| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 278, Issue 33, 30441-30449, August 15, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
¶
From the
Department of Cell Biology, University
Medical Center, University of Nijmegen, The Netherlands and
Division of Cardiovascular Diseases, Departments
of Medicine, Molecular Pharmacology and Experimental Therapeutics, Mayo
Clinic, Rochester, Minnesota 55905
Received for publication, March 27, 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Adenylate kinases (AK, EC 2.7.4.3
[EC]
), an evolutionary conserved family of
enzymes that catalyzes the reaction ATP + AMP 
2 ADP
(7), have been implicated in
cellular adenine nucleotide homeostasis
(8). cDNAs for five isoforms of
AK (AK1–AK5) along with the variant of AK1 (AK1
, a membrane-bound
form with a presumed role in cell cycle regulation) have been cloned from
metabolically active tissues
(9–12).
Mammalian skeletal muscle is particularly rich in AK1, the major isoform of
the family (9), present in the
sarcoplasm, and clustered along the myofibrillar I-band or bound as AK1
to membranes
(13–15).
By donating the energy of the -phosphoryl group of ATP/ADP to the
cellular energetic pool, AK isoenzymes protect cells against energy
deprivation in periods of high metabolic demand
(6,
16–20).
The different intracellular localizations and distinct kinetic properties of
AK isoforms permit the formation of a coordinated enzymatic network for
nucleotide-mediated metabolic signaling, coupling myofibrillar, nuclear, or
sarcolemmal energy-dependent processes with mitochondrial energetics
(21–23).
Creatine kinases (CK, EC 2.7.3.2
[EC]
) catalyzing the reaction
MgADP– + CrP2– + H+
Cr + MgATP2– belong to a smaller and
evolutionary younger family of enzymes with a role in high energy phosphoryl
transfer and cellular energy buffering
(1,
5,
24). Creatine kinases are
foremost found in cells with high peak demands in metabolic energy such as the
brain, heart, or skeletal muscle
(1,
5). In skeletal muscle, the
principal CK isoform is the cytosolic isoform, MM-CK, a homodimer mainly
present as a soluble protein in the cytosol and bound to the myofibrillar M-
and I-bands (13) as well as to
the sarcoplasmic reticulum membranes
(25). Skeletal muscle also
contains an additional mitochondrial CK isoform (ScCKmit), which amounts to
1–10% of the total CK activity depending on the type of muscle fiber
(26,
27). This CK member associates
and functionally interacts with the adenine nucleotide translocator and
voltage-dependent anion channel in the mitochondrial inner and outer membrane
(28–30),
providing an efficient ATP export and metabolic signal reception pathway
(31).
AK and CK in concert with nucleoside diphosphokinase (NDPK) and the enzymes
that function in the glycolytic phosphotransfer pathway form the cellular
energetic infrastructure responsible for effective handling and distribution
of high energy phosphoryl (
P) groups throughout the structured muscle
environment (5,
6,
24,
32–34).
In this network, AK- and CK-mediated reactions play a complementary and
functionally alternate role
(18,
27,
35,
36). By pharmacological
inhibition of the CK circuit, it has been demonstrated that an increase in
AK-mediated phosphotransfer may compensate for the loss of CK-activity
(17). Likewise, in skeletal
muscles carrying a null mutation in either the M-CK or AK1
gene, leading to complete lack of corresponding protein expression and
activity, an adaptive rewiring of flux through the remaining intact
phosphotransfer circuit occurs
(6,
18,
19). In addition,
M-CK and AK1 mutant muscles respond with similar but not
identical ultrastructural and molecular adaptations, suggesting an inherent
plasticity of the bioenergetic network
(6,
24,
37–39).
Although progress has been made in our understanding of individual phosphotransfer reactions, the consequences of combined deletion of major AK and CK isoforms remain unknown. Here, we report on the effects deleting both the AK1 and M-CK proteins in a single cell-type skeletal muscle fiber of a double knock-out mouse. Use of this model provides us with a unique opportunity to assess the significance of the activities of mitochondrial CK, glycolytic enzymes, and NDPK-mediated phosphotransfer reactions in muscle physiology and metabolism. Monitoring intracellular phosphotransfer kinetics by [18O]phosphoryl labeling revealed that lack of AK1 and M-CK resulted in a serious impairment of communication between ATP-generating and ATP-consuming cellular sites. As a consequence, AK1/M-CK-deficient muscles had a reduced ability to sustain the dynamic fluctuations in ATP/ADP/AMP nucleotide metabolism and overall cellular ATP turnover during functional load despite increased high energy phosphoryl flux through alternative glycolytic and guanylate phosphotransfer pathways. These data provide further evidence for the existence of a fully integrated high energy phosphoryl transfer system with a high degree of functional plasticity required for optimal muscle energetics.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
High energy Phosphoryl Transfer—ATP turnover and phosphoryl
flux through AK, CK, and glycolytic systems were measured in intact
gastrocnemius-plantaris-soleus (GPS) muscle complex using the
[18O]phosphoryl labeling technique
(16,
17,
21,
41). Mice were anesthetized
with pentobarbital (Beuthanasia D) (100 mg/kg intraperitoneal) and injected
with heparin (50 units of intraperitoneal) prior to muscle dissection.
