| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
O2 and intramuscular 31P metabolite kinetics during high-intensity exercise in humans
St. George's Hospital Medical School, Departments of 1Physiology, 2Biochemistry, and 3Pharmacology, London SW17 0RE; 4Centre for Exercise Science and Medicine, University of Glasgow, Glasgow G12 8QQ, United Kingdom; 5Canadian Centre for Activity and Ageing, University of Western Ontario, London, Ontario, Canada N6G 2M3; and 6Division of Physiology, Department of Medicine, University of California, San Diego, La Jolla, California 92093-0623
Submitted 18 October 2002 ; accepted in final form 7 May 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
O2) was measured
breath by breath (by use of a turbine and mass spectrometer) simultaneously
with intramuscular phosphocreatine (PCr) concentration ([PCr]),
[Pi], [ATP], and pH (by 31P-MRS) and arterialized-venous
blood sampling. DCA had no effect on the time constant (
) of either
O2 increase or PCr
breakdown [O2 45.5
± 7.9 vs. 44.3 ± 8.2 s (means ± SD; control vs. DCA);
PCr 44.8 ± 6.6 vs. 46.4 ± 7.5 s; with 95% confidence
intervals averaging < ±2 s]. DCA, however, resulted in significant
(P < 0.05) reductions in 1) end-exercise [lactate] (-1.0
± 0.9 mM), intramuscular acidification (pH, +0.08 ± 0.06 units),
and [Pi] (-1.7 ± 2.1 mM); 2) the amplitude of the
fundamental components for [PCr] (-1.9 ± 1.6 mM) and
O2 (-0.1 ± 0.07
l/min, or 8%); and 3) the amplitude of the
O2 slow component. Thus,
although the DCA infusion lessened the buildup of potential fatigue
metabolites and reduced both the aerobic and anaerobic components of the
energy transfer during exercise, it did not enhance either
O2 or [PCr],
suggesting that feedback, rather than feedforward, control mechanisms dominate
during high-intensity exercise. magnetic resonance spectroscopy; kinetics; dichloroacetate; O2 uptake; fatigue; phosphocreatine concentration
O2) are integral to the
ability to sustain muscular exercise. However, at exercise onset,
O2 responds relatively
slowly in response to the increased demand for ATP turnover. Therefore
anaerobic mechanisms, such as depletion of high-energy phosphate stores
[predominantly the local phosphocreatine (PCr) pool] and anaerobic glycolysis
with lactate (Lac) accumulation provide supplementary energy to fuel the
contractile process. Together with a contribution from stored O2,
this anaerobic component of the energy provision has been termed the
O2 deficit.
It has recently been demonstrated that O2 availability is
unlikely to mediate O2
control, at least for moderate exercise
(12); rather, the fundamental
component of the control is more likely to reside in the muscles' ability to
utilize O2. Classical theories typically characterize
O2 control as an
"inertial" feedback mechanism operating through features of ATP
splitting, such as [ADP] (8,
where brackets denote concentration), the "phosphorylation
potential" (47) or the
change in Gibbs free energy of cytosolic ATP hydrolysis. More recently,
however, it has been suggested that feedforward mechanisms (with respect to
ATP utilization) may play an important role in
O2 control via the rate
of substrate provision to the electron transport chain
(21,
22,
45,
46).
Activation of the pyruvate dehydrogenase (PDH) enzyme complex before
exercise in humans, via dichloroacetate (DCA) administration, has been shown
to reduce both the magnitude of the lactic acidemia and the degree of
intramuscular [PCr] depletion for a given work rate (e.g., Refs.
45,
46). Thus, although PDH is
known to be activated acutely during exercise (e.g., Ref.
10), these observations
suggest that its activation may be relatively slow. This could introduce a
"stenosis" for substrate provision at exercise onset that may
constrain the rate at which muscle
O2 develops.
The demonstration by Timmons and colleagues
(45,
46) of a reduced reliance on
anaerobic components of the energy demands of the exercise strongly suggests
that O2 [and consequently
pulmonary O2 uptake
(O2)] kinetics may
actually be faster with prior PDH activation. It is hypothesized, therefore,
that substrate availability may normally limit the rate of oxidative
phosphorylation at exercise onset via the rate of PDH activation. Releasing
this "stenosis" would result in more ADP and inorganic phosphate
(Pi) being converted to ATP oxidatively and less ADP being
rephosphorylated at the expense of PCr.
