The Carcinogenic Activities of Certain Halogen Derivatives
of 4-Dimethylaminoazobenzene
in the Rat*
J. A. MILLER,
PH.D.,
R. W. SAPP, M.S.,
AND E. C. MILLER,
(From the McArdle Memorial Laboratory, Medical School, University of Wisconsin,
Previous
studies
from
this laboratory
(17, 19)
have demonstrated
that methyl, chioro, or nitro
groups in the @‘,
3', or 4' positions of the hepatic
carcinogen,
4-dimethylaminoazobenzene,
pro
duced parallel alterations in the carcinogenicity of
this dye. In each series the 4'-isomer was the least
active and the 3'-isomer was the most active. Fur
thermore, except for the very high activity of the
3'-methyl derivative, the activities of the various
derivatives with substituents
in any one position
TABLE
1
ANALYSES ON THE PREVIOUSLY UNDE
SCRIBED AMINOAZO DYES
PER CENT N (Dus&ts)@
Tha
M.P.
5C.
FoUND
6, Wisconsin)
ity of 4-dimethylaminoazobenzene.
The present
paper reports mainly on the activity of this dye
after substitution in the @‘,
3', or 4' positions with
a) the 0, p-directing fluoro, F-, group which forms
the strongest known bond with carbon, and b) the
highly stable rn-directing and ring-deactivating
trifluoromethyl,
CF3-, gi@oup. The results ob
tamed with the former compounds bear on the
problem of whether the occurrence of a benzidine
rearrangement
(4, 11) in the metabolism of ami
noazo dyes could be of importance in the induc
tion of tumors by these dyes. In addition the
carcinogenic activities of several other dyes re
lated to 4-dimethylaminozaobenzene
are recorded.
METHODS
ORET
(coan.)
Madison
PH.D.
ICAL
Preparation of compound.s.—A1lbut one of the
Derivatives of 4-dimethyl
aminoazobenzene:
compounds fed in these experiments were prepared
110.0—111.517.10,17.6817.27128.0—124.017.50,17.66“153.5—154.517.29,17.41“118.0—119.014.16
2'-fluoro
in this laboratory. The melting points of the corn
3'-fluoro
pounds
which have been described previously
4'-fluoro
agreed well with those in the chemical literature
2'-trifluoromethyl
79.014.43,14.77“173.0—173.514.31,14.33“129.0—130.014.42,14.5714.29144.0—145.012.26,12.4912.7983.5—84.517.8
S'-trifiuoromethyl
and are given in the text. Since several of the com
4'-trifluoromethyl
pounds have not been described previously, the
2',5'-dichloro
2',4',6'-trichloro
nitrogen contents and corrected melting points of
2',3-Dimethyl-4-mono
the new compounds are listed in Table 1. For
methylaminoazoben
these determinations
100 to @OO
mg. of each corn
zene
4-Formylmonomethyl110 . 5—111.5 17 .94, 17 .97
“
pound was purified by treating a hot solution in
aminoazobenzene
benzene with charcoal, passing the cooled solution
S These determinations were made by the Clark Microanalyt.ical Labora
through 6―X 1―column of alumina, and finally
tory, 1044 West Main St.reet, Urbana, Illinois.
crystallizing
were very
similar.
Apparently
the position
of the
group in the molecule had a greater effect on the
carcinogenicity
of the dye than did the type of
group involved.
An interpretation
of these results as represent
ing the effects per se of the nitro and methyl
groups on the activities of these dyes in vivo must
be made with caution since these groups may be
converted in vivo to amino and carboxyl groups,
respectively.
Accordingly
we are studying the
effects of more stable groups on the carcinogenic
*
This
investigation
was
supported
in
part
by
grants
Of the amines required
dimethylaniine,'
ether
(Skelly
for these syntheses
p-fluoroaniline,2
aniline,'
m-trifluoromethylan
-
ilme,2 2, 4, 6-tribromoaniline,' 2, 4, 6-trichloroaniline,' 2,
5-dichloroaniline,'
and
o-toluidine'
were
obtained
corn
mercially. Dr. R. G. Jones of the Eli Lilly Research Lab
oratories very generously supplied us with the o-and p
trifluoromethylanilines;
his syntheses of these difficultly
obtainable amities have been published (10). The o- and
1 Eastman
from
the National Cancer Institute, U.S. Public Health Service and
the Jane Coffin Childs Memorial Fund for Medical Research.
from benzene-petroleum
Solve B). The melting points of the compounds
were raised no more than 1 or @O
C. by this
purification.
2 Fluor-Syn
Kodak
Company,
Laboratories,
Rochester
1128
North
4, N.
Water
Y.
St.,
Milwaukee
2, Wis.
65@
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1949 American Association for Cancer Research.
MILLER
et aL—Derivatives
653
qf 4-Dimethylaminoazobenzene
m-fluoroaniines were prepared by the general methods
of Schiemann and Pillarsky (@3),but the yields of the
ative formed by coupling diazotized o-toluidine with
nitrobenzenediazoniurn
was rearranged (19) by heating it at 47 to 50°C. for 70
minutes in the presence of excess N-methyl-o-toluidine
and the hydrochloride of this amine. The dye was pre
borofluorates
used
in these
preparations were increased by following the modifica
tions introduced by Starkey (24). m-Hydroxydimethyl
aniline was prepared from m-nitroaniine' by methyla
tion with dimethyl sulfate,' reduction of the nitro
group with stannous chloride in concentrated HC1, and
finally replacement of the amino group with a hydroxy
group (1).
N-methyl-o-toluidine
cipitated
(19). The diazoamino
as the hydrochloride
compound
which was then decom
posed with excess alkali. Aften recrystallization from
benzene-petroleum ether the yield was 40 per cent of
theoretical.
