STUDIES I N CARCINOGENESIS
11. THEDETECTION
OF DIBENZANTHRACENE
IN MOUSETUMORS
INDUCED
BY THIS HYDROCARBON
EGON LORENZ
AKD
M. J . SHEAR
(From the Ofire of Cancer Investigationr, U . S. Puhlic Health Service, Harvard Medical School,
Boston)
.
I n an investigation of the r6le of 1 :2 :5 :6-dibenzanthracene in the production of tumors Chalmers ( 1934) examined, spectrographically, tissues of
fowls that had received injections of this hydrocarbon dissolved in fat. Similar spectrographic studies, on mouse tissues, have been carried on in this
laboratory and obstacles similar to those reported by Chalmers have been
encountered. Chalmers found that some extracts contained substances which
showed “ intense absorption in the ultraviolet.” Such substances would mask
the dibenzanthracene spectrum ; consequently the spectrographic analysis of
those extracts was being held in abeyance, Chalmers stated, until the interfering substances could be removed. He found that extracts of tissue into which
dibenzanthracene had been injected two years previously showed no selective
absorption even though the extracts were concentrated “ t o the point of general absorption in the ultraviolet.” Chalmers then determined, on successive
days, the quantity of dibenzanthracene remaining after injection and found
that most of the hydrocarbon disappears from the site of injection within
a week.
Berenblum and Kendal ( 1934) injected a colloidal aqueous preparation
of dibenzanthracene into the breast muscle of the fowl and, by means of the
fluorescence spectrum, examined the muscle extracts for the presence of the
hydrocarbon. The method was said to be sensitive to 0.02 mg. These investigators stated: “ Since each of the 7 birds initially received 0.6 mg. of dibenzanthracene, while less than 0.02 mg. was present in the breast muscle at the
end of the experiment, most of the dibenzanthracene must have disappeared
from the site of injection during the course of the experiment.”
Substances which produce continuous absorption in the ultraviolet, thus
masking the possible presence of small amounts of dibenzanthracene, were
found in tissue extracts in the present study, as in Chalmers’ investigation.
Inasmuch as some progress has been made towards the removal of interfering
substances from extracts of mouse tumors, the results obtained so far are being
reported here.
In brief, the lipoid fraction was extracted from tumors produced in mice
with dibenzanthracene; the lipoids were then treated with various solvents in
an attempt to separate the hydrocarbon from spectroscopically interfering
substances. Alcohol extraction separated the dibenzanthracene from much
of the interfering lipoids. The alcoholic solution, after dilution with water,
was next treated with hexane, which removed the dibenzanthracene from the
aqueous alcohol, leaving behind a further quantity of interfering lipoids. The
333
334
EGON LORENZ AND M. J. SHEAR
procedures employed were guided by spectrum analyses. Further removal of
the remaining interfering material was effected by chilling in solid CO, and
by adsorption on Al,O:;.
I t was found that tumors arising six to eight months after injection of
a lard solution of 1 :2 :5 :6-dibenzanthracene contained detectable amounts
of this hydrocarbon. Fifth and sixth generation transplants of these tumors,
however, failed to give any evidence of the presence of dibenzanthracene.
EXPERIMENTAL
STUDIES
Technic: Two methods are available for the spectrum analysis of such
carcinogenic compounds as dibenzanthracene, one using the fluorescence and
the other the absorption spectrum. Although the first method is somewhat
simpler than the second, since the fluorescence spectra of carcinogenic compounds lie mostly in the visible region of the spectrum and since the compound
fluoresces in the solid state, it is not suitable for quantitative analysis or for
identification, because the fluorescence spectra of many of the carcinogenic
compounds show relatively broad and diffuse bands in the same region of
the spectrum. Furthermore, for the detection of minute quantities (of the
order of magnitude of a fraction of a milligram) whose fluorescence intensity
is too weak to be detected visually, the exposure of photographic plates would
require an undesirably long time. Therefore analysis by absorption sRectra
was chosen in this investigation as this method makes possible the exact identification of the compound in question, is rapid, allows quantitative determination, and is superior in sensitivity to other methods. These statements, of
STUDIES I N CARCINOGENESIS
335
course, apply only to compounds having an absorption spectrum with sharp
and narrow bands of high absorbing power, as is the case with the carcinogenic
compounds investigated in this laboratory.
