STUDIES ON GOSSYPOL.

I. THE PREPARATION, PURIFICATION, AND SOME OF THE PROP-

ERTIES OF GOSSYPOL, THE TOXIC PRINCIPLE
OF COTTONSEED.

BY E. P. CLARK.*
(From the Food Research Division of the Chemical and Technological Research
Branch of the Bureau of Chemistry and Soils, United
States Department of Agriculture, Washington.)

(Received for publication, October 3, 1927.)

The condition known as cottonseed injury, which often follows

the incautious feeding of cottonseed meal to livestock, has been
shown by Withers and Carruth (1) and by Schwartze and Alsberg
(2) to be the toxic effect of a peculiar phenolic body known as
gossypol.
Undoubtedly this substance was first prepared in a fairly pure
condition by Marchlewski (3) who gave it its name. His prepara-
tion contained loosely bound acetic acid and had a melting point
of 188°C. He reported analyses of a number of sa,mplesbut was
not able to establish with certainty a molecular formula, although
he suggested C13H1404 or C32H31010.He made no crystalline deriv-
atives of the material but showed that it was not a glucoside and
in all probability contained no alkoxyl groups. He recorded some
color reactions and its behavior toward certain reagents.
No further outstanding chemical work was published upon the
subject until Carruth (4) communicated his interesting results in
1918. This investigator reported improvements in the prepara-
tion of gossypol, thus making possible more extended studies.
A summary of his results is as follows: He found that gossypol was
a phenolic substance, C&H3009 or C&HZ809. When crystallized
from acetic acid it contained 1 molecule of loosely bound acid
which could be removed by dissolving the substance in ether, add-
* This work was done under a research fellowship supported by the Inter-
state Cottonseed Crushers’ Association.
725

This is an Open Access article under the CC BY license.

726 Studies on Gossypol. I

ing water, and then removing the ether by evaporation. The

acetic acid remained in the water and could be titrated. In this
way a value near 532 was obtained for its molecular weight. De-
terminations of this constant by physicochemical methods were
not conclusive. Values ranging from 300 to 595 resulted by the
use of different methods and solvents. When gossypol was dis-
solved in aniline, or when gossypol and aniline were dissolved in
an appropriate solvent, a characteristic crystalline product was
formed. Carruth assumed the compound to have the formula
of C&H2809.2COH5NH2. No carbon or hydrogen figures were
given on this material, but those for nitrogen were reported
from 3.97 to 4.84 per cent. Crystalline acyl derivatives were not
obtained, but analysis of the amorphous acetyl derivative indi-
cated the presence of four acetyl groups. When gossypol was
heated to its decomposition point a new substance called B gossy-
pol was formed. This was obtained crystalline and apparently
represented gossypol less 2 mols of water.
The above constitutes essentially all that is recorded concerning
the chemistry of this interesting substance. For this reason,
and also because of the relation of gossypol to the so called cotton-
seed injury, a new chemical study of the substance has been under-
taken. The hope is enterta.ined that the information thus obtained
will be useful in the problem of detoxication of cottonseed meal.
The present communication is a preliminary report upon this work.
It deals with the preparation, purification, and analysis of gossypol
as well as certain of its derivatives.
The data submitted herewith indicate that the molecular formula
for gossypol is C30H300g. This formula differs from Carruth’s,
but, in t,he light of the experience gained in the purification of the
substance, it seems doubtful whether Carruth’s material was en-
tirely pure. As recorded in the experimental part, a reddish brown
impurity was associated with the gossypol, when prepared by Car-
ruth’s method, and was removed only with great difficulty by a
large number of solvents suitable for its recrystallization. A com-
bination of solvents was finally found, however, which gave the
desired result, but again difficulty was encountered in obtaining
the material free from the solvent used in its purification. In
a number of cases, a small indefinit,e amount of solvent remained
with the crystals. This could not be removed in a 0.1 mm. vacuum
 E. P. Clark 727

