INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

CREATINE KINASE (CK), an enzyme central to cellular high-energy phosphate metabolism catalyzes the following reaction where PCr is phosphocreatine and Cr is creatine. The high level of CK activity found in skeletal muscle ensures that, when high-energy phosphate production is necessary during repetitive contractile activity, ATP levels will be maintained at the expense of PCr. This role of CK as a temporal ATP buffer is widely accepted (12, 24). Indeed, recent studies have demonstrated that, in the complete absence of CK activity, an immediate marked decline in muscle force and power during repetitive activation is observed (19, 21, 27), clearly indicating an important role for CK in muscle function.

In mature skeletal muscle CK is expressed as two distinct isoforms (24, 26) that differ in their subcellular localization (24, 30) and together are postulated to provide spatial ATP buffering between sites of energy production and utilization (1, 24). Sarcomeric mitochondrial CK (ScCKmit) is localized to the mitochondrial intermembrane space and concentrated at "energy-transfer" contact sites, where the inner and outer mitochondrial membranes are in close apposition (24, 30). Studies suggest that ScCKmit is 1) associated with the adenine nucleotide transporter on the inner mitochondrial membrane (16) and a voltage-sensitive channel (porin) on the outer mitochondrial membrane (7) and 2) involved in the channeling of high-energy phosphates produced by oxidative phosphorylation into the cytosol as PCr (24, 30). In contrast, the myofibrillar (M-CK) isoform is cytosolic, a portion of which is localized at subcellular sites of energy utilization including the 1) actomyosin ATPase; 2) sarcoplasmic reticulum Ca²⁺-ATPase; and 3) sarcolemmal Na⁺-K⁺-ATPase. M-CK is also colocalized with glycolytic enzymes at sites of glycolysis on the I band. Despite extensive investigation demonstrating an association of CK isoforms at different subcellular sites and speculation that CK subcellular compartmentalization provides the optimal microenvironment for efficient substrate (ADP, Cr) and product (ATP, PCr) channeling (1, 24), the extent to which intracellular localization of M-CK and ScCKmit isoforms is functionally linked to overall muscle performance remains unclear.

Most skeletal muscles express predominantly the M-CK isoform (>95% of total CK activity) with limited ScCKmit expression (1-2% of total CK activity) (24, 26). The adult diaphragm (Dia), in contrast, expresses both M-CK and ScCKmit in abundance (26, 27), with ScCKmit accounting for up to 25% of total CK activity in this muscle (26, 27). The Dia from transgenic mice, in which CK expression has been altered via targeted mutagenesis, provides a powerful tool for examining how the activity of the two isoforms and, by inference, their intracellular localization, modulate muscle performance in vitro (19-23). In the present study, we determined the isometric contractile and fatigue properties of 1) control CK-sufficient (Ctl), 2) M-CK-deficient (M-CK[/]), and 3) combined M-CK/ScCKmit-deficient null mutant (CK[/]) Dia to assess the effect of the absence of M-CK activity on Dia performance in vitro. Moreover, we explored the metabolic consequences of isolated M-CK and combined M-CK/ScCKmit deficiency on the Dia by determining the fiber type composition, succinic dehydrogenase (SDH) activity, adenylate kinase (AK) activity, and glycogen content across all three study groups. We hypothesized that both M-CK[/]- and CK[/]-deficient Dia would have diminished force generation during repetitive activation relative to CK-sufficient Ctl Dia, supporting a functional linkage between M-CK and Dia performance.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Studies were conducted on adult (90- to 110-day old) M-CK[/]-deficient transgenic, CK[/] double-deficient transgenic, and CK-sufficient Ctl mice, all of which had a mixed genetic (C57B1/6 × 129/Sv) background. The M-CK[/]-deficient and CK[/] double-deficient transgenic lines were generated by Steeghs et al. (19-21) and van Deursen et al. (22, 23) as previously described and confirmed by PCR analysis. Eight animals from each study group were anesthetized with pentobarbital sodium (60 mg/kg ip), and individual segments of Dia were excised for 1) in vitro assessment of isometric contractile and fatigue properties and 2) measurement of CK activity. Dia from additional animals in each study group were excised for quantitative measurements of AK activity [n = 10/group], glycogen content [n = 9/group], and fiber-specific SDH activity [n = 6/group]. The experiments were approved by the Magee-Womens Hospital and the Research Institute Institutional Animal Care and Use Committee.