Isolated mouse GPS muscle was preincubated for 12 min at room temperature in a
medium containing (in mM) 137 NaCl, 5 KCl, 2 CaCl2, 1
MgCl2, 1 NaH2PO4, 20 HEPES, 0.05 EDTA, 5
glucose, and 24 NaHCO3 (pH 7.4). The muscle complex was then
rapidly transferred into a medium enriched with 30% 18O-water
(Isotec Inc.) and paced at 2 Hz exactly as described previously
(6). After 3 min of
18O labeling, muscles were freeze-clamped using liquid
N2 and extracted in a solution containing 0.6 M
HClO4 and 1 mM EDTA. Protein content was determined
using a protein assay kit (Bio-Rad). Cellular ATP, ADP, GTP, GDP, inorganic
phosphate, creatine phosphate (CrP), and glucose 6-phosphate (Glc-6-P) were
purified and quantified using high performance liquid chromatography
(17,
41). To obtain information on
basal levels and turnover rates of phosphoryl-containing metabolites, isolated
mouse GPS muscle was treated identically but without pacing. Samples
containing phosphoryls of 
-ATP, -ATP, -ADP, -GTP,
-GTP, -GDP, inorganic phosphate, and CrP as glycerol 3-phosphate
were converted to trimethylsilyl derivatives. 18O enrichment of
phosphoryls in glycerol 3-phosphates was determined with a Hewlett-Packard
5980B gas chromatograph mass spectrometer
(6).
High Energy Phosphoryl Transfer Rates—Total cellular ATP
turnover was estimated from the total number of 18O atoms that
appeared in phosphoryls of Pi, CrP, -ATP, -ATP,
-ADP, -GTP, -GTP/GDP, and Glc-6-P. Net AK-, CK-, and
hexokinase-catalyzed phosphotransfers were determined from the rate of
appearance of 18O-containing phosphoryls in -ATP and ADP and
CrP and Glc-6-P, respectively
(6,
17,
41,
42).
Metabolite Levels—ATP, ADP, GTP, and GDP levels were quantified in muscle perchloric extracts by use of high pressure liquid chromatography (21, 41). AMP, ADP, CrP, muscle lactate, and glucose 6-phosphate levels were determined using coupled enzyme assays (21, 24). Muscle inorganic phosphate level was determined using the EnzChek phosphate assay kit (Molecular Probes).
Zymogram Analysis and Enzyme Activity—Homogenates from freshly excised GPS muscles (10% w/v) were prepared in SETH buffer (in mM: 250 sucrose, 2 EDTA, 10 Tris-HCl (pH 7.4)) at 4 °C. GPS extracts were diluted 1:1 in 30 mM Na3PO4 buffer (pH 7.4) containing 0.05% v/v Triton X-100, 0.3 mM dithiothreitol, and a complete protease inhibitor mixture (Roche Applied Science). Muscle extracts were incubated for 30 min at room temperature and centrifuged for 20 min at 11,000 x g, and an aliquot (1–5 µl) was applied to agarose gels (Alameda, CA). AK1 and creatine kinase isoenzymes were separated electrophoretically and stained for enzyme activity (6, 43). CK and AK activities were measured from whole hind limb muscle extracts (6) using a CK NAC (N-acetyl-L-cysteine)-activated kit (Roche Applied Science) and a coupled enzyme system, respectively (21).
Western Blot Analysis—Skeletal muscles were excised, pulverized with a mortar and pestle using liquid N2, and extracted in a buffer containing 50 mM NaCl, 60 mM Tris-HCl (pH 7.5), 5 mM EDTA, and 0.2% Triton X-100 to which a complete protease inhibitor mixture was added. Extracts were centrifuged (10 min, 8000 x g, 4 °C), and proteins were separated on 10% SDS-polyacrylamide gels before being electrophoretically transferred onto nitrocellulose membranes. M-CK and AK1 proteins were detected using antibodies raised against chicken M-CK (44) and mouse AK1-glutathione S-transferase fusion proteins (15). Aldolase antibody (Rockland, Gilbertsville, PA) was used as a control. Immunocomplexes were visualized by chemiluminescence using a secondary antibody coupled to horseradish peroxidase and exposure to Kodak X-Omat AR films.
