Biochemical and Biophysical Research Communications 405 (2011) 651–656

Contents lists available at ScienceDirect

Biochemical and Biophysical Research Communications

journal homepage: www.elsevier .com/locate /ybbrc

Electron microscopic evidence for the myosin head lever arm mechanism inhydrated myosin filaments using the gas environmental chamber

Hiroki Minoda a,b, Tatsuhiro Okabe a, Yuhri Inayoshi a, Takuya Miyakawa c, Yumiko Miyauchi c,Masaru Tanokura c, Eisaku Katayama d, Takeyuki Wakabayashi e, Tsuyoshi Akimoto f, Haruo Sugi f,⇑a Department of Applied Physics, Tokyo University of Agriculture and Technology, Koganeishi, Tokyo184-8588, Japanb CREST, Japan Science and Technology Agency, Kawaguchi, Saitama 332-0012, Japanc Department of Applied Biological Chemistry, Graduate School of Agricultural and Life Sciences, University of Tokyo, Bunkyo-ku, Tokyo 113-0032, Japand Graduate School of Medicine, Institute of Medical Science, University of Tokyo, Minato-ku, Tokyo 108-8639, Japane Department of Biosciences, School of Science and Engineering, Teikyo University, Utsunomiya, Tochigiken 320-8551, Japanf Department of Physiology, School of Medicine, Teikyo University, Itabashi-ku, Tokyo 173-8605, Japan

a r t i c l e i n f o a b s t r a c t

Article history:Received 20 January 2011Available online 31 January 2011

Keywords:Muscle contractionMyosin filamentMyosin headTransmission electron microscopeGas environmental chamber

0006-291X/$ - see front matter � 2011 Elsevier Inc. Adoi:10.1016/j.bbrc.2011.01.087

⇑ Corresponding author. Address: 3-6-22 Sakae, NiJapan. Fax: +81 484 78 4079.

E-mail address: sugi@kyf.biglobe.ne.jp (H. Sugi).

Muscle contraction results from an attachment–detachment cycle between the myosin heads extendingfrom myosin filaments and the sites on actin filaments. The myosin head first attaches to actin togetherwith the products of ATP hydrolysis, performs a power stroke associated with release of hydrolysis prod-ucts, and detaches from actin upon binding with new ATP. The detached myosin head then hydrolysesATP, and performs a recovery stroke to restore its initial position. The strokes have been suggested toresult from rotation of the lever arm domain around the converter domain, while the catalytic domainremains rigid. To ascertain the validity of the lever arm hypothesis in muscle, we recorded ATP-inducedmovement at different regions within individual myosin heads in hydrated myosin filaments, using thegas environmental chamber attached to the electron microscope. The myosin head were position-markedwith gold particles using three different site-directed antibodies. The amplitude of ATP-induced move-ment at the actin binding site in the catalytic domain was similar to that at the boundary between thecatalytic and converter domains, but was definitely larger than that at the regulatory light chain in thelever arm domain. These results are consistent with the myosin head lever arm mechanism in musclecontraction if some assumptions are made.

� 2011 Elsevier Inc. All rights reserved.

1. Introduction

More than 50 years have passed since the monumental discov-ery that muscle contraction results from relative sliding betweenactin and myosin filaments, coupled with ATP hydrolysis [1,2].The myofilament sliding is caused by the attachment–detachmentcycle between the myosin head extending from myosin filamentsand the sites on actin filaments. The myosin head (M), in the formof M�ADP�Pi first attaches to actin (A), and exerts a power stroke,associated with release of Pi and ADP, so that at the end of thepower stroke, M forms rigor-like linkage with A. Upon binding witha new ATP, M detaches from A to exert a recovery stroke, associatedwith the reaction, M�ATP ? M�ADP�Pi, and reattaches to actin [3].

