(CANCER RESEARCH 52. 5668-5673. October 15, 1992]
Sequence Specificity of Aflatoxin B,-induced Mutations in a Plasmid Replicated in
Xeroderma Pigmentosum and DNA Repair Proficient Human Cells
Dan D. Levy, John D. Groopman, Susan E. Lim, Michael M. Seidman, and Kenneth H. Kraemer1
Laboratory of Molecular Carcinogenesis [D. D. L., M. M. S., K. H. K.], National Cancer Institute, Bethesda, Maryland 20892; The Johns Hopkins University School
of Hygiene and Public Health [J. D. G., S. E. LJ, Baltimore, Maryland 21205; and Otsuka Pharmaceutical Company [M. M. S.J, Rockville, Maryland 20850
ABSTRACT
The mutagenic spectrum induced by aflatoxin-DNA lesions in DNA
repair deficient and repair proficient human cells was investigated. The
reactive metabolite aflatoxin B,-8,9-epoxide was synthesized and re
acted in vitro with the shuttle vector plasmid pS189. Plasmids were
transfected into human fibroblasts and allowed to replicate, and the
recovered plasmids were screened in indicator bacteria for plasmid sur
vival and mutations in the \ufil marker gene. Sequence data were ob
tained from 71 independently arising mutants recovered from DNA
repair deficient xeroderma pigmentosum (XP) cells |XP12BE(SV40)|
and 60 mutants recovered from a DNA repair proficient cell line
(GM0637). Plasmid survival was lower and mutation frequency higher
with the XP cells, and the mutation hotspots differed substantially for
the 2 cell lines. Most mutations (>90°/o)were base substitutions at G:C
pairs, only about one-half of which were ( ¡:(
' •¿'!
:A transversions, the
expected predominant mutation. One-third of the mutations at GG sites
and none of those at isolated Gs were G:C—A:T transitions. Tandem
base substitutions also occurred only at GG sites and were found only
with XP cells. The location of mutation hotspots with either cell line did
not correlate with the level of modification within the sequence as as
sessed by a DNA polymerase stop assay. These results suggest that the
DNA repair deficiency associated with XP can influence not only the
overall frequency of mutations but also the distribution of mutations
within a gene. The finding of transition mutations exclusively at GG
sites may be of predictive value in attempts to link dietary aflatoxin
exposure to cancers associated with specific mutations in the c-ras on
cogene and the p53 tumor suppressor gene.
INTRODUCTION
AFB)2 is the best studied of a series of natural products of
Aspergillus flavus, a ubiquitous mold commonly found as a
contaminant of food crops. AFB, has been shown to be carci
nogenic in many animal species (1), and consumption of food
contaminated with aflatoxin has been associated with a high
incidence of hepatocellular carcinoma in Asia and Africa (2, 3).
AFB, is lexicologically inert but when enzymatically oxidized
to the 8,9-epoxide, a rapid reaction occurs with guanine resi
dues in DNA.
Aflatoxin-DNA adducts are rapidly removed from normal
cells and tissues. In cell lines derived from patients with XP,
aflatoxin-DNA adducts are far more stable (4). XP is a rare
genetic disease characterized by a high incidence of skin cancers
in sun exposed areas of the body and a defect in the excision
repair of certain types of DNA damage (5, 6). Large "bulky"
cells. In contrast, strand breaks and oxidative damage induced
by agents such as ionizing radiation appear to be repaired
normally.
Synthesis of the reactive intermediate, AFB,-8,9-epoxide,
was recently achieved (8), thus allowing the in vitro preparation
and characterization of aflatoxin-modified DNA. We studied
the mutations induced by aflatoxin-DNA lesions in both the
presence and the absence of DNA repair during replication in
human cells. pS189 (9) is a shuttle vector carrying both mam
malian and bacterial origins of replication. Mutations induced
in the marker supF gene during replication in human cells do
not influence replication in either the human cells or bacteria
but permit screening in host bacteria unable to metabolize
X-gal because of a suppressible mutation in the lac gene. We
found reduced plasmid survival and increased mutagenesis with
the XP cell relative to the repair proficient cells. Most of the
mutations in both cell lines were at G:C base pairs, although
there were significant differences in the classes of mutant plas
mids and the locations of the mutations. A DNA polymerase
stop assay was used to evaluate the relative amount of adduci
formation within the supF marker gene. There was no associa
tion between the DNA adduci level and ihe mutation frequency
at each G:C pair wilhin ihe sequence in either cell line.
MATERIALS
AND METHODS
Cells and Cell Culture. SV40 transformed DNA repair deficient
human fibroblasts (XP12BE(SV40), complementation
group A,
GM4429] and DNA repair proficient fibroblasts (GM0637) were ob
tained from the Human Genetic Mutant Cell Repository (Camden, NJ)
and grown in Dulbecco's modified Eagle's medium supplemented with
10% fetal bovine serum and glutamine in an 8% CC>2humidified incu
bator as described previously (10).
Plasmid Modification and Characterization. Shuttle vector pS189
(9), purified by CsCl gradient centrifugation, was dissolved in water
(1 mg/ml) and modified by reaction with 1-2 ^1 AFB|-8,9-epoxide at
various concentrations in acetone. The AFB|-DNA was dissolved in
water and aliquoted before freezing on dry ice and storage at —¿35°C.
