Am J Physiol Heart Circ Physiol
279: H1849–H1857, 2000.
Mitochondrial NAD(P)H, ADP, oxidative phosphorylation,
and contraction in isolated heart cells
ROY L. WHITE1 AND BEATRICE A. WITTENBERG2
Department of Physiology, Temple University School of Medicine,
Philadelphia, Pennsylvania 19140; and 2Department of Physiology and
Biophysics, Albert Einstein College of Medicine, Bronx, New York 10461
1
Received 5 May 1999; accepted in final form 5 April 2000
restwork transition; steady-state adenosine 5⬘-triphosphate
utilization; substrate supply; phosphocreatine; lactate; 2-deoxy-D-glucose; reduced nicotinamide adenine dinucleotide;
reduced nicotinamide adenine dinucleotide phosphate
DURING INCREASED CARDIAC OUTPUT the contractile apparatus, together with the pumps and exchangers necessary to maintain transmembrane ionic gradients, create an increased demand for ATP. Recent NMR data in
the whole heart have shown that when steady-state
ATP utilization is increased, an increased rate of ATP
synthesis can be maintained without detectable change
in the steady-state levels of high-energy phosphate (7).
To maintain constant mitochondrial NADH (NADHm)
levels during increased work output when the rate of
NADHm oxidation is increased, NADHm must be re-
Address for reprint requests and other correspondence: B. A.
Wittenberg, Dept. of Physiology and Biophysics, Albert Einstein
College of Medicine, 1300 Morris Park Ave., Bronx, NY 10461
(Email: [email protected]).
http://www.ajpheart.org
plenished by increased rates of NADH synthesis.
NADH supply is governed by the rates of the mix of
mitochondrial dehydrogenases drawing on a common
pool of NAD⫹ and the relative concentrations of the
substrates of the mitochondrial dehydrogenases (26).
The beating heart, when perfused with blood containing both glucose and fatty acid, preferentially uses
fatty acids as a substrate (12). The heart becomes more
dependent on glucose oxidation, however, at high work
loads (20).
NADHm is most accurately and reliably measured in
isolated heart cells, where oxygen delivery and substrate delivery can be rigorously controlled and the
optical path length for fluorescence measurements is
relatively constant for any given field of view (24).
NADHm in isolated heart cells is decreased by 30%
during 4-Hz electrical pacing with glucose and glutamine as substrates (24). To test whether the decrease
in NADH is due to a limitation in the rate of supply of
reducing equivalents from exogenous substrates, we
supplemented the extracellular glucose with fatty acids or the ketoacids acetoacetate or ␤-hydroxybutyrate.
We chose glucose plus ␤-hydroxybutyate as our standard reference condition.
We have shown that in the presence of ample oxygen
and physiological substrates, the rate of supply of reducing equivalents may limit the steady-state replenishment of NADHm during rapid pacing; the consequent decrease in NADHm is accompanied by
decreased contractility. NADHm and the contractile
amplitude during pacing may be enhanced by substrate manipulation or by a large increase in free ADP
(ADPfree).
MATERIALS AND METHODS
Preparation of Isolated Heart Cells
Heart cells were prepared from adult male rats (350–450
g) by a modification of the procedure reported previously (25).
The isolation medium, HEPES-buffered minimal essential
medium (MEM), contained (in mM) 117 NaCl, 5.7 KCl, 4.4
NaHCO3, 1.5 NaH2PO4, 1.7 MgCl2, 21.1 HEPES, 10 taurine,
The costs of publication of this article were defrayed in part by the
payment of page charges. The article must therefore be hereby
marked ‘‘advertisement’’ in accordance with 18 U.S.C. Section 1734
solely to indicate this fact.
0363-6135/00 $5.00 Copyright © 2000 the American Physiological Society
H1849
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on June 17, 2017
White, Roy L., and Beatrice A. Wittenberg. Mitochondrial NAD(P)H, ADP, oxidative phosphorylation, and contraction in isolated heart cells. Am J Physiol Heart Circ
Physiol 279: H1849–H1857, 2000.—To examine the relationship between mitochondrial NADH (NADHm) and cardiac
work output, NADHm and the amplitude and frequency of
the contractile response of electrically paced rat heart cells
were measured at 25°C. With 5.4 mM glucose plus 2 mM
␤-hydroxybutyrate, NADHm was reversibly decreased by
23%, and the amplitude of contraction was reversibly decreased by 27% during 4-Hz pacing. With glucose plus 2 mM
pyruvate or with 10 mM 2-deoxy-D-glucose, NADHm was
maintained during rapid pacing, and the contractile amplitude remained high. Phosphocreatine levels decreased with
2-deoxy-D-glucose administration but not with rapid pacing.
Respiration increased to meet the increased ATP demand at
30°C. The data suggest that 1) when NADHm is decreased
during rapid pacing with defined substrates, the amplitude of
contraction is decreased; 2) the amplitude of contraction
during electrical pacing does not change with rate of pacing
when both the ATP and NADHm levels are continuously
replenished; and 3) the replenishment of NADHm during
pacing with physiological substrates may be rate-limited by
substrate supply to mitochondrial dehydrogenases. During
activation of mitochondrial dehydrogenases, or a significant
increase in free ADP induced by 2-deoxy-D-glucose, this rate
limitation is bypassed or overcome.
H1850
NADH, ADP, AND CONTRACTION OF HEART CELLS
5.4 glucose, and 2 L-glutamine and amino acids and vitamins;
pH was adjusted to 7.2, and osmolarity was adjusted to 292
mosM. Retrograde aortic perfusion of the heart with 0.05%
collagenase (Boehringer) was followed by mechanical dissociation of cells from the tissue matrix, followed by separation
of intact cells with Percoll. The final yield from each heart
was 10–18 ⫻ 106 cells, of which 80–90% were rectangular
and functionally intact (sarcomere length, 2.0 ␮m; synchronous contraction with electrical stimulation). Heart cell protein was 4.82 ⫾ 1.60 mg protein/106 cells (n ⫽ 25).