over both sulfuric acid and potassium hydroxide and at tempera-

tures which would insure the integrity of the substa.nce. Several
solvents, however, were found which, under certain conditions,
gave material having uniform composition. The criterion used
to verify this was to obtain such preparations from different sol-
vents as gave check analyses. When this was accomplished, car-
bon and hydrogen figures were obtained which indicated the
formula given above. Various derivatives, prepared from the
purified gossypol, also gave values upon analysis which checked
the formula indicated. Among these were t.he so called “acetate;”
i.e., gossypol + 1 molecule of acetic acid, anhydrogossypol (Car-
ruth’s B gossypol), the aniline condensation product, the dioxime,
and a hexaacetyl derivative.
The gossypol obtained in this work was a bright canary-yellow
crystalline substance with a melting point of 214°C. (corrected).
In solution no optical activity was exhibited. Molecular weight
determinations by the camphor method of Rast (5) gave results of
the order required for the formula suggested. Also, the acetic acid
in the so called acetate was found to be 10.51 per cent, which gave
a molecular weight of 511 for gossypol, assuming 1 mol of the latter
to be associated with 1 mol of acetic acid. Gossypol was soluble
in ether and acetone, but only sparingly soluble in other common
organic solvents. It was insoluble in water, but dissolved readily
in dilute ammonia and sodium carbonate. It was soluble in aque-
ous fixed alkalies; but its solution, with an excess of alkali, slowly
decomposed, giving a brown solution which gradually became a
beautiful purple. However, when potassium hydroxide was added
to an aqueous suspension of finely ground gossypol, in amounts
slightly more than enough to cause complete solution, the excess
alkali could be titrated and the acid value of the substance was
thus determined. Assuming that 2 equivalents of potassium hy-
droxide were used, molecular weight values of 505 and 507 were
obtained. Its solution in concentrated sulfuric acid had a deep
scarlet color. Upon diluting this acid liquid with water, the gossy-
pol was thrown out unchanged. It was readily soluble in saturated
methyl alcoholic hydrochloric acid. The color of the solution was
first brown which gradually turned deep green and finally intense
blue. When heated to its melting point, gossypol lost 2 mols of
water, producing a new crystalline body, anhydrogossypol,
728 Studies on Gossypol. I
C,aI-Iz,Os. This substance possessed quite different solubilities,
but still retained the characteristic color reactions of the original
material. Oximation of gossypol produced a colorless dioxime.
This seemed to indicate that the chromophore is a quinone group.
As stated by Carruth, gossypol combined with aniline to form a
deep orange-colored crystalline compound. This substance was
found to be a condensation product formed by 1 mol of gossypol
uniting with 2 mols of aniline, with the elimination of 2 mols of
water. Its formula was found to be C&H40N20s, and it is probably
a Schiff base type of compound.
Acetylation of gossypol was carried out with acetic anhydride,
with use of sodium acetate and also pyridine as catalysts. With
sodium acetate, by employing the conventional proportions of
reagents (1 part of acetate, 4 parts of substance, and 16 parts of
acetic anhydride) and by boiling the mixture for 10 minutes, a
non-crystallizable product was obtained which analyzed for 30.8
per cent acetyl. But when a large excess of the reagents was used
and the boiling continued for 2 hours, a material resulted, about
10 per cent of which was obtained as beautiful crystalline plates.
Acetylation in pyridine solution gave this same product but the
yield was better. This substance, when analyzed for acetyl by
Perkin’s method (6) for 0-acyl compounds, gave from 24.9 to
25.9 per cent acetyl. These figures agreed quite well for a tetra-
acetyl gossypol (25.07 per cent). Carbon values, however,
were those of a hexaacetyl body. It was then found that if Per-
kin’s method (6) for N-acyl compounds was used, the percentage
of acetyl required for a hexaacetyl derivative was obtained. The
crystalline material was therefore a hexaacetyl gossypol, indicat-
ing that in the parent substance there were six hydroxyl groups.
Thus all the oxygen-in gossypol has been accounted for. 2 atoms
are present as carbonyl groups, as was shown by the formation of a
dioxime, and 6 are present as hydroxyl groups. It is to be em-
phasized, though, that two of these hydroxyl groups behave quite
differently from the remaining four. This was manifest by the
drastic treatment necessary to remove the last two acetyl groups,
as just mentioned, and also by the fact that two hydroxyl groups
are capable of being titrated as an acid.
Finally it is desired to call attention to some toxicity tests in
which rats were used as experimental animals. These tests were
 E. P. Clark 729
conducted solely for the purpose of checking the toxicity of several
preparations during the course of their purification. In this work,
however, certain observations were made which seemed to be of
sufficient interest to warrant reporting.
White rats, whose average weight was 190 gm., were injected
intraperitoneally with cottonseed oil solutions of the various
TABLE I.
Toxicity Tests Made upon Different Preparations of Gossypol.
The material, dissolved in cottonseed oil, was given intraperitoneally
to white rats, the average weight of each being 190 gm.

