114R. E. H. NICHOLAS AND C. RIMINGTON
5. Specimens from Prof. C. J. Watson's labor-
atory have been examined. The uroporphyrin
methyl ester I had m.p. 292-293° and behaved as
uroporphyrin I containing a trace of uroporphyrin
III on paper chromatography. The 2080 ester
behaved chromatographically as a hexacarboxylic
Wewish tothankDrFalkand Miss Benson forperforming
the chromatographic examinations by their method, and
also Prof. C. J. Watson and Dr J. Canivet for kindly placing
materials at our disposal. A generous grant to one of us
(C. R.) from the Trustees of the Nuffield Foundation has
made possible the establishment ofa Unit for the Investiga-
tion of Pyrrole Pigment Metabolism and this grant is
Chu, T. C., Green, A. A. & Chu, E. J. (1951). J. biol. Chem.
Falk, J. E. & Benson, A. (1953). Biochem. J. 55, 101.
Falk, J. E. & Willis, J. B. (1951). Aust. J. 8Ci. Re8. A, 4,579.
Fischer, H. (1915). Hoppe-Seyl. Z. 95, 34.
Gray, C. H. (1950). Arch. intern. Med. 85,459.
Grinstein, M., Schwartz, S. & Watson, C. J. (1945). J. biol.
Chem. 157, 323.
Jope, E. M. & O'Brien, J. R. P. (1945). Biochem. J. 39, 239.
Kennard, 0. & Rimington, C. (1953). Biochem. J. 55, 105.
Macgregor, A., Nicholas, R. E. H. & Rimington, C. (1952).
Arch. intern. Med. 90, 483.
McSwiney, R. R., Nicholas, R. E. H. & Prunty, F. T. G.
(1950). Biochem. J. 46, 147.
Mertens, E. (1936). Hoppe-Seyl. Z. 238, I.
Mertens, E. (1937). Hoppe-Seyl. Z. 250, 57.
Nicholas, R. E. H. (1951). Biochem. J. 48, 309.
Nicholas, R. E. H. & Rimington, C. (1949). Scand. J. clin.
Lab. Invest. 1, 12.
Nicholas, R. E. H. & Rimington, C. (1951 a). Biochem. J.
Nicholas, R. E. H. & Rimington, C. (1951 b). Biochem. J. 48,
Rimington, C. (1939). Proc. Roy. Soc. B, 127, 106.
Rimington, C. (1952). Sem. Hop. Parie (in the Press).
Rimington, C. & Miles, P. A. (1951). Biochem. J. 50, 202.
Rimington, C. & Sveinsson, S. L. (1950). Scand. J. clin. Lab.
Invest. 2, 209.
Waldenstrom, J. (1934). Acta med. scand. 83, 281.
Waldenstrom, J. (1935). Dt8ch. Arch. klin. Med. 178, 38.
Waldenstrom, J. (1936). Hoppe-Seyl. Z. 239, 111.
Waldenstrom, J., Fink, H. & Hoerburger, W. (1935).
Hoppe-Seyl. Z. 233, 1.
Watson, C. J., Schwartz, S. & Hawkinson, V. (1945). J. biol.
Chem. 157, 345.
Wells, G. & Rimington, C. (1953). Brit. J. Dcrm. (in the
Westall, R. G. (1952). Nature, Lond., 170, 614.
The Protein Components of the Isolated Myofibril
BY S. V. PERRY
Department of Biochemi8try, Univer8ity of Cambridge
(Received 10 February 1953)
The myofibrili s unique among the organized com-
ponents of cells in that it requires a common sub-
strate, adenosine triphosphate (ATP), for enzymic
and mechanical activity.
survive when the myofibril
from the cell, and although it is true that the main
protein component, actomyosin, has similar charac-
teristics, extraction of this complex completely
disorganizes the myofibrillar structure and results
in modification of both enzymic and, mechanical
Schick & Hass (1951) investigated the solubility
in salt solutions of varying ionic strength of myo-
fibrils obtained by a procedure which involved
tryptic digestion ofthe muscle tissue. These workers
used an arbitrary method of microscopic examnina-
tion to determine when solution was complete, but
in view of the pronounced effect oftrypsin onmyo-
sin (Gergely, 1950; Perry, 1951), and on proteins
Both these properties
is removed intact
generally, it is difficult to decide to what extent the
constituent proteins were modified during the pre-
paration of the myofibrils. Information about the
protein components of the myofibril other than
actomyosin is scanty; for example nothing is known
about the nature ofthe various banded features, all
of which must play a part in the function of the
contractile unit.Earlier electron-microscope in-
vestigations (Rosza, Szent-Gyorgyi & Wyckoff,
1950; Perry & Home, 1952) have provided evidence
for the washing out ofsome band components with
no apparent modification of the underlying basic
The present communication is concemed with
further investigations of the properties of myo-
fibrils prepared without the use of enzymes. In
particular, the nature of the protein components
extracted from the myofibril under varying ionic
conditions has been investigated.