To
characterize the adducts formed, a portion of each sample was hydrolyzed and analyzed (11) for specific adducts by reverse phase chromatography using a CIS ODS-Ultrasphere column (Rainin Instrument
Co., Woburn, MA). Chromatography was performed at 35°Cusing a
Beckman model 344 liquid Chromatograph coupled to a Hewlett-Pack
ard 1040 diode array detector using a 25-min gradient of 6-15%
ethanol/20 HIMtriethylammonium formate (pH 3.0) at 1.0 ml/min. The
retention times of the AFB,-DNA adducts were determined using au
DNA lesions such as UV induced dipyrimidine dimers (5, 6) thentic standards (11). Formation of sites at which adduction led to the
loss of the nucleoside moiety (apurinic sites) was assessed by quantitatand aflatoxin (4, 7) are repaired poorly or not at all by XP
ing AFB]-N7G in the supernatants of the reaction mixtures. This treat
ment resulted in the generation of 5.4 to 24 AFB^N'G adducts/
plasmid with no detectable (<2%) ring-opened AFB|-Fapyr adducts. To
Received 4/17/92; accepted 7/29/92.
ensure that minor adducts were not a significant factor in mutagenesis
The costs of publication of this article were defrayed in part by the payment of
experiments, plasmid was modified with a higher concentration of eppage charges. This article must therefore be hereby marked advertisement in accord
oxide, leading to 146 lesions/plasmid. In this experiment, 1.3% of the
ance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 To whom requests for reprints should be addressed.
lesions were AFB,-Fapyr adducts. In addition, the supernatant of the
2 The abréviationsused are: AFB,. aflatoxin B(; AFBi-Fapyr, 8,9-dihydro-8reaction mixture was analyzed and a small amount of AFB|-N7-G
(jV5-formyl-2,5.6-triamino-4-oxopyrimidine-Ar5-yl)-9-hydroxy-aflatoxin B,; AFBiN'G, 8.9-dihydro-8-(.V7-guanyl)-9-hydroxy aflatoxin B,: XP, xeroderma pigmen
detected, indicating that a maximum of 1.5 apurinic sites/plasmid (1%
of the adduci sites) may have been generated.
tosum.
5668
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
SEQUENCE
SPECIFICITY
OF AFLATOXIN
Mutagenesis Assay. Plasmid with or without AFB, modincation
was transfected into fibroblasts using CaP04 as described previously
(10). Briefly, 2 fig of plasmid were transfected into IO6cells in a 60-mm
dish. After 6 h, the medium was replaced and the cells allowed to grow
an additional 40 h. The plasmids were then recovered using an alkaline
lysis procedure, unreplicated plasmid removed by digestion with Dpnl,
and the plasmids used to transform MB7070 indicator bacteria. The
transformed bacteria were diluted and plated on agar plates containing
ampicillin, isopropyl thiogalactoside, and \-gal. Wild type (blue) and
mutant (white and light blue) colonies were counted to determine the
plasmid survival and mutation frequency. Mutant plasmids were cloned
and sequenced using Sequenase and the protocols of the supplier (U. S.
Biochemicals). Mutations detected with this methodology are represen
tative of all mutations formed during replication of the modified plas
mids in the human cells since the screening assay is sensitive to muta
tions at virtually every base pair in the SupF tRNA marker gene,
including at least 48 of the 50 G:C pairs (9). Statistical analysis was by
Fisher's exact test.
B, MUTAGENESIS
100'=.
o
n
Ë
Adduct Location. The intensity of aflatoxin modification of specific
sites in the region between base pairs 70 and 190 in pS189 was evalu
ated using a polymerase stop assay (12, 13). Aflatoxin modified or
control plasmid (0.2 Mg/ml) was treated with 0.2 NNaOH at 37°Cfor 10
min to denature the DNA for sequencing. High performance liquid
Chromatographie analysis revealed that >90% of the adducts were con
verted to AFB|-Fapyr adducts with no detectable (<0.2%) depurination. A 4-fold Mexcess of primer labeled at the 5' end (6000 Ci/mmol
[7-12P]ATP) was then used to sequence 2 ng.of plasmid DNA. Primers
annealing to the coding and noncoding strands were used in separate
reactions. The T7 polymerase (Sequenase version 2), blocked by the
AFB,-Fapyr lesion, stopped 1 base 5' to each band in the adjacent ddC
sequencing lane as has been described previously (14). Additional stops
directly opposite the adducts were also observed, also as shown previ
ously for aflatoxin adducts (15, 16). Use of the Klenow fragment of
DNA polymerase I at 37°Cand Taq polymerase at 75°Cproduced
similar results (data not shown). Stop bands were precisely located by
performing dideoxy-DNA sequencing on adducted DNA (15), while
quantitation of the stops was performed by sequencing samples of the
same modified DNA without dideoxynucleotides. Autoradiograms
were analyzed on a Molecular Dynamics densitometer using the man
ufacturer's ImageQuant software. The intensity of replication blockage
at each position at a given dose was compared with the corresponding
intensity at other doses. The correlation coefficient, r, for these pairwise
comparisons was >0.97, indicating excellent correlation. Inclusion of
a fourth independently sequenced sample (at the highest aflatoxin
dose) was averaged after pairwise comparisons showed good correlation
(r > 0.85).
