-Proc. Nat. Acad. Sci. USA
Vol. 70, No. 8, pp XXXXXXXXXX, August 1973
Carcinogens are Mutagens: A Simple Test System Combining Live
Homogenates for Activation and Bacteria for Detection
(frameshift mutagens/aflatoxin
enzo(a)pyrene/acetylaminofluorene)
BRUCE N. AMES, WILLIAM E. DURSTON, EDITH YAMASAKI, AND FRANK D. LEE
Biochemistry Department, University of California, Berkeley, Calif. 94720
Contributed by Bruce N. Ames, May 14, 1973
ABSTRACT 18 Carcinogens, including aflatoxin Bi,
enzo(a)pyrene, acetylaminofluorene, benzidine, and di-
methylamino-trans-stilbene, are shown to be activated by
liver homogenates to form potent frameshift mutagens.
We believe that these carcinogens have in common a ring
system sufficiently planar for a stacking interaction with
DNA base pairs and a part of the molecule capable of being
metabolized to a reactive group: these structural features
are discussed in terms of the theory of frameshift muta-
genesis. We propose that these carcinogens, and many
others that are mutagens, cause cancer by somatic muta-
tion. A simple, inexpensive, and extremely sensitive test fo
detection of carcinogens as mutagens is described. It con-
sists of the use of a rat or human liver homogenate fo
carcinogen activation (thus supplying mammalian metab-
olism) and a set ofSalmonella histidine mutants for muta-
gen detection. The homogenate, bacteria, and a TPNH-
generating system are all incubated together on a petri
plate. With the most active compounds, as little as a few
nanograms can be detected.
We have previously described the use of a set of mutants of
Salmonella typhimurium for detecting and classifying chemical
mutagens with great simplicity and sensitivity (1, 2). With
this test we have also shown that the active forms of a large
number of known carcinogens are mutagens XXXXXXXXXXThe active
forms of carcinogens such as aflatoxin, polycyclic hydrocar-
ons, dimethylnitrosamine, and various aromatic amines
are formed by mammalian metabolism, in particular by the
TPNH-dependent microsomal enzymes of liver XXXXXXXXXXThe
principal limitation of any bacterial system for detecting
carcinogens as mutagens is that bacteria do not duplicate
mammalian metabolism in activating carcinogens. Mam-
malian-liver homogenates have been used by Garner et al.,
(6) to activate aflatoxin B1 to a compound lethal to our bac-
terial tester strain lacking excision repair, by Malling (12)
to activate dimethylnitrosamine to a compound that reverts
one of our bacterial tester strains, and by Slater et al. (13)
to activate dimethylnitrosamine to a compound lethal fo
acteria lacking polymerase I. In this study we have extended
this work and shown that carcinogens can be detected as mu-
tagens simply and with great sensitivity by incubation of the
carcinogen, a rat or human liver homogenate, and our bac-
terial tester strain together on a petri plate.
MATERIALS AND METHODS
Compounds. Glucose-6-phosphate, TPN, TPNH, and 2-
naphthylamine were obtained from Sigma. Benzo(a)pyrene,
2-acetylaminofluorene, and benzidine were from Aldrich. Di-
A
eviation: Me2SO, dimethylsulfoxide.
methylsulfoxide (Me2SO), spectrophotometric grade, was ob-
tained from Schwarz/Mann, sodium phenoba
ital from
Mallinckrodt, aflatoxin B1 from Calbiochem, and 3-methyl-
cholanthrene from Eastman; 7,12-dimethylbenz(a)anthracene
was a gift of P. L. Grover. Schuchardt (Munich) was the
source for the other carcinogens.
Bacterial Strains used are mutants of S. typhimurium LT-2
and have been discussed in detail (2).
Source of Liver. Male rats (Sprague-Dawley/Bio-1 strain,
Horton Animal Laboratories) were maintained on Purina
laboratory chow. A week before they were killed, thei
drinking water was made 0.1% in sodium phenoba
ital (14).