No. 0 f ,zlesth: 0
No.

DC&. Material used. DOS?. rats result Remarks.

used. ing.
-.
1967 mg. PC r
lw.
Feb. 9 Crude gossypol 25 1 0 0
acetate. 50 2 2 2, 2 Acute.
100 1 1 ‘L
1
_
May 4 Gossypol ‘recrys- 25 2 6, 13 All but 13
tallized from 50 2 2, 2 day animal
chloroform. 75 2 1, 1 acute.
100 2 1, 1
- _
May 17 Gossypol Vtwice 20 2 7, 10
recrystalli ze d 30 2 3, 5 All acute
from ether 40 1 3
and petroleum 50 1 1 type.
ether.
- -
June 20 Purified gossy- 10 6 17, 18 2 day and 3
pol recrystal- 15 6 10, 11, 12, day ani-
lized from 12, 13 mals acute,
ether (analyt- 20 6 2, 3, Q, all others
ically pure). 10, 10 chronic.
-

preparations. There was no apparent difference between the

toxicity of the crude gossypol acetate and the analytically pure
gossypol. Rats receiving doses of 20 mg. or more per kilo of body
weight died in from 20 hours to 13 days. Of those receiving from
10 to 20 mg. per kilo, 50 per cent survived; but in the remaining,
deaths occurred in from 2 to 18 days. Though t.he percentage of
deaths at 15 mg. per kilo equals that at 20 mg. per kilo, death
730 Studies on Gossypol. I

resulting from the latter dosage was the result of acute poisoning,
whereas death at 15 mg. per kilo was the result of chronic poisoning.
Postmortem findings in the acute type of poisoning were oily
droplets and plaques in the peritoneum, congested lungs, hemor-
rhagic condition of all the viscera, and cardiac dilatation. In the
chronic type of poisoning the outstanding indication was intestinal
impaction. The cause of delayed death was stated by Dr. For-
mad, pathologist of the Bureau of Animal Industry, “to be prob-
ably due to a paralyzing effect upon the sympathetic nervous
system.” Dr. Formad further stated that “it seemed probable
that there was a tonic spasm of the sphincter ani accompanied
by general decreased intestinal tonus which led to decreased rate
of elimination of feces and finally to an intestinal impaction. This
latter, in turn, produced pressure upon the diaphragm which was
transmitted to the heart and lungs and served as a probable cause
of death.”
In connection with the toxicity work just mentioned, assistance
was obtained from Dr. James C. Munch and Miss Georgiana S.
Gittinger of the Pharmacology Laboratory of this bureau. It is
a pleasure for the writer to express h&e his thanks to these workers
for their kind cooperation.
EXPERIMENTAL.

