Muscle Research

It is one of the oldest and newest lines of biological
inquiry, promising an insight into the nature of life.

Albert Szent-GyGrgyi

If science is the art of measuring, then
muscle has no equal as a material in the
study of life, for there is no other tissue
whose function is connected with equally
extensive and intensive changes in chem-
istry, physical state, energy, and dimen-

26 SEPTEMBER 1958

sions. This is why physiology, up to the
turn of the century, was mainly mus-
cle physiology. After muscle had been
pushed into the background by enzymes
and hormones for a while, the develop-
mcnt of modern physical methods once

699


inquiry. Fortunately, research is simpli-
fied to some extent by the fact that
energy production and energy consump-
tion are separated. This separation en-
ables the researcher to work on one of
the two processes independently. What
is driving the muscle machine is, accord-
ing to our present knowledge, the fret
energy released by the splitting of the
terminal "high-energy phosphate-bond,
- I'," of adenosine triphosphate (ATP j ,
which is created at the cxpensc of fer-
mentation and oxidation. Oxidative phos-
phorylation is linked to the mitochon-
dria, while contraction is the function of
the contractile filaments.

There are many approaches to muscle.
We can inquire, for instance, into the
physical changes accompanying contrac-
tion,  measuring  heat production or
changes in elastic properties, as A. V.
Hill and his associates have done. We
can inquire into the nature of the single
parts of the contraction cycle, asking
how depolarization is produced on the
muscle membrane, how this depolarizs-
tion is propagated, how it triggers the
function of the contractile matter inside
the fiber, and how the contracted muscle
returns to its resting state. We may in-
quire into the nature of the contractile
material and the changes which it un-
dergoes in contraction and subsequent
relaxation, and we may inquire into the
feedback mechanisms which adjust mo-
tion to the physiological requirements.
Since each of these partial processes rep.
resents a more or less self-consistent field
of inquiry, it is impossible to cover all
of them within the boundaries of a short
article. Accordingly, I shall limit myself
to one aspect only, one to which most
of my personal experience relates: the
problem of the mechanochemical cou-
pling and the nature of the main con-
tractile protein, myosin.

Early Work on Myosin

Myosin has been known for almost a
century, having been discovered by W.
Kiihne, who showed that a great amount
of a protein can be extracted from mus-
cle by a strong salt solution. This pro-
tein precipitated on dilution of the salt
present and was found in the 1930's by
Edsall, Muralt, H. H. Weber, and others
to consist of rod-shaped molecules. When
I embarked on muscle research two dec-
adcs ago it became increasingly clear that
what was driving contraction was the
- P of adenosine triphosphate. Engel-
hardt and Ljubimowa (I) had just dis-

covered that myosin could split this bond
and thus release the energy which it
needed for its contraction. The idea of

a "contractile enzyme" was most cxcit-
ing. None of us had much doubt, then,
that contraction had to be some sort of
a folding, elicited in the myosin rodlets
by the ATP molecule at certain points,
and we were looking forward to ihe pos-
sibility of describing this reaction soon
by a simple chemical equation.

The only trouble was that myosin
would not contract outside the body.
My associates, Banga and Straub, and I
showed (2) that this failure was due to
the fact that the contractile protein was
not myosin but actomyosin, a complex
of myosin with a hitherto unknown pro-
tein,  "actin." About the same time
Schramm and Weber (3) showed "myo-
sin" to be dishomogeneous in the ultra-
centrifuge. Under the clcctron micro-
scope (Ardenne and Weber, 4) the faster
scdimenting fraction was found to con-
sist of filaments which wcrc, evidently,
filaments of actomyosin.

In the resting muscle there seems to
be no interaction between actin and
myosin, the formation of actomyosin be-
ing brought about by "excitation." The
association of actin and myosin goes
hand in hand with the increase in elastic
modulus which characterizes the "active
state" of A. V. Hill (5). Once it has
been formed in the presence of physio-
logical concentrations of ATP and ions,
actomyosin has to go over into its con-
tracted state. The energy spent in this
process can be used to lift a weight-
that is, to do work.