RESULTS
Shuttle vector plasmid pS189 was modified in vitro by reac
tion with AFB|-8,9-epoxide. High performance liquid Chro
matographie analysis of acid hydrolysates of the modified plas
mids revealed the presence of up to 24 AFB,-N7-G adducts/
plasmid ( 15 nmol/mg DNA). Plasmids with various amounts of
modification were then passaged through DNA repair deficient
XP fibroblasts or DNA repair proficient human fibroblasts,
allowed to replicate, and then purified and used to transform
Lesions/Plasmid
Fig. I. Plasmid survival. Relative number of bacterial colonies counted after
transfection of AFBi-epoxide modified pSI89 into XP (•)or DNA repair profi
cient (A) fibroblasts. Following replication, progeny plasmids from each dish were
used (o transform indicator bacteria, and dilutions of the transformed bacteria
were plated as described in "Materials and Methods."
indicator bacteria to screen for mutations in the marker supF
gene. Mutant colonies were then expanded and the mutations
determined by dideoxy DNA sequencing.
The sensitivity of the XP cells to this carcinogen is shown
by the 4-8-fold decrease in plasmid survival at each dose com
pared to the DNA repair proficient cell line (Fig. 1). The plas
mid mutation frequency increased with increasing adduct load
(Table 1) with 7-40-fold increases compared to unmodified
control plasmids. The hypermutability of XP cells is shown by
the 2-3-fold greater mutation frequency at each dose relative to
the repair proficient cells.
Mutations were identified by DNA sequencing (Table 2).
There were 71 independent mutant plasmids carrying 87 point
mutations recovered from the XP cells. Most of the mutant
plasmids (76%) contained a single point mutation in the supF
marker gene, while a few had multiple mutations (2 or 3 sub
stitutions located within this region >4 bases apart from one
another). Surprisingly, 15% had tandem mutations (2 or 3 point
substitutions <4 bases apart). The DNA repair proficient cells
yielded 60 independent mutant plasmids. There were no plas
mids with tandem base substitutions from these cells. The
higher frequency of plasmids with multiple mutations found in
Table I Mutation frequency of AFB,-treated plasmid replicated in xeroderma pigmentosum or normal fibroblasts
Mutation frequency in plasmids described in Fig. I.
XP12BE
GM0637
coloniesMutant4IK
of
No. of AFB]
adducts/plasmid0
coloniesMutant5
of
AFB,adducts/plasmid01324No.
of
frequency
IO40.67
x
6
1622No.
17Total60,000
13.400
10.100
6,430Mutation
6.0
18
26No.
25
80.300
28
65.300
28Total11,800 33,600Mutation
frequency
IO40.42
x
3.0
4.3
8.3
5669
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
SEQUENCE
SPECIFICITY
OF AFLATOX1N B, MUTAGENESIS
Table 2 Classes of mutations in AFB¡-treatedplasmid replicated in xeroderma
pigmentosum or normal fibroblasto
No. of plasmids with mutations" (%)
Base
substitutionsSingle
substitutionsTandem
base
substitutionsMultiple
base
substitutionsOther
base
mutationsrIndependent
XP12BE
GM0637
(76)11
(15)*4
(6)2
(3)71
(80)0
(0)8
(13)4
(7)60(100)
plasmids sequenced54
(100)48
" Mutant plasmids from experiments described in Fig. 1 and Table 1.
bP< 0.008 versus GM0637.
c Other mutations: XP12BE, a single base insertion and a single base deletion;
GM0637. 2 rearrangements and 2 single base substitutions near single base frame
shifts (see also legend to Fig. 2).
Table 3 Types of mutations in AFB ¡-treatedplasmid replicated in xeroderma
pigmentosum or normal fibroblasts
No. of base changes" (%)
XP12BE
GM0637
Base substitutions
G:C—T:A transversions
G:C—C:G transversions
G:C—A:T transitions
substitutions*Frameshifts
A:T
(55)
17 (20)
16 (18)
4(5)1
(68)
10 (14)
9 (13)
2(3)1
1 base deletion
insertionTotal48
1 base
(D
(D87(100)48
1
(1)71
1
(D
(100)
" Single, tandem, and multiple mutations found in the plasmids described in
Table 2.
*A:T substitutions: XP12BE. 3 A:T—T:A and 1 A:T—G:C; GM0637, 1
A:T—T:A and 1 A:T—C:G.
the DNA repair proficient cells (13%) has been reported previ
ously for UV-irradiated plasmids (17).
Virtually all (>90%) of the mutations were base substitutions
occurring at G:C base pairs (Table 3). The predominant muta
tion, occurring in slightly over half of all cases, was the G:C—
T:A transversion. Most of the other mutations were evenly
divided between G:C—C:G transversions and G:C—A:T tran
sitions. Frame shift mutations were rare, occurring in roughly
SUPPRESSOR
2% of the mutations, and half of those were associated with a
nearby base substitution. The occurrence of each type of muta
tion was similar in the 2 cell lines. The relative proportions of
G:C^T:A and other types of mutation did not vary among the
different AFB,-epoxide doses for either cell line (data not
shown).
The types of mutations were not evenly distributed in the
supF gene (Fig. 2). G:C—A:T transitions never occurred at
isolated Gs, but were all located within runs of Gs (G:C—A:T
transitions constituted 0 of 43 mutations at isolated Gs and 18
of 54 mutations at Gs in runs, P < 10~5). The frequency of
G:C^A:T transitions was 3 times as high in the plasmids with
tandem mutations as in the plasmids containing single base
substitutions (7 of 23 tandem mutations were G:C—-A:Ttran
sitions versus 6 of 54 single base substitions, P = 0.04), pre
sumably related to the finding of tandem substitutions only at
adjacent Gs.