For the preparation of 4-formylmonomethylamino
(16) was
4-Dirnethylaminoazobenzene
(m.p. 117-118°), @-hy azobenzene, 4-monomethylaminoazobenzene
droxy-4-dimethylaminoazobenzene
(m.p. 163—165°),refluxed in toluene (15 ml. per gm. of dye) for 4 hour
and the fluoro and trifluoromethyl derivatives of 4-di
with three equivalents of 88 per cent formic acid. The
toluene was then distilled off slowly and, when less than
rnethylaminoazobenzene
were prepared
by diazotizing
1 mole of the appropriate primary amine with 1 mole of half of the toluene remained, a second and identical
sodium nitrite (25 per cent aqueous solution) in the charge of toluene and formic acid was added. After all
presence of 2.5 moles of hydrochloric acid and about
of the toluene had been removed by distillation, the
30 moles of water at 0 to 3°C. For coupling, the diazo
residue was taken up in ethanol, precipitated by the
solution was added to a cooled (15°C.) 70 per cent
addition of water, and finally recrystallized from eth
ethanol solution of 1 mole of the desired tertiary amine anol-water. 4-Monomethylaminoazobenzene was easily
in the presence
of 2.5 moles
of anhydrous
sodium
ace
regenerated from the formyl derivative by acid hydroly
tate. In the case of the fluoro and trifluoromethylan
sis. The preparations
of 4'-nitro-4-dimethylaminoazo
iines it was necessary to heat the acid mixture to dis
benzene (m.p. 227—229°
C.) and 2, 4'-diamino-5-dimeth
solve the amine. The fine crystals of amine hydrochlo
ylaminobiphenyl
trihydrochlonide
have been described
ride which appeared when the solution was cooled previously (19).
rapidly, went slowly into solution upon diazotization.
Method of atsay.—Young adult male albino rats3
The trifluoromethyl and 2-hydroxy derivatives of 4-di
weighing 160 to 190 gm. were housed in screen-bottom
rnethylaminoazobenzene
were recrystallized
from eth
cages in groups of 5 to 7 with food and water available
anol-water mixtures, while 4-dimethylaminoazobènzene ad libitum. All of the compounds were fed in a semi
and its fluoro derivatives were recrystallized from hen
synthetic diet (13, diet 3) consisting of crude casein, 12;
zene-petroleum ether mixtures. The yields of the Vitab rice bran concentrate,
2 ; salts, 4 ; glucose mono
recrystallized dyes were approximately 80 per cent of hydrate (cerelose), 77; and corn oil, 5. After analysis of
the dietary constituents sufficient crystalline riboflavin
theoretical.
2, 5-Dichloroaniine
was diazotized by .dissolving
was added to bring the total content of the diet to the
0.1 mole in 450ml. of 10 N hydrochloric
acid and 210 ml.
of water, cooling to 10°C., and then adding the sodium
nitrite solution (0.1 moles) within 30 seconds (22). The
diazo solution and dirnethylanili@iewere then coupled as
described above to give 2', 5'-dichloro-4-dimethylami
noazobenzene. After recrystallization
from benzene and
petroleum ether the yield was 68 per cent of theoret
ical.
Fifty
per cent yields
aminoazobenzene
of 2', 4', 6'-tribromo-4-dimethyl
(m.p. 162—162.5°
C.) and 2', 4', 6'-
trichloro.4-dimethylaminoazobenzene
were obtained
by
the general procedure outlined by Saunders (22). One
tenth mole of sodium nitrite dissolved in 70 ml. of con
centrated
sulfuric
acid was added
0.1 mole of the halogenated
slowly
to a solution
of
amine in 40 ml. of concen
trated sulfuric acid while the temperature was main
tained below 20°C. The diazo solution was diluted by
pouring it slowly into 2 to 3 times its volume of crushed
ice and water and the mixture was added with efficient
stirring to 350 ml. of cold 70 per cent ethanol containing
0.1 mole of dimethylaniine and 82 gm. of anhydrous
sodium
acetate.
The dyes were recrystallized
from a
mixture of benzene and petroleum
ether.
2', 3-Dimethyl-4-aminoazobenzene
(“o-aminoazotol
uene―or “2-amino-5-azotoluene―)
obtained commer
cially' was recrystallized from ethanol-water to a m.p.
desired
level.
The
rats
of the
ceived only 1 mg. of riboflavin
first
series
(Table
2) ne
per kg. of diet for the
first 8 days. At that time the weight losses were so severe
that the level was raised to 2 mg. per kg. of diet for the
remainder of the first series and for the two subsequent
experiments (Tables 3 and 4). In a fourth series the
rats fed 4-dimethylaminoazobenzene
and one of the
groups fed the 4'-fluoro derivative received 2 mg. of
riboflavin per kg. of diet while the diet for the second
group fed 4'-fluoro-4-dimethylaminoazobenzene
con
tamed 10 mg. of riboflavin per kg. 4'-Nitro-4-dimeth
ylaminoazobenzene and 2, 4'-diamino-5-dimethylami
nobiphenyl tnihydrochlonide were added to the diets
after being ground in a mortar with a small quantity of
glucose; the other compounds were dissolved in the
corn oil of the diet with mild heat. Each rat received 1
drop of halibut liver oil per month.
In the first series (Table 2) each compound was fed to
13 or 14 rats at a level of 2.67 millimoles per kg. of diet
(equivalent to 0.060 percent of 4-dimethylaminoazoben
zene). At the end of 8 months the rats fed 4-dimethyl
aminoazobenzene and its 4'-fluoro derivative were ex
amined for tumors by laparotomy. Those fed the 4'-fiu
oro derivative were then transferred to the basal diet
(no dye added) for 2 months, while the other group was
fed 4-dimethylaminoazobenzene
for an additional
of 101 to 102°C. 2', 3-Dimethyl-4-monomethylami
noazobenzene
was prepared
from the diazoamino
deny
Holtzman Rat Co.,Madison,Wis.
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1949 American Association for Cancer Research.