It is now generally agreed that the kinetics of O2 exchange,
both at the lung and exercising muscle, are composed of what may be termed
1) a "fundamental" (i.e., phase II) kinetic component,
the functional "gain" (i.e.,
O2/
)
and time constant () of which are largely independent of work rate
(); and 2) a second slow component of
delayed onset [manifest in
O2
(O2 sc),
O2, and [PCr]
([PCr]sc)], at high-intensity work rates associated with a lactic
acidemia (e.g., Refs. 3,
7,
22,
32,
33,
39,
40,
51). The
O2 sc has been shown to
correlate both in magnitude and time course with arterial [Lac] (unlike the
fundamental component) (e.g., Refs.
32,
41). We therefore hypothesized
that reduced reliance on anaerobic energy provision at exercise onset,
consequent to DCA administration (e.g., Refs.
45,
46), would 1) reduce
the of the early fundamental
O2 kinetics and
2) reduce the magnitudes of the
O2 sc and
[PCr]sc.
It is of interest, therefore, that neither Grassi et al.
(14) in dogs nor Bangsbo et
al. (2) in humans were able to
demonstrate any alteration in either the magnitude or the time course of the
O2 response to exercise
consequent to DCA administration. Neither was an alteration in the [PCr] and
[Lac] responses demonstrable (similar to Gibala and Saltin, Ref.
11), i.e., in contrast to
other previous reports (20,
31,
45,
46). It is important to point
out, however, that the measurement of
O2 is technically
challenging and relies on numerous assumptions. In addition, the temporal
resolution of the [PCr] response profile is generally poor when assessed from
muscle biopsies, especially when the variables of interest may be changing
rapidly; furthermore, a single biopsy site may not be representative of the
entire muscle under investigation. Interrogation of a relatively larger muscle
mass by 31P magnetic resonance spectroscopy (31P-MRS)
attempts to overcome this problem and also provides adequate time resolution
to estimate the parameters of phosphate metabolite kinetics with appropriate
confidence.
We therefore determined the kinetic features of both
O2 and [PCr]
simultaneously in humans, using breath-by-breath gas-exchange measurement and
31PMRS, respectively, during knee extensor exercise after an
infusion of either DCA or a saline placebo (Con).
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
Exercise. The methods of exercise inside the bore of the 1.5-T
superconducting magnet (Signa Advantage, GE, Milwaukee, WI) have previously
been described by Whipp et al.
(53). Briefly, each subject
was initially habituated to the exercise in the Laboratory of Human Physiology
in the same prone position and using the same ergometer as used for the MR
studies. This allowed both 1) subject familiarization and 2)
selection of the appropriate-intensity work rate for the subsequent
31P-MRS experiments (i.e., a level that caused the subject to
fatigue after 
10-12 min at an estimated mean power of 87 ± 16 W;
see Ref. 53 for
calculation).
Simultaneous determinations of
O2 and intramuscular
[PCr], [ATP], and [Pi] were made from 48 tests (8 in each subject).
The subjects lay prone inside the bore of the magnet with their feet suspended
in the rubber stirrups of a custom-designed plastic insert
(53). This ergometer permitted
high-intensity square-wave exercise to be undertaken by means of rhythmic
alternate-leg knee-extensions of constant excursion and frequency (in response
to an audible cue). The contraction phase of the knee extensors of the
nondominant leg was synchronized to occur in unison with the
31P-MRS interrogation of the quadriceps of the relaxed dominant
leg. The subject was strapped down to the scanner table by means of a broad
nondistensible strap placed over the hips. Each exercise session consisted of
two square-wave tests (4 min rest, 8 min exercise, 10 min rest). After an
initial test (test 1), the subjects were removed from the magnet and
sat quietly for at least 1 h before reentering the magnet to complete a second
identical test (test 2).
Infusions. Before test 1 in each session, subjects were infused with either sodium DCA or a saline control (Con). Sodium DCA was prepared by the St. George's Hospital Medical School pharmacy at a concentration of 25 mg/ml (pH 7.0), in the same fashion as Timmons et al. (46). The DCA was delivered intravenously (50 mg/kg body mass) over a 30-min period via an infusion pump. The same volume of saline was delivered in the Con condition. Test 1 began 30 min after the end of the infusion. Each subject completed two DCA sessions and two Con sessions, each on different days. For one DCA session and one Con session in each subject, arterialized-venous blood was sampled from an indwelling venous catheter (20 g) in the dorsum of a heated hand (28) for [Lac] determination (Analox, GM7, London, UK) at 4 and 2 min before exercise onset; at 2, 4, 6, and 8 min during exercise; and at 2, 5 and 10 min postexercise.
31P-MRS sequence. A single-pulse 31P-MRS acquisition was employed using a surface coil (8-in. transmit and 5-in. receive), tuned to a frequency of 25.85 MHz for phosphorus, placed under the quadriceps muscle of the dominant leg midway between the knee and hip joints (53). The coil was securely fastened to the table.