Measurement of NADHm in Cells
Fig. 1. Changes in steady-state mitochondrial NADH and cell length during pacing
with glucose ⫹ ␤-hydroxybutyrate (A), ⫹
pyruvate (B), and ⫹ 2-deoxy-D-glucose (C).
Left, bar graphs showing steady-state NADH
fluorescence in individual cells at defined periods during the pacing protocol (means ⫾ SE
from 6–30 cells). *NADH was significantly
different from initial value during 0.4-Hz pacing. Right, edge detector traces showing amplitude and frequency of contraction during
pacing. A: substrates were 5.4 mM glucose
and 2 mM ␤-hydroxybutyrate. Mitochondrial
NADH was significantly and reversibly lowered during 4-Hz stimulation compared with
0.4-Hz stimulation. The amplitude of contraction was also significantly and reversibly decreased during 4-Hz pacing compared with
0.4-Hz pacing. Cells were able to contract and
relengthen synchronously with either 0.4- or
4-Hz pacing. B: substrates were 5.4 mM glucose and 2 mM pyruvate. Mitochondrial
NADH was not significantly different during
4-Hz stimulation compared with 0.4-Hz stimulation. The amplitude of contraction was not
detectably reduced during 4-Hz pacing. Cells
were able to contract and relengthen synchronously with either 0.4- or 4-Hz pacing. C: ATP
utilization was enhanced and glycolysis was
inhibited by the addition of 10 mM 2-deoxyD-glucose; 5.4 mM glucose and 2 mM ␤-hydroxybutyrate were the substrates. Mitochondrial NADH was not significantly
different during 4-Hz stimulation compared
with 0.4-Hz stimulation. The amplitude of
contraction was not detectably reduced during 4-Hz pacing. Cells were able to contract
and relengthen synchronously with either
0.4- or 4-Hz pacing.
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on June 17, 2017
This method has been described previously (24). The
NADHm level is defined as the ratio of NADH fluorescence
measured under a defined condition divided by the maximal
NADH observed in the presence of 10 ␮M rotenone. A Nikon
inverted microscope equipped with an epifluorescence attachment and dual, matched photomultiplier tubes was used
as a microfluorimeter to record intrinsic fluorescence spectra
from myocytes at two wavelengths (415 and 470 nm). Fluorescence emissions from single myocytes (steady-state measurements averaged over 6–30 cells; see Fig. 1, left) or groups
of 12–15 myocytes (kinetic measurements; see Figs. 2 and 3)
were isolated using a mask with a rectangular opening of
constant size in the light path. NADH fluorescence was
measured near 470 nm. Cell fluorescence not specific to
NADH (background) was measured at 415 nm, a wavelength
isoemissive for oxymyoglobin and deoxymyoglobin. This
serves as a reference to compensate for differences in cell
thickness and light scattering and was subtracted from the
reading at 470 nm. Fluorescence measurements were made
during a 15-ms, 350-nm illumination of cells. Separate measurement of the NADH of round cells gave the fully oxidized
NADH value. The two photomultiplier signals were collected
by a Macintosh computer (Apple Computer, Cupertino, CA)
connected to a MacLab (CB Sciences, Dover, NH) data acquisition system. Steady-state NADH fluorescence was measured in individual cells at defined periods during the pacing
protocol. Cells were superfused with HEPES medium containing 3 mM CaCl2 and equilibrated with 50% O2-45%
N2-5% CO2. The control values (shown in Table 1) were the
calculated average of initial and recovery values. Occasionally (10–20% of experiments) a continual drop in NADHm
was observed among the control, rapid pacing, and recovery.
This was attributed to a decline in the viability of the cell
during the experimental protocol, and these data were rejected. In the presence of either acetate or iodoacetate, repeated experiments showed that NADH during recovery was
NADH, ADP, AND CONTRACTION OF HEART CELLS
H1851
significantly lower than that observed initially, even though
a significant decrease in NADH was observed during rapid
pacing compared with recovery. In these cases the final
recovery value was used as the control.
bipolar stimulation pulses. Stimulus voltage was increased
until cells contracted synchronously (usually ⬍160 V/cm).
Pacing Protocol
The frequency of contraction was controlled by the rate of
stimulation. The extent of the contraction was independent
of delivered stimulus voltages. Resting cell length was monitored, and if the resting length of the cell changed during
recording, the data were discarded. The frequency and amplitude of contraction of individual cardiac myocytes were
recorded using a high-speed video camera and video motion
edge detector (Crescent Electronics, Orem, UT) in the Nikon
microfluorimeter chamber. The video edge detector measured
the time between two defined changes in contrast (the ends of
the cell, i.e., the time recorded is a measure of cell length) in
a scan line of a video image. During pacing, this measurement was recorded continuously and was updated for each
field (240 s⫺1) of video data from the Pulnix camera. Cell
length, when plotted as a function of time, provides a measure of diastolic and systolic length, the amplitude of contraction, and the number of contractions. Cells contracted
strongly (⬃20% of the total length of the cell) when stimulated at 4 Hz in control experiments.
Cells were equilibrated for 20 min with stimulation at 0.4
Hz. After 10 min at 4 Hz (comparable to the rate of a normal
rat heart) and, finally, to demonstrate reversibility, we then
reduced the rate of electrical stimulation to 0.4 Hz for 20 min.
