Original Article

Contribution of Calcium Influx in Mediating

Glucose-Stimulated Oxygen Consumption in
Pancreatic Islets
Ian R. Sweet1 and Merle Gilbert1

In brain, muscle, and pancreatic islets, depolarization induces an increase in respiration, which is dependent on
calcium influx. The goal of this study was to assess the
quantitative significance of this effect in islets relative to
glucose-stimulated ATP turnover, to examine the molecular
mechanism mediating the changes, and to investigate the
functional implications with respect to insulin secretion.
Glucose (320 mmol/l) increased steady-state levels of cytochrome c reduction (32 66%) in isolated rat islets, reflecting
an increased production of NADH, and oxygen consumption
rate (OCR) by 0.32 nmol/min/100 islets. Glucose-stimulated
OCR was inhibited 30% by inhibitors of calcium influx (diazoxide or nimodipine), whereas a protein synthesis inhibitor
(emetine) decreased it by only 24%. None of the inhibitors
affected cytochrome c reduction, suggesting that calciums
effect on steady-state OCR is mediated by changes in ATP
usage rather than the rate of NADH generation. 3-isobutyl-1methylxanthine increased insulin secretion but had little
effect on OCR, indicating that the processes of movement and
exocytosis of secretory granules do not significantly contribute to ATP turnover. At 20 mmol/l glucose, a blocker of
sarcoendoplasmic reticulum calcium ATPase (SERCA) had
little effect on OCR despite a large increase in cytosolic
calcium, further supporting the notion that influx of calcium,
not bulk cytosolic calcium, is associated with the increase in
ATP turnover. The glucose dose response of calcium influx
dependent OCR showed a remarkable correlation with insulin secretion, suggesting that the process mediating the effect
of calcium on ATP turnover has a role in the amplification
pathway of insulin secretion. Diabetes 55:3509 3519, 2006

ytosolic calcium is a major mediator governing

the amount of insulin released in response to a
glucose challenge (1 4). The effect of calcium
in mediating energy turnover has been a point
of particular interest due to the role of the metabolic

From the 1Robert H. Williams Laboratory, Department of Medicine, University

of Washington, Seattle, Washington.
Address correspondence and reprint requests to Ian R. Sweet, Robert H.
Williams Laboratory, HSB K-165, Box 357710, University of Washington, 1959
NE Pacific St., Seattle, WA 98195-7710. E-mail: isweet@u.washington.edu.
Received for publication 25 March 2006 and accepted in revised form 11
September 2006.
CPA, cyclopiazonic acid; IBMX, 3-isobutyl-1-methylxanthine; ISR, insulin
secretion rate; KATP channel, ATP-sensitive K channel; OCR, oxygen consumption rate; SERCA, sarcoendoplasmic reticulum calcium ATPase; TCA,
trichloroacetic acid.
DOI: 10.2337/db06-0400
2006 by the American Diabetes Association.
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.

DIABETES, VOL. 55, DECEMBER 2006

coupling factors ATP and ADP in regulating calcium influx

by closing ATP-sensitive K channels (KATP channels). It
has been proposed that calcium may increase ATP production by enhancing the metabolic generation of NADH
mediated by calciums activation of certain key dehydrogenases that regulate the rate of the trichloroacetic acid
(TCA) cycle (513). Or, alternatively, calcium may mediate
an increase in ATP usage associated with stimulation of
ion pumping, biosynthesis, and/or the movement and
exocytosis of insulin granules (14 18). A number of characteristics required for the operation of calcium-induced
stimulation of the generation of NADH have been elegantly
demonstrated; studies have shown that glucose can elevate mitochondrial calcium (8,19), activate mitochondrial
dehydrogenases (6,7), and increase NAD(P)H (20). However, in the context of total cellular ATP turnover during
the second phase of insulin secretion, it is difficult to
assess the quantitative significance of these affects. To
accomplish this, oxygen consumption rate (OCR), a reflection of the rate of electron transport, is an optimal
parameter. Although an initial study found no effect of
calcium on OCR (21), in recent studies using a novel
oxygen microsensor implanted into single islets, blocking
calcium influx with nifedipine clearly increased oxygen
tension in the extracellular space within the islets stimulated with 10 mmol/l glucose (22), reflecting a lowered
OCR. Moreover, oscillations in oxygen tension were reciprocally correlated with glucose-stimulated cytosolic calcium (23). Finally, similar to brain (24) and muscle (25),
depolarization of the membrane by potassium decreased
oxygen tension (26). Thus, it appears to be firmly established that calcium influx has an effect on oxygen consumption; however, questions remain regarding the
mechanism mediating the interaction and regarding how
much of the ATP generated in response to glucose is
utilized by the calcium-dependent process(es). In general,
very little is known regarding the processes that govern
ATP usage in -cells (27), and the contributions of protein
synthesis, ion transport, and exocytosis of insulin granules
to total glucose-stimulated ATP production have not been
quantified. Due to the dual control of OCR by both energy
supply and demand, it is important to appreciate that OCR
measurements alone cannot distinguish the mechanism
that is driving any changes in OCR.
Electron transport integrates the driving forces of energy supply (as reflected by reducing equivalents mainly in
the form of NADH) and energy demand (primarily mediated by the strong potency of ADP to stimulate electron
transport [28]) and, with oxidative phosphorylation, generates the appropriate amount of ATP to maintain the
3509

CALCIUM AND ATP USAGE

FIG. 1. Schematic depicting the hypothesized mechanisms mediating

the effect of calcium on ATP turnover. OCR is under dual control both
by processes that increase the generation of NADH and by processes
that utilize ATP. Whether calcium stimulates ATP turnover was tested
by measuring OCR, and the contribution of the two driving forces
mediating any change in OCR was distinguished by assessing cytochrome c reduction, a measure of NADH generation.