A series of axial gradient recall echo images of the thigh were acquired to
ensure reproducible positioning of the radio frequency coil. Shimming was
performed by using the proton signal of muscle water acquired from the
quadriceps. The 31P-radio frequency excitation pulse was set at a
level to give maximum [PCr] signals at a 2,000-ms repetition rate from an
80-mm-thick axial slice of muscle. Free induction decays for 31P
spectra were collected every 2,000 ms throughout the entire protocol
(rest-exercise-recovery) with a spectral width of 2,500 Hz and 1,024 data
points. 31P-MRS data were averaged over eight acquisitions
(providing a 31P spectrum every 16 s) to estimate the relative
signal intensities of the three ATP peaks (
, 
, and 
), PCr,
and Pi every 16 s.
Signal intensities of each resonance were calculated automatically (as a
batch job of the time-course data) by means of the time-domain
variable-projection fitting program, VARPRO
(48) and the Java-based MR
user interface (jMRUI) (30),
using appropriate prior knowledge of the ATP multi-plets
(44). The longitudinal
relaxation time (T1) saturation factor was assumed to remain
constant for each resonance throughout the experiment. Intramuscular pH
(pHi) was estimated from the chemical shift of the Pi
peak relative to the PCr peak in the 31P spectrum, by using the
relationship determined by Moon and Richards
(29)
 | (1) |
is the chemical shift of Pi relative to PCr. The
concentration of PCr (and Pi) was calculated from the
PCrto--ATP (and Pi-to--ATP) peak area ratio, by
assuming that the area of -ATP represented a concentration of 8.2 mM
(17).
Pulmonary gas-exchange measurement.
O2 was determined breath
by breath (CaSE, Gillingham, Kent, UK; by use of the algorithms of Ref.
4) simultaneously with the
phosphate metabolite determination
(53). Inspired and expired
volumes were measured by a custom-designed nonmagnetic turbine and a volume
measuring module (VMM, Interface Associates, Laguna Niguel, CA) calibrated
with a 3-liter syringe before each experiment (Hans Rudolph, Kansas City, MO).
O2, CO2, and N2 concentrations were measured
by using a quadrupole mass spectrometer (QP9000, CaSE), calibrated against
precision-analyzed gas mixtures. Gas was drawn continuously from the
mouthpiece along the extended 45-ft sampling capillary line, which had a 5-95%
rise time of <80 ms
(53).
Kinetic analysis. Kinetic analyses were performed on the
O2 and [PCr] responses by
nonlinear least-squares fitting (Origin, Microcal)
(38). Responses from each
subject were analyzed individually. After any extraneous outlying values were
edited out (see Ref. 37), the
O2 and [PCr] responses
for each test were interpolated on a second-by-second basis and time aligned
(exercise onset corresponding to time zero). The responses during
like conditions (i.e., Con or DCA) were averaged, and then the ensemble was
time averaged (10 s for
O2 and 15 s for [PCr]) to
produce a single response for each subject during both the Con and DCA
trials.
The fundamental (phase II) on-transient response was assumed to be that
portion of the response that conformed to a simple exponential, beginning at
time zero for [PCr] (Eq. 2) and after the phase I (or
"cardiodynamic") duration for
O2 (Eq. 3)
 | (2) |
and
 | (3) |
O2 0 and
PCr0 are the values of
O2 and [PCr] at
t = 0, O2
ss and [PCr]ss are the asymptotic values above
baseline to which the fundamental phase of
O2 and [PCr] project,
is the time constant of the responses, and is a delay term
similar to (but not equal to) the phase I-phase II transition time
(50).
The O2 deficit of the fundamental (O2Df)
was calculated as
 | (4) |
' is the
O2 mean response time
estimated by using the function in Eq. 3 with constrained to
the onset of exercise (i.e., = 0) and limiting the data field from
exercise onset to the onset of the slow component (i.e., including phase I)
(see Ref. 50).
As previously described
(38) the fitting strategy was
designed to identify, a posteriori, the onset of any slow component in the
O2 and [PCr] responses.
The fitting window was lengthened iteratively (beginning at the fundamental
onset to, initially, 60 s) until the exponential model-fit demonstrated a
discernible departure from the measured response profile. The goodness of fit
was determined by 1) the maintenance of the flat profile of the
residual plot, and 2) the demonstration of a local threshold in the
2 value. The magnitude of the slow component for both
O2 and [PCr] was then
estimated from the difference between the steady-state amplitude of the
fundamental (i.e., O2
ss and [PCr]ss) and the amplitude at 8 min of
exercise.