The rats XXXXXXXXXXg) were killed by a blow to the head and
cervical dislocation; the liver was removed and placed in a
sterile, ice-cold beaker. A portion of human liver was obtained
from an autopsy of a 77-year-old man who had died 7 h
earlier of heart failure.
Preparation of Liver Homogenate Fraction "S-9". We have
used the procedure of Garner et al. (6). All steps were per-
formed at 0-4o with cold and sterile solutions and glassware.
The liver (rat livers were 10-25 g each) was washed in an
equal volume of 0.15 M KCl, minced with sterile scissors
in three volumes of 0.15 M KCl (3 ml/g of wet liver), and
homogenized with a Potter-Elvehjem apparatus with a
Teflon pestle. The homogenate was centrifuged (Sorvall
RC2-B) for 10 min at 9000 X g, and the supernatant, which we
call the S-9 fraction, was decanted and saved. 1 ml of S-9
fraction contained microsomes from 250 mg of wet liver; the
protein concentrations were fairly constant from preparation
to preparation except for the human S-9 fraction which was
about half, perhaps due to difficulties in homogenization be-
cause of its fi
ous nature. The fresh S-9 fractions (rat and
human) were distributed in 2-ml portions in small plastic
tubes (2-ml liquid nitrogen storage tubes/4-Shore-USA, La
Jolla, Calif.), quickly frozen in dry ice, and stored at -80° in
a Revco freezer. As required, sufficient S-9 fraction was
thawed (at room temperature) and kept in ice; the unused
portion was discarded at the end of the day.
Mutagenesis Test wvith the S-9 Fraction. The method without
the liver activation system has been described in detail (2).
The only modification is the addition of S-9 Mix to the top
agar. The S-9 Mix contains per ml: 0.3 ml of S-9 fraction, 8
mM MgCl2, 33 mM KCl, 5 mM glucose-6-phosphate, 4 mM
TPN, and 100 mM sodium phosphate (pH 7.4). To 2 ml of
2281
Fatma Alqadi
Fatma Alqadi
Fatma Alqadi
Fatma Alqadi
Fatma Alqadi
Fatma Alqadi
Fatma Alqadi
Fatma Alqadi
Fatma Alqadi
Fatma Alqadi
2282 Genetics: Ames et al. Proc. Nat. Acad. Sci. USA XXXXXXXXXX)
Carcinogen
NH2 2-minoant
NH2 2-ainoflu4
o LY ITNHCOCs 2-acetylm
fluorene
H 2N H~~~N"2 benzidine
&.&NH2 4-aminobip
~ r 4-amino-t
C-C.. stilbene
T ~4-dimethyl@ Ad ~~~~trans-St
,0NH2 p-(phenyls
(.,NmN aniline
4- (o-tolyl
@[NsN CH3 toluidi
XaXN(CH3) N,N-dimet}
ft NdN p- (r- toI
aniline
"H2 2-naphthyl
JH2
o[Ino l-aminopyx
NH2
6-aminochi
enzo(a)p
3 3-methyl-cholantl
7.12-dimebenz (a).