What made actin exciting was the fact
that it allowed us to produce and study
motion and contraction in vitro, and
bolstered our hopes that soon we would
know all about the process. If ATP was
added to actomyosin in the test tube, the
actomyosin underwent violent physical
changes which consisted in the shorten-
ing of its filaments and the loss of its
hydrophilous character. The analogy be-
tween these in vitro reactions and mus-
cular contraction could be brought closer
by showing that a muscle, thoroughly ex-
tracted with glycerol, is still capable of
contracting and developing maximal ten-
sion on addition of physiological concen-
trations of ATP (6) (Glycerol destroys
the liner mechanisms but leaves acto-
myosin intact.) So the conclusion could
be drawn that muscular contraction, es-
sentially, is an interaction of aclin, myo-
sin, ATP, and ions. I will omit the dis-
cussion of actin and limit myself to
myosin.

Complex Nature of Myosin

The first experimental evidence that
the situation was not as simple as we
believed and that myosin is not a homo-
geneous rodlet was obtained by Gergely
(7) and Perry (8), who showed that
trypsin decreased the viscosity of myosin
solutions without decreasing its ATP-asc
activity. The myosin, thus treated, could
be separated into two fractions, only one
of which showed enzymic activity. After
studies pursued with Mihalyi (Y), the
final analysis of this change was given by
A. G. Szent-Gyorgyi (9), who showed
that the "myosin molecule" is disinte-
grated by trypsin into six subunits, mero-
myosins, which were shunted in a row,
in series. There arc two different kinds
of such subunits. One kind was thicker
and sedimented faster than the other and
was, accordingly, called "H" (heavy),
while the other was slender and had a
lower molecular weight and was called
"L" (light), The H meromyosin had the
full ATP-ase activity of the whole myo-
sin molecule and interacted with actin,
while the L seemed to be involved in
shortening. The nature of the links hold-
ing the mcromyosins together has not
yet been cleared up definitely. All the
same, these findings made it certain that
the myosin particle is not a homogenc-
ous rodlet but consists of different parts
with different structures and functions.
The L meromyosin has a high, the II
a low, a-helix content (Cohen et al., 20).
That these subunits are, in one way or
another, preformed in myosin is also
shown by their different amino acid
turnover numbers (Velick, 1 I ).

The situation was somewhat simpli-
fied by Laki and Carroll's (12) finding
that carefully extracted myosin had only
half of the previously accepted molecu-
lar weight; "old myosin" was thus a
dimer formed in vitro after extraction.
As far as its dissociating action on acto-
myosin is concerned, ATP seems to re-
act with myosin in stoichiometric propor-
tions (Hanson and Mommacrts, 13). To
compensate for this simplification, it was
found that the meromyosins themselves
are built of a great number of much
smaller subunits into which they disinte-
grate if they are acted upon by urea.
The L type disintegrates completely, the
H partially (A. G. Szent-Gyorgyi and
Borbiro, 24). The molecular weight of
these sub-subunits, "protomyosins," is
about l/100 that of myosin. What is dis-
turbing about this finding is the fact that
urea is known to split hydrogen bonds
only, leaving covalent bonds intact. If

SCIENCE, VOL. 128


we define a molecule as a structure with
a covalent backbone, then the "myosin
molecule" is no molecule at all but a
complex system of small units held to-
gcther by secondary forces, like H-bonds,
van dcr Waals attractions, or dipole mo-
ments.

