|
|
|
|||||||
|
|||||||||
(Received for publication, October 20,
1994; and in revised form, March 3, 1995) From the
The hypothesis tested was whether creatine kinase (CK)
equilibrates with its substrates and products in the cytosol as if in
solution. We used the creatine analogs cyclocreatine (cCr) or
One of the key tenets of bioenergetics is the near-equilibrium
property of creatine kinase (CK) ( By a
variety of criteria, heterogeneity of cell types in individual muscles
have been
documented(15, 16, 17, 18, 19) ,
and chemical contents of these cell types have shown that their
metabolite contents differ. Furthermore it has been shown that these
differences are reflected in chemical potentials of predominantly fast versus slow muscles (-68 to -61 kJ/mol
ATP)(20) . In addition to heterogeneity of metabolite
distribution, there is also considerable information on the existence,
properties, and intracellular distribution of CK isoforms (3, 21, 22) . Although the major isoform is
MM-CK (which is thought to be in solution in the cytoplasm as well as
bound to the M-line of myofibrils), there is also a substantial amount
of ``mitochondrial'' isoforms existing in dynamic multimers
in the intermembrane space of mitochondria(3) . Based on the
existence of multiple isoforms of CK taken together with heterogeneity
of metabolite contents(23, 24) , it has been argued
that the simple concept of CK equilibration, with its substrates and
products mixing in the bulk cytoplasm as if in solution, is not
completely valid. In view of the large number of physiological and
biochemical studies that have assumed a simple behavior of the cytosol
(with respect to CK function), we investigated this issue by
capitalizing upon the competitive inhibitor nature of two creatine
analogs ( Feeding synthetic analogs of creatine (which as competitive
inhibitors are not expected to disturb equilibration) partially
displaces PCr and Cr from the cell. By choosing duration of feeding
such that there were approximately equal concentrations of total
creatine and total analog with minimal adaptive changes, (
In our experiments, combined reactions are
measured in solution at the temperature of interest with a similar set
of measurements then performed on muscles of animals treated with
analogs. If the reaction achieved equilibration with its substrates in
muscle as in solution, the value for K
consider the reaction for the analog (X) which is phosphorylated
at equilibrium (PX),
At pH 7.0, equilibrium expression for and are written as
follows.
With adenylates at equilibrium with both analog and creatine
species, then we can define a K
In muscle, when the mixture of substrates and a competitive
inhibitor comes to equilibrium with the same adenylate pool, the
observed mass action ratio (
Muscles were
superfused with phosphate-free saline equilibrated with 95%
O Two methods of quantification were used for spectral analysis. In
solutions where signal-to-noise and spectral resolution were high and
base lines strictly flat, spectral peaks were integrated by summing
digitized data symmetrically around each peak. Each integral value was
expressed as fraction of the total phosphorus integral within the
spectrum and reported in absolute chemical content based upon ATP
concentration measured by optical spectroscopy as described above. In
isolated hindlimb muscles, time domain fitting of free induction decay
was used (45) in a commercially available package (FITMASTERS,
Philips Medical Systems). As with solution data, integrals were
expressed as fractional areas and normalized to chemical content by
HPLC-determined ATP content in µmoles
Figure 1:
Two such experiments (with least variance) were used to
construct HPLC standards for phospho-analog in HPLC assays. These
standards are essential for measuring PX concentrations in
freeze-clamped muscles and for cCr treated animals where The values in all
experiments for K
Figure 2:
Fractional peak areas of EDL and
SOL muscles of control and analog-fed mice are presented in Table 2. Control values for PCr, P
Figure 3:
Endogenous substrates for CK (except ADP)
as determined by anion (PCr and ATP) and cation (free creatine)
exchange HPLC for control (open bars) and
The contents of phosphorylated and free analog from
Figure 4:
Synthetic substrates for CK in
Figure 5:
Comparison of HPLC (closed bars)
and NMR (open bars) ratios of PCr/ATP, PCr/
Figure 6:
Combined equilibrium constants for cCr/Cr
and
The result in To test whether there is any
significant binding of
Rats fed One of the key tenets of bioenergetics of excitable tissues
is the near-equilibrium property of CK coupled with the assumption that
cytosolic metabolites involved in the reaction are freely mixing and
hence available to the enzyme. The existence of multiple isozymes of CK
and demonstration of their localization into subcellular compartments,
perhaps with their associated substrate
pools(23, 24) , has suggested that the concept of
freely diffusing metabolites mixing cytoplasm may be too simplistic. In
light of the importance of this concept to interpretation of
biochemical data, in particular by In cCr experiments, The
remaining discussion considers three further issues that show cell to
cell differences in ATP/ADP are the only physiologically important
factors determining whether observed mass action ratio from tissue
equals K
To test whether quantitative agreement
occurs in muscle between NMR and HPLC of their respective perchloric
acid extracts, a similar analysis was performed as in solutions for in vivo data from control and treated muscles. From Fig. 5, PCr/ATP for all muscles are in excellent agreement using
The second possibility to explain lack of equilibration is
based upon enzyme kinetics and suggests that CK fluxes under conditions
of our muscle experiments may not be sufficient to maintain equilibrium
in resting muscle in vivo in the presence of competitive
inhibitors during 3 weeks of feeding or following metabolic
perturbations during tissue handling in the experiment. Arguments
against this hypothesis are derived from in vivo muscle
experiments from other investigators.
