DERIVATIVES

(From

OF KERATIN

BY DAVID R. GODDARD*
AND
the Laboratories
of The Rockefeller
New

LEONOR
Institute

MICHAELIS
for

Medical

Research,

York)

(Received for publication,

July 5, 1935)

A-SH

+ R-X

--+ A-S-R

+ HX

This reaction occurs with great easeat neutral to mildly alkaline

reaction, and has already been applied to proteins for different
reasonsby Mirsky and Anson (4) and Goddard and Schubert (5).
Although organic halogen compounds can react with amino groups
according to Michaelis and Schubert (6), we will show that it has
been possible under certain conditions to substitute completely
* Sational

Research Fellow in the Biological

361

Sciences.

Downloaded from http://www.jbc.org/ by guest on July 3, 2016

In an earlier paper the authors (1) expressed the view that keratins are fibrous proteins whose characteristic properties are essentially determined by the S-S groups of cystine which act as very
firmly established cross links uniting the elementary fibers of polypeptide chains. This view was based on the action of certain
alkaline reductants which could be shown to reduce disulfides and
convert keratin into an amorphous protein soluble in weak alkalies,
digestible by true proteases, and in which the sulfur is in the sulfhydryl state. The important r81eof the disulfide bond had already
been emphasized by Speakman and Huist (2) and A&bury (3) on
the basis of their chemical and physical studies, including x-ray
diffraction patterns of keratins.
Since keratins have a very high percentage of disulfide sulfur
(10 to 15 per cent cystine) and may readily be reduced to sulfhydryl proteins by alkaline thioglycolate, we realized that this
reduced protein might be a useful material for study as an example
of a sulfhydryl protein. We were particularly interested in studying the properties of derived proteins formed by substitution in
the sulfhydryl group by reaction with organic halogen compounds
according to the following scheme.

362

Derivatives

of Keratin

Downloaded from http://www.jbc.org/ by guest on July 3, 2016

the sulfhydryl groups of the protein of reduced wool keratin, without substituting amino groups.
The proteins studied are the sulfhydryl protein obtained by the
alkaline thioglycolate reduction of wool, this protein after it has
been reoxidized to the disulfide form, and the substituted proteins
formed by reaction of the sulfhydryl protein with iodoacetic acid,
iodoacetamide, iodoethyl alcohol, and cu-bromopropionic acid, at
mildly alkaline reaction.
The disulfide protein may be formed by
reoxidation of the sulfhydryl protein at neutral reaction by the air
or ferricyanide.
As to the nomenclature of these proteins, we may
designate the reduced keratin as kerateine, and the reoxidized,
amorphous kerateine as metakeratin.
By reaction of kerateine
with iodoacetate we obtained a carboxymethylkerateine;
with
cu-bromopropionic acid, cY-carboxyethylkerateine;
with iodoethyl
alcohol, hydroxyethylkerateine;
with iodoacetamide,
carbamylmethylkerateine.
None of the derived keratins has been obtained in a strictly pure
state, nor may they be considered as chemical entities.
It is
not even likely that the keratin of a single hair is a homogeneous
chemical substance. Furthermore,
it should be borne in mind
that the reduction of the disulfide group by thioglycolate
is a
reversible reaction, and its completeness depends on the excess of
the reductant applied, and that the reaction with the halogen compounds can proceed only to that extent to which the S-S group has
been previously reduced. In the preparation of a substituted
protein the non-reduced, non-substituted
residue can be determined as cystine by applying the Folin-Marenzi
method to the
hpdrolysate.
The substituted sulfur does not react in this method.
The analytical data for these proteins and wool are presented
in Table I. It will be seen that the S and N are similar to the
values of wool, and secondly, that nearly all the cystine has been
reduced and substituted.
Proceeding to the isoelectric points and solubilities of these
derived proteins, it should be interesting to compare native wool
with the derived proteins in this respect. There is no agreement
as to the isoelectric point of native wool. Elijd and Silva (7), from
swelling experiments and changes in the pH of solutions in which
wool was allowed to come to equilibrium,
determined the isoelectric point to be 4.9. This is in agreement with the recent result

3.31

Carboxymethylkerateine, unfractionatedt
Carboxymethylkerateine, Fraction A
Carboxymethylkerateine, Fraction B
Carboxymethylkerateinet
Carbamylmethylkerateine
a-carboxyethylkerateine

12.2, 12.1

16.50
16.54

0.75
0.36

17.35 D 2.74
2.40

1.35

2.66

15.34

15.50
15.30

0.46

1.73

15.90

0.71

0.26

0.97

14.95

0.21

0.38

3.25

3.24

3.12

( 2ystine
N

3.3-3.6

0.80, 0.76

0.76

3.e3.3

0.50

4.54.9

3.1-3.7

5.0-5.3

3.74.3

3.8-4.3

4.5-4.7

4.6-4.9

4.9(?)