PROTEINS OF THE MYOFIBRIL
Preparation of myofibrils. Myofibrils were prepared as
previously described (Perry, 1952), using as the medium for
suspension 0 078M-borate buffer, pH 7-1 (30 ml. 0*05m.
boraxand 970 ml. 0 2M-boricaciddilutedto2*5 1.; Palitzsch,
1922). In certain cases, to reduce the amount of protein
leaching out of the myofibrils, the preparation was carried
out in a solution of 0 025m-KC1 and 0-039m-borate buffer,
pH 7X1. All manipulations were carried out in the cold
Extractionofmyofibrils. Myofibrils weretreated at 0° with
the appropriate salt solutions so that extraction was carried
out atprotein concentrations ranging from 0 5 to 1.0mg./ml.
The original suspension was always in borate buffer and in
consequence a low concentration of the latter was usually
present during the first extraction. After a given extraction
time, the myofibrillar residues were sedimented by centri-
fugation at 00 for 10 min. at 100000g. To determine the
amount ofinsoluble fraction, the residue was resuspended in
salt solution and the extraction repeated 2-3 times. Finally
total N estimations were carried out on the appropriate
Viscosity measurements. The usual procedure was to mix
5 ml. of extract with 2 ml. of 0 16m-phosphate buffer,
pH 7 0, and take 3 ml. of this solution for viscosity deter-
mination at 0° in an Ostwald viscosimeter (time of flow for
water, 40-50 sec.). Total concentration of protein in the
viscosimeter was 3-4 mg./ml. ATP (0 03 ml.) was added in
the viscosimeter to give a final concentration of 0-0004m.
Adenosine triphosphata8e activity. The method was es-
sentially that described earlier (Perry, 1951); glycylglycine
buffer was used at pH 7-4 and glycine atpH 9 0.
Collagen estimation. Myofibrils were treated with 01
NaOH for 20 hr. at 150, centrifuged for 10 min. at 100000g
and N determinations carried out on the clear supernatant
and the original whole extract. Nitrogenous material not
rendered soluble by this procedure was considered to con-
sist of collagen and elastin (Lowry, Giligan & Katersky,
1941; Robinson, 1952).
Estimations. The acid-labileP contentofthemyofibril and
derived fractions was estimated as described in an earlier
communication (Perry, 1952). Total nucleic acid (including
phosphoprotein) P values were determined bythe method of
Schmidt & Thannhauser (1945), and N byamicro-Kjeldahl
technique. Protein concentration was arbitrarily taken as
6 x nitrogen concentration.
Materials. The ATP was prepared routinely in this
department by Mr E. J. Morgan. It was freed fromheavy-
metal impurities by treatment with 8-hydroxyquinoline,
stored as the barium salt and converted to the sodium salt
for use. Dr T.-C. Tsao kindly providedtheelectrophoretic-
ally pure, recrystallized sample oftropomyosin.
Electrophoresis. Experiments were carried out with the
usual Longsworth & McInnes Tiselius apparatus, and also
with the Perkin-Elmer model 38 instrument. Conductivities
were determined on the protein solutions as prepared for
Myofibril proteins soluble in 0-078M-borate, pH7-1
Freshly prepared myofibril suspensions in 0-078m-
borate buffer contained verylittle dissolvedprotein,
for the addition of 10% (w/v) trichloroacetic acid
(TCA) to an equal volume of the clear supernatant
obtainedwhenthesuspensions had beencentrifuged
for 10 min. at 10000 g, usually gave no detectable
turbidity. During storage, at 4° in the presence of
toluene, protein passed into solution with the result
that after 15 days 11-12% of the total nitrogen of
the suspension had become soluble.
storage under sterile conditions brought
additional nitrogen into solution, and dialysis ofthe
solution showed that it did not contain nitrogen
compounds of low molecular weight. This solution
ofmyofibrillar proteins is subsequently referred to
as the soluble fraction.