Function and Chemistry

As a rulr, new knowledge leads to
a better understanding. With muscle,
things seem to go in the opposite direc-
tion, and one may ask whether the real
difficulty does not lie in an inadequacy
of our basic concepts. Present biochem-
istry stands under the domination of
classical chemistry, according to which
two molecules must come into bodily
contact to be able to interact. This woud
mean that the ATP molecules can in-
duce changes in the contractile protein
only at the points at which they are
bound and split. The fact that only the
H meromyosin splits ATP, while it is
the L which seems to be more directly
involved in contraction, suggests the in-
adequacy of the classical concept, mak-
ing some sort of a migration of energy
seem likely. This calls to mind the case
of the Bacillus proteus. This bacterium
has long flagellums, about as long as a
sarcomere. These flagellums move the
bacterium by means of the undulatory
motion passing along their whole Icngth.
According to their x-ray spectrum, as
shown by the studies of Astbury, Beigh-
ton, and Weibull (15), these very thin,
threadlike structures arc closely related
to rnyosin and have about the same di-
ameter as the contractile filaments of
muscle. Thus, in them we see "biolog-
ical movement stripped to its barest cs-
sentials." Since these flagellums are too
thin to allow us to suppose that circula-
tion takes place inside them, the cncrgy
which moves them must be fed into
them at their basal end and then, some-
how, must migrate along their Icngth.
Perhaps we have taken a much too nar-
row view of life in trying to explain all
its reactions in terms of classical chem-
istry. In order to understand we might
have to descend from the dimension of
macromolecules to those of electrons,
from classical chemistry to quantum me-
chanics, taking into account factors such
as molecular excitations, the resonance
transfer of their energy, solid-state phys-
ics, the electromagnetic field and its per-
turbations, long-range water structures,
and, possibly, proton conduction. Every-
thing seems possible at present. Our
knowledge of muscle is in the liquid state.

26 SEPTEMBER 1958

Function and Structure

Looking out for some more solid hold,
one can try to correlate the known chem-
ical data with the classical microscopic
structure of muscle. Such an attempt was
made lately by Holtzer and Marshall
(16)) who applied Coons' (17) "fluorcs-
cent antibody method" to muscle, inject-
ing the various muscle proteins and their
subunits into rabbits and then making
visible the immune bodies thus produced
by couphng them with a fluorescent dye.
These workers found that the different
immune bodies were bound differently
by the different parts of the sarcomcre.
The "myosin antibody" was bound by the
A-band. This finding supported earlier
findings (Ambcrson, 18; Hassclbach, 19;
Hanson and Huxley, 20) that myosin is
located in the A-band. The "L-antibody"
was bound by the lateral parts of the
`A-band, while the "H-antibody"  was
bound by the narrow M-band, lying in
the middle of the sarcomerc, suggesting
that this band is its location, and there
may be no such thing as myosin in mus-
cle at all. What we called "myosin"
might have been an aggregate of mero-
myosins formed after their extraction.

Another approach was opened by the
polarization microscope of Shinya InouP
(21). This instrument, with its high reso-
lution and clean polarization optics, re-
veals new structural details and shows
new cross bands. It also indicates that
the A-band contains a relatively great
quantity of a structural protein which ic
neither myosin nor actin and which

The muscles which move our body con-
sist of fibers of the dimension of a human
hair. Under the microscope (schematic
representations above) these fibers are
found to he built of darker, denser, doubly
refractant segments (the anisotropic "A-
hands"), and lighter, less dense segments
with poor double refraction (the isotropic
Y-bands"). In the middle of the
I-hands are the "Z-membranes." The seg-
ments enclosed by two Z-membranes are
called "sarcornercs." In the middle of the
A-hand there is a thin membrane, the
"M-membrane," delimited on either side
by a narrow zone of small density, the
"H-hand."

seems to bc identical with the "X-pro-
tein" (22). The microscope also shows
that muscle fibrils from which myosin
has been extracted bind H meromyoGn
with preference in the M-band.

In considering the problem of corrc
lating structure with function and them
istry, one's thoughts naturally turn to the
electron microscope, which extended the
domain of morphology into macromo-
lecular dimensions. The first attempt on
this line is linked to the names of Hall,
Jakus, and Schmitt (23), who showed
that the muscle fiber, csscntially, is a
bundle of a great number of thin fila-
ments which do not bend or fold in con-
traction. New details were revealed lately
by the admirable pictures of H. Huxley
(24) which show the presence of two
kinds of filaments in cross-striated mus-
cle. There are thicker "primary" fila-
ments, located in the .4-band, and twice
as many thinner "secondary" filaments
reaching from the Z-band to the H-band,
In cross sections the thinner filaments
were found to surround the thick ones
in a hexagonal array.