Actomyosin Contraction—Actomyosin contraction was measured using an established superprecipitation method (45, 46). Excised femoral quadriceps muscles (muscles from two mice combined) were pulverized with mortar and pestle using liquid N2 and extracted in a buffer containing 0.6 M KCl, 0.04 M NaHCO3, 0.01 M Na2CO3, 4 mM dithiothreitol, and a complete protease inhibitor mixture. Extracts were homogenized (3 x 5 s) with a blender (Polytron), maintained at 4 °C for 10–15 min, and centrifuged (30 min, 9500 x g, 4 °C). The supernatant was gently diluted with 10 volumes of ice-cold water containing 0.5 mM dithiothreitol and maintained at 4 °C for 30 min. The solution was centrifuged (15 min, 7000 x g,4 °C), and the precipitate dissolved in a buffer containing 0.6 M KCl, 20 mM Tris-HCl (pH 7.4), and the complete protease inhibitor mixture. Skeletal muscle actomyosin was stored at –20 °C after mixing with an equal volume of glycerol. Superprecipitation was measured in a buffer containing 0.8 mg/ml skeletal muscle actomyosin, 0.8 mM CaCl2, 2 mM MgCl2, 50 mM KCl, 1 mM EGTA, and 20 mM Tris-HCl (pH 6.8) at 25 °C. The change in absorbance at 660 nm was followed after the addition of ADP (0.2 or 2 mM), phosphoenolpyruvate (PEP; 2 mM), CrP (2.0 mM), or ATP (2 mM). Three measurements for each genotype (muscles from 2 mice/sample combined) were performed and statistically compared.
Statistics—Data are presented as mean ± S.E. Student's t test for unpaired samples was used for statistical analysis, and p < 0.05 was considered significant.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
Cellular Energetics in Resting and Contracting MAK=/=
Muscle—Genetic deletion of M-CK and AK1 produced a dramatic
reduction in the amount of 18O labeling of the pools of CrP,
-ATP, and -ADP in skeletal muscle. Compared with wild type,
18O-metabolic labeling of CrP was decreased from 8.8 ± 0.4
to 5.4 ± 0.2% in MAK=/= resting non-contracting muscle, a
reduction of 39% (p < 0.05, n = 6)
(Fig. 2A). Stimulation
of contractile activity further aggravated the difference in metabolic
labeling of CrP between wild-type and mutant muscles, changing from 17.2
± 1.2% in wild-type to 8.0 ± 0.5% in knock-out muscle
(p < 0.05, n = 6), a reduction of 54%
(Fig. 2A). On average,
muscle stimulation at 2 Hz increased 18O labeling of CrP by 95% in
wild type but gave only a 49% elevation in MAK=/= mutant muscle.
This increase in [18O]CrP labeling despite the absence of M-CK
could be attributed to metabolic flux through the mitochondrial ScCKmit
reaction, which is still functionally active in MAK=/= mutant
muscles. Indeed, MAK=/= muscles had a residual CK activity of 1.5
± 0.1 µmol CrP·min–1·mg
protein–1 (n = 3) due to the presence of
ScCKmit, which is at 8% of total CK activity in wild-type controls.
AK-catalyzed phosphotransfer activity assessed by 18O incorporation
in the -phosphoryl group of ATP was 1.8 ± 0.2% in non-contracting
wild-type muscle and increased to 6.3 ± 0.5% in contracting muscle
(Fig. 2B). However,
AK-catalyzed -phosphoryl labeling was dramatically reduced in
MAK=/= muscles when compared with wild type (by 68 and 85% in
resting and contractile state, respectively) with values of 0.6 ± 0.06%
in non-contracting muscle and 0.9 ± 0.1% in 2-Hz paced muscle
(p < 0.05, n = 6)
(Fig. 2B). These
results indicate that both CK- and AK-mediated phosphoryl exchange rates
increase with metabolic demand, but absolute and relative capacities for
up-regulation are severely impaired when M-CK and AK1 are both lacking.