The confidence limits for the parameter estimation
(37) and the 2
values were also obtained; confidence was set at 95% and tolerance at 5%
(i.e., P < 0.05). The differences between parameter values were
examined by Student's t-test or repeated-measures ANOVA with
Scheffé's post hoc testing where appropriate. Values are given as means
± SD or 95% confidence intervals where indicated, and a P
value of <0.05 was used as the criterion for the rejection of the null
hypothesis.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
O2
and [PCr] kinetics. Differences between Con and DCA were reproducible
across each of the four trials for both infusions with the trials randomized
and the subject blind to the infusate. Averaging of
O2 and 31P-MRS
responses allowed high-confidence discrimination of their kinetic features.
Representative examples of the kinetic responses of the on-transient of
O2, [PCr],
[Pi], and pH are shown in Fig.
2 during Con (Fig.
2, left) and DCA (Fig.
2, right) for a single subject.
O2 was characterized by
an early cardiodynamic region (phase I) followed by the fundamental (phase II)
kinetics and the slow component. [PCr] was well described by two phases: the
fundamental and the slow component. [PCr] began to fall immediately at the
onset of exercise and fell to a level that was consistent with the expected
fundamental exponential; i.e., during the first 16 s [PCr] did not fall less
or more than expected from the remainder of the exponential response in all
cases. Once phase aligned, the [PCr] and
O2 on-transient kinetics
were essentially indistinguishable from each other
(Fig. 2A, for example)
during both Con and DCA conditions in all subjects. The [Pi]
on-transient response, however, was more complex (similar to Ref.
40): [Pi] typically
rose toward a new steady state but then either attained the expected
asymptote, continued to increase, or began to fall. Regardless, the
end-exercise [Pi] was lower in all subjects for the DCA tests (12.6
± 2.8 mM) compared with Con (14.2 ± 3.3 mM) (P =
0.05).
|
Examples of the exponential model fits to the fundamental region of the
O2 and [PCr] responses
are shown in Fig. 3A
for both DCA and Con. The associated confidence intervals suggest that the
responses were similar to within a 95% confidence of 3.2 ± 1.2 s (i.e.,
< ±2 s, or ±4%). Average fundamental on-transient
values for all six subjects were
O2 = 44.3 ±
8.2 s and [PCr] = 46.4 ± 7.5 s for DCA, compared with
O2 = 45.5 ±
7.9 s and [PCr] = 44.8 ± 6.6 s for Con
(Table 1). ANOVA revealed no
difference between
O2 for DCA and Con,
no difference between [PCr] for DCA and Con, and no difference between
O2 and [PCr]
for either DCA or Con.
|
|
In addition, the fundamental amplitude of
O2
(O2 ss) was
reduced in the DCA condition, relative to Con, for all but one subject; in
Fig. 3B, the Con and
DCA O2 responses are
superimposed to illustrate the most striking example of this effect. Thus
O2 ss averaged
0.84 ± 0.21 l/min during Con but only 0.77 ± 0.20 l/min during
DCA (P < 0.05; Table
1). There was no difference in the baseline
O2, however.
The DCA-related reduction in the fundamental
O2 response amplitude
(O2 ss) was
accompanied by a similar reduction in [PCr]ss (P
< 0.05; Fig. 3B).
This was the case for every subject and represented a 1.9 mM
"sparing" of PCr on average (i.e., [PCr]ss for
DCA = 9.0 ± 4.1 mM vs. for Con = 10.9 ± 3.3 mM)
(Table 1). Resting [PCr] and
[PCr]/[ATP] was, on average, unaffected by DCA administration, although there
was a variation of 9% among subjects. Consequently, although the
confidence of the absolute [PCr] estimation may be poor, the relative changes
within subjects in this repeated-measures design are more robust. The
fundamental [PCr] asymptote, therefore, averaged 20.4 ± 2.2 mM with DCA
compared with 19.0 ± 1.9 mM for Con.
The reduction in the fundamental amplitude of the
O2 and [PCr] responses
had no effect on the time course of their change. That is, when the responses
were normalized to the fundamental amplitude (i.e.,
O2 ss,
[PCr]ss), their kinetics were not altered
(Fig. 3C, for the same
subject as shown in Fig.
3B).