Inspection of the locations of the mutations (Fig. 2) reveals
several interesting features. With the XP cells, there were 7
hotspots (loci with >5% of the total mutations). The most
prominent was at base pair 123, at which there was a remark
able diversity of changes: roughly equal numbers of transition
and transversion mutations. Of the other hotspots, loci 109,
113, 139, 160, and 168 all had more than one type of base
change while only at base pair 159 were all of the changes
identical. The tandem base substitutions were also not ran
domly distributed; each occurred at a G:C site with at least one
other tandem base substitution and most (7 of 11) occurred at
base pairs 108-110 or 175-176.
The mutation spectrum from repair proficient cells was quite
different from that with the XP cells, despite the overall simi
larity of the types of mutations. With the repair proficient cells,
the hotspots were at base pairs 127, 144, 164, 172, and 123.
The sites with the most mutations from plasmids replicated in
the repair proficient cells, base pairs 164 and 127, were not
frequently mutated in the XP12BE spectrum. Conversely, while
position 123 had twice as many mutations as any other hotspot
with the XP cells, it had only the fifth greatest number of
mutations in the spectrum from the repair proficient cells.
IRNA (99-183)
IS'
CODING STRAND
I
I
I
I
I
I
I
I
I
I
90
100
110
120
130
140
150
160
170
180
I
I
I
I
I
I
I
I
I
ATTACCTGTCKTrOGGCrrCCCGAOCGGCCAAAOGGAOCAOACTCTAAATCTGCCGTCATCGACTTCGAAGGTTCGAATCCTTCCCCCACCACCATC
TAT
AAT
A
TTrr
AAA0GI1ATAATCCTTTCAT
G
ocAAT
TAT
AG A
AAT1
TAG
TTTTT1
IXPI2BE
CAOI1A
FIBROBLASTS1
TTAG
1C
CTGMO637
1AT
AAT
TTAACAAACTTTTT11TCTTAA11A
CAAAAG
3'
11
11
11
I
AA
AA
1AA
ACATA A
T-ATTTTTTCAA
AAA
FIBROBLASTSU
Fig. 2. AFB, mutation spectrum. Top, mutations generated in XP12BE cells following replication of pS189 with 5.4, 16, and 22 aflatoxin adducts/plasmid. Only
single and tandem base substitutions are shown. Possible sibling mutant plasmids, those containing mutations identical to other plasmids arising from the same plate
of cells, are not shown. Other mutations: 70 G—T. 144 G—C; 118 C—A, 176 C—A; 133 C—G, 185 C—T; 133 C—G, 155 C—T; 172—5delete C; 175 or 176 insert
A; 13 possible sibling mutant plasmids. Bottom, mutations generated in DNA repair proficient GM0637 cells following replication of pS189 with 7.9, 13, and 24
adducts/plasmid. Only single base substitutions are shown. Other mutations: 62 C—T, 133 C—A:95 C—A,97 G—A, 109 C—A; 104 G—C, 123 G—T; 124 G—A,
129 G—C; 133 C—G, 141 G—A; 133 C—A, 142 C—T, 169 C—A;155 C—T, 163 C—G, 185 C—T; 158 delete A, 159 G~T; 169 C—A, 182 C—A; 174 C—A, 176
or 177 insert A; 2 rearranged plasmids with no supF and 13 possible sibling mutant plasmids.
5670
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
SEQUENCE SPECIFICITY OF AFLATOX1N B, MUTAGENESIS
Plasmids from the repair proficient cell had very few mutations
of any type at the sites of the tandem mutations in the XP
spectrum, suggesting that the tandem mutations were generated
by a different mechanism than the single base substitutions.
The types of mutations at the hotspots in the plasmids also
differed with the 2 cell lines. Only 2 of 5 hotspots had mutations
other than G:C—-T:A substitutions with the repair proficient
cells, whereas the other types of base substitution mutations
were common in 6 fo 7 plasmid hotspots from the XP cells. To
evaluate the diversity of changes at hotspots, the ratio of
G:C^T:A substitutions to the total at that site was computed.
For the XP cells, the result (28 of 43) was not significantly
different from the ratio for mutations at all locations, but for
the repair proficient cells 19 of 21 of the substitutions at
hotspots were G:C—>T:A substitutions, significantly higher
than for the overall spectrum in that cell line or for the XP cells
(P < 0.03). Thus, while the overall G:C^T:A substitution rate
was not significantly different for the 2 cell types, this may
mask the differences in the processing of lesions at individual
sites.
To determine whether the sites of mutations in either spec
trum correlated with the location of aflatoxin adducts, a polymerase termination assay was performed. By using primers for
each strand labeled with high specific activity, information for
every G in the sequence was obtained, showing marked differ
ences in the extent to which AFB, lesions blocked the polymerase at various Gs (Fig. 3). The relative intensity of blockage
at each site along the sequence did not vary in plasmids with
different amounts of modification. Runs of 2 or more Gs ap
peared to be hotspots for lesions, as has been reported earlier
using both a polymerase stop assay (14, 16) and assays using
strand breakage at the lesion site (18, 19). Only 18% of the
blockage occurred at isolated Gs on the coding strand, although
they represented 38% of the Gs in the sequence and 34% of the
mutations. Thus, isolated Gs showed disproportionately low
levels of polymerase blockage, but their contributions to muta-
lions were similar to their relative abundance in the sequence.