Fig. 1 gives the absorption spectrum calculated for a molar solution of
1 :2 :5 :6-dibenzanthracene in ether. The log. of the extinction coefficient
(log. K ) is plotted against wave numbers (cm-l). The extinction coefficient
was measured by means of a Judd-Lewis photometer in connection with a
quartz spectrograph. As source of illumination a tungsten arc was used.
Due to the sharpness of the absorption bands of 1 :2 :5 :6-dibenzanthracene,
this method of determining the extinction coefficient is not free from error;
the total error in the value of the extinction coefficient is estimated as being
10 per cent, which, however, is satisfactory
of the order of magnitude of
for the purpose of this paper.
The absorption spectrum of 1 :2 :5 :6-dibenzanthracene lies completely in
the ultraviolet. Three different groups of bands can be distinguished: one
from 25,000to 28,000cm:' ( h = 4000 to 3570A); one from 28,000to 33,000
cm.-l ( h = 3570 to 3030A) and one from 33,000 to 38,000cm:l (A = 3030
to 2630A). The band at 33,620 cm.-l (A = 2975A) has at its maximum a
log. K of 5.82. This is the most persistent band of the spectrum and is the
last to disappear with increasing dilution of the solvent. It was used in the
quantitative analysis of minute amounts of dibenzhnthracene, as will be shown
later.2
As already mentioned, the absorption spectrum of dibenzanthracene lies
completely in the ultraviolet. In absorption spectrum analysis, the standard
procedure consists in measuring photo-electrically or photographically the
extinction as a function of wavelength. The identification, as well as the
determination of the amount present, is then based on such measurements.
However, the photo-electric method is tedious and the photographic method
requires expensive equipment. Inasmuch as the absorption spectrum of dibenzanthracene consists of well defined, sharp bands, it is possible to obtain
sufficient data for identification and quantitative analysis by simply photographing the absorption spectrum under certain standard conditions, as will
be shown later. This is the method adopted for the rapid identification of
the compounds.
A large Gaertner quartz spectrograph with a slit width of approximately
0,015mm. was used. The source of light consisted of a water-cooled hydrogen
discharge tube, operated at 1500 volts a.c. and 200 ma. Using Eastman
'(Process " plates only a few seconds of exposure were necessary to obtain
a continuous spectrum reaching from h = 4000 to 2200 A. As absorption
cells, a 1 cm. and a 3 cm. cell were used. Each of these cells had a volume
somewhat less than 10 C.C.
The blackening curve of the photographic plate and the difference in
extinction of the different bands of the dibenzanthracene spectrum (see Fig.
1) make it impossible to obtain the complete absorption spectrum with one
1 We are indebted to Professor G. S. Forbes of the Division of Chemistry of Harvard University for kindly allowing us to use his spectroscopic apparatus for the measurement of the
extinction coefficient.
2 The absorption spectrum of 1:2:5:6-dibenzanthracene was first measured by Clar (1929).
Curves of the extinction coefficient of the third group of bands of dibenzanthracene were given
by Chalmers.
336
EGON LOREN2 AND M. J. SHEAR
concentration or with one cell thickness. For a cell thickness of 1 cm. the
first group of bands will appear with concentrations of approximately 0.2 mg.
per c.c., the second group with concentrations of approximately 0.05 mg. per
c.c., and the third group with concentrations of 0.01 mg. (or less) per C.C.
The dibenzanthracene employed was a yellow preparation, m.p. 258-260",
obtained from the Eastman Kodak Co. Purification by treatment with sulphuric acid, according to the method of Cook, Hieger, Kennaway, and Mayneord (1932), removed all of the yellow color, the resulting white crystals
melting at 263.5'-264.5'
(corr.). Purification by adsorption ' on activated
alumina, according to the technic of Winterstein and Schon (1933), removed
most of the color; the purest fraction was only faintly yellow and had a map.
of 265.5'-266.0' (corr.). Although treatment with sulphuric acid gave colorless crystals, the m.p. was not as high as that of the faintly yellow crystals
obtained by the adsorption method. All specimens showed identical absorption spectra; the yellow Eastman preparation showed, in addition, a faint
continuous absorption in the blue region.