Steady-state measurements of single cells are reported in
Fig. 1 (left). The continuous time course of NADHm fluorescence (shown in Figs. 2 and 3) was obtained by continuous
sampling of NADH fluorescence from a field of 12–15 isolated
heart cells. Conditions were optimized for survival of the
cells, with maximum sustainable contractile amplitude. The
pacing protocol could be repeated three or four times without
loss of cell viability. Cells (2 ⫻ 104 total cells) on the microscope stage were superfused with HEPES-buffered MEM
(see composition in Preparation of Isolated Heart Cells) supplemented with 15 mM NaHCO3, 3 mM CaCl2, and 33 nM
dobutamine at 25°C. ␤-Hydroxybutyrate or other additional
substrates were given at a dosage of 2 mM when present. The
medium was equilibrated with a gas mixture containing 50%
O2-5% CO2 (balance N2) to maintain the pH at 7.2 at a
temperature of 25°C. Control experiments showed no difference in NADH levels measured with 95% O2 compared with
the 50% O2 superfusion. Dobutamine was added to enable
the cells to follow 4-Hz pacing. Calcium levels were maximized to levels that did not cause calcium overload. Platinum-iridium electrodes at the bottom of the dish delivered
Measurement of Heart Cell Contractions
Measurement of Respiration in Cell Suspensions
The method has been described previously (3–5). Steadystate oxygen consumption rates were determined with a
large-diameter, polarographic oxygen electrode (model 2110,
Orbisphere, Geneva, Switzerland) in a thermostated brass
block held at 30°C. Experiments were carried out under
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on June 17, 2017
Fig. 2. A: time course of mitochondrial NADH fluorescence: effects of substrate during 4-Hz pacing of heart
cells. NADH fluorescence from 12–15 isolated heart
cells was sampled continuously for 25 min with different substrate regimens. F, 2 mM ␤-hydroxybutyrate
(␤OH butyrate) ⫹ 5.4 mM glucose; ‚, 10 mM 2-deoxyD-glucose (deoxyglucose) ⫹ 5.4 mM glucose ⫹ 2 mM
␤-hydroxybutyrate. NADH was decreased significantly
and reversibly during 4-Hz pacing only when ␤-hydroxybutyrate ⫹ glucose were the substrates. When
ATP utilization was increased and glycolysis was inhibited with 2-deoxy-D-glucose, mitochondrial NADH did
not decrease during 4-Hz stimulation. B: comparison of
effects of 2-deoxy-D-glucose and iodoacetate on highenergy phosphate levels of paced heart cells. Steadystate high-energy phosphate levels were measured after 10 min of 4-Hz stimulation. Cells superfused with
MEM containing glucose ⫹ ␤-hydroxybutyrate and 3
mM CaCl2 retained high phosphocreatinine (PCr)-toATP levels (PCr/ATP ratio) near resting levels during
4-Hz stimulation. The PCr/ATP ratio was markedly
depleted, however, when 2-deoxy-D-glucose was added
to the medium (left). If glycolysis was stopped by iodoacetate, endogenous glycogen was rapidly depleted, yet
the PCr/ATP ratio was not diminished (right). This
suggests that the observed decrease in PCr, with an
increase in free ADP, is the result of increased ATP
utilization induced by 2-deoxy-D-glucose rather than of
decreased glycolysis.
H1852
NADH, ADP, AND CONTRACTION OF HEART CELLS
cal stimulation in the absence of cells. Stimulus pulse duration was 0.05 ms at 90 V.
Measurements of High-Energy Phosphate Levels
Analytical procedures were used as described previously
(see Ref. 5).
RESULTS
We have examined NADHm, the contractile amplitude
of contraction, high-energy phosphate levels, and the rate
of steady-state respiration in heart cells with different
substrates and different rates of ATP utilization.
Effect of Substrates
Fig. 3. Time course of mitochondrial NADH fluorescence: the effect
of pyruvate addition on NADH recovery after pacing. Substrates
were 5.4 mM glucose and 2 mM ␤-hydroxybutyrate. NADH fluorescence from a field of 12–15 isolated heart cells was sampled continuously for 40 min. The stimulation rate was increased from 0.4 to 4
Hz for 10 min, and mitochondrial NADH decreased. When the
stimulation rate was decreased back to 0.4 Hz, mitochondrial NADH,
in this case, did not fully return to control levels. After the addition
of pyruvate at 19 min, mitochondrial NADH was increased to higher
levels. The addition of 10 ␮M rotenone at 29 min increased mitochondrial NADH to the fully reduced form.
conditions compatible with cell viability. HEPES-buffered
MEM supplemented with 1 mM CaCl2 and 33 nM dobutamine in the chamber was supplemented with other substrates as specified, and the mixture was equilibrated with
20% O2-balance N2. Cells (3–10 ⫻ 105 total cells) in the
respiration chamber were stirred constantly. Sixty-seven
percent of the total cells in the chamber were rectangular at
the end of an experiment. Because only rectangular cells
contract synchronously in response to electrical stimulation,
respiration values were normalized to the number of rectangular cells. This calculation may overestimate the normalized resting respiration, because damaged rounded cells can
still take up oxygen. The respiration chamber was equipped
with Ag-AgCl stimulating electrodes for pacing the cells. No
electrolytic generation of gases was observed during electri-
Table 1. Effects of substrate on contractile amplitude and on NADHm during 4-Hz pacing
NADHm
Cell Treatment
Control
fluorescence
Contraction amplitude
at 4-Hz pacing (compared
with control), %
Contraction Amplitude
at 4-Hz Pacing (compared
with 0.4 Hz), %
Glucose ⫹ ␤-hydroxybutyrate
Glucose ⫹ octanoate
Glucose ⫹ acetoacetate
Glucose ⫹ acetate
Glucose ⫹ lactate
Glucose ⫹ pyruvate
Glucose ⫹ ␤-hydroxybutyrate ⫹ insulin
Glucose ⫹ ␤-hydroxybutyrate ⫹ iodoacetate
Glucose ⫹ ␤-hydroxybutyrate ⫹ 2-deoxy-D-glucose
0.393 ⫾ 0.03
0.219 ⫾ 0.03
0.399 ⫾ 0.04
0.344 ⫾ 0.03
0.413 ⫾ 0.04
0.409 ⫾ 0.079
0.422 ⫾ 0.06
0.303 ⫾ 0.05
0.375 ⫾ 0.04
77 ⫾ 6* (28)
55 ⫾ 6* (18)
51 ⫾ 7* (5)
25 ⫾ 2* (4)
94 ⫾ 8 (27)
104 ⫾ 5 (13)
64 ⫾ 3* (6)
60 ⫾ 6* (15)
94 ⫾ 8 (12)
73 ⫾ 2 (7)
73 ⫾ 1 (3)
70 ⫾ 5 (3)
Not tested
91 ⫾ 3 (4)
96 ⫾ 3 (3)
Not tested
Not tested
86 ⫾ 4 (3)
All values are means ⫾ SE; nos. in parentheses are no. of cells. Glucose concentration was 5.4 mM; all other substrate concentrations were
2 mM. Iodoacetate concentration when present was 100 ␮M, insulin when present was 100 ␮U/ml, and 2-deoxy-D-glucose was 10 mM. Control
values of mitochondrial NADH (NADHm) are the average of NADH before and after stimulation when these values were not significantly
different. With iodoacetate and acetate, NADHm during recovery was always significantly lower than before rapid pacing. Here the final
recovery value was taken as the control. * Significantly different (P ⱕ 0.05) from control.