What aspects of metabolite distribution could explain
the observed disequilibration with Results with cCr and
previously established equilibration of CK with The following equations illustrate the effects of analog, Cr,
and adenylate distribution within the sample on calculation of
where
If we constrain each fractional volume to its thermodynamic
equilibrium by setting the solution combined equilibrium constant K
As in Case 1, it follows from these equations that as long as fPCr for both volume fractions is identical then
where K
Volume 270,
Number 21,
Issue of May 26, 1995 pp. 12428-12438
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
P NMR STUDIES USING CREATINE ANALOGS (*)
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
APPENDIX
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
-guanidopropionate (GPA) to test if mass action ratios
(
) for CK in muscle could be predicted from combined equilibrium
constants (K
) measured in solutions mimicking
the intracellular environment. Mice were fed cCr or GPA and their
muscles assayed for substrates and products of the CK reaction by P NMR spectroscopy and high performance liquid
chromatography. After three weeks of feeding, was
indistinguishable from K in cCr-treated muscles
demonstrating both PCr/Cr and phospho-analog/analog must have
equilibrated with a constant and uniform cellular ATP/ADP ratio. In
GPA-treated muscles, was smaller than K due to a
higher content of muscle GPA. Feeding GPA for
9-12 weeks resulted in a closer agreement between K and , suggesting ATP/ADP ratios are not
uniform within the muscle perhaps due to transient metabolic stress in
some cells. From this analysis it follows that calculation of free ADP
from the CK equilibrium for a heterogeneous population of cells with
respect to total Cr and ATP content is correct only if chemical
potentials of these cells are uniform.
)coupled with cytosolic
substrates which are freely mixing and available to the
enzyme(1, 2, 3, 4) . Three criteria
have been used to demonstrate that a near-equilibrium condition exists
within the cytoplasm: (a) kinetic limitations are considered
minimal because concentrations of substrates are in the approximate
range of their K
values(5, 6) ; (b) CK activity is far
in excess of maximal ATPase activity within the cell(1) ; (c) P NMR spin transfer methods have shown that
forward and reverse fluxes for CK are equal (7) and are far
greater than net ATPase activity(8) . Thermodynamic control of
PCr/Cr by ATP/ADP through CK equilibration implies that ATP/ADP is the
same in all loci within the cell(1, 9, 10) .
Alternatively, the cytoplasmic fraction of ATP and ADP must be so large
that other ``compartments'' (which may or may not be at the
same chemical potential) do not influence physicochemical properties of
bulk cytoplasm. For a simple system which is not diffusion limited, it
has been shown that this near-equilibrium formulation can be used for
calculation of ADP
and ATP chemical potential
(
G
)(11, 12, 13, 14) .
For complex populations of cells, comprising tissues or organs, a
crucial assumption for proper calculation of ADP and
G (which are derived values from measured
parameters (PCr, Cr, ATP, P
, and pH) on a population of
cells) is that not only is ATP/ADP within each cell uniform,
but it is uniform between cells as well. -guanidopropionic acid (GPA) and cyclocreatine
(cCr)) to establish a criterion for testing CK function in
vivo. Application of creatine analogs to study cellular functions
of CK was pioneered by Fitch et al.(25, 26, 27) and
Walker(28, 29, 30, 31) . Analogs of
creatine deplete total Cr over a period of several weeks when fed to
rodents and chickens. Both phosphorylated and free forms of analog
accumulate in brain, heart, and skeletal
muscle(27, 28, 29, 32, 33, 34, 35, 36, 37) .
As competitive inhibitors, these compounds can be exploited to answer
questions in cellular biology concerning organization of enzymatic
activity in cells, because these analogs are also utilized by cells as
substrates for cellular energy
metabolism(34, 37, 38, 39, 40, 41) .
So it is possible to test whether competitive inhibitors and endogenous
substrates obey simple rules of enzyme kinetics and equilibration. 
)we optimized the ability to measure mass action ratios for
CK and to test whether ratios in tissue equaled those predicted from
combined equilibrium constants measured in solutions mimicking the
intracellular ionic environment. There are a number of reasons why this
equality may not hold true: (a) analogs and their
phosphorylated forms may not mix and equilibrate with endogenous
substrates and products of CK (as would occur in a solution); (b) mixing occurs, but metabolites may be separated in
compartments with significantly different ATP/ADP or PCr/Cr ratios; or (c) ATP/ADP ratios between individual muscle cells may not be
uniform. Our results show that in steady state and during long-term
exposure to these analogs none of these possibilities is true, except
for an instructive case of feeding GPA for short (3 week)
duration, where heterogeneity of ATP/ADP ratios must be considered.
Measurements of Apparent Equilibrium
Constants
K for the CK reaction in the
presence of its endogenous substrates and a competitive inhibitor with
CK in a solution in vitro can be defined as the sum of two
separate reactions, each with its own uniquely defined equilibrium
constant but coupled by substrates common to both reactions. Our
approach was to measure this K in solution and
the corresponding mass action ratio in intact muscle as a test of
equilibration in vivo. K can be
expressed as the ratio of individual equilibrium constants for the
combined reaction. This derivation has the advantage of simplifying
chemical analyses, since substrates common to both reactions, ADP, ATP
(and Mg
), as well as pH, factor out of algebraic
expressions; its limitations and implications are discussed in detail
(see ``Discussion'') and illustrated mathematically (see
``Appendix''). in
solution and the mass action ratio in muscle should agree within
experimental error. The following equations specify the approach taken. 
as
follows.