phosphorylation potential (ATP/ADP/Pi) at a set-point

(14,29,30). Either a measure of mitochondrial NADH or the
phosphorylation potential must be concomitantly measured (31,32). Since the thermodynamically relevant free
ADP is difficult to measure, we have assessed the reductive state of cytochrome c (14), a component of the
electron transport chain that is in thermodynamic equilibrium with mitochondrial NADH (33,34), as a measure of
calciums effect on NADH generation.
The conceptual framework of the investigation is depicted in Fig. 1, and we sought to: 1) quantify the contribution of calcium and protein synthesis to glucosestimulated OCR, 2) determine whether the increase in
OCR in response to calcium influx is mediated by an
increase in the generation of NADH or ATP usage, 3) test
whether movement and exocytosis of insulin granules
and/or calcium pumping into the endoplasmic reticulum is
mediating calcium-dependent OCR, and 4) determine the
relation between calcium influx dependent OCR and insulin secretion rate (ISR).
RESEARCH DESIGN AND METHODS
Krebs-Ringer bicarbonate solution was used for the perifusion analysis and
contained 2.6 mmol/l CaCl2/2H2O, 1.2 mmol/l MgSO4/7H2O, 1.2 mmol/l
KH2PO4, 4.9 mmol/l KCl, 98.5 mmol/l NaCl, and 25.9 mmol/l NaHCO3 (all from
Sigma-Aldrich, St. Louis, MO) supplemented with 20 mmol/l HEPES/NaHEPES (Roche, Indianapolis, IN) and 0.1% BSA (Serological, Norcross, GA).
Antimycin A, potassium cyanide, nimodipine, diazoxide, clonidine, 3-isobutyl1-methylxanthine (IBMX), A23187, cyclopiazonic acid (CPA), emetine, and
glybenclamide were purchased from Sigma-Aldrich.
Rat islet isolation and culture. Rat islets were harvested from male BB rats
(250 g, bred in the laboratory of Dr. ke Lernmark, University of Washington, Seattle, WA) (35) and anesthetized by intraperitoneal injection of sodium
pentobarbital (35 mg/230 g body wt). All procedures were approved by the
University of Washington Internal Animal Care and Use Committee. Islets
were prepared and purified as described (14,36). Islets were cultured for 18 h
at 37C before the experiments in RPMI-1640 containing 11.1 mmol/l glucose
supplemented with 10% heat-inactivated fetal bovine serum (Gibco, Grand
Island, NY).
3510

Perifusion system. A perifusion system was used that allows for simultaneous measurement of OCR and cytochrome c reduction and collection of
outflow fractions for subsequent measurement of ISR. Handpicked rat islets
(n 750) were loaded into the chamber with Cytodex beads (Amersham
Biosciences, Piscataway, NJ) and sandwiched between two layers of Cytopore
beads (Amersham Biosciences), as previously described in detail (14,37,38).
Flow rate was set to 80 l/min for all perifusion experiments; chamber
volume was 400 l. Data were corrected for delays in the flow system by
referencing the insulin measurements to glucose measured in the outflow
fractions, and the OCR and cytochrome c reduction data were referenced to
the response to antimycin A, which was given at the end of each flow culture
experiment, as previously described (14,29,30). However, data were not
corrected for dispersion of the signals, and, consequently, the temporal
resolution of the kinetic responses is limited to 4 5 min.
Measurement of OCR. OCR was calculated as the flow rate times the
difference between inflow and outflow oxygen tension, which was measured
by detecting the phosphorescence lifetime of an oxygen-sensitive dye that was
painted on the inside of the perifusion chamber (38). Unlike the previous
report (38), phosphorescent lifetimes were monitored using either a PMOD
5000 Frequency Domain Phosphorometer (Oxygen Enterprises, Philadelphia,
PA) or an MFPF-100 multifrequency phase fluorometer lifetime measurement
system made by TauTheta Instruments (Boulder, CO) where the end of the
excitation light guide (one-eighth inch; Edmund Industrial Optics, Barrington,
NJ, or a 2-mm fiberoptic patch cord; TauTheta part no. SFO-026) illuminated
by a 405-nm light emitting diode was just touching the outside of the glass
opposite where the dye was painted and the detecting light guide situated at
a 90 angle. For the Oxygen Enterprise System, a 48-kHz, 16-bit Sigma-Delta
digitizer was used to average the phosphorescence signal (700 nm) over 10 ms
per scan. Ten scans were performed at each measurement point, and the
signal was averaged over a 100-ms interval; lifetime was calculated as
described (39) for each interval, with measurements being repeated every
30 s. The MFPF100 measured the luminescent lifetime of phosphorescence
using an avalanche photodiode where the light emitting diode was sinusoidally modulated at 5 kHz and phase shift measured over an interval of 0.5 s
and repeated every 30 s.
Measurement of cytochrome c reduction. Absorption due to cytochrome c
was measured by light transmission at 550 nm through the bed of islets/
Cytodex beads, as described previously (14).
Imaging and quantification of cytosolic calcium. Cytosolic calcium was
measured as reflected by fluorescence detection as described by Hille and
colleagues (40), except that fura-2AM was used instead of indo-1. Briefly, islets
were loaded with dye by incubating them in 4 mol/l fura-2AM (Invitrogen,
Carlsbad, CA) for 20 min at 37C. Subsequently, the islets were pipetted into
a temperature-controlled perifusion dish (Bioptechs, Butler, PA) next to a
semicircle-shaped section of 650 micron-OD glass capillary tubing that maintained the position of the islets. The perifusion dish was mounted onto the
stage of a Nikon Eclipse TE-2000-U inverted microscope. Fluorescent emission was detected at 510 nm by a Pixelfly camera (PCO Imaging, Kelheim,
Germany) during alternating excitation at either 340 or 380 nm.
Protein synthesis. The amount of L-[35S]-methionine incorporated into the
TCA-precipitable fraction was determined as follows. Islets were preincubated for 60 min in Krebs-Ringer bicarbonate solution containing 20 mmol/l
glucose and subsequently for 60 min in the presence of [35S]-methionine (4.1
Ci), either with or without 5 mol/l nimodipine or in the presence of varying
amounts of emetine. Thereafter, 300 l of 25% TCA was added. Samples were
vortexed for 10 s and then centrifuged for 3 min at 3,270g. The supernatant
was removed, 400 l of 5% TCA was added, and the samples were again
vortexed and centrifuged. After aspirating the supernatant, 50 l of 0.5% SDS
was added and samples placed on a rocker at 4C until the pellet dissolved
(2 h). The amount of cell-associated radioactivity was determined as
previously described (41).
Insulin measurements. Insulin was measured by enzyme-linked immunosorbent assay per the manufacturers instruction (ALPCO, Windham, NH).
Protocols and data analysis. The basic protocol for the perifusion experiments entailed a 90-min baseline period at 3 mmol/l glucose, followed by
stimulation with 20 mmol/l glucose (45 min) and 45 min in the presence of an
effector of the process of interest. In some experiments, a second agent that
reversed the effect of the inhibitor was added to the perifusate during an
additional 40-min period. To assess the effects of agents, steady-state values of
OCR and cytochrome c reduction were calculated by averaging a portion of
each kinetic curve obtained before and after the agent came in contact with
the islets in the chamber. Typically, this portion was the last 15 min before the
next change in perifusate composition, except where noted. Thus, steadystate averages were calculated at 3 mmol/l glucose from 15 to 0 min, at 20
mmol/l glucose from 30 to 45 min, for agent 1 from 75 to 90 min, and for agent
2 from 115 to 130 min. ISR, which does not actually reach a steady state, was
DIABETES, VOL. 55, DECEMBER 2006