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on June 17, 2017
Glucose plus ␤-hydroxybutyrate. MITOCHONDRIAL
Steady-state measurements of NADHm in single
isolated heart cells are shown in Fig. 1 (left). With
glucose and ␤-hydroxybutyrate as substrates, NADHm
was significantly and reversibly decreased by ⬃23%
after 10 min of rapid pacing when compared with slow
pacing (P ⱕ 0.05, see Fig. 1A and Table 1). A significant
decrease was also observed with glucose supplemented
with fatty acids or ketoacids (see Table 1). NADH
measurements in the presence of insulin were not
significantly different at rest or during pacing (see
Table 1). In comparison, Fig. 2A shows a continuous
trace of NADHm as a function of time measured in a
field containing 12–16 cells. Figure 2 illustrates that a
distinct overshoot of NADH occurred during recovery.
Such overshoots in response to changes in stimulation
rate have been reported previously (2).
CONTRACTILE RESPONSE. The effects of electrical pacing
on the rate and magnitude of contraction of single cells
are shown in Fig. 1 (right). The cells were able to follow
both the slow and fast stimulus rates under all of the
conditions explored in these experiments. With glucose
plus ␤-hydroxybutyrate, however, the amplitude of
contraction was significantly and reversibly decreased
(28 ⫾ 2%, n ⫽ 7) at the higher rate of stimulation,
whereas diastolic length was maintained.
NADH.
H1853
NADH, ADP, AND CONTRACTION OF HEART CELLS
Table 2. Summary of high-energy phosphate levels, NADHm , and respiratory rate with and
without 2-deoxy-D-glucose, and with and without 4-Hz pacing
PCr/ATP
Cell Treatment
PCr, mM
ATP, mM
0 Hz
Glucose ⫹
␤-hydroxybutyrate
Glucose ⫹
␤-hydroxybutyrate ⫹
2-deoxy-D-glucose
Glucose ⫹ lactate
14.5 ⫾ 1.0†
7.8 ⫾ 0.4†
1.79 ⫾ 0.05† (17)
1.85 ⫾ 0.04‡ (20)
0.72 ⫾ 0.09*† (6)
0.68*‡
1.71 ⫾ 0.08‡
1.84 ⫾ 0.06† (5)
1.77 ⫾ 0.04‡ (5)
1.85 ⫾ 0.08† (6)
1.72 ⫾ 0.06‡
Glucose ⫹ pyruvate
5.5 ⫾ 0.9*†
13.6 ⫾ 1.0†
15.8 ⫾ 1.0†
7.6 ⫾ 0.4†
7.4 ⫾ 0.6†
8.6 ⫾ 0.3†
4 Hz
ADPfree at
0 Hz, ␮M
NADHm
at 4 Hz,
fraction
of control
51 ⫾ 1.4†
216 ⫾ 17*†
0.83*‡
0.77‡
Rate of O2 Uptake,
nmol/min per 106 cells
0 Hz
4 Hz
34 ⫾ 4† (6)
46 ⫾ 4†
48 ⫾ 4*† (3)
61 ⫾ 1*†
33† (1)
45†
33 ⫾ 3† (3)
42 ⫾ 5†
0.94*‡
55 ⫾ 1.8†
47 ⫾ 2.0†
0.94‡
1.04‡
HIGH-ENERGY PHOSPHATE LEVELS. The phosphocreatine
(PCr)-to-ATP (PCr/ATP) ratio measured in resting
cells with glucose plus ␤-hydroxybutyrate in the superfusion chamber at 25°C was 1.85 ⫾ 0.04 and was not
significantly decreased during stimulation at 4 Hz
(1.71 ⫾ 0.08, n ⫽ 20). The PCr/ATP ratio measured in
stirred suspensions of cells at 30°C (1.79 ⫾ 0.05, n ⫽
17) was not different from that measured in superfused
cells at 25°C. At 30°C, PCr was 50.5 ⫾ 3.5 nmol/mg
protein, ATP was 27.4 ⫾ 1.4 nmol/mg protein, and the
calculated ADPfree was 51 ␮M (n ⫽ 17) (see Table 2).
STEADY-STATE RATE OF OXYGEN UPTAKE. The cells maintained a steady rate of oxygen consumption in the
absence of electrical or chemical stimulation. The
steady-state rate of respiration of a stirred suspension
of heart cells at 30°C was reversibly increased from
34 ⫾ 6 nmol to 46 ⫾ 4 nmol O2/min per 106 rectangular
cells (n ⫽ 6) during 4-Hz electrical stimulation (Table
2) (see Fig. 4 for a typical trace). The magnitude of this
response did not differ among all substrate regimens
tested. The steady-state rate of respiration increased to
a value of 118 ⫾ 2 nmol O2/min per 106 cells (n ⫽ 3)
with 3% ␮M carbonyl cyanide m-chlorophenylhydazone (CCCP), an uncoupler of oxidative phosphorylation.
Glucose plus pyruvate or lactate. NADHm AND CONTRACTILE RESPONSE. In contrast to the results obtained
with glucose plus ␤-hydroxybutyrate, neither NADHm
nor the amplitude of contraction decreased significantly during fast pacing with glucose plus pyruvate as
a substrate (Fig. 1B). Similar results were observed
with glucose plus lactate as a substrate (Table 1).
The effect of pyruvate addition is also demonstrated
in the continuous NADHm trace in Fig. 3. NADHm
dropped during rapid pacing but did not return to
initial values when pacing was subsequently slowed to
0.4 Hz (Fig. 3). This phenomenon was observed in
⬃20% of the preparations. NADHm increased to a new
steady state after the addition of 2 mM pyruvate. A
similar effect was observed when lactate was added
(data not shown).