The influence of the solvent was studied next. Ether, alcohol, chloroform, and hexane were the solvents used. With all of them identical spectra
were obtained for dibenzanthracene; furthermore, no shift of wavelength or
change in the shape of the bands could be observed.
A rough check on the validity of Beer's law was made. As was to be
expected, it was valid for the low concentrations of dibenzanthracene employed.
For the identification of dibenzanthracene and for the determination of
the amount present in solution, the following method was adopted. Standard
solutions in ether were prepared containing from 0.25 mg. per C.C. to 4 X
mg. per C.C. of the hydrocarbon. Standard spectra were taken with these
solutions. The 1 cm. absorption cell was used, except for the lowest concentration, for which the 3 cm. absorption cell was employed. On each photographic plate (Eastman " Process ") 6 spectra of the same solution were
taken with different times of exposure and, in addition, 2 spectra of the solvent
alone. A comparison wavelength scale which was calibrated with the standard
emission lines of the mercury arc was printed adjacent to each spectrum.
These spectra were compared with spectra taken of the unknown s ~ l u t i o n . ~
If dibenzanthracene was present, it was thus possible to identify it readily
and to make a rough estimate of its concentration.
The presence of other substances does not seriously interfere so long as
their spectra do not show either detailed band structure in the region of the
dibenzanthracene spectrum or too strong a continuous absorption. The interfering substances present in tissue extracts may be removed to a considerable extent by chemical procedures and will be described below.
Since the second group of bands of the dibenzanthracene spectrum (A =
3570 to 3030A) appears on the photographic plates for concentrations down
to 0.01 mg. per c.c., it is possible to identify readily dibenzanthracene in
unknown solutions when it is present in this or greater amounts; it is necessary only to compare the bands of the spectrum of the unknown solution as
4
This specimen was purified by Dr. E. B. Hershberg, who also determined the melting points.
The conditions of the developing process were, of course, kept constant for all plates.
STUDIES IN CARCINOGENESIS
3 37
to wavelength and width with those of the standard spectrum of a dibenzanthracene solution of appropriate concentration.
As the concentration decreases down to 0.01 mg. per c.c., the second group
of bands disappears, but the third group then becomes evident. The structure
of this third group of bands is not sufficiently detailed for the ready identification of dibenzanthracene in an unknown solution. However, this third
group may be used for identification in the special case where it is known
that the solution contains no other dissolved substance having absorption
bands in that particular region of the spectrum. In this case it can be identified down to concentrations of 4 X
mg. per C.C. with the 3 cm. cell. Of
course, it is possible to use a still longer cell, but the extent to which this
procedure may be employed is limited by the increase in the continuous absorption of other dissolved substances and of the solvent itself, and by the
amount of sample available for analysis.6
As already mentioned, a rough estimate of the amount of dibenzanthracene
present was made by this method. To obtain more precise quantitative data,
the following procedure was chosen: after obtaining an approximate value of
the amount present, solutions of dibenzanthracene were prepared in which
the concentrations varied in steps of 10 per cent above and below the determined approximate value. Then spectra of these solutions were printed on
a photographic plate with the solution of unknown concentration intercalated
so that above and below a spectrum of unknown concentration there were
spectra of known concentrations differing by 10 per cent. Time of exposure
and length of cell were kept constant. Bands of two adjacent spectra showing the same photographic density indicated that the concentration of both
solutions was the same. Continuous absorption in the spectrum of the unknown in the neighborhood of the bands would, of course, decrease the
accuracy, resulting in too large a value.
This method of identifying 1:2 :5 :6-dibenzanthracene holds, of course, not
only for this substance but also for all carcinogenic substances so far studied
in this laboratory, since they all show absorption spectra with many bands.