Solubifity

Notice the conspicuous loss of

Soluble in 0.1 M Na2C03 or NH,OH,

insoluble in M/~O HCI or 0.1 M
sodium acetate

Soluble inO. M NazCOJ and NHdOH;

insoluble in N/30 HCl
Soluble in 0.1 M PITa& and NHdOH;
insoluble in ~/30 HCl
Soluble in 0.1 M sodium acetate or
N/m HCl
Soluble in 0.1 M sodium acetate and
$30 HCl
Soluble in 0.1 M sodium lactate and
lactic acid
Soluble in 0.1 M sodium acetate and
M/SO HCl
Insoluble in sodium acetate; soluble
in aa/3 HCl
Soluble in 0.1 M sodium acetate and
~130 HCl

in M NaOH or HCl

Points)

Insoluble

of Isoelechc

kmelectric
point

Column

0.81,0.78

0.81,0.79

Amino

I
Cent, Except

TABLE

0.80

(Per

16.32

1.42

12.2*

Cystine

16.50

11.6

16.29

1rotal

Derivatives

* Total cystine = 12.2; of this 88 per cent was in the form of cysteine.
t This was not the preparation from which Fractions A and B were prepared.
$ This preparation was allowed to react with iodoacetate for 40 hours at pH 9.0 to 9.5.
sulfur.
$ The increased N content is due to the amide N of iodoacetamide.

Hydroxyethylkerateine

3.33
3.29
4.50
4.55
2.00
2.00
2.78
2.34
2.12
2.15
2.17
2.73
2.76
2.77

3.37

Metakeratin

Kerateine

3.35
3.33
3.34

Native wool

of Keratin

jr0td s

Analysis
-

Downloaded from http://www.jbc.org/ by guest on July 3, 2016

St
K;
K
e
El;

.F

&
is&
P
K

F
0

364

Derivatives

of Keratin

Downloaded from http://www.jbc.org/ by guest on July 3, 2016

of Dumanski
and Dumanski
(8) of 4.9 obtained by deflections of
wool fibers in an electrical field. Speakman (2) denies that wool
has any distinct isoelectric point and acknowledges
only a wide
isoelectric zone of 5.0 to 7.0. Harris (9) has determined the isoelectric point of wool as 3.4 by cataphoresis
of wool particles.
This disagreement makes the comparison with the derived proteins
difficult.
In these derived proteins there is no difficulty in determining the isoelectric point with the flocculation method (10).
Kerateine and metakeratin
prepared from wool have isoelectric
points in the region 4.6 to 4.7. Carboxymethylkerateine
and
a-carboxyethylkerateine
have much lower isoelectric points, as
would be expected from introduction
of the carboxyl groups.
Carbamylmethylkerateine
has a slightly higher isoelectric point,
while hydroxyethylkerateine
has the same isoelectric point as
kerateine.
The change in isoelectric points and solubilities
of
these derived proteins are what would be expected from the chemical nature of the substituted
groups.
A comparison of the solubilities of these compounds shows the
following features.
Wool is insoluble in all acids and bases except
in so far as it is hydrolyzed.
Wool dissolves readily in alkaline
solutions of NaCN, Na&, and thioglycolate
(l), but in all these
Metakeratin
cases the fibrous pattern is irreversibly
destroyed.
and kerateine of wool are dissolved readily in weak (0.1 M) NaOH,
Na2C03, and NH,OH.
They are insoluble in sodium acetate,
water, and neutral salts, and slightly soluble in dilute HCl.
Stable
neutral solutions containing several gm. of metakeratin
in 100 cc.
may be obtained by dialyzing the alkaline solution.
These solutions are precipitated by traces of acetic or other acids, and once
precipitated are as difficulty soluble as the undialyzed metakeratin.
Carboxymethylkerateine
and carboxyethylkerateine
are very
readily soluble not only in solutions of NaHC03 and dilute HCl,
but even in a solution of sodium acetate.
Carbamylmethylkerateine has the same solubility in alkalies as kerateine, but is much
more soluble in dilut,e HCI than is kerateine.
The solubility of
hydroxyethylkerateinc
is similar to kerateine.
Determinations
of free amino N of some of our preparations were
performed by the method of Van Slyke (11) in.the constant volume
apparatus.
The main purpose of this analysis was to show
whether under the conditions of our work a substitution
of the H