Sliding Filaments

On stretching, the two kinds of fila-
mcnts were found to be sliding past one
another, making the H-zone and I-band
wider. Building on these observations,
Hanson and Huxley (25) proposed a new
theory of contraction according to which
what happens in this process is the oppo-
site of what happens on stretching: the
secondary filaments are pulled in bc-
twecn the primary ones with a consccu-
tive gradual narrowing of the I-band,
which disappears altogether when the
%-membrane reaches the A-band. A. F.
kIuxley's (26) motion pictures of living
muscle strongly plead for this mechanism
of contraction, which explains also the
puzzling fact that thcrc is no change in
x-ray periodicities in initial states of co,,-
traction: the muscle shortens but its
filaments do not.

No doubt, this theory signifil,s an im-
portant step in the study of muscle. It
gives a clear picture of the mechanics
and the morphological changes taking
place in the contraction of cross-striated
muscle, offering a solid foundation for
further discussion. But do we really un-
derstand muscle now? Far from it; mus-
cle has remained just as much a mystery
as it was before. We still do not know
what happens when ATP is split and
how its energy is, eventually, convcrtcd
into the pull exerted on the secondary

701


filaments. The in vitro reaction of actin,
myosin, and ATI' shows that there are
interactions between thcsc substances
leading to violent physical changes.
Though physical (A. F. Huxley, 27)
and chemical (H. H. Weber, 28) theo-
ries are not lacking, the nature of these
interactions is still unknown. They rep-
resent the primary happening and form
the core of the problem of muscular con-
traction. 1Yithin the framework of the
macromolecular arrangements of cross-
striated muscle, they cause the secondary
iilaments to bc pulled in between the pri-
mary ones, but if this "pulling in" is all
there is to it, then shortening should stop
at 30 to 40 percent-as soon as the
Z-membranes reach the A-band. All the
same, muscle can go on shortening up to
80 percent, producing tension all the
time. These high degrees of shortening, in
cross-striated muscle, may not be physio-
logical,  corresponding to the "delta
state" of Ramsay, in which changes be-
gin to be irreversible (5). All the same,
for the theory they are of prime import.
Smooth muscles which have no cross
bands, and, accordingly, no periodic dou-
ble array of filaments, also contract up
to 80 percent, though they do so at a
slower rate. Similarly, actomyosin fila-
ments can contract under the influence
of ATP up to 80 percent, though "slid-
ing" makes no sense at all here. So it
seems that the sliding of filaments is
linked to the specific steric arrangements
in cross-striated muscle, where this slid-
ing makes rapid shortening possible,
being the secondary consequence of
changes which we fail to understand.

Conclusion

So we can sum up by saying that WC
still do not understand muscle and do
not know how ATP is driving it. It may
be true not only that our outlook on bio-
logical action is too narrow, but also that

our knowledge of muscle structure is too
incomplete. Important structures, such as
the "cndoplasmic reticulum" (Porter
and Pallade, 29), have been discovered
lately, and there is no reason to believe
that this structure is the last unknown.
Important protein fractions (22) wait
for identification, while other fractions,
such as Bailey's tropomyosin (30) have
not yet been fitted into the muscle ma-
chine, The dimensions indicate that the
myosin filaments are many molecules
thick. So we have to suppose that, just
as protomyosins have to join in a very
specific way to form a myosin molecule
(if there is such a substance at all),
so the myosin molecules have to join
in a very specific way to build a filament
-structural details, without the detailed
knowlcdgc of which we can hardly hope
to understand function. The painstaking
and extensive application of current
methods may yield a great deal of im-
portant new information, but it is pos-
sible that entirely new approaches arc
needed. Such new approaches are being
opened in various quarters. Koshland's
(31) application of the isotope tcch-
niques has already led to surprising new
data. The magnetic anisotropy of mus-
cle, discovered recently by Arnold,
Mueller, and Steele (32) in my labora-
tory, may lead to new clues.

Thcrc is a certain urgency about solv-
ing all these riddles, for only a better
understanding of muscle can enable us
to cope with its disorders, which cause
so much suffering. The number of dys-
trophic patients in this country alone
goes into the hundred thousand, and so
does the number of lives lost because of
hormonal disturbances of the membrane
activity of uterus muscle cells (Csapo,
33). We can hope that a better under-
standing of muscle will not only spare
human suffering and frustration but that
it will bring us closer, also, to the under-
standing of the basic principles on which
life is built.

