Flexibility of the myosin molecule was studied by an in vitro motility assay in terms of the direction of actin movement. Actin filaments can move in both directions on tracks of heavy meromyosin made on a nitrocellulose surface, and, furthermore, along the native thick filaments passing over their central bare zone. These observations indicate that the myosin molecule has a considerable flexibility in interacting with actin filaments.

The recent development of in vitro motility assay systems has revealed that the polarity of actin filaments is important in determining the direction of actin-myosin sliding movement. We have developed an in vitro motility assay system for actin and myosin fragments (Kron et al. 1990). This assay system uses a flow cell consisting of a nitrocellulose film cast on a coverslip. Myosin fragments are put on the nitrocellulose film and the movement of rhodamine-phalloidin labelled actin filaments are observed by fluorescence microscopy. In the first series of experiments, we demonstrated that SI attached to nitrocellulose film can support the movement of actin filaments (Toyoshima et al. 1987).

In this assay system, there may be a carpet of myosin fragments or a lawn of them on the nitrocellulose film. Actin filaments move smoothly and nearly always continuously at a constant speed. Actin filaments follow winding paths and frequently change direction, but never move backward reversing the polarity of the movement. Apparently, the direction of movement is determined by the polarity of the actin filament. Since the movement is really smooth even with SI, there must be considerable flexibility in myosin heads.

To see the rigor binding of heavy meromyosin (HMM) to actin filaments, the surface was examined by electron microscopy. Fig. IA shows a negatively stained image of HMM-bound surface. This field corresponds to about 0.25 μ m2, and from the density estimate about 250 HMM molecules are in it. We can see arrowhead-like patterns along actin filaments. Thus, HMM can form rigor complexes even though they were scattered and tethered to the surface.

A similar surface was examined by freeze-etching and rotary-shadowing (Fig. IB). HMM heads appeared to bind to actin filaments from the top as well as from the bottom. Nothing special was apparently caused by tethering HMM to the surface. These observations suggest that HMM heads on the nitrocellulose surface can move to bind actin filaments with considerable freedom.

The next set of experiments was to investigate the movement of decorated actin filaments (Toyoshima et al. 1989). First, the decorated filaments were formed in a test tube by mixing stoichiometric amounts (head to actin monomer) of HMM and fluorescent actin. By placing decorated filaments on the nitrocellulose film we were able to make tracks of HMM. Since the HMM binds to actin filaments in a specific configuration, polarity of myosin heads must be the same. Then BSA solution was applied to block the free surface from nonspecific binding of actin filaments. Finally, ATP was added to start the movement.

On adding ATP, actin filaments moved continuously following the same path with one filament length, then dissociated from the surface when the trailing end of the filament passed the opposite end of the track. This was expected because the viscosity of the solution was very high and thus the inertia could be neglected.

After initial actin filaments have gone, HMMs are left on the nitrocellulose surface forming tracks. We have confirmed that the tracks are well preserved by examining the same field with a fluorescence microscope (to see the initial location) and with an electron microscope using immuno-gold labelling (to see the tracks).

Patterns of actin movement observed on the tracks are summarized in Fig. 2. After initial actin filaments have moved and dissociated, other filaments released from HMM tracks elsewhere in the flow cell came to the tracks and followed the same tracks very accurately (Fig. 2). Other filaments came to the track from different directions and followed the same path toward the other end (Fig. 2A). Sometimes, after having moved and dissociated at the end of the track, actin filaments flipped in solution and returned back to the other end (Fig. 2B). Further, two or more filaments came to the same track from different directions and passed over each other (Fig. 2C).

Fig. 1.

Electron micrographs of actin filaments on HMM lawn. (A) Negatively stained with uranyl acetate. (B) Freeze-etched and rotary shadowed.

Fig. 1.

Electron micrographs of actin filaments on HMM lawn. (A) Negatively stained with uranyl acetate. (B) Freeze-etched and rotary shadowed.

Fig. 2.

Various patterns of actin movement observed on the tracks of HMM.

Fig. 2.

Various patterns of actin movement observed on the tracks of HMM.

These observations suggest that HMMs are tethered to the nitrocellulose with their S2 portion, and their heads (SI portion) can rotate freely to support the movement of actin in both directions (Fig. 3).

Fig. 3.

A cartoon showing how HMM can support the movement of actin filaments in both directions.

Fig. 3.

A cartoon showing how HMM can support the movement of actin filaments in both directions.