This suggests that, overall, there was little correlation between
the intensity of AFB, adduction and the mutation frequency at
different Gs in the gene.
Empirical rules have previously been derived to predict afla
toxin binding to a G based on its nearest neighbors ( 18,20). The
G in the sequence CGA was predicted to be of relatively low
reactivity. This sequence appears 6 times in supF as part of the
palindrome TCGA (the Gs are at positions 149, 155, and 163 of
the coding strand and 150, 156, and 164 of the noncoding
strand) as well as at position 111. Mutations can be detected at
all of the G:C pairs in these sequences (9), and there was no
overall strand bias to AFB,-induced mutations. Consistent with
predictions of AFB, binding, the level of polymerase blockage
at each of the 7 CGA triads in the pS189 sequence was similar
within each strand and modest relative to blockage at other
sequences. However, the level of mutations in these identical
sequences was quite variable. Positions 156 and 164 were
hotspots, with 7 and 5 mutations in the 2 cell lines, and there
were 3 mutations at position 155 and none at 149, 150, 163, or
111.
Examining levels of AFB,-induced polymerase blockage and
mutation at individual sites reinforced the conclusion that there
was no correlation between the degree of blockage and the
mutation frequency for either cell line. For example, the Gs at
base pairs 116 and 123 were among the most intense blockage
sites on the noncoding strand. While there were many muta
tions at position 123 in both cell lines, only 1, in the repair
proficient cells, was detected at position 116. Conversely, po
sitions 127, 139, and 163 were among the least modified on the
coding strand. Position 127 was a hotspot for the repair profi
cient cells and position 139 for the XP12BE cells, but there
were no mutations in either spectrum at position 163 (although
mutations can be detected there) (9).
DISCUSSION
We have examined the pattern of mutagenesis induced in an
AFB| modified shuttle vector during replication in repair defi
cient and proficient human fibroblasts. Plasmid survival was
lower and mutation frequency higher with the XP cells. These
findings are consistent with previous reports of the sensitivity
of XP cells to AFB, (4, 7). The differences between the 2 cell
types were not limited to an increase in mutation frequency.
While the types of mutations were similar for AFB,-modified
plasmids replicated in the 2 types of cells, there were differ
ences in both the classes and locations of the mutations,
suggesting that repair or other metabolic differences can mod
ulate the mutability of a given base within a gene. These differ
ences may have implications for the likelihood of oncogenic
transformation.
The most common mutation induced by aflatoxin was the
G:C—«T:A
transversion base substitution. This is consistent
with the "A-rule" (12) that noninstructional lesions most fre
quently result in the insertion of an A during DNA replication.
However, almost half of the mutations were base substitutions
that did not follow this prediction. In this respect, aflatoxin is
unlike UV (10) and cisplatin (21), for which the A-rule accu
rately predicts >80% of the mutations. Aflatoxin is similar to
Fig. 3. AFB|-induced polymerase slops. All sites at which T7 DNA polymerase
other polycyclic aromatic carcinogens that have been studied
was blocked by \l H, lesions on the coding strand (top) and noncoding strand
using this shuttle vector system in a DNA repair proficient
(bottom) in the supFgene in pS189. Bars, SE of data averaged from 3 samples with
human embryonic cell line (Ad293) by reacting the synthesized
8. 13, and 24 adducts/plasmid plus an independent experiment with plasmid at
epoxide with the plasmid (13, 22-25). In each case, the vast
the highest adduci dose.
5671
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
SEQUENCE
SPECIFICITY
OF AFLATOX1N B, MUTAGENESIS
majority of mutations were single base substitutions at G:C—•fluence the likelihood that mutation will occur at a specific base
base pairs. The fraction of G:C—<T:Atransversions was similar
pair. This may help explain the observation that hepatocellular
for AFB, (55 and 67%), 5-methylchrysene (57%) (22), benzocarcinoma is the only tumor type elevated in populations ex
[ajpyrene (63%) (23), 1-nitropyrene (61%), 1,6-dinitropyrene
posed to dietary aflatoxins (2), despite the ability of a number of
(64%) (24), aminofluorene (65%), acetylaminofluorene (65%) cytochromes P450, present in a variety of tissues, to activate
(13), and methylbenzathracene (56%) (25). The only exceptions
aflatoxin (33). In contrast, the degree of damage at a site did not
were 4 stereoisomers of benzo[c]phenanthrene (26), for which,
correlate with the appearance of mutation hotspots, in agree
consistent with substantial adduction of Ade, 24-68% of the ment with several other studies in which the correlation be
tween adduct frequency and mutation at a site was found to be
mutations were base substitutions at A:T pairs. Even for those
compounds, the A-rule correctly predicts 59% of the mutations.
poor for DNA damage following treatment with UV (34),
Our results are significantly different from 2 prior reports of acetylaminofluorene (35), benzo[a]pyrene diolepoxide (23), or
aflatoxin mutagenesis in bacteria. In one, 89% of mutations in 1-nitro- or 1,6-dinitropyrene epoxides (24).