The myosin head (or myosin subfragment 1) consists of the cat-alytic domain, which contains the actin binding and ATPase sites,and the lever arm domain, which connects the myosin head to

ll rights reserved.

izashi, Saitamaken 352-0014,

myosin filament via myosin subfragment 2 (myosin S 2). The twodomains are connected by the converter domain. It has beensuggested mainly from crystallographic studies on the nucleotidedependent structural changes of the myosin head detached frommyosin filaments, that the myosin head power stroke is producedby active rotation of the lever arm domain around the converterdomain, while the catalytic domain remains rigid [4–7]. However,it is not clear whether the myofilament sliding is actually causedby the lever arm mechanism in muscle, where the myosin headsare not detached from, but are firmly connected to, myosinfilaments.

To elucidate the above problem, a most straightforward way isto directly record ATP-induced movement at different regionswithin individual myosin heads extending from hydrated myosinfilaments, using the gas environmental chamber (EC or hydrationchamber), which enables us to record dynamic conformationalchanges in hydrated biological macromolecules, retaining theirfunction in aqueous solution in the electron microscope [8].With this novel method, Sugi et al. [9,10] have succeeded inrecording the ATP-induced myosin head movement in the

652 H. Minoda et al. / Biochemical and Biophysical Research Communications 405 (2011) 651–656

myosin–paramyosin core complex, and in synthetic myosin fila-ments consisting of rabbit skeletal muscle myosin–myosin rodmixture. We have obtained the following important informationabout the myosin head flexibility in hydrated myosin filaments:(i) in the absence of ATP, the time averaged myosin head positiondoes not change appreciably with time, indicating that the myosinhead fluctuates around a stable equilibrium position; (ii) in re-sponse to ATP, the myosin heads move away from the filamentbare region, indicating that the ATP-induced myosin head move-ment corresponds to the recovery stroke; (iii) after exhaustion ofATP, the myosin heads return to their original position beforeATP application. These findings indicate that, even in the absenceof actin filaments, the myosin heads can undergo their functionalconformational changes without being guided by actin filamentsand without being masked by random thermal fluctuations. It fol-lows from this that individual myosin heads may be fairly rigidwhen they are connected to myosin filaments, so that their func-tional movement is not masked by thermal fluctuation.

In the present study, we attempted to examine the validity ofthe myosin head lever arm hypothesis in hydrated myosinfilaments by recording the ATP-induced movement at different re-

Fig. 1. (A–C) A gallery of electron microscopic images of rotary shadowed antibody 1, 2 oror (C) antibody 3. IgG molecules are indicated by arrowheads. Scale bar, 100 nm. The sizefrom the published paper [12]. (D) Binding specificity of antibody 3 for the myosin regulatransferred onto PVDF membrane. In lane 1, all the myosin subunits, i.e. the heavy chain(d), are visualized by Coomassie staining. In lane 2, the regulatory light chain is specificallbinding of antibody 3 to the two peptides.

gions within individual myosin heads, by using three different site-directed antibodies, which will hereafter be called antibody 1, anti-body 2 and antibody 3, respectively. Antibodies 1 and 2 have al-ready been used successfully by Sutoh et al. [11] to obtainmolecular maps of the myosin head, while antibody 3 has beenprepared by us by the method similar to that of Sutoh et al. [11].Here we report that the results obtained are entirely consistentwith the myosin head lever arm mechanism, provided someassumptions are made.

2. Materials and methods

2.1. Gas environmental chamber

The gas environmental chamber (EC) used was identical to thatof Sugi et al. [9,10]. The EC is a cylindrical compartment with upperand lower windows to pass the electron beam. Each window iscovered with a carbon sealing film (thickness, 15–20 nm)supported on a copper grid with seven apertures, providing sevenviewing areas. The specimen was placed on the lower carbon film

3 (IgG)–myosin complexes. IgG is the affinity-purified (A) antibody 1, (B) antibody 2of IgG molecule is larger than that of myosin head. The panels A and B are provided

tory light chain. The proteins in the myosin sample are separated on SDS–PAGE, and(a), the essential light chain (b), the regulatory light chain (c), and other light chainy visualized by Western blot using antibody 3. We also obtained ELISA data showing

H. Minoda et al. / Biochemical and Biophysical Research Communications 405 (2011) 651–656 653

with a thin layer of experimental solution (thickness, <250 nm),which was in equilibrium with the circulating water vapor(pressure, 60–80 torr; temperature, 26–28 �C). The EC containedan ATP-containing microelectrode with its tip immersed in theexperimental solution to apply ATP to the myosin filaments ionto-phoretically [9,10]. The EC was attached to a 200-kV transmissionelectron microscope (JEM 2000EX, JEOL).