I.R. SWEET AND M. GILBERT

FIG. 2. Control studies: concomitant measurement of OCR, cytochrome c reduction, and ISR in the absence of effectors. Islets were perifused with
buffer containing 3 mmol/l glucose for 90 min. Glucose was increased to 20 mmol/l, and parameters were measured continuously for a time
spanning both experimental protocols used in subsequent experiments (average of six perifusions).
also averaged for each perifusion composition, except time windows of 25
to 0, 20 45, 6590, and 105130 min were used.
The following equations were then used to calculate the steady-state
changes induced by agent 1 relative to glucose-stimulated OCR, cytochrome c
reduction, and ISR.
OCRagent 1 OCR20 mM glc
OCRagent 1
100
100
OCRglc
OCR20 mM glc OCR3 mM glc

(1)

(%Cyt c Red) Cyt c Redagent 1 Cyt c Red20 mM glc

(2)

100

ISRagent 1
ISRagent 1 ISR20 mM glc
100
ISR20 mM glc
ISR20 mM glc

(3)

When a second agent was tested, the steady-state changes were calculated as
follows:
100

OCRagent 2 OCRagent 1
OCRagent 2
100
OCR glc
OSR20 mM glc OCR3 mM glc

(%Cyt c Red) Cyt c Redagent 2 Cyt c Redagent 1

100

ISRagent 2
ISRagent 2ISRagent 1
100
ISR20 mM glc
ISR20 mM glc

(4)
(5)
(6)

For OCR and cytochrome c reduction data, unpaired t tests were used to
compare steady-state changes in response to an agent to control studies at
identical time points where the agents were not present. For the ISR data, due
to the variability in responses that were typically seen, the requirement for
normally distributed data was not met, and an unpaired Mann-Whitney U test
DIABETES, VOL. 55, DECEMBER 2006

was used. Calculations for both tests were carried out using Kaleidagraph
(Synergy Software, Reading, PA).

RESULTS

Control studies and response to glucose. To validate

the protocols that were used, control experiments were
carried out in the absence of added agents. All measured
parameters increased in response to 20 mmol/l glucose
(Fig. 2), and plateaus in OCR and cytochrome c reduction
were reached within 15 min, which remained very stable
for the duration of the protocol. Steady-state OCR and
cytochrome c reduction data averaged from 75 to 90 or 115
to 130 min, which is 2% different from the average of data
obtained between 30 and 45 min (Table 1), demonstrating
the validity of comparing these time periods to evaluate
the effect of agents given during the second or third
periods.
For all measurements made in response to changing
from 3 to 20 mmol/l glucose (n 40), OCR, cytochrome c
reduction, and ISR increased from 0.32 0.02 to 0.64
0.026 nmol/min/100 islets (mean SE), 31.6 1.4 to
66.1 2.8%, and 0.32 0.049 to 2.02 0.16 ng/min/100
islets, respectively.
Effect of calcium influx. Two blockers of calcium uptake were used: nimodipine, a blocker of L-type calcium
3511

CALCIUM AND ATP USAGE

TABLE 1
Contribution of calcium and insulin secretion to glucose-stimulated OCR, cytochrome c reduction, and ISR
Target process
Agent
Calcium influx
Nimodipine (n 7)
Diazoxide (n 9)
Glibenclamide (n 9)*
A23187 (n 5)
Insulin secretion
Clonidine (n 4)
IBMX (n 4)
Calcium uptake by ER
CPA (n 8)
Protein synthesis
Emetine (n 6)
Control (agent 1)
None (n 6)
Control (agent 2)
None (n 6)

OCR
(% change from
Eqs. 1 and 4)

Cytochrome c reduction
(% change from
Eqs. 2 and 5)

ISR
(% change from
Eqs. 3 and 6)

27.2 7.4 (0.007)

28.6 2.1 ( 0.001)
39.9 6.8 ( 0.001)
20.0 4.3 (0.004)

2.6 1.9 (NS)

5.1 1.6 (0.021)
7.8 2.3 (NS)
2.9 4.0 (NS)

85.2 4.7 (0.0002)

82.2 4.6 (0.002)
141.4 28.2 (0.002)
88.1 46.9 (NS)

12.9 3.2 (0.007)

2.7 3.9 (NS)

0.2 2.4 (NS)

4.8 0.6 (0.025)

78.1 7.6 (0.019)

96.0 16.4 (0.019)

2.8 4.6 (NS)

1.4 1.1 (NS)

20.1 4.9 (0.005)

281.2 109.7 (0.003)

1.6 2.6 (NS)

70.2 19.4 (NS)