Age of preparation (days)
Fig. 1. Rate of solution of protein N on storing myofibriI
suspensions in 0-078m-borate buffer, pH 7-1, at 4°. The
results obtained with five preparations are shown, each
preparation being indicated by a different set of points.
Fig. 1 illustrates the results obtained with five
different myofibril preparations and shows that
afteracomparatively rapidsolution ofproteininthe
first day or two, the rate of extraction decreases
until after 15 days it is veryslow. Whenasuspension
ofmyofibrils, which had been stored until 11-2% of
the total protein had gone into solution, was centri-
fuged down and resuspended in fresh borate buffer,
the rate ofsolutionwas not increased. This indicates
that the process was not a distribution of myo-
fibrillar proteins between the gel and sol phases but
more soluble in the buffer than the bulk ofthe myo-
The soluble fraction was obtained as a clear,
slightly viscous solution of protein with minimum
in the range pH 4-85-4.
saturation with ammonium sulphate little protein
practically complete at 50% saturation.
This fraction showed certain properties which
suggested that it might consist, in part, of the
globulin X complex. According to Weber & Meyer
(1933) globulin X is precipitated at ionic strength
< 0-005 and there is evidence that in living muscle
S. V. PERRY
onlypart is in solution inthesarcoplasm. Dialysis of
the soluble fraction against distilled water adjusted
to pH 7, or against phosphate buffers containing
equimolar amounts of KH2PO4 and K2HPO4 and
ranging in total ionic strength from 0 001 to 0-1, did
not precipitate any significant amount of protein.
On this evidence it was concluded that the soluble
fraction did not contain globulin X.
Adeno8ine triphosphata8e activity of 8oluble frac-
tion. No adenosinetriphosphatase (ATPase) activity
was shown by the soluble fraction in the presence of
calcium ions at pH's 7-4 and 9-0, nor in the presence
of magnesium ions at pH 7-4. For example, when
12-14% of the total protein had leached out of the
myofibril, the whole suspension
activity which was entirely
localized in the readily centrifuged myofibrillar
residue. These facts indicate that myosin was not
a component of the soluble fraction.
Ab8ence of F-actin. Precipitation of the soluble
fraction with O-OlM-acetate buffer, pH 5-1, and
dissolution in a small volume ofdistilled water with
sufficient 0-1M-sodium bicarbonate to bring the pH
back to 7, gave a very viscous solution. This was not
F-actin as it did not form actomyosin on addition
to a solution of freshly prepared myosin in 0-5m-
potassium chloride, i.e. ATP did not produce a drop
in viscosity when added to the combined solutions.
Tropomyosin and the 8olublefraction. Addition of
potassium chloride to the soluble fraction, as
obtained fromthemyofibrils bycentrifugation, orto
the concentrated solution obtained after isoelectric
precipitation, gave a marked drop in viscosity
(Fig. 2). This salt effect was reversible and strongly
suggested the presence of tropomyosin, the only
protein so far isolated from muscle which exhibits
these properties and which is presumed to be myo-
fibrillar in origin on the basis of the procedure
employed for its extraction from washed muscle
mince (Bailey, 1948).
As usually obtained, the soluble fraction con-
tained 1-5-2-0 mg. protein/ml., and concentration
was necessary before electrophoretic investigation
could be undertaken. This was accomplished by
suspending the solution (containing 0-05M-phos-
phate buffer, pH 7-0) in a dialysis bag in front ofan
electric fan. When this process was carried out inthe
coldroom at 4° thevolume was reduced to one-sixth
in 36 hr. without any significant amount of de-
Examination of this concentrated
solution in the Tiselius electrophoresis apparatus
revealed two main peaks (Fig.
invariably seen, but in some preparations another
component, moving more slowly than the two main
peaks, and present in much smaller amount, could
be detected. The slower ofthetwomain components
was estimated to be present in twice the amount of
the faster. Addition of a sample of recrystallized
3). These were
tropomyosin to the soluble fraction did not result in
the appearance of a new peak but, as Fig. 3 clearly
shows, the fast component increased in amount
relative to the slower. This was considered as good
evidence for the identity ofthe fast component with
tropomyosin. In Table 1 are presented the mobili-
ties of the two main components of the soluble
KCI concentration (M)
Fig. 2. Effect of KCI on the viscosity of a solution of the
soluble fraction in 0-078m-borate buffer, pH 7-1. Protein
concentration approx. 0-5 mg./ml.