2.2. Synthetic myosin filaments

Synthetic myosin filaments consisting of myosin–myosin rodmixture were prepared by the method of Sugi et al. [10]. Briefly,myosin and myosin rod, obtained from rabbit psoas muscle, weremixed at a molar ratio of 1:1, and slowly polymerized by dialysisagainst a solution of low ionic strength to obtain thick bipolarmyosin filaments. Colloidal gold particles (diameter, 20 nm; coatedwith protein A; EY Laboratories) were attached to three differentsites within individual myosin heads, using three different site-di-rected antibodies, described in the text. The antibodies were ob-served to attach to only one of the two myosin heads probablybecause of steric hindrance [9–11]. The Mg-ATPase of the filaments(0.14–0.15 s-1) was not appreciably affected by these antibodies.

2.3. Recording and analysis of the filament images

To avoid electron beam damage to the filaments (15), observa-tion and recording were made with a total incident electron dosebelow 6–7 electrons/nm2 [12]. The filament images were recordedwith the imaging plate (IP) system (PIX system, JEOL) under amagnification of 10,000� (exposure time, 0.18 s). The pixel sizeon the IP records was 2.5 � 2.5 nm. Both the vertical and horizontalcoordinates in each pixel were divided into 10 subdivisions, so thatthe center of mass position of each gold particle was determined as

Fig. 2. Myosin head structure showing approximate regions of attachment of antibodies(CAD) comprises the 25 K (green), the 50 K (red) and part of the 20 K (dark blue) fragmen20 K fragment and the essential (ELC, light blue) and the regulatory (RLC, magenta) ligpeptides around Lys 83 and that of two peptides (Met58 � Ala70 and Leu106 � P(http://pymol.sourceforge.net). (For interpretation of the references to colour in this figu

the coordinates (two significant figures) within a single pixelwhere the center of mass position was located. The center of massposition for each selected gold particle was determined as thecoordinates (two significant figures) within a single pixel wherethe center of mass position was located, and these coordinates rep-resenting the center of mass position of the particles i.e. the posi-tion of the myosin heads were compared between two IP records ofthe same filaments. The absolute coordinates common to the twoIP records were obtained based on the position of natural markers(bright spots on the carbon film). Other details of the method havebeen described [9,10].

3. Results

3.1. Position-marking of the three different regions within individualmyosin heads

It has been shown that antibody 1 attaches to the junctionalpeptide between the 50 and 20 K heavy chain segments, locatednear the actin binding site in the catalytic domain of the myosinmolecule, while antibody 2 attaches to the reactive lysine residue(Lys83) (Sutoh et al. [11]), which is located at the boundary be-tween the catalytic and converter domains [13]. In addition tothese two antibodies [11], we prepared another site-directed anti-body (antibody 3) which attaches to two peptides (Met58 � Ala70and Leu106 � Phe120) of the myosin regulatory light chain in thelever arm domain [14]. The two different sites of attachment ofantibody 3 were effective in facilitating its attachment to the reg-ulatory light chain. As shown in Fig. 1, antibody 3 molecules (IgG)are confirmed to actually attach to the head-rod junction of themyosin molecule (C), i.e. the myosin head lever arm domain, whileantibodies 1 and 2 attach to the head of myosin molecule (A and B).

1, 2 and 3, indicated by numbers 1, 2 and 3, 30 , respectively. The catalytic domaints of myosin heavy chain, while the lever arm domain (LD) comprises the rest of theht chains. CAD and LD are connected via the converter domain (COD). Location ofhe120) in LD are colored yellow. Figure prepared by using software PyMOLre legend, the reader is referred to the web version of this article.)

654 H. Minoda et al. / Biochemical and Biophysical Research Communications 405 (2011) 651–656

In addition, Western blot analysis showed that antibody 3 specifi-cally bound only to the regulatory light chain among all myosinsubunits.