HIGH-ENERGY PHOSPHATE LEVELS AND RESPIRATION. With
pyruvate plus glucose as a substrate, at 30°C, the
PCr/ATP ratio was 1.85 ⫾ 0.08 (n ⫽ 6) and ADPfree was
47 ␮M. With glucose plus lactate, the PCr/ATP ratio
was 1.84 ⫾ 0.06 (n ⫽ 5) (see Table 2). Respiration of
cells in the presence of glucose plus lactate was not
Fig. 4. Effects of 2-deoxy-D-glucose on respiration of paced heart cells.
Respiration of stirred cell suspensions is reported in 2 individual experiments. Respiration was measured in the presence of glucose ⫹ ␤-hydroxybutyrate (E) and when glycolysis was inhibited by 2-deoxy-Dglucose in the presence of the same substrates (●). At rest, steady-state
rate of respiration was considerably higher in the presence of 2-deoxyD-glucose. The upward-pointing arrows at 2 min indicate the start of
4-Hz stimulation. The downward-pointing arrows indicate cessation of
stimulation. Respiration was increased to a new steady state during
4-Hz stimulation. Average data are presented in Table 2. Respiration
increased about 12 nmol O2/min per 106 cells⫺1 in glucose ⫹ ␤-hydroxybutyrate (n ⫽ 6). In the presence of 2-deoxy-D-glucose, in contrast,
respiration was higher at rest and increased about 13 nmol O2/min per
106 cells (n ⫽ 3) during rapid pacing. We conclude that the magnitude
of the respiratory stimulation was essentially the same during 4-Hz
pacing with and without inhibition of glycolysis.
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on June 17, 2017
Values are presented as means ⫾ SE; nos. in parentheses are no. of determinations. * Value is significantly different from that obtained
in the absence of deoxyglucose. † Data were obtained at 30°C. ‡ Data were obtained at 25°C. Statistics for NADHm are presented in Table 1.
ADPfree, free ADP; PCr, phosphocreatine; PCr/ATP, phosphocreatine-to-ATP ratio.
H1854
NADH, ADP, AND CONTRACTION OF HEART CELLS
significantly different from that observed with glucose
plus ␤-hydroxybutyrate under comparable conditions
(Table 2).
Effects of Enhanced ATP Utilization and Pi
Sequestration with 2-Deoxy-D-Glucose
Fig. 5. Schematic representation of
some relevant sarcoplasmic and mitochondrial reactions that contribute to
the mitochondrial NADH pool. Pathways are shown for the metabolism of
glucose, lactate, pyruvate, and ␤-hydroxybutyrate from the sarcoplasm to
the mitochondria. Inhibition of glycolysis by 2-deoxy-D-glucose requires ATP
to form 2-deoxy-D-glucose 6-phosphate
in a reaction with hexokinase, thus
effectively sequestering Pi. This reaction occurs in the sarcoplasm, and the
product competitively blocks breakdown
of the physiological intermediate glucose-6-phosphate. 2-Deoxy-D-glucose-6
phosphate is not further metabolized.
Iodoacetate, in contrast, blocks glycolysis at the triosephosphate dehydrogenase step. During glycolysis, reducing
equivalents formed in the sarcoplasm at
the glyceraldehyde-3-phosphate dehydrogenase step and/or lactate dehydrogenase
(when lactate is present) step enter the
mitochondrial NADH pool via the malateaspartate shuttle. Mitochondrial dehydrogenases of the tricarboxylic cycle, as well
as pyruvate dehydrogenase and ␤-hydroxybutyrate dehydrogenase, also contribute to the mitochondrial NADH pool.
DISCUSSION
Mitochondrial NADH
In the heart, NADH fluorescence arises from the
mitochondrial compartment (23) and reflects the redox
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on June 17, 2017
The addition of 2-deoxy-D-glucose or iodoacetate inhibited the glycolytic pathway (see Fig. 5). In addition,
2-deoxy-D-glucose caused a substantial increase in ATP
utilization for the phosphorylation of a large pool of
2-deoxy-D-glucose.
NADHm and contractile response. After a transient
overshoot, NADHm returned to control values during
rapid pacing when cells were superfused with medium
containing 10 mM 2-deoxy-D-glucose with glucose plus
␤-hydroxybutyrate (Figs. 1C and 2A). In contrast,
NADHm was significantly reduced during rapid pacing
when iodoacetate was used to inhibit glycolysis (Table
2). With 2-deoxy-D-glucose, the cells were able to follow
rapid rates of stimulation, and the amplitude of contraction at 4 Hz was little reduced compared with that
at 0.4 Hz (Fig. 1C and Table 1).
High-energy phosphate levels. A large significant decrease in the PCr/ATP ratio was observed in the presence
of 2-deoxy-D-glucose (see Fig. 2B and Table 2). The PCr/
ATP ratio in resting cells was decreased to 0.7 in the
microfluorimeter chamber at 25°C with 10 mM 2-deoxyD-glucose compared with 1.85 in the absence of 2-deoxyD-glucose. The PCr/ATP ratio was 0.8 when the pacing
rate was increased to 4 Hz in the presence of 2-deoxy-D-
glucose, and again there was no further increase in
ADPfree during rapid pacing. Similarly, in heart cell suspensions stirred at 30°C, the PCr/ATP ratio was significantly decreased from 1.79 to 0.72 ⫾ 0.09 (n ⫽ 6) with
2-deoxy-D-glucose, ATP was 26.6 ⫾ 1.4 nmol/mg protein
(7.6 mM), and PCr was 19.1 ⫾ 3.0 nmol/mg protein (5.5
mM); the calculated ADPfree was 216 ␮M (see Table 2).
No significant drop in the PCr/ATP was observed when
glycolysis was inhibited with iodoacetate (see Fig. 2B).
