Muscle contraction requires high energy fluxes, which are supplied by MM-CK (muscle-type creatine kinase) which couples to the myofibril. However, little is known about the detailed molecular mechanisms of how MM-CK participates in and is regulated during muscle contraction. In the present study, MM-CK is found to physically interact with the slow skeletal muscle-type MyBPC1 (myosin-binding protein C1). The interaction between MyBPC1 and MM-CK depended on the creatine concentration in a dose-dependent manner, but not on ATP, ADP or phosphocreatine. The MyBPC1–CK interaction favoured acidic conditions, and the two molecules dissociated at above pH 7.5. Domain-mapping experiments indicated that MM-CK binds to the C-terminal domains of MyBPC1, which is also the binding site of myosin. The functional coupling of myosin, MyBPC1 and MM-CK is further corroborated using an ATPase activity assay in which ATP expenditure accelerates upon the association of the three proteins, and the apparent Km value of myosin is therefore reduced. The results of the present study suggest that MyBPC1 acts as an adaptor to connect the ATP consumer (myosin) and the regenerator (MM-CK) for efficient energy metabolism and homoeostasis.
- ATPase activity
- energy homoeostasis
- muscle-type creatine kinase (MM-CK)
- slow skeletal muscle-type myosin-binding protein C (MyBPC)
In skeletal and cardiac muscle cells, movement of cross bridges extending from the thick to the thin filaments causes sliding of the filaments, and thus muscle contraction and relaxation. In addition to the primary thick and thin filament components myosin and actin, the sarcomere also contains accessory proteins. MyBPC (myosin-binding protein C), one of the accessory proteins, is a major component of the thick filament, and is believed to play a significant role in sarcomeric structure and regulation of muscle contractility. MyBPC is localized in the C-zone, a region in the central two thirds of the A-band [1,2]. Three isoforms of MyBPC exist in adult muscle: fast skeletal (MyBPC2), slow skeletal (MyBPC1) and cardiac (MyBPC3). MyBPC proteins mainly consist of ten globular domains named sequentially from C1 to C10, among which seven are Ig-I-like domains, and the other three are FnIII (fibronectin III) domains . Early studies suggested that MyBPC binds to the LMM (light meromyosin) of the myosin rod , and that the binding site lies in the C10 domain, a C-terminal Ig-I-like domain conserved in all three types of MyBPC [5,6]. The protein has drawn more interest since MyBPC3 was identified through gene association studies to be involved in a heart disease [FHC (familial hypertrophic cardiomyopathy)] . Contrary to other extensively studied sarcomeric proteins, the precise roles of MyBPC, especially the skeletal muscle isoforms MyBPC1 and MyBPC2, in sarcomeric structure and contractility regulation, is still elusive.
It has been well-documented that, during muscle contraction, the principal thick filament component myosin provides the energy necessary for filament sliding by hydrolysing ATP into ADP through its ATPase activity . In order to keep the sustained muscle activity in which ATP is rapidly consumed, the rapid renewal of ATP is extremely important. Actually, the total ATP level is maintained almost constant during physiological functions. The renewal of ATP is controlled by the PCr (phosphocreatine)–CK [Cr (creatine) kinase, EC 22.214.171.124] system, which connects the intracellular ATP microdomains/compartments at the sites of both ATP production and utilization [9–11]. In the PCr–CK network, four CK isoforms are functionally and/or structurally coupled either to sites of energy production or energy utilization [11–13]: two mitochondrial forms are functionally related to oxidative phosphorylation and production of PCr, whereas two cytosolic isoenzymes use PCr to buffer the fluctuation of the intracellular ATP/ADP ratio. The energy homoeostasis is thus achieved by the reversible reaction from MgATP and Cr to MgADP and PCr catalysed by these distinctly localized CKs. Among the four CK isoenzymes, the MM-CK (muscle-type CK) specifically binds to the myofibril M-line and is associated with the actin-activated myosin ATPase as an intramyofibrillar ATP regenerator [10,14–17]. The binding of MM-CK to the M-line is precisely regulated by pH , substrates [18,19] and the oxidative status of the enzyme . In addition to the M-line, MM-CK also exists in the I-band of the sarcomere .