Fig. 2 is a diagram of the myosin head structure, consisting ofthe catalytic (CAD) and the lever arm (LAD) domains, connectedvia the converter domain (COD). The essential (ELC) and regula-tory (RLC) light chains are located in the lever arm domain.Approximate attachment regions of antibodies 1, 2 and 3 are indi-cated by numbers 1, 2 and 3, 30, respectively. Gold particles(diameter, 20 nm) were attached to individual myosin heads asposition markers using one of the three antibodies, so that the at-tached particles served as markers indicating ATP-induced move-ment at the three different regions within individual myosinheads. In the electron microscopic field (magnification,10,000�), each gold particle consisted of 20–50 dark pixels, andwas clearly distinguished from the background [10]. The centerof mass position of the gold particle was taken as the myosinhead position [9,10].

3.2. Direction of the ATP-induced myosin head movement at both sidesof the filament bear region

Based on the stability of the myosin head position time-aver-aged over 0.18 s [10], we examined the ATP-induced myosin headmovement by comparing two imaging plate (IP) records of the

Fig. 3. Examples of IP records showing the reversal in the direction of ATP-induced myothe same filament with gold particles, taken before ATP application (red) and after A(C) Diagram showing the reversal in the direction of ATP-induced myosin head movemenB. Open and filled circles (diameter, 20 nm) are drawn around the center of mass positiolocation of the bare region is indicated by vertical broken line. (For interpretation of the rthis article.)

same filaments, taken before and after application of ATP [9,10].The amplitude of ATP-induced myosin head movement was mea-sured both on segments of the filaments separated from other fil-aments and on segments of the filaments partly overlapped otherfilaments. Measurements of ATP-induced myosin head movementwere also made on the filaments, which clumped and overlappedwith other filaments, so that the filament profile was not clear.As has been the case for the myosin heads, position-marked withantibody 1 at the catalytic domain [10], the myosin heads, posi-tion-marked with antibodies 2 at the catalytic-converter domainboundary, moved mostly in one direction in response to iontopho-retically applied ATP, at variable angles to the filament long axis,probably reflecting (1) irregular myosin head arrangement in thesynthetic filament and (2) mechanical restriction of ATP-inducedmovement in the myosin heads located at the lower filament sur-face facing the carbon sealing film [10].

As shown in Fig. 3, we also succeeded in recording reversal inthe direction of ATP-induced movement of the myosin heads, posi-tion-marked with antibody 2 at the catalytic-converter domainboundary, across the filament bare region where the myosin headpolarity was reversed. Fig. 3A and B show a pair of IP records of thesame filament with the myosin heads, position-marked with goldparticles at both sides of the bare region. Record A (colored red)was taken before ATP application, while record B (colored blue)was taken after ATP application. Fig. 3C is a diagram showing the

sin head movement across the filament bare region. (A and B) a pair of IP records ofTP application (blue). The myosin head were position-marked with antibody 2.

t across the filament bare region, obtained by comparing the records shown in A andns of the same particles before and after ATP application, respectively. Approximateeferences to colour in this figure legend, the reader is referred to the web version of

H. Minoda et al. / Biochemical and Biophysical Research Communications 405 (2011) 651–656 655

ATP-induced myosin head movement, obtained from the change inthe center of mass position of each particle between the records Aand B. It can be seen that, at both sides of the bare region, the myo-sin heads move away from the bare region at variable angles to thefilament long axis [9,10].

On the other hand, we could not obtain records clearly showingthe reversal in the direction of ATP-induced movement in the myo-sin heads, position-marked with antibody 3 at the lever arm do-main, because the amplitude of movement was mostly verysmall. Nevertheless, the present results indicate that the myosinhead lever arm domain also moves away from the bare region.Since both the catalytic domain up to its boundary to the converterdomain is supported by the lever arm domain, which is furtherconnected to the filament via myosin S 2, it may be obvious thatthe lever arm domain also moves away from the bare region aswith the whole catalytic domain. It may be therefore safe to con-

Fig. 4. (A–C) Histograms showing the amplitude distribution of ATP-induced myosin he(C), each obtained from the ATP-induced changes in the center of mass position of each gothe absence of actin filament (D), and in the presence of actin filament (E). Approximaterespectively.

sider that the ATP-induced myosin head movement correspondsto the recovery stroke.