D. R. Goddard

and L. Michaelis

365

Downloaded from http://www.jbc.org/ by guest on July 3, 2016

atoms of the amino group occurred by reaction of kerateine with

iodoacetate, which could be imagined according to t.he results of
Michaelis and Schubert with simple amino acids (6). The figures
show definitely that no detectable substitution
of amino groups
occurred, even in a period of 40 hours at a pH about 9.0 to 9.5.
This is of interest to the general problem as to the point of attack
of iodoacetate on the enzymes which it poisons.
As regards the percentage of amino nitrogen, Van Slyke and
Birchard (12) found that in several proteins the free amino N was
equal to half the lysine nitrogen, with the exception of gliadin in
which it is much higher.
The free amino N, as determined by the
Van Slyke met,hod on metakeratin and several derived kerateines,
is somewhat higher than would be expected from the published
values of the lysine content of wool of 2.2 (13), 2.3 (14), and 2.8 (15)
per cent. If we take the nitrogen content of wool as 16.5, then the
percentage of free amino nitrogen referred to total nitrogen should
be 1.8 to 2.3 per cent, depending upon the value assumed to be the
correct lysine value. We find t,hat 4.7 to 4.9 per cent of the total N
is free amino N in our preparations.
It is not likely that the whole substance of a hair, even disregarding any pigment, represents a single homogeneous protein.
The
usual method of separating a natural protein mixture into its component parts is the fractional precipitation
with salts.
Such a
method obviously cannot be used for an insoluble protein such as
keratin.
Even for kerateine and metakeratin
the solubility is too
low for an attempt of fract.ionation.
But the much wider range
of solubility of some of the derived proteins makes possible fractional precipitation.
Carboxymethylkerateine
was fractionated
with ammonium sulfate from solution of the protein in 0.1 M
sodium acetate.
The fractionation
led to two fractions;
one
which contained the major bulk of the original protein and had
a similar content of S, N, and amino N as the original protein; and
a second sniall fraction, Fraction B, which showed higher S and
lower N and amino N content.
Fraction A is completely precipitated from solution by 35 per cent saturation with (NHJzSO,
and gives a compact, cheese-like precipitate.
This fraction is
completely insoluble at its isoelectric point, it does not dissolve
appreciably in dilute lactic acid or sodium lactate, but does dissolve readily in dilute sodium acetate or dilute HCl.
The other

Derivatives

of Keratin

EXPERIMENTAL

The Parent
Protein-For
the preparation
of the substituted
proteins a single batch of wool protein, essentially metakeratin
still containing some unoxidized kerateine, was used. This protein is spoken of as the parent
protein.
It was prepared essentially as described by Goddard and Michaelis (1).
100 gm. of
defatted but otherwise
native, untreated wool are dissolved by
mechanical shaking in 2 liters of 0.5 M disodium thioglycolate
(if