|
Moreover, a prominent difference in total ATP levels was seen in
non-contracting muscles with 27.5 ± 0.9 nmol ATP·mg
protein–1 in wild type and only 21.9 ± 1.2
nmol ATP·mg protein–1 in MAK=/=
muscle (p < 0.05, n = 6 each). Conspicuously, in
contracting double knock-out MAK=/= muscles, total ATP levels were
well maintained compared with wild type (26.8 ± 0.3 nmol ATP·mg
protein–1 versus 25.2 ± 1.6 nmol
ATP·mg protein–1, respectively; p
> 0.05, n = 6 each). These results suggest that coupling between
metabolic energy salvage systems in MAK=/= muscles may be less
tight under resting conditions but becomes partly restored under conditions of
increased energy demand. It is conceivable that different thresholds in
ATP/ADP/AMP ratios are necessary to switch-on alternative pathways for energy
homeostasis. Indeed, the ATP/ADP ratio, a kinetic index of P utilization
and replenishment, was decreased from 8.4 ± 0.2 to 5.9 ± 0.1 in
non-contracting muscles and from 6.5 ± 0.2 to 5.6 ± 0.1 in
contracting MAK=/= muscles when compared with wild-type controls,
respectively (p < 0.05, n = 6). Moreover, the ADP/AMP
ratio was significantly higher in non-contracting MAK=/= muscle and
MAK=/= muscle under workload when compared with wild type: 9.4
± 0.4 versus 8.0 ± 0.3 (p < 0.05,
n = 6) and 8.3 ± 0.3 (n = 6) versus 6.7
± 0.4 (n = 5, p < 0.05), respectively. The ATP/AMP
ratio in resting muscle was lower in MAK=/= muscle than in wild
type (54 ± 3 versus 63 ± 5, respectively; n =
6, p < 0.02), whereas in the contracting wild-type (n =
5) and mutant muscle (n = 6), the ATP/AMP ratio was equally
maintained (42 ± 4 versus 42 ± 2). These results
indicate that in the absence of AK1 and M-CK, the dynamics of adenine
nucleotide metabolism is affected with an apparently less efficient P
trafficking among separate adenine nucleotide pools.
Upon initiation of contraction, CrP levels in wild-type muscles decreased from 62.2 ± 2.6 nmol CrP·mg protein–1 in the resting state to 31.5 ± 3.0 nmol CrP·mg protein–1 in the active state, a reduction of 49%. In double mutants, a reduction only by 21% was observed (from 57.4 ± 1.4 nmol CrP·mg protein–1 in muscle at rest to 45.5 ± 3.4 nmol CrP·mg protein–1 in active muscle; p < 0.05, n = 6) (Fig. 2C). This indicates that the total CrP level does not correlate with muscle genotype when measured at rest but is differentially affected by the absence of M-CK/AK1 under conditions of higher metabolic demand.
Intracellular levels of Pi resulting from hydrolytic cleavage of ATP were similar in wild-type and MAK=/= muscles at rest (39.0 ± 1.8 versus 39.5 ± 0.6 nmol Pi·mg protein–1). Upon functional load, however, the level of Pi increased by 20% in MAK=/= but was still significantly lower compared with the 44% increase of Pi in wild-type muscle (46.8 ± 2.2 versus 56.7 ± 2.7 nmol Pi·mg protein–1 in MAK=/= versus wild type; p < 0.05, n = 6) (Fig. 2D). Accordingly, the CrP/Pi ratio, an index of the cellular energetic status, was significantly higher in the contracting MAK=/= muscle (1.0 ± 0.1 versus 0.6 ± 0.1; p < 0.05, n = 6) compared with wild-type controls (n = 6). In non-contracting muscle, this ratio was maintained at similar levels of 1.6 ± 0.1 versus 1.5 ± 0.1 in wild type and MAK=/=, respectively (p > 0.05, n = 6). These results indicate that, in the absence of M-CK and AK1, the dynamics of CK (i.e. mitochondrial CK-driven) catalysis and Pi utilizing and/or producing reactions are sufficient to maintain cellular metabolite levels at rest but become inadequate with increased metabolic demand.