) for the combined reaction will equal K. Unlike solution experiments in which there
is no net ATP utilization (ATPase flux) and ATP/ADP does not change,
equilibration may not be exact. In muscle, even at rest, there is a
small net ATPase activity. In this case, equilibration will only occur
if forward and reverse fluxes of CK are far in excess of steady state
fluxes of ATP through ATP synthesis and ATPase reactions.Measurements of Combined Equilibrium Constants in
Solution
K for CK in the presence of
creatine and either analog was measured using a combination of P NMR spectroscopy and HPLC methods. Solutions were
constructed to model cytoplasmic conditions at 23 °C and contained
in mM: 100 KCl, 5.3 ATP, 15 PCr, 100 MOPS, 4 EGTA, 1
KHPO
, 92
KCH
O, and 70 Tris at pH 7.0 as is
standard for skinned fiber experiments(42) . Aliquots of 3.5 ml
each were used for NMR experiments to which were added unphosphorylated
forms of either GPA or cCr prior to adjusting final volume and pH
of each solution. At the start of NMR experiments and before addition
of CK, a 50-µl aliquot of this mixture was assayed for ATP content
using an optical spectrophotometer at 259 nm. Absorbance due to analogs
was negligible at this wavelength. A control NMR spectrum was acquired
under fully relaxed conditions, then the phosphorylation reaction was
initiated by adding 2 mg/ml of rabbit CK to each tube (260 units of CK
activity/ml). Serial spectra were acquired during the time course to
equilibration and were halted when change in peak areas for PCr and
phospho-analog differed by less than 5% from previous acquisition. At
equilibration, samples were removed from the magnet, kept at constant
temperature, and CK-denatured by addition of 4% SDS/ml, chilled to 4
°C and centrifuged to remove precipitated SDS and
protein(43) . This method was used to avoid acid hydrolysis of
PCr which could significantly alter the result. Supernatant samples
were frozen at -70 °C for later analysis by HPLC.Materials
-Guanidopropionic acid was
synthesized from -alanine and cyanamide as described
previously(40) . Cyclocreatine was synthesized from
chloroacetate and ethylamine as reported by Griffiths and Walker (36) and recrystallized twice from hot water. Product purity
and verification of structure were assessed by proton NMR spectroscopy
and elemental analysis (Galbraith, Knoxville, TN). HPLC standards for
PcCr and GPAP were produced from equilibration solutions and
calibrated in the following manner. Quantification of PCr was
accomplished by integration of P NMR spectra of solutions
prior to addition of CK and after equilibration with analog (PCr and
PX). As the total integral of this region (PCr + PX) did not
change, the amount of PX equals initial less final PCr area. Resonances
were scaled to absolute chemical content by using ATP content measured
spectrophotometrically at 259 nm. These calibrated solutions provided
standards for HPLC quantification of PcCr and GPAP. Rabbit
creatine kinase, ATP, ADP, PCr, and Cr were obtained from Sigma.Analog Administration
Male Swiss Webster mice were
obtained at age 21 days and allowed to acclimate to the cage for a
period of 1 week prior to feeding. Analogs were administered by mixing
with standard rodent chow (GPA 2% w:w, cCr 1% w:w; BioServ,
Frenchtown, NJ). Animals were allowed food and water ad
libitum. Feeding experiments were performed for a period of 3
weeks at which time tissues from control, GPA, and cCr fed mice
were excised from animals under surgical anesthesia (ketamine and
xylazine). EDL (fast-twitch) and SOL (slow-twitch) muscles were allowed
to recover in a bath of physiologic saline from surgical perturbations
for 30 min, after which they were either removed from the bath,
blotted, and rapidly frozen at -80 °C with brass Wollenberger
tongs or placed in a custom-built P NMR probe for analysis
after which these samples were also frozen. In both cases, HPLC
analysis of neutralized perchloric acid extracts were performed.Nuclear Magnetic Resonance Spectroscopy
Phosphorus
NMR spectroscopy was performed on a 7T GN 300 (General Electric) using
either a 10-mm probe (for model solutions) or a custom-built phosphorus
probe for isolated mouse muscles(44) . For solutions, data were
acquired with a 
/2 pulse width (18 µs at 90 watts), 15-s
delay, and 4096 data points. Transformed data were the sum of 64
acquisitions which were apodized with a 3-Hz exponential filter prior
to the Fourier transform. For muscle samples, data were acquired with a
/2 pulse width (6.6 µs at 50 watts), 15 s delay, and 2048 data
points. Summed data were the result of either 300-400 data
transients apodized with a 15-Hz exponential filter prior to the
Fourier transform. The coil for this probe was a six-turn solenoid made
from 30-gauge Formvar-coated copper wire, driven by a balance matched
tank circuit, serially tuned to the phosphorus frequency. Muscle
samples were held in place within a 1.5-mm inner diameter capillary.
Magnetic field homogeneity was shimmed on the available proton signal
prior to the start of the experiment to a line width of less than 0.07
ppm (for solutions) and 0.1 ppm (for muscle)., 5% CO and contained (mM) 116 NaCl,
4.6 KCl, 26.2 MOPS (titrated to pH 7.4 with NaOH), 2.5
CaCl, 1.2 MgSO, and gentamycin (10 mg/l) at pH
7.4. Since these preparations gave reproducible P spectra
for up to 4 h, we concluded that they were in a metabolic steady state
during the 2-h duration of our experiment. Muscles were freeze-clamped
after completion of spectral acquisition. Contralateral muscles were
prepared similarly and frozen when its mate was put into the probe.
Muscles frozen about 30-45 min after dissection rather than after
3 h in the NMR probe showed no differences in metabolite contents. 
gww
.
ATP content by P NMR spectroscopy was the average of
, 
, and ATP resonances.HPLC Analysis
Chromatographic analysis was
performed on stable neutralized perchloric acid extracts as described
previously (46) . Tendons were removed from frozen muscle
before weighing. Total creatine content was the sum of Cr and PCr;
total analog content for each compound was calculated similarly.