Preparation of Gossypol.
The gossypol used in this work was prepared by the method of
Carruth (4). Ground cottonseed meats were first freed from oil
by extraction with petroleum ether. 3 kilos of the air-dried ma-
terial thus obtained were percolated, at a moderate rate, with
ethyl ether until 7.5 liters of extract were obtained. Two such lots
were combined and the ether was distilled until the residue had a
volume of 1 liter. This was filtered and further concentrated, under
reduced pressure, to 400 cc. To this solution, 300 cc. of acetic acid
were added and the mixture was set aside to crystallize. After 2
hours, the so called gossypol acetate which had formed was freed
from the mother liquor, washed with acetic acid, and dried. 80
kilos of ground seed gave 385 gm. of crystals, a yield of 0.48 per
cent. The gossypol acetate consisted of somewhat imperfectly
shaped plates which melted at a temperature between 184-187%.
Its solution in ether had a brownish red color, owing to an impurity
 E. P. Clark 731

which was difficult to remove by recrystallization. It was found,

however, that the material could be purified by liberating the
loosely bound acetic acid and crystallizing the free gossypol from
appropriate solvents. To accomplish this the following procedure
was adopted. 50 gm. of the crude crystals were suspended in 400
cc. of ether. Approximately 4 volumes of water were added, and
the mixture was stirred until all the substance dissolved in the
ether layer. The contents of the beaker were then heated on the
steam bath until the ether evaporated and the water had a tem-
perature of about 60°C. Under this treatment the gossypol
became a porous mass and could be broken up. To facilitate this,
it was cooled by adding crushed ice. The gossypol was then fil-
tered from the water containing the acetic acid, washed with cold
water, and dried. The yield was 45 gm.
To purify the crude product it was dissolved in ether and filtered.
2 volumes of petroleum ether (b.p. 3Fj-70”) were added to the fil-
trate, and the solution was concentrated, under reduced pressure,
until crystals began to form. The liquid was removed from the
flask, and 4 volume of petroleum ether was added which caused
crystallization to take place very rapidly. The yield was generally
70 per cent. The mother liquors were concentrated, as described
above, until crystallization began. Upon standing some time, a
second crop of crystals was obtained, amounting to 15 per cent of
the original material. The latter was treated as new material and
was further purified by repeating the above process. The result-
ing mother liquors were quite darkly colored. They were evapo-
rated, in vacua, to dryness, dissolved in ether, and acetic acid
was added. In this way some crude acetate was always recovered
which was reworked.
The first crop of crystals was recrystallized a second time to
free them completely from the red impurity. The resulting prod-
uct, as well as its solution in ether, had a bright canary-yellow
color. It was not used for analysis, however, as it contained an
indefinite amount of the solvent employed in its crystallization,
and this could not be readily removed. It was therefore recrystal-
lized from different media which gave uniform preparations of
definite composition.1
* Of the methods to be described, only the last two are practical from the
preparative standpoint. The ether method is recommended.
732 Studies on Gossypol. I
The first solvent used was dilute alcohol. A boiling saturated
solution of the purified gossypol in 95 per cent alcohol was diluted
with water until a slight turbidity occurred. The liquid was set
aside and was occasionally stirred during crystallization. The
product obtained had a distinct green color but gradua,lly turned
an orange color during several days’ exposure to the air. This
air-dried material consisted of irregularly shaped fractured plates
which began to soften at 190°C. and melted with decomposition
at 199” (corrected). A molecular weight determination made by
the camphor method of Rast (5) gave the following results:
12.3 mg. of substance dissolved in 136.8 mg. of camphor, A = 6.3’.

M = (40,000) (12.3) = 571

(136.8) (6.3) -

0.1982 gm. substance gave 0.1067 gm. Hz0 and 0.5045 gm. CO,.

For CX,HX,OS. Calculated. M 518.4. C 69.48 per cent, H 5.83 per cent
Found. “ 571. “ 69.44 “ “ “ 6.01 “ “

