1⫹/⫺ mice and wild-type littermates (10 wk old), db/db
mice (7 wk old) or NSY mice (20 wk old) fasted overnight, as
previously reported (22). One gram per kilogram of glucose into
db/db mice or 2 g/kg glucose into Kv2.1⫹/⫺ mice, wild-type mice,
and NSY mice was injected ip, followed by blood sampling from
the tail vein. Saline (0.1 mL per 10 g body weight), GxTx (100
nmol/kg, ip), liraglutide (Novo Nordisk) (3 nmol/kg, ip), or li-
raglutide with GxTx was administrated 30 minutes before the
glucose challenge. Liraglutide at doses of 30 ␮g/kg (⬃10 nmol/
kg) and greater is reported to substantially reduce blood glucose
in mice (23). Hence, in the present study, we used 3 nmol/kg as
doi: 10.1210/en.2014-1728 endo.endojournals.org 117

a subthreshold dose. In oral glucose tolerance tests (OGTT), 2 coCard DIA meter (Arkray), and insulin concentrations were
g/kg glucose was orally injected into the wild-type and Kv2.1⫹/⫺ measured using an ELISA kit (Morinaga Institute of Biological
mice. Blood glucose concentrations were measured using a Glu- Science).

A Statistical analysis
Data represent the means ⫾ SEM.
Statistical analyses were performed using
the unpaired or paired Student’s t test or
a one-way ANOVA followed by Bonfer-
roni multiple comparison tests. Values of
P ⬍ .05 were considered statistically sig-
nificant. A ␹2 test was used for compar-

Downloaded from https://academic.oup.com/endo/article-abstract/156/1/114/2800622 by guest on 30 January 2020

ison of incidence of [Ca2⫹]i oscillation in
␤-cells.

Results
Blockade of Kv2.1 channel
enhances exendin-4-induced
insulin release in mouse
isolated islets
Effects of Ex-4 and GxTx on the
insulin release in mouse isolated is-
lets were measured under basal (2.8
mM) and stimulatory (8.3 mM) glu-
cose conditions. Insulin release from
islets under batch-incubation condi-
tions was larger with 8.3 mM glu-
cose than 2.8 mM glucose (P ⬍ .01,
Figure 1A). The glucose-induced in-
sulin release tended to be enhanced
by a low concentration of either Ex-4
(30 pM) or GxTx (3 nM) but was not
statistically significant (Figure 1A).
Combination of Ex-4 (30 pM) and
GxTx (3 nM) significantly enhanced
the glucose (8.3 mM)-induced insu-
lin release, without affecting the in-
sulin release at 2.8 mM glucose (Fig-
ure 1A). Furthermore, GxTx (3 nM)
enhanced the insulinotropic action
of Ex-4 at 30 –100 pM, whereas it
did not affect the effect of Ex-4 at
Figure 3. Kv2.1 channel blockade cooperates with Ex-4 to induce [Ca2⫹]i increases in mouse higher concentrations of 1–10 nM,
␤-cells. Representative traces of [Ca2⫹]i in single ␤-cell are expressed by dual-wavelength fura-2
fluorescence ratio (F340/F380). A and B, Either Ex-4 (30 pM) or GxTx (3 nM) slightly increased showing a leftward shift of the con-
[Ca2⫹]i in ␤-cell at 8.3 mM glucose. C and D, The substimulatory concentrations of Ex-4 (30 pM) centration-response curve of Ex-4
and GxTx (3 nM) in combination at 8.3 mM glucose evoked marked increases in [Ca2⫹]i (C), with EC50 values of 127.6 pM for
which were completely inhibited by nifedipine (Nif) (10 ␮M) (D). E, The cooperative effect of Ex-4
control and 34.1 pM for GxTx, re-
(30 pM) and GxTx (3 nM) was not observed at basal 2.8 mM glucose. F, AUC of [Ca2⫹]i increase
was significantly enhanced by a combination of Ex-4 and GxTx in a glucose-dependent manner, spectively (Figure 1B). These results
and this enhancement was completely inhibited by Nif. G, Long-term treatment of cells with Ex-4 indicate that GxTx potentiates the
(30 pM) and GxTx (3 nM) in combination elicited [Ca2⫹]i oscillations in ␤-cells at 8.3 mM glucose.
Ex-4 action to promote glucose-in-
H, Incidence of ␤-cells with [Ca2⫹]i oscillations was increased by the combination treatment. The
numbers above each column indicates the number of ␤-cells with [Ca2⫹]i oscillations over that duced insulin release. GxTx at a
examined. *, P ⬍ .05 vs 8.3 mM glucose (n ⫽ 60 –104 single ␤-cells). higher concentration of 1 ␮M by it-
118 Sukma Rita et al Kv2.1 Blockade Enhances Insulinotropic GLP-1 Action Endocrinology, January 2015, 156(1):114 –123