In Vitro Combined Equilibrium Constants
P NMR spectroscopy was used to measure ATP, PCr,
and PX in each solution during approach to equilibrium which reached
completion within the first spectrum for cCr (less than 10 min), yet
took approximately 8 h for GPA (data not shown). Phosphorylation
rates based on time to equilibration after addition of CK were
substantially faster for cCr than GPA as expected from kinetic
constants for these compounds(47) . Fig. 1shows spectra
of representative solutions at equilibrium. Note that phosphorylated
analog species and PCr are resolved, and ADP was below limits of
detection. Summed areas of PCr and phospho-analog resonances in final
equilibrium solutions were the same as total PCr peak areas in initial
spectra prior to addition of CK, and there was no change in ATP
resonance intensities. This stoichiometry demonstrates that the
increase in phospho-analog equaled the decrease in PCr. Besides
verifying that equilibrium was reached, P NMR experiments
provided HPLC standards for phospho-analogs, which were not otherwise
available. At the conclusion of each NMR experiment reaction mixtures
were analyzed by HPLC for their components of free analog necessary for
proper calculation of K; HPLC analysis of
phosphorylated species provided a cross-calibration of methods. Both
HPLC and P NMR measurements permit separate calculation of
a PCr/ATP ratio. Correlation between PCr/ATP ratio for NMR versus PCr/ATP by HPLC were linear with unity slope and intercepts that
were not significantly different from zero at a level of p < 0.05. Regression equations were for solutions containing cCr (n = 7) and GPA (n = 8),
respectively.
P NMR
solution spectra for determination of combined equilibrium constants. A depicts a typical spectrum for GPA experiments.
Chemical shift values were 2.5, -2.54, -2.98, -5.0,
-10.0, and -18.5 for P, PCr, GPAP, and
, , and resonances of ATP, respectively
. B depicts solutions used for cCr. Note that chemical shifts are
identical with the exception of PcCr at -2.38 and absence of
GPAP. Solutions contained (in mM): 100 KCl, 5.3 ATP, 15
PCr, 100 MOPS, 4 EGTA, 1 KHPO, 92
KCHO, 70 Tris (pH 7.0), and 1 mg/ml
CK. Note chemical shift differences of PCr to GPAP (0.44) and PcCr
(0.16) and that there is no ADP detected in either solution, indicating
that the source of phosphate for each analog is from PCr
hydrolysis.
P
NMR chemical shift differences are small. with CK for both analogs are
given in Table 1. K values for both
analogs with CK agree with previous
reports(34, 40, 48) .
Analysis of Mouse Muscles
Representative P NMR SpectroscopyP NMR spectra of resting EDL muscles from both control and
analog-fed animals are shown in Fig. 2. There is an accumulation
of phosphorylated synthetic analog within 21 days of feeding as
indicated by the chemical shift change in the region of PCr
(-2.54 ppm). Chemical shift differences ( = 0.44 ppm)
between PCr and GPAP (-2.98 ppm relative to phosphoric acid)
were readily identifiable using 15 Hz apodization with near base-line
resolution achieved using only 2 Hz filtering (see inset of panel B).
In contrast, chemical shift differences ( = 0.16 ppm)
between PcCr (-2.38 ppm relative to phosphoric acid) and PCr were
much smaller with GPA and not resolved as well in vivo due to inherently broader line widths compared with solutions. The inset of Fig. 2C contains the same region
reprocessed with 2-Hz exponential filtering and shows PCr as a shoulder
upfield from the PcCr resonance.
P NMR spectra of control (A) and GPA (B)- and cCr (C)-treated
extensor digitorum longus muscles from mouse hindlimb. Data acquisition
parameters were as follows: /2 pulse width (6.6 µs), 15-s
predelay, 5-KHz sweep width, and 400 acquisitions. Summed data were
filtered with a 15-Hz exponential prior to the Fourier transform. Inset of each panel contains an expanded region surrounding
the PCr resonance which was processed using 2-Hz line broadening to
show chemical shift differences of PCr to GPAP (0.44, B)
and PCr and PcCr (0.16, C). Peak assignments are as follows: P
= inorganic phosphate, PCr = phosphocreatine, GPAP =
-guanidophosphonic acid phosphate, PcCr =
phosphocyclocreatine, and , , and reson
ances of
ATP.
, and ATP are
consistent with previous reports using Swiss-Webster (20) and
C-57 strains of mice (49, 50) from this laboratory.
Metabolite Quantitation by HPLC
Quantitative
measurements of all metabolites of CK reaction were performed using
HPLC methods because: (a) creatine and free analog, necessary
for calculation of K, are not detected by P NMR, nor are their resonances resolved by in vivo
H NMR spectroscopy; (b) ATP
content is needed to scale chemical content in vivo by NMR;
and (c) chemical shift differences in PCr and PcCr are small,
so HPLC estimates provide important cross comparisons to quantify PcCr
concentrations in muscle. Fig. 3presents levels of PCr, Cr, and
ATP in neutralized perchloric acid extracts determined by HPLC in
control (open bars) and GPA (closed bars)- and
cCr (left-hatched bars)-treated muscles. Total Cr values for
control SOL and EDL were 19.4 ± 1.04
µmolgww and 28.7 ± 1.71
µmolgww and decreased in both treatment
groups after 21 days of analog feeding to 28 and 22% of control values
for GPA SOL and EDL and to 37 and 42% of control for cCr SOL and
EDL, respectively. GPA-treated mice contained only 29 and 31% of
PCr in SOL and EDL, respectively. In cCr fed mice, PCr values were only
32 and 38% of control levels of PCr in SOL and EDL, respectively. The
contents of free Cr decreased in both EDL and SOL with both analogs to
approximately 50% of control levels. There was also a concurrent
decrease in ATP concentrations for each analog treatment, however
decreases were larger in GPA group (54 and 73% of control for SOL
and EDL) than in cCr group (75 and 95% of control for SOL and EDL).