Loss of AK- and CK-supported Contraction in MAK=/= Actomyosin—In skeletal muscle, M-CK is clustered at the M-line and I-band of myofibers where it regenerates ATP to sustain myosin ATPase activity (13, 24). AK1 also localizes on the myofibrillar I-band where it may work as an ATP regenerator to support muscle contraction (13, 21, 47). To determine whether this cellular AK1 and M-CK partitioning has functional significance, we compared the contractile properties of actomyosin complexes isolated from wild-type and MAK=/= muscles. Using wild-type actomyosin complexes, in the presence of 0.2 mM ADP, AK1-supported induction of contraction was observed (Fig. 3A). Similar contractile characteristics were seen in the presence of 2 mM CrP based on the M-CK activity that remained in the actomyosin complex after isolation. However, the AK1- and M-CK-supported contractions were abolished in actomyosin complexes isolated from MAK=/= muscles (Fig. 3A). Lack of contraction in wild-type actomyosin complexes in the presence of ADP was also observed with pharmacological inhibition of the AK circuit by diadenosine pentaphosphate, Ap5A (data not shown). The addition of 0.2 mM exogenous ATP as a "direct" substrate caused wild-type actomyosin complex to contract with a maximum amplitude of 0.51 ± 0.02 absorbance units·min–1·mg protein–1 (n = 3) (Fig. 3B). The onset of contraction was, however, delayed in comparison when ATP was delivered via AK1 and M-CK enzymes that co-purified with the actomyosin complex (Fig. 3, A and B). On average, in wild type, the times to onset of contraction after addition of 0.2 mM ADP and 2.0 mM CrP (plus 0.2 mM ADP) were tcontr, ADP = 28 ± 14s(n = 3) and tcontr, CrP = 15 ± 4s(n = 3), respectively, whereas after direct addition of 0.2 mM ATP, a significantly longer tcontr, ATP = 334 ± 13s(n = 3; p < 0.01) was required to initiate contraction. These results suggest that the actomyosin-associated M-CK and AK1 enzymes serve to channel ATP molecules and thereby provide the myosin ATPase microenvironment with preferential access to metabolic energy.
|
Although lack of cytosolic M-CK and AK1 phosphotransfer blunted CrP- and ADP-supported actomyosin contraction, activation of pyruvate kinase-catalyzed ATP production by the glycolytic substrate phosphoenolpyruvate (PEP) secured contraction in both wild-type and MAK=/= actomyosin complexes (Fig. 3C). PEP at 2.0 mM produced contraction in MAK=/= actomyosin at approximately similar rates as in wild type, i.e. 0.064 ± 0.008 (n = 3) in wild-type and 0.05 ± 0.01 (n = 2) absorbance units·min–1·mg protein–1 in mutant actomyosin (Fig. 3C). Thus, the ability of actomyosin to contract using ATP delivered through the glycolytic pathway was not affected in the MAK=/= muscle, suggesting that alternative safeguarding routes for high energy phosphoryl transfer are operational in the absence of M-CK and AK1.
Glycolytic and Guanylate Metabolism in MAK=/= Muscle— Studies in AK1 or M-CK single mutant mice and cell model systems suggest a compensatory interchange between the AK/CK circuits and the glycolytic machinery for energy production (4, 6, 16, 33, 39, 48). We determined here that 18O-metabolic labeling of Glc-6-P by hexokinase, the initial step in the glycolytic cascade, was significantly increased from 8.1 ± 0.9% (n = 6) in the wild-type to 18.9 ± 0.9% (n = 6) in MAK=/= muscle (p < 0.05) at a labeling rate of 1.0 ± 0.1 and 3.6 ± 0.5 nmol·min–1·mg protein–1 (p < 0.05, n = 6) (Fig. 4A), respectively. In line with this finding, the total Glc-6-P content in MAK=/= muscle contracting at 2 Hz was higher than in wild type (12.9 ± 1.3 versus 5.9 ± 0.3 nmol Glc-6-P·mg protein–1, respectively, p < 0.05, n = 6). These results implicate elevated glycolytic phosphotransfer activity that could compensate for lack of metabolic flux through the AK1- and M-CK-mediated reactions. However, lactate levels remained unchanged in the contracting MAK=/= muscle (50 ± 2 versus 47 ± 2 nmol lactate·mg protein–1 in mutant versus wild-type muscles; n = 6), suggesting a coordinated elevation of pyruvate oxidation, i.e. mitochondrial Krebs cycle activity in conjunction with accelerated aerobic glycolysis.
|
Remodeling of intracellular high energy phosphoryl fluxes through guanylate
phosphotransfer circuits may provide additional compensation for deficits in
the cellular phosphotransfer network
(6,
18). Guanine and adenine
nucleotide metabolism is intertwined through NDPK, guanylate kinase, and
succinate thiokinase-catalyzed reactions
(42,
49–51).
Notably, the percentage of -GTP 18O labeling was
significantly higher in MAK=/= muscle (22 ± 2% in
MAK=/= versus 13 ± 1% in wild-type, respectively;
p < 0.05, n = 6) (Fig.