654
(‘ancer Reseanh
month, examined
by laparotomy
then fed the basal diet for @2
months.
a second time, and
@‘,
4', 6'-Tribromo-,
@2',5'-dichloro- , and 4 ‘-nitro-4-dimethylaminoazoben
additional ‘2
months. Since the rats would not survive
with continuous feeding of @‘,
4', 6'-trichloro4-dimeth
they
were transferred
to the basal
diet for 1 month periods at the end of the fifth and sev
enth
months
of dye-feeding.
The
which the dye was actually
@,,
total
ingested
3.-Dimethyl-4-aminoazobenzene
methyl
period
during
was 10 months.
and
derivative
were fed continuously
its
N-mono
for 14 months.@
In the second series (Table 3) 2.40 millimoles of
4-dimethylaminoazobenzene
and its three fluoro deny
atives were fed per kg. of diet for 3 months while 2-hy
(lroxy-4-dimethylaminoazobenzene
was fed at a level of
@2.67millimoles
per kg. for 8 months.
At the end of the
(lye-feeding
periods
the livers were examined
by lap
arotomy,
and after 2 months
on the basal diet the rats
were killed for a final tumor count.
In the third series (Table 4) 4-dimethylaminoazoben
zene and its three tnifluoromethyl derivatives were at
first fed at a level of 2.67 milhimoles per kg. Due to the
poor health
of the rats fed the 2'- and 3'-tnifluoromethyl
derivatives it was necessary after 5 weeks to transfer the
rats fed 4-dimethylaminoazohenzene
and the tnifluoro
methyl derivatives to the basal diet for 10 days, and
then feed a reduced level of these four dyes (2.13
millimoles per kg.) for the remainder of the experiment.
After 4-dimethylaminoazobenzene
had been fed for a
total of 5 months and the tnifluoromethyl derivatives
and 4-formylmonomethylaminoazobenzene
(2.67 milli
moles per kg.) for 8 months, the livers were examine@I by
laparotomy
and the animals
maintained
on the
diet for 2 months.
2, 4'-Diamino-5-dimethylaminobi
basal
phenyl tnihydrochloride
was fed at the higher level of
5.34 millimoles per kg. for 10 months. Each group in
series two and three
consisted
of 14 rats.
In the fourth series both 4-dimethylaminoazoben
zene and its 4'-fluoro derivative were fed at a level of
2.40 millimoles per kg. of diet. Three rats from each of
the groups fed 4-dimethylaminoazobenzene
fluoro-4-dimethylaminoazobenzene
and 4'-
in the low riboflavin
(2 mg. per kg.) diet were killed for hepatic
(2) and protein-bound
dye (12, US) analyses
riboflavin
at 1, 2, 3, 4,
6, 8, 13, and 18 weeks. These analyses were also made on
the
livers
of 3 rats
fed
tional 2 months before being killed for the final tumor
count.
zene were fed for 8 months before the rats were subjected
to laparotomy and transferred to the basal diet for an
ylaminoazohenzene,
containing the same levels of riboflavin for an addi
4'-fluoro-4-dimethylaminoazo
After
the dyes
which
had not previously
been
tested for carcinogenic activity were fed for at
least 2 weeks in the above experiments the absorp
tion of the compounds from the gastrointestinal
tract was investigated
by determining
the dye
content of 24 hour fecal samples. The feces were
air-dried at 37°C., powdered, and extracted with
acetone in a Soxhlet apparatus
for at least 24
hours. An aliquot of the acetone extract was
evaporated to dryness in vacuo, the residue was
dissolved
in 7 N HC1, and the light absorption
of
this solution was compared with that of a standard
solution of the compound fed.
RESULTS
The results
of the three
series
in which
a total
of 14 aminoazo dyes and one possible non-azo
rnetabolite
of 4-dimethylaminoazobenzene
were
tested for carcinogenicity are tabulated in Tables
2 to 4. The tumor incidences of the three control
groups fed 4-dimethylaminoazobenzene
agreed
well with those obtained previously (19, 21). The
other compounds had activities ranging from zero
to nearly twice that of the reference compound.
Although the carcinogenicity of 2', 3-dimethyl
4-aminoazobenzene
(“o-aminoazotoluene “)
for the
rat liver has been definitely
established
(8), no
tumors4 were obtained in this experiment
when it
was fed at a level of 2.67 millimoles per kg. of
diet for 14 months (Table 2, group 7). Likewise, no
liver tumors4
were found
when its N-methyl
deny
ative, 2' , 3-dimethyl-4-monomethylaminoazoben
zene (group 8) was fed for 14 months at the same
level. However, two of the rats fed the latter corn
pound had tumors at other sites ; one developed a
carcinoma arising from the sebaceous glands near
the eye after 12 months while a mixed basal and
squamous cell carcinoma arising from the ear
duct was found in a second rat after 14 months.5
This
ear tumor
was similar
to those
found
after
benzene in the high riboflavin (10 mg. per kg.) diet and
the basal diet containing low or high levels of riboflavin
administration
of certain amines such as 2-ace
tylaminofluorene
(26), certain heterocyclic ana
for 3, 8, and 18 weeks. To determine
logues
the effect of the
level of dietary riboflavin on the carcinogencity of 4'fluoro-4-dimethylaminoazobenzene
the livers
13 rats fed the dye in the diets containing
high levels of riboflavin,
laparotomy
4 Note
respectively,
of 17 and
the low and
were examined
by
at 13 weeks. They were then fed basal diets
added
in proof:
After
these
rats
had
been
fed
the
dye for a total of 16 months they were continued on the basal
diet for 2 more months and then killed. By this time 2', 3-di
methyl-4-aminoazobenzene
had produced benign hepatomas
in 2 of the remaining 9 rats; 2 of the 10 rats that had been
fed
the
N-methyl
derivative
developed
similar
tumors.
of the
latter
compound
(14),
and
the
4-
aminostilbenes (7).