Steady-state rate of oxygen utake. The rate of respiration of unstimulated heart cells was increased to 48
⫾ 3.7 nmol O2/min per 106 rectangular cells⫺1 (n ⫽ 3)
in the presence of 2-deoxy-D-glucose, and the rate of
respiration during rapid pacing was reversibly and
significantly increased to 61 ⫾ 1.3 nmol O2/min per 106
rectangular cells (n ⫽ 3) (see Table 2). An illustrative
example is shown in Fig. 4 (top trace). Thus, although
initial respiratory rates are higher and ADPfree was
greatly increased, electrical stimulation at 4 Hz with
2-deoxy-D-glucose caused an increase in respiration
(61 ⫺ 48 ⫽ 13 nmol O2/min per 106 rectangular cells) of
the same order of magnitude as that observed in the
absence of 2-deoxy-D-glucose (46 ⫺ 34 ⫽ 12 nmol O2/
min per 106 rectangular cells).
NADH, ADP, AND CONTRACTION OF HEART CELLS
creased by ⬃13 nmol O2/min per 106 rectangular cells
in unstimulated heart cells, corresponding to an additional synthesis of 64 nmol ATP/min per 106 rectangular cells to meet the additional ATP demand. Stimulation at 4 Hz again caused an additional increase in
respiration of 13 nmol O2/min per 106 rectangular cells,
corresponding to a total rate of ATP synthesis of 279
nmol ATP/min per 106 rectangular cells. The increment in respiration with 4-Hz pacing, over and above
that observed with 2-deoxy-D-glucose, was similar to
that observed during rapid pacing without 2-deoxy-Dglucose. The 2-deoxy-D-glucose model permits us to
study the steady state at higher rates of oxidative
phosphorylation and at high ADPfree levels. 2-Deoxy-Dglucose 6-phosphate, formed by phosphorylation of
2-deoxy-D-glucose, is not further metabolized. ATP synthesis must increase not only to compensate for the loss
of glycolytically generated ATP, but also to meet the
new demand for ATP to phosphorylate an (essentially)
infinite supply of 2-deoxy-D-glucose.
It is noteworthy that PCr levels fall to almost onethird of initial levels when ATP utilization is stimulated by 2-deoxy-D-glucose but not when ATP utilization is increased by electrical pacing (Table 1 and Fig.
2B). Because both oxygen consumption and the frequency of contraction can be increased during pacing,
it is unlikely that sequestration of Pi is limiting the
rate of ATP synthase activity. The increase in ADPfree
cannot be attributed to inhibition of glycolysis, because
PCr levels are not decreased in heart cells when glycolysis is inhibited by iodoacetate (Fig. 2B). This suggests
that with 2-deoxy-D-glucose, ATP synthase activity is
not sufficiently rapid to maintain ATP levels at this
rate of ATP utilization. In contrast, during electrical
pacing in either the absence or presence of 2-deoxy-Dglucose, when intracellular calcium is markedly increased (9), PCr does not drop. Here, ATP synthase
activity is increased sufficiently to maintain ATP levels. These results are in accord with suggestions that
calcium may directly increase the activity of ATP synthase (22).
NADH regeneration. The rate of delivery of reducing
equivalents to NAD⫹ at the inner mitochondrial membrane is the sum of mitochondrial and cytoplasmic
dehydrogenase activities (26) (Fig. 5). To test whether
the decrease in NADH observed during pacing is due to
a limitation in the rate of supply of reducing equivalents, we supplemented the extracellular substrate
supply with fatty acids (acetate or octanoate) or ketoacids (acetoacetate or ␤-hydroxybutyrate). Substrate
transport into the cell is not limited, because there
is a high-affinity monocarboxylate transport carrier
present (16) and because insulin (which increases glucose entry into the cells) does not affect NADH. With
␤-hydroxybutyrate as a substrate, there is an additional NADHm-generating step, ␤-hydroxybutyrate dehydrogenase, and NADH levels at both rates of pacing
are increased (14, 25). In the presence of a fatty acid
substrate, metabolite flux occurs mainly through the
␤-oxidation pathway; pyruvate dehydrogenase is inhibited, and glycolytic flux is negligible (1).
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on June 17, 2017
state of the free NADH pool of the mitochondrial matrix (10). Cytosolic NADH reflects the lactate-to-pyruvate ratio and is higher in the presence of lactate than
of pyruvate (19). Because we find that the fluorescence
intensity is not affected by changes in the lactate-topyruvate ratio, the fluorescence signal we detect is
primarily from NADHm. The PCr/ATP ratio, a sensitive measure of hypoxia, remains high in superfused
cells, and oxygen consumption can be increased severalfold over that observed in electrically stimulated cell
suspensions by the addition of CCCP, an uncoupler of
oxidative phosphorylation, demonstrating that oxygen
supply is not limiting.
NADH oxidation. Steady-state oxygen uptake was
measured at different states of work output: heart cells
at rest, heart cells stimulated at 4 Hz, heart cells
uncoupled by CCCP, and heart cells treated with 2-deoxy-D-glucose. During 4-Hz stimulation, the rate of
oxygen utilization increased from the resting level of
34 to 46 nmol O2/min per 106 rectangular cells (see
Table 2). The stimulated rate was ⬃26% of the maximum rate achievable when oxidative phosphorylation
was uncoupled with CCCP (118 nmol O2/min per 106
cells). The magnitude of the steady-state incremental
oxygen uptake during 4-Hz stimulation was not affected by inhibition of glycolysis, decreased PCr levels,
or the NADHm level (see Table 2 and Fig. 4). Thus the
increase in oxygen consumption and, thereby, NADH
oxidation was closely correlated to the rate of stimulation.
Our values of oxygen consumption for resting cardiac
myocytes are of the same order of magnitude as those
reported earlier (11, 18, 21). The fractional increment
in oxygen utilization measured in stirred cell suspensions during 4-Hz pacing reported here is less than
that estimated in electrically stimulated single cells on
the stage of the microfluorimeter (25). The current data
give a more precise measure of the actual increment of
oxygen uptake during stimulation, because measurements in the steady-state chamber permit a more accurate measurement of the rate of oxygen uptake (3–5,
21).
