PLOS ONE 2016; doi: 10.1371/journal pone.0162003

Tension Recovery following Ramp-Shaped Release in High-Ca and Low-Ca Rigor Muscle Fibers: Evidence for the Dynamic State of AMADP Myosin Heads in the Absence of ATP 

Haruo Sugi1, Maki Yamaguchi2, Tetsuo Ohno2, Takakazu Kobayashi3, Shigeru Chaen4, Hiroshi Okuyama2

 1Department of Physiology, Teikyo University School of Medicine, Tokyo, Japan, 2Department of Physiology, Jikei University School of Medicine, Tokyo, Japan, 3Department of Electronic Engineering, Shibaura Institute of Technology, Tokyo, Japan, 3Department of Integrated Science in Physicsand Biology, College of Humanities and Science, Nihon University, Tokyo, Japan

Correspondence should be addressed to Haruo Sugi, Department of Physiology, Teikyo University School of Medicine, Tokyo, Japan, E-mail:



Based on the ATPase reaction steps of actomyosin in solution, it is generally believed that myosin heads (M), in the form of MADPPi first binds with actin filaments (A), binds with actin filaments (A) to perform power stroke associated with reaction, AMADPPi → AM + ADP + Pi. In this scheme, AM is a high-affinity complex. Biochemical and electron microscopic studies on extracted protein samples, actin-binding sites on myosin heads are located at both sides of junctional peptide between 50K and 20K segments of myosin heavy chain. We have shown, however, that a monoclonal antibody (IgG) to the junctional peptide has no effect on Ca2+-activated muscle fiber contraction and in vitro actin-myosin sliding, although it covers the actin-binding sites in myosin heads. This indicates that, during muscle contraction, myosin heads do not pass through the static rigor AM configuration, determined using extracted protein samples. To determine the actual state of myosin heads at the end of power stroke in muscle, we studied mechanical responses of single skinned muscle fibers to ramp-shaped releases (0.5% of Lo, complete in 5ms) in high-Ca (pCa,4) and low-Ca (pCa,>9) rigor states. The fibers showed initial elastic tension drop, which was followed by small but definite tension recovery towards a steady level. The tension recovery persisted over many minutes in high-Ca rigor fibers, while it decreased quickly in low-Ca rigor fibers. The tension recovery was not affected by EDTA (10mM, with MgCl2 removed) in high-Ca rigor fibers, but was eliminated quickly in low-Ca rigor fibers. These results can be accounted for by assuming that, in high-Ca rigor fibers, AMADP myosin heads have long lifetimes and dynamic properties, which show up as the tension recovery following ramp-shaped releases, strongly suggesting that at the end of power stroke in muscle, myosin heads do not take the form of AM complex, determined using extracted protein samples.



Muscle is a machine to convert chemical energy of ATP hydrolysis into mechanical work. Muscle contraction results from cyclic attachment and detachment between myosin heads (M) extending from myosin filaments and actin filaments (A), coupled with ATP hydrolysis. Based on biochemical studies on ATPase reaction steps of actomyosin in solution, M first attaches to A in the form of MADPPi, and perform power stroke associated with reaction, AMADPPi → AM + ADP + Pi [1]. In this scheme, AM is a high-affinity complex, and generally believed to correspond to rigor AM linkage in rigor muscle after removal of external ATP [2]. Recently, however, we have found that a monoclonal antibody, which attaches to junctional peptide between 50K and 20K segments of myosin heavy chain [3], which covers actin-binding sites in myosin head [4], has no effect on Ca2+-activated muscle contraction [5]. This indicates that, contrary to general view, myosin heads do not pass through rigor AM configuration during muscle contraction.


To give information about the state of myosin heads immediately after completion of power stroke during muscle contraction, we compared the mechanical response of single skinned rabbit psoas muscle fibers to sudden decrease in fiber length between hi-Ca and low-Ca rigor states. The high-Ca rigor state was produced by first activating the fibers in contracting solution (pCa, 4) and then put them into high-Ca rigor solution (pCa, 4), from which ATP was removed. In this state, myosin heads was expected to form linkages with actin after completion of their last power stroke. The low-Ca rigor state was produced by transferring relaxed fibers from relaxing solution (pCa,>9) to low-Ca rigor solution (pCa,>9). Establishment of rigor state was confirmed by measuring in-phase and quadrature stiffness to reach steady values.


We applied ramp-shaped releases (0.5% of fiber slack length Lo, complete in 5ms) to rigor fibers without giving damage to them. Figure 1 A shows tension changes in the fiber when a series of release-restretch cycles are applied to high-Ca rigor fibers. It was found that the initial elastic drop in rigor tension was followed by small but distinct tension recovery towards steady level, which resembled quick tension recovery in intact frog muscle fibers in response to quick release [6], though the time scale was ~three orders of magnitude slower in the former than in the latter. During restretch of the fibers to the initial length (complete in >20ms), the tension rose during stretch and decayed exponentially after completion of stretch. The method of estimation of the amplitude of tension recovery is illustrated in Figure 1B. The tension immediately before release and immediately after release are defined as T0 and T1, respectively, while the steady level of tension attained during tension recovery is defined as T2. The amplitude of tension recovery Trec is expressed relative to T0, as Trec = (T2―T1)/To. Trec was maximum for the first release, applied 10―15s after transfer of the fibers to rigor solution, and gradually decreased with repeated application of release-restretch cycle. In 28 muscle fibers studied, generating maximum Ca2+-activated tension of 50―80kN/m2 (20oC), the average Trec value was 0.14 ± 0.05 (mean ± SD, n=25). The magnitude of initial elastic tension drop (T0―T1)/P0, serving as a measure of apparent rigor fiber stiffness, was smallest for the first release, ranging from 0.5 to 0.7.