The polyhalogen derivatives 2', 4,' 6'-tnibromo-,
2', 4', 6'-tnicholoro-, and 2', 5'-dichloro-4-dimeth
ylaminoazobenzene
(Table 2, groups 3—5)were
all inactive even after being fed for 8 to 10 months,
although
the tnichloro derivative
did produce
mild cirrhosis and cysts in the liver after being
are indebted to Drs. H. P. Rusch and J. M. Price for
the histological examinations
of the tumors reported in this
paper.
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1949 American Association for Cancer Research.
@
@—ION=NOMcH*
..@
MILLER
et al.---Deriratires
(group
6) was
likewise
genie in these tests; this compound was fed to corn
plete the series of derivatives containing a nitro
tnibromo
and 4'-nitro
derivatives
655
fluoro derivatives (Table 3, groups 10 and 11) ap
peaned to be equally active and almost twice as
carcinogenic as 4-dimethylaminoazobenzene
. 2'Fluoro-4-dimethylaminoazobenzene
(group
12)
induced tumors a little faster than 4-dimethylami
noazobenzene in this experiment, but the differ
ence was small. In the same experiment 2-hydroxy
4-dimethylaminoazobenzene
(group 13), a possi
ble metabolite of the parent compound, did not
non-cancino
group on the prime ring (19). The inactivity
j.@
of 4-DimethylaminGazobenzene
fed for 8 to 10 months. 4'-Nitro-4-dimethylamino
azobenzene
@.
of the
is of doubtful
significance,
however; in each case approximately
half of the dye fed was apparently
not absorbed
since it was excreted in the feces unchanged. On
the other hand no more than 3 per cent of the 2',
4', 6'-tnicholoro-4-dimethylaminoazobenzene
and
produce any tumors or gross liver damage
after
being fed for 8 months; the 2'-, 3'-, and 4'-hydroxy
ported in this paper were excreted in the feces.
In the first series 4'-fiuoro-4-dimethylaminoazo
derivatives are also non-carcinogenic (19).
While replacement of a hydrogen on the ben
zene ring with a fluorine atom enhanced the activ
benzene
ity of 4-dimethylaminoazobenzene,
less than
1 per cent
(group
of the other
2) induced
compounds
liver tumors
re
in 7 of 12
less active
than
4-dimethylaminoazobenzene
replacement
of
the three hydrogens of the methyl groups of 3'and 2'-methyl-4-dimethylaminoazobenzene
by flu
onine destroyed their activities (Table 4, groups 16
and 17; cf. Table 5). 4'-Tnifluoromethyl-4-dimeth
ylaminoazobenzene
(group 15) was likewise in
active when fed for 8 months. @fh rats receiving
the 2'- and 3'-trifluoromethyl
derivatives were in
rats by 3 months while none of the rats fed 4dimethylaminoazobenzene
had gross tumors at
that time. Since all other substitutions
in the 4'
position have resulted in compounds considerably
(17,
19), it was of interest to feed the 2'- and 3'-fiuoro
derivatives as well. In this series the 4'- and 3'TABLE
2
THE CARCIN0GENIcITIzs OF CERTAINHALOGENDERIVATIVES
OF 4-DIMETHYLAMLNOAZOBENZENE
—
Time
GroupCompound
fed
(2.67 rrillmoles per kgm. diet)—
Incidence
dye*
of
liver tumorsGross
cirrhosis
@endof
feeding
@yefedmonii@
mild
13l2@moderate
4-D@n@ninoazobenzene
23
severe—
F(j@NN(jN(CH@
2@/
@-F@ora-4-dm@h@ninoazobenzene
3
Cl (3N=N(JN(CM3)@
2moderate
none
lot
—
Br€@1=N@O@N(cH3)2
4———,I;'9@
none—
@‘9/mild
8@/
@Tñ&omo4&@n@
Ø4=NON(CH3)249,/
5—
.,@N--@-6O@NON=N(j*J(CH@
@N!t!O-4-dmeth.*Imnoazobenzene
€@N@
7
14-@/‘N8
@i@@eth*4-arr5flOozobenzene,,,I@-9-@N—
——
-@-c@3=N€@H3‘I
-%N
@Number
of rats with tumors over nurrter of rats alive at end d feeding dye
tSee text
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1949 American Association for Cancer Research.
TABLE
3
THE CARCINOGENICITISH
OF CERTAIN MONOFLUORO
DasuvATIvES
OF 4-DIMETHYLAMINOAZOBENZENE
in
Compound
Group.
-
dye-
dietTime
monthsmilimoles
fedLevel
@-
‘14mild4-Dimethylominoozobenzene10F(j@NN(j@N(CH3)2
cirrhosis
it
per kgm.fed3
[email protected])/
*
cidence of liver tumors
InGross
5
8
end
10feeding
148/
severeI
/3mild
i-fluoro-4-dimethylaminoozobenzene%12/
I
severe12
3'-Auoro-4-dimethylaminoazobenzeneNI@,@/7moderate
—ON-NQ@CH@
mild
1fl@th@lOIThnoazobenzene‘III%%moderate
—(j@N@N(jN(GH@
@
@
13
P@.