Although MM-CK clearly plays a crucial role in energy homoeostasis during muscle contraction, the detailed molecular mechanism of how MM-CK functionally couples to myosin remains elusive. Hornemann et al. [14,15] showed that the M-line protein myomesin and M-protein interacted with MM-CK via four conserved lysine residues of MM-CK. However, no direct enzymatic evidence is available for the coupling of the catalysis of MM-CK and ATPase through these proteins. Moreover, it is unclear whether MM-CK also provides energy supply to the other parts of the myofibril and whether there are other proteins providing direct connection between MM-CK and myosin. In the present study, we found that MM-CK and MyBPC1 were physically and functionally associated, and that MyBPC1 acted as an adaptor to bridge MM-CK and myosin. These discoveries provide new insights into the molecular mechanism of sarcomere function and diseases caused by the imbalance between ATP production and utilization in muscle or other tissues with a high energy demand.
Mouse anti-MM-CK antibody, anti-FLAG antibody and M2-agarose-affinity gel were obtained from Sigma, and rabbit anti-MM-CK antibody was from Bioworld Technology. Anti-HA (haemagglutinin) antibody and anti-Myc antibody were purchased from Santa Cruz Biotechnology and Cell Signaling Technology respectively. Rabbit anti-GST (glutathione transferase) antibody was a gift from Dr Yinghua Chen (School of Life Sciences, Tsinghua University, Beijing, China). MF20 and ALD66 antibodies were purchased from the Developmental Studies Hybridoma Bank. Protein A/G Plus-agarose beads were from Santa Cruz Biotechnology. Glutathione–Sepharose 4B and Superdex G-200 were from Amersham Pharmacia Biotech. Skeletal myosin, Triton X-100, leupeptin, PMSF and PNPase (purine nucleoside phosphorylase) were from Sigma. MESG (7-methyl-6-thioguanosine) was obtained from Berry Associates. ATP, ADP, Cr and PCr were from Ameresco. All other chemicals were of analytical grade.
Yeast two-hybrid screen
The yeast two-hybrid screening was performed following the manufacturer's protocol using the Clontech BD Matchmaker GAL4 two-hybrid system and a human skeletal muscle cDNA library (Clontech). The full-length coding sequence of human MM-CK was cloned into the GAL4 DB (DNA-binding) vector pGBKT7,which was later transformed into the yeast strain AH109. Then the bait-bearing strain was co-transformed with a human skeletal muscle cDNA library, which was cloned into the GAL4 DNA AD (activation domain) vector pACT2. Lower stringency selection was performed on −Leu/−Trp/−His agar plates. The positive clones were then subjected to the higher stringency selection on −Ade/−Leu/−Trp/−His plates. After selection, X-α-Gal (5-bromo-4-chloro-3-indolyl α-D-galactopyranoside) was added to the plate for the colorimetric detection of the MEL1 reporter-gene product, α-galactosidase. Positive clones were finally isolated and verified by sequencing.