GPA (closed
bars)- and cCr (hatched)-treated EDL and soleus. Data are
presented as mean + S.E. with units of
µmolesgww.
GPA (closed bars) and cCr (left-hatched bars) fed animals
are presented in Fig. 4. Accumulation of each phospho-analog
reached similar levels in both EDL and SOL during 21 days of feeding
while levels of free analog were 11-12-fold lower in cCr-treated
muscles than those from mice-administered GPA.
GPA (closed bars)- and cCr (hatched)-treated EDL and
soleus. PcCr and GPAP were determined by anion exchange HPLC (left panel), whereas free cCr and GPA
were determined
using a cation exchange column (right panel). In addition to
the measured value for free analog, B shows predicted values (cross-hatched) for each analog calculated from solution
equilibrium constants and therefore contain no error bars. Data are
presented as mean + S.E. with units of
µmolesgww.
Cross-calibrations
Integration of areas from P NMR spectra yields in vivo values for
metabolite ratios which can be compared with HPLC data obtained from
extracts of the same or contralateral muscle from the same animal under
stable resting conditions. Therefore, we have independent validation by
NMR and HPLC of relative metabolite content with respect to phosphorus
containing metabolites. As with solution experiments, the test of
quantitative agreement between the two techniques was performed by
measuring PCr/ATP in addition to PCr/PX in NMR spectra and HPLC.
Results of these comparisons are presented in Fig. 5. Note that
compounds had no difference in mean values for either method used. That
NMR ratios (closed bars) are indistinguishable from those
determined by HPLC (open bars) demonstrates that
phosphorus-containing substrates for CK are 100% NMR visible, both
under normal resting states and under conditions of partial creatine
depletion. Quantitative agreement in stoichiometry with PCr, PX, and
ATP by both methods in control and treated muscles (where P is lowered by creatine depletion and incorporated into PX)
suggests that, although P was only measured by the NMR
method, all P is NMR-visible in these muscles.
GPAP, and
PCr/PcCr (presented as PCr/PX). ATP values for NMR were calculated as
mean of + + ATP resonances. Value
s are
presented as mean + S.E.
In Vivo Tests of Equilibration
Fig. 6contains results of calculations of K for CK using HPLC analysis of neutralized
perchloric acid extracts. The upper panel contains data from
cCr-treated mice at 3 weeks of feeding and compares mass action ratios
for EDL (cross-hatched), SOL (right-hatched) with K values determined from solution experiments (left-hatched) which were 33.9 ± 2.15, 33.6 ±
1.67, and 34.3 ± 3.26, respectively. The lower panel contains data for GPA-treated mice presented in the same
order. Values of for EDL and SOL and K in
solution are 2.05 ± 0.16, 1.58 ± 0.11, and 3.06 ±
0.31, respectively. Agreement of for EDL and SOL in neutralized
perchloric acid extracts with solution K in the
instance of cCr-treated muscles shows that these substrates equilibrate
in the cytoplasm. This does not rule out existence of smaller local
``pools'' of metabolites but contributions of these pools to
total metabolite content would have to be large to influence this
result.
GPA/Cr via CK reaction compared for in vitro (solution) and in vivo. Values are presented as mean
+ S.E. (n = 6-8 for each group). The upper panel represents data from cCr experiments, whereas the lower panel contains data from GPA experiments: left-
hatched bars = solutions, right-hatched bars = soleus, and cross-hatched bars = EDL. The lower panel contains data from experiments performed at 3
weeks as well as on a longer time course (9-12 week) GPA
feeding (solid bars) to test for transient changes in tissue
. Asterisks were placed over 3-week data to indicate that
these values deviate significantly from solution K.
GPA-treated muscles suggests there is a
lack of equilibration between PCr/Cr and GPAP/GPA, since
for both tissues is considerably lower than solution K values. Because P NMR spectra
and HPLC results agree in both analog treatments, the apparent lack of
equilibration can only be explained by either low free Cr or high
GPA values (see ). Fig. 4shows depletion of
free Cr occurs to approximately the same extent in EDL and SOL of both
analog-treated groups; therefore lack of equilibration may result from
the presence of too much GPA. This finding leads to the hypothesis
that there may be some nonspecific binding of GPA in the intact
cell which is released during perchloric acid extraction and
contributes to HPLC measurement.GPA to subcellular fractions we performed
an equilibrium binding experiment using differential centrifugation
with a whole muscle homogenate. Rabbit tibialis anterior muscle was
homogenized (1:10 w/v) in 125 mM Tris buffer (pH 7.0). After
an initial incubation period of 30 min, the homogenate was divided into
4 equal volumes. Either GPA or cCr (approximately 40-50
mM) was added to a pair of samples by diluting 1:1 with stock
solutions of each analog (also in 125 mM Tris), then incubated
for an additional period of 30 min at 25 °C. Differential
centrifugations were performed at 600, 5,000, and 13,000 
g, and supernatants and pellets of each fraction were
subjected to perchloric acid extraction. The test for binding of analog
as compared with Cr was to measure the ratio of synthetic analog to
endogenous Cr in each fraction. Neutralized extracts were assayed by
cation exchange HPLC for free GPA, cCr, and Cr. If some GPA
binding were sufficient to account for the observed lack of
equilibration, then there should have been excess GPA in at least
one of the fractions, and a different ratio of GPA/Cr should have
been detected. Cyclocreatine was used as a negative control for this
experiment. The results presented in Table 3were decisive: the
ratio of free Cr to analog was constant in all fractions tested. On
this basis we exclude an unsuspected selective binding of GPA as
an explanation for apparent disequilibration of GPA experiments in
animals fed the diet for 3 weeks.