(j@NN(jN(CH@J2
2-Hydroxy-4-dmethlam@oa@ob@e@
2.67
8
o/
0
none
‘14
@‘Number
of rats with tumors over numberof rats
dyeTABLE
alive at end
feeding
4THE
TRIFLUOROMETHYLI)ERIvATIvER
CARCINOGENICITIES
OF CERTAIN
4-DIMIrniYLAMIN0Az0BENzENE—
OF
Level in Time
Compound feddiet
Group—
@
millimoles
mos@Incidence.@
per kgm.fed
..j@—
14@NNQN(CH@
moderate15F@C@N=N(jN(C@
of liver tumors@'Gross
cirrhosis
ut end of
feeding
compoundmonths
10
2.l35,@mIld
4- Dirnethylaminoazobenzene2.671
4@-TrifluoroniethyI-4-dimethylaminoazobenzene
F3b@@
@-Trfluoromethy@4-dth@brinoa@benzene
I
@‘
CF3
7NH0
<IjNN(jN(CH@
/9N—18c@NzNoN@:@
2@TrifkioromethyI-4-dimethybminoazobenzene.@894'94'nonel6No/@0%NI
/90/
mild—
!f9!mylmaiximethylaminoazobenzene
/ll4/ /llnone
@
195@34—;-none
3HCI
@
@0@[email protected]/
Number of rats with tumors over number of rats c@iveat end of feeding compound
The higher level was fed for the fi rst 5 weeksand the lower level for the remainderof the experiment
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1949 American Association for Cancer Research.
MILLER
poor health
during
the first month
of the expeni
ment so that it was necessary to feed the basal
diet for 10 days and then continue with a reduced
level of dye. 4-Formylmonomethylaminoazoben
zene (group 18), a possible intermediate
formation
of 4-monomethylaminoazobenzene
the dimethyl
active
than
compound
either
in the
from
in vivo, was much less
of these
compounds.
Although
it was fed for 8 months, the compound induced
liver tumors in only 4 of 11 rats by 10 months.
2, 4'-Diamino-S-dimethylaminobiphenyl
tnihydro
chloride (group 19), which has been suggested as a
possible metabolite
of 4-dimethylaminoazoben
zene (3, 4), did not induce
@
657
et al.—Derivatives of 4-Dimethylaminoazobenzene
any liver tumors
of hepatic riboflavin which can be maintained
when either 4-dimethylaminoazobenzene
(13) or
its 3'-methyl derivative is fed (6).
Like the other dimethylaminoazo
dyes which
have been studied (15) 4'-fluoro-4-dimethylami
noazobenzene gave rise to a considerable level of
protein-bound
aminoazo dye in the liver. In this
experiment the level of bound dye reached a maxi
mum between 3.5 and 4 weeks when either 4-di
methylaminoazobenzene
or its 4-'fluoro derivative
was fed in the low riboflavin diet (Fig. 1). After
the maximum levels were reached, the level of
bound dye derived from the 4'-fluoro derivative
even
though it was fed at a level of 5.34 millimoles per
kg. of diet for 10 months. This confirms an earlier
test in which the compound was fed for only 4
months (19). An adenocarcinoma arising from the
small intestine was found in one of the surviving
rats which were killed at 12 months.
The liver tumors induced by the three fluoro
dyes and 4@formyhnonomethylaminoazobenzene
were histologically similar to those induced by
058- 4-0@neft@@nmoaabeRere
0.021
0
0.0 IC
0.001
4-dimethylaminoazobenzene.5
Studies on 4'-fluoro-4-dimethytaminoazobenzene.
2
4
8
10
2
14
16
8
20
Weeks
—Like 4-dimethylaminoazobenzene
(21) the car
cinogenicity of 4'-fluoro-4-dimethylaminoazoben
aminoazobenzene and its 4'-fluoro and 4'-methyl (15) deriva
zene was markedly
FIG. 1.—The protein-bound
dye-time curves for 4-dimethyl
tives. The relative carcinogenic
activities
of the diet. Thus when the 4'-fluoro derivative was
fed in a diet containing 2 mg. of riboflavin per k'g.
for 35 months, the tumor incidence was 41 per
cent at 3.5 months and 82 per cent after 2 months
on the basal diet. The corresponding incidences for
the rats fed 10 mg. of riboflavin per kg. of diet
were 0 and 15 per cent respectively. Further, the
livers of the rats fed the high level of roboflavin
showed little or no gross damage while the livers of
those fed the lower level were moderately cirrhotic.
In this experiment only one of 11 rats fed 4-dirneth
in parentheses.
level of the bound
ylaminoazobenzene
level of bound dye after the maximum was found
when the very active 3'-methyl-4-dimethylami
noazobenzene was fed (15) . The levels of bound
dye in the livers of the rats fed 4'-fluoro-4-dimeth
ylaminoazobenzene
in the high riboflavin diet
for 3, 8, or 18 weeks were similar to those observed
when the low riboflavin diet was fed. The period of
altered
for
by the riboflavin
8.5
tumors at that time.
The ingestion of either
benzene
or
its
4'-fluoro
months
had
level
gross
4-dimethylaminoazo
derivative
in
the
low
riboflavin diet decreased the level of hepatic ribo
flavin from 14.4 @tgm.per gm. at the beginning of
the experiment to approximately
11.6 and 10.6
@gm.per gm. by 3 and 8 weeks respectively. The
livers of rats fed the same diet without the dye con
tained 12.8 and 17.3 @gm.of riboflavin per gm. at
3 and 8 weeks respectively. Ingestion of the high
riboflavin diet increased the level of hepatic
riboflavin to 17.5 and 18.1 @igm.per gm. for the
rats fed the basal diet and 14.3 and 15.4 jsgm. per
gm. for those fed 4'-fluoro-4-dimethylaminoazo
benzene
at 3 and 8 weeks respectively.
High
levels of dietary riboflavin
also increase the level
The micromolar
of the dyes are given
dye derived
from 4'-fluoro-4-dimethylaminoazobenzene was calculated
from the quantity of steam-distillable amine obtained follow
ing reduction
with tin and hydrochloric
general procedure
amine
described
was obtained
previously
from a solution
acid according
to the
(15). One micromole
of the bound
of
dye which
had a light absorption (log lo/l) at 520 ms of 12.0 when dis
solved in 4.5 ml. of the hydrochloric acid-ethanol solution. The
constant for 4-dimethylaminoazobenzene
is also 12.0 under
these conditions (15).
fell more rapidly than that formed from 4-dimeth
ylaminoazobenzene
dye-feeding
required
; a similar
before
sharp
drop
the maximum
in the
level of
protein-bound dye derived from the 4'-fluoro de
rivative was obtained fits in reasonably well with
similar data on the C-monomethyl
derivatives of
4-dimethylaminoazobenzene
(15). Thus the bound
dye-time curve for the very active 3'-methyl deriv
ative (activity = 10—12)reached a maximum after
2 weeks of dye-feeding while the bound dye-time
curves
with 4-dimethylaminoazobenzene.