As a matter of fact, the method holds for all substances showing selective
absorption bands.
Absorption Spectra of Tumor Extracts
After studying the characteristics of the absorption spectrum of pure
dibenzanthracene solutions, and after determining that this method was capable of identifying and measuring minute amounts of the hydrocarbon, tumor
tissue was examined for the presence of this compound.
Tumors were produced in mice by the subcutaneous injection of a lard
solution containing 4 mg. dibenzanthracene per c.c.; the mice were given two
ilijections of 0.25 C.C.each. When, after six to eight months, the tumors had
attained a diameter of 2 to 3 cm., the healthy tumor tissue, freed from visible
lard, was removed and dried at about 65". The dry tumor tissue, weighing
9.8 gm., was extracted with ether in a Soxhlet apparatus for 1.5 days. A
6In employing a 3 cm. cell with a volume of 1 C.C. instead of 10 C.C.the total amount of
dibenzanthracene detectable is of the order of magnitude of 4 X lo-'' mg.
338
EGON LORENZ AND M. J. SHEAR
number of the organs and tissues of these mice were also treated and extracted
in the same way.
Preliminary Tests: The ether extract of the tumor tissue was turbid and
therefore could not be used in the spectrographic examination. The solvent
was therefore removed by evaporation; on extracting the residue with ether,
a clear solution was obtained. This, too, was unsatisfactory because of the
strong continuous absorption in the ultraviolet.
The analogous extracts of spleen and of muscle also showed this pronounced continuous absorption. These tissue extracts were then saponified
and the unsaponified residues taken up in ether. However, there was no
important reduction in the continuous absorption. This appeared to indicate
that the interfering substances were not fats. Addition of acetone and of
alcohol to aliquot portions of the ether solution of the unsaponified matter
gave no precipitation, indicating that neither lecithin nor cephalin was responsible for the continuous absorption.
A commercial sample of cholesterol was then dissolved in ether in concentrations of 1 mg. and 0.2 mg. per C.C. The stronger solution gave some
continuous absorption in the short ultraviolet but the range and intensity of
the absorption were such as to indicate that the continuous absorption of the
tissue extracts was not due to their cholesterol content.
Inasmuch as saponification had not removed the interfering material from
the extracts of spleen and of muscle, the tumor extract was not subjected to
saponification. The ether was removed from the tumor extract by evaporation and the residue treated with ethyl alcohol at 60". The alcohol extract
was cooled and filtered, giving a strongly yellow solution (I). The residue
was then extracted with ether; this, too, gave a deep yellow solution (11).
The residue from the alcohol and ether extractions was treated with chloroform; a pale yellow solution (111) was obtained. (See Chart 1.)
Although all three fractions contained material which gave continuous
absorption in the ultraviolet, the second group of bands of dibenzanthracene
was seen in both of the first two extracts. Most of the hydrocarbon (0.7 mg.)
went into the alcohol extract, a small amount (0.1 mg.) went into the ether
extract, and none was detected in the chloroform extract. (See Chart 1.)
The ether was removed from extract I1 and the residue treated with
alcohol; this alcoholic extract (IIa) contained almost all the dibenzanthracene
(0.1 mg.) present in the ether extract (11). A second extraction of the
residue with alcohol (IIb) removed somewhat less than 0.02 mg. of dibenzanthracene. The residue was taken up in ether (IIc) ; no dibenzanthracene
was found in this fraction.
Hence, extraction of the tumor lipoids with three successive portions of
ethyl alcohol removed all of the dibenzanthracene, thus separating it from a
large amount of interfering material.
The alcoholic extract showed the presence of dibenzanthracene unmiitakably, together with a pronounced diminution in the continuous absorption.
For some purposes this is all the purification required. However, on analysis
of similar extracts of 5th and 6th generation dibenzanthracene tumors, no
hydrocarbon was detected: the second group of bands of the hydrocarbons
could not be found and the third group, if present, was masked by continuous
339
STUDIES IN CARCINOGENESIS
absorption. Therefore, for the detection of less than 0.01 mg. dibenzanthracene per c.c., the residual continuous absorption had to be eliminated.