Fig. 3. Electrophoresis diagrams of the soluble fraction
extracted firom rabbit myofibrils, after 270min.
Soluble fraction; (b) soluble fraction with added tropo-
pH 7-1; 0-25M-NaCl.
On addition of potassium chloride to a final con-
centration of 0-5m the viscosity of the soluble
fraction dropped to a very low level (Fig. 2), and in
view ofthe electrophoretic evidence it was assumed
that at low salt concentrations the viscosity of the
extract was almost entirely due to the tropomyosin
present and that the main component contributed
PROTEINS OF THE MYOFIBRIL
little. On this asumption a value for the tropo-
myosin concentration of the soluble fraction was
obtained by comparing the reduction of viscosity
obtained on addition of 0-5M-potassium chloride
with that produced under identical conditions with
Table 1. Electrophoretic mobilitiea of component8
of the solublefraction of rabbitmyofibrils
(Electrophoresis carried out at 00
phosphate buffer, pH 6-5, and 0-22M-KCI. Values are the
average of two independent determinations.)
in 0 05M-sodium
a sample ofpure tropomyosin. Table 2, showing the
results of such determinations on several prepara-
tions, indicates that about 30% of the soluble
fraction consists of tropomyosin. This proportion
Phosphorus content of the solublefractwon. Hamoir
(1951) has shown that tropomyosin from carp and
rabbit muscle can form a crystallizable complex
with nucleic acid, but it is not yet clear whether this
protein exists in situ as such a complex or whether
the nucleic acid is picked up during the extraction
Consequently, it was of interest to
determine the amountofnucleic-acidphosphorus in
the soluble fraction, for under the conditions of
extraction tropomyosin is removed from the myo-
fibril without exposure to contamination by the
nucleic acid of intracellular components such as
nuclei and sarcosomes.
The soluble fraction contained variable amounts
of free inorganic phosphate which presumably had
leached out of the myofibril. Total nucleic-acid
phosphorus values also showed some variation, the
highest value obtained being 0-081 % (see Table 3).
If, in the extreme case, which is rather unlikely, all
this phosphorus occurred as nucleic acid associated
with tropomyosin in the soluble fraction, the com-
plex would contain 2-5-3-0% nucleic acid. Nucleo-
tropomyosin isolated by Hamoir (1951) from carp
Table 2. Tropomyosin extractedfrommyofibril8stored in 0-078M-borate buffer, pH 7-1
(as % soluble
stayed fairly constant during the extraction of the
soluble fraction, and the maximum amount of
tropomyosin extracted represented 4-0 % of the
Table 3. Nucleic-acid phosphorus content of
(All fractions were extracted with cold trichloroacetic
acid and fat solvents according to the procedure ofSchmidt
& Thannhauser (1945). The P figure represents the total
amount of ribonucleic and deoxyribonucleic acids and
phosphoprotein in the fraction analysed.)
Whole myofibrils after
removal of the soluble
total proteins of the myofibril. On the basis of his
organic solvent extraction method Bailey (1948)
estimated tropomyosin to account for 2 6% of the
total proteins of rabbit muscle, i.e. 4-0% of the
myofibril if the latter contains 65% of the total
protein of muscle.
muscle contained considerably more nucleic acid
Effect of low concentrations of potassium chloride
on the extraction ofthe solublefraction. To investigate
theeffect oflowconicentrations ofpotassiumchloride
on the extraction ofthe soluble fraction, samples of
Table 4. Protein passing into solution on storing
myofibrils in media of low ionic strength
Composition of medium
addition of TCA
to the supernatant
a freshly prepared myofibril, suspensionwere left for
15 days in various solutions, the compositions of
which are shown in Table 4. At the end of this
period, the suspensions were centrifugedfor 10 min.
at 10000 g and a sample ofeach supernatant added
S. V. PERRY
to an equal volume of 10% TCA. The relative
obtained, which represent
mately the relative amounts of protein which had
gone into solution, are shown in Table 4. It was
readily apparent that the minimum amount of
protein had leached out when the myofibrils were
stored in a solution containing 0-039M-borate
buffer, pH 7-1, and 0-025M-potassium chloride.
Fig. 4. Electronmicrographofrabbit myofibrils afterextrac-
tion for 17 days at 4° with 0-078M-borate buffer, pH 7-1.
Grold-palladium shadowed. Magnification, x 9250.