4B), indicating an increased flux through
GTP-metabolizing enzymes such as succinyl-CoA synthase in the Krebs cycle
and/or NDPK. Moreover, in support of enhanced guanylate phosphotransfer, an
increased turnover of -phosphoryls in GTP and GDP was observed in
MAK=/= muscle. The percentage of combined -GTP/GDP
18O labeling was 2.8 ± 0.1 (n = 3) and 4.5 ±
0.3 (n = 3) in wild-type and MAK=/= skeletal muscles,
respectively (p < 0.05; Fig.
4C), indicating increased guanylate kinase-catalyzed
phosphotransfer in the MAK=/= muscle. Consequently, the increased
guanine nucleotide contribution to cellular energetics resulted in a 2-fold
reduction in the ATP/GTP ratio in mutant muscle (41 ± 3 versus
78 ± 3 in MAK=/= and wild-type muscle, respectively;
p < 0.05, n = 6). Thus, in response to combined AK1/M-CK
deficiency, mutant muscles demonstrate elevated glycolytic and guanine
nucleotide phosphotransfer metabolism.
Disrupted Intracellular Energetic Communication in
MAK=/= Muscles—Efficient removal of ADP and
Pi released in the hydrolysis of ATP is a prerequisite for optimal
muscle function (52,
53). Pi produced by
cellular ATPase activity in one compartment must be captured by ATP-supplying
pathways to regenerate ATP in another compartment. The
[18O]phosphoryl labeling technique permits assessment of metabolite
exchange among different cellular sites
(54). The
[18O]Pi/[18O]-ATP ratio, an index for
intracellular phosphotransfer communication, was significantly reduced in
MAK=/= muscles (0.45 ± 0.03 versus 0.66 ±
0.03; p < 0.05, n = 6), i.e. by 32% compared
with the wild type (Fig.
5A). Although other factors could contribute to the
observed changes, reduction of the
[18O]Pi/[18O]-ATP ratio is in line
with the compromised ability of mutant muscle to maintain metabolite levels
and ratios. In this regard, the lower
[18O]Pi/[18O]-ATP ratio in contracting
MAK=/= muscle was accompanied by increased 18O labeling
of -ATP (22.2 ± 0.8% in MAK=/= versus 17.9
± 0.6% in wild type, p < 0.05, n = 6)
(Fig. 5B). These
results suggest an increased ATP supply capacity (apparently mitochondrial)
yet perturbed energetic communication between cellular locales of ATP
production and consumption in MAK=/= muscles.
|
Phosphotransfer Dynamics in MAK=/=
Muscle—Metabolic demand and/or stress may cause redistribution
of fluxes through phosphotransfer pathways
(6,
16,
18,
19,
42,
54,
55). Therefore, we measured
the relative contribution of AK- and CK- and alternative phosphotransfer
reactions to total cellular phosphotransfer under resting and working
conditions. Absence of both AK1 and M-CK did not affect total ATP turnover in
wild-type or MAK=/= resting muscle, which were 11.4 ± 0.2
versus 11.5 ± 0.5 nmol
ATP·min–1·mg
protein–1, respectively (p > 0.05,
n = 6) (Fig.
6A). However, in the active contracting state, total
cellular ATP turnover increased by 114% in wild-type but only by 70% in
MAK=/= muscle, suggesting a lower energetic capacity of AK1 and
M-CK deficient myocytes. Indeed, in contracting MAK=/= muscle,
total ATP turnover was lowered by 20% compared with wild type, i.e.