Fractionation of the AZcohoE Extracts: The alcohol extract ( I ) was treated
with an equal volume of hexane, and enough water added to cause separation
into two layers, both of which were strongly yellow. Spectrographic examina-
removed
ether
alcohol extraction
chloroform extraction
I
0.7 mg DBA
11
0.1 mg. DBA
I
I
I
added hexane
and water
1
I
I
1
removed
ether
I
I Ia
(alcohol
extract)
0.1 mg. DBA
IiI
0.00 mg. DBA
I
lib
IIC
(alcohol
(ether
extract)
extract)
0.02 mg. DBA 0.00 mg. DBA
Ib
Ia
(hexane)
(aq.alc.1
0.6 mg. DBA 0.1 mg. DBA
added
hexane
Ic
(hexane)
0.1 mg. DBA
Id
(aq.alc.)
0.01 mg. DBA
added
hexane
I
I
Ih
Ig
(hexane)
(aq.alc.)
0.01 mg. DBA 0.00 mg. DBA
CHART
I. SUMMARIZING TIIE PROCEDURE
ANTHRACENE (DBA), PRESENT
IN
EMPLOYED
I N SEPARATING THE 1 : 2 ; 5 : 6-DIBENZDIBENZANTHRACENE-INDUCED
TUMORS,
FROM
SPECTROSCOPICALLY
INTERFERING
MATERIAL
tion showed that most of the dibenzanthracene (0.6 mg.) had gone into the
hexane layer (Ia), leaving only a small amount (0.1 mg.) in the aqueous
alcohol (Ib) , The continuous absorption-producing material was about
equally divided between the two solvents.
The aqueous alcohol layer (Ib) was then treated with fresh hexane. This
time the hexane layer was colorless, while the alcohol layer remained strongly
yellow. Again most of the dibenzanthracene (0.1 mg.) went into the hexane
340
E G O N LORENZ A N D M. J. S H E A R
(Ic), leaving only a trace (0.01 mg.) in the alcohol (Id). Most of the
continuous absorption remained in the alcohol, whereas the hexane layer
showed little.
The aqueous alcohol fraction Id was then extracted with two fresh portions of hexane; these hexane extracts (Ig) were colorless, while the alcohol
(Ih) retained the deep yellow color. The hexane (Ig) extracted all of the
small amount of dibenzanthracene (0.01 mg.) remaining in Id, while the
residual alcohol (Ih) contained no detectable hydrocarbon. Furthermore,
the hexane extract (Ig) showed very little continuous absorption, while the
alcohol layer (Ih) showed it strongly.
The Hexane Extract: The dibenzanthracene was now concentrated in the
hexane extract but, in spite of the large amount of interfering material which
had been eliminated by the manipulation, continuous absorption was still
present. Since in the previous steps much of the continuous absorption had
stayed in the aqueous alcohol layer, the hexane fraction was now treated with
fresh alcohol and enough water added to cause separation into two layers.
However, this step did not effect further separation of the hydrocarbon from
the interfering substances.
Attempts at further separation were made, without much success, by the
employment of xylene, of cyclohexane, and of methyl alcohol. I n the course
of the manipulations a considerable amount of red wax, insoluble in organic
solvents, separated out from the methyl alcohol solution; it dissolved in aqueous NaOH and did not contain dibenzanthracene. This fraction exhibited
some general absorption, but the fraction which contained the dibenzanthracene still showed more general absorption than was considered desirable.
Since no marked advantage was gained by the use of these solvents and
procedures, the partially purified dibenzanthracene fractions were all combined and subjected to a repetition of the earlier steps. The solvents were
removed by evaporation and the residue extracted three times with hot ethyl
alcohol. The alcohol extracts were combined and treated with hexane and
a little water, as before. Almost all of the dibenzanthracene was removed by
three extractions with hexane; it required three additional extractions with
hexane, however, to remove the last traces of the hydrocarbon from the
aqueous alcohol.