1.9 3.0

1.7 1.9

28.1 13.3

0.9 3.5

1.5 2.0

2.1 16.9

Data are average SE or average SE (P value). P values were generated either from an unpaired t test (OCR and cytochrome c data) or
Mann-Whitney U test (ISR data), relative to the control experiments. Steady-state calculations were made on data from each experiment (the
composites of which are shown in Figs. 27) using Eqs. 1 6. Glybenclamide was tested following diazoxide (Fig. 3B), and IBMX was tested
following clonidine (Fig. 4). For A23187, since the effect on OCR was transient, OCRagent1 was calculated as the average from 53 to 60 min
(Fig. 5). ER, endoplasmic reticulum.

channels, and diazoxide, which opens KATP channels,

thereby preventing activation of the L-type channels by
glucose. After stimulating with glucose, nimodipine (Fig.
3A) and diazoxide (Fig. 3B) decreased OCR by 27.2 7.4
and 28.6 2.1% of glucose-stimulated OCR, respectively.
Glibenclamide reversed the effect of diazoxide (Fig. 3B). In
the presence of 5 mol/l nimodipine, OCR in response to
a change in glucose was 0.15 0.03 nmol/min/100 islets
(Fig. 4); the ionophore A23187, which acts as a calciumspecific channel, increased cytosolic calcium and stimulated OCR, partially reversing the effect of nimodipine.
Thus, it appears calcium influx is associated with changes
in OCR (Table 1).
Altering calcium influx with either nimodipine or
A23187 did not have an effect on cytochrome c reduction,
suggesting that their effects on OCR were not mediated by
a change in metabolic rate via a change in NADH. Although the effects were small, in eight of nine experiments,
cytochrome c decreased in response to diazoxide (P
0.021). It should be recognized that these measurements
are close to the detection limits of the measurement, and
in light of this, the kinetic profile characterized by a 15-min
delay in response to diazoxide was not considered meaningful (Fig. 3B). The differential effect of nimodipine
versus diazoxide on cytochrome c reduction did not
correspond to a significantly different decrement of OCR
induced by the two agents. The effect of A23187 on ISR
was small, probably reflecting the relatively small increase
in cytosolic calcium elicited by the agent.
Effect of protein synthesis on OCR and cytochrome c
reduction. At a concentration that inhibited protein synthesis by 84 0.4% (n 3), 10 mol/l emetine decreased
glucose-stimulated OCR by 20% but had no effect on cytochrome c reduction (Fig. 5). Since the concentration of
emetine used only produced an 84% inhibition of protein
synthesis, an estimate of the total contribution of protein
synthesis to glucose-stimulated OCR is 20% per 0.84, or 24%.
3512

Effect of calcium influx on protein synthesis. To test

whether calcium influx increased OCR by stimulating
protein synthesis, incorporation of 35S-methionine into
TCA-precipitable protein was measured in the presence
and absence of 5 mol/l nimodipine. Nimodipine actually
increased the rate constant of label incorporation by 69%
(from 4.4 to 7.5 l/min/100 islets), ruling out the possibility
that nimodipine decreased OCR by inhibiting protein
synthesis.
Effect of movement and exocytosis of secretory granules on OCR. Since diazoxide and nimodipine inhibit both
calcium processing and insulin secretion, experiments
were carried out to directly test the contribution of
movement and exocytosis of secretory granules to OCR.
Clonidine (1 mol/l), an 2 adrenergic agonist, inhibited
glucose-stimulated ISR and had a small inhibitory effect on
OCR (Fig. 6 and Table 1). The presence of IBMX subsequently stimulated ISR but did not increase OCR, suggesting that the contribution of the process of insulin secretion
to ATP turnover is small. The small effect of IBMX on
cytochrome c, although statistically significant, was not
considered likely to be meaningful due to the small
number and its lack of effect on OCR. In addition, a similar
lack of effect of 5 mmol/l caffeine on OCR and cytochrome
c reduction was observed (data not shown).
Effect of SERCA on OCR. To test the contribution of
pumping cytosolic calcium into the endoplasmic reticulum, experiments were carried out using an agent that
blocks uptake by SERCA (50 mol/l CPA; Fig. 7). CPA led
to a rapid increase in ISR and cytosolic calcium but
initially led to little change in OCR or cytochrome c
reduction. There was, however, a small delayed decrease
in OCR 15 min after the addition of CPA (OCR
0.010 0.015 nmol/min/100 islets, n 8), which did not
reverse upon removal of the CPA. The protocol shown in
Fig. 7 was carried out but with 3 or 10 mmol/l glucose
substituted for 20 mmol/l, and similar results were obDIABETES, VOL. 55, DECEMBER 2006

I.R. SWEET AND M. GILBERT

FIG. 3. Effect of inhibition of calcium influx on OCR, cytochrome c reduction, and ISR. Forty-five minutes after islets were stimulated with 20
mmol/l glucose, calcium influx was inhibited either by direct inhibition of L-type calcium channels (5 mol/l nimodipine; A) or by opening the KATP
channels (50 mol/l diazoxide; B) (average of seven [A] or nine [B] perifusions). In the latter case, glybenclamide (1 mol/l) was subsequently
added to the inflow buffer.

DIABETES, VOL. 55, DECEMBER 2006

3513

CALCIUM AND ATP USAGE

FIG. 4. Effect of increasing influx of calcium via an ionophore, in the presence of an L-type calcium channel blocker. Islets were perifused in the
presence of 3 mmol/l glucose and 5 mol/l nimodipine for 90 min; at t 0, glucose concentration was raised to 20 mmol/l for 45 min; subsequently,
10 mol/l A23198 was added. OCR (top), ISR (middle), and cytochrome c reduction (data not shown) were measured concomitantly (average of
five perifusions). Bottom: Detection of fluorescence from a calcium-sensitive dye, using a similar temporal protocol except that each time period
lasted 10 min as indicated (average of responses by three islets).