19.5 ± 1.4 and 24.3 ± 0.7 nmol
ATP·min–1·mg
protein–1 (n = 6, p < 0.05)
(Fig. 6C),
respectively. In wild-type muscle, the estimated AK flux increased from 1.9
± 0.1 nmol·min–1·mg
protein–1 in resting muscle to 5.6 ± 0.4
nmol·min–1·mg
protein–1 in contracting muscle (n = 6,
p < 0.05), an increase of 194%, which could account for 17 and 23%
of total cellular ATP turnover, respectively
(Fig. 6, B and
D). Although the estimated CK flux was increased by 42%
from 9.4 ± 0.3 (n = 6) to 13.3 ± 0.9 (n = 6)
nmol·min–1·mg
protein–1 in resting and contracting wild-type
muscle, respectively (p < 0.05)
(Fig. 6), the contribution of
CK-mediated phosphotransfer to total cellular ATP turnover decreased from 83
to 55% in response to muscle workload. In the MAK=/= knock-out
muscle, the residual AK flux was 0.48 ± 0.05 in the resting state and
0.80 ± 0.08 nmol·min–1·mg
protein–1 in the contracting state. Under both
conditions, AK-catalyzed phosphotransfer activity could account for 4% of
total ATP turnover, indicating that a very limited AK capacity remained in
MAK=/= muscles (Fig. 6,
A and C). Indeed, residual AK activity in double
mutant muscles was 44 ± 9 nmol
ATP·min–1·mg
protein–1 (n = 5), which is
1–2% of total AK activity in the wild type. In mutant muscle, CK
flux increased from 5.1 ± 0.2 at rest to 6.0 ± 0.4
nmol·min–1·mg
protein–1 upon activation of contractions and
accounted for 45 and 31% cellular ATP turnover, respectively
(Fig. 6). Thus, the ablation of
AK1 translated into a 75% decline in the contribution of AK catalysis to
cellular ATP processing, a contribution further reduced (by 82%) under
increased workload. The absence of M-CK translated into a 46 and 44% reduction
in the contribution of CK catalysis to cellular ATP processing in resting and
contracting muscles, respectively. Therefore, in the active state, the
combined AK and CK contribution to cellular high energy phosphoryl transfer
was reduced from 78% in the wild-type to 35% in MAK=/= muscle. This
phosphotransfer deficit was partially compensated by glycolytic,
hexokinase-catalyzed, and guanine nucleotide-dependent phosphoryl exchange
mechanisms (Fig. 4).
Contribution of such mechanisms increased from 9% in wild type to 39% in
MAK=/=, leaving approximately 25–30% total metabolically
active ATP unexplained. This could be accounted for by the remaining
phosphotransfer systems, such as the glyceraldehyde-phosphate
dehydrogenase/phosphoglycerate kinase enzyme couple
(19,
33,
39,
56) and pyruvate kinase
(39) or by diffusional type of
ATP replenishment due to cytoarchitectural rearrangements
(57).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Skeletal muscle devoid of M-CK and AK1 showed a severe decline in total
muscle AK and CK activities, which translated into a 54 and 85% reduction in
CrP and ATP -phosphoryl turnover, respectively. The remaining CK
activity can be attributed to the ScCKmit isoform still present in
mitochondria of MAK=/=. The reduction in -phosphoryl turnover
is apparently less severe in MAK=/= compared with the single
AK1 knock-out muscle where a 99% reduction in -phosphoryl
turnover was observed (6),
suggesting that MAK=/= muscles are more proficient in recruiting a
functional reserve of the -phosphoryl transfer capacity. Increased flux
capacity through the AK3-catalyzed GTP:AMP phosphotransfer reaction could
underlie this phenomenon as the high energy phosphate from GTP is transferred
to AMP followed by conversion to ATP by oxidative phosphorylation in the
mitochondrial matrix (9), in
line with the increased -GTP turnover in MAK=/= muscles.
The presence of cytosolic CK and AK within local microenvironments of
actomyosin ATPases facilitates transfer of high energy phosphoryls to
regenerate ATP, thus supporting the kinetic and thermodynamic efficiency of
muscle contraction (5,
47). Indeed, resting and
contracting MAK=/= knock-out muscles displayed overall lower
ATP/ADP and higher ADP/AMP ratios compared with wild type. This suggests that
AK1 and/or M-CK is not only necessary for local ATP regeneration but also for
equilibrating adenine nucleotide gradients throughout the entire cytosolic
pool. Inefficient local and global communication in MAK=/= muscle
is manifested in a decreased overall cellular ATP turnover and lower
Pi levels in response to increased workload. This may result from
insufficient delivery of ATP to cellular ATPase sites despite an increased ATP
synthesis capacity as indicated from increased 18O labeling of
-ATP.
The interaction between AK-CK and glycolytic phosphotransfer systems is critical for energy supply and local ATP regeneration in different cell types (4, 6, 16, 33, 39, 48, 53). In muscles deficient in both AK1 and M-CK, glycolysis is probably the only remaining major cytosolic phosphotransfer circuit. Glycolytic metabolism could safeguard maintenance of appropriate ATP/ADP ratios in the myosin ATPase microenvironment required for efficient contractile cycles (58, 59). Indeed, the overall concentration and turnover rate of the first glycolytic intermediate, Glc-6-P, was more than doubled in MAK=/= mice. Contribution of glycolytic hexokinase-catalyzed phosphotransfer increased from 8% in wild-type to 37% in MAK=/= muscles, whereas combined activities of guanylate phosphotransfer enzymes such as succinate thiokinase, NDPK, and guanylate kinase could account for only 0.6 and 1.8% of total cellular ATP turnover in wild-type and MAK=/= muscles, respectively. Therefore, enhanced ATP production and distribution by glycolytic enzymes may have a general compensatory role in sustaining physiological relevant ATPase activity in MAK=/= muscle.