GPA over longer time
periods did not exhibit this apparent lack of
equilibration(40) , suggesting that the present results showing
excess free GPA might be transient, and if measured later in the
time course, would not be present. To test this hypothesis, we fed mice
over equivalent time periods as in rat studies (9-12 weeks) and
performed the same HPLC measurements as in the present study. The
results presented in Fig. 6show a marked deviation of mass
action ratios from K in muscles from GPA
animals with several weeks of feeding but with typical feeding regimens
used, 9-12 weeks, mass action ratios are statistically
indistinguishable from K, indicating that
GPA effects on are transient.
P NMR spectroscopy, we
employed two competitive inhibitors of CK, cCr, and GPA to
investigate these issues. Feeding synthetic analogs of creatine (which
as competitive inhibitors are not expected to disturb equilibration)
partially displaced the content of PCr and Cr from the cell. By
choosing the duration of feeding such that there were approximately
equal concentrations of total Cr and total analog, it was possible to
have an accurate measure of mass action ratios for CK in resting muscle
and test the hypothesis that these ratios equal those predicted from K measured in solution as would be the case if
the reaction is equilibrating and the cytosol is freely mixing with
respect to substrates. in muscle was
indistinguishable from those predicted by K in
solution. Therefore results with this analog could not disprove the
central hypothesis of this work. In contrast, in GPA experiments
(3 weeks of analog feeding), in muscle was clearly different from
those predicted by K in solution. However, with
longer periods of feeding in muscle were indistinguishable from
that predicted by K in solution, as has been
shown in other preparations(40) . Thus our results show that
muscles in the presence of GPA undergo a significant bioenergetic
transformation from the onset of analog feeding to the time they
demonstrate large phenotype changes in contractile and metabolic
proteins(51, 52) . The primary observation of
near-equilibration of CK with its natural substrates and exogenous
competitive inhibitors is consistent with the idea that cytoplasmic
phosphorylation potential displays thermodynamic characteristics of
substrates and enzymes freely mixing. Thus, these data establish a
solid biochemical basis for the widely used concept of simple solution
dynamics for interpretation of P NMR data. measured in solution. This conclusion
is derived from the principle that PCr/Cr (and also PX/X) ratios are
set by cell ATP/ADP ratios (which are determined by cytosolic chemical
potential) acting through CK thereby rendering calculation of
independent of total Cr or analog
content(1, 9, 10) . By extension, this
conclusion can be related to calculation of free ADP in whole tissue
where metabolite heterogeneity is present either as a characteristic of
phenotype (20) or induced via differential cellular responses
to metabolic stress (e.g. fiber recruitment during exercise)
or abnormal physiologic states (e.g. localized ischemia).Quantitative Analysis
Analysis of solution NMR
experiments test validity of using both HPLC and NMR methods to assay
mass action ratios in muscle. Absolute concentrations for PCr and PX
were obtained from NMR by ratio to the mean of ATP resonances, which
was measured by optical spectroscopy, yielding mM concentrations of PCr and PX in each solution. For HPLC, both ATP
and PCr were determined from enzymatically calibrated standards.
PCr/ATP, obtainable from each method, was used as a method of
cross-calibration. In solution these ratios of chemical contents are
very well correlated for both analogs, since a plot of PCr/ATP has a
slope of unity and a zero intercept. Our conclusion from these
experiments is that both assays are fully internally consistent and
proper concentrations can be derived from both methods for phosphorus
metabolites in solution.
P NMR and anion exchange HPLC. Quantitative agreement with
NMR and chemical methods illustrates that artifacts of freeze clamping
and extraction are absent using our technique irrespective of total Cr
content. Calculation of might also be influenced by NMR
visibility of PCr and PX. NMR visibility of ATP and P have
been examined in other tissues such as heart and liver, particularly
under different metabolic states (53, 54, 55, 56) . Our analysis
using metabolite ratios allowed unambiguous investigation of this issue
in muscle. To maintain agreement with stoichiometry of the chemical
results, any NMR ``invisible'' portion of these metabolites
would require an equivalent fraction of the other metabolites to be
``invisible'' as well. We illustrate this important point as
follows: consider PCr/ATP in control EDL (from Fig. 3) with
approximate metabolite contents of 20 µmolegww PCr and 5 µmol
gww ATP. If a
``pool'' of ATP or PCr corresponding to 20% of total cellular
content was not NMR visible by some mechanism (e.g. rapid T relaxation), this possibility is equivalent to 4
µmol of PCr being invisible but only 1 µmol of ATP. This
``NMR-invisible'' pool would have to be nonstoichiometric (4
PCr/1 ATP) to be consistent with observed ratios and these
``invisible fractions'' would have to be extracted and
quantified to the same extent for HPLC and NMR ratios to agree.
Additionally, in muscles from treated animals, PCr/PX ratios for both
methods also agree. Therefore, any NMR-invisible fraction which exists
must influence invisible metabolites in this pool via an identical
nonstoichiometric mechanism. We consider this scenario which so affects
equal fractions of PCr, PX, and ATP untenable and favor the simpler
explanation that phosphorus metabolites involved in the CK reaction,
namely PCr, ATP (and in analog-treated muscles PX) must be fully
NMR-visible.
The results with
cCr-treated muscles and quantitative agreement of HPLC and NMR lead to
the conclusion that there is thermodynamic control of PCr/Cr and
PcCr/cCr through CK in the cytoplasm. These results argue that the
apparent lack of equilibration in GPA EquilibrationGPA-treated muscles at the
3-week time point cannot be explained by artifacts of the methodology.