The steady-state rate of oxygen uptake gives a direct
measure of the rate of oxidative phosphorylation in a
tightly coupled system. Assuming a p/o ratio of 2.4
molecules of ATP per oxygen atom (13), the oxygen
uptake during 4-Hz stimulation corresponds to a rate
of ATP synthesis by oxidative phosphorylation of 215
nmol ATP/min per 106 rectangular cells. The increase
in respiration during 4-Hz pacing is not accompanied
by any change in ADPfree, in agreement with observations in the whole heart (7, 8) and in heart cells (21).
These data show that homeostatic mechanisms maintain highly regulated ATP and ADPfree levels in the
cells during increased activity.
With ␤-hydroxybutyrate, lactate, or pyruvate,
ADPfree was 45–55 ␮M (Table 2), not far from the
Michaelis-Menten constant of 20–30 ␮M for ADPfree
for oxidative phosphorylation in mitochondria (7). With
2-deoxy-D-glucose, ADPfree was increased to 216 ␮M,
and the rate of oxygen uptake was significantly in-
H1855
H1856
NADH, ADP, AND CONTRACTION OF HEART CELLS
Contractile Response During Pacing
The most sensitive index of contractile dysfunction
in these experiments is the amplitude of contraction of
single cells in response to rapid electrical pacing. During rapid pacing with glucose plus fatty acids, the
amplitude of contraction was significantly and reversibly decreased (Fig. 1A and Table 1) when NADHm was
decreased, even though the rate of oxygen uptake was
significantly increased (Table 2) and the PCr/ATP ratios were not detectably affected. The effective work
output is diminished even though the rate of oxygen
consumption at comparable ADPfree does not differ
detectably from that when NADHm levels are high.
This may reflect a decreased rate of ATP synthesis per
mole of O2 consumed, as a consequence of reduced
mitochondrial membrane proton electrochemical potential (⌬␮H⫹) when NADH levels fall, because the
⌬␮H⫹ and NADH are in near equilibrium (6).
Contractile amplitude has been shown to decrease
with either decreased intracellular pH or with decreased systolic calcium as a result of direct effects on
the contractile mechanism. However, during stimulation intracellular pH does not decrease significantly
(24), and intracellular calcium is significantly increased.
A decrease in the strength of contraction is not observed when NADHm is maintained during pacing,
either by the enhanced activation of pyruvate dehydrogenase or by a marked increase in ADPfree during
2-deoxy-D-glucose treatment. Thus, when ATP levels
and NADHm are nearly constant, the rate of oxidative
phosphorylation and work output can be increased by
rapid pacing, even at significantly decreased PCr levels, as long as oxygen supply is not limiting.
The following conclusions can be stated in summary:
1) During 4-Hz pacing, the rate of oxygen uptake of
isolated heart cells is increased to meet increased ATP
demand without a change in steady-state high-energy
phosphate levels. The amplitude of contraction during
rapid electrical pacing does not become limited as long
as NADHm levels are maintained. 2) With physiological substrates, NADHm and the contractile amplitude
of contraction are both decreased during pacing. We
suggest that the rate of regeneration of NADHm during
rapid pacing is insufficient to replenish the steadystate levels of NADHm. With glucose and fatty acid
substrates, a rate-limiting reaction(s) in the glycolytic
cycle, the aspartate-malate shuttle, pyruvate dehydrogenase, the tricarboxylic acid cycle, and/or ␤-oxidation
may limit the supply of carbon substrates to mitochondrial dehydrogenases. 3) The addition of either of the
substrates that activate pyruvate dehydrogenase or 2deoxy-D-glucose, which increases ADPfree fourfold,
overrides the rate limitation and maintains both
NADHm and the amplitude of contraction during pacing. 4) When NADH falls during pacing, the increment
in the oxygen consumption rate is not detectably different from that observed when NADH levels are maintained during pacing, but the effective work output is
decreased, suggesting that the efficiency of oxygen utilization for contractile work is diminished. This may
reflect a decreased rate of ATP synthesis per mole of O2
consumed, as a consequence of a reduced ⌬␮H⫹, when
NADH levels fall. 5) The rate of ATP synthesis by ATP
synthase is enhanced to meet the ATP demands of
4-Hz stimulation without a change in high-energy
phosphate levels. With 2-deoxy-D-glucose, however, a
metabolic ATP demand of similar magnitude results in
a large decrease in PCr. Because the PCr drop was not
observed when glycolysis was inhibited by iodoacetate,
the data are in accord with the suggestion that increased intracellular calcium levels during electrical
pacing may directly enhance the activity of ATP synthase.
We gratefully acknowledge Drs. Jonathan B. Wittenberg and
Robert K. Josephson for helpful discussions. We thank Laura Cipriani for technical assistance.
The work was supported by National Heart, Lung, and Blood
Institute Grants HL-42074 (to R. L. White) and HL-19299 (to B. A.
Wittenberg).
REFERENCES
1. Bowker-Kinley MM, Davis WI, Wu P, Harris RA, and
Popov KM. Evidence for existence of tissue-specific regulation
of the mammalian pyruvate dehydrogenase complex. Biochem J
329: 191–196, 1998.
2. Brandes R and Bers DM. Increased work in cardiac trabeculae
causes decreased mitochondrial NADH fluorescence followed by
slow recovery. Biophys J 71: 1024–1035, 1996.
3. Cole RP, Sukanek PC, Wittenberg JB, and Wittenberg BA.
Mitochondrial function in the presence of myoglobin. J Appl
Physiol 53: 1116–1124, 1982.
4. Degn H and Wohlrab H. Measurements of steady-state values
of respiration rate and oxidation levels of respiratory pigments
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on June 17, 2017
Under these conditions, we found that steady-state
NADHm levels are decreased ⬃50% during rapid stimulation in the presence of glucose plus fatty acids or
ketones (Fig. 1A and Table 1). The data are consistent
with the notion that the drop in NADHm with glucose
and fatty acid substrates during rapid stimulation reflects an insufficient rate of supply of reducing equivalents to NAD⫹ to maintain steady-state NADHm levels. The simplest explanation of the data is that during
pacing, delivery of reducing equivalents to mitochondrial NAD⫹ is limited by a rate-limiting reaction(s).