Its optical properties were found to be as follows:2 pale yellow

plates, in part obliquely rhombic in outline. The substance dis-
solved in organic liquids, so its refractive indices were determined
by immersion in solutions of mercuric potassium iodide in glycerol.
a = 1.605; /3 = 1.740; y was much greater than the highest liquid
available, but it was determined as approximately 1.53 by immer-
sion in molten mixtures of piperine with antimony-arsenic iodide.
For identification purposes the lowest of these values was found to
be the most useful, as it was shown by the majority of the grains
in one crystallographic direction.
Another portion of the purified gossypol was recrystallized from
ether and methanol in t,he following manner. 4 volumes of me-
thanol were added to a filtered ethereal solution of gossypol.
The solvents were removed by distillation, under reduced pressure,
until crystallization began. The contents of the flask were removed
2 The optical properties of this and of several of the compounds, to be
described later, were kindly determined by Dr. Edgar T. Wherry of this
bureau, and Mr. George L. Keenan of the Food, Drug and Insecticide
Administration. The writer wishes at this time to express his thanks to
Dr. Wherry and Mr. Keenan for their valuable cooperation.
 E. P. Clark 733

and crystallization was allowed to proceed to completion. The

preparation obtained in t,his way had a somewhat greenish color
but became bright yellow upon exposure to the air during a week.
In the latter condition, the substance consisted of irregularly
shaped fractured plates which began to darken and contract at
205°C. and melted with decomposition at .214’ (corrected). Its
optical properties were found to be the same as those recorded for
the previous preparation. Molecular weight determination by
the Rast method gave the following.
10.7 mg. of substance dissolved in 110 mg. of camphor, A = 6.8”.

M = (40,000) (10.7)
(110) (6.8) = 572*

I. 0.1092 gm. substance gave 0.0585 gm. Hz0 and 0.2778 gm. COz.
II. 0.1973 “ “ “ 0.1059 “ “ “ 0.5032 “ “

For CsoHaoOs. Calculated. M 518.4. C 69.48 per cent, H 5.83 per cent.
Found. I. “ 572. “ 69.40 “ “ “ 5.99 “ “
II. “ 69.58 “ “ “ 6.01 “ “

A third method of obtaining solvent-free gossypol was to crystal-

lize the purified substance from ether. This was done by stirring
on a steam bath an ethereal solution of the purified material wit.h
4 volumes of water until the ether had evaporated and the water
had a temperature of approximately 60°C. Upon cooling, the gos-
sypol was obtained as a bright yellow porous mass which consisted
of very small irregularly shaped plates. It began to soften and
cont.ract at 202” and melted with decomposition at 214” (corrected).
The optical properties of this preparation could not be determined,
owing to the minuteness and physical state of the crystals. Molec-
ular weight determinations were made by quickly dissolving a
weighed amount of substance in an excess of 0.1 N KOH, and
titrating the excess alkali with phenolphthalein as an indicator.
I. 0.3373 gm. material used 13.3 cc. 0.1 N KOH.
II. 0.4169 “ “ “ 16.5 “ 0.1 “ “
Molecular weight found (assuming 1 mol of gossypol requires 2 mols of
KOH). I, 507; II, 505.
I. 0.2055 gm. substance gave 0 1091 gm. Hz0 and 0.5229 gm. COZ.
II. 0.1958 “ “ “ 0.1040 “ “ “ 0.4979 “ “
Found. I. C 69.42 per cent, H 5.94 per cent.
II. “69.37 ” ” (‘ 5.94 “ (‘
734 Studies on Gossypol. I

So Called Acetate Prepared from Purified Gossypol.

As this compound is important in the preparation and identifica-
tion of gossypol, a sample was made from purified starting mate-
ria,l in order that accurate data concerning it could be determined.
The substance was obtained by adding acetic acid to an ethereal
solution of gossypol whereupon crystallization began immediately.
It consisted of bright yellow broad plates with notched ends, oc-
casionally rhombs, which began to soften at 180°C. a.nd melted
with decomposition at X39-190°C. (corrected). The acetic acid
content of the substance was determined by dissolving the material
in ether, adding water, and then removing the ether by evapora-
tion on the steam bath. When the gossypol was separat,ed
by filtration, the acetic acid in the aqueous filtrate was titrated.
0.4996 gm. acetate required 8.75 cc. 0.1 N KOH.
I. 0.1074 gm. acetate gave 0.0585 gm. Hz0 and 0.2600 gm. COZ.
II. 0.1929 “ (‘ “ 0.1043 “ “ ‘I 0.4668 “ ‘I
For C30H3008.CHICOOH.
Calculated. CH,COOH 10.38%, C 66.41%>, I-1 5.92%.
Found. I. “ 10.51?7@ “ 66.04yc, ” 6.09%.
II. “ 66.02%, “ 6.05%.