Downloaded from http://www.jbc.org/ by guest on July 3, 2016

fraction, Fraction B, is soluble in 35 per cent saturated (NH&S04,

but is completely precipitated
as a very sticky, non-flocculent
precipitate
at 60 per cent saturation.
It is not completely insoluble even at its isoelectric point and readily dissolves in an
excess of dilute lactic acid or even in sodium lactate.
The high
sulfur content of Fraction
B indicates that the native keratin
probably consists of two fractions, and not that the substitution
leads to two different products.
One of the characteristic
properties of keratin is its complete
resistance to digestion by trypsin and pepsin.
But all of the
proteins prepared from wool, either as sulfhydryl,
disulfide, or
substituted
proteins, are as readily digested by trypsin and pepsin
as are typical proteins.
The resistance of keratin to digestion is
not dependent upon the disulfide state as such, since metakeratin
is readily digested, but upon the fibrous property of the keratin,
as brought about by the spatial arrangement of the S-S bonds.
Once this bond has been broken, the protein becomes digestible
by true proteases no matter what other reactions the sulfur may
undergo.
As regards the oxidation and reduction of S of these proteins,
native keratin may only be reduced at a pH of 10 or higher, but
metakeratin
is reduced by the same agents at a pH of 7 to 8.
Kerateine is readily oxidized to metakeratin,
even dialysis of the
protein suspended in acetate buffer at pH 4.6 leads to nearly
complete oxidation in 3 days.
In an alkaline solution oxidation
is much more rapid.
Within the short time required to dissolve
kerateine in ammonia, to filter, and to reprecipitate
(20 to 30
minutes), complete oxidation takes place. Such a high degree of
autoxidizibility
is as a rule not encountered in other sulfhydryl
proteins according to A. E. Mirsky
(personal communication).

D. R. Goddard and L. Michaelis

367

Mirsky and Anson (4) modification

of the Folin-Marenzi

method,

contained 12.3 per cent of total cystine (cystine + cysteine) and

3.17 per cent of cysteine. So, the sulfur, which had been practically completely reduced, has been reoxidized during the preparation to 75 per cent.
Kerateine-10
gm. of parent protein were freshly reduced under
oxygen-free nitrogen with 4.65 gm. of thioglycolic acid at a pH
just pink to phenolphthalein
for 2 hours. This protein was then
precipitated with trichloroacetic acid, collected on the centrifuge,
washed with acetone, transferred to a mortar, and ground three
times with acetone and three times with acid acetone, and dried
in a vacuum desiccator.
Metakeratin-The
parent protein, as stated above, is a mixture
of kerateine and metakeratin.
An attempt to convert this completely to metakeratin by oxidation with 3 per cent HzOl at a
pB of about 3.5 led to a loss of cystine, and on acidification H2S
could be smelled. The cystine content before oxidation of the
protein was 12.27 and after 10.51 per cent.
Metakeratin
without loss of sulfur may be prepared from
kerateine by oxidation in mildly alkaline solution with potassium
15 gm. of parent protein were suspended in 300 cc.
ferricyanide.
1 The c,ystine content here includes the cysteine, for the protein hydrolysate is oxidized, while still acid, with 3 per cent HZOn. Then the solution is
reduced with sulfite and analyzed by the Folin-Marenzi procedure.

Downloaded from http://www.jbc.org/ by guest on July 3, 2016

the pH is maintained between 11 to 12, this takes about 3 hours).

The undissolved residue is removed by centrifugation followed by
filtration, the filtrate is precipitated with acetic acid, collected on
the centrifuge, washed with acetone, and ground in a mortar
with acetone. After freeing from acetone by a vacuum, the
protein suspension is dialyzed in cellophane tubes until free of
salts and thioglycolic acid. The dialyzed protein is washed with
acetone and dried in a vacuum desiccator.
It is obtained as a
fine white powder containing about 4 per cent moisture, nearly
insoluble in water, but slightly soluble in acids, and very (but
slowly) soluble in alkali. This powder has essentially the same
content of nitrogen, sulfur, and cystine as the wool from which
it is prepared.
It becomes to a great extent reoxidized on prepaA
typical
preparation, hydrolyzed and analyzed by the
ration.