This procedure demonstrated that the dibenzanthracene could be extracted
regularly with ethyl alcohol, and could be removed completely from the alcohol
by hexane. In the course of this repetition of the procedure further quantities of interfering material were removed, but the solution still showed some
continuous absorption in addition to the characteristic bands of dibenzanthracene.
At this point treatment of the material in CH,OH solution with dilute
aqueous NaOH in the cold removed a considerable amount of the yellow color
and some of the continuous absorption.
Since the method of partition between different solvents diminished the
continuous absorption considerably, it was applied to extracts of transplanted
dibenzanthracene tumors.
Absorption Spectra of Lipoids of Transplanted Tumors: Tumors induced
by the same lard solution were transplanted into other mice for several gen-
STUDIES I N CARCINOGENESIS
341
erations. When the tumor was in its 5th generation, 36 gm. of healthy tumor
tissue was employed in the spectrum analysis. The tissue, treated as before,
gave 6 gm. of dry material, which was extracted with ether for two days. The
tumor lipoids were then extracted five times with hot alcohol. The alcohol
extract was next extracted six times with hexane as described above. No
evidence of dibenzanthracene was found.
A similar procedure was followed with 148 gm. of healthy tumor tissue
from 6th generation transplants of the same tumor; this gave 25 gm. of dry
material. After application of the same extraction procedure that was employed with 5th generation material, the hexane fractions of both the 5th and
6th generation tumors were combined. No dibenzanthracene was detected.
I t should be noted, however, that the combined hexane fractions from 184 gm.
of tumor tissue showed sufficient continuous absorption to obscure the third
group of absorption bands of dibenzanthracene, if present. It could be concluded, however, that if any dibenzanthracene were present, the total amount
was less than 0.02 mg. in 184 gm. of tissue. T o rule out the possible presence
of still smaller amounts of dibenzanthracene, to the limit of sensitivity of the
spectrographic method, the last traces of continuous absorption would have
to be removed.
The method of partition between different solvents did not effect complete removal of the interfering material. Other technics were therefore required. The method of partition between solvents had the advantage that
none of the dibenzanthracene was lost. Since employment of other methods
of separation might result in loss of some of the hydrocarbon, the methods
were first applied to a solution of liver lipoids to which a known amount of
dibenzanthracene had been added.
Twenty-four grams of mouse liver lipoids were dissolved in 225 C.C. of
ether and, after 50 C.C. was set aside as a control, dibenzanthracene was added
to the rest to give a concentration of 0.003 mg. of the hydrocarbon per C.C.
Aliquot portions of this solution were shaken up with various adsorbents and
centrifuged. Some adsorbents removed both the interfering substances and
the dibenzanthracene; others removed a large part of the hydrocarbon and
only a little of the interfering material. The best results were obtained with
ignited A120,, one treatment removing a good deal of the continuous absorption and leaving most of the dibenzanthracene in the solution. However,
repeated treatment with A120, removed further amounts of dibenzanthracene.
Chilling of a CH,OH solution of the liver lipoids also was effective in
separating the dibenzanthracene from a large part of the interfering lipoids.
An aliquot portion of the ether solution of liver lipoids containing dibenzanthracene was evaporated to dryness and taken up in CH,OH. The solution
was then chilled in solid CO, and filtered in an apparatus packed in the same
refrigerant. A large amount of lipoid material was thus removed, without
removing detectable amounts of the dibenzanthracene.
A combination of the chilling and adsorption was tried on a solution containing only dibenzanthracene, in a concentration of 0.003 mg. per C.C. No
hydrocarbon came out of the solution upon chilling. Treatment of this solution with ignited A1208was also found to be satisfactory in that none of the
hydrocarbon was removed by adsorption.
3 42
EGON LORENZ AND M. J. SHEAR
The same combination of chilling and adsorption was then tried on an
aliquot portion of the dibenzanthracene-liver lipoid solution, after removing
the ether and dissolving in CH,OH. Most of the continuous absorption and
only a small part of the dibenzanthracene was removed by one ' treatment.
The combination of chilling and adsorption on Alsos was next applied to
the combined fractions of the partially purified 5th and 6th generation extracts.