the lad gene were G:C—T:A transversions (27). That study,
In analyzing these data, we have excluded plasmids with mul
however, was restricted to the 70 loci that could be identified by tiple (nontandem) mutations since they appeared to be gener
screening using rescue of amber and ochre nonsense mutations.
ated by different processes than the single and tandem muta
More recent work, using DNA sequence data from the lacZ
tions. The location of the mutations in plasmids with multiple
mutations did not coincide with the hotspots of single base
gene carried in bacteriophage M13 (28), yielded results signif
substitutions, and while the overall frequency of each type of
icantly different from both the earlier study and the results
reported here. The frequency of G:C—T:A transversions was substitution was similar, G:C—A:T transition mutations were
not confined to GG sites as they were in plasmids with a single
similar to ours (57% of all base substitutions), but significantly
fewer (6%) G:C—-C:G transversion and significantly more
or tandem mutation. The multiple mutations were not likely to
(39%) G:C^A:T transition substitutions were detected. SOS
have been caused by multiple lesions on a single plasmid since
they did not increase with increasing lesion dose, and several
induction of the bacteria caused minor changes in the results.
Tandem base substitutions were rare in both normal and SOShad mutations that would have required lesions on opposite
strands. These findings are consistent with similar results with
induced bacteria.
One of the most striking differences between the aflatoxinUV-irradiated plasmids in the same repair proficient cell line,
induced mutations in AFB,-treated plasmids replicated in the 2 perhaps due to the action of an error-prone polymerase during
cell lines was the large number of tandem base substitutions at excision-repair of the DNA (17).
The sequence specificity of mutation type we found at GG
GG sites in the XP cells and their complete absence in the
repair proficient cells. This class of mutation is common for sites in AFB i-modified pS189 appears to be reflected in afla
lesions that span more than one base, such as UV-induced
toxin carcinogenesis. Studies of AFB, hepatocarcinogenesis in
rat (36) and trout (37) showed that oncogenic mutations at GG
lesions (10) and cisplatin adducts (21), but rare for polycyclic
sites in codons 12 and 13 of ras gene were primarily G:C—»T:A
aromatic carcinogens (<6%) (13, 22-26), at least in repair pro
(10 of 12) with an occasional (2 of 12) G:C—A:T and no
ficient cells. The presence of tandem base substitutions exclu
sively in the XP cells suggests that their generation may have G:C—C:G substitutions. Mutated p53 tumor suppressor genes
from hepatocellular carcinomas from humans exposed to di
been related to the repair defect in these cells. The mechanism
etary aflatoxins (38, 39) also are in agreement with our data. In
for their generation is unclear, however, the finding of adduci
these tumors, 10 of 13 p53 mutations were G:C—T:A substi
bypass during replication via misinsertion opposite an unmod
ified base 5' to the adduci has been reported for methylbenzatutions at the final base of codon 249 (AGG) and 1 was a
G:C-*C:G transversion at the same base pair. A G:C—A:T
thracene-G adducts (29). Furthermore, the polymerase stop as
transition at that p53 mutation hotspot would result in inser
say clearly indicated a sequence related variability of the degree
to which an aflatoxin adduci could block replication at the base tion of the wild-type amino acid and thus would not contribute
5' to the adduct, consistent with the finding of these tandem
to oncogenic transformation. In summary, in this nonselective
plasmid system we found that roughly two-thirds of the muta
mutations only at adjacent Gs. It is unlikely that these muta
tions induced by AFB, modification of pS189 were G:C—>T:A
tions were induced by 2 adducts on adjacent bases since there
transversions and that the remaining mutations were predom
was no dose related increase in the frequency of tandem substi
inantly G:C^C:G transversions at isolated Gs or G:C—>A:T
tutions, nor were there any dose related changes in the poly
merase stop patterns. Finally, while adjacent Gs appear to be transitions at adjacent Gs. This sequence specificity appears to
hotspots for adduct formation (18), poly d(G-C) is a much
be consistent with the data emerging from studies of mutated
better substrate for aflatoxin binding than poly(dG)-poly(dC)
oncogenes sequenced from tumors associated with aflatoxin
carcinogenesis.
(30), suggesting that such vicinal adducts are rare.
The differences in mutant classes and mutation hotspots in
these cell lines are similar to those previously reported for UV- ACKNOWLEDGMENTS
damaged plasmids in the same cells (10), suggesting that the
The authors wish to thank Dr. H. V. Gelboin for his insightful
DNA repair defect is responsible. The relative intensity of
organization, encouragement, and support of this project.
hotspots has also been shown to be modified when UV-damaged plasmids were replicated in virally transformed lymphoblasts and fibroblasts from the same individual (31), or even by REFERENCES
modulating intracellular calcium levels in the same cell line3
1. Busby, W. F.. and Wogan, G. N. Aflatoxins. In: C. E. Searle (ed.), Chemical
Carcinogens, pp. 945-1136. Washington. DC: American Chemical Society,
(32). Thus, naturally occurring or virally induced (e.g., hepatitis
1984.
B infection) variations in physiology and metabolism may in2. Hollstein. M., Sidransky, D., Vogelstein, B., and Harris, C.C. p53 mutations
in human cancers. Science (Washington DC), 253: 49-53, 1991.
3. Ozturk, M., eÃ-al. p53 mutation in hepatocellular carcinoma after aflatoxin
exposure. Lancet. 338: 1356-1359, 1991.