self markedly enhanced the 8.3-mM glucose-induced
insulin release, and Ex-4 (30 pM) failed to further po-
tentiate it (Figure 1A), suggesting that Ex-4 (30 pM)
partially inhibits Kv2.1 channels as one of its action
mechanisms.
Exendin-4-induced insulin release is enhanced in
isolated islets from Kv2.1ⴙ/ⴚ mice
In isolated islets from Kv2.1⫹/⫺ mice, whose Kv cur-
rents in ␤-cells were reduced by 26%, the glucose (8.3
islets, the addition of GxTx at 30 nM failed to further
increase the glucose-induced insulin release (data not
shown), confirming that GxTx promoted insulin release
primarily via suppressing Kv2.1 channels (24).
␤-Cell Kv currents are partially blocked by a low
concentration of GxTx (3 nM) and Kv2.1 hetero-KO
The Kv channel currents in mouse ␤-cells under con-
ventional whole-cell clamp were measured in the presence
of 100 ␮M tolbutamide to inhibit the KATP channel and

Downloaded from https://academic.oup.com/endo/article-abstract/156/1/114/2800622 by guest on 30 January 2020

mM)-induced insulin release was significantly greater
than that from the wild-type littermate (Figure 1C),
whereas the basal levels of insulin release at 2.8 mM glu-
cose were not altered. The glucose (8.3 mM)-induced in-
sulin release in Kv2.1⫹/⫺ islets was further elevated by a
low concentration of Ex-4 (30 pM), which did not signif-
icantly alter the glucose-induced insulin release in wild-
type islets (Figure 1C), supporting the potentiation of Ex-4
effects by blockade of Kv2.1 channels. Basal insulin re-
lease at 2.8 mM glucose was not affected by Ex-4 (30 pM)
in wild-type and Kv2.1⫹/⫺ islets (Figure 1C). In Kv2.1⫺/⫺
A B
thereby exclude involvement of this channel in the cur-
rents. In the presence of 5.6 mM glucose, a constant de-
polarizing pulse from ⫺70 to 0 mV every 20 seconds
evoked outward Kv currents, and the currents decreased
by 14% upon exposure to GxTx (3 nM) (Figure 2A). At a
holding potential of 0 mV, GxTx (3 nM) significantly de-
creased the current densities to 28.8 ⫾ 2.5 pA/pF from
33.6 ⫾ 3.3 pA/pF (P ⬍ .05, n ⫽ 9). In the ␤-cells of
Kv2.1⫹/⫺ mice, Kv currents were diminished by 26% com-
pared with wild-type littermates (Figure 2B). At a holding
potential of 0 mV, the current densities in the ␤-cells of
Kv2.1⫹/⫺ and wild-type mice were
25.2 ⫾ 2.5 pA/pF and 34.2 ⫾ 2.8 pA/
pF, respectively (P ⬍ .05, n ⫽ 20).

Blockade of Kv2.1 channels

potentiates Ex-4-induced [Ca2ⴙ]i
increases in mouse ␤-cells
Effects of the GLP-1 analog, the
Kv2.1 channel blocker, and their
combination were examined on
␤-cell [Ca2⫹]i, a principal mediator
of insulin release. A rise in the glu-
cose concentration from 2.8 to 8.3
mM elicited [Ca2⫹]i increases in a bi-
phasic manner, transient large in-
C D E creases followed by sustained mod-
erate increases. During the period
with the sustained [Ca2⫹]i increases
in the presence of 8.3 mM glucose,
Ex-4 (30 pM) (Figure 3A), and GxTx
at 3 nM that reduced ␤-cell Kv cur-
rents by 14% were added (Figure
3B). These treatments tended to in-
crease [Ca2⫹]i in mouse single
␤-cells. However, the area under the
Figure 4. Ex-4-induced [Ca2⫹]i increases are augmented in ␤-cells from Kv2.1⫹/⫺ mice. A and B,
At 8.3 mM glucose, low concentration of Ex-4 (30 pM) evoked a large increase in [Ca2⫹]i in
curve (AUC), which corresponds to
Kv2.1⫹/⫺ ␤-cells (B), in contrast to a slight increase in [Ca2⫹]i in wild-type ␤-cells (A). C and D, an integrated [Ca2⫹]i increase for 3
AUC of 8.3 mM glucose-induced [Ca2⫹]i increases for 20 minutes in Kv2.1⫹/⫺ ␤-cells was minutes during stimulation, did not
significantly greater than that in wild-type ␤-cells (D), whereas the basal [Ca2⫹]i levels at 2.8 mM
show statistically significant differ-
glucose was identical (C). E, AUC of Ex-4-induced [Ca2⫹]i increases in Kv2.1⫹/⫺ ␤-cells was
significantly greater than that in wild-type ␤-cells. *, P ⬍ .05; **, P ⬍ .01 vs wild-type (n ⫽ 55– ence (Figure 3F) with either Ex-4 or
60 single ␤-cells). GxTx.
doi: 10.1210/en.2014-1728 endo.endojournals.org 119