The magnitude of the [PCr]sc
(Table 2) was approximately
halved from -0.66 ± 0.7 (Con) to -0.28 ± 0.3 mM (DCA). However,
this difference was in the responses of only three of the subjects, with
effectively no change in the remainder. The
O2 sc, however, was
reduced in all subjects from Con to DCA: from 108 ± 67 to 80 ±
51 ml/min, respectively (P < 0.05)
(Table 2).
|
Estimated ATP turnover rate. We could discern no effect of DCA on
the resting [PCr] or [PCr]/[ATP]. The initial rate of [PCr] change
(d[PCr]/dt) at exercise onset (argued to reflect the ATP turnover
rate: e.g., Ref. 23) as
determined from the exponential fit of the fundamental phase
(37) was reduced by 20%
in the DCA condition: 11.3 ± 3.7 (DCA) vs. 14.4 ± 2.6 mM/min
(Con) (P < 0.05; Table
2).
Components of the O2 deficit for the
fundamental phase. Despite the fundamental
O2 being unchanged
between DCA and Con, the mean response time (') tended to be
greater (P > 0.05) for the DCA condition (1.10 ± 0.26 min)
than for Con (1.02 ± 0.23 min). This reflected the significantly
(P < 0.05) longer
O2 delay (;
Eq. 3) for DCA (24 ± 7 s) compared with Con (16 ± 7 s)
(Table 1).
The lengthening of ' and the reduction of
O2 ss with DCA
effectively offset each other, such that the O2 deficit for the
fundamental phase (Eq. 4) was not affected:
O2Df = 0.86 ± 0.37 liters for DCA compared with
0.87 ± 0.38 liters for Con (Table
2).
Blood lactate and intramuscular pH responses. End-exercise [Lac]
was reduced (P < 0.05) with DCA in every subject, from 2.7
± 1.8 mM (range 1.2 to 6.0 mM) for Con to 1.8 ± 1.1 mM (range
1.0 to 3.9 mM) with DCA (Fig.
4, Table 2). During
exercise, [Lac] was reduced by 20% with DCA at every time point
(P < 0.05) after minute 2, the preexercise resting [Lac]
was also reduced by 0.15 ± 0.22 mM with DCA.
|
Resting pHi (Eq. 1) was not different between the Con
and DCA conditions, averaging 7.06 ± 0.03 and 7.07 ± 0.01,
respectively. After exercise onset, pHi characteristically
manifested a small alkaline shift (peaking at 30-45 s); this shift was
greater during the DCA condition. Mean intramuscular pH subsequently fell in
both conditions to end-exercise values of 6.91 ± 0.14 in Con and 6.99
± 0.06 during DCA (P < 0.05;
Table 2). Determination of
pHi, however, was complicated by the inconsistent behavior of the
Pi peak at this exercise intensity. Four of the six subjects
expressed a clearly discernible splitting of the Pi peak.
Interestingly, in accordance with a DCA-related reduction in acidosis, the
degree of the chemical shift of the "acid" Pi peak was
reduced after DCA administration in all these four cases (a typical example is
shown in Fig. 5). The more
alkaline of the two Pi peaks at end-exercise averaged 7.02 ±
0.02 during Con and 7.07 ± 0.01 during DCA, and the acid Pi
peak region averaged 6.57 ± 0.29 in Con and 6.73 ± 0.20 during
DCA.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
O2 kinetics at exercise
onset, with a concomitant reduction in the acetyl group deficit (e.g., Refs.
45,
46) and speeding of the
O2 response.
However, conflicting results in several recent studies have clouded this
issue. On the one hand, for example, in some reports neither PCr utilization
nor blood [Lac] during exercise was observed to be significantly reduced after
DCA administration in humans
(2,
11,
21) or for electrically
induced exercise in dog gracilis muscle
(14). These findings contrast
with those of Timmons and colleagues
(45,
46), Howlett et al.
(20), and Parolin et al.
(31). The explanation for
these differences may be due, in part, to the intensity of the exercise
studied (2). For example, both
Howlett et al. (21) and
Bangsbo et al. (2) used
severe-intensity exercise (52)
characterized by a progressively increasing blood [Lac]. Despite almost
complete activation of PDH before exercise onset after DCA administration,
these authors were unable to determine any differences in PCr breakdown or
blood [Lac]. In contrast, the studies of Timmons et al.
(46), Howlett et al.
(20) and Parolin et al.
(31) demonstrated significant
reductions in PCr utilization, glycogen utilization, and blood [Lac]
accumulation after DCA administration. These studies were undertaken at a
lower intensity with exercise durations of between 8 and 15 min; i.e., within
the heavy-intensity domain
(52). In the present study, we
also utilized heavy-intensity exercise, to allow the
O2 sc to develop to a
discriminable level (work rates being chosen to elicit a fundamental amplitude
of O2 equivalent to
70% maximal O2). In
addition, in agreement with Timmons et al.