3S. Seetharam and M. M. Seidman. unpublished observations.
5672
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
SEQUENCE
SPECIFICITY
OF AFLATOXIN
4. I radon. S. A., Tyrrell, R. M., and Cerimi, P. A. Excision repair of aflatoxin
Bl-DNA adducts in human fibroblasts. Cancer Res., 41: 5125-5129. 1981.
5. Cleaver, J., and Kraemer, K. H. Xeroderma pigmemosum. In: C. R. Scriver,
A. L. Beaudet, W. S. Sly, and D. Valle (eds.). The Metabolic Basis of Inher
ited Disease, Ed. 6, pp. 2949-2971. New York: McGraw Hill, 1989.
6. Kraemer, K. H., Lee, M. M., and Scotto, J. Xeroderma pigmentosum: cuta
neous, ocular and neurologic abnormalities in 830 published cases. Arch.
Dermatol., 123: 241-250, 1987.
7. Mahoney, E. M., Ball J. C, Swenson. D. H., Richmond, D., Mäher,V. M.,
and McCormick, J. J. Cytotoxicity and mutagenicity of aflatoxin dichloride
in normal and repair deficient diploid human fibroblasts. Chem.-Biol. Inter
act., 50: 59-76, 1984.
8. Baertchi, S. W., Raney, K. D., Stone. M. P.. and Harris, T. M. Preparation
of the 8,9-epoxide of the mycotoxin aflatoxin Bl: the ultimate carcinogenic
species. J. Am. Chem. Soc., 110: 7929-7931, 1988.
9. Kraemer, K. H., and Seidman, M. M. Use of supF, an Escherichia coli
tyrosine supressor tRNA gene, as a mutagenic target in shuttle-vector plasmids. Mutât.Res., 220: 61-72, 1989.
10. Bredberg, A., Kraemer, K. H., and Seidman. M. M. Restricted ultraviolet
mutational spectrum in a shuttle vector propagated in xeroderma pigmento
sum cells. Proc. Nati. Acad. Sci. USA, 83: 8273-8277, 1986.
11. Groopman, J. D., Croy, R. G., and Wogan, G. N. In vitro reactions of
aflatoxin Bl-adducted DNA. Proc. Nati. Acad. Sci. USA, 78: 5445-5449,
1981.
12. Sagher, D., and Strauss, B. Insertion of nucleotides opposite apurinic/
apyrimidinic sites in deoxyribonucleic acid during in vitro synthesis: unique
ness of adenine nucleotides. Biochemistry, 22: 4518-4526, 1983.
13. Mah, M. C-M., Boldt. J., Culp, S. J.. Mäher,V. M., and McCormick, J. J.
Replication of acetylaminofluorene-adducted plasmids in human cells: spec
trum of base substitutions and evidence of excision repair. Proc. Nati. Acad.
Sci. USA, 88: 10193-10197, 1991.
14. Jabobsen, J. S., Refolo. L. M., Conley, M. P., Sambamurti. K.. and Humayun, M. Z. DNA replication blocking properties of adducts formed by anatoxin Bl-2,3-dichloride and aflatoxin Bl-2,3-oxide. Mutât.Res., 179: 89101, 1987.
15. Puisieux, A., Lim, S., Groopman, J.. and Ozturk, M. Selective targeting of
p53 gene mutational hotspots in human cancers by etiologically defined
carcinogens. Cancer Res., 51: 6185-6189. 1991.
16. Marien, K., Mathews, K.. van Holde, K., and Bailey. G. Replication blocks
and sequence interaction specificities in the codon 12 region of the c-Ha-ras
proto-oncogene induced by four carcinogens in vitro J. Biol. Chem., 264:
13226-13232, 1989.
17. Seidman, M. M.. Bredberg, A., Seetharam, S., and Kraemer. K. H. Multiple
point mutations in a shuttle vector propagated in human cells: evidence for an
error-prone polymerase activity. Proc. Nati. Acad. Sci. USA, 84:4944-4948,
1987.
18. Benasutti, M.. Ejadi, S.. Whitlow, M. D.. and Loechler, E. L. Mapping the
binding site of aflatoxin Bl in DNA: systematic analysis of the reactivity of
aflatoxin Bl with guanines in different DNA sequences. Biochemistry, 27:
472-481, 1988.
19. Marien, K., Movers, R., Loveland, P., Van Holde, K., and Bailey, G. Repli
cation blocks and sequence interaction specificities in the codon 12 region of
the c-Ha-rai proto-oncogene induced by four carcinogens in vitro. J. Biol.
Chem., 261: 7455-7462, 1987.
20. Muench, K. F., Misra, R. P., and Humayun, M. Z. Sequence specificity in
aflatoxin Bl-DNA interactions. Proc. Nati. Acad. Sci. USA, 80: 6-10, 1983.
21. Bubley, G. J., Ashburner. B. P., and Teicher, B. A. Spectrum of c/s-diamminedichlorideplatinum(II)-induced
mutations in a shuttle vector propagated
in human cells. Mol. Carcinog., 4: 397-406, 1991.
B, MUTAGENESIS
22. Bigger, C. A. H.. Flickinger, D. J., Strandberg. J.. Pataki, J., Harvey. R. G.,
and Dipple, A. Mutational specificity of the an//-l,2-dihydrodiol 3.4-epoxide
of 5-methylchrysene. Carcinogenesis (Lond.), 12: 2263-2265, 1990.