A B C Ex-4-induced [Ca2ⴙ]i increase is

augmented in ␤-cells from
Kv2.1ⴙ/ⴚ mice
To support the potentiation of
[Ca2⫹]i increase by a combined treat-
ment with a Kv2.1 channel blocker
and GLP-1 agonist, we next exam-
ined effects of Ex-4 on the ␤-cell
[Ca2⫹]i from Kv2.1⫹/⫺ mice in
which the gene coding Kv2.1 is par-

Downloaded from https://academic.oup.com/endo/article-abstract/156/1/114/2800622 by guest on 30 January 2020

tially deleted and ␤-cell Kv currents
D E F were reduced by 26%. Glucose (8.3
mM)-induced [Ca2⫹]i increase was
moderately enhanced in Kv2.1⫹/⫺
␤-cells compared with those in wild-
type ␤-cells (Figure 4D), whereas
basal [Ca2⫹]i at 2.8 mM glucose was
identical (Figure 4C). In the presence
of 8.3 mM glucose, Ex-4 (30 pM)
only slightly increased [Ca2⫹]i in
Figure 5. Kv2.1 channel blockade and Ex-4 cooperatively induce [Ca2⫹]i increases in a PKA- wild-type ␤-cells (Figure 4A) but
dependent manner in mouse ␤-cells. A and B, Synergistic action of Ex-4 and GxTx to increase
evoked large [Ca2⫹]i increases in
[Ca2⫹]i at 8.3 mM glucose (A) was attenuated by the pretreatment with a PKA inhibitor H-89
(10 ␮M) (B). C, AUC of [Ca2⫹]i increases by Ex-4 and GxTx was significantly inhibited by H- Kv2.1⫹/⫺ ␤-cells (Figure 4B). The
89. D, PKA activator, Phe-cAMP (3 ␮M), slightly increased [Ca2⫹]i and the Epac activator, AUC of Ex-4-induced [Ca2⫹]i in-
CPT-cAMP (3 ␮M), little increased [Ca2⫹]i in the ␤-cells. E, In the presence of GxTx (3 nM), creases was significantly augmented
the PKA activator evoked marked increases in [Ca2⫹]i, whereas the Epac activator had little
effect on [Ca2⫹]i in ␤-cells. F, AUC of [Ca2⫹]i increases was significantly elevated by Phe- in Kv2.1⫹/⫺, compared with wild-
cAMP with GxTx but not by CPT-cAMP with GxTx. *, P ⬍ .05 vs control of 8.3 mM glucose type, ␤-cells (Figure 4E).
(n ⫽ 48 –56 single ␤-cells).
Kv2.1 channel blockade and Ex-
In contrast, when Ex-4 and GxTx at low concentra- 4 cooperatively increase [Ca2ⴙ]i
tions at which each agent alone was ineffective were com- in a protein kinase A (PKA)-dependent manner in
bined, they acted synergistically to evoke marked [Ca2⫹]i ␤-cells
increases (Figure 3C) with significant increases in the AUC Activation of GLP-1 receptor is linked to cAMP sig-
of [Ca2⫹]i (Figure 3F). Higher concentrations of each drug naling and leads to the activation of PKA and exchange
alone, Ex-4 at 1 nM and GxTx at 100 nM, significantly proteins directly activated by cAMP (Epac). The synergis-
increased [Ca2⫹]i at 8.3 mM glucose (data not shown). tic action of Ex-4 and GxTx to increase [Ca2⫹]i was at-
Nifedipine (10 ␮M), an L-type Ca2⫹ channel blocker, tenuated by pretreatment with a PKA inhibitor, H-89 (10
completely inhibited the [Ca2⫹]i increase induced by the ␮M) (Figure 5, A and B), with a significant reduction of the
combination of Ex-4 (30 pM) and GxTx (3 nM), suggest-
AUC of [Ca2⫹]i increases (Figure 5C). Furthermore, a low
ing that this combination enhanced Ca2⫹ influx through
concentration of 6-Phe-cAMP (Phe-cAMP) (3 ␮M), a PKA
voltage-dependent L-type Ca2⫹ channels (Figure 2, D and
activator, only slightly increased [Ca2⫹]i in ␤-cells (Figure
F). In 2.8 mM glucose, the combination of Ex-4 and GxTx
5D). In the presence of GxTx (3 nM), however, this PKA
had no effect on [Ca2⫹]i in ␤-cells (Figure 3, E and F).
activator induced marked increases in [Ca2⫹]i (Figure 5, E
Furthermore, long-term treatment of cells with Ex-4 (30
pM) and GxTx (3 nM) in combination elicited [Ca2⫹]i and F), showing a synergistic effect of PKA activator and
oscillations in many ␤-cells at 8.3 mM glucose (Figure GxTx. In contrast, synergistic effect was not observed
3G), and the incidence of ␤-cells with [Ca2⫹]i oscillations between GxTx and an Epac activator 8-pCPT-2⬘-O-
was significantly increased by the combination treatment Me-cAMP-AM (CPT-cAMP) (3 ␮M) (Figure 5, E and
(Figure 3H). The combination of Ex-4 (30 pM) and GxTx F). These results suggest that the blockade of Kv2.1
(3 nM) did not evoke [Ca2⫹]i increases at 8.3 mM glucose channels may interact with the GLP-1 receptor-oper-
in the cells that did not respond to 8.3 mM glucose but ated PKA signaling route, thereby inducing [Ca2⫹]i in-
responded to tolbutamide (data not shown). creases in ␤-cells.
120 Sukma Rita et al Kv2.1 Blockade Enhances Insulinotropic GLP-1 Action Endocrinology, January 2015, 156(1):114 –123