Localization of glycolytic enzymes within myofibrils at the I- and M-bands
as well as their binding to mitochondrial outer membrane provides a network
capacity for transferring and distributing high energy phosphoryls
(13,
60,
61). This global glycolytic
function integrating mitochondrial ATP production with cytosolic sites of ATP
consumption may underlie increased coupling between high energy phosphoryl
flux through glycolytic metabolism and muscle contraction in MAK=/=
mice. Indeed, the apparently increased glycolytic flux in MAK=/=
skeletal muscle was paralleled by the increased ATP synthesis rate and
enhanced metabolic labeling of -GTP. This result can be explained by
elevated Krebs cycle activity
(42) coupled to increased
oxidative phosphorylation activity. The finding of a significant increase in
the volume of intermyofibrillar mitochondria in MAK=/=
muscles2 similar to
that seen in AK1 and M-CK single mutants
(6,
24) and in other muscles with
compromised energy metabolism
(62–64)
supports this conclusion. Thus, it is conceivable that MAK=/=
muscles combine the advantages of increased glycolytic flux and improved
mitochondrial potential for aerobic ATP production to compensate for the loss
of M-CK and AK1 supported high energy phosphoryl transfer activities.
Measurements in isolated actomyosin showed that the onset of contraction
was 12–20-fold faster if ATP was provided via AK1 or M-CK rather
than by direct ATP supply through diffusional nucleotide exchange. In this
regard, previous functional studies on fast-twitch muscles of mice lacking
M-CK have demonstrated lack of burst activity
(24) and compromised ability
to sustain contraction force
(43), in line with loss of
preferential access to ATP via the CK reaction. Similarly, AK1 activity
facilitates muscle relaxation, indicating that removal of contraction-produced
ADP by AK1 and its phosphorylation to ATP takes place much faster than
diffusional exchange of ADP to ATP
(47). As ultra-fast local and
global ATP regeneration is required to sustain contractile activity in
fast-twitch muscle (24,
52), this would underscore the
particular importance of CK- and AK-catalyzed phosphotransfer in these muscle
types. Moreover, mice carrying a single mutation in the AK1 or
M-CK gene show more pronounced adaptations at the molecular and
architectural level in their fastrather than slow-twitch fibers
(24,
37–39).
Thus, tight coordination between cellular sites of ATP consumption and ATP
generation by the complete set of high energy phosphoryl transfer pathways
would not only be essential for safeguarding cellular energetic economy
(6,
33,
54,
65) but also for preserving
the differential physiological contractile characteristics of slow- and
fast-twitch muscles. Studies in vivo with measurement of
physiological performance of intact muscle will be eventually necessary to
reveal further details.
In summary, this study demonstrates that lack of the M-CK and AK1 catalytic capacity abolished CK- and AK-mediated efficient high energy phosphoryl delivery to skeletal muscle actomyosin ATPase accompanied with compromised ability to sustain ATP turnover and to maintain nucleotide ratios in response to functional load. An apparent compensatory redistribution of high energy phosphoryl flux through glycolytic and guanylate phosphotransfer pathways lessened the genetic stress-induced energetic burden. The results provide reverse genetics-based evidence for the physiological significance of an integrated cellular phosphotransfer and energy-homeostasis network composed of complementary enzymatic pathways.
|
FOOTNOTES |
|---|
¶ To whom correspondence should be addressed: Dept. of Cell Biology, Nijmegen Center for Molecular Life Sciences University Medical Center, Medical Faculty University Nijmegen, P. O. Box 9101, 6500 HB Nijmegen, The Netherlands. Tel.: 31-24-3614329/3614287; Fax: 31-24-3615317; E-mail: b.wieringa{at}ncmls.kun.nl.
1 The abbreviations used are: CK, creatine kinase; AK, adenylate kinase;
ScCKmit, skeletal muscle containing an additional mitochondrial CK isoform;
NDPK, nucleoside diphosphokinase; GPS, gastrocnemius-plantaris-soleus; CrP,
creatine phosphate; Glc-6-P, glucose 6-phosphate; PEP, phosphoenolpyruvate;
MM-CK, muscle-type dimeric cytosolic CK.
2 E. Janssen, unpublished observation.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