(ac
tivity = 6),
2'-methyl-4-dimethylaminoazoben
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1949 American Association for Cancer Research.
Cancer Re,s'earch
658
zene (activity
= 2—3),and 4'-methyl-4-dimethyl
arninoazobenzene
after
(activity
approximately
spectively.
< 1) reached
4, 8, and
maxima
< 21 weeks,
re
support
f2-Methyl-4-dimethylaminoazobenzene
of the benzidine
hypothesis.
(activity = 0) and 3-methyl-4-monomethylami
noazobenzene (activity < 1) gave rise to maximum
hand, the very high activities
fluoro derivatives
constitute
hound
against
dye levels after
12 weeks of feeding.
DISCUSSION
The data presented above on the activities of
the halogen-containing
derivatives of 4-dimethyl
aminoazobenzene
have a bearing on the benzi
dine-rearrangement
hypothesis
of Elson and
Warren
(4). These
workers
found
that
the urine
from rats fed azobenzene, a non-carcinogenic corn
pound, contained a substance which yielded ben
zidine or 4, 4'-diaminobiphenyl
on treatment
TABLE
DzmvATIvES
3@J@ 2@
@ct@3L
dye
•
61
@L
@i:@
_c@3:
—
4'
3'
3
0
0
<I
0-12
2-3
0
5
3
-2
0
5-6
2
0
the
carbon-fluorine
substitution
4-dimethylaminoazobenzene
of the
fluorine
did not decrease
the
of this dye encourages one to attempt
tests of the benzidine
rearrangement
by
preparing
polyfluoro
derivatives
substituted in such a way that they could not un
dergo this rearrangement
upon reduction to the
hydrazo
The
data
compounds.
presented
in
this
and
previous
work
0
0
7
0
nogenicity
are
dent
that
0
alike
in
0
graded
(3) also considered
for
(17, 19) concerning the effect of substitution
in the
. ‘,
.
.
.
rings of 4-dimethylaminoazobenzene
on its carci
with acid. This substance was assumed to be a
derivative of hydrazobenzene and it was suggested
that a rearrangement
of this compound to benzi
dine might also take place in vivo. Elson and
Hoch-Ligeti
that
corresponding
0
0
2
fact
hypothesis
(unsubstituted
sithetituent
P2r@tL2!L HO-
the
activity
further
4-(jN@N@N(CH@)2
Activities
hypothesis
atom into any of the positions on the prime ring of
OF 4-DIMETHYLAMINOAZOBENZENE
Pek,tiu.
. .——...-
the
On the other
of the 4'- and 2'positive evidence
bonds in these compounds are among the strongest
known chemical linkages. Since para rearrange
ments usually take place with more facility than
those directed to the ortho positions, it might be ex
pected according to this theory that the 4'-fluoro
derivative would have a very low activity ; instead
it proved to be more active than the parent dye or
the 2' derivative. Furthermore,
the direct tests of
the rearrangement
product itself for carcinogenic
activity in the rat liver have so far yielded nega
tive results (Table 4, group 19 and (19) ). In any
case
5
THE CARcIN0GENIcmES OF VARIOUSRING-SUBSTITUTED
@
pound can be disregarded since no more than half
of the compound fed was absorbed, the inactivity
of the trichloro derivative is negative evidence in
it probable
that
4-dimethylarninoazobenzene
undergoes a similar
rearrangement
in vivo to yield 2, 4'-diamino-5-di
methylaminobiphenyl.
This compound, a substi
tuted p-phenylene .diamine, was found to be a
potent enzyme inhibitor (3) and hence might be
involved in the carcinogenic process induced by
the parent dye. Since at least one of the para or
ortho positions of aromatic hydrazo compounds
must be unsubstituted
for the benzidine rearrange
ment to take place (9), the substitution of one or
more groups in the 2', 4', and 6' positions of 4-di
methylaminoazobenzene
should prevent, or at
least hinder, this reaction from occurring in vivo.
Actually neither 2', 4', 6'-trichloroor 2', 4',
6'-tribromo-4-dimethylaminoazobenzene
was ac
tive. While the inactivity of the tribromo corn
summarized
chloro,
in
nitro,
producing
similar
carcinogenicities.
Table
5.
and methyl
series
of
Hydroxy
It
is
evi
groups
act
compounds
and
of
trifluoro
methyl groups are also alike in that they destroy
the ability of the molecule to act as a carcinogen.
The effect of the hydroxy group may be to provide
a point of attack for metabolic breakdown of the
dye before it can act carcinqgenically.
In fact,
4'-hydroxy-4-dimethylaminoazobenzene
is formed
from 4-dimethylaminoazobenzene
during its me
tabolism by rat liver homogenates (20) and there is
evidence that demethylated
hydroxyazo deriva
tives
are present
in the urine
of rats
fed the dye
(18). It is not known how the strongly rn-directing
and ring-deactivating
trifluoromethyl
group de
stroys
the
carcinogenic
activity
of
the
dye.
Equally difficult to explain are the observations
that the fluoro group in the 8' and 4' positions in
creases the activity of the molecule while the same
group in the 2'-position did not alter the activity
of the dye. This is particularly striking in the case
of the 4'-derivative since other substituents in this
position
either
abolish
or greatly
diminish
the ac
tivity
of 4-dimethylaminoazobenzene.
It is inter
esting
that the activity
derivative,
of the 4'-fiuoro
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1949 American Association for Cancer Research.