More precise investigation indicated that only 4%
ofthe total myofibrillar protein passed into solution
during storage for 15 days in this medium. This
solution was adopted for routine preparations when
it was necessary to reduce the leaching out ofmyo-
fibrillar protein to a minimum. It was noted that
myofibrils prepared in this manner sedimented
faster and formed a more compact mass in the
centrifuge tube than those prepared in 0-078M-
borate, and it was more difficult to free the prepara-
tions from nuclei and other rapidly sedimenting
cell components. During preparation in 0-078M-
borate the myofibrils became increasingly difficult
to centrifuge down at the speeds used (Perry, 1952)
and presumably in this buffer some hydration and
Myofibrils from which the soluble fraction had
been extracted were similar in general microscopic
appearance to the freshly prepared material.
methylene blue was used as an aid for more careful
observation, the contrast between the heavily
stained bands (which frequently appeared as a
double band, see below) and the lightly stained
bands was greater in the stored than in the fresh
preparations. This appearance ofincreased contrast
was possibly due to the fact that, comparedwith the
fresh material, myofibrils from which the soluble
fraction had been extracted stained very lightly
indeed in the less dense part of the structure.
Fig. 5. Electron micrograph of rabbit myofibrils after
30 days extraction with 0-078M-borate buffer, pH 7-1.
Gold-palladium shadowed. Magnification, x7600.
Electron-microscope examination revealed that
although there was evidence of longitudinal fila-
structures within the myofibril,
details could no longer be recognized. A character-
istic feature of these extracted myofibrils was the
appearance of regular dense bands (Figs. 4, 5). Two
ofthese bands occurred in each sarcomere, although
sometimes a pair was close enough together to
appear as one broader band. These bands are con-
sidered to correspond to the outer edges of the A
band or perhaps to the A substance which has
moved into the I band, for myofibrils isolated by the
method described are usually slightly contracted
(Perry, 1952). Features of the A band such as the
M line and H sub-lines (Perry & Home, 1952) were
never seen in electron microscope examinations of
myofibrils stored long enough to extract the soluble
fraction. Further work needs to be done to decide
the precise location within the myofibril of the
proteins of the soluble fraction.
Double refraction of the myofibril. Examination of
a number of myofibril preparations was kindly
carried out with the polarizing microscope by
PROTEINS OF THE MYOFIBRIL
Table 5. Effect ofATP on the viscosity of Weber's solution extracts of myofibril8
Dr M. Mitchinson ofthe Zoology Department. These
investigations were of a qualitative nature but
all preparations had pronounced
double refracting bands. No obvious diminution of
the relative intensity ofthese bands could be seen in
preparations which had been stored at 40 for 11-50
days. This suggests that the proteins of the soluble
fraction are not responsible for the birefringent
properties of the myofibril.
Protein8 extracted by salt solutions of
higher ionic strength (> 0-5)
> 0-5M extracted considerable amounts of protein
to give a viscous, sometimes slightly turbid, solu-
tion. These solutions were shown to contain acto-
myosin by the fall in viscosity produced on the
addition of low concentrations of ATP. Myosin
almost free from actin is obtained by short ex-
tractions of minced fresh whole muscle with 0-5M-
potassium chloride, whereas the isolated myofibrils
produced viscous actomyosin solutions immediately
they were treated with potassium chloride at this or
higher concentrations. In addition, a solution con-
taining 0-47M-potassium chloride,
pyrophosphate and 0-05M-phosphate buffer,pH 6-3,
such as was used by Hasselbach & Schneider (1951)
to extract myosin free from actin from whole
muscle, was found to extract considerable amounts
of actomyosin from the myofibril
Weber's solution (0-6M-potassium chloride, O-1M-
sodium carbonate, 0-04m-sodium bicarbonate) was
a very efficient agent for the extraction of acto-
myosin. This medium also extracted actomyosin
from myofibrils which had been stored for 13 days
and which had been washed free from the soluble
fraction (Table 5).
When the process of salt extraction of the myo-
fibrils was followed microscopically, cross striations
were seen to disappear, although the myofibrils
persisted in outline for some time, and finally the
residues or ghosts which remained could be centri-
fuged down by high-speed centrifugation. Lithium
,chloride, which Bate-Smith (1937) found to be an
efficient extractant for theproteinsofwholemuscle,
brought an average of 89 % ofthe totalmyofibrillar
protein into solution and less (82%) when used in
phosphate. Potassium chloride was less effective
than lithium chloride and similarly its power of
bringing myofibrillar protein into solution was
reduced by 0-067M-dipotassium hydrogen phos-
In general, both these salts
extracted rather more protein from myofibrils
which had been stored for some time under sterile
containing potassium iodide (Table 6), were the
most efficient in extracting protein from the myo-
fibril. Exhaustive extraction with these solutions
reduced the insoluble residue to 2-4% of the total
nitrogen of the myofibril.