Furthermore, since PCr and PX values from HPLC are the same as those
measured in vivo, lack of equilibration observed in 3-week
GPA feeding must be explained in terms of either free Cr or free
GPA content. One possible explanation for apparent lack of
equilibration of GPA with CK might be that a fraction of free
GPA, which is present in muscle but bound to an unknown location
thus unavailable for reaction, is subsequently released during
perchloric acid extraction. This notion is similar to binding of ADP to
actin. The key tenet of this hypothesis is that CK actually
equilibrates with both GPA and GPAP and any bound and unbound
pools have equilibrated as well, but extraction of the muscles with
perchloric acid extraction releases bound metabolite which is then
assayed by HPLC. Predicted values for unphosphorylated analog are
presented in Fig. 4for both analogs (cross-hatched
bars) as well as HPLC measured values for GPA (closed
bars) and cCr (right-hatched bars). The difference
between predicted value and that measured depicts the amount that must
be bound (by hypothesis) and subsequently released during extraction to
account for disequilibration in GPA treated muscles. This putative
bound fraction, calculated from the combined equilibrium constants for
each analog (assuming the dependent variable is the unphosphorylated
analog) is 50-100% less than what was actually measured in
perchloric acid extracts of EDL and SOL by HPLC. Thus, the prediction
from equilibrium binding experiments is that if GPA had some
binding affinity for a subcellular component, a different ratio of Cr/X
should have been detected in at least one of the fractions tested.
Results of equilibrium binding experiments (see Table 3)
demonstrate that Cr/X is constant regardless of the fraction sampled.
The 600 g pellet of intact cells and tissue fragments
contained only 10% of the Cr and analog. The remainder was found in the
supernatant of each differential centrifugation, demonstrating that
these compounds behave similarly and are primarily found in solution.
Finally, if nonspecific binding were occurring, animals subjected to
longer feeding time courses (possessing higher total analog
concentrations) would also exhibit an apparent disequilibration with
GPA, perhaps to a greater extent. But GPA fed animals on
longer feeding regimes have tissue mass action ratios which are not
distinguishable from solution K values at 3
weeks. One would expect the opposite result if binding were a plausible
explanation. We therefore exclude nonselective binding of GPA as
the explanation for the apparent disequilibration after 3 weeks of
feeding.P NMR spectroscopy
performed on intact hindlimb of GPA-treated rats have shown that
GPAP levels decrease 20% (approximately 5 mM) during
acute exercise in rat hindlimb muscles and return to initial levels
within 13 min(40) . This translates into a phosphorylation rate
of 0.37 mM/min for rat fast-twitch muscles. During the course
of our NMR experiments (3 h), approximately 60 mM of GPAP
could have been produced. This flux is far in excess of what is
required to meet equilibrium conditions assuming rat and mouse
fast-twitch muscles have similar CK activities. Thus a kinetic
limitation under resting conditions cannot explain the observations for
GPA-treated muscles.Heterogeneity of Metabolite Distribution
In
solution experiments, equilibration of CK with a uniform ATP and ADP
concentration is certain even in the presence of competitive inhibitors
(see and ). In intact tissues this may not be
completely valid because of heterogeneity. There are two possibilities
where PCr/Cr and PX/X might equilibrate with different ATP/ADP ratios
in intact muscle: (a) either ATP/ADP in some subcellular
location of some or all cells is different than bulk cytoplasm or (b) a subpopulation of cells within the tissue might be at a
different cytosolic ATP/ADP ratio than the rest. These alternatives
reflect intracellularversusintercellular compartments. Creatine kinase localization within the cell,
including mitochondrial isoforms, have been reported (3) , but
their existence does not de facto demonstrate that associated
substrates are segregated from the cytosol as well. Heterogeneity of
chemical contents within muscle fiber types has been reported
previously(20) , raising the possibility that apparent lack of
equilibration with GPA is due to intercellular heterogeneity of ATP/ADP ratios, perhaps induced by exposure to
analog itself.GPA at 3 weeks? From our
analysis based upon an expansion of from the
``Experimental Procedures'' (see in the
``Appendix''), lack of agreement for solution K values with from muscles of
GPA-treated animals cannot be explained by varying distributions
of Cr or its analogs. In fact, observed disequilibration can only be
explained if ATP/ADP ratios are no longer constrained to be equal (e.g.fPCr
fPCr, Appendix, and ) between volume fractions. Although adenylates do not
directly enter into any of these calculations, under equilibrium
conditions, fractional PCr content is determined by ATP/ADP through CK (9, 10) and therefore ultimately sets PX/X ratio as
well. Thus, if ATP/ADP, or equivalently fPCr, in 1 volume
fraction is significantly different from the other then
no longer equals K
. This is possible
while physicochemical constraints for thermodynamic equilibrium are
still met in each volume fraction comprising the sample. This analysis
demonstrates that apparent lack of equilibration observed is a
physiologic phenomenon resulting from different fractional PCr contents
(hence different ATP/ADP ratios) in each volume fraction. Thus
measurements from whole muscle may appear to violate CK equilibration, i.e. does not equal K, despite the fact that each volume fraction
is in equilibrium as defined by and in the
``Appendix.'' We favor the hypothesis that GPA is
selectively effecting a set of muscle fibers in a cell-specific manner,
and the nature of this perturbation is related to metabolism of Cr
ultimately altering cellular energetics. This fiber-specific effect may
in part explain adaptive changes in Cr analog administration in murine
muscle(51, 52, 57) .GPA in other
preparations exposed to long duration feeding (confirmed here in mouse
hindlimb muscles) are highly significant for describing and
understanding functional organization of CK in muscle cells. That we
observed muscle to be indistinguishable from solution K indicates that these major bioenergetic
reactions function in cells of resting muscles as they do in
homogeneous solutions in vitro, despite the presence of
mitochondrial and cytoplasmic isoforms, their partial binding to
macromolecular structures, and compartmentalization of the cell volume
(sarcoplasmic reticulum, mitochondria, etc.). Corollaries of this
conclusion, supported by other studies of CK function, are several: (a) calculations of metabolically active ADP concentration
from CK equilibration are valid; (b) concentrations of PCr,
phospho-analog, and ATP measured by P NMR are equal to
those measured in perchloric acid extracts, thus there is no evidence
for NMR-invisible pools of these metabolites; (c) the net
quantitative contribution of metabolites in diffusion-limited
compartments to total measured quantities is small; (d)
effects of such compartmentalization, however important for certain
functions, are negligible with respect to total cellular bioenergetics
and metabolism. Thus our experiments unambiguously demonstrate that
thermodynamic characteristics of the cytosol can be predicted as if
these metabolites were freely mixing in solution. Finally, if there is
an apparent lack of equilibration due to two compartments in the muscle
with differing ATP/ADP ratios, bioenergetics can still be solved by
knowledge of either ATP/ADP (or fPCr) in one of the two
components.