(46), Howlett et al.
(20), and Parolin et al.
(31), we observed, by our
noninvasive monitoring technique, a significant reduction in PCr breakdown
after DCA administration that was sustained throughout the 8-min exercise
period and accompanied by a reduced blood [Lac] from 2 min to end
exercise.
The observation of a DCA-induced reduction in the exercise-induced PCr
breakdown should, according to the hypothesis of Greenhaff and colleagues
(15,
46), result in a speeding of
the response kinetics of
O2 and therefore
O2. In the present study,
however, this was not the case. We could demonstrate no reduction in the
fundamental for either [PCr] or
O2 after DCA infusion,
despite less PCr utilization (e.g., Fig.
3). We contend that this lack of effect was not a reflection of
poor kinetic discriminability, as the values were discerned with a 95%
confidence intervals averaging 4 s (i.e., ±2 s; see below). In
turn, this suggests that traditional phosphate-linked control mechanisms for
O2 (e.g., Refs.
8,
47) may well predominate, as
the tight coupling between the kinetics of
O2 and [PCr] remained
similar in the presence of DCA. Interestingly, the DCA-related
"improvement" in aerobic function, manifest at the muscle as a
reduction in anaerobic components of the O2 deficit, was not the
case for the O2Df at the lung. This was due to a longer
delay time for O2
( in Eq. 3; Table
1) in all cases during the DCA condition, necessitating (as [PCr]
and [Lac] contributions were reduced) a greater contribution to the energy
transfer from stored O2.
Interestingly, however, despite the lack of effect of DCA on
O2, the fundamental
O2 amplitude was reduced
by some 8%, in a similar fashion to the creatine supplemented cycle ergometry
of Jones et al. (22). This was
a surprising feature of our results, which was unlike the priming effect of
prior high-intensity exercise demonstrated in a previous study using similar
techniques, in which
O2 was reduced
(38). We also found that the
initial rate of [PCr] change (d[PCr]/dt), considered to reflect the
total exercise-induced rate of ATP turnover (e.g., Ref.
23), was also reduced (by
20%) subsequent to DCA administration. Greenhaff et al.
(15) also showed that the
anaerobic component of ATP production was reduced after DCA administration by
a similar degree and was maintained over the first 5 min of exercise. The
mechanisms for this apparent reduction in ATP requirement at constant
,reflected in both the d[PCr]/dt and
O2, are at present
unclear. Contributing factors could include 1) a shift to a greater
contribution from muscle fibers with lower
"O2 gain,"
although whether this would reflect greater contribution of type I or type II
fibers is currently disputed (see Ref.
3 for discussion); 2)
a reduction in pHi (in either the highly acid-producing muscle
fibers or in the muscle as a whole); 3) a greater reliance on
oxidative carbohydrate metabolism (a "complete" shift from fatty
acid to carbohydrate oxidation would only account for an 6% improvement,
i.e., P-to-O2 ratio of 5.6 and 6, respectively); and 4) a
greater mechanical efficiency of performing the task (all of our subjects
spontaneously reported that the exercise felt "easier" after DCA
administration, despite their remaining blinded to the experimental condition
until all exercise bouts were completed).
Potential mechanisms of lessening fatigue with DCA administration.
An alteration in the fatigue status or fatigue index could provide an
important mechanism by which DCA would improve the overall energy status of
exercising muscle. Grassi et al.
(14) found that the fatigue
index in electrically stimulated dog gracilis muscle was improved by 8%
after DCA infusion. This, however, was manifest via a better maintenance of
force production at the same
O2 with DCA rather than
by a reduced O2. The net
effect, intriguingly, is similar to the present study, as we have demonstrated
an 8% reduction in
O2 at a constant work
rate, and, hence, constant force production.
In the present study, a reduction in fatigue in either fibers with low
O2 gain (i.e., low
O2/)
or a reduced reliance on those highly acid-producing fibers may lead to an
improvement in oxidative efficiency of the exercise; more strictly, after DCA
infusion, this would be expressed as a lesser reduction in oxidative
efficiency as exercise progresses.