MILLER
like that of the parent dye, can be strongly inhib
ited by feeding a high level of riboflavin. On the
other hand, the activity of S'methyl-4-dimethyl
aminoazobenzene,
another derivative that is more
active than the parent dye, is not greatly reduced
by dietary additions of this vitamin (5). However,
like the other highly active aminoazo dyes (15) the
amount of protein-bound dye formed from 4'-flu
oro-4-dimethylaminoazobenzene
mum
within
diminishes.
1 month
This
rises
of dye-feeding
is quite
unlike
the
to a maxi
and then
bound
dye
time curve of 4'-methyl-4-dimethylaminoazoben
zene, a very weak carcinogen ( (15) and Fig. 1).
Although it was expected from the data in the
literature (8) that 2', 3-dimethyl-4-aminoazoben
zene (“o-aminoazotoluene―) would be a very
weak carcinogen under our conditions, the substi
tution of a N-methyl group did not increase its
activity for rat liver to a measurable
extent.
While this result might not have been expected in
view of the great increase in activity effected by
the monomethylation
of 4-aminoazobenzene
(17,
19, 25) it was known that a methyl group in the
S position does diminish strongly the activity of
4-monomethylaminoazobenzene
(19) . Similarly,
the low activity of 4-formylmonomethylaminoa
zobenzene, a possible metabolite of 4-dimethylami
noazobenzene,
was
659
et al.—Derirative8 of 4-Dimethylaminoazobenzene
surprising.
Evidently
the
N-formyl group in this compound is not readily
removed in vivo to yield the highly active dye,
aminoazobenzene,
whereas the 92'-fluoro derivative
had essentially the same activity as the parent dye.
The activity of the 4'-fluoro derivative was strong
ly inhibited by high levels of dietary riboflavin,
and the level of protein-bound
dye formed in the
liver from this dye reached a maximum after 1
month of feeding.
3. The substitution
of the stable rn-directing
tnifluoromethyl group in the 92', 3', or 4' positions
yielded dyes that were inactive even after 8
months of continued administration.
4. Two dyes that presumably could not undergo
the benzidine rearrangement
in vivo, 92', 4', 6'-tni
bromo- and 92',4',6'-trichloro-4-dimethylaminoazo
benzene, were inactive, but the tribromo deriva
tive was only poorly
absorbed
from
the intestinal
tract. However, the benzidine rearrangement
prod
uct of 4-dimethylaminoazobenzene,
2, 4'-diami
no-5-dimethylaminobiphenyl,
was
inactive
even
when fed at a high level for 10 months. In addi
tion, the high activities of the 2'- and 4'-fluoro
derivatives are considered to constitute positive
evidence against the benzidine rearrangement
hypothesis.
5. Two possible metabolites of 4-dimethylarni
noazobenzene were tested : 4-formylmonomethyl
aminoazobenzene
had only a low carcinogenic
activity and 92-hydroxy-4-dimethylaminoazoben
zene was inactive. The following miscellaneous
derivatives
were inactive under the conditions
4-monomethylaminoazobenzene.
used : 2', 3-dimethyl-4-arninoazobenzene
While it is difficult to evaluate the importance
of steric hindrance in determining the carcinogenic
activities of the compounds listed in Table 5,
this factor should probably receive more atten
tion. The “space-filling―
properties of the groups
under discussion appear in many cases to be of as
much importance as their effects on the electronic
noazotoluene―), 2', 3-dimethyl-4-monomethylami
noazobenzene,
and 4'-nitro-4-dimethylaminoazo
benzene.
6. The following new compounds are described:
2'-fluoro-, 3'-fluoro-, 4'-fluoro-, 2'-tnifiuoromethyl-,
3'-tnifluoromethyl-,
4'-tnifluoromethyl-,
92', 4', 6-'
trichloro-, and 2', 5'-dichloro-4-dimethylamino
azobenzene ; 92', 3-dimethyl-4-monomethylamino
azobenzeneand 4-formylmonomethylaminoazoben
zene.
structure
of the whole molecule.
These
considera
tions may apply, for example, to such inactive
dyes as 2', 4', 6'-tnichloro-4-dimethylaminoazo
benzene and the trifluoromethyl
derivatives
of
4-dimethylaminoazobenzene.
SUMMARY
1. The carcinogenic activities of 15 halogen and
other derivatives
of the hepatic
carcinogen,
4-dimethylaminoazobenzene,
were determined in
the rat. Some of these compounds were fed to test
the hypothesis
(Elson and associates) that the
parent dye undergoes a benzidine rearrangement
in vivo which is concerned in the carcinogenic
process induced by the dye.
2. The substitution
of the stable o, p-directing
fluoro group in the 3' or 4' positions yielded dyes
which were nearly twice as active as 4-dimethyl
(“o-ami
REFERENCES
1. BOBRANSKII, B. R., and EKER, I. M. Preparation
of the
Dimethylcarbamic
Ester of m-Hydroxyphenyltrimethyl
ammonium
Methyl
Sulfate (Prostigmine,
Proeserine).
J. Applied Chem. (U.S.S.R.), 14:524—527,1941. Chem.
Abstr.,36:3159—3160, 1942.
2. CONNER, R. T., and STRAUB, G. J. Combined Determina
tion of Riboflavin and Thiamin in Food Products. Indust.
and Engin. Chem., Anal. Ed., 13:885—388,1941.
3. Ei.sox, L. A., and Hocn-Liosrri,
C. The Inhibition
of
Urease and Succinoxidase
by Metabolic
Products
of
p-Dimethylaminoazobenzene
and by
Amines. Biochem. J.,40:880—391,1946.
Some Related
4. Eiso@, L A.,and WAuuxm, F. L. The Metabolism ofAzo
Compounds. I. Azobenzene. Biochem. J.,38: 217-220,1944.