Possible sites of a rate-limiting reaction could be in
glycolytic and the aspartate-malate shuttle pathways,
through pyruvate dehydrogenase (when glucose is a
substrate), or in the tricarboxylic acid cycle (15, 17).
This limitation in the rate of formation of NADHm
with abundant oxygen and high ATP levels during
pacing is not observed in the presence of lactate, pyruvate, or 2-deoxy-D-glucose. An addition of lactate or
pyruvate leads to an increase in the steady-state intracellular concentration of pyruvate, which leads to activation of pyruvate dehydrogenase. The rate-limiting
step(s) in the rate of NADHm formation during pacing
is then overcome or bypassed. The rate-limiting step is
also overcome or bypassed when ADPfree is substantially increased by 2-deoxy-D-glucose.
NADH, ADP, AND CONTRACTION OF HEART CELLS
5.
6.
7.
8.
9.
11.
12.
13.
14.
15.
16. Poole RC, Cranmer SL, Halestrap AP, and Levi AJ. Substrate and inhibitor specificity of monocarboxylatetransport into
heart cells and erythrocytes. Further evidence for the existence
of two distinct carriers. Biochem J 269: 827–829, 1990.
17. Randle PJ, England PJ, and Denton RM. Control of the
tricarboxylate cycle and its interactions with glycolysis during
acetate utilization in rat heart. Biochem J 117: 677–695, 1970.
18. Rose H, Strotmann KH, Popping S, Fischer Y, Kulsch D,
and Kammermeier H. Simultaneous measurement of contraction and oxygen consumption in cardiac myocytes. Am J Physiol
Heart Circ Physiol 261: H1329–H1334, 1991.
19. Scholz TD, Laughlin MR, Balaban RS, Kupriyanov VV,
and Heineman FW. Effect of substrate on mitochondrial
NADH, cytosolic redox state, and phosphorylated compounds in
isolated hearts. Am J Physiol Heart Circ Physiol 268: H82–H91,
1995.
20. Starnes JW, Wilson DF, and Erecinska M. Substrate dependence of metabolic state and coronary flow in perfused rat heart.
Am J Physiol Heart Circ Physiol 249: H799–H806, 1985.
21. Stumpe T and Schrader J. Phosphorylation potential, adenosine formation, and critical PO2 in stimulated rat cardiomyocytes. Am J Physiol Heart Circ Physiol 273: H756–H766, 1997.
22. Territo PR, Mootha VK, French SA, and Balaban RS.
Calcium activation of cardiac oxidative phosphorylation. Direct
evidence for activation of the F0F1ATPase (Abstract). Biophys J
76: A295, 1999.
23. Vuorinen KH, AlaRami A, Yan Y, Ingman P, and Hassinen
IE. Respiratory control in heart muscle during fatty acid oxidation. Energy state or substrate-level regulation by Ca2⫹? J Mol
Cell Cardiol 27: 1581–1591, 1995.
24. White RL and Wittenberg BA. NADH fluorescence of isolated
ventricular myocytes: effects of pacing, myoglobin and oxygen
supply. Biophys J 65: 196–204, 1993.
25. White RL and Wittenberg BA. Effects of calcium on mitochondrial NAD(P)H in paced rat ventricular myocytes. Biophys J 69:
2790–2799, 1995.
26. Williamson DH, Lund P, and Krebs HA. The redox state of
free nicotinamide adenine dinucleotide in the cytoplasm and
mitochondria of rat liver. Biochem J 103: 514–527, 1967.
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on June 17, 2017
10.
at low oxygen tensions. A new technique. Biochim Biophys Acta
245: 347–355, 1971.
Doeller JE and Wittenberg BA. Myoglobin function and energy metabolism of isolated cardiac myocytes: effect of sodium
nitrite. Am J Physiol Heart Circ Physiol 261: H53–H62, 1991.
Doumen C, Wan B, and Ondrejickova O. Effect of BDM,
verapamil, and cardiac work on mitochondril membrane potential in perfused rat hearts. Am J Physiol Heart Circ Physiol 269:
H515–H523, 1995.
From AHL, Bache RJ, Zhang J, and Ugurbil K. Nuclear
magnetic resonance studies of bioenergetics in normal and abnormal myocardium. In: NMR in Physiology and Biomedicine,
edited by Gillies RG. New York: Academic, 1994, p. 413–437.
From AHL, Zimmer D, Michurski SP, Mohanakrishnan P,
Ulstad VK, Thoma WJ, and Ugurbil K. Regulation of the
oxidative phosphorylation rate in the intact cell. Biochemistry
29: 3731–3743, 1990.
Hansford RG. Role of calcium in respiratory control. Med Sci
Sports Exerc 26: 44–51, 1994.
Hassinen I, Ito K, Nioka S, and Chance B. Mechanism of
fatty acid effect on myocardial oxygen consumption. A phosphorus NMR study. Biochim Biophys Acta 1019: 73–80, 1990.
Haworth RA, Hunter DR, Berkoff HA, and Moss R. Metabolic cost of the stimulated beating of isolated adult rat heart
cells in suspension. Circ Res 52: 342–351, 1983.
Kim DK, Heineman FW, and Balaban RS. Effects of hydroxybutyrate on oxidative metabolism and phosphorylation potential
in canine heart in vivo. Am J Physiol Heart Circ Physiol 260:
H1767–H1773, 1991.
Kingsley-Hickman PB, Sako EY, Ugurbil K, From AHL,
and Foker JE. 31P NMR measurement of mitochondrial uncoupling in isolated rat hearts. J Biol Chem 265: 1545–1550, 1990.
Laughlin MR, Taylor J, Chesnick AS, and Balaban RS.
Nonglucose substrates increase glycogen synthesis in vivo in dog
heart. Am J Physiol Heart Circ Physiol 267: H217–H223, 1994.
O’Donnell JM, Doumen C, LaNoue KF, White LT, Yu X,
Alpert NM, and Lewandowski ED. Dehydrogenase regulation
of metabolite oxidation and efflux from mitochondria in intact
hearts. Am J Physiol Heart Circ Physiol 274: H467–H476, 1998.
H1857