It seems unlikely that DCA would directly affect muscle recruitment through a neuromuscular adaptation, although this should, of course, not be ruled out. The mechanism is more likely to reside within the muscle itself. The main factors that are thought to lead to fatigue within exercising muscle include alterations in Ca2+ release and sensitivity, [Pi], and pHi (e.g., Refs. 1, 49). Although we have no measure of the former, we have shown improvements in both the [Pi] and the pHi responses during the DCA condition. That is, DCA administration reduced the pHi decrement on exercise, the magnitude of the [Pi] response, and the splitting of the Pi peak in the 31P-MRS spectrum. The latter has been suggested to reflect the magnitude of fast-twitch fiber recruitment (e.g., Ref. 55), and, as such, it could be inferred here that there was a reduced reliance on poorly efficient, glycolytic muscle fibers. However, our laboratory (among others) has recently suggested (41) that the split Pi peak is more likely to reflect regional distributions of muscle force production within the region of 31P-MRS interrogation (i.e., the quadriceps). Nevertheless, the reduction in the split Pi peak magnitude (in each of the 4 subjects in whom this phenomenon was manifest) would suggest either that there was a more homogenous force production profile with the exercising limb (with consequent lower force per fiber recruited) and/or that there was a reduced recruitment of poorly efficient, acid-producing fibers during exercise, presumably consequent to reduced fatigue.
DCA-induced effects on fatigue-inducing variables: [Pi] and pH. The increase in [Pi] has been suggested to be closely related to fatigue during exercise (49). We demonstrated here that there was a reduction in the end-exercise [Pi] with DCA, presumably because of the reduced [PCr] breakdown. However, the kinetics of [Pi] are complex. They may depend on multiple factors, not all related to [PCr] metabolism, e.g., Pi-to-phosphate-monoester trapping (5) and Pi uptake in the sarcoplasmic reticulum (49). Nevertheless, significant (P < 0.05) reductions in both [Pi] and calculated diprotonated [Pi] observed in the present study would be consistent with reduced fatigue.
Another possibility is the direct effect of DCA in reducing the acid stress
in the exercising muscle. It has been suggested that the [PCr] asymptote
during exercise will be related to both the of
O2 (e.g., Ref.
38) and the pHi
(e.g., Ref. 9). Because
O2 (and by inference
O2) was not altered
by DCA infusion in our study (or in Refs.
2 and
14), it is reasonable to
assume that the lesser reduction in pHi seen during the DCA
condition would necessitate a lesser reduction in [PCr]. This interaction of
pHi and [PCr] responses during high-intensity exercise (that
generates a metabolic acidosis) has been suggested to be necessary to maintain
the provision of ADP to the mitochondrion
(9). Because [ADP] is regulated
through the creatine kinase (CK) reaction and is dependent on both
pHi and [PCr], it is therefore possible that, to maintain [ADP]
supply, the acid-reducing effect of DCA administration would necessitate an
altered [PCr] amplitude via the CK equilibrium. At constant [ADP], therefore,
a reduced acid shift of 0.1 pH unit would lead to a reduction in the [PCr]
decrement by 1.5 mM (23),
i.e., close to that observed here. The [PCr]sc may be a
manifestation of this effect, as we have previously proposed
(39,
40). It has recently been
suggested, however, that CK may not be in equilibrium at times when [PCr] or
pHi are rapidly changing
(24); thus the traditional
calculation of [ADP] assuming CK equilibrium may not be valid, especially in
the early kinetic region. However, in the present study, end-exercise [ADP]
(in which [PCr] and pHi were more stable) was significantly reduced
during DCA (compared to Con) by 30%.
O2
sc and [PCr]sc. It has been suggested
that the O2 sc is
manifest via a mechanism related to progressive fast-twitch fiber recruitment
during high-intensity exercise
(51). Consistent with this,
direct measures of O2
(34) and more recent
31P-MRS approaches
(38,
39) have demonstrated that the
O2 sc originates (to a
large extent, e.g., 85-90%) within the exercising muscle. Furthermore,
Casaburi et al. (7) and Poole
et al. (33) showed that the
reduction in the O2 sc
observed with training was associated with the reduction in blood [Lac]. It
was therefore hypothesized in the present study that the reduction in the
exercise-induced acid stress observed with DCA administration (e.g., Refs.
20,
45,
46) would be associated with a
reduction in the magnitude of the
O2 sc. This was indeed
the case: the O2 sc was
reduced by 28% on average and this change was accompanied by a 32%
reduction in end-exercise blood [Lac] between the Con and DCA exercise bouts.
However, the increase in blood [Lac] was small in the present study, probably
because of the relatively small muscle mass involved in the exercise. The
close relationship between the alteration in blood [Lac] and the magnitude of
the O2 sc, however, is
consistent with the findings of Casaburi et al.
(7) and Poole et al.
(33) and with the hypothesis
that DCA administration may have reduced the concentration of metabolites
predisposing the exercising muscle to fatigue. Hence, DCA may have reduced the
requirement for recruitment of low oxidatively efficient muscle fibers
(presumably type II, although see Ref.