significantly suppressed the blood

A B
glucose increases at 60 and 120 min-
utes and enhanced insulin responses
at 15 minutes during IPGTT (Figure
6, A and B). In the OGTT, further-
more, the Kv2.1⫹/⫺ mice exhibited
improved glucose tolerance at 60
minutes (Figure 6C) and increased
plasma insulin levels at 15 minutes
after glucose challenge (Figure 6D),

Downloaded from https://academic.oup.com/endo/article-abstract/156/1/114/2800622 by guest on 30 January 2020

compared with wild-type mice. This
result indicates that Kv2.1 channel
regulates the insulinotropic actions
of endogenous GLP-1 and possibly
C D other incretin hormones released by
an oral glucose challenge.
We next conducted an IPGTT us-
ing GxTx and liraglutide in type 2
diabetic db/db mice and NSY mice.
After overnight fasting, either GxTx
or liraglutide, or their combination,
was administered to db/db mice aged
7 weeks or NSY mice aged 20 weeks
30 minutes before a glucose chal-
lenge. A single administration of ei-
ther GxTx (100 nmol/kg, ip) or lira-
glutide (3 nmol/kg, ip) did not
Figure 6. Genetic blockade of Kv2.1 channel and GLP-1 agonist cooperatively improves glucose significantly suppress blood glucose
tolerance and increases the plasma insulin level. A and B, In an IPGTT, after overnight fasting, 2
g/kg glucose was ip injected into wild-type and Kv2.1 ⫹/⫺
mice, whereas test agents were
increases in IPGTT. In contrast, a
administered 30 minutes before the glucose injection. In Kv2.1 ⫹/⫺
mice, a low dose (3 nmol/kg, combination of GxTx and liraglu-
ip) of liraglutide, which was ineffective in wild-type mice, significantly suppressed blood glucose tide significantly suppressed blood
increases at 60 and 120 minutes and enhanced insulin responses at 15 minutes of IPGTT. *, P ⬍
glucose increases in both db/db mice
.05; **, P ⬍ .01 vs control (n ⫽ 8 –10). C and D, In an OGTT, after overnight fasting, 2 g/kg
glucose was orally injected into mice. Kv2.1 ⫹/⫺
mice exhibited improved glucose tolerance at 60 (Figure 7A) and NSY mice (Figure
minutes (C) and increased plasma insulin levels at 15 minutes after a glucose challenge (D), 7C). During the IPGTT, plasma in-
compared with wild-type mice. *, P ⬍ .05; **, P ⬍ .01 vs wild-type (n ⫽ 7). sulin levels exhibited no significant
increase in response to either GxTx
Genetic and pharmacological blockades of Kv2.1 or liraglutide. GxTx and liraglutide in combination, how-
channel and GLP-1 agonist cooperatively improve ever, exhibited a magnificent increases in plasma insulin
glucose tolerance and increase plasma insulin levels at 15 and 30 minutes after a glucose challenge, com-
To assess the effect of Kv2.1 blockade and GLP-1 ag- pared with the saline-injected control db/db (Figure 7B)
onist in combination in vivo, we examined the effects of and NSY mice groups (Figure 7D).
liraglutide, a long-acting GLP-1 agonist, in Kv2.1⫹/⫺
mice. In Kv2.1⫹/⫺ mice after overnight fasting, increases in
blood glucose levels at 15–120 minutes and plasma insulin Discussion
responses at 15 and 30 minutes of IPGTT were not sig-
nificantly different from those of wild-type mice (Figure 6, The present study demonstrated that the GLP-1 receptor
A and B), although insulin responses at 15 and 30 minutes agonist Ex-4 and Kv2.1 channel blocker GxTx at sub-
tended to be slightly enhanced. In wild-type mice, the threshold concentrations, when combined, markedly in-
administration of liraglutide (3 nmol/kg, ip) failed to sig- creased insulin release in islets and evoked [Ca2⫹]i in-
nificantly alter blood glucose and plasma insulin levels creases in ␤-cells in a glucose-dependent manner. These
during IPGTT (Figure 6, A and B). In contrast, the admin- results indicate that the GLP-1 agonist and Kv2.1 channel
istration of this low dose of liraglutide in Kv2.1⫹/⫺ mice blocker act synergistically to activate ␤-cells and induce
doi: 10.1210/en.2014-1728 endo.endojournals.org 121