5. GIESE, J. E.,Ci.@tTToN,
C. C.,Mru.sia,E. C.,and BAU
MANN,
C. A. The
Effect
of Certain
Diets
on Hepatic
Tumor
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1949 American Association for Cancer Research.
660
Cancer Research
Formation
Due to m'-Methyl-p-Dimethylaminoazoben
zene and o'-Methyl-p-Dimethylaminoazobenzene.
Cancer
Research,6:679-684,
1946.
6. GRIFFIN, A. C., CLAYTON,C. C., and BAUMANN,C. A. The
Effects of Casein and Methionine
on the Retention
of
Hepatic Riboflavin
and on the Development
of Liver
Tumors in Rats Fed Certain Azo Dyes. Cancer Research,
9:82-87,1949.
7. HADDOW, A., HARRIS, R. J. C., KON, G. A. R., and ROE,
E. M. F. The Growth-Inhibitory and Carcinogenic Prop
erties of 4-Aminostilbene
and Derivatives.
Phil.
Roy. Soc. London, Series A, 241:147—195,1948.
Carcinogenicity of p-Dimethylaminoazobenzene
Trans.
8. HARTWELL,J. L. Survey of Compounds Which Have Been
Tested for Carcinogenic
C-Monomethyl
Derivatives.
Cancer Research 9@336—343,
1949.
16. Miu.zai, J. A., and BAUMANN,C. A. The Determination of
p-Dimethylaminoazobenzene,
p-Monomethylaminoazo
benzene, and p-Aminoazobenzene
in Tissue. Cancer Re
search,
5:157—161,
1945.
17. Miu.zii, J. A., and BAUMANN, C. A. The Carcinogenicity
of Certain Am Dyes Related to p-Dimethylaminoazoben
zene. Cancer Research, 5:227-234,1945.
18. MILLER, J. A., and MILLER, E. C. The Metabolism and
Activity, pp. 196-203. Washington,
and Re
lated Compoundsin the Rat. Cancer Research, 7:39-41,
1947.
19. Mnmi,
D. C.: Federal Security Agency. U.S. Pub. Health Service,
J. A., and Miusm,
E. C. The Carcinogenicity
of
Certain Derivatives of p-Dimethylaminoazobenzene in the
1941.
9. JACOBSON,P., and Ku@z, R. V. Umlagerung des p-Di
methylamidohydrazobenzols.
Ann. Chem., Leipzig, 303:
853—361.1898.
Rat. J. Exp. Med., 87: 139—156,1948
20. MUELLER, G. C., and MILi.mi, J. A. The Metabolism of
4-Dimethylaminoazobenzene
by Rat Liver Homogenates.
10. JONES, R. G. Ortho and Pam Substituted Derivatives of
J. Am. Chem. Soc., 69:2346—2350, 1947.
21. RUSCE, H. P., BAUMANN, C. A., Miu.mi,
J. A., and
KLINE, B. E. Experimental Liver Tumors. In: F. R.
11. KIRBY,A. H. M. Studies in Carcinogenesis with Am Corn
Moulton, Am. Assn. Adv. Science Research Conference on
Benzotrifluoride.
pounds. II. The Action of Azo Compounds in Mice, and
the Bearing Thereof on Theories of Azo Dye Carcinogene
sis. Cancer Research, 5:683—696, 1945.
12. [email protected],E. C.,and MILLER, J.A. The Presenceand Sig
nificance of Bound Aminoazo Dyes in the Livers of Rats
Fed
p-Dimethylaminoazobenzene.
Cancer
Research,
7:468—480, 1947.
13. Mum@, E. C.,MILLER, J.A.,KUNE, B. E.,and RUSCH,
H. P. Correlation of the Level of Hepatic Riboflavin with
the Appearance of Liver Tumors in Rats Fed Aminoazo
Dyes. J. Exp. Med., 88:89-98,1948.
14. MILLER, E. C.,MILLER, J.A.,SANm@n,R. B.,and BROWN,
R. K. The Carcinogenic Activities of Certain Analogues of
2-Acetylaminofluorene
in the Rat. Cancer Research,
9:504—509, 1949.
15. Miu@un, E. C.,MILLER, J.A.,S@&pp,
R. W., and WESmi,
G. M. Studies on the Protein-Bound
Aminoazo Dyes
Formed in vivo from 4-Dimethylaminoazobenzene
and Its
.
J.
Biol.
Chem.,
176:585—544,1948.
Cancer, pp. 267—287.Washington, D. C., 1945.
22. SAUNDERS, K. H. The Aromatic Diazo Compounds and
Their Technical Applications, p. 6. New York: Longmans,
(;@n&Co., 1936.
23. Scmzas@NN, G., and PILLARSEY, R. Vber Aromatische
Fluorverbindungen, V. Mitteil. Die Difluorbenzole; mit
einem Anhange: Analytische
Erfahrungen.
Ber., 62:3035—
3043, 1929.
24. STAJIKEY,E. B. p-Dinitrobenzene.
In Blatt,A. H., editor.
OrganicSyntheses, Coil. Vol. II., pp. 225—227.
New York:
John Wiley & Sons, 1943.
25. Suoiuit&, K., HALTER,C. R., KENSLER, C. J., and [email protected],
C. P. Observations
on Rats Fed with Compounds Re
lated to Dimethylaminoazobenzene.
Cancer Research,
5:235—238. 1945.
26. WusoN, R. H., DEEDS, F., and Cox, A. J. The Toxicity
and Carcinogenic Activity of 2-Acetaminofluorene. Cancer
Research, 1:595-608, 1941.
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1949 American Association for Cancer Research.
The Carcinogenic Activities of Certain Halogen Derivatives of
4-Dimethylaminoazobenzene in the Rat
J. A. Miller, R. W. Sapp and E. C. Miller
Cancer Res 1949;9:652-660.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/9/11/652
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1949 American Association for Cancer Research.