Its optical properties were found to be as follows: bright yellow

plates, often obliquely rhombic. Oily immersion media were
applicable for this substance. o( = 1.550; p = 1.750; y was higher
than the highest liquid available, but it was determined approxi-
mately as 1.84 in piperine iodide melts. As in the case of pure
gossypol, the lowest index was the most useful for identification
purposes.

Anhydrogossypol.
Gossypol in Q gm. portions, contained in medium sized test,-
,tubes, was heated with stirring in an oil bath to 215°C. As soon as
a uniform mass was obtained and effervescence ceased, it was
removed from the bath and cooled. The melt was quickly dis-
solved in ether and removed from the test-tube. Anhydrogossy-
pol began to crystallize immediately and the process was complete
in a few minutes. The compound was filtered, washed with ether,
and dried. The yield was 45 per cent of theory. Purification was
 E. P. Clark 735
accomplished by dissolving it in a large volume of dry ether and
slowly evaporating this solution on the steam bath until crystal-
lization began. The product consisted of orange-colored rods
which softened and partly melted at 229-230°C.) then melted com-
petely to a black liquid quite sharply at 268°C. (corrected). It
was very sparingly soluble in common organic solvents, but pos-
sessed the color reactions of gossypol.
I. 0.2019 gm. substance gave 0.0994 gm. Hz0 and 0.5510 gm. COZ.
II. 0.1410 IL (‘ ” 0.0698 “ ” “ 0.3842 “ “
For ChH2606. Calculated. C 74.66 per cent, H 5.43 per cent.
Found. I. “ 74.45 “ “ “ 5.51 “ “
I‘ 74.34 LL c< “ 5.54 ‘< I‘
II.

Its optical properties were found to be as follows: brownish yellow

rods with oblique pyramidal terminations. O( = 1.560; /? and y
were much higher than 1.75, the highest liquid available. In
addition to the a! value, which was shown lengthwise of the rods,
the peculiar red and green polarization colors shown between
crossed nicols were characteristic of this substance.

Dianiline Gossypol.

This substance was conveniently prepared by dissolving gossypol

in 10 times its weight of boiling aniline. On cooling the solution,
dianiline gossypol separated as a beautiful orange-colored crystal-
line mass. The crystals were freed from the mother liquor and
were thoroughly washed with ether. The yield was practically
quantitative. It was purified by recrystallization from benzene.
A portion of the crude material was dissolved in boiling benzene,
and the solution filtered and concentrated, under slightly diminished
pressure, to about Q its original volume. The substance crystal-
lized immediately upon removing the concentrate from the flask
and allowing it to cool. When heated rapidly its melting point
was 302-303°C. with decomposition (corrected). In ordinary
light it consisted of yellow rods with all the refractive indices
greater than 1.75 and accordingly indeterminate. Between
crossed nicols it showed parallel extinction and anomalous blue
and green polarization colors owing to peculiar dispersion. This
may be of use in recognizing the substance. Molecular weight
 Studies on Gossypol. I

determinations by the camphor method of Rast gave the following

values :
I. 19.9 mg. dissolved in 195.3 mg. of camphor, A = 6”.

M = (40,000) (19.9)
(195.3) (6.0) = ‘*‘*

II. 16.9 mg. dissolved in 182.3 mg. of camphor, A = 5.7”.

M = (40,000) (16.9) = 650

(182.3) (5.7) *

I. 52.9 mg. substance required 4.67 cc. ~/30 acid (Kjeldahl).$

II. 51.3 “ ‘( “ 4.55 “ N/30 “ “
I. 0.1040 gm. “ gave 0.0565 gm. Hz0 and 0.2869 gm. CO*.
II. 0.1202 “ “ “ 0.0665 “ “ “ 0.3328 “ “

For CazH40Nz06.
Calculated. M 668.5. C175.43 per cent, H 6.03 per cent, N 4.19 per cent.
Found. I. “ 680. ‘c.75.26 ‘d ‘6 I‘ 6.08 ‘6 6‘ u 4.12 iI ‘6
“ 75.53 “ “ “ 6.19 “ ‘I ‘I 4.14 ‘L “
II. “ 650.