23. Yang, J-L., Mäher,V. M., and McCormick, J. J. Kinds of mutations formed
when a shuttle vector containing adducts of (+/-)-7/j,8a-dihydroxy-9a,10aepoxy-7,8,9,10-tetrahydrobenzo[a|pyrene
replicates in human cells. Proc.
Nati. Acad. Sci. USA, 84: 3787-3791, 1987.
24. Boldt. J., Mah, M. C., Wang, Y. C., Smith, B. A., Beland, F. A., Mäher.V.
M., and McCormick J. J. Kinds of mutations found when a shuttle vector
containing adducts of 1,6-dinitropyrene replicates in human cells. Carcino
genesis (Lond.), 12: 119-126, 1991.
25. Bigger, C. A. H., Flickinger, D. J.. St. John. J.. Harvey, R. A., and Dipple, A.
Preferential mutagenesis at G.C base pairs by the an(i-3.4-dihydrodiol-l,2epoxide of 7-methylbenz[a|anthracene. Mol. Carcinog., 4: 174-179, 1991.
26. Bigger, C. A. H., St. John, J., Yagi, H., Jerina, D. M., and Dipple, A.
Mutagenic specificities of four stereoisomeric benzo[c]phenanthrene dihidrodiol epoxides. Proc. Nati. Acad. Sci. USA, 89: 368-372, 1992.
27. Foster, P. L., Eisenstadt, E., and Miller, J. H. Base substitution mutations
induced by metobolically activated aflatoxin Bl. Proc. Nati. Acad. Sci. USA,
80: 2695-2698. 1983.
28. Sambamurti, K.. Callahan, J., Luo. X.. Perkins, C. P.. Jacobsen, J. S., and
Humayun, M. Z. Mechanisms of mutagenesis by a bulky DNA lesion at the
guanine N7 position. Genetics. 120: 863-873. 1988.
29. Hruszkewycz, A. M., and Dipple, A. Bypass of a hydrocarbon adduci in an
oligonucleotide template mediated by mispairing adjacent to the adduci.
Carcinogenesis (Lond.), 12: 2185-2187, 1991.
30. Yu, F-L., Bender. W.. and Gerónimo, I. H. Base and sequence specificities of
aflatoxin Bl binding to single- and double-stranded DNAs. Carcinogenesis
(Lond.), //: 475-478. 1990.
31. Seetharam. S., Kraemer, K. H., Waters. H. L., and Seidman, M. M. Muta
tional hotspot variability in an ultraviolet treated shuttle vector plasmid
propagated in xeroderma pigmentosum and normal human lymphoblasts and
fibroblasts. J. Mol. Biol., 212: 433-436, 1990.
32. Seetharam. S., and Seidman, M. M. Modulation of an ultraviolet mutational
hotspot in a shuttle vector Xeroderma cells. Nucleic Acids Res., 19: 16011604, 1991.
33. Aoyama, T., Yamano, S., Guzelian, P. S., Gelboin, H. V., and Gonzalez, F.
J. Five of 12 forms of vaccinia virus-expressed human hepatic cytochrome
P450 metabolically
4790-4793.
1990." activated aflatoxin Bl. Proc. Nati. Acad. Sci. USA, 87:
34. Brash, D. E., Seetharam. S., Kraemer. K. H., Seidman, M. M., and Bredberg,
A. Photoproduct frequency is not the major determinant of ultraviolet mu
tation hotspots or coldspots in human cells. Proc. Nati. Acad. Sci. USA, 84:
3782-3786,1987.
35. Burnouf, D., Koehl, P., and Fuchs, R. P. P. Single adduci mutatgenesis:
strong effect of the position of a single acetylaminofluorene adduci within a
mutation hotspot. Proc. Nati. Acad. Sci. USA, 86: 4147-4151. 1989.
36. McMahon, G., Davis, E. F., Huber. L. J., Kim, Y., and Wogan, G. N.
Characterization of c-Ki-ras and N-ras oncogenes in aflatoxin Bl-induced rat
liver tumors. Proc. Nati. Acad. Sci. USA, 87: 1104-1108, 1990.
37. Chang, Y-J., Mathews, C., Mangold, K., Marien, K., Hendricks, J., and
Bailey, G. Analysis of ras gene mutations in rainbow trout liver tumors
initiated by aflatoxin Bl. Mol. Carcinog., 4: 112-119, 1991.
38. Hsu. I. C., Metcalf, R. A. Sun, T., Welsh, J. A., Wang, N. J., and Harris, C.
C. Mutational hotspot in the p53 gene in human hepatocellular carcinomas.
Nature (Lond.), 350: 427-428, 1991.
39. Bressac. B., Kew, M., Wands, J., and Ozturk, M. Selective G to T mutations
of p53 gene in hepatocellular carcinoma from southern Africa. Nature
(Lond.). 350: 429-431, 1991.
5673
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1992 American Association for Cancer Research.
Sequence Specificity of Aflatoxin B1-induced Mutations in a
Plasmid Replicated in Xeroderma Pigmentosum and DNA Repair
Proficient Human Cells
Dan D. Levy, John D. Groopman, Susan E. Lim, et al.
Cancer Res 1992;52:5668-5673.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/52/20/5668
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. © 1992 American Association for Cancer Research.