A B possibly other incretin hormones

and that blockade of Kv2.1 chan-
nels provides a novel tool to enhance
the insulinotropic potency of GLP-
1-based drugs.
GLP-1 agonists have been shown
to improve glucose intolerance in
type 2 diabetic patients (25) with a
lower risk of hypoglycemia than
sulfonylureas. It has been shown,

Downloaded from https://academic.oup.com/endo/article-abstract/156/1/114/2800622 by guest on 30 January 2020

however, that GLP-1 agonists often
causes gastrointestinal adverse events
such as nausea and vomiting and that
their incidence is dose dependent
C D (14, 15). The present study showed
that the use of Kv2.1 channel
blocker, even at subthreshold con-
centrations, markedly reduces the
dose of GLP-1 agonists required for
their insulinotropic and blood glu-
cose-lowering effects. This finding
opens a new clinical avenue to reduce
the dose and associated adverse ef-
fects of GLP-1 agonists while achiev-
ing the same glycemic control in type
2 diabetes. It has been reported that
both the plasma GLP-1 level and in-
Figure 7. GxTx and liraglutide in combination improve glucose tolerance and increase plasma
sulinotropic GLP-1 action are re-
insulin in type 2 diabetic model mice. In an IPGTT, after overnight fasting, 1 g/kg glucose was ip duced in type 2 diabetic patients,
injected into db/db mice and 2 g/kg glucose to NSY mice, and test agents were administered 30 contributing to impaired insulin re-
minutes before the glucose injection. A and B, In db/db mice, GxTx (100 nmol/kg, ip) and
lease and hyperglycemia (26 –28). In
liraglutide (3 nmol/kg, ip) individually failed to significantly alter blood glucose (A) and plasma
insulin levels (B), but when combined, they significantly improved glucose tolerance at 30, 60, these cases, DPP-4 inhibitors could
and 120 minutes (A) and increased plasma insulin levels at 15 and 30 minutes after a glucose restore the plasma GLP-1 level and
challenge (B), compared with the saline-injected control group. C and D, In NSY mice, GxTx (100 Kv2.1 channel blocker could restore
nmol/kg, ip) and liraglutide (3 nmol/kg, ip) individually failed to significantly alter blood glucose
(C) and plasma insulin levels (D). When combined, they significantly improved glucose tolerance the GLP-1 action, and these two pro-
at 30, 60, and 120 minutes (C) and increased plasma insulin at 15 and 30 minutes (D), compared cesses could collaborate to augment
with saline-injected control. *, P ⬍ .05; **, P ⬍ .01 vs control (n ⫽ 10). the eventual insulinotropic GLP-1
effect in islet ␤-cells. Thus, the clin-
insulin secretion. Moreover, Kv2.1 channel blockade and ically applicable Kv2.1 channel blocker, if available in the
GLP-1 agonist cooperatively improve glucose tolerance future, could amplify the potency of DPP-4 inhibitors. We
and increase plasma insulin in vivo. A low dose of lira- here propose a novel strategy to intensify the therapy of
glutide, which was ineffective in wild-type mice, signifi- GLP-1 agonists and DPP-4 inhibitors by using Kv2.1
cantly suppressed blood glucose increases and enhanced channel blockers. In addition, DPP-4 inhibitors are known
insulin responses to IPGTT in Kv2.1⫹/⫺ mice. Kv2.1⫹/⫺ to elevate the circulating levels of not only GLP-1 but also
mice further exhibited improved glucose tolerance and other incretin hormones, which include glucose-depen-
increased plasma insulin levels during an OGTT com- dent insulinotropic polypeptide and pituitary adenylate
pared with wild-type mice. Subthreshold doses of GxTx cyclase-activating polypeptide (29, 30) that act through
and GLP-1 receptor agonist liraglutide in combination cAMP signaling in islet ␤-cells. The insulinotropic effects
markedly increased plasma insulin and improved glucose of these hormones could also be enhanced by blockade of
tolerance in diabetic db/db mice and NSY mice. These Kv2.1 channels. This possibility remains to be studied.
results indicate that Kv2.1 channels physiologically The effects of GLP-1 are mediated by the stimulation of
limit insulinotropic actions of endogenous GLP-1 and adenylate cyclase and a subsequent increase in intracellu-
122 Sukma Rita et al Kv2.1 Blockade Enhances Insulinotropic GLP-1 Action
lar cAMP, which lead to the activation of the PKA and/or
Epac pathways (31–33). Our present study indicates that
the synergistic effect of substimulatory concentrations of
GxTx and Ex-4 to increase [Ca2⫹]i in ␤-cells was attenu-
ated by pretreatment with a PKA inhibitor H-89. Further-
more, GxTx together with a substimulatory concentration
of PKA activator Phe-cAMP (3 ␮M), but not the Epac
activator CPT-cAMP (3 ␮M), evoked [Ca2⫹]i increases in
␤-cells. These results suggest that the Kv2.1 channel block-
ade interacts with the PKA, but not Epac, branch of cAMP
signaling in ␤-cells. Additionally, it is possible that the
relatively low concentrations of GLP-1 agonists used in
our study produces cAMP to a marginal level incapable of
activating Epac because higher concentrations of cAMP
are required for activating the Epac pathway (34, 35).
However, our data cannot exclude an additional in-
volvement of Epac signaling because interplay of PKA
and Epac was suggested to be implicated in a cAMP-
dependent stimulation of [Ca2⫹]i increases in ␤-cells
and insulin secretion (36).
Regarding the further mechanism for interaction be-
tween PKA and the Kv channel, we have previously re-
ported that a PKA activator Phe-cAMP attenuates ␤-cell
Kv currents (37). Consistent with our observations, the
activation of the GLP-1 receptor phosphorylates Kv2.1 in
␤-cells via PKA (38) and reduces the Kv currents to an-
tagonize ␤-cell repolarization in a cAMP/PKA-dependent
manner (18, 39). Under these conditions, additional treat-
ment of cells with substimulatory concentration of GxTx
may further inhibit the Kv2.1 channels and induce a
greater depolarization, leading to enhanced Ca2⫹ influx
and insulin secretory response. As an additional or alter-
native target process in ␤-cells, cAMP signal-mediated
promotion of insulin exocytosis (16) and/or facilitation of
L-type Ca2⫹ channel activities (40) could be involved,
which remains to be established. On the other hand, we
have reported that transient receptor potential melastatin
2 (TRPM2), a nonselective cation channel, mediates the
GLP-1 signal to promote insulin release (41, 42). Hence,
it is possible that the activation of TRPM2 channels by
GLP-1 could cooperate with the blockade of Kv2.1 chan-
nels, resulting in the effective depolarization of ␤-cells.
This potential implication of the TRPM2 channels in the
synergistic effects of GLP-1 agonist and GxTx remains to
be determined.
In conclusion, the combination of low doses of a GLP-1
agonist and Kv2.1 channel blocker had strong in vivo in-
sulinotropic and blood glucose-lowering effects in type 2
diabetic mice. Blockade of the Kv2.1 channels enhances
the potency of GLP-1-based drugs, which can lower their
dose required to improve glucose tolerance in type 2 dia-
betes and thereby reduce the frequency of their adverse
Endocrinology, January 2015, 156(1):114 –123
effects. Our study provides a novel avenue to make the
GLP-1-based diabetes therapy more effective and reliable.
Acknowledgments
We thank S. Ookuma, Y. Hobo, M. Warashina, and C. Sakamoto
(Jichi Medical University) for technical assistance.
This study was subsidized by the JKA through its promotion
funds from KEIRIN RACE (to T.Y.).
Downloaded from https://academic.oup.com/endo/article-abstract/156/1/114/2800622 by guest on 30 January 2020
Address all correspondence and requests for reprints to:
Katsuya Dezaki, PhD, Associate Professor of Physiology, Jichi
Medical University, Yakushiji 3311-1, Shimotsuke, Japan.
E-mail: dezaki@jichi.ac.jp.
Other correspondence should be addressed to the following:
Katsuya Dezaki, E-mail: dezaki@jichi.ac.jp; and Toshihiko
Yada. E-mail: tyada@jichi.ac.jp.
This work was supported by a Grant-in-Aid for Scientific
Research (B) (to T.Y.), (C) (to K.D.), and for Challenging Ex-
ploratory Research (to T.Y.) from the Japan Society for the Pro-
motion of Science; Strategic Research Program for Brain Sciences
(Grant 10036069) by the Ministry of Education, Culture, Sports,
Science, and Technology of Japan (to T.Y.); Ministry of Educa-
tion, Culture, Sports, Science, and Technology of Japan-Sup-
ported Programs for the Strategic Research Foundation at pri-
vate universities in 2011–2015 and 2013–2017 (to T.Y.); by
grants from the Japan Diabetes Foundation, Takeda Science
Foundation, and Salt Science Research Foundation (to K.D. and
T.Y.).
Disclosure Summary: The authors have nothing to disclose.
References
1. Roe MW, Worley JF 3rd, Mittal AA, et al. Expression and function
of pancreatic ␤-cell delayed rectifier K⫹ channels. Role in stimulus-
secretion coupling. J Biol Chem. 1996;271:32241–32246.
2. MacDonald PE, Ha XF, Wang J, et al. Members of the Kv1 and Kv2
voltage-dependent K⫹ channel families regulate insulin secretion.
Mol Endocrinol. 2001;15:1423–1435.
3. Yan L, Figueroa DJ, Austin CP, et al. Expression of voltage-gated
potassium channels in human and rhesus pancreatic islets. Diabetes.
2004;53:597– 607.
4. MacDonald PE, Sewing S, Wang J, et al Inhibition of Kv2.1 voltage-
dependent K⫹ channels in pancreatic ␤-cells enhances glucose-de-
pendent insulin secretion. J Biol Chem. 2002;277:44938 – 44945.
5. MacDonald PE, Wheeler MB. Voltage-dependent K⫹ channels in
pancreatic ␤ cells: role, regulation and potential as therapeutic tar-
gets. Diabetologia. 2003;46:1046 –1062.
6. Herrington J, Sanchez M, Wunderler D, et al. Biophysical and phar-
macological properties of the voltage-gated potassium current of
human pancreatic ␤-cells. J Physiol. 2005;567:159 –175.
7. Herrington J, Zhou YP, Bugianesi RM, et al. Blockers of the delayed-
rectifier potassium current in pancreatic ␤-cells enhance glucose-
dependent insulin secretion. Diabetes. 2006;55:1034 –1042.
8. Tamarina NA, Kuznetsov A, Fridlyand LE, Philipson LH. Delayed-
rectifier (KV2.1) regulation of pancreatic ␤-cell calcium responses to
glucose: inhibitor specificity and modeling. Am J Physiol Endocrinol
Metab. 2005;289:E578 –E585.
doi: 10.1210/en.2014-1728
9. Jacobson DA, Kuznetsov A, Lopez JP, Kash S, Ammälä CE, Philip-
son LH. Kv2.1 ablation alters glucose-induced islet electrical activ-
ity, enhancing insulin secretion. Cell Metab. 2007;6:229 –235.
10. Jensen MV, Haldeman JM, Zhang H, et al. Control of voltage-gated
potassium channel Kv2.2 expression by pyruvate-isocitrate cycling
regulates glucose-stimulated insulin secretion. J Biol Chem. 2013;
288:23128 –23140.
11. Zhang YX, Liu Y, Dong J, et al. An exploratory study of the asso-
ciation between KCNB1 rs1051295 and type 2 diabetes and its re-
lated traits in Chinese Han population. PLoS One. 2013;8:e56365.
12. Drucker DJ. Glucagon-like peptides. Diabetes. 1998;47:159 –169.
13. Drucker DJ. Minireview: the glucagon-like peptides. Endocrinol-
ogy. 2001;142:521–527.
14. Garber A, Henry R, Ratner R, et al. Liraglutide versus glimepiride
monotherapy for type 2 diabetes (LEAD-3 Mono): a randomised,
52-week, phase III, double-blind, parallel-treatment trial. Lancet.
2009;373:473– 481.
15. Nauck M, Frid A, Hermansen K, et al. Efficacy and safety compar-
ison of liraglutide, glimepiride, and placebo, all in combination with
metformin, in type 2 diabetes: the LEAD (liraglutide effect and ac-
tion in diabetes)-2 study. Diabetes Care. 2009;32:84 –90.
16. Shibasaki T, Takahashi H, Miki T, et al. Essential role of Epac2/
Rap1 signaling in regulation of insulin granule dynamics by cAMP.
Proc Natl Acad Sci USA. 2007;104:19333–19338.
17. Yada T, Itoh K, Nakata M. Glucagon-like peptide-1-(7–36)amide
and a rise in cyclic adenosine 3⬘,5⬘-monophosphate increase cyto-
solic free Ca2⫹ in rat pancreatic ␤-cells by enhancing Ca2⫹ channel
activity. Endocrinology. 1993;133:1685–1692.
18. MacDonald PE, Salapatek AM, Wheeler MB. Glucagon-like pep-
tide-1 receptor activation antagonizes voltage-dependent repolariz-
ing K⫹ currents in ␤-cells: a possible glucose-dependent insulino-
tropic mechanism. Diabetes. 2002;51:S443–S447.
19. Dezaki K, Hosoda H, Kakei M, et al. Endogenous ghrelin in pan-
creatic islets restricts insulin release by attenuating Ca2⫹ signaling in
␤-cells: implication in the glycemic control in rodents. Diabetes.
2004;53:3142–3151.
20. Yoshida M, Dezaki K, Yamato S, et al. Regulation of voltage-gated
K⫹ channels by glucose metabolism in pancreatic ␤-cells. FEBS Lett.
2009;583:2225–2230.
21. Dezaki K, Kakei M, Yada T. Ghrelin uses G␣i2 and activates voltage-
dependent K⫹ channels to attenuate glucose-induced Ca2⫹ signaling
and insulin release in islet ␤-cells: novel signal transduction of ghre-
lin. Diabetes. 2007;56:2319 –2327.
22. Dezaki K, Sone H, Koizumi M, et al. Blockade of pancreatic islet-
derived ghrelin enhances insulin secretion to prevent high-fat diet-
induced glucose intolerance. Diabetes. 2006;55:3486 –3493.
23. Rolin B, Larsen MO, Gotfredsen CF, et al. The long-acting GLP-1
derivative NN2211 ameliorates glycemia and increases ␤-cell mass
in diabetic mice. Am J Physiol Endocrinol Metab. 2002;283:E745–
E752.
24. Li XN, Herrington J, Petrov A, et al. The role of voltage-gated po-
tassium channels Kv2.1 and Kv2.2 in the regulation of insulin and
somatostatin release from pancreatic islets. J Pharmacol Exp Ther.
2013;344:407– 416.
endo.endojournals.org 123
25. Perry CM. Liraglutide: a review of its use in the management of type
2 diabetes mellitus. Drugs. 2011;71:2347–2373.
26. Nauck M, Stöckmann F, Ebert R, Creutzfeldt W. Reduced incretin
effect in type 2 (non-insulin-dependent) diabetes. Diabetologia.
1986;29:46 –52.
27. Kjems LL, Holst JJ, Vølund A, Madsbad S. The influence of GLP-1
on glucose-stimulated insulin secretion: effects on ␤-cell sensitivity
in type 2 and nondiabetic subjects. Diabetes. 2003;52:380 –386.
28. Vilsbøll T, Krarup T, Deacon CF, Madsbad S, Holst JJ. Reduced
postprandial concentrations of intact biologically active glucagon-
like peptide 1 in type 2 diabetic patients. Diabetes. 2001;50:609 –
613.
29. Ahrén B. DPP-4 inhibition and islet function. J Diabetes Investig.
Downloaded from https://academic.oup.com/endo/article-abstract/156/1/114/2800622 by guest on 30 January 2020
2012;3:3–10.
30. Omar B, Ahrén B. Pleiotropic mechanisms for the glucose-lowering
action of DPP-4 inhibitors. Diabetes. 2014;63:2196 –2202.
31. Drucker DJ. The biology of incretin hormones. Cell Metab. 2006;
3:153–165.
32. Holst JJ. The physiology of glucagon-like peptide 1. Physiol Rev.
2007;87:1409 –1439.
33. Damdindorj B, Dezaki K, Kurashina T, et al. Exogenous and en-
dogenous ghrelin counteracts GLP-1 action to stimulate cAMP sig-
naling and insulin secretion in islet ␤-cells. FEBS Lett. 2012;586:
2555–2562.
34. Enserink JM, Christensen AE, de Rooij J, et al. A novel Epac-specific
cAMP analogue demonstrates independent regulation of Rap1 and
ERK. Nat Cell Biol. 2002;4:901–906.
35. de Rooij J, Rehmann H, van Triest M, Cool RH, Wittinghofer A,
Bos JL. Mechanism of regulation of the Epac family of cAMP-de-
pendent RapGEFs. J Biol Chem. 2000;275:20829 –20836.
36. Song WJ, Mondal P, Li Y, Lee SE, Hussain MA. Pancreatic ␤-cell
response to increased metabolic demand and to pharmacologic
secretagogues requires EPAC2A. Diabetes. 2013;62:2796 –2807.
37. Dezaki K, Damdindorj B, Sone H, et al. Ghrelin attenuates cAMP-
PKA signaling to evoke insulinostatic cascade in islet ␤-cells. Dia-
betes. 2011;60:2315–2324.
38. Kim SJ, Widenmaier SB, Choi WS, et al. Pancreatic ␤-cell prosur-
vival effects of the incretin hormones involve post-translational
modification of Kv2.1 delayed rectifier channels. Cell Death Differ.
2012;19:333–344.
39. MacDonald PE, Wang X, Xia F, et al. Antagonism of rat ␤-cell
voltage-dependent K⫹ currents by exendin 4 requires dual activation
of the cAMP/protein kinase A and phosphatidylinositol 3-kinase
signaling pathways. J Biol Chem. 2003;278:52446 –52453.
40. Suga S, Kanno T, Nakano K, Takeo T, Dobashi Y, Wakui M. GLP-
I(7–36) amide augments Ba2⫹ current through L-type Ca2⫹ channel
of rat pancreatic ␤-cell in a cAMP-dependent manner. Diabetes.
1997;46:1755–1760.
41. Uchida K, Dezaki K, Damdindorj B, et al. Lack of TRPM2 impaired
insulin secretion and glucose metabolisms in mice. Diabetes. 2011;
60:119 –126.
42. Yoshida M, Dezaki K, Uchida K, et al. Involvement of cAMP/EPAC/
TRPM2 activation in glucose- and incretin-induced insulin secre-
tion. Diabetes. 2014;63:3394 –3403.