tained: at all glucose concentrations, ISR increased in

response to CPA (ISR3mmol/l 0.66 0.10 ng/min/100
islets, n 4; ISR10mmol/l 4.3 1.3 ng/min/100 islets,
n 4; ISR10mmol/l 4.0 0.48 ng/min/100 islets, n 8),
but no effect on OCR during the first 15 min was observed.
OCR was subsequently inhibited irreversibly (OCR3mmol/l
0.018 0.007 nmol/min/100 islets, n 4; OCR10mmol/l
0.051 0.017 nmol/min/100 islets, n 4). Since CPAs
effect on calcium was reversible and occurred within 15
3514

min, it seems that its delayed effect on OCR was not

associated with changes in ATP usage due to calcium
uptake into the endoplasmic reticulum. Consistent with
this interpretation, 5 mol/l thapsigargin increased ISR but
had no effect on OCR for the entire 45 min at either 10 or
20 mmol/l glucose (data not shown).
Influence of glucose on calcium influx dependent
OCR and ISR. The protocol used in Fig. 3A was repeated
at different levels of glucose (3, 8, 12, and 20 mmol/l).
DIABETES, VOL. 55, DECEMBER 2006

I.R. SWEET AND M. GILBERT

FIG. 5. Inhibition of protein synthesis by emetine decreased glucose-stimulated OCR. Forty-five minutes after islets were stimulated with 20
mmol/l glucose, protein synthesis was inhibited with 10 mol/l emetine (average of six perifusions).

Steady-state values were calculated for each time period

as described in RESEARCH DESIGN AND METHODS, as the changes
in OCR in response to glucose and nimodipine (the denominator [OCRglc] and numerator [OCRnim] in Eq. 1)
and the change in ISR in response to glucose (ISRglc
ISR20mmol/l glc ISR3mmol/l glc). The values were normalized by dividing by the value of the parameter obtained at
20 mmol/l glucose. The glucose dose response of
OCRnim, but not glucose-stimulated OCRglc, correlated
well with ISRglc (Fig. 8). In other experiments (data not
shown), neither ISR nor OCR increased when glucose was
raised from 20 to 30 mmol/l.
DISCUSSION

Although it is not an ATP-dependent process, it has been

observed, using an elegant oxygen sensor implanted into
single islets, that calcium influx is associated with changes
in respiration in islets (22,23,26). These findings are similar
to those found in brain (24) and muscle (25), and it
appears that there is a calcium-activated process that
increases energy turnover, which is ubiquitous in electriDIABETES, VOL. 55, DECEMBER 2006

cally excitable cells. We have used a perifusion system

that simultaneously assesses OCR, cytochrome c reduction, and ISR in isolated rat islets (14,38) to further
investigate the properties and implications of this process.
Dual control of OCR by NADH generation and energy
state: effect of calcium influx. OCR and ATP production
are under dual control in many cell types (29,42,43),
including the pancreatic -cell (14). An increase in substrate availability can thermodynamically drive electron
transport and ATP production by mass action due to
increases in NADH/NAD and the reductive state of the
cytochromes in the electron transport chain. Conversely,
an increase in the rate of processes that utilize ATP will
lead to a decrease in ATP/ADP/Pi and, subsequently, an
increase in electron transport and ATP production to
replenish the energy state (29). The concomitant measurement of cytochrome c reduction (a measure normally in
equilibrium with mitochondrial NADH/NAD levels [33,34])
allows for the distinction between substrate-driven and
energy demand driven OCR (14), and we used this approach to determine the extent that calcium affects each
3515

CALCIUM AND ATP USAGE

FIG. 6. Effect of an 2-adrenergic agonist and a phosphodiesterase inhibitor on OCR, cytochrome c reduction, and ISR. Forty-five minutes after
stimulating islets with glucose, ISR was inhibited by 1 mol/l clonidine and subsequently stimulated with 0.25 mmol/l IBMX (average of four
perifusions).

regulatory arm under steady-state conditions. Glucose

induced the expected increase in cytochrome c reduction,
and antimycin A caused a decrease, but neither agent that
directly effected calcium influx (nimodipine and the calcium ionophore A23187) or protein synthesis rate had an
effect on cytochrome c reduction. This suggests that an
increase in NADH generation is not the dominant driving
force that is mediating calcium-dependent steady-state
OCR in intact islets, but, rather, calciums effect on OCR is
mediated by a change in ATP usage.
In contrast to nimodipine, the decrement in steady-state
OCR induced by diazoxide was paralleled by a small but
statistically significant decrease in cytochrome c reduction; both effects of which were reversed by glibenclamide.
Presumably, the difference in the responses to nimodipine
and diazoxide is related to the latters effect on membrane
potential through the action of the KATP channel. However,
since both agents elicited similar changes in OCR, the
magnitude of the change in cytochrome c induced by
diazoxide (only 5%) apparently did not translate to significant changes in OCR.
The scenario that the calcium effects on OCR are
mediated by ATP usage is consistent with the conclusions
3516

of Henquin and colleagues (18) who found that both

nimodipine and diazoxide increased the total islet content
of ATP/ADP. If calcium increased ATP production by
stimulating the TCA cycle, then blocking calcium influx
would be expected to lower ATP/ADP. It has recently been
reported (44) that oscillations of NADP(H) in mouse
pancreatic islets required calcium and could be blocked by
diazoxide and nifedipine. However, the steady-state values
did not change much, which is in agreement with our
measurements of cytochrome c reduction. Note that the
measurements made by our perifusion system have a
resolution of 4 5 min and cannot resolve oscillations in
cytochrome c reduction or OCR with periods less than
this, and, in addition, the amplitude of the changes in
NADP(H) were 10% of the glucose response, a change
that would be undetectable by our method for measuring
of cytochrome c reduction. While it has been found that
NAD(P)H levels in single -cells decreased for the duration of a 1- to 2-min study in the presence of EGTA (20)
and it was concluded that calcium plays a role in increasing the generation of NADH, it appears that the changes
we observed in OCR that are sustained 45 min after a
DIABETES, VOL. 55, DECEMBER 2006