Salt extraction, other than with potassium iodide
and Weber's solutions, invariably left a residue
which centrifuged down as a jelly-like mass and
which was shown by electron microscope examina-
tion to be composed mainly of fibrous material
which could not be characterized. No myofibrils
could be recognized, however, if the extraction had
Although potassium chloride readily extracted
actomyosin, the process was by no means complete;
even after fifteen extractions with M-potassium
chloride the residue still showed myosin-ATPase
activity. Lithium chloride wasmuch moreeffective,
for a short extraction with this salt brought the
whole of the myosin-ATPase into solution. The
averaging 11 % of the nitrogen of the whole myo-
fibril, contained 16-7 % total nitrogen and 0-14%
The nature of this fraction is unknown, but it
probably consists of components of the banded
structures ofthe myofibril, e.g. from theAband and
the Z membrane. The presence of actin in this
residue seems unlikely, as lithium chloride brings
into solution all the labile phosphorus, and therefore
presumably all oftheactin ofthe myofibril (Table 7).
If the residue was treated with acetone and pro-
ceeded with as for the preparation of actin (Straub,
1942), none of this protein could be prepared.
Attempts to prepare actin from fresh myofibrils by
the Straub method were uniformly unsuccessful. It
or salt solutions
S. V. PERRY
Table 6. Protein extractedfrom myofibria8 by 8alt 8Olutions of higher ionic 8trength
0*47M-KCI, 0 01M-Na2P2O7,
0 05M-phosphate buffer, pH 6-23
No. of times
1 (2 hr.)
1 (3 days)
1 (20 hr.)
i1 (2 hr.)
1 (22 hr.)
Table 7. Effect of salt extraction on the bound nucleotideofthe myofibril
(The extractions were carried ouit at 00 and the residue separated and washed with extractant once or twice by centrifu-
gation for 10 min. at 10 000g.)
(mg. 10.min. P/mi.) (mg. 10-min. P/ml.)
appeared that most of the actin is removed by the
salt extraction of the myofibrils in the first stage of
the actin preparation. This is yet another example
of the differences in extractability of the proteins
actin and myosin in the isolated myofibril as com-
pared with fresh minced muscle.
Collagen content of myofibril preparations. Under
the conditions ofestimation a very small amount of
material remained insoluble, but as the nitrogen of
myofibrils prepared both by the collagenase and the
non-enzymic methods was rendered completely
soluble by 0-1N-sodium hydroxide the preparations
were considered free from collagen and elastin.
The labile phosphorus of the myofibril
In an earlier communication (Perry, 1952) it was
shown that nucleotide, mainly in the form ofadeno-
sine diphosphate, was closely associated with the
myofibril and was responsible for the small acid-
1-25M-KCI, 1 hr.
1-25M-KCI, 2 hr.
0-90M-LiC1, 2 hr.
I OM-KCl, 1 hr.
0-90M-LiCl, 1 hr.
Soluble fraction removed and
10-min. P estimated; residue
labile phosphorus content of these preparations.
On storage of myofibril suspension in borate buffer
this labile phosphorus remained constant inamount
and was invariably associated with the insoluble,
myofibril fraction. Table 7 shows that even after
storage until 13 % ofthe total protein had gone into
solution no significant amount of labile phosphorus
could be detected in the soluble fraction. This labile
phosphorus was stable in the presence of lithium
chloride and potassium chloride, although the
extraction of the myofibril by these salts did affect
its distribution between the soluble and the in-
effective in extracting the bound nucleotide. Treat-
ment for aperiod of 1 hr. at 00 was adequate to bring
85-100% of the labile phosphorus into solution,
whereas potassium chloride was less effective and in
various experiments under similar conditions with
this salt, 25-80% was extracted.
PROTEINS OF THE MYOFIBRIL
One of the objects of the present investigation was
to obtain a quantitative assessment of the various
proteins present in the myofibril isolated free from
methods which have been used for such studies on
(1951), could not be satisfactorily applied to the
isolated myofibril and this aim has not yet been
achieved. Whereas myosin practically free from
actin can be obtained from fresh whole muscle
mince without much difficulty, extraction of the
isolated myofibrilundersimilar conditions produced
actomyosin rich in actin.