There are several pieces of evidence that the myosin head can rotate freely around its axis. For example, two heads of HMM can bind to consecutive actin monomers located on the same strand (Craig et al. 1980), showing that the two heads are translationally related in this bound configuration. Further, a single IgG molecule crosslinks two heads of myosin (Winkelmann and Lowey, 1986; Miyanishi et al. 1988), showing that the two heads can be arranged so that the same epitopes on the two heads face each other. All this means that the myosin head can rotate around its axis more than 180 degrees, and suggests that myosin heads rotate very rapidly to find the correct binding site on the actin molecule.

From experiments to measure tension development at various sarcomere lengths (Gordon et al. 1966), it has been suggested that myosin heads produce force so that thin filaments slide to the center of the sarcomere, and that thin filaments cannot interact productively with another half of myosin heads over the central zone. The question arises whether actin filaments can move along isolated thick filaments up to the center or over the center to the other end.

To answer this question, we isolated long native thick filaments; long enough to be observed in a light microscope. Crustacean skeletal muscles are known to have long sarcomeres and long thick filaments. As material, we used crayfish claw muscles which have sarcomere lengths of about 10 um. To isolate thick filaments, claw muscle is homogenized gently and centrifuged at a low speed. The supernatant contains isolated thick filaments.

Electron microscopy showed many narrow thick filaments of about 5 to 6 μm in length. They have the central bare zone of 0.2μm. The thickness of the isolated thick filaments is about 25 nm and comparable to that of microtubules. Thus, the isolated native thick filaments can be visualized under dark field illumination as has been shown for microtubules (Miki-Noumura and Kamiya, 1976).

When applied to glass slides, fluorescent actin filaments as well as thick filaments are visible under dark field illumination. As the rhodamine-phalloidin labelled actin filaments emit red light, we can distinguish actin filaments from thick filaments by eye. Unfortunately, the SIT camera is color blind, everything becomes black and white. Nevertheless, the intensity of the portion where an actin filament is moving along the filament becomes higher; thus it is possible to extract the image of actin filaments by subtracting the image when no actin filament is moving as the reference. After storing sequences of images using a frame grabber, the computer painted actin and myosin filaments with different colors.

On the native thick filaments, actin filaments moved from one end to the other end passing over the central bare zone. No difference in velocity was found between the movements towards the center (regular orientation) and from the center (so-called wrong orientation). Although the velocity is the same in both orientations, we do not know if myosin can produce the same amount of force in both directions.

In conclusion, the direction of the movement is determined by the polarity of actin filaments and myosin has a large flexibility. It is marvelous that myosin can produce this large force in spite of its large flexibility. This flexibility must be taken into account when exploring the cross-bridge mechanism.

Craig
,
R.
,
Szent-Gyôrgyi
,
A. G.
,
Beese
,
L.
,
Flicker
,
P.
,
Vibert
,
P.
and
Cohen
,
C.
(
1980
).
Electron microscopy of thin filaments decorated with a Ca2+-regulated myosin
.
J. molec. Biol.
140
,
35
55
.
Gordon
,
A. M.
,
Huxley
,
A. F.
and
Julian
,
F. J.
(
1966
).
The variation in isometric tension with sarcomere length in vertebrate muscle fibers
.
J. Physiol.
184
,
170
192
.
Kron
,
S. J.
,
Toyoshima
,
Y. Y.
,
Uyeda
,
T. Q. P.
and
Spudich
,
J. A.
(
1990
).
Assays for actin sliding movement over myosin coated surfaces
.
Meth. Enzymol. (in press)
.
Miki-Noumura
,
T.
and
Kamiya
,
R.
(
1976
).
Shape of microtubules in solutions
.
Expl Cell Res.
97
,
451
453
.
Miyanishi
,
T.
,
Toyoshima
,
C.
,
Wakabayashi
,
T.
and
Matsuda
,
G.
(
1988
).
Electron microscopic study on the location of 23K and 50K fragments in skeletal myosin head
.
J. Biochem.
103
,
458
462
.
Toyoshima
,
Y. Y.
,
Kron
,
S. J.
,
McNally
,
E. M.
,
Nibbling
,
K. R.
,
Toyoshima
,
C.
and
Spudich
,
J. A.
(
1987
).
Myosin subfragment-1 is sufficient to move actin filaments in vitro. Nature
328
,
536
539
.
Toyoshima
,
Y. Y.
,
Toyoshima
,
C.
and
Spudich
,
J. A.
(
1989
).
Bidirectional movement of actin filaments along tracks of myosin heads
.
Nature
341
,
154
156
.,
Winkelmann
,
Í). A.
and
Lowey
,
B. J.
(
1986
).
Probing myosin head structure with monoclonal antibodies
.
J. molec. Biol.
188
,
595
612
.