,11"In, aflatoxin
OCH
Q1 sterigua
OCH3
118 S-9
thracene 20
20 -
0 +
orene 10
10 -
0 +
mino- 50
so
0 +
50
50 -
O +
phenyl 100
100 _
0 +
one- 10 +
10
0
lamino- 10
tilbene 10
0 +
azo)- 100 +
100 -
0O
,lazo)-o- 10 +
,no 10
0+
Ayl- 100 +
lylazo)- 100
I 0 +
lamine 100 +
100
0 +
ene 10 +
10 -
0. +
TABLE 1. Activation of carcinogens to mutagens
nj~zai~aewrwi per pzaHisi~fne Revertnte per powt
TA1S35 TA1536 TA1537 TA1S38
XXXXXXXXXX
T 27
XXXXXXXXXX
XXXXXXXXXX
21 0 iT 39
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
38 0 10
XXXXXXXXXX
XXXXXXXXXX
92 2 a
XXXXXXXXXX
23 842
7 17
10 42
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXXS3
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
94 1 8 T7
XXXXXXXXXX
330
16
41
66
42
2 34
0 8
2 18
1 136
0 23
1 9
ysene XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
)yrene 5 +' XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXXS 88
threne so XXXXXXXXXX
XXXXXXXXXX
ethyl XXXXXXXXXX
)anthracene XXXXXXXXXXis
XXXXXXXXXX
BI 1 +t XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
totystin* 0.1 .t XXXXXXXXXX
XXXXXXXXXX
XXXXXXXXXX
147
31
81
21
398
59
29
638
30
46
505
44
110
21
27
88
20
36
266
26
26
121
8
32
Revertant colonies (his +) on each plate were scored after 2 days.
Numbers underlined are judged to be significantly different from
the controls. Rat-liver homogenate (8-9 Mix) was added to the
plates, where indicated. Human S-9 fraction gave qualitatively
similar results when 2-aminoanthracene, the fluorene derivatives,
4-aminobiphenyl, 6-aminochrysene, and aflatoxin B1 were assayed
with TA1538. The amount chosen for each carcinogen was the one
that gave maximum mutagenesis of several amounts tried (see
Fig. 1). All of the compounds were added from solutions in Me2SO
usually 1 mg/ml; the control, lacking compound, had an equiva-
lent addition of Me2SO.
* The 8-9 preparation was from a 200-g rat that was injected
intraperitoneally, 24 hr before it was killed, with 16 mg of 3-meth-
ylcholanthrene dissolved in corn oil.
t One-third the normal amount of 8-9 fraction was used in the
9-9 Mix.
t We are indebted to Dennis Hsieh for a sample and for suggest-
ing we test sterigmatocystin.
molten top agar at 450 are added 0.1 ml of the bacterial teste
strain culture (2 to 3 X 109/ml), up to 0.1 ml of a solution
(Me2SO or water) of the compound to be tested, and 0.5 ml
of 8-9 Mix; then the tube is rotated quickly and the contents
are poured on the agar plate. The additions and pouring should
take less than a minute. The colonies on the plates (his+
evertants) are counted after a 2-day incubation at 37°.
RESULTS
Combined System for Carcinogen Activation and Bacterial
Mutagenesi8. We use the TPNH-dependent microsomal-en-
zyme systems in liver homogenate to activate carcinogens.
The activated carcinogens are detected as mutagens. The
test assays, on a petri plate, the number of revertant colonies
induced by mutagens in the set of histidine-requiring mutants.
Two technical improvements greatly simplify the use of the
combined bacterial and liver system. The TPNH-generating
system and liver homogenate traction (S-9 Mix) can be in-
cubated directly on the petri plate along with the compound
to be tested and the bacterial tester strain. Both rat and
human liver preparations can be frozen for many months
without loss of activity.
Carcinogens Detected as Mutagens. Table 1 shows that rat-
liver homogenates can activate 18 different aromatic type
carcinogens to mutagens. The compounds were chosen fo
testing because they were known to be carcinogenic in hu-
mans or in animals (7, 15). Control values are presented both
for the number of revertant colonies on plates with compound
and no S-9 Mix, and for the number of colonies on plates
A B
~15-N
*10
~~~~~~~~~~~
C XXXXXXXXXX
Liver Homogenate Per Assay (pil of S-9) Ad g of Compound
Fig. 1. Effect of 8-9 fraction and carcinogen concentration on
mutagenesis of TA1538. (A) The procedure was as described in
Methods except that the amount of S-9 fraction in the Mix was var-
ied. As the human 8-9 had half the protein of the rat S-9, the re-
sults with the human material are plotted at half the 8-9 amount
actually used. 4-Aminobiphenyl (100 Isg per assay) had about 0.1
the activity of