In vitro mechanical measurements. The methods for measuring in vitro mechanical properties of the Dia muscle have been previously described (28). Briefly, Dia muscle strips (~2 mm wide) were cut from the midcostal region of the right hemidiaphragm, with fiber attachments at the rib and central tendon left intact. The muscle segment was mounted in a vertical tissue chamber, which was constantly perfused with mammalian Ringer solution aerated with 95% O₂-5% CO₂ and maintained at 37°C (28). The monitored PO₂, PCO₂, and pH were 400-460 Torr, 35-40 Torr, and 7.35-7.40, respectively. The costal margin was fixed by use of a vascular clamp mounted in series to a micropositioner near the base of the tissue chamber. A small piece of aluminum foil was glued to the central tendon with cyanoacrylate and then attached to the force transducer (model 300B, Cambridge Technology) via fine wire. This provided a noncompliant attachment to the force transducer and prevented tearing of the central tendon. The muscle bundles were stimulated directly (Grass model S-88 stimulator and current amplifier) by use of monophasic rectangular pulses of cathodal current (1.0-ms duration) delivered through platinum plate electrodes placed ~1 cm apart. Muscle bundles were positioned midway between the two electrodes. To ensure supramaximal stimulation, current was increased by 50% over the current necessary to obtain peak twitch force (~250-300 mA). Muscle fiber length was adjusted incrementally by using a micropositioner until maximal isometric twitch force responses (P_t) were obtained [i.e., optimal fiber length (L_o)]. Twitch contraction (CT) and half relaxation times were also determined. Maximum tetanic force (P_o) was assessed by stimulating the muscle bundle at 200 Hz delivered in a 1-s train. Force signals were displayed on a digital oscilloscope (model 1602, Gould), digitized at 500 Hz, and stored on a computer disk file.

Determination of CK activity and isoenzyme distribution. Total CK activity was determined at 25°C by using a coupled enzyme assay (26, 27). The ATP generated by the CK reaction was used in a hexokinase/glucose-6-phosphate dehydrogenase-coupled enzyme system, which ultimately yields a reduced NADP (NADPH) proportional to CK activity (Sigma Diagnostics, St. Louis, MO). For this analysis, a 5- to 10-mg muscle sample was homogenized for 15 s in a 1:100 (wt/vol) dilution of CK extraction buffer containing 26 mM Tris, 0.3 M sucrose, 1% NP-40, and 20 mM -mercaptoethanol at pH 8.0. Homogenates were diluted to 1:100 in extraction buffer. Thirty-microliter aliquots of diluted homogenate were added to 1 ml of CK assay buffer at 25°C containing (in mM) 130 KCl, 10 Tris (pH 7.4), 1 MgCl₂, 2 AMP, 5 glucose, 0.7 NADP, 1.5 ADP, and 9 PCr, as well as 50 µM diadenosine pentaphosphate, 1.3 U of hexokinase, and 0.5 U of glucose-6-phosphate dehydrogenase. AMP and diadenosine pentaphosphate were included to inhibit AK from producing ATP. The rate increase in absorbance at 340 nm due to production of NADPH through the coupled enzyme reaction was used to determine CK activity. Care was taken to make sure the rate obtained changed linearly with the volume of homogenate added. Protein content was determined by using the Lowry method (8), and CK activity was expressed as micromoles per milligram protein per minute.