I.R. SWEET AND M. GILBERT

FIG. 7. Effect of inhibition of SERCA. Using the flow culture system, islets were perifused in the presence of 3 mmol/l glucose for 90 min; at t
0, glucose was raised to 20 mmol/l for 45 min; subsequently, 50 mol/l CPA was added. OCR (top), ISR (middle), and cytochrome c reduction (data
not shown) were measured concomitantly (average of four perifusions). Bottom: Detection of fluorescence from a calcium-sensitive dye, using
a similar temporal protocol except that each time period lasted 20 min (average of responses by three islets).

change in calcium influx were not influenced by this

mechanism.
Energy demands of insulin secretion are small. Having
concluded that calcium influx can induce changes in OCR
by altering rates of ATP usage, we set out to identify the
process that is mediating the increased utilization and
began by manipulating known affects of calcium. There
was a small effect of clonidine on OCR; however, there
was no effect of IBMX in the face of dramatic changes in
insulin secretion. The inhibition of OCR by clonidine is
difficult to interpret but may be due to changes in memDIABETES, VOL. 55, DECEMBER 2006

brane potential and calcium influx. However, the lack of

effect of IBMX on OCR clearly demonstrates that the
energy costs associated with exocytosis of insulin granules are small. Thus, the increase in ATP production in
response to glucose is not simply related to the energy
needed for secretion, as has been suggested (8,45).
Role of SERCA in calcium-dependent OCR. A plausible
candidate for a process that is mediating ATP usage in
response to calcium influx is the SERCA. However, CPA, a
blocker of SERCA, while dramatically increasing cytosolic
calcium and insulin secretion, had little effect on OCR.
3517

CALCIUM AND ATP USAGE

FIG. 8. Glucose dependency of OCR, ISR, and calcium influx dependent OCR. Using the protocol in Fig. 3A, steady-state values of glucosestimulated OCR (OCRglc) and (ISRglc) and the decrement in OCR in
response to blocking calcium influx by nimodipine (OCRnim) were
calculated (as described in RESULTS) from perifusions done at 3 (n 5),
8 (n 5), 12 (n 7), or 20 mmol/l (n 8) glucose. Data were
normalized by dividing by the value of the parameter obtained at 20
mmol/l glucose.

This result has two important implications. First, is that

SERCA makes only a small contribution to glucose-stimulated ATP usage at steady state. We did not anticipate this
result but, in hindsight, seems very plausible in the scenario where in the initial phase of glucose stimulation of
calcium influx, calcium pumping into the endoplasmic
reticulum is significant but, over time, the endoplasmic
reticulum fills up and SERCA is no longer active. Second,
and more far-reaching, it clearly demonstrates the lack of
relation between bulk cytosolic calcium and OCR and,
consistent with the lack of effect of calcium on cytochrome c reduction, argues against a significant role for
calcium stimulation of the TCA cycle in dictating mean
OCR under steady-state conditions. This indicates that the
calcium influx dependent process involves local stimulation, perhaps in microdomains in the vicinity of the
membrane, as has been suggested (46). We have no
explanation for the delayed irreversible effect of CPA on
OCR, but since no such effect was observed in response to
thapsigargin, it is presumably not related to blocking
SERCA.
Estimation of ATP usage from OCR data. If, as in the
-cell, little lactate is produced (47), glucose-stimulated
ATP production will be mostly generated by oxidative
phosphorylation. Therefore, ATP production will be proportional to OCR (less the affect of mitochondrial uncoupling). Although, following glucose stimulation, there is a
transient period when ATP production outpaces ATP
usage, a steady state is soon reached where the two rates
are equal. Under these conditions, the contribution of
processes to ATP usage can be estimated by inhibiting that
process and measuring the decrease in OCR (16,17). Thus,
the calcium-mediated process and protein synthesis uses
at least 30 and 24% of the total ATP generated in response
to glucose, respectively. The actual contribution to ATP
usage would be higher if there is significant mitochondrial
uncoupling or if inhibiting ATP utilization by one process
3518

lead to an increase in ATP usage by another process (17).

Since nimodipine increased protein synthesis, the latter
may be occurring and 30% represents an underestimation.
The mechanism of calcium influx dependent ATP
usage. Having ruled out exocytosis of insulin granules,
other plasma membraneassociated candidates mediating
calciums effects on OCR could involve ion pumping
(including plasma membrane calcium ATPase and Na-Ca
exchanger in conjunction with the Na-K ATPase). However,
in experiments not presented, ouabain, a blocker of Na-K
ATPase, had no effect on OCR, lessening the likelihood that
the Na-Ca exchanger is mediating significant changes in ATP
usage. Another possible mechanism mediating calcium-dependent OCR is an effect of calcium on biosynthesis, most
importantly protein synthesis. However, we found that nimodipine did not inhibit protein synthesis, which is consistent
with studies where it has been observed that calcium has
little effect on insulin biosynthesis rates (48). An intriguing
possibility that phospholipase C activation could be involved
will be considered in future studies, based on data showing
that blocking calcium influx also blocks phospholipase C
activation by glucose (49).
Implications for glucose-stimulated ISR. How calcium
induces movement and exocytosis of insulin granules is
not understood and is critical to the understanding of the
control of insulin secretion. Studies relating the ratio of
ATP/ADP to insulin secretion at the full range of glucose
concentrations indicate that the largest increases in insulin secretion occur at glucose concentrations 10 mmol/l,
where ATP/ADP is only increasing marginally (50). This is
consistent with our dose responses of OCR and ISR (Fig.
8) and could imply that while ATP/ADP is the sole relevant
regulator of membrane potential via KATP channels, an
additional process contributes to the amplifying pathway
of insulin secretion (1), whose relative contribution to
total insulin release increases as a function of glucose
concentration with a right-shifted glucose dose response.
In light of the fact that the actual process of exocytosis of
insulin granules uses only small amounts of ATP and the
lack of effect of blocking SERCA on OCR, our data
demonstrating that ATP usage for the calcium influx
dependent process is greater than that for protein synthesis, and the remarkable correlation between its glucose
dependency and that of ISR (Fig. 8) support the concept
that the process contributing to the amplifying pathway is
intimately linked to calcium influx. Since the presence of
increased glucose is essential for the actions of secretogogs, such as GLP-1, arginine, and acetylcholine, it seems
that insulin secretion does not occur unless the calcium
influx dependent process has been activated. It has been
known since 1966 that the presence of calcium is essential
for the secretion of insulin to occur (51). We speculate that
identification of the process that is utilizing ATP in response to calcium influx will answer the question of how
this occurs.
ACKNOWLEDGMENTS