The human MM-CK-coding region was cloned into the expression vector pET21b, expressed in Escherichia coli BL21 (DE3)-pLysS (Stratagene) and purified as described previously . The entire coding sequence of human MyBPC1 consists of ten domains, which were subcloned into different combinations and named as follows: hMC1–3 (residues 1–430), hMC4–5 (residues 431–622), hMC6–7 (residues 623–815), hMC8–10 (residues 816–1124), hMC1–5 (residues 1–622) and hMC6–10 (residues 623–1124). They were all cloned into the GST fusion vector pGEX 6P-1 (Amersham Biosciences), and expressed in E. coli BL21 cells. Purification of GST-tagged protein was carried out on a glutathione–Sepharose 4B column according to standard protocols. The fusion proteins were further separated by a Superdex G-200 gel-filtration column. Binding studies were carried out in 500 μl of binding solution [20 mM Mops (pH 7.0), 20 mM NaCl, 5 mM MgCl2 and 20% glycerol] with 10 μl of washed glutathione beads, 75 μg of MM-CK and 100 μg of GST or GST–MyBPC1 fusion proteins. Following incubation overnight at 4 °C, the beads were carefully rinsed three times using the binding solution. The binding products were detected by Western blotting using anti-GST and anti-MM-CK antibodies. The appropriate ionic strength was screened, and no significant difference was found when the buffer contained 20–150 mM NaCl. To investigate the effect of the CK substrate on the binding, ATP, ADP, Cr or PCr was added respectively in the GST-pulldown binding solutions with a final concentration of 5 mM and then the pH of the binding system was adjusted to pH 7.0. The pH-dependence of the binding was performed by adjusting the pH of the binding buffer ranging from 6.0 to 8.0. Quantification analysis of the Western blot bands was carried out using the Band Scan software (Glyko)
Co-IP (co-immunoprecipitation) assays
Prior to transfection, 6×105 HEK (human embryonic kidney)-293T cells (A.T.C.C.) were seeded in a 60-mm dish for 24 h. The coding sequences of MM-CK, MyBPC1 and its mutants were subcloned into the eukaryotic expression vector pCMV5 containing HA, FLAG or Myc tags. The LMM region of hMHC-1 [human slow skeletal muscle MHC (myosin heavy chain)-1] was fused into the pCMV5 vector containing the FLAG tag. The plasmids were co-transfected using Vigofect transfection reagent (Vigorous) according to the manufacturer's instructions. As for the Co-IP of MyBPC1 and MM-CK, 8 μg of FLAG-tagged wild-type or mutated MyBPC1 and 2 μg of Myc- or HA-tagged MM-CK incubated with 4 μg of Vigofect were added to one 60-mm dish with the HEK-293T cells. To examine the interaction of the three proteins, 3 μg of FLAG–hMHC-1, 5 μg or 10 μg of Myc–hMC8–10 and 3 μg of HA–MM-CK were co-transfected. After 20–24 h transfection, the HEK-293T cells were grown to 80–90% confluence. The cells were washed with ice-cold PBS buffer and harvested using 600 μl of IP (immunoprecipiation) lysis buffer. The IP lysis buffer consisted of 10 mM Tris/HCl (pH 7.4), 20 mM NaCl, 10 mM EDTA, 10 mM EGTA, 1% Triton X-100, 1 mM PMSF and 1 mg/ml leupeptin. The cell lysates were sonicated (at 20% duty cycle for 10 pulses) and centrifuged for 30 min at 15000 g at 4 °C. A 50 μl aliquot of supernatant was used to test the expression of protein, whereas the remaining supernatant was incubated with 10 μl of anti-FLAG M2-agarose-affinity gel or HA-antibody-linked Protein A/G Plus-agarose beads overnight on ice. The beads were washed three times gently with the IP lysis buffer. Both the supernatants and pellets were examined by immunoblotting using anti-FLAG, anti-Myc or anti-HA antibodies.
The Co-IP assays of endogenous MyBPC1 and MM-CK were performed using C2C12 mouse skeletal myoblast cell lines (A.T.C.C.). To induce differentiation, cells were cultured in DMEM (Dulbecco's modified Eagle's medium; Gibco, Invitrogen) supplemented with 1% FBS (fetal bovine serum; Gibco, Invitrogen) for 96 h. C2C12 myotube lysates were prepared and incubated with 2 μg of antibodies against MyBPC1 (ALD66) or control-linked Protein A/G Plus-agarose beads overnight at 4 °C. Agarose beads were washed as described above, separated by SDS/PAGE (10% gel), and immunoblotted with anti-MyBPC1 and anti-MM-CK antibodies.
The co-sedimentation experiment was carried out according to a method described previously with some modifications [5,23]. Briefly, skeletal myosin stored in glycerol was dialysed against myosin-binding buffer [0.1 M KCl, 20 mM imidazole/HCl (pH 7.0) and 1 mM DTT (dithiothreitol)] to obtain reconstituted filaments. The filaments were then pelleted by centrifugation (15000 g for 30 min) and solubilized in the high-ionic-strength solution [0.6 M KCl, 20 mM imidazole/HCl (pH 7.0) and 1 mM DTT]. Myosin was then aliquoted (100 μl aliquots) and the KCl concentration was diluted from 0.6 M to 0.1 M by adding distilled water. The final concentration of myosin and MM-CK in all experiments was 3 μM and 1.1 μM respectively. The MyBPC1 mutants hMC1–3, hMC4–5 or hMC8–10 were added to a final concentration of 2.5 μM. The solution was incubated on ice for 1 h. Myosin filaments associated with the binding proteins were then centrifuged in an Airfuge (Beckman Instruments) at 22 psi (1 psi=6.9 kPa) for 20 min at 4 °C. Finally, the pellets were resuspended in 50 μl solutions containing 0.1 M KCl, 20 mM imidazole/HCl and 0.5 mM PMSF. Supernatants and pellets were resolved by SDS/PAGE (10% gel). Western blotting was used to analyse the precipitates using the MM-CK-specific antibody.
Mice were killed humanely and the soleus muscle, which mainly consists of slow twitch muscle fibres, was dissected from the hind limbs. Animal experimentation was conducted in the animal facility of Tsinghua University and approved by the Institutional Animal Care and Use Committee of Tsinghua University. The longitudinal and transverse sections were prepared by following the standard procedures and fixed on microscope slides (SuperfrostPlus, Fisher). For immunostaining, the slides were microwaved for 15 min in antigen-retrieval buffer [10 mM sodium citrate (pH 6.0)], cooled for 20 min at room temperature (25 °C), and washed three times with PBST (PBS containing 0.1% Tween 20). The sections were permeabilized by incubating the slides in 1% Triton X-100 in PBS at 4 °C for 30 min, followed by three washes in PBST. Subsequently, sections were blocked in blocking buffer (PBS containing 10% goat serum and 0.2% Triton X-100) for 1 h. Double staining was carried out with a rabbit anti-MM-CK antibody (1:100 dilution) and a mouse monoclonal antibody against MyBPC1 (slow type, ALD66) or against the mouse myosin heavy chain (sarcomeric, MF20) antibody at a 1:100 dilution in blocking buffer. After three washes in PBST, slides were incubated with Cy3 (indocarbocyanine)-conjugated anti-mouse and Cy5 (indodicarbocyanine)-conjugated anti-rabbit secondary antibodies (Jackson ImmunoResearch Laboratories) for 1 h. Samples were detected using the Zeiss LSM 710 confocal microscope system (Carl Zeiss). Pictures were taken as sequential laser scans to prevent potential overlapping of the fluorescence signals.
ATPase activity measurement
The ionic strength of skeletal muscle cells was calculated using a computer program developed previously  using the absolute stability constants reported by Fabiato . The reaction buffer consisted of 30 mM Tes (pH 7.1), 10 mM EGTA, 130 mM KCl, 3.16 mM free Mg2+ and 31.6 μM free Ca2+. Myosin Ca2+-dependent ATPase activity was determined using a continuous spectrophotometric assay , which is a coupled enzyme system using PNPase and its chromogenic substrate MESG for the quantification of Pi produced in the ATPase reaction . All experiments were performed at 25 °C in a 0.4 ml reaction mixture containing 100 μM MESG, 0.1 mg/ml PNPase, 50 nM skeletal myosin, 0.2 mg/ml CK and 12.2 mM PCr in the reaction buffer. After 5 min of equilibration at 25 °C, the reactions were initiated by the addition of ATP unless otherwise indicated. The time-course absorbance change at 360 nm was monitored on an Ultraspec 4300 spectrophotometer (Pharmacia Biotechnology). Initial velocities were calculated from the linear slope of progress curves obtained. Quantification of phosphate release was determined using the molar absorption coefficient at 360 nm of 11200 M−1·cm−1 for the reaction at pH 7.1 . The concentration of MESG was determined by the absorbance at 331 nm using a molar absorption coefficient of 32000 M−1·cm−1. All experiments were repeated at least three times.
Statistical analysis was performed using the GraphPad Prism software (GraphPad Software). The unpaired two-tailed Student's t test was used to compare the sets of data assuming Gaussian distribution, and a P value of less than 0.05 was considered significant.
MyBPC1 interacts with human MM-CK through its C-terminal domains
Yeast two-hybrid screening in the human skeletal muscle cDNA library was carried out using human MM-CK as a bait. One of the colonies grew well on the −Ade/−Leu/−Trp/−His plate, which was further confirmed using the α-galactosidase assay (Figure 1A). Sequence analysis showed that the cDNA from the corresponding colony encoded the C-terminus of MyBPC1 with amino acid residues ranging from 1060 to 1141. The interaction was further confirmed in the HEK-293T cells by Co-IP experiments, in which the full-length MyBPC1 was fused with a FLAG tag and co-transfected with Myc-tagged MM-CK into the HEK-293T cells. As expected, MM-CK co-precipitated with the full-length MyBPC1 (Figure 1B). To investigate whether the endogenous MyBPC1 interacts with MM-CK, the Co-IP assay was performed using C2C12 skeletal myotubes. As shown in Figure 1(C), MM-CK was detected in the IP product and was absent in the control immunoprecipitates. These studies demonstrated the physical interaction between the two proteins.
A GST-pulldown assay was performed to identify the core sequences essential for the direct interaction between MyBPC1 and MM-CK. The ten domains of MyBPC1 were divided into four truncated forms (Figure 2A). As shown in Figure 2(B), both hMC6–7 and hMC8–10 could pull down MM-CK, suggesting that MyBPC1 interacted with MM-CK through its C-terminal domains 6–10. The domain-mapping results were further confirmed by Co-IP assay in HEK-293T cells. Consistent with the GST-pulldown results, MM-CK was co-immunoprecipitated with the truncated proteins hMC6–10, hMC6–7 and hMC8–10, but not with hMC1–5 (Figure 2C). Similar results were obtained in the opposite Co-IP direction (Figure 2D).
The interaction between MyBPC1 and MM-CK is Cr- and pH-dependent
The binding of MM-CK to the M-line is precisely regulated by many factors [14,18–20], and thus a screen of the effects of substrates and pH on the MyBPC1–CK interaction using the GST-pulldown assay was performed to elucidate their physiological relevance. The experiments were carried out by adding each of the four CK substrates, ATP, ADP, Cr or PCr to the binding solutions of the GST-pulldown assay with a physiological concentration of 5 mM . No interaction was observed in the presence of 5 mM Cr, whereas the other three substrates did not affect the interaction of CK with hMC8–10 (Figure 3A). Furthermore, the interaction was found to be dependent on the Cr concentration (Figures 3B and 3C), where the increase in Cr concentration led to the dissociation of the MyBPC1–CK complex.
The intracellular pH of muscle cells is between 6.0 and 8.0 . Consistent with previous observations , the interaction between MM-CK and MyBPC1 was sensitive to pH changes (Figures 3D and 3E). MM-CK and MyBPC1 optimally bound in the pH range from 6.25 to 7.25, and the interaction was significantly weakened when the system pH changes from the optimal range. This observation further confirmed that the interaction between MyBPC1 and MM-CK might be functionally correlated.
Characterization of the interaction of MM-CK, MyBPC1 and myosin in vitro and in vivo
MyBPC is structurally and functionally associated with myosin under physiological circumstances , and CK is known to be involved in muscle energy metabolism [10,14–17]. Thus both in vitro and in vivo binding assays were carried out to explore the possible association of the three proteins. An in vitro co-sedimentation assay was employed using the purified myosin, MyBPC1 and MM-CK. At a physiological ionic strength, skeletal myosin reconstitutes into filaments, and MyBPC co-sediments to the pellets. The primary myosin-binding site of MyBPC resides in the C-terminal Ig-I-like domain C10 [4–6]. Co-incidently, the results of the present study showed the co-existence of myosin and hMC8–10 in the pellets (Figure 4A, top panel). In the negative control, most of the hMC1–3 and hMC4–5 molecules appeared in the supernatants (Figure 4A, middle panel). A little hMC1–3 remained in the pellets (indicated by the arrow), because in addition to the primary binding site C10, the N-terminal domains C1 and C2 of MyBPC also bind the myosin S2 (subfragment 2) region [32,33]. MM-CK was also seen to associate with myosin and hMC8–10 (Figure 4A, bottom panel). Notably, both myosin and hMC8–10 were essential for the binding of MM-CK, implying that MM-CK did not directly interact with myosin, but was mediated by MyBPC1. A 45 kDa protein was detected in our SDS/PAGE studies as a component of commercial myosin, and it was later identified to be actin by mass spectrometric analysis (results not shown). However, the existence of actin did not affect the conclusions above, as revealed by the control experiments using the N-terminal fragments of MyBPC1.
Co-IP experiments were performed to verify the physical association of myosin, MyBPC1 and MM-CK. The LMM region of hMHC-1 encoding the MyBPC-binding part of myosin was cloned , and a Co-IP assay was conducted together with MM-CK and hMC8–10. The results indicated that hMC8–10 strongly interacted with the slow-type myosin, whereas MM-CK could not bind to myosin in the absence of MyBPC1 (Figure 4B). Meanwhile, more MM-CKs were immunoprecipitated with the increasing amount of hMC8–10 applied.
The association of MM-CK with MyBPC1 and myosin was further investigated in situ in the slow skeletal muscle sections. In longitudinal sections (Figure 4C, a and c), MM-CK mainly localized at the I-band of adjacent sarcomeres and the flanking area, similar to that described in a previous study . The M-line distribution was not detected in the present as well as a previous study due to its weak staining in the intact muscle sections , mostly because of the difference in sample preparation procedures. Consistent with previous reports [1,34], MyBPC1 localized to a doublet in longitudinal sections. The distribution of MHC was almost in-line with that of MyBPC1. At higher magnification, MM-CK showed thick staining at the I-band with relatively faint staining at the periphery of the I-band, extending to the MyBPC1 and MHC distribution area (Figure 4C, a and c, arrows). In transverse sections (Figure 4C, b and d), the anti-MM-CK antibody showed staining at the periphery of the fibre at the sarcolemma (arrows), as well as a banding pattern in the cytoplasm. MyBPC1 and MHC antibodies exhibited similar staining patterns and overlapped with MM-CK at both the sarcolemma and cytoplasm. Thus both the in vitro and in vivo studies suggested that MM-CK, MyBPC1 and myosin were physically coupled.
Functional coupling of myosin ATPase and MM-CK mediated by MyBPC1
An ATPase activity assay was performed to further investigate the physiological implications of the interaction of MyBPC1 with MM-CK and myosin. As indicated in Figure 5(A), in the absence of MyBPC1 and CK, the myosin ATPase activity was (1.2±0.2)×10−2 μM Pi/s at an ATP concentration of 0.625 μM (column 1). Upon the addition of MM-CK alone to the reaction solutions, the ATPase activity of myosin was significantly increased by approximately 3.3-fold (column 2). In this two-enzyme system, myosin ATPase catalyses ATP to produce ADP and Pi, whereas MM-CK rephosphorylates ADP by PCr to provide ATP for the consumption of myosin ATPase. The rate-limiting factor is more likely to be the rate of ADP release from the ATPase and the diffusion of ADP to the rephosphorylation site. The addition of hMC8–10 alone does not improve the ATPase activity of myosin (column 3, P> 0.05). When both MM-CK and hMC8–10 were added in the reaction solution of myosin (column 5), the activity was increased by approximately 2-fold as compared with the solution containing myosin and CK, possibly caused by the proximity of the two enzymes upon the addition of MyBPC1, as shown above. Meanwhile, the addition of hMC1–3 did not affect the ATPase activity of myosin in the presence of CK (column 4).
To further characterize the functional coupling of MM-CK to myosin through MyBPC1, the myosin ATPase activity was measured by varying the initial ATP concentration from 0.31 μM to 16 μM (Figure 5B). The reaction followed Michaelis–Menten kinetics. The discrepancy between the two lines showed an ATP-concentration dependence, where a greater discrepancy was observed at low initial ATP concentrations. The apparent Km values for the ATPase were 2.7±0.2 μM and 6.7±0.5 μM for solutions in the presence or absence of MyBPC1 respectively. The 2.5-fold reduction of the apparent Km value in the presence of MyBPC1 suggested that the physical binding of CK, myosin and MyBPC1 greatly influenced the enzymatic kinetics of the ATPase activity of myosin.
MyBPC1 is a novel adaptor for the association of MM-CK with myosin
It has been well recognized that both ATP and the PCr–CK system is subcellularly compartmentalized, where PCr and Cr act as the shuttle molecules to connect the intracellular microdomains of ATP production and utilization . At myofibrils, the M-line-bound CK acts as an ATP regenerator and enhances myofibrillar ATPase activity [10,35–37]. MM-CK also exists at the sarcomeric I-band  and the outer face of the sarcoplasmic reticulum . Myomesin was the main molecule identified thus far that mediates the physical binding of MM-CK to myofibrils . In the present study, MyBPC1 was identified as a novel binding partner of CK. MyBPC has been proposed to be involved in sarcomere assembly, and its importance in muscle contraction has been revealed by FHC, a disease caused by inherited mutations in MyBPC . The findings of the present study support the involvement of MM-CK in energy metabolism at myofibrillar regions other than the M-line, and provide insights into the physiological functions of these two important molecules in the sarcomere.
In intact longitudinal skeletal muscles, MM-CK shows intense staining at the I-band, and partially extends into the C-zone where it co-localizes with MyBPC1 and MHC, with a relatively faint staining. The I-band localization of MM-CK has been revealed to be mediated by PFK (phosphofructokinase) and aldolase in a previous study . In the present study, the partial and relative weak co-staining of MM-CK with MyBPC1 and myosin suggested that their interaction might be transient or regulated by some factors different from the I-band localization. This speculation agreed with the fact that their association was regulated by one of the CK substrates, Cr, and environmental pH.
Although MyBPC1 could interact with MM-CK, the structural basis of their interaction remains elusive. Domain-mapping experiments suggested that the isolated fragments hMC6–7 or hMC8–10 participated in the direct binding with MM-CK. MyBPC1 has been proposed to take an extended conformation [14,39], in which domains 1–5, 8 and 10 are Ig-I-like domains, and domains 6, 7 and 9 are FnIII domains. Given the fact that hMC1–5 did not interact with MM-CK, whereas both hMC6–7 and hMC8–10 did, the FnIII domains might be the potential binding sites. However, the results of the present study indicated that both the isolated domain 10, and domains 8 and 9 could also interact with MM-CK, although to a lesser extent (results not shown). Therefore the linear organization of the ten domains of MyBPC1 might provide multiple MM-CK-binding sites. Meanwhile, it has been proposed that MyBP-C interacts with myosin through its C-terminal domain 10, and its N-terminal end may extend from the filament to bind the neighbouring molecules, such as actin filaments [6,31]. Thus the potential CK-binding sites at the C-terminal end of MyBPC1 might facilitate the close spatial localization of MM-CK and myosin, thereby further enhancing the catalytic coupling between the two enzymes.
Functional implications of the association of MM-CK with myosin mediated by MyBPC1
Myosin ATPase catalyses the conversion of ATP into ADP and phosphates, whereas MM-CK regenerates ATP for myosin from ADP and PCr. At the sites with a high energy demand, the existence of the physically associated CK could provide a rapid regeneration of ATP. This is extremely important for the muscle cells since the microstructure of the myofilament lattice impedes the diffusion of molecules . The bound CK not only helps to overcome the rate-limiting step of the diffusion of the substrate ATP and product ADP in the crowded and structured cytosol, but also affects the ATPase activity of myosin kinetically. Notably, our activity assay was performed in a homogeneous and separated in vitro system in the absence of the geometrical constraints of the sarcomere and associated proteins. The diffusion rates of CK and its substrates in this in vitro system might be faster than in the actual crowded biosystem, and thus the influence of the diffusion distance should be more significant for the in vivo conditions.