The values of Trec and To, determined by repeated release-restretch cycle, decreased gradually with time, but Trec was observed over many minutes until To decreased to <10% of the initial value. The possibility that the long lasting occurrence of Trec is due to incomplete removal of ATP was excluded by the result that 10mM EDTA (with external MgCl2 removed) had no effect on Trec, since EDTA chelates Mg ions to eliminate remaining ATP, if any, to bind with myosin heads. The long lasting Trec may therefore be accounted for by assuming that AKADP myosin heads in high-Ca rigor fibers have long lifetimes to produce Trec.


Meanwhile, Trec was also observed in Low-Ca rigor muscle fibers, though it decreased more rapidly with repeated release-restretch cycle, and eventually disappeared after application of 5-10 release-restretch cycles (Figure 3A). If the fibers were previously kept in relaxing solution containing 10mM EDTA (with MgCl2removed), and subjected to release-restretch cycles, the rigor tension dropped from To to T1, and stayed at the level of T1 without exhibiting tension recovery (Figure 3B). It was noticed that the value of (To-T1)/To, serving as a measure of apparent fiber stiffness, was close to unity for the first release.


The definite difference in the mechanical properties between high-Ca and low-Ca rigor fibers can be explained in terms of different population of AMADP myosin heads. When the Ca2+-activated fibers are put into high-Ca rigor solution, a large proportion of myosin heads may be in the state of AMADP, which have long life-times in rigor fibers. The applied release may be mostly taken up by the distortion of AMADP myosin heads, and the elastic restoration of their configuration would show up as tension recovery following release, as illustrated in Figure 4. The applied stretch, on the other hand, would be taken up by elastic sarcomere structures due to nonlinear myosin head elasticity.


On the other hand, low-Ca rigor state is produced by removing ATP from relaxing solution (pCa, .9), in which actin-myosin interaction is inhibited by tropomyosin around actin filaments, so that myosin heads should override tropomyosin to form rigor linkages with actin. In low-Ca rigor fibers, most myosin heads may form rigor linkages with actin without performing power stroke coupled with ATP hydrolysis, and the proportion of AMADP myosin heads would be much smaller than in high-Ca rigor fibers. In the presence of EDTA, free Mg ions necessary for formation of AMADP myosin heads are quickly removed from low-Ca rigor solution. Consequently, low-Ca rigor fibers contain little or no AMADP myosin heads to result in disappearance of tension recovery following applied releases.


If this explanation is correct, the larger apparent stiffness of low-Ca rigor fibers in the presence of EDTA (Figure 3B) may result from higher stiffness of static AM myosin heads, which do not exhibit tension recovery and would gradually change to the AM rigor linkages with configuration similar to that determined using extracted protein samples [2]. This view is consistent with X-ray diffraction pattern from rigor muscles, which are prepared in low-Ca conditions and show marked mass transfer of myosin heads to actin [7]. 




Figure 1. (A) Tension responses of a single skinned muscle fiber in high-Ca rigor state to a series of release-restretch cycles. The fiber was subjected to a series of ramp-shaped release (0.5% of Lo, complete in 5ms) and restretch (complete in >20ms) cycles. Downward and Upward arrows indicate times of application of contracting and high-Ca rigor solutions, respectively. In this and Figure 2, downward and upward arrowheads below tension records indicate times of application of release and restretch, respectively. (B) Diagram illustrating method of estimating amplitude of tension recovery Trec. Upper and lower traces show length and tension changes, respectively. The fiber is first activated in contracting solution, and then transferred into high-Ca rigor solution. On application of release, intial elastic tension drop from To to T1 is followed by tension recovery from T1 to T2. The amplitude of tension recovery relative to To is expressed as Trec = (T2-T1)/To. 



Figure 2. Disappearance of tension recovery following ramp-shaped releases in a low-Ca rigor fiber in the presence of EDTA (10mM). Note that initial elastic drop in tension is not followed by tension recovery. 



Figure 3. Diagrams illustrating possible mechanism of tension recovery following ramp-shaped releases in high-Ca rigor muscle fibers. Left: A myosin head with ADP attached to its CAD forms rigor linkage with an action monomer (shaded) in actin filament. Arrow indicates direction of displacement on application of release. Right: Elastic distortion of a myosin head causing initial drop in rigor tension. Arrow indicates direction of limited elastic restoration of myosin head to show up as tension recovery.




1. Lymn RW, Taylor EW, “Mechanism of adenosine triphosphate hydrolysis by actomyosin” Biochemistry, vol.10, pp.4617―4624. 1971
2. Bagshaw CR, “Muscle Contraction” , London: Chapman & Hall, 1993
3. Sutoh K, Tokunaga M, Wakabayashi T, “Electron microscopic mapping of myosin head with site-directed antibodies”, Journal of Molecular Biology, vol. 206, pp.353―369. 1989
4. Levine BA, Moir AJG, Trayer IP, Wiliam RJP, “Nuclear magnetic resonance studies of calcium-modulated proteins and actin-myosin interaction” In: Squire JM editor, “Molecular Mechanism in Muscle Contraction”, Houndmills: Macmillan Press, pp.171―209. 1990
5. Sugi H, Chaen S, Kobayashi T, Abe T, Kimura K, Saeki Y, et al., “Definite differences between in vitro actin-myosin sliding and muscle contraction as revealed using antibodies to myosin head”, PLOS ONE e93272, 2014
6. Huxley AF, Simmons RM, “Proposed mechanism of force generation in striated muscle”, Nature, vol. 233, pp.533―538. 1971
7. Squire JM, “The Structural Basis of Muscle Contraction”, New York: Plenum, 1981