This study was funded by grants from the National Institutes of Health (DK17047, DK063986, and P30 DK17047S1).
Special thanks to Drs. Duk-Su Koh and Bertil Hille for
help on the calcium studies. Rats used in this study were
kindly provided by Dr. ke Lernmark.
DIABETES, VOL. 55, DECEMBER 2006

I.R. SWEET AND M. GILBERT

REFERENCES
1. Henquin JC, Ravier MA, Nenquin M, Jonas JC, Gilon P: Hierarchy of the
beta-cell signals controlling insulin secretion. Eur J Clin Invest 33:742
750, 2003
2. Prentki M, Matschinsky FM: Ca2, cAMP, and phospholipid-derived
messengers in coupling mechanisms of insulin secretion. Physiol Rev
67:11851248, 1987
3. Mears D: Regulation of insulin secretion in islets of Langerhans by
Ca(2)channels. J Membr Biol 200:57 66, 2004
4. Satin LS: Localized calcium influx in pancreatic beta-cells: its significance
for Ca2-dependent insulin secretion from the islets of Langerhans.
Endocrine 13:251262, 2000
5. Hansford RG: Relation between mitochondrial calcium transport and
control of energy metabolism. Rev Physiol Biochem Pharmacol 102:172,
1985
6. McCormack JG, Longo EA, Corkey BE: Glucose-induced activation of
pyruvate dehydrogenase in isolated rat pancreatic islets. Biochem J
267:527530, 1990
7. Rutter GA, Pralong WF, Wollheim CB: Regulation of mitochondrial glycerol-phosphate dehydrogenase by Ca2 within electropermeabilized insulin-secreting cells (INS-1). Biochim Biophys Acta 1175:107113, 1992
8. Rutter GA, Theler JM, Murgia M, Wollheim CB, Pozzan T, Rizzuto R:
Stimulated Ca2 influx raises mitochondrial free Ca2 to supramicromolar levels in a pancreatic beta-cell line: possible role in glucose and
agonist-induced insulin secretion. J Biol Chem 268:2238522390, 1993
9. Kennedy ED, Wollheim CB: Role of mitochondrial calcium in metabolismsecretion coupling in nutrient-stimulated insulin release. Diabetes Metab
24:1524, 1998
10. Ainscow EK, Rutter GA: Mitochondrial priming modifies Ca2 oscillations
and insulin secretion in pancreatic islets. Biochem J 353:175180, 2001
11. Kennedy HJ, Pouli AE, Ainscow EK, Jouaville LS, Rizzuto R, Rutter GA:
Glucose generates sub-plasma membrane ATP microdomains in single
islet beta-cells: potential role for strategically located mitochondria. J Biol
Chem 274:1328113291, 1999
12. MacDonald MJ, Brown LJ: Calcium activation of mitochondrial glycerol
phosphate dehydrogenase restudied. Arch Biochem Biophys 326:7984, 1996
13. Gilon P, Henquin JC: Influence of membrane potential changes on cytoplasmic Ca2 concentration in an electrically excitable cell, the insulinsecreting pancreatic B-cell. J Biol Chem 267:2071320720, 1992
14. Sweet IR, Cook DL, DeJulio E, Wallen AR, Khalil G, Callis J, Reems J:
Regulation of ATP/ADP in pancreatic islets. Diabetes 53:401 409, 2004
15. Porterfield DM, Corkey RF, Sanger RH, Tornheim K, Smith PJ, Corkey BE:
Oxygen consumption oscillates in single clonal pancreatic -cells (HIT).
Diabetes 49:15111516, 2000
16. Buttgereit F, Brand MD: A hierarchy of ATP-consuming processes in
mammalian cells. Biochem J 312:163167, 1995
17. Wieser W, Krumschnabel G: Hierarchies of ATP-consuming processes:
direct compared with indirect measurements, and comparative aspects.
Biochem J 355:389 395, 2001
18. Detimary P, Gilon P, Henquin JC: Interplay between cytoplasmic Ca2 and
the ATP/ADP ratio: a feedback control mechanism in mouse pancreatic
islets. Biochem J 333:269 274, 1998
19. Kennedy ED, Rizzuto R, Theler JM, Pralong WF, Bastianutto C, Pozzan T,
Wollheim CB: Glucose-stimulated insulin secretion correlates with
changes in mitochondrial and cytosolic Ca2 in aequorin-expressing INS-1
cells. J Clin Invest 98:2524 2538, 1996
20. Pralong WF, Spat A, Wollheim CB: Dynamic pacing of cell metabolism by
intracellular Ca2 transients. J Biol Chem 269:27310 27314, 1994
21. Hutton JC, Malaisse WJ: Dynamics of O2 consumption in rat pancreatic
islets. Diabetologia 18:395 405, 1980
22. Jung SK, Aspinwall CA, Kennedy RT: Detection of multiple patterns of
oscillatory oxygen consumption in single mouse islets of Langerhans.
Biochem Biophys Res Commun 259:331335, 1999
23. Jung SK, Kauri LM, Qian WJ, Kennedy RT: Correlated oscillations in
glucose consumption, oxygen consumption, and intracellular free Ca(2)
in single islets of Langerhans. J Biol Chem 275:6642 6650, 2000
24. Erecinska M, Nelson D, Chance B: Depolarization-induced changes in
cellular energy production. Proc Natl Acad Sci U S A 88:7600 7604, 1991
25. Barnes WS: Effects of Ca(2)-channel drugs on K()-induced respiration
in skeletal muscles. Med Sci Sports Exerc 25:473 478, 1993
26. Kennedy RT, Kauri LM, Dahlgren GM, Jung SK: Metabolic oscillations in
-cells. Diabetes 51 (Suppl. 1):S152S161, 2002
27. Erecinska M, Bryla J, Michalik M, Meglasson MD, Nelson D: Energy
metabolism in islets of Langerhans. Biochim Biophys Acta 1101:273295, 1992
28. Chance B: Electron transfer: pathways, mechanisms, and controls. Annu
Rev Biochem 46:967980, 1977
DIABETES, VOL. 55, DECEMBER 2006