368

Derivatives

of Keratin

Downloaded from http://www.jbc.org/ by guest on July 3, 2016

of Hz0 and M NaOH was added with constant stirring until a

faint pink color developed with phenolphthalein;
after some time
the protein was completely dissolved.
0.2 M potassium ferricyanide was added with a burette until the yellow color did not
disappear instantly.
The protein was dialyzed until acid to
litmus; this gave an opalescent solution free of precipitate, but
easily coagulated by traces of acetic acid. The protein was
precipitated
with dilute acetic acid, washed with acetone, and
dried in a vacuum desiccator.
Carboxymethylkerateine
and Its Fractionation-50
gm. of parent
protein were suspended in 1000 cc. of water, 20 cc. of M NaOH
added, and nitrogen led through
the mixture.
32.5 gm. of
thioglycolic acid were neutralized to phenol red and 250 cc. of
3.4 M phosphate buffer, pH about 7.4, added to it; this solution
was added to the protein.
After 3 hours 75 gm. of recrystallized
iodoacetic acid, neutralized to phenol red, were added to 250 cc.
of the same buffer, and added to the protein-thioglycolate
solution.
After 2 hours the protein was precipitated by adding enough
solid (NH&S04
to give a 50 per cent saturated solution.
The
protein was filtered off, washed with 50 per cent saturated
(NH*)&SOd, and redissolved in 1500 cc. of Hz0 + 20 cc. of M
NaOH.
It was again reduced with thioglycolic acid and allowed
to react with iodoacetic acid exactly as in the procedure above.
The protein was precipitated
with (NH&SO4
at 50 per cent
saturation
and collected on a centrifuge, and the supernatant
liquid was discarded.
This protein was then dissolved in 1200 cc.
of 0.1 M sodium acetate and fractionated
by precipitation
at
35 per cent saturation
with (NH&SOa.
A second fraction was
obtained from the filtrate by60 per cent saturat,ion with (NH&S04.
The fractions were purified by repeating this procedure several
times.
The first fraction, Fraction A, represents the bulk of the
material; the final yield was 30 gm. of dry protein.
The other
fraction, Fraction B, gave a yield of only 2 gm. of dry protein.
Another preparation
of carboxymethylkerateine
was prepared
from freshly reduced parent protein by treatment
with iodoacetate at pH 9.0 to 9.5 for 40 hours, the iodoacetate being added
in intervals and to a large excess, in an attempt to substitute
the
amino groups as well as the -SH
groups.
The analysis shows
that no detectable substitution
of the free amino groups occurred,

D. R. Goddard

and L. Michaelis

369

Downloaded from http://www.jbc.org/ by guest on July 3, 2016

and that the long alkaline treatment caused a loss in sulfur, as

is shown in Table I.
a-Carboxyethyl-,
Carbamylmethyl-,
and HydroxyethylkerateineThese three preparations
were prepared in almost identical manner. 10 gm. of parent protein were suspended in 100 cc. of HzO,
for each preparation,
and sufficient M NaOH to give a light pink
color with phenolphthalein,
without
att.empting to dissolve the
protein completely.
4.65 gm. of thioglycolic acid neutralized to
phenolphthalein
were added to each, and the flasks stood for
3 hours under a stream of oxygen-free
nitrogen.
Then either
9.1 gm. of cr-bromopropionic
acid (neutralized to phenolphthalein)
or 11.0 gm. of iodoacetamide, or 9.5 gm. of freshly distilled iodoethyl alcohol were added. From time to time additional NaOH
was added to maintain the pH at a light pink to phenolphthalein.
After 2 hours the preparations
made with iodoacetamide
and
iodoethyl alcohol were precipitated with acetic acid, washed with
acetone, and then the suspensions dialyzed against distilled Hz0
for 3 days.
The preparation
with cY-bromopropionic
acid stood
overnight, was reduced in a similar manner as above, and treated
The protein
a second time with bromopropionic
acid for 4 hours.
was then dialyzed for 3 days, precipitated
with trichloroacetic
acid, washed with acetone, and dried in a vacuum desiccator.
The analytical results are listed in Table I.
Isoelectric Points-The
isoelectric points were estimated by
the method of Michaelis and Rona (10). The proteins with the
higher isoelectric points were dissolved in 0.1 M sodium acetate and
precipitated with 0.1 M acetic acid (final concentration
of acetate
ion &s 0.01 M), while the proteins of the isoelectric zone not
covered by the pH range of acetate buffer were dissolved in 0.1 M
sodium lactate and precipitated
with lactic acid. pH of the
mixtures was determined with the glass electrode.
Digestion-Purified
trypsin
(16) and Fairchilds
pepsin were
used. The soluble proteins were digested with great rapidity
(no precipitate
with trichloroacetic
acid or sulfosalicylic
acid
after 10 to 15 minutes at 37).
Those proteins not entirely
soluble at this pH were digested more slowly but at a rate comparable to other coagulated proteins.
Methods of Analysis-All
the proteins were analyzed for nitrogen
by Teorells (17) micro-Kjeldahl
method, for sulfur by the method

370

Derivatives

of Keratin

SUMMARY

Keratin of wool is reduced by sodium thioglycolate to kerateine.