Relerences

I. W. A. Engelhardt and M. N. Ljuhimowa.
Nnlurc 144, 669 ( 1939).
2. A. Scent-GyGqyi, Ckemirtry of Murcular
Contraction (Academic Press, New York.
1947); Act0 Physiol. Srand. 9. suppl. 25
,1945,

\_" .I,.

3. G. Schramm and H. H. Weher, K&&d-Z
Ino, 242 (1942).

4. M. v. Ardenne and H. II. Weber, Koiloid-Z.
97, 322 (1941).

5. W. F. II. M. Mommaer ts, Murculor cmtrac-
lion (Interscience, New York, 1950).
G, A. Szent-Gvhrevi. Bid. 1
       ~I~ _, I    Dull. 9G, 140 (1949).
7. J. Gcrq _ely, Federation Proc. 9, I76 (19.50) :
IO ) 188 (1951).
a. s. V. Perry. Biockem. 1. 48. 257 (19511.
9. A. G. Szent-Gydrgyi, Aduonrer in Enzvnol.
16, 313 (1955).

10. C. Cohen and A. G Szent-Gyorgyi, 1. Am.
  Chenr. SW. 79, 248 (1957).
II. S. F. Velirk, fliockim. et Biopkyr. Aclo 20,
  228 (195G).

12. K. Laki and W. R. Carroll, Nnlure 175. 389
  ( 195.5)

13. W. F. H. M. Mommaerts and J, Hanson, I.
  Grn. Phyriol. 39, 831 (1956).
14. A. G. Scent-GyBrgyi and M. Borhirn, Arch.
  Diockem. Diopkyr. 60, 180 (1956).
15. W. T. Asthury, E. BeiShton, C. Weihull, Syuz-
pmia Sot. Explf. Iliof. 9, `282 (1955).
1G. H. Holtzer and M. Marshall, Proc. Biopkyr.
  Sm., Barton, Mar,., 5-7 Feb. 1958 (1958), p.
  35.

17. A. H. Coons, Fluorescent Antibody Methods.
  General Cyfockcmical Metkodr (Academic
  Press, New York, 1958).
18. W. R. Amberson, R. D. Smith, B. Cbinn, S.
Himmellarb, J. Metcalf, Dial. IluN. 97, 231
  (1949).
19. W. Hasselhach,  Z. Norurforrck. Bh, 449
  (1953).

20. J, Hanson and H. E. Huxley, Nature 172, 530
  (1953).

21. S. InouC and W. L. Hyde, .I. Biophyr. Dia-
  ckmt Cytol. 3, 83 (1957).
22. A. G. Scent-GyBrgyi, D. Ma&, A. Srent-
  GyBrgyi, Biockim. CL Biopkyr. Acfa 16, 339
  (1955).

23. C. E. Ilall, M. A. Jakus, F. 0. Schmitt, Iliol.
  Bull. 90, 32 (1946).

24. H. E. Huxlev. 1. Lliookvr. Biockem. Cylol. 3,

631 (1957). ..   -

25. J, Hanson a,,d H. E. Huxley, SymQosia Sot.
  Exptl. Bid. 9, 228 (1955).
26. A. F. Huxley and R. E. Taylor, h'aturc 17G,
  I(168 (19.55).

27. A. F. Huxley, Motion picture displayed al
  Waads Hale, Mass., 30 June 1958.
28. II. II. Weber. The Motility of Murcle Ceflr

(Harvard Univ. Press, &&ridge, Mass.,
1958).

29. K. R. Porter and G. E. Pallade, J. Iliopk)`,.
  Biockem. Cytol. 3, 269 (1957).
30. K. Bailey, "Structure proteins, 11. Muscle,"
  in The Proteins. II. Neurath and K. Bailey.
  Edr. (Academic Press, New York, 1954), vul.
  2, pt. h, p. 952.

31. Personal communication.
32. W. Arnold, H. Mueller, R. Steele, Proc. Nell
  Acad. Ser. U.S. 44. I 119581.



33. A. Csapo, Nnlur~ 173, lOi (19'14); `ice also
  Sci. Amrnran 198, 411 (1958).

SCIENCE, VOL. 128