Determination of muscle SDH activity. Total muscle SDH activity was calculated on the basis of 1) the relative contribution of fiber types to total muscle CSA and 2) specific fiber type SDH measurements, the latter determined by using a microdensitometric procedure (3). This quantitative SDH technique has been described in detail in previous papers (3, 18, 28). Briefly, muscle segments for SDH analysis were excised and quickly frozen at L_o [1.5 × excised length (18)] in isopentane cooled to its melting point by liquid nitrogen. Dia muscle samples were taken from the right midcostal region. Serial cross sections of muscle fibers were cut at 6-µm thickness by using a cryostat (model 2800E, Reichert-Jung, Buffalo, NY) kept at 20°C and classified as fiber type I, IIa, IIx, or IIb on the basis of 1) pH lability of myofibrillar ATPase staining (18) and 2) immunohistochemistry (18). Alternate sections were reacted for SDH in the presence or absence (i.e., tissue blank) of enzymatic substrate (succinate). The stoichiometric reduction of nitro blue tetrazolium (NBT) to its diformazan (NBT-dfz) was used as an indicator of the SDH reaction. NBT-dfz is a colored compound (peak absorbance wavelength of 570 nm), the concentration of which is directly proportional to measured light absorbance [optical density (OD)]. Microscopic images of the muscle fibers were digitized into an array of 1,024 × 1,024 picture elements (pixels) at eight-bit resolution (256 gray levels) by using an image-processing system calibrated for photometry (0.02-2.00 OD units; MegaVision 1024XM). From the digitized images, the boundaries of individual muscle fibers (100-150/image) were outlined and the average fiber OD was calculated. In previous studies (3) the SDH reaction has been shown to be linear for at least a 9-min period, and the fiber OD at time 0 is the same as that of the tissue blank. On this basis a single end-point reaction time of 5 min was used to determine the maximum velocity of SDH enzyme activity within individual Dia fibers, expressed as millimoles fumarate per liter tissue per minute.

Determination of AK activity. Muscle samples were diluted 1:100 (wt/vol) and homogenized as in the CK assay. The enzymatic analysis was performed on samples with a final dilution of 1:1,000 (wt/vol). Homogenate samples of 30 µl were analyzed by using the spectrophotometric method of Passonneau and Lowry (13).

Determination of glycogen content. Muscle samples were diluted 1:6 (wt/vol) with 0.6 N perchloric acid and immediately homogenized. Muscle homogenates were incubated at 37°C for 2 h in amyloglucosidase to hydrolyze glycogen. Glycogen content was measured by using the protocol of Kepler and Decker (6).

Statistical analysis. Statistical methods [Mintab Statistical Software, PC Version Release 8.0, Minitab, State College, PA (15)] included a two-way analysis of variance for comparisons of animal weight, L_o, and in vitro contractile properties across CK phenotypes. CK, AK, and SDH activities, as well as glycogen content, were compared across study groups by using a Kruskal-Wallis test. Where significance across CK phenotypes was observed, post hoc analysis was performed by using Scheffé's test. To compare differences in force generation across time and CK phenotypes during repetitive activation, curve fitting and ANOVA were performed, followed by a post hoc analysis to identify differences at individual time points. Statistical significance was established at P < 0.05. Data are reported as means ± SD.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Physical characteristics and isometric contractile properties. Animal and Dia characteristics as well as Dia isometric contractile properties are presented in Table 1. Animal body weight, Dia L_o, CT, half relaxation time, and P_t-to-P_o ratio did not differ across CK phenotypes. Dia P_o normalized to muscle CSA (P_o /CSA), however, was slightly lower (~16%) in CK[/] mice compared with M-CK[/] and Ctl Dia (P < 0.05).