29. Wilson DF, Owen CS, Holian A: Control of mitochondrial respiration: a

quantitative evaluation of the roles of cytochrome c and oxygen. Arch
Biochem Biophys 182:749 762, 1977
30. Deeney JT, Prentki M, Corkey BE: Metabolic control of beta-cell function.
Semin Cell Dev Biol 11:267275, 2000
31. Panten U, Christians J, von Kriegstein E, Poser W, Hasselblatt A: Effect of
carbohydrates upon fluorescence of reduced pyridine nucleotides from
perifused isolated pancreatic islets. Diabetologia 9:477 482, 1973
32. Panten U, Zunkler BJ, Scheit S, Kirchhoff K, Lenzen S: Regulation of energy
metabolism in pancreatic islets by glucose and tolbutamide. Diabetologia
29:648 654, 1986
33. Erecinska M, Veech RL, Wilson DF: Thermodynamic relationships between the oxidation-reduction reactions and the ATP synthesis in suspensions of isolated pigeon heart mitochondria. Arch Biochem Biophys
160:412 421, 1974
34. Wilson DF, Stubbs M, Veech RL, Erecinska M, Krebs HA: Equilibrium
relations between the oxidation-reduction reactions and the adenosine
triphosphate synthesis in suspensions of isolated liver cells. Biochem J
140:57 64, 1974
35. MacMurray AJ, Moralejo DH, Kwitek AE, Rutledge EA, Van Yserloo B,
Gohlke P, Speros SJ, Snyder B, Schaefer J, Bieg S, Jiang J, Ettinger RA,
Fuller J, Daniels TL, Pettersson A, Orlebeke K, Birren B, Jacob HJ, Lander
ES, Lernmark A: Lymphopenia in the BB rat model of type 1 diabetes is due
to a mutation in a novel immune-associated nucleotide (Ian)-related gene.
Genome Res 12:1029 1039, 2002
36. Matsumoto S, Shibata S, Kirchhof N: Immediate reversal of diabetes in
primates following intraportal transplantation of porcine islets purified on
a new histidine-lactoioniate-iodixanol gradient (Abstract). Transplantation 67:S220, 1999
37. Sweet IR, Cook DL, Wiseman RW, Greenbaum CJ, Lernmark A, Matsumoto
S, Teague JC, Krohn KA: Dynamic perifusion to maintain and assess
isolated pancreatic islets. Diabetes Technol Ther 4:6776, 2002
38. Sweet IR, Khalil G, Wallen AR, Steedman M, Schenkman KA, Reems JA,
Kahn SE, Callis JB: Continuous measurement of oxygen consumption by
pancreatic islets. Diabetes Technol Ther 4:661 672, 2002
39. Vinogradov SA, Fernandez-Seara MA, Dugan BW, Wilson DF: Frequency
domain instrument for measuring phosphorescence lifetime distributions
in heterogeneous samples. Rev Sci Instrum 72:3396 3406, 2001
40. Chen L, Koh DS, Hille B: Dynamics of calcium clearance in mouse
pancreatic -cells. Diabetes 52:17231731, 2003
41. Sweet IR, Cook DL, Lernmark A, Greenbaum CJ, Wallen AR, Marcum ES,
Stekhova SA, Krohn KA: Systematic screening of potential beta-cell
imaging agents. Biochem Biophys Res Commun 314:976 983, 2004
42. Wilson DF, Erecinska M, Drown C, Silver IA: Effect of oxygen tension on
cellular energetics. Am J Physiol 233:C135C140, 1977
43. Wilson DF, Rumsey WL: Factors modulating the oxygen dependence of
mitochondrial oxidative phosphorylation. Adv Exp Med Biol 222:121131,
1988
44. Luciani DS, Misler S, Polonsky KS: Ca2 controls slow NAD(P)H oscillations in glucose-stimulated mouse pancreatic islets. J Physiol 572:379 392,
2006
45. Jouaville LS, Pinton P, Bastianutto C, Rutter GA, Rizzuto R: Regulation of
mitochondrial ATP synthesis by calcium: evidence for a long-term metabolic priming. Proc Natl Acad Sci U S A 96:1380713812, 1999
46. Barg S, Ma X, Eliasson L, Galvanovskis J, Gopel SO, Obermuller S, Platzer
J, Renstrom E, Trus M, Atlas D, Striessnig J, Rorsman P: Fast exocytosis
with few Ca(2) channels in insulin-secreting mouse pancreatic B cells.
Biophys J 81:3308 3323, 2001
47. Sekine N, Cirulli V, Regazzi R, Brown LJ, Gine E, Tamarit-Rodriguez J,
Girotti M, Marie S, MacDonald MJ, Wollheim CB, et al.: Low lactate
dehydrogenase and high mitochondrial glycerol phosphate dehydrogenase
in pancreatic beta-cells: potential role in nutrient sensing. J Biol Chem
269:4895 4902, 1994
48. Curry DL, Bennett LL, Grodsky GM: Dynamics of insulin secretion by the
perfused rat pancreas. Endocrinology 83:572584, 1968
49. Yamazaki H, Philbrick W, Zawalich KC, Zawalich WS: Acute and chronic
effects of glucose and carbachol on insulin secretion and phospholipase C
activation: studies with diazoxide and atropine. Am J Physiol Endocrinol
Metab 290:E26 E33, 2006
50. Detimary P, Jonas JC, Henquin JC: Possible links between glucose-induced
changes in the energy state of pancreatic B cells and insulin release:
unmasking by decreasing a stable pool of adenine nucleotides in mouse
islets. J Clin Invest 96:1738 1745, 1995
51. Grodsky GM, Bennett LL: Cation requirements for insulin secretion in the
isolated perfused pancreas. Diabetes 15:910 913, 1966
3519