This can be oxidized to metakeratin
which differs from native
keratin by its amorphous character, solubility in alkali, and by
its digestibility
with pepsin or trypsin.
The H of the -SH group of kerateine was substituted by treatment with iodoacetate, cr-bromopropionate,
iodoacetamide,
and
iodoethyl alcohol. The derived proteins thus obtained differ
distinctly in their solubilities and in their isoelectric points. They
are all digested by pepsin or trypsin.
No detectable substitution
of amino hydrogen occurred during the treatment,
and under
proper treatment no loss of sulfur occurred. The derivative
obtained with iodoacetic acid is soluble enough to allow the attempt of a fractionation with ammonium
sulfate. This led to
two fractions differing widely in solubility and sulfur content.
BIBLIOGRAPHY

1.
2.
3.
4.
5.
6.
7.
8.
9.
10.

Goddard, D. R., and Michaelis, L., J. Biol. Chem., 166,605 (1934).

Speakman, J. B., and Huist, M. C., Tr. Faraday
Sot., 29,148 (1933).
Astbury, W. T., Tr. Faraday Sot., 29,193 (1933).
Mirsky, A. E., and Anson, M. L., J. Gen. Physiol., 18,308 (1935).
Goddard, D. R., and Schubert, M. P., Biochem. J., 29, 1009 (1935).
Michaelis, L., and Schubert, M. P., J. Biol. Chem., 106, 331 (1934).
Eliid, E., and Silva, E., 2. physik. Chem., 137, 142 (1923).
Dumanski, A., and Dumanski, 0. A., Kolloid-Z.,
66,24 (1934).
Harris, M., Bur. Standards
J. Research, 8,779 (1932).
Michaelis, L., and Rona, P., Biochem. Z.. 97,38 (1910).

Downloaded from http://www.jbc.org/ by guest on July 3, 2016

of Elek and Hill (IS), and for cystine by the Folin and Marenzi (19)
procedure, with a Zeiss-P&rich
photometer.
If the protein
contained both cystine and cysteine the acid hydrolysates were
oxidized with 3 per cent HzOz, and then the excess H202 was
removed by the sulfite used in the cystine reduction.
Cysteine
was analyzed for by the iodoacetate method of Mirsky and Anson
(4). Amino nitrogen was determined by the method of Van
Slyke (11) in the constant volume apparatus.
The proteins were
dissolved by grinding in a mortar with 10 cc. of 0.1 M NaOH and
diluted to 50 cc. 5 cc. of solution containing 20 mg. of protein
were used for analysis; the nitrite and protein were added prior
to the addition of the acid. The time of reaction with nitrous
acid was 30 minutes.

D. R. Goddard
11.
12.
13.
14.
15.
16.
17.
18.
19.

and L. Michaelis

371

Van Slyke, D. D., J. Biol. Ch.em., 9, 185 (1911); 63,425 (1929).

Van Slyke, D. D., and B&hard,
F. J., J. Biol. Chem., 16,539 (1913-14).
Vickery, H. B., and Block, R. J., J. Biol. Chem., 86,107 (1930).
Stewart, A. M., and Remington, C., Biochem. J., 26,2189 (1931).
Marston, H. R., Australian
Council
SC. Znd. Research,
Bull.
38 (1928).
Anson, M. L., andMirsky,
A. E., J. Gen. Physiol.,
17,151 (1933).
Teorell, T., Actamed. &and., 68,305 (1928).
Elek, A., and Hill, D. W., J. Am. Chem. Sot., 66,3479 (1933).
Folin, O., and Marenzi, A. D., J. Biol. Chem., 83, 103 (1929).

Downloaded from http://www.jbc.org/ by guest on July 3, 2016

DERIVATIVES OF KERATIN
David R. Goddard and Leonor Michaelis
J. Biol. Chem. 1935, 112:361-371.

Access the most updated version of this article at

http://www.jbc.org/content/112/1/361.citation

Click here to choose from all of JBC's e-mail

alerts
This article cites 0 references, 0 of which can be
accessed free at
http://www.jbc.org/content/112/1/361.citation.full.h
tml#ref-list-1

Downloaded from http://www.jbc.org/ by guest on July 3, 2016

Alerts:
When this article is cited
When a correction for this article is posted