|
|
|
|||||||
|
|||||||||
(Received for publication, May 19,
1994; and in revised form, November 16, 1994) From the
Benzo[a]pyrene-7,8-dihydrodiol 9,10-epoxide
(BPDE), a metabolite of the widespread environmental pollutant
benzo[a]pyrene, is mutagenic in both bacterial and
mammalian systems. Toward understanding the mutagenic effects of
different stereoisomers of BPDE at specific sites in DNA, six
stereochemically defined BPDE adducts were constructed on adenine N
Polycyclic aromatic hydrocarbons (PAH) ( There is considerable evidence to
demonstrate that primary nucleotide sequence can modulate the
stereo-selectivity and distribution of BPDE lesions in modified
DNA(8, 9, 10) . Thus, carcinogenic species
may be biased toward or against certain bases because of the
stereoelectronic effects of adjacent bases or the stereochemistry of
the ultimate carcinogen(11) . Furthermore, preferential
site-specific mutation at a particular position could be due to the
relative stability of the adduct or the lack of significant structural
distortions in the nucleotide caused by the carcinogen(12) .
The interaction of these PAH diol epoxides with numerous sites in DNA
involves a nucleophilic attack in every case at the benzylic carbon of
the epoxide resulting in a S The heterogeneity of base adduction by BPDE
is significantly reduced by synthesizing optically pure (+)- or (-) -anti or -syn enantiomers.
Variations in biological activity between enantiomers within a given
test system are likely to be due to different conformations assumed by
these adducts. Studies that differentiate between effects due to
adducts produced by cis and trans addition of anti- and/or syn-BPDE to DNA are also gaining
importance(16, 17) . Studies examining the metabolism
of BP in 3-methylcholanthrene-treated rats have shown four possible
BPDE isomers. The ratios of formation for
(+)-anti-BPDE/(-)-syn-BPDE/(-)-anti-BPDE/(+)-syn-BPDE
were 214:36:24:1,
respectively.(18, 19, 20, 21) .
However the adduct-forming potentials of these BPDE isomers on adenine
have not been firmly established. In spite of the availability of
considerable information on both the binding spectra and mutational
specificity of BPDE, little is known about the relationship between
these two factors within specific sequences. Template-directed
mutagenesis employing oligodeoxynucleotides bearing stereo-specific and
site-specific lesions offer the possibility of correlating a
stereochemically-defined adduct with a particular mutation spectrum.
The mutagenicities of these stereoisomers, however, are different in
bacterial and mammalian cells (22, 23, 24) .
Single-stranded vectors carrying a defined, uniquely located lesion are
powerful tools for investigating mutagenic mechanisms in vivo both in prokaryotic and eukaryotic
systems(25, 26) . The objective of this study was
to correlate in vivo and in vitro replication
competence with BPDE adduct chirality. Toward this goal, six
stereochemically-defined BPDE adducts were constructed on adenine
N
BPDE-adducted N-ras 61-containing 11-mers
were synthesized by the method of Kim et al. (28) and
purified as described by Latham et al.(30) . Both the
nonadducted and the BPDE-adducted N-ras 61-containing
oligonucleotides were phosphorylated with T4 polynucleotide kinase (New
England Biolabs Inc., Beverly, MA). A 100-fold molar excess of the
different N-ras 61 11-mers relative to the amount of
linearized vector were individually ligated together in the presence of
a 2-fold excess of a 51-mer scaffold(25, 30) . Each
reaction was incubated overnight at 16 °C with a total of 400 units
of T4 DNA ligase (New England Biolabs Inc.).
Ligation
efficiencies were further confirmed by electron microscopy subsequent
to the removal of the 51-mer scaffold. Samples were prepared using the
formamide modification of the basic protein (Kleinschmidt) technique as
described by Davis et al.(31) . Grids were rotary
shadowed with platinum/palladium in a ratio of 80:20 and examined in
either a Philips 300 or 410 electron microscope. Projected images were
traced from photographic negatives, and lengths were determined with a
map measure.
Figure 1:
Structures of adenine N
The purities of 11-mers containing
the isomeric dA-BPDE adducts were also analyzed on 15% polyacrylamide
sequencing gels subsequent to labeling the 5` terminus using T4
polynucleotide kinase and [
Figure 2:
Polyacrylamide gel electrophoresis of
11-mers containing a single isomeric dA N
Figure 3:
Schematic representation for the
insertion of N-ras 61 oligodeoxynucleotides in the M13mp7L2
genome and the detection of point
mutations.
Figure 4:
Determination of specific point mutations
at position 2 of N-ras codon 61. 11-mers containing the
N-ras 61 codon, differing only by A
To further determine the
limits of sensitivity of this differential hybridization, four 51-mer
sequences bearing the N-ras 61 codon within them but differing
only by a single nucleotide at position 2 were constructed. Dot-blot
assays were performed with these DNAs ranging from 1 pg to 5 µg.
DNA amounts as low as 1 ng hybridized with their complementary
17-mers with high specificity and were distinctly detected after an
overnight exposure. DNAs ranging from 0.5 to 5 µg further exhibited
an intense signal with their corresponding probes under exposures as
short as 15 min (data not shown). These control experiments firmly
establish the viability of phage containing any type of point mutation
and our ability to identify any of these mutations resulting as a
consequence of replication past a BPDE-adducted site.
In concert with the
results of Southern blot analyses for the determination of ligation
efficiencies, independent confirmation was obtained through direct
visualization of doubly-ligated molecules by electron microscopy. Using
denaturing electron microscopic methodologies, the appearance of
circular single-stranded DNA molecules was direct evidence for the
presence of doubly ligated 11-mers into the EcoRI restriction
site of M13mp7L2 DNA. Linear single-stranded DNA molecules represented
either the vector DNA with no insert or those that were singly ligated
to an 11-mer. At least 200 molecules were scored for each of the six
stereoisomerically-defined BPDE-adducted DNAs and the nonadducted
N-ras 61-11-mer template samples. The average
full-length circular forms of ligated vector DNA with the insert was
1.76 ± 0.18 µm (Fig. 5). As discussed above, although
full-length linear DNA molecules were detected by electron microscopy,
it was not possible to distinguish singly-ligated molecules from the
linearized vector alone due to the small insert size. This electron
microscopic study provides an alternative methodology for verification
of the double ligation event.
Figure 5:
Ligated circles of BPDE-adducted 11-mer
within the single-stranded M13mp7L2 vector as determined by electron
microscopy. The BPDE-modified 11-mers were inserted into an unique EcoRI site within M13mp7L2 single-stranded DNA, resulting in
covalently closed circular molecules.
Similar to the ligation efficiencies,
the plaque-forming abilities of all the six stereochemically defined
BPDE-adducted DNAs were distinctly lower than the corresponding values
of the unmodified template. Furthermore, a wide spectrum of
plaque-forming abilities was observed, ranging from 5.8% for
the(-)-anti-trans-BPDE adduct to 22.8% for the
(+)-anti-cis-BPDE adduct ( Table 1and Table 3).
Figure 6:
Synthesis of site-specific BPDE-adducted
33-mers employed for in vitro studies.
Equimolar ratios of templates (nonadducted and adducted) and
Figure 7:
A kinetic analysis of primer extension
reactions with templates containing various BPDE stereoisomers. The
sequence of the template-primer complex is represented at the top. The adducted site is designated by an asterisk.
Chain elongation studies were performed at 2, 5, 10, and 30 min. Lane1, N-ras 61-33-mer; lane2, (+)-anti-trans-BPDE-33-mer; lane3, (-)-anti-trans-BPDE-33-mer; lane4, (-)-syn-trans-BPDE-33-mer; lane5, (+)-syn-trans-BPDE-33-mer; lane6, (+)-anti-cis-BPDE-33-mer; lane7,
(-)-anti-cis-BPDE-33-mer.
Figure 8:
Differential blockage of replication at
the adducted site and 1 base 3` to the site of lesion. Lanes2-7 contain templates with different stereoisomeric
BPDE adducts. Lane1, N-ras 61-33-mer; lane2, (+)-anti-trans-BPDE-33-mer; lane3, (-)-anti-trans-BPDE-33-mer; lane4, (-)-syn-trans-BPDE-33-mer; lane5, (+)-syn-trans-BPDE-33-mer; lane6, (+)-anti-cis-BPDE-33-mer; lane7, (-)-anti-cis-BPDE-33-mer.
These data were taken from a 30-min primer extension reaction with the
Klenow fragment. The 27-mer and 28-mer represent the partially extended products up to 1 base 3` to the
adducted site and opposite the site of lesion, respectively.
Percentages of fully extended primer and blockage at various positions
were analyzed by VISAGE gel electrophoresis analysis system and were
tabulated as shown below.
Extended products were
quantitated by densitometric analysis of the autoradiographs. A 30-min
time point was chosen to measure the amount of fully-extended or
truncated products formed with nonadducted and adducted templates (Fig. 8). Based on the autoradiographic signal, 99.3% of the
extended product of the nonadducted template was of full length (Fig. 8, lane1). With the adducted templates
that inhibited replication beyond 1 base 3` to the site of lesion,
approximately 99.7% of the partially extended products were represented
by this premature termination (Fig. 8, lanes2, 5, and 7). Two of the lesions
((-)-anti-trans- and
(-)-syn-trans-) responsible for significant pause sites
1 base downstream of the adduct revealed polymerized products amounting
to two-thirds of all the extended products (Fig. 8, lanes3 and 4). The remaining one-third was
represented by those product molecules that terminated opposite the
corresponding adduct. The (+)-anti-cis-BPDE enantiomer
exhibited roughly a 14:1 ratio of partially extended products 3` to the
adducted position to those opposite the lesion (Fig. 8, lane6). In essence, all of the BPDE-adducted templates proved
to be poor substrates for in vitro replication with Klenow
fragment in contrast to the relatively efficient replication competence
of four of these six adducted sequences in repair-deficient E. coli cells after adjusting for ligation efficiencies. Specific chiral interactions of different diastereomers of
BPDE with nucleophilic sites in DNA cause genotoxic lesions that lead
to consequences such as mutations, which can initiate a cancerous
response in cells, subsequently leading to alterations in gene
expression(23, 34, 35) . The focus of this
study was to observe the role of six stereochemically-defined BPDE
adducts both in vivo and in vitro when anchored at a
specific position on DNA. The mutagenic potential of dA-BPDE adducts
that were determined by in vivo studies included the relative
lethality of these lesions as well as the type and incidence of point
mutations arising from replication of the damaged DNA. An additional
parameter investigated by in vitro observations involved the
role of Klenow fragment in translesion synthesis on encountering the
various stereospecific BPDE adducts attached to the DNA template. To
assess the relative abilities of these stereoisomers to adversely
influence replication, it is critical that the modified
oligonucleotides be of utmost purity(33) . Toward this goal,
synthesis of oligodeoxynucleotides bearing adducts at exocyclic amino
sites of adenine was carried out by the postoligomerization method, and
the integrity of the samples was determined by a variety of analytical
techniques(27, 28) . The resultant high purity of the
adducted oligonucleotides establishes that within our ability to
determine, the mutational frequencies observed were due to the adducts
alone and not as a result of contaminants (Fig. 2). In
vivo studies involved repair-deficient cells that had a recA A wide spectrum of
lethality was observed within the various adducted templates examined
even after adjusting for ligation efficiencies. When considering
ligation efficiencies and plaque-forming abilities as contributors to
survival, the data suggest that percentage lethality follows the rank
order of(-)-anti-trans- >
(-)-syn-trans- >
(+)-anti-trans- >
(+)-syn-trans- > (+)-anti-cis- In vivo mutagenesis of all of the BPDE adducts studied revealed only
A In vivo analyses of
translesion synthesis of the BPDEadducted templates were complemented
by in vitro studies. Previous in vitro primer
extension assays with BPDE-adducted templates involved almost
exclusively guanine residues (16, 51, 52) .
These bulky DNA adducts are known to block DNA replication with certain
enzyme systems either at or 1 base prior to the site of the adduct in
the template(53) . Recent in vitro studies with
oligonucleotides containing stereospecific trans-adducts of anti- and syn-BPDE on adenine indicated that the
polymerase (Sequenase) was completely arrested at 1 base 3` to the
adduct(54) . The present study is a further effort to
investigate how BPDE-adenine adducts behave upon encountering
polymerases in an in vitro system. Although replication in E. coli cells is predominantly performed by DNA polymerase
III, the difficulty of assembling all the core proteins involved in the
proper functioning of this holoenzyme is a major limiting factor.
Therefore, the alternate choice of Klenow fragment was made due to its
wide usage with various adducted templates. Kinetic analyses in this
investigation revealed that the Klenow fragment had no capacity to
perform translesion synthesis even after an incubation period of 30
min. Temporal (from 2-30 min) comparison of the patterns of the
partially extended products from each adducted template indicated that
the primer extensions were completed as early as 2 min. All of the
template and primer were utilized completely as indicated by the fact
that no primer was left at the 17-mer position when equimolar
substrates were employed in the reactions (Fig. 8). Quantitative analyses of the truncated products exhibited almost
total termination 3` to the adducted site with
(+)-anti-trans-, (+)-syn-trans-,
and (-)-anti-cis-adducted templates, whereas with
(-)-anti-trans-,(-)-syn-trans- and
(+)-anti-cis-BPDE adducts a stop site was observed 3` to
the adducted site, followed by complete blockage opposite the lesion (Fig. 8). Besides the bulky nature of the adducts that could be
responsible for this resistance to in vitro replication, the
orientation and tilt of the pyrenyl ring relative to the modified
strand could be a causal factor for stalling and abrupt cessation of
primer extension with the Klenow fragment. It is not unlikely that
those adducts impeding the enzyme progress beyond 1 base 3` to the
lesion site (Fig. 8, lanes2, 5, and 7) could be stereo-selectively distinct in angle positioning
from the isomers that allow incorporation of a nucleotide opposite the
adduct (Fig. 8, lanes4 and 6). This
dichotomy is further exacerbated by the fact that isomers categorized
under the first group exhibit a S configuration at the C-10 of
BPDE, whereas the isomers encompassing the second group revealed a R conformation at the same position (Fig. 1). There is
evidence to show that PAH with 10S configuration point in the
opposite direction to those with the 10R configuration (3) . The precedence of exonuclease digestion controlled by
adduct orientation relative to the 5`
Volume 270,
Number 10,
Issue of March 10, 1995 pp. 4990-5000
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
a
pyrene-7,8-dihydrodiol
9,10-Epoxide Adducts on Adenine N
at the Second
Position of N-ras Codon 61 (*)
ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES
at position 2 of the human N-ras 61
codon within an 11-base oligonucleotide fragment. Both the nonadducted
and BPDE-adducted N-ras 61 11-mers were inserted into a unique EcoRI site in single-stranded M13mp7L2 DNA and utilized for in vivo studies. The ligation efficiencies of BPDE-adducted
11-mers into the single-stranded vector were determined by Southern
hybridization and confirmed by electron microscopy. Repair-deficient
AB2480 E. coli cells were transformed with adducted and
nonadducted DNA samples. The resultant plaque-forming abilities were
used to evaluate the replication competence of the various BPDE adducts
with respect to the nonadducted 11-mer. Point mutations due to aberrant
replication at the adducted site were identified by the technique of
differential DNA hybridization. All of the six BPDE adducts examined
were mutagenic in vivo, generating exclusively A
G
mutations at frequencies ranging from 0.26 to 1.20%. In vitro replication studies using these BPDE-adducted 11-mers involved
primer extension assays with Klenow fragment. All of the BPDE-modified
templates demonstrated distinct blockage at the adducted site and/or 1
base 3` to the adducted site, allowing essentially no translesion
synthesis to form fully extended polymerization products in
vitro.
)are
pervasive in the environment, arising during combustion processes. Some
of these PAH are carcinogens. Benzo[a]pyrene (BP) is
one such PAH that has received intense study in an attempt to define
mechanisms of genotoxicity. This compound is metabolically activated to
bay region 7,8-dihydrodiol 9,10-epoxides that initiate mutagenesis and
carcinogenesis by covalently binding to
DNA(1, 2, 3) . The mutagenic potential of
these diol epoxides is dependent on a variety of interactions including
ones between the carcinogen and the template, the nature of the
polymerase involved in replication past the adduct, and the efficiency
of DNA repair within the cell. DNA lesions caused by such carcinogens
if improperly repaired, may be converted to permanent changes in the
genome during DNA replication. These changes involve base
substitutions, deletions, or frameshift mutations that eventually can
lead to neoplastic transformation(2) . Incubation of BPDE with
DNA containing ras genes can result in
oncogenesis(4, 5) . The mechanism by which these ras genes are activated in tumor cells often involves a single
point mutation, usually resulting in the alteration of amino acid
residue 12 or 61 of the protein encoded by these genes(4) . In vitro mutagenesis experiments showed that activating
mutations could also occur at codons 13, 59, and
63(6, 7) .
at position 2 of N-ras codon 61 within an
11-base oligodeoxynucleotide by the postoligomerization
strategy(27, 28) .
Bacterial Strains and Culture Conditions
The
strain of Escherichia coli used for extraction of
single-stranded M13mp7L2 DNA was UT481 (met thy
(lac-pro) hsdR
BamHI hsd M+ sup DTn10/F` tra D 36 pro AB
lacI
ZM15). Cells were grown in minimal
medium at 37 °C, and overnight shaking cultures (200 rpm) were
diluted 1:100 in Luria-Bertani broth and grown to an A
of 0.2. M13mp7L2 phage were used to transfect these cells.
Bacterial cells were harvested at the end of 16 h, and single-stranded
phage DNA was isolated from the supernatant. Repair-deficient AB2480 (uvrA16 recA13, courtesy of Dr. A. Ganesan, Stanford
University) E. coli cells were utilized to study the
replication competence and mutagenesis of N-ras 61-adenine
-M13mp7L2 templates by transfecting
these cells with DNA bearing different stereoisomeric adducts of BPDE.
Shaking cultures were routinely grown at 37 °C to an A of 0.35 for transfection, while the feeder
cells were grown to stationary phase.Isolation of Single-stranded DNA and Insertion of N-ras
61-containing Oligonucleotide
Single-stranded M13mp7L2 DNA (a
gift from Dr. C. Lawrence, University of Rochester) was isolated as
described by Sambrook et al.(29) . A unique EcoRI site within an engineered hairpin loop was utilized to
linearize the phