|
|
|
|
|
|
|
||
Original Article |
Anders.Arner{at}mphy.lu.se
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: myosin isoforms • phosphate • ATP • ADP • force-velocity relation
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) or into "phasic" and "tonic" types (Somlyo and Somlyo 1968) based on the characteristics of the contractile patterns of the intact smooth muscle. Most likely the membrane properties and excitation pathways are coordinated with the properties of the contractile machinery, and as pointed out by Drs. Somlyo and co-workers (Horiuti et al. 1989), the phasic and tonic muscle types have different turnover characteristics of the actin–myosin interaction. The difference in shortening velocity and rate of tension development between fast and slow smooth muscles is equal to, or even greater than, that between fast and slow skeletal muscle types (Malmqvist and Arner 1991). It is an open question whether smooth muscle can be divided into discrete groups or if the smooth muscle cells and tissues exhibit a continuous distribution in cross-bridge turnover kinetics. A division of smooth muscle into well-defined muscle types (e.g., based on kinetics of the actin–myosin interaction) would be important, but is not available at the present.
The molecular mechanisms responsible for the difference in cross-bridge kinetics between smooth muscles are not resolved. Alterations in the loop 1 at the 25/50-kD junction of the myosin II molecule alter the kinetics of myosin (Sweeney et al. 1998). Evidence from in vitro motility experiments shows that a seven amino acid insert in this region modulates the filament velocity and duty cycle of the actin–myosin interaction of smooth muscle (Kelley et al. 1992; Rovner et al. 1997; Lauzon et al. 1998). This effect has been difficult to clearly demonstrate in the organized contractile system of smooth muscle. Comparative studies on smooth muscle preparations have reported a correlation between contractile kinetics and expression of the myosin heavy chain insert as well as with the isoform distribution of the essential light chains of myosin (Malmqvist and Arner 1991; Fuglsang et al. 1993; Sjuve et al. 1996; Matthew et al. 1998). Forced expression experiments in chicken embryonic smooth muscle cells suggest that the essential light chains can modulate contraction kinetics (Huang et al. 1999). Although the myosin heavy chain insert, and/or light chain isoforms, influence the smooth muscle myosin kinetics, it is unknown which reactions in the cross-bridge cycle determine the kinetics of the organized contractile system in fast and slow smooth muscles.
The actin–myosin interaction in smooth muscle is considered to occur according to the general kinetic scheme proposed for skeletal muscle, although several of the rate constants are slower in smooth muscle (Marston and Taylor 1980; Cremo and Geeves 1998). One important property of smooth muscle myosin is the high affinity of ADP for the actin–myosin complex in vitro (Cremo and Geeves 1998). Recently, ADP binding to the subfragment 1 of smooth muscle myosin has been shown to induce structural changes in the myosin molecule (Whittaker et al. 1995; Gollub et al. 1996), and kinetic and structural data from smooth muscle myosin suggest the presence of two different actin-myosin-ADP states on the pathway to the ADP release (Rosenfeld et al. 2000). The low, micromolar, dissociation constant for ADP in smooth muscle was first demonstrated in smooth muscle fibers in rigor using competition with pyrophosphate or ATP (Arheden and Arner 1992; Nishiye et al. 1993). In a study of smooth muscle preparations in rigor, ADP binding did not alter force (Dantzig et al. 1999), suggesting that the structural changes in the isolated myosin molecule upon ADP binding are not readily detected in the organized contractile system. However, a recent report has suggested that mechanical changes can be induced in rigor (Khromov et al. 2001).
The finding that ADP inhibits the rate of relaxation from active contractions (Khromov et al., 1995) suggests that ADP binds strongly to cross-bridge states present during active cross-bridge cycling and that a high ADP affinity may have a functional role in modulating mechanics of active contractions. The ADP release/binding reactions are generally considered to occur at the end of, or after, the cross-bridge working stroke (Cooke and Pate 1985), and the release of ADP is considered to be rate-limiting for shortening velocity (Siemankowski et al. 1985). Thus, it is possible that the tight ADP binding of the actin–myosin complex has a key role in determining contractile kinetics of smooth muscle.
We have previously shown that ADP inhibits shortening velocity in Ca2+-activated skinned smooth muscle (Arner et al. 1987), but quantitative data regarding ADP effects on the force-velocity relation in a fully activated smooth muscle have not been presented. Differences in contractile kinetics between fast (phasic) and slow (tonic) smooth muscles have been suggested to be due to differences in ADP binding (Fuglsang et al. 1993; Khromov et al. 1996). However, these studies have compared muscles in rigor and in the dephosphorylated state during relaxation. To answer the question whether reactions involved in binding or release of ADP, or possibly of ATP or phosphate, are responsible for the difference in shortening velocity between fast and slow smooth muscles, we have performed a detailed study of the interaction between actin and myosin in the organized contractile system by determining force-velocity relationships in fully thiophosphorylated preparations of a fast (taenia coli) and a slow (aorta) smooth muscle from guinea pigs at different concentrations of ADP, ATP, and phosphate. Preliminary results of this work have been presented previously (Malmqvist et al. 1999).
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
400 g and were killed by cervical dislocation. The muscle tissues were cut out under a microscope and chemically skinned using Triton X-100 as described by Arner and Hellstrand 1985. The preparations were then stored at –18°C until they were used. Immediately before each experiment, fiber preparations were cut out. The preparations from the taenia coli had a diameter of 0.15 mm and a length of 2–4 mm. A 1.2–1.8-mm-long segment of the aorta was cut open and mounted with the circular muscle layer in the long axis of the preparation, which gave preparations of 2–3-mm length with the full thickness of the media layer (0.05 mm).
Solutions
The solutions used for the skinned muscle preparations contained 30 mM N-Tris-(hydroxymethyl)methyl-2-aminoethane-sulfonic acid, 4 mM EGTA, and 2 mM free Mg2+. All solutions were adjusted to pH 6.9 with KOH and to an ionic strength of 150 mM using KCl. In the experiments, the concentrations of MgATP, MgADP, and inorganic phosphate (Pi) were varied. When ADP-depletion/ATP-generation was used, 12 mM phosphocreatine (PCr) and 0.5 mg/ml creatine kinase (CK) were added to solutions or, in the ATPase-determining experiments, 10 mM phosphoenol pyruvate and 20 U/ml pyruvate kinase. Experiments using ADP were performed in the presence of 0.2 mM of the myokinase inhibitor AP5A (Feldhaus et al. 1975). The standard relaxation and activation solutions contained 3.2 mM MgATP. The composition of all solutions was calculated using a computer program and stability constants essentially as described by Fabiato and Fabiato 1979 and Fabiato 1981. All experiments were performed at room temperature (22°C). All chemicals were purchased from Sigma-Aldrich and Boehringer. Calmodulin was a gift from Dr. E. Thulin (Department of Physical Chemistry II, Lund University).
Quick-release experiments
The muscle preparations were wrapped with small clips of aluminum foil at each end and mounted between a stainless steel pin attached to a force transducer (model AE 801; SensoNor a.s.) and an isotonic lever (Arner and Hellstrand 1985). The lever arm could be released with electromagnetic relays. The afterload on the muscle was determined by a spring load on the lever. After release, a stable afterload was established within 5 ms. After each quick release, force and muscle length were recorded for 1 s using a sampling rate of 1 kHz on an RTI-800 Analogue Devices A/D board in a personal computer. The shortening velocity was determined at different points in time after the release by analysis of the length responses as described by Arner 1982. The length of the muscle was measured at the end of the experiment using a microscope with an ocular scale and the velocity values were expressed in muscle lengths (ML) per second. The following Hill (Hill 1938) equation was fitted to the force and velocity (V) data:
 | (1) |
In 
, a and b are constants, P the afterload, and Po the isometric force at each contraction. The maximal shortening velocity (Vmax) was then given by extrapolation of the fitted curve to P/Po = 0.
The preparations were initially mounted in a relaxing solution (1 nM free [Ca2+], pCa 9) and stretched to a passive force of 0.1 mN. Thereafter, the muscles were maximally activated using a repeated thiophosphorylation procedure (Arheden et al. 1988). Thiophoshorylaton of the regulatory light chains was done by treating the muscle for 10–15 min using a calcium-containing (pCa 4.5) rigor solution with 0.5 µM calmodulin and 1.8 mM ATP-
-S. After a 5-min period in calcium-free rigor solution, a contraction was initiated by transfer to an ATP containing solution. When force had reached a plateau, the muscle fiber was transferred to fresh solution and a series of 15–30 releases to different afterloads was performed. The isometric force was measured immediately before the beginning of each release series. Before the next contraction and force-velocity determination, the fiber was again treated with thiophosphorylation. A maximum of five (taenia coli) and six (aorta) contractions and force-velocity determinations were made on each preparation. In general, force and velocity data were normalized to values obtained in each fiber preparation during a standard reference contraction at saturating (3.2 mM) [MgATP] in the presence of PCr and CK. The different solutions were applied at random order. Six sets of experiments were performed: (1) varied [MgATP] in the presence of the PCr/CK system; (2) varied [MgATP] in the absence of the PCr/CK system; (3) varied [MgATP] in the presence of MgADP in the absence of the PCr/CK system (taenia coli only); (4) varied [MgADP] at constant [MgATP] (taenia coli: 1 and 6 mM; aorta: 10 mM); (5) 0 and 20 mM Pi at 3.2 mM MgATP in the presence of the PCr/CK system (aorta only); and (6) in rigor solutions (0 mM MgATP with 50 U/ml hexokinase and 10 mM glucose). In some of these experiments, apyrase (20 mg/ml) was included in rigor solutions.
For analysis of the [MgATP] dependence of the maximal shortening velocity (Vmax), a hyperbolic equation was used to determine the apparent binding constant (Km). Max denotes the Vmax at saturating [MgATP].
 | (2) |
For analysis of [MgADP] dependent inhibition of velocity, the following equation was used to determine the apparent inhibition constant (Ki):
 | (3) |
Isometric Force Experiments
In one series of experiments, the influence of varied Pi concentrations on active force was determined in the taenia coli and aorta. The preparations were mounted in 0.5- ml plastic baths for isometric force registration using AE 801 force transducers. Activation with thiophosphorylation was performed as described above. Four preparations were studied in parallel and force was measured at Pi concentrations in the range 0–40 mM in solutions with 3.2 mM [MgATP] and PCr/CK system. Pi was introduced in the contraction solution and force was determined at the plateau of each active contraction.
Determination of Tissue ATPase Activity
In a series of experiments performed to examine ATPase activity, skinned muscle fibers from the taenia coli and the aorta were mounted in 0.5-ml cups and attached to Grass FT03 force transducers. The experiments were performed essentially as described by Arner and Hellstrand 1985. The fibers were incubated in contraction and relaxation solutions containing phosphoenolpyruvate and pyruvate kinase for 15 min. Incubation solutions and blanks were frozen. The release of pyruvate, which is proportional to the production of ADP, was determined using a fluorimetric assay. The values from the incubation solutions and the blanks were compared with a standard curve with known concentrations of ADP. ATPase activity was related to the wet weight of the preparations.
Determination of Myosin Isoforms
To examine the content of myosin essential light chain 17a and 17b in the skinned taenia coli and the aorta fibers (compare with Malmqvist and Arner 1991), the proteins were separated according to isoelectric point (first dimension), using ampholytes in the pH range 4.5–6. For the second dimension, the proteins were separated according to molecular mass using 15% polyacrylamide gels. The gels were stained with Coomassie blue and evaluated with densitometric scanning.
Statistics
All values are given as mean ± SEM with the number of observations within parenthesis. Curve fitting to the hyperbolic Hill 1938 force-velocity relation was performed using a nonlinear procedure to minimize the perpendicular distance between data points and the fitted curve (Fletcher and Powell 1963) as described previously (Arner 1982, Arner 1983). All other curve fitting was performed using the nonlinear fitting routines implemented in Sigmaplot for Windows (Jandel Corp.).
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) that the protocol with repeated thiophosphorylation gives reproducible contractions with <5% difference in maximal shortening velocity (Vmax) and force between the first and seventh contraction in the taenia coli muscle. In the present study, we evaluated whether force and velocity were changed with the number of contractions for the aorta and found that Vmax and force at optimal [MgATP] for the last contraction (No. 3 or 5) in the series were essentially similar to the initial values (Vmax = 95 ± 4; force = 112 ± 5% of initial value, n = 6).
|
equation () as shown in Fig. 2. As seen in Fig. 2 B the force-velocity relation of the aorta was less concave compared with that of the taenia coli (Fig. 2 A) especially at low MgATP concentrations, which corresponds to high a/Po and low b values in the Hill 1938 equation. The variation in the parameters a/Po and b was large, and their interpretation in smooth muscle is unclear. Therefore, we did not analyze these parameters of the Hill equation. The fitted Hill equation deviated very little from the data points at lower relative afterloads, and we used the equation to obtain estimates of the Vmax.
|
; Arner and Hellstrand 1985). In accordance with our previous analysis (Arner and Hellstrand 1985), we used the velocity values determined 100 ms after release. Velocity values presented here refer to this time point, unless stated otherwise.
We have previously reported (Malmqvist and Arner 1991) that the guinea pig taenia coli has 20% of the basic essential myosin light chain (LC17b). This was confirmed in the present study (25% LC17b). Previous studies have reported 60% LC17b in rat and rabbit aorta (Malmqvist and Arner 1991) and in rabbit femoral artery (Fuglsang et al. 1993). In the present study, we found 70% LC17b in the guinea pig aorta, which is consistent with the previous data from large elastic arteries.
We observed, when we determined the [MgATP] dependence of Vmax that the relation did not extrapolate to zero Vmax at zero [MgATP]. Therefore, we also performed quick-release experiments in rigor and found a significant shortening with an apparent Vmax under these conditions. The Vmax in rigor at 100 ms after release, relative to that at optimal [MgATP], was 39 ± 1.2% (n = 14) for the taenia coli and 55 ± 3.5% (n = 7) for the aorta. Quick-release experiments on preparations fixed with glutaraldehyde at the end of the experiments showed no shortening response, excluding that the shortening observed in rigor muscles was due to compliance of transducer, lever arm, or fiber attachment. A significant shortening response in rigor muscles was also observed in preparations where the ends had been fixed with cellulose acetate glue and using the slack test method. This shows that the response in rigor was not due to the aluminum clips or to the isotonic quick-release method. To ensure that the shortening in rigor was not due to ATP contamination in the solutions the rigor experiments were performed in solutions supplemented with glucose/hexokinase and apyrase as described in MATERIALS AND METHODS. To further ensure that shortening in rigor was not due to active cross-bridge cycling, we performed experiments in the presence of 1 mM vanadate, an inhibitor of active cross-bridge cycling in smooth muscle (Jaworowski et al. 1999). The force and Vmax in rigor were unchanged when vanadate was added in the aorta (n = 2) and taenia coli (n = 2). Addition of 5.32 mM MgADP to rigor did not alter the maximal shortening velocity in rigor (n = 2) in the aorta.
We have interpreted the shortening in rigor conditions as a result of viscous phenomena in the preparation. Since passive force in the relaxed state (pCa 9 solution) was low (taenia coli: 0.7 ± 0.5; aorta: 4.3 ± 1% of maximal active tension, n = 10), elastic or viscous elements in parallel with the contractile apparatus would not contribute to force or shortening. Therefore, we assumed a visco-elastic element in series with the contractile component. The shortening responses in rigor after a quick release consisted of an initial elastic recoil, and a subsequent "viscous" phase of slower shortening. The calculated maximal shortening velocity in rigor decreased with time in an approximately similar manner as in the active contraction (Vmax at 500 ms after release relative to that at 100 ms, taenia coli rigor: 20.1 ± 1.5; aorta rigor: 20.6 ± 3.2; taenia coli active: 31.9 ± 0.7; aorta active: 23.6 ± 1.7%, n = 6). Thus, the time constant of the visco-elastic element is approximately the same as that of the active response. Therefore, the visco-elastic component cannot be simply eliminated by measuring velocity at a different time after release. The amplitude of the viscous shortening phase in rigor, which reflects the properties of the spring in the visco-elastic component, was not linearly dependent on the amplitude of the force step. The resulting strain-force relationship was nonlinear, with an increasing stiffness at increasing strain. The behavior could be adequately described by an exponential spring similar to that proposed for series and parallel elastic components (Arner and Hellstrand 1985). A simple visco-elastic element, composed of a viscous element and a linear spring, would have a linear dependence of velocity on force. In a visco-elastic element containing an exponential spring, the dependence of velocity on force becomes complex. Using an analytical solution to the differential equation describing a model with an element in series with the contractile component composed of a linear viscous element (force is proportional to velocity) and a nonlinear (exponential) spring in parallel (Voigt configuration), we simulated the dependence of maximal velocity on the isometric tension before release. The model predicts that the relation between maximal shortening velocity and isometric tension is nonlinear; velocity determined at higher isometric tension levels has a very weak dependence on force. The rigor isometric force in our experiments was in the range 28–55% (taenia coli) and 60–78% (aorta) of maximal active isometric force (at 3.2 mM MgATP). Linear regression of the data gave a very weak dependence of Vmax on force (taenia coli: r2 = 0.26 and aorta: r2 = 0.02) with curves that intercepted with the velocity-axis clearly above zero velocity. These data are consistent with the predictions of the model discussed above. Based on these considerations, an estimate of the Vmax of the contractile component can be obtained by subtracting the rigor Vmax, without correction for the isometric force level. Using this correction the [MgATP] dependence of Vmax extrapolated to zero velocity at zero [MgATP] and was adequately described by a hyperbolic equation (), as discussed below (Fig. 3). After correction, the Vmax of the taenia coli was 0.11 ML/s and for the aorta 0.015 ML/s. This makes the difference in Vmax between the two muscle types slightly greater than evident in Fig. 2. In the following data presentation, we have subtracted the rigor velocity from the Vmax values.
|
) was about fourfold higher. A fit to the whole data set gave Km values of 351 ± 76 µM for the taenia coli and 84 ± 31 µM for the aorta. Fig. 3 B shows the corresponding force values. Force was lower at reduced [MgATP], but appeared to be less influenced by a change in [MgATP] than Vmax. Force in rigor was 0.69 ± 0.03 (6) for the aorta and 0.42 ± 0.02 (14) for the taenia coli, relative to the corresponding values at optimal (3.2 mM) [MgATP]. In separate experiments, we determined active force normalized to cross-sectional area (determined by dividing preparation wet weight with length and density) and found that the force of the aorta preparations at maximal activation and optimal [MgATP] was lower than that of the taenia coli (4.5 ± 0.8 (6) vs. 59 ± 19 (4) mN/mm2). It should be noted that these values were not corrected for tissue content of smooth muscle myosin. The absolute forces that the preparations developed in these experiments were 6.4 ± 0.8 (4) mN (taenia coli) and 1.7 ± 0.3 (6) mN (aorta).
As shown in Fig. 4 A, the Vmax of the skinned muscles was dependent on the presence of the phosphocreatine/creatine kinase system (PCr/CK). Removal of the backup system resulted in a significant reduction of the maximal shortening velocity at lower ATP concentrations. This effect was more pronounced in the aorta compared with the taenia, showing that lowered ATP/ADP ratios in the muscle fiber influence velocity more in the slow smooth muscle. The apparent Km for MgATP increased 1.8-fold in the taenia coli and 34-fold in the aorta.
|
) the approximate [MgADP] and [MgATP] in the center of the taenia coli preparations could be calculated from preparation dimensions and ATPase (Cooke and Pate 1985). At 3.2 mM MgATP in the bathing medium, [MgATP] and [MgADP] in the center on the taenia coli were calculated to 2.6 and 0.6 mM, respectively. Since the aorta preparation was thinner and had a lower ATPase activity, the concentrations in the center of the preparation would be similar to those in the bathing medium (3.2 and 0.04 mM for [MgATP] and [MgADP], respectively) when calculated as above. These calculations show that the differences in dependence of the backup system between the aorta and the taenia coli preparations are not due to differences in tissue ATPase or ADP/ATP diffusion, but rather reflect a difference in ATP binding or sensitivity to ADP as discussed below.
To investigate the effects of ADP on the ATP dependence of Vmax, force-velocity relations were determined at different ATP concentrations in the presence of 2.66 and 5.32 mM MgADP in the taenia coli preparation (Fig. 5). The ATP dependence of Vmax (Fig. 5 A) was shifted towards higher [MgATP] in the presence of ADP. Addition of 2.66 mM [MgADP] did influence the extrapolated Vmax at saturating [MgATP] to a minor extent and the apparent Km for MgATP increased to 0.811 mM (fit to ). The inhibition of Vmax at the high MgADP concentrations (5.32 mM) could not be reversed by increased [MgATP], suggesting noncompetitive effects of ADP at higher concentrations. Increasing [MgADP] resulted in a decrease in active force (Fig. 5 B).
|
50% of maximal. This behavior was observed both in the aorta and the taenia coli (Fig. 6). Even at lower [MgATP] (1 mM, data not shown), addition of ADP did not reduce velocity to zero in the taenia coli; Vmax was inhibited at saturating [MgADP] to 50% of the value at zero MgADP. The data for the whole range of ADP concentrations in Fig. 6 could not be directly described using simple Michaelis-Menten's kinetics since velocity was not inhibited to zero at saturating ADP. If we only use the initial part of the data at nonsaturating ADP concentrations (where Lineweaver-Burke plots were linear) fitting to gave apparent Ki values for ADP of 10 µM in the aorta and 360 µM in the taenia coli. Although the Ki values at present cannot be directly interpreted, the analysis show that a pronounced difference in the ADP binding exists between the two muscles. In the aorta, force was slightly inhibited in the highest MgADP concentration interval, but was essentially unchanged in the range of MgADP concentrations where the inhibition of velocity occurred. Thus, velocity could be decreased by ADP by 30%, at essentially unchanged force in the aorta.
|
15%, velocity was slightly, but not significantly, increased. Maximal shortening velocities of aorta preparations at 3.2 mM MgATP and PCr/CK were with 20 mM Pi 0.017 ± 0.002 ML/s and without Pi 0.014 ± 0.001 ML/s.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
; White et al. 1993), whereas the taenia coli belongs to the fast smooth muscles (Malmqvist and Arner 1991). The difference in maximal shortening velocity (Vmax) is about sevenfold (Fig. 3), which is greater than the difference in shortening velocity between fast- and slow-twitch skeletal muscles (e.g., rabbit psoas versus soleus; Pate et al. 1992). Our quick-release experiments on muscles in rigor revealed an important technical aspect of experiments on smooth muscle preparations, i.e., a significant contribution of viscous elements on the shortening responses. A more detailed analysis revealed that the viscous responses in rigor could be described by a series-coupled element with a nonlinear spring and a viscous element. This model showed a weak dependence of viscous maximal velocity on isometric tension in rigor and, therefore, we subtracted a constant velocity from the active contractions. After this correction, the dependence of Vmax on [MgATP] could be adequately described by a hyperbolic equation. At present, we do not have any data regarding the nature of the viscous component, it might reside outside of the contractile machinery, in the cytoskeleton, in the cell–cell interactions, or be a part of the cross-bridge interaction.
The phosphate (Pi) release is considered to be associated with force generation in skeletal and smooth muscles (Hibberd et al. 1985; Österman and Arner 1995). The phosphate release is not rate-limiting for the maximal shortening velocity in the aorta (present study) and the taenia coli (Österman and Arner 1995), suggesting that the actin-myosin-ADP (A-M-ADP) states binding Pi do not resist shortening and are thus not responsible for the slower velocity in the aorta. Also, Vmax is not simply a function of force, since a reduction in isometric force (by Pi), does not reduce Vmax. Thus, the difference in cross-bridge kinetics between smooth muscle tissues does not only involve alterations in reactions rate-limiting for shortening velocity, but also in the Pi-release reactions associated with force generation.
The apparent Km for the ATP effects on Vmax in the slow aorta muscle was about fourfold lower than that of the fast taenia coli smooth muscle, with apparent Km values of 84 and 351 µM, respectively. This finding is consistent with results from slow and fast rabbit skeletal muscles, where the Km for MgATP was lower in the slow muscle (semimembranous, 18 µM; psoas, 150 µM; Pate et al. 1992). If it is assumed that ADP and ATP binds to actin-myosin (A-M) states generated after the power stroke and before detachment, increased ADP and decreased ATP would increase the population of A-M and A-M-ADP states. These states would resist shortening until detached by ADP release and binding of ATP or by direct dissociation by the filament sliding (Cooke and Pate 1985). If the rate-limiting step for shortening velocity occurs before the ATP-induced detachment, ATP has to be reduced to a lower concentration in a slow muscle to influence shortening velocity. Since the apparent Km was lower in the slow aorta compared with the taenia coli (Fig. 3, present study) we can thus exclude that differences in the ATP-induced detachment reaction are responsible for the difference in Vmax between the fast and slow smooth muscles.
Actually a comparison of the ratios Vmax/Km between muscles gives information on the difference in apparent second-order rate constant for ATP-induced dissociation, assuming similar sarcomere equivalent lengths and cross-bridge attachment ranges. This ratio was 50% lower in the aorta, which could be consistent with a slightly lower dissociation constant for ATP. The second- order rate constant for the ATP-induced dissociation from rigor has been suggested to be lower in smooth muscle fiber preparations compared with skeletal muscle (Somlyo et al. 1988), and has been reported to be about threefold lower in tonic compared with phasic smooth muscles (Khromov et al. 1996). Our data are not inconsistent with a difference in ATP binding, but show that a difference in the ATP-dependent dissociation cannot explain the difference in shortening velocity between fast and slow smooth muscles at least in the physiological range of ATP concentrations.
Previous studies on skinned smooth muscles in rigor have shown a strong MgADP binding, with a Kd of 1 µM (Arheden and Arner 1992; Nishiye et al. 1993), compared with a Kd of 60 µM in skeletal muscle (Schoenberg and Eisenberg 1987). Although a strong ADP binding has been demonstrated in dephosphorylated smooth muscles during relaxation from active contractions (Khromov et al., 1995) quantitative data regarding ADP and ATP binding to phosphorylated myosin during cross-bridge cycling have been lacking. Experiments regarding ADP effects in muscle fiber preparations during active contraction are complicated by the fact that ADP-depleting/ATP-generating backup systems cannot be used. The ATPase activity of the smooth muscle is low, our preparations were made small, and we used comparatively high [MgADP] and [MgATP]. Therefore, we predict that effects of gradients are minimal and have used the bath concentrations of the substrates and products in our analysis. The comparatively high MgADP dependence of the slow aorta muscle was not due to gradients created by diffusion or tissue ATPase, since the ATPase activity was lower and dimensions of the aorta preparations were smaller. When we performed experiments on the aorta preparations, removal of the PCr/CK backup system gave a prompt reduction in Vmax. Thus, this finding suggests that the shortening of the slow, more economical, aorta is highly dependent on the backup system. Since our data regarding [MgATP] variations show that the apparent Km for ATP is lower in the slow muscle, the influence of the backup system cannot be explained by ATP depletion in the muscle tissue, but rather by ADP accumulation. Since the ATPase activity of the aorta is lower than that of the taenia coli, the effect of the backup system most likely reflects a significant difference in the effects of [MgADP] on the shortening velocity between the slow and fast smooth muscles.
Even though the inhibition of Vmax by MgADP occurred at low concentrations, the velocity was only inhibited to 50% of maximal at saturating [MgADP] in the presence of MgATP (Fig. 6). This behavior is clearly different from the inhibition of velocity when ATP was reduced, where Vmax approached zero (i.e., the apparent Vmax in rigor). We assume that addition of ADP generates a population of A-M-ADP states that oppose shortening. It could be possible that the situation at saturating [MgADP] is a rigorlike state with altered viscous properties giving an apparent velocity that is higher than that in rigor. This seems very unlikely since we found that addition of MgADP to rigor did not increase velocity. A second possible explanation could be that the MgADP binding is weaker at higher MgADP concentrations, a situation where velocity is decreased. This finding is difficult to explain since a lower velocity would shift the distribution of cross-bridge strain towards lower strain in the negative direction, which according to general models of muscle contraction would increase the binding affinity of MgADP (Pate and Cooke 1989). A more complex model would be to consider cooperative phenomena between the myosin heads. It has been recently proposed for skeletal muscle that the binding of the two myosin heads to actin occurs in a coordinated cooperative manner (Conibear and Geeves 1998). Our finding from the smooth muscle that ADP can only inhibit velocity to 50% might be consistent with a model where only one of the two myosin heads can initiate and perform the power stroke reactions, and where a subsequent attachment of the second head promotes ADP release of the leading head.
The interpretation of the Ki values is not straight forward since we could not completely inhibit shortening velocity and we cannot at this stage present a complete model to analyze the behavior. However, if we analyze the initial part of the ADP inhibition data (Fig. 6) we find an apparent Ki value of 10 µM in the aorta and 360 µM for the taenia coli. This suggests that the binding of ADP to cycling cross-bridges differs with a factor of 40 between the slow and fast smooth muscle types. This result is consistent with the finding of a fourfold difference in binding of ADP to rigor cross-bridges between a fast/phasic (Kd = 4.9 µM rabbit bladder) and a slow/tonic (Kd = 1.1 µM rabbit femoral artery) smooth muscle (Fuglsang et al. 1993). The larger difference in our Ki values could reflect a difference in the distribution of strain between the fast (taenia coli) and slow (aorta) smooth muscles during active shortening.
In the in vitro motility assay, the velocity of actin over smooth muscle myosin is influenced by ATP and ADP concentrations. It has been shown that the Km for ATP is 40 µM and the Ki for ADP is 0.24 mM using turkey gizzard myosin at 30°C (Warshaw et al. 1991). Addition of phosphate at high [MgATP] did not influence the velocity. These data are generally consistent with our observations in smooth muscle fibers. In comparison, we find a slightly higher Km for MgATP (351 µM) and a higher apparent Ki for ADP (360 µM) for the guinea pig taenia coli muscle fibers at 22°C. It should be recognized that the interaction between actin and myosin in the in vitro motility assay does not mimic all aspects of the actin–myosin interaction in the muscle fiber. In vitro motility assay studies have shown that an insert in the myosin heavy chain influences velocity (Kelley et al. 1992; Rovner et al. 1997; Sweeney et al. 1998) and duty cycle (Lauzon et al. 1998). The velocity variation is about twofold or less in these studies. This is in contrast with the results from smooth muscle fibers where the variation in velocity is more than fivefold (Malmqvist and Arner 1991; present study). Another difference between the in vitro motility assay and studies of velocity is that the ADP dependence of velocity in in vitro motility assay was similar with and without insert (Rovner et al. 1997), whereas the ADP dependence differs markedly between fast and slow muscles both under isometric and isotonic conditions (Fuglsang et al. 1993; Khromov et al. 1996; present study). Thus, it is likely that the structure and organization of the contractile filaments in the smooth muscle fiber have a major influence of kinetics of the cross-bridge cycle, possibly by altering strain dependence of reactions in the cross-bridge cycle.
In the intact smooth muscle, intracellular [ADP] has been shown to be in the submillimolar range in the relaxed state and to increase during active contraction and metabolic inhibition (Hellstrand and Paul 1983; Krisanda and Paul 1983; Hellstrand and Vogel 1985; Fisher and Dillon 1988). In the aorta, we find an apparent Ki for ADP and shortening velocity in the micromolar range. Interestingly, the effect of ADP on velocity in the aorta muscle occurred at almost unchanged force (Fig. 6), a phenomenon similar to the "latch" behavior observed in intact muscle. ADP in the vicinity of the contractile proteins, thus, might have a role in modulating the cross-bridge turnover primarily in the slow and economical smooth muscle types. In the living muscle, force can be supported by unphosphorylated cross-bridges, and it should be noted that our experiments were performed in maximally phosphorylated muscles. However, biochemical data (Greene and Sellers 1987) and studies on smooth muscle in rigor (Arheden and Arner 1992) do not suggest a large effect of myosin light chain phosphorylation on the ADP binding to the A-M state. The slow muscle is less affected by increased inorganic phosphate. Although the effects of phosphate occur at comparatively high concentrations, the lower phosphate sensitivity could be another mechanism, in addition to ADP binding, for the slow muscle to maintain tone during sustained contractions and in situations with impaired energy supply.
|
ACKNOWLEDGMENTS |
|---|
This work was supported by grants from the Swedish Medical Research Council (No. 04X-12584 to U. Malmqvist and No. 04X-8268 to A. Arner) and the Medical Faculty Lund University.
Submitted: 4 December 2000
Revised: 6 March 2001
Accepted: 27 March 2001
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Facebook 
Reddit 
Technorati 
Twitter What's this?
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