3.3. Amplitude of the ATP-induced movement at three different regionswithin the myosin head

Fig. 4 A, B and C present histograms showing the amplitudedistribution of ATP-induced movement of the myosin heads, posi-tion-marked with antibody 1 (A), antibody 2 (B) and antibody 3 (C),respectively. The histograms were constructed from 1692 mea-surements on nine different pairs of IP records obtained from 10different filaments in A, from 1112 measurements on five differentpairs of IP records obtained from four different filaments in B, andfrom 981 measurements on five different pairs of IP records ob-tained from five different filaments in C. The large number of mea-surements in A compared to those in B and C is due to inclusion of

ad movement, position-marked with antibody 1 (A), antibody 2 (B) and antibody 3ld particle. (D and E) Diagrams illustrating the myosin head lever arm mechanism inattachment regions of antibodies 1, 2 and 3 are indicated by numbers 1, 2 and 3, 30 ,

656 H. Minoda et al. / Biochemical and Biophysical Research Communications 405 (2011) 651–656

measurements made in the previous study with antibody 1 [10].The wide scatter of amplitudes including values >12.5 nm has beendiscussed previously [10]. In both A and B, the amplitude distribu-tion showed a peak at 2.5–5 nm.

The mean amplitude of ATP-induced movement was6.14 ± 0.09 nm (mean ± s.e.m., n = 1692) at the actin binding sitein the catalytic domain, and 6.14 ± 0.22 nm (n = 1112) at the cata-lytic-converter domain boundary (B), indicating no significant dif-ference in the amplitude of ATP-induced movement between thetwo regions in the catalytic domain. On the other hand, theamplitude distribution profile of ATP-induced movement at theregulatory light chain in the lever arm domain showed a peak at0–2.5 nm, and the average amplitude of ATP-induced movementwas 3.55 ± 0.11 nm (n = 981), being significantly smaller than thecorresponding values in the catalytic domain (t-test, P < 0.01).

4. Discussion

Based on the small myosin head flexibility stated above, thepresent results can well be accounted for by the lever arm mecha-nism if it is assumed that the cyclic conformational changes of themyosin head are almost the same in both the presence and absenceof actin filaments, as illustrated in Fig. 4D and E. In the ATP-in-duced myosin head movement in the absence of actin filaments,i.e. the myosin head recovery stroke (Fig. 4D), the myosin head isinitially in the post-power stroke configuration (solid line), andon binding with ATP it undergoes a conformational change to reachthe pre-power stroke configuration with hydrolysis products(Pi and ADP) bound to it (broken line). During the recovery stroke,the myosin head lever arm would rotate not only actively aroundthe converter domain but also passively around the boundary be-tween the lever arm domain and myosin S 2. As a result, the ampli-tude of ATP-induced movement is definitely larger at the tworegions in the catalytic domain (indicated by numbers 1 and 2)than at the regulatory light chain in the lever arm domain (3 and30), in complete agreement with the results shown in Fig. 4.

In the myosin head power stroke by the lever arm mechanism(Fig. 4E), the myosin head is thought to attach to actin filamentsinitially with hydrolysis products bound to it (solid line), and thenundergoes a conformational change, coupled with release of Pi andADP, to reach the post power-stroke configuration (broken line).By comparing the diagrams in Fig. 4D and E, one may wonderwhy the lever arm domain can rotate around the boundary to myo-sin S 2 in the absence of actin filament, which is expected to com-press the attached myosin head towards the myosin filamentbackbone. It may be that the lever arm rotation around the con-verter domain and around the boundary to myosin S 2 is tightlycoupled with each other by some unknown mechanism. Muchmore experimental work is needed to elucidate this mechanism.

Meanwhile, recent improvement of the atomic force micro-scope (AFM) has made it possible to visualize processive move-ment of myosin V along actin filaments fixed on a mica surface[15]. This technique may not, however, be applied to record

mechanically unrestricted movement of myosin heads extendingfrom the upper surface of bulky myosin filaments, fixed to carbonsealing film. We are currently planning to develop movie recordingsystem by coupling the EC system with high-speed movie system.