The amount of continuous absorption was thereby reduced to the point where
less than 0.01 mg. of dibenzanthracene should have been detected if present.
However, dibenzanthracene was not detected.
A fresh sample of 120 gm. of healthy tumor tissue obtained from 6th
generation transplants was treated according to the scheme outlined above,
including the ehilling and adsorption with A1,03. No dibenzanthracene
was detected.
DISCUSSION
I n 59 gm. of healthy tumor tissue obtained six to eight months after the
treatment of mice with dibenzanthracene in lard solution, 0.8 mg. of the
hydrocarbon was still present according to spectrum analysis. After repeated transplantation of these tumors no dibenzanthracene was found in 36
gm. of tissue from 5th generation or in 148 gm. of tissue from 6th generation
dibenzanthracene tumors. A second batch of 120 gm. of tissue from 6th
generation dibenzanthracene tumors also gave no indication of the presence
of this hydrocarbon.
Dibenzanthracene was readily detected, using a 1 cm. cell, in solutions
containing 0.001 mg. of the hydrocarbon per C.C. The presence of lipoids
which produce continuous absorption in the ultraviolet masks the presence of
small amounts of dibenzanthracene. However, even after most of the continuous absorption had been eliminated, no dibenzanthracene was detected in
5th and 6th generation tumors. In view of the fact that continuous absorption was still present in the purified tumor extracts, it cannot be stated that
no dibenzanthracene was present. The amount of continuous absorption
was, however, small enough to enable 0.01 mg. of dibenzanthracene to be
detected.
It is therefore estimated that the amount of dibenzanthracene, if present
at all, was less than 0.01 mg. in the 36 gm. of 5th generation transplants and
in 148 gm. of 6th generation transplants of dibenzanthracene tumors.
Chalmers found that dibenzanthracene disappears rapidly from the site
of injection. Berenblum and Kendal did not detect any dibenzanthracene
eighteen months after injection. The question arose as to whether the hydrocarbon had been removed from the site of injection or whether it had undergone chemical change. I n the experiments described here, however, dibenzanthracene was readily detected in tumors six to eight months after injection
of a lard solution of the hydrocarbon. Furthermore, dibenzanthracene recovered from cholesterol-dibenzanthracene pellets (Shear, 1935), after con:
tact with subcutaneous tissue for a year or more, showed no evidence of
chemical change.
8 Repeated treatment with A1,0., removed further amounts of dibenzanthracene until, after the
5th treatment, about half of the dibenzanthracene was removed; there was, however, no concurrent
diminution in the amount of continuous absorption.
STUDIES IN CARCINOGENESIS
343
SUMMARY
1. A method is described for the identification and quantitative determination of 1:2 :5 :6-dibenzanthracene by means of absorption spectrum analysis.
2 . Dibenzanthracene may be readily identified down to a concentration
of approximately 0.01 mg. per C.C. If known to be present, it can be detected
down to a concentration of approximately 4 X lo-' mg. per C.C.
3. Substances present in the lipoid fraction of tumors and showing continuous absorption in the region of the dibenzanthracene absorption spectrum
can be largely removed by chemical procedures.
4. Tumors induced by a lard solution of dibenzanthracene were found to
contain an appreciable amount of the hydrocarbon.
5 . Fifth and sixth generation transplants of these dibenzanthracene-induced tumors failed to show the presence of dibenzanthracene.
REFERENCES
BERENBLUM,
I., AND KENDAL,L. P.: Brit. J. Exper. Path. 15: 366-371, 1934.
CHALMERS,
J. G.:Biochem. J. 28: 1214-1218, 1934.
CLAR,E.: Berichte 62: 350-359, 1929.
COOK,J. W.,HIEGER,I., KENNAWAY,
E. L., AND MAYNEORD,
W. V.: Proc. Roy. SOC.,Series
B 111: 455-484, 1932.
SHEAR,M. J.: Am. J. Cancer 26: 322-332, 1936.
WINTERSTEIN,
A.,AND SCHON,K.: Ztschr. f . physiol. Chem. 230: 146-158, 1934.