Gossypol Dioxime.
30 cc. of an alcoholic solution of hydroxylamine, made by ex-
actly neutralizing with sodium hydroxide the hydrochloric acid in
3 gm. of hydroxylamine hydrochloride, were added to a suspension
of 5 gm. of gossypol in 50 cc. of alcohol. The solution was warmed
and stirred until all the gossypol was dissolved. The solution was
then heated to boiling and set aside during the night to crystallize.
The oxime was washed with alcohol, then with water, again with
alcohol, and then dried. The yield was 4 gm. It was recrystal-
lized by dissolving 1 gm. in 125 cc. of boiling alcohol, filtering, and
adding an equal volume of water. Crystallization began at once
3 Attention is called at this point to the necessity of using mercury as a
catalyst in the Kjeldahl nitrogen determinations upon this substance.
When copper was used as a catalyst, low values for nitrogen, ranging from
3.6 to 3.8 per cent (depending upon the period of digestion) were obtained.
In some cases the digestion was prolonged 6 hours after the solution cleared.
The low nitrogen figures upon this compound reported by Carruth and by
Schwartze are accounted for by the fact that copper was used as a catalyst
by them in their determinations (personal communications).
 E. P. Clark 737
and was complete within an hour. The yield was 85 per cent of
theory. The material consisted of thick rhomboidal plates which
began to darken and contract around 225°C. but did not melt at
315’C. It had a faintly yellowish green color, but under the micro-
scope the crystals were colorless.
I. 49.9 mg. substance required 5.28 cc. N/30 acid (Kjeldahl).
<‘
II. 50.6 ” ‘I 5.35 ‘I N/30 ” (’
I. 88.3 L( “ gave 48.0 mg. Hz0 and 210.8 mg. CO*.
II. 93.5 (‘ “ ” 51 .3 “ “ “ 223.9 “ “
For C&,H32N208.
Calculated. C 65.66 per cent, H 5.88 per cent, N 5.11 per cent.
Found. I. ” 65.13 ” “ “ 6.08 “ ” ” 4.94 ” “
“ 65.33 <C “ I‘ 6.14 CL “ LC 4.94 “ ‘I
II.

Acetyl Gossypol.
A mixture of 1 gm. of gossypol, 0.25 gm. of anhydrous sodium
acetate, and 4 cc. of acetic anhydride were boiled under a reflux
for 10 minutes. The resulting solution was diluted with an equal
volume of acetic acid, cooled, and poured into 300 cc. of water.
A light cream-colored product was precipitated which was freed
from the mother liquor, washed with water, and dried.
0.4030 gm. substance required 28.9 cc. 0.1 N KOH (Perkin’s 0-acyl
method).
Acetyl found. 30.86 per cent.

3 gm. of gossypol acetate, 40 cc. of acetic anhydride, and 1.5

gm. of sodium acetate were boiled for 2 hours, after which the
solution was treated in the manner described above. Yield, 4
gm. This material was dissolved in 75 cc. of boiling methanol.
The next day a small amount of a crystalline material had sepa-
rated. This was dissolved in boiling ethyl acetate, filtered, and 2
volumes of methanol were added. The solution was concentrated,
under reduced pressure, until crystals began to form. After several
hours the crystalline product was collected, washed with methanol,
and dried. Yield, 0.4 gm. The material consisted of thin plates,
nearly rectangular in outline and practically colorless. It began to
soften and darken at a temperature of 255”C., and melted with
decomposition at 276-277°C. (corrected). QI = 1.515; /? = 1.580;
y = 1.645. The highest index was best adapted for the identifi-
738 Studies on Gossypol. I