View this table: [in this window] [in a new window]   Table 1.   Physical characteristics and isometric contractile properties of Ctl, M-CK [/]-deficient, and CK [/] double-deficient diaphragm

Total CK activity and isoenzyme phenotype. Total CK activity was 729 ± 200 µmol · mg protein¹ · min¹ in Ctl Dia and was composed of ~25% ScCKmit and 75% CK-MM isoforms, as indexed by densitometry (Fig. 1). In contrast, total CK activity in M-CK[/] Dia was 147 ± 38 µmol · mg protein¹ · min¹, averaging 20% of Ctl levels. Only the ScCKmit isoform was noted on CK zymogram gels of M-CK[/] Dia (Fig. 1). Neither the M-CK nor ScCKmit isoform was noted on CK zymogram gels of CK[/] Dia (Fig. 1); total CK activity in CK[/] Dia was 8.5 ± 2.3 µmol · mg protein¹ · min¹, averaging 1.1% of Ctl levels. We surmise that the residual CK activity in CK[/] Dia originates from immature satellite cells and the smooth muscle lining of vasculature in this muscle.

View larger version (28K): [in this window] [in a new window]   Fig. 1.   Representative agarose gels stained for creatine kinase (CK) activity that demonstrate CK isoenzyme distribution {sarcomeric mitochondrial CK (ScCKmit) and cytosolic myofibrillar (M-CK; MM) isoforms in control CK-sufficient (Ctl), M-CK-deficient (M-CK[/]), and combined M-CK/ScCKmit-deficient null mutant (CK[/]) diaphragm (Dia)}. In whole muscle homogenates, neither the MM nor ScCKmit isoform was noted in CK[/] Dia as opposed to abundant expression of both isoforms in Ctl Dia. Whole muscle homogenate of M-CK[/] Dia expressed only ScCKmit isoform activity. Mitochondrial subfraction of Ctl and M-CK[/] Dia contained only ScCKmit isoform, whereas cytosolic mitochondrial-free subfraction contained only M-CK in Ctl Dia and neither M-CK nor ScCKmit in M-CK[/] Dia.

Isometric force generation during repetitive activation. Force production during repetitive isometric tetanic activation in CK[/], M-CK[/], and Ctl Dia is shown in Fig. 2. Force declined significantly and precipitously in CK[/] Dia during the early phase of repetitive activation compared with M-CK[/] and Ctl Dia. Subsequently, force declined in M-CK[/] and Ctl Dia as well, and forces stabilized at ~30-45% of baseline force in all three groups during the latter stages of repetitive contraction. The pattern of force decline did not differ significantly between M-CK[/] and Ctl Dia at any point during the activation paradigm. In contrast, CK[/] Dia force output remained significantly lower than in both M-CK[/] and Ctl Dia until the 120-s time point of repetitive activation. At the 120-s time point, all three CK phenotypes generated comparable force.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Changes in force in Ctl (solid line), M-CK[/] (short dashed line), and CK[/] (long dashed line) Dia during repetitive isometric activation. First stimulation and every other 1 thereafter are shown for entire 2-min period of repetitive activation. Pattern of force decline did not differ significantly between M-CK[/] and Ctl Dia at any point during activation paradigm. Significantly lower isometric force was seen in CK[/] as opposed to Ctl and M-CK[/] Dia after initial stimulus until 120-s time point of repetitive activation paradigm. At 120-s time point, all three CK phenotypes generated comparable force.

Fiber type proportions. The proportions of different fiber types in the Dia across CK phenotypes are reported in Table 2. The majority of fibers across all three CK phenotypes was classified as IIx, whereas type IIa fibers comprised approximately one-quarter and type I fibers only one-tenth of the fiber subgroups. Type IIb fibers were not observed in any of the three CK phenotypes. M-CK[/] Dia was composed of significantly more type IIa fibers and fewer type IIx fibers than Ctl and CK[/] Dia.

View this table: [in this window] [in a new window]   Table 2.   Fiber type proportions and CSAs of Ctl, M-CK [/]-deficient, and CK [/] double-deficient mouse diaphragm

Fiber CSA. The CSA of different fiber types are also reported in Table 2. The mean CSA of type IIx fibers was ~1.5 times that of type I or IIa fibers, regardless of CK phenotype. Type I and IIa fiber CSA were significantly lower in M-CK[/] and CK[/] compared with Ctl Dia, and type IIx fiber CSA was significantly lower in CK[/] compared with M-CK[/] and Ctl muscle.