.
From expansion of in ``Experimental
Procedures,'' consider 2 volumes which comprise the entire sample.
This expression can be expanded to include mass action ratios for more
than 2 fractional volumes. Contribution of each volume fraction to the
total mass action ratio can be written as
follows,
is the mass action ratio for the total
sample volume (V), is the first fractional
volume, and is the second fractional volume and these are
constrained to be + = V
.
Mass action ratios for each fractional volume (
and 
) can be written as
follows.
Case 1: Distribution of Total Analog
From the total
analog content for each fractional volume (TX), content of PX and X can
be expressed in terms of PCr and Cr content as
follows.
= 
=
, we see that for any analog concentration (TX),
the PX/X ratio is now entirely defined by the only independent
variables in and , namely PCr and Cr. Under
these circumstances the PX/X ratio reflects the concentrations of PCr
and Cr in each volume fraction independent of total analog content.
More importantly, one can also conclude that as long as concentrations
of PCr and Cr are equal in both fractional volumes then K = . These conclusions
are independent of total analog present or its distribution between
volume fractions.
Case 2: Distribution of Total Creatine
Independent
variables from and can be expressed as
fractional PCr (fPCr) content with respect to total Cr present
as
follows.
= , inde
pendent of
total Cr content of either volume fraction or distributions. Thus K remains equal to 
.Case 3: Distribution of Adenylates
If adenylates
are in equilibrium PCr and Cr within each volume fraction, then we
write,
and K are equilibrium constants for CK for each fractional volume
. Thus
only if ATP/ADP ratios for both frac-tional volumes are equal will
PCr/Cr and PX/X ratios agree and thus K = .
)GPA, -guanidopropionate; X, creatine analog; PX,
phosphorylated analog; HPLC, high performance liquid chromatography;
MOPS, 4-morpholinepropanesulfonic acid; EDL, extensor digitorum longus;
SOL, soleus.)
We gratefully acknowledge the technical assistance of
Rudolph Stuppard for HPLC analysis of tissue extracts and Dr. Thomas W.
Beck for assistance in solution NMR experiments and HPLC. Many thanks
to Drs. P. Bryant Chase, W. Ross Ellington, and Jeroen Jeneson for
critical reading of the manuscript and providing thoughtful comments.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  R. S. O'Connor, C. M. Steeds, R. W. Wiseman, and G. K. Pavlath Phosphocreatine as an energy source for actin cytoskeletal rearrangements during myoblast fusion J. Physiol., June 15, 2008; 586(12): 2841 - 2853. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. J. Brault, K. A. Abraham, and R. L. Terjung Phosphocreatine content of freeze-clamped muscle: influence of creatine kinase inhibition J Appl Physiol, May 1, 2003; 94(5): 1751 - 1756. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H J A in't Zandt, A J C de Groof, W K J Renema, F T J J Oerlemans, D W J Klomp, B Wieringa, and A Heerschap Presence of (phospho)creatine in developing and adult skeletal muscle of mice without mitochondrial and cytosolic muscle creatine kinase isoforms J. Physiol., May 1, 2003; 548(3): 847 - 858. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Mark. E. Trump, C. C. Hanstock, P. S. Allen, D. Gheorghiu, and P. W. Hochachka An 1H-MRS evaluation of the phosphocreatine/creatine pool (tCr) in human muscle Am J Physiol Regulatory Integrative Comp Physiol, March 1, 2001; 280(3): R889 - R896. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Wyss and R. Kaddurah-Daouk Creatine and Creatinine Metabolism Physiol Rev, July 1, 2000; 80(3): 1107 - 1213. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. F. Watchko, M. J. Daood, B. Wieringa, and A. P. Koretsky Myofibrillar or mitochondrial creatine kinase deficiency alone does not impair mouse diaphragm isotonic function J Appl Physiol, March 1, 2000; 88(3): 973 - 980. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Maril, H. Degani, E. Rushkin, A. D. Sherry, and M. Cohn Kinetics of cyclocreatine and Na+ cotransport in human breast cancer cells: mechanism of activity Am J Physiol Cell Physiol, October 1, 1999; 277(4): C708 - C716. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. J. Labella, M. J. Daood, A. P. Koretsky, B. B. Roman, G. C. Sieck, B. Wieringa, and J. F. Watchko Absence of myofibrillar creatine kinase and diaphragm isometric function during repetitive activation J Appl Physiol, April 1, 1998; 84(4): 1166 - 1173. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. W. Hochachka and M. K. P. Mossey Does muscle creatine phosphokinase have access to the total pool of phosphocreatine plus creatine? Am J Physiol Regulatory Integrative Comp Physiol, March 1, 1998; 274(3): R868 - R872. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| All ASBMB Journals | Molecular and Cellular Proteomics |
| Journal of Lipid Research | ASBMB Today |