The conventional method of actomyosin pre-
paration involves the prolonged extraction ofwhole
muscle and produces a very impure product. As
a starting material for actomyosin preparation,
isolated myofibrils have many advantages; for
example, one short extraction produces an acto-
myosin completely free from sarcoplasmic proteins.
In fact, if methods can be devised for obtaining
myosm and actin independently from the isolated
myofibrils, the preparation of these proteins in the
purified form would be considerably simplified.
It is generally assumed that the presence ofATP
in the muscle cell is necessary for the extraction of
myosin free from actin. This explanation follows
naturally from the behaviour of ATP with the
isolated proteins, but it is not entirely satisfactory
as the endogenous ATP is rapidly broken down in
muscle homogenates (Bailey & Marsh, 1952).
may be that in the cell myosin can only be extracted
without appreciable contamination with actinwhen
active phosphorylation systems
maintain the bound nucleotide of the myofibril
(Perry, 1952) in the maximally phosphorylated
state, i.e. mainly as ATP. This state ofaffairs would
be characteristic of relaxed muscle or of muscle
which has been freshly minced, where the oxidative
and glycolytic systems are still directed into the
production of energy-rich phosphate. It would not
apply to muscle which has gone into rigor or to the
isolated myofibril; indeed, in the latter case, the
diphosphate. The nature of the mechanisms in-
volved in the phosphorylation ofthe bound nucleo-
tide is now under investigation.
Of the two main components which passed into
solution on continued extraction of myofibrils with
solutions of low ionic strength, one has unequi-
vocally been shown to be tropomyosin.
nucleic acid is associated with this fraction and, as it
is unlikely that extraction with 0O078x-borate
buffer, pH 7-1, would dissociate a tropomyosin-
nucleic acid complex, it is reasonable to conclude
that tropomyosin does not occur in the myofibril as
see Hasselbach & Schneider
are working to
a complex containing large amounts ofnucleic acid
such as has been isolated by Hamoir (1951) from
carp muscle. The other component which was
extracted with tropomyosin has not yet been
identified. It was clearly neither myosin nor the
normal form of G-actin. The absence of actin from
the soluble fraction was further suggested by the
fact that no labile phosphorus appeared in this
fraction during the extraction process. The possi-
bility cannot be excluded that this unidentified
protein was a form of inactive or denatured actin,
but the fact that it was always found in constant
argues against this explanation. Itmay be that this
component is identical with that protein invariably
found in Straub's actin preparations and which has
been assumed, without any clear reason, to be
Repeated extraction with neutral salts such as
potassium chloride or lithium chloride failed to
bring 9-17 % of the total myofibrillar nitrogen into
solution, but if Weber's or potassium iodide solu-
tions were used 96-98% of the nitrogen became
soluble, and indicated that the myofibril must con-
tribute little if anything to the so-called stroma
protein which accounts for 17% of the total nitro-
gen of rabbit white muscle and 27% in the red
variety (Weber & Meyer, 1933; cf. Hasselbach &
Schneider, 1951). The sarcoplasmic granules, or
sarcosomes, which are very resistant to agents of
solution and not easily removed from the muscle
residue, probably contribute to the insoluble stroma
fraction. In mixed rabbit muscle the sarcosomes
make 3-4% of the total muscle protein nitrogen,
but in red muscle such as that obtained from pigeon
breast, the figure is considerably.higher (Chappell,
1952). This fact might account, in part at least, for
the higher stroma values for red muscle. In rabbit
white musclethebulk ofthestromaproteinmust be
derived from insoluble nuclear residues, the sarco-
lemmaand extracellular tissue. It isverylikelythat
part of this stroma fraction consists of protein
which has been denatured during the attainment of
the degree of conmninution necessary for adequate
extraction. Disintegration, particularly in media of
low ionic strength, can produce very appreciable
denaturation ofthe myofibrillar components.
1. On prolonged extraction of isolated rabbit-
musclemyofibrilswith 0-078 M-borate buffer,pH 7-1,
approximately 13% of the myofibrillar protein
passed into solution.
2. This fraction soluble in 0078m-borate buffer
contained two main components, one of which was
tropomyosin, whereas the other has not yet been
identified. Proteins shown not to be present in the
S. V. PERRY
soluble fraction were globulin X,.myosin, F-actin
3. The bound nucleotide of the myofibril re-
mained associated with the myofibril on prolonged
extraction with 0 078M-borate at pH 7-1, but
passed into solutiononextractionwith saltsolutions
of higher ionic strength.
4. Exhaustive extraction with potassium chloride
and lithium chloride failed to dissolve 9-17 % of
the myofibrillar protein, whereas extraction with
Weber's and potassium iodide solutions reduced the
insoluble residue to 2-4%.
TheauthorwishestoexpresshisthankstoMrR. W. Home
for taking the electron micrographs and to Dr V. E.
Cosslett for making available the facilities of the Electron
Microscope Section, Cavendish Laboratory.
Bailey, K. (1948). Biochem. J. 48, 271.
Bailey, K. & Marsh, B. B. (1952). Biochim. biophy8. Acta, 9,
Bate-Smith, E. C. (1937). Proc. Roy. Soc. B, 124,136.
4Chappell, J. B. (1952). Personal communication.
Gergely, J. (1950). Fed. Proc. 9, 176.
Hamoir, G. (1951). Biochem. J. 48, 146.
Hasselbach, W. & Schneider, G. (1951). Biochem. Z. 321,
Lowry, 0. H., Gilligan, D. R. & Katersky, E. M. (1941).
J. biol. Chem. 189, 795.
Palitzsch, S. (1922).
Bull. Inst. Oc4snogr. Monaco, no.
Perry, S. V. (1951). Biochem. J. 48, 257.
Perry, S. V. (1952). Biochem. J. 51, 495.
Perry, S. V. & Home, R. W. (1952). Biochim. biophy8. Acta,
Robinson, D. R. (1952). Biochem. J. 52, 621.
Rosza, G., Szent-Gyorgi, A. & Wyckoff, R. W. G. (1950).
Exp. Cell. Re8. 1, 194.
Schmidt, G. & Thannhauser, S. J. (1945). J. biol. Chem. 161,
Schick, A. F. & Hass, G. M. (1951). J. exp. Med. 91, 655.
Straub, F. B. (1942). Stud. Int. med. Chem. Univ.Szeged.2,
Weber, H. H. & Meyer, K. (1933). Biochem. Z. 266, 137.
A New Colorimetric Reagent for Carbohydrates
BY E. LUNT AND D. SUTCLIFFE
Department ofExperimental Pathology, Univer8ity ofBirmingham
(Received 18 December 1952)
Methods for the estimation of carbohydrates, and
particularly polysaccharides, have been somewhat
tedious till the introduction by Dreywood (1946)
and Morris (1948) of the reagent anthrone in con-
centrated sulphuric acid. This provided a sensitive,
facile andgeneralmethod forcarbohydrate analysis,
since simple dilution of the reagent (2 volumes) by
the test sample (1 volume) produces a strongly acid
medium sufficiently hot to hydrolyse a polysac-
charide, which then gives a colour reaction with the
anthrone equivalent to that of its constituent units.
Thus the reagent may be calibrated against stan-
dards of the constituent monosaccharide, since the
colorimetric yield is known to be quantitative.
Unfortunately the anthrone reagent is somewhat
unstable, necessitating frequent preparations, in-
cluding standards for calibration. A search was
therefore made for a similar but stable reagent.
Various simple polyphenols, substituted phenols
and aromatic amines were tried, but those which
were chromogenic were also unstable in sulphuric
acid. In an endeavour to stabilize resoreinol, one of
the most promising reagents when used in place
of anthrone, the hydroxyl groups were protected
by benzoylation. When resorcinol dibenzoate in
ethanolic solution and sulphuric acid were added in
succession to test samples, the benzoyl groups were
removed by hydrolysis with the development of
a satisfactory colour, but benzoic acid crystallized
out on cooling. The ditoluene-p-sulphonate proved
resistant to hydrolysis andgave scarcelyany colour.
0-Substitution having failed, the nuclear-substi-
(RDSA) was prepared, and found to be stable in
aqueous solution, although insoluble in concen-
trated sulphuric acid. The disulphonic acid was
about as sensitive a colorimetric reagent as the
parent phenol, giving an intense golden-brown
colour with carbohydrates. This reaction has been
investigated so as to provide a reliable and simple
method for carbohydrate estimation which would
be useful when only very small volumes or very
dilute solutions were available for analysis. Under
the conditions of the reaction all the isomeric
hexoses tested gave a peak at 490mp.,but the
sensitivity varied with the
(Fig. 1). Forany series ofdeterminations, therefore,
the appropriate hexose must be used as standard