Fiber SDH activities. Fiber-specific SDH activities are shown in Table 3 and were comparable across type I, IIa, and IIx fibers within a given CK phenotype. However, fiber-specific SDH activity was significantly higher in M-CK[/] and CK[/] compared with Ctl Dia (Table 3). Thus total muscle SDH activity was significantly higher in M-CK[/] and CK[/] compared with Ctl Dia (Table 3). There were no significant differences between M-CK[/] and CK[/] Dia fiber-specific or total muscle SDH activities.

View this table: [in this window] [in a new window]   Table 3.   SDH activities of different fiber types and total diaphragm muscle of Ctl, M-CK [/]-deficient, and CK [/] double-deficient mice

AK and glycogen measurements. AK activity and glycogen content for each Dia CK phenotype are shown in Table 4. Both AK activity and glycogen content were significantly higher in M-CK[/] and CK[/] transgenic Dia compared with Ctl Dia. No sigificant differences were noted, however, between M-CK[/] and CK[/] Dia in AK activity or glycogen content.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

The principal findings of this study are that 1) isolated M-CK deficiency alone is not associated with any greater impairment of force-generating capacity during repetitive activations than that observed in Ctl Dia, whereas 2) combined M-CK and ScCKmit deficiency is associated with a profound early sustained impairment of Dia force generation during repetitive isometric contractions. These results demonstrate that Dia function is not absolutely dependent on the presence of M-CK. Although both M-CK[/] and CK[/] Dia manifest metabolic adaptations, these adaptations did not differ in magnitude between transgenic lines, suggesting that abundant ScCKmit expression alone may provide energy-buffering capacity to counter the effects of isolated M-CK deficiency in M-CK[/] Dia. These findings challenge the concept that intracellular localization of M-CK is functionally linked to overall muscle performance in the Dia.

View this table: [in this window] [in a new window]   Table 4.   Adenylate kinase activity and glycogen content of Ctl, M-CK [/]-deficient, and CK [/] double-deficient mouse diaphragm

Prior studies designed to assess the physiological role of CK in muscle have utilized nonspecific inhibitors of total CK activity and substrate analogs of Cr [-guanidinoproprionic acid (-GPA)] and suggest an important role for CK in muscle energy metabolism (5). Studies in -GPA-fed animals demonstrate a loss of force-generating capacity during the early phase of isometric activation that likely reflects diminished skeletal PCr and ATP stores, whereas these same muscles demonstrate an ability to sustain force during the latter phases of repetitive activation, mirroring an increase in mitochondrial capacity and glycogen content in response to chronic -GPA exposure (11, 17). These studies, although evincing the importance of CK in skeletal muscle high-energy metabolism, do not provide insights regarding how the activity of individual CK isoenzymes or their intracellular localization modulates skeletal muscle performance because their methods employed global inhibition of CK activity.

To address the question of how the activity of individual CK isoenzymes and, by inference, their intracellular localization, modulates skeletal muscle performance in vitro we 1) studied the Dia, a muscle that expresses both M-CK and ScCKmit in abundance (26, 27) and 2) used targeted mutagenesis to selectively inhibit individual CK isoform activity in the intact muscle (19-23). Our study results confirm that CK[/] Dia expresses neither M-CK nor ScCKmit and that M-CK[/] Dia expresses only ScCKmit. Moreover, Dia ScCKmit expression is limited to the mitochondria, i.e., we did not detect any evidence for ectopic localization of ScCKmit protein in the cytosol of M-CK[/] transgenic animals. This observation is consistent with the findings of Steeghs et al. (19-21) and van Deursen et al. (22, 23) in M-CK[/] limb muscle, the premise that mitochondrial localization sequences are dominant and the conclusion that the M-CK[/] knockout did not perturb the ScCKmit mitochondrial localization signals.