3 for a different
consideration). These alterations were also accompanied by a 0.4 mM reduction
in the [PCr]sc, consistent with our previous findings
(38). However, when the
O2 sc and
[PCr]sc values were expressed as a percentage of the fundamental
amplitude, their magnitude between Con and DCA conditions was not
significantly different. This suggests that the fundamental gain may play a
role in modulating the magnitude of the
O2 sc, although there are
currently no sufficiently persuasive data to support this hypothesis.
Other effects of DCA. It is important to note that DCA
administration has been reported to have effects other than simple activation
of the PDH complex. DCA has been shown, for example, to have inotropic and
hemodynamic effects in resting adults
(6,
27,
43). Ludvik et al.
(27) found that cardiac index
was increased by almost 20% at rest and that this was accompanied by a similar
reduction in peripheral resistance. More recently, however, Bangsbo et al.
(2) found no difference in
either resting or exercising thigh blood flow, in response to DCA
administration, using the knee extensor model; however, they also found no
effect on [PCr] and [Lac] during exercise. We, however, had no measure or
reasonable estimate of cardiac output or muscle blood flow in these studies;
others have shown that increasing muscle blood flow did not speed the kinetics
of O2 at moderate
exercise (12) or consistently
so at high work rates (13) in
the dog.
Discriminability. The issue of kinetic resolution is naturally of
importance. As previously shown by Lamarra et al.
(25) and Rossiter et al.
(37), the confidence with
which the time constants for the exponential kinetics can be established
depends on an interplay among the response amplitude, the time constant, and
the characteristics of the system "noise." In these studies we
have attempted to minimize these confounding factors by 1) performing
simultaneous O2 and [PCr]
measurement, 2) maximizing the amplitude of the response by selecting
both a large muscle mass exercise and a high-intensity work rate, and
3) performing repeated measures and appropriate averaging and
parameter estimation techniques, such that the average 95% confidence
intervals were < ±2 s. For the 2 mM difference in [PCr]
expressed over, for example, 6 kg of muscle, the 12 mmol would be
equivalent to some 45 ml of O2 at a P-to-O2 ratio of 6.
Because the O2 deficit at the muscle is equal to
O2 ss x
, the O2 equivalent of 45 ml conservation in [PCr] with an
unchanged fundamental of 0.75 min
(Table 1) would be equivalent
to 45 x 4/3, or 60 ml/min in
O2 ss. This is close to
our measured change in O2
ss. Were to have changed without a change in amplitude (i.e.,
800 ml/min) as originally hypothesized, then the change in would
have been 45/800, or 3 s, which is discriminable by our methods.
In summary, in line with the suggestions of Timmons and colleagues
(45,
46), it appears that the
muscle O2 deficit (as reflected by [PCr] and blood [Lac]) was
reduced after DCA administration. However, although the (and
') were similar for both [PCr] and
O2, the amplitude of both
responses was reduced after DCA administration. This was such that the close
kinetic coupling of O2 to
[PCr] remained intact, suggesting that the availability of key TCA
intermediates does not exert significant flux control to
O2 (and by inference
O2): because the
mechanism is capable of constraint does not mean that it is necessarily
constraining. It appears that the traditional view of feedback control via
[PCr] or related phosphate compounds (e.g., [ADP], Ref.
8; or the
"phosphorylation potential," Ref.
47) would provide an adequate
explanation for the kinetics of
O2. However, the
DCA-derived reduction in the acid stress to the exercising muscle due to the
prior release of the flux "stenosis" of PDH may allow a
significant reduction in the buildup of fatigue metabolites such as
Pi, H+, or H2PO4-, as
suggested by the data in the present study. This in turn would necessitate a
lessened requirement for continued (and additional) fiber recruitment
throughout the exercise period, leading to an improved
O2/W with
DCA infusion.
|
DISCLOSURES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
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.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONDISCLOSURESACKNOWLEDGMENTSREFERENCES |
|---|
O2 during heavy exercise.
J Appl Physiol 62:
199-207, 1987.
O2 onkinetics in isolated
in situ canine muscle. J Appl Physiol
85: 1394-1403,
1998.
O2 on-kinetics in canine
muscle contracting at peak
O2. J Appl
Physiol 89:
1293-1301, 2000.
) and [PCr]
(
) on-transient responses to the high-intensity knee extensor exercise
after infusion with saline control (Con) or DCA. [PCr] is phase shifted to
facilitate comparison. B: simultaneously determined intramuscular
[Pi] dynamics in Con and DCA. C: calculated intramuscular
pH responses during Con and DCA. All responses are from the same
representative subject.
) trials. C: comparison of the
, and pHLO, 
,
respectively).