The association of CK with myosin via MyBPC1 is precisely regulated by both Cr and intracellular pH. The dose-dependence of the MM-CK–MyBPC1 interaction on Cr suggests the possibility of the Cr concentration being a sensor to monitor the energy status. When ATP and Cr are produced to exceed the consumption rate of myosin ATPase, MM-CK could catalyse the reverse reaction, so as to maintain a constant ATP level in muscle cells. Simultaneously, the excess Cr impedes the association of CK to the filament and further blocks the shuttling of high-energy phosphate donors and acceptors. This feedback regulation might be achieved by a conformational change induced by the binding of Cr. Actually, previous structural studies have indicated that the incorporation of the transition state analogue TSAC (Cr-nitrate-MgADP) induces considerable conformational changes involving the movement of two loops (residues 60–70 and 323–332) [41,42]. However, it is difficult to speculate the exact binding sites of MyBPC1 at CK by domain mapping since the structure and stability of the two domains of CK are closely correlated [43,44]. Similar to the binding of MM-CK with myosin through myomesin , the interaction between MM-CK and MyBPC1 was also dependent on pH. The pH range in which binding occurred is exactly within the intramuscular oscillation pH range [30,45], which varies between rest and activation. The binding is strong in moderate and slightly acidic pH, and is completely lost in basic pH. Possibly acidic pH is a signal to recruit CK, since rapid hydrolysis of ATP under the high energy-demanding condition produces ADP and H+ quickly, and thereby reduces the intracellular pH.
A working model of the functional association of myosin, MyBPC1 and MM-CK
In summary, in the present study we found that MM-CK could bind to the myosin filament through MyBPC1 dynamically during the resting and activation states of muscle, and a simplified model is proposed in Figure 6. In slow skeletal muscle cells, MyBPC1 resides in the C-zone of the sarcomere, where actin and myosin overlaps. During contraction, myosin ATPase catalyses ATP to form ADP and H+, which subsequently results in an acidic intracellular microenvironment. This physiological condition favours the cytosolic CK interaction with MyBPC1. By catalysing the reaction of H+, ADP and PCr to ATP and Cr, MM-CK could ensure a sufficient ATP supply for the ATPase under a high-energy turnover state. The utilization of H+ produced by ATPase also protects the muscle cells against extreme low pH conditions which may be harmful. After activation, the increase in intracellular pH, probably by the consumption of H+ during MM-CK catalysis, weakens the loading of MM-CK to the myosin. It is also possible that under extreme/pathological conditions, such as ischaemia, the depletion of PCr leads to an excess of Cr , which blocks the binding of MM-CK to the filament. The free MM-CK will catalyse the reverse reaction (PCr formation) to restore the excessive energy into the PCr pool under mild basic conditions. In such a dynamic way, MM-CK dissociates from the myosin filament at rest and re-associates at high workload to fulfil the different requirements of energy metabolism in muscle cells.
Zhe Chen, Yong-Bin Yan and Hai-Meng Zhou conceived the study. Zhe Chen, Tong-Jin Zhao, Yong-Bin Yan and Hai-Meng Zhou designed the experiments. Zhe Chen, TongJin Zhao, Yan-Song Gao and Fan-Guo Meng performed the in vitro experiments. Zhe Chen and Jie Li performed the in vivo experiments. Zhe Chen, Yong-Bin Yan and Hai-Meng Zhou analysed the data. Zhe Chen and Yong-Bin Yan wrote the paper.
This work was supported by the National Key Basic Research Project of China [grant number 2007CB914401]; the National Key Basic Research and Development (973) Program of China [grant numbers 2006CB503905, 2010CB912402]; the National Natural Science Foundation of China [grant number 30970635], and funds from Jiaxing, Zhejiang.
Abbreviations: CK, creatine kinase; Cr, creatine; Co-IP, co-immunoprecipitation; DTT, dithiothreitol; FHC, familial hypertrophic cardiomyopathy; FnIII, fibronectin III; GST, glutathione transferase; HA, haemagglutinin; HEK-293T, cell, human embryonic kidney-293 cell expressing the large T-antigen of simian virus 40; hMHC-1, human slow skeletal muscle MHC (myosin heavy chain)-1; IP, immunoprecipitation; LMM, light meromyosin; MESG, 7-methyl-6-thioguanosine; MHC, myosin heavy chain; MM-CK, muscle-type CK; MyBPC, myosin-binding protein C; PBST, PBS containing 0.1% Tween 20; PCr, phosphocreatine; PNPase, purine nucleoside phosphorylase
- © The Authors Journal compilation © 2011 Biochemical Society