We observed that isometric twitch characteristics were not different across wild-type, M-CK[/], and CK[/] Dia, consistent with reports on limb muscle from these lines (19, 21). In contrast, P_o /CSA was significantly lower in CK[/] Dia compared with Ctl and M-CK[/] Dia. The mechanism underlying the slight reduction of P_o /CSA in CK[/] Dia was not addressed by the present study, but this finding is consistent with the aforementioned observations in CK[/] limb muscle (medial gastrocnemius) (21), CK[/] Dia (27), and -GPA-treated rats (9).

The present study also demonstrated that, in the combined absence of ScCKmit and M-CK activity, Dia muscle was unable to sustain tetanic force production at Ctl levels during the early phase of repetitive isometric activation. Tetanic force was already significantly depressed in the second series of isometric tetanic activation, and this depression was persistent throughout the period of repetitive stimulation. The rapidity of this decline was marked and consistent with previous reports on CK[/] limb muscle during repetitive activation (19, 21), possibly reflecting the high energetic demands of repetitive isometric tetani and the impaired ATP-buffering capacity of CK[/] Dia. The mechanisms whereby CK deficiency imparts an adverse effect on Dia contractile performance remain unclear but may involve altered excitation-contraction coupling via impaired sarcoplasmic reticulum Ca²⁺ uptake and release (19, 25). Forces subsequently stabilized at the same level as Ctl and M-CK[/] Dia by 2 min of repetitive activation are also consistent with previous observations in CK[/] limb muscle (19, 21).

In marked contrast to CK[/] Dia, the absence of M-CK alone was not associated with any greater impairment of Dia force-generating capacity during repetitive isometric activation compared with Ctl Dia. This novel observation runs counter to our original hypothesis and differs from reports on the contractile performance of limb muscle from M-CK[/] null mutant mice in vitro (19, 21-23). Medial gastrocnemius from M-CK[/]-deficient mice fails to sustain twitch and tetanic force production during the early phase of repetitive activation, whereas forces stabilize at the same level as Ctl during the latter stages of repetitive contraction (19, 21-23). Moreover, the absence of ScCKmit alone [isolated ScCKmit null mutant (20)] in medial gastrocnemius muscle is not associated with any significant change in force-generating capacity during any stage of repetitive isometric activation compared with Ctl muscle (20), suggesting that the absence of M-CK accounted for the impairment of function in both M-CK[/] and CK[/] limb muscle.

The discrepancy in contractile performance between Dia and limb muscle from the M-CK[/] null mutant mice during repetitive activation may relate to the relatively abundant expression of ScCKmit in the Dia (26, 27) as opposed to in the medial gastrocnemius (1-2% of total CK phenotype) (21-23). Indeed, the limb skeletal muscle studies do not lend themselves as well to an examination of how the activity of the two CK isoforms and their intracellular localization modulate muscle performance because medial gastrocnemius muscle expresses M-CK almost exclusively, with little to no ScCKmit. It may be that the reduced performance of CK-deficient transgenic muscle (whether limb muscle or Dia) is linked to a general decrease in total CK activity irrespective of CK phenotype. A recent study of transgenic mouse limb muscle mutants with graded reductions in M-CK activity showed a strong positive correlation between the level of M-CK expression and the ability to maintain maximal muscle force during the initial phase of isometric activation (23). Moreover, only when total CK activity had fallen below 16-34% of Ctl levels were marked impairments of contractile performance during repetitive activation noted (23). Furthermore, in skeletal muscle lacking M-CK but expressing the brain isoform of CK, contractile activity is normal even though brain CK did not localize to myofibrils and was at ~30% of wild-type levels (14).

We observed that ScCKmit activity in M-CK[/] Dia was ~20% of Ctl Dia total CK activity. This level is within the range van Deursen and colleagues (23) documented as being sufficient to sustain contractile performance during repetitive activation. Similarly, this could explain why ScCKmit [/]-deficient limb muscle is able to sustain contractile performance at wild-type levels (20). The present data are consistent with the notion that reduced contractile performance of CK-deficient Dia muscle during repetitive activation is linked to a general decrease in total CK activity, irrespective of CK isoform phenotype, and challenge the concept that intracellular localization of CK isoenzymes at sites of energy production and energy utilization is functionally linked to overall muscle performance.

Both M-CK[/] and CK[/] Dia manifested metabolic adaptations, including an increase in oxidative capacity and glycogen content. The increase in Dia SDH activity was ~140% of Ctl levels in both M-CK[/] and CK[/] lines and was noted across all fiber types (I, IIa, IIx). This enhanced Dia oxidative capacity is consistent with previous observations of increased mitochondrial enzyme activity and an expanded mitochondrial volume in M-CK[/] (22, 23) and CK[/] (21) limb muscle. However, the increase in Dia oxidative capacity in M-CK[/] and CK[/] animals does not correlate with the differences in fatigue resistance during repetitive activation. These results suggest that, in these animals, factors other than oxidative capacity are involved in the determination of Dia fatigue resistance. Glycogen content was also significantly greater in M-CK[/] and CK[/] Dia compared with Ctl levels, consistent with previous observations in limb muscle from these same transgenic lines (21, 22) and skeletal muscle of -GPA-fed animals (17). Again, the increased glycogen content of the M-CK[/] and CK[/] Dia was not associated with improved fatigue resistance. However, the higher SDH activity and glycogen content of the Dia in M-CK[/] and CK[/] mice suggest that they are exposed to a prolonged metabolic stress similar to exercise and that they are able to adapt to altered CK activity by activating pathways of aerobic metabolism and gluconeogenesis.

AK activity was also increased in M-CK[/] and CK[/] Dia above levels in Ctl animals. It has been suggested that AK may act as a buffer for adenine nucleotides in skeletal muscle in a coordinated manner with CK catalysis (4, 31). If AK does have a role in energy phosphate transport because of an interaction and communication with CK, one might expect an opposite but complementary change in AK enzyme expression. This view is supported by the small but significant increase in AK activity observed in both M-CK[/] and CK[/] Dia and previous work by others (4, 14, 31). Such enhanced AK activity could contribute with ScCKmit to maintain the supply of high-energy phosphates to counter M-CK deficiency in M-CK[/] Dia. The observed adaptations in CK-deficient Dia (increased SDH and AK activity and glycogen content), however, do not differ in magnitude between M-CK[/] and CK[/] Dia, suggesting that they do not account for the difference between M-CK[/] and CK[/] contractile performance.

Two different models of the CK-PCr system have been proposed in the literature. In one, often referred to as the PCr-CK "shuttle" hypothesis, it is asserted that CK isoenzymes are highly compartmentalized at sites of energy production and utilization (1, 24). The diffusion of PCr-Cr and the specific localization of CK at the mitochondria, sites of glycolysis, the actomyosin ATPase, sarcoplasmic reticulum Ca²⁺-ATPase, and sarcolemmal Na⁺-K⁺-ATPase are envisioned to provide the optimal microenvironment for efficient substrate and product channeling and enhanced muscle performance (1, 24). In contrast, an alternative model asserts that CK-catalyzed fluxes are maintained near equilibrium in the cytosol by facilitated diffusion, i.e., the specific localization of CK isoforms is not essential for muscle performance (10, 12, 29). Although there is now considerable evidence of CK isoform localization within muscle cells (24), such localization does not necessarily indicate that CK isoforms (either ScCKmit at the mitochondria intermembrane space or M-CK localized to intracellular sites) are "segregated" from the cytosol (29). Data from the present study demonstrate that Dia muscle function is not absolutely dependent on the presence of M-CK. Moreover, the present results suggest that residual CK activity in the form of ScCKmit may play a crucial role in sustaining M-CK[/] Dia isometric function as Dia devoid of both M-CK and ScCKmit (CK[/]) manifests a profound short-term impairment of contractile performance. Thus the findings of the present study are more consistent with a cytoarchitectural arrangement in Dia muscle that does not require cytosolic CK. It may be that an elaborate mitochondrial network with abundant ScCKmit, relatively short physical distances between sites of ATP production and consumption, and/or small changes in other metabolic enzymes contribute to the resiliency of contractile performance in the M-CK[/] Dia.