To summarize, the present results are entirely consistent withthe myosin head lever arm mechanism in muscle contraction, ifthe ATP-induced myosin head movement recorded in hydratedmyosin filaments is assumed to be a reversal of the myosin headpower stroke. The present work may therefore constitute the firstelectron microscopic evidence for the myosin head lever armmechanism. Finally, we emphasize that our methods with the ECis an extremely powerful tool in elucidating mysteries in the fieldof life sciences.

Acknowledgments

We thank Japan Electron Optics Laboratory, Ltd. for generouslyproviding facilities, which made our work possible, andDrs. K. Nagayama and T. Etoh for financial support, and Dr. K. Sutohfor advice in preparing antibody 3. We are also indebted to Drs. K.Sutoh and M. Tokunaga for allowing us to use their published elec-tron micrographs of antibody–myosin complexes.

References

[1] A.F. Huxley, R. Niedergerke, Interference microscopy of living muscle fibres,Nature 173 (1954) 971–973.

[2] H.E. Huxley, J. Hanson, Changes in the cross striations of muscle duringcontraction and stretch, Nature 173 (1954) 973–976.

[3] R.W. Lymn, E.W. Taylor, Mechanism of adenosine triphosphate hydrolysis byactomyosin, Biochemistry 10 (1971) 4617–4624.

[4] T.P.Q. Uyeda, P.D. Abramson, J.A. Spudich, The neck region of the myosin motordomain acts as a lever arm to generate movement, Proc. Natl. Acad. Sci. USA 93(1996) 4459–4464.

[5] M.A. Geeves, K.C. Holmes, Structural mechanism of muscle contraction,Ann. Rev. Biochem. 68 (1999) 687–727.

[6] S. Fisher, D. Windschugel, D. Horak, K.C. Holmes, Structural mechanism of therecovery stroke in the myosin molecular motor, Proc. Natl. Acad. Sci. USA 102(2005) 6873–6878.

[7] H.E. Huxley, M Reconditi, A. Stewertand, et al., X-ray interference studies ofcross-bridge action in muscle contraction: evidence from quick releases, J. Mol.Biol. 363 (2006) 743–761.

[8] A. Fukami, S. Murakami, Progress of electron microscopy of wet specimensusing film-sealed environmental cell, J. Electron. Microsc. 28 (1991) S41.

[9] H. Sugi, T. Akimoto, K. Sutoh, et al., Dynamic electron microscopy ofATP-induced myosin head movement in living muscle thick filaments,Proc. Natl. Acad. Sci. USA 94 (1997) 4378–4382.

[10] H. Sugi, H. Minoda, Y. Inayoshi, et al., Direct demonstration of the cross-bridgerecovery stroke in muscle thick filaments in aqueous solution by using thehydration chamber, Proc. Natl. Acad. Sci. USA 105 (2008) 17396–17401.

[11] K. Sutoh, M. Tokunaga, T. Wakabayashi, Electron microscopic mappings ofmyosin head with site-directed antibodies, J. Mol. Biol. 206 (1989) 357–363.

[12] H. Suda, A. Ishikawa, A. Fukami, Evaluation of the critical electron dose on thecontractile activity of hydrated muscle fibers in the film-sealed environmentalcell, J. Electron Microsc. 41 (1992) 223–229.

[13] A. Muhlrad, Y.M. Peyser, M. Nili, K. Ajtai, et al., Chemical decoupling of ATPaseactivation and force production from the contractile cycle in myosin by sterichindrance of lever-arm movement, Biophys. J. 84 (2003) 1047–1056.

[14] R.A. Milligan, Protein–protein interactions in the rigor actomyosin complex,Proc. Natl. Acad. Sci. USA 93 (1996) 21–26.

[15] N. Kodera, D. Yamamoto, R. Ishikawa, T. Ando, Video imaging of walkingmyosin V by high-speed atomic force microscopy, Nature 468 (2010) 72–76.