cation of the substance. It was shown in one crystallographic

direction by the majority of the plates.
I. 0.4280 gm. substance required 24.76 cc. 0.1 N KOH (Perkin’s
0-acyl method).
II. 0.2507 gm. substance required 14.9 cc. 0.1 N KOH (Perkin’s O-
acyl method).
III. 0.3024 gm. substance required 23.05 cc. 0.1 N KOH (Perkin’s N-
acyl method).
Acetyl found. I, 24.88 per cent; II, 25.56 per cent; III, 32.8 per cent.
I. 90.3 mg. substance gave 45.1 mg. Hz0 and 216.0 mg. CO*.
II. 138.3 (‘ (‘ ,‘ 70.5 “ C‘ ‘I 331.1 “ ‘I
For C&,HUOS(CHZCO)G.
Calculated. Acetyl33.5 per cent, C 65.44 per cent, H 5.50 per cent.
Found. I. Acetyl(II1) 32.8%, C 65.26%, H 5.59%.
II. “ 65.31V& “ 5.70$&.

The acetylation of gossypol in pyridine solution gave the same

compound as above but the yield was better. This procedure was
carried out as follows: A solution of 5 gm. of gossypol in 30 cc. of
pyridine was treated with 10 cc. of acetic anhydride. The mixture
was heated to its boiling point and was then allowed to stand for
24 hours. Upon pouring the reaction mixture into water 8.8 gm.
of cream-colored material were obtained. The crystalline hexa-
acetate was obtained by dissolving the substance in methanol and
allowing it to crystallize during a day. Upon recrystallization, as
outlined in the above experiment, 1.6 gm. of pure acetyl body were
obtained.
The optical properties and melting point were the same as given
above.
A molecular weight determination by the camphor method gave
the following :
14.4 mg. of substance dissolved in 184.7 mg. of camphor, A = 4°C.

M = (40,000) (14.4) = 780

(184.7) (4) -

For CLH,,Ola. Calculated. M = 770.5.

SUMMARY.

1. Gossypol, the toxic principle of cottonseed, has been pre-

pared in a state of analytical purity, and its molecular formula
 E. P. Clark 739
has been ascertained to be C30H3008. Certain properties of this
purified material have been recorded.
2. Analytical data, including optical properties in certain cases,
have also been obtained upon several derivatives of gossypol.
Those included are the so called acetate, anhydrogossypol, the
dianilide, the dioxime, and the hexaacetyl derivative.
3. The dianilide, which is important because of its use in the
quantitative determination of gossypol, has not been found to be
a salt of gossypol (1 mol of gossypol + 2 mols of aniline), as Car-
ruth has suggested, but a condensation product formed by 1 mol of
gossypol condensing with 2 mols of aniline, with the elimination of
2 mols of water.
4. Of the 8 oxygen atoms in gossypol, 2 have been shown to be
present as carbonyl groups, whereas the remaining 6 have been
shown to be present as hydroxyl groups. Two of these hydroxyl
groups behave differently from the remaining four, being much
more acidic and requiring drastic treatment for the hydrolysis of
their acetyl derivative.
5. A short report concerning some interesting physiological
effects of gossypol upon rats has been made.

BIBLIOGRAPHY.

1. Withers, W. A., and Carruth, F. E., J. Agric. Research, 1915, v, 261; 1918,
xii, 83.
2. Schwartze, E. W., and Alsberg, C. L., J. Pharmacol. and Exp. Therap.,
1919, xiii, 504; 1921, xvii, 344; J. Agric. Research, 1923, xxv, 285; 1924,
xxviii, 173, 191. Schwartze, E. W., J. Oil and Fat Ind., 1926, iii, 173.
(A quite complete bibliography of the whole problem of cottonseed
meal poisoning will be found in references 1 and 2.)
3. Marchlewski, L., J. prakt. Chem., 1899, Ix, 84.
4. Carruth, F. E., J. Am. Chem. Xoc., 1918, xl, 647.
5. Rast, K., Rer. &em. Ges., 1922, Iv, 1051.
6. Perkin, A. G., J. Chem. Sot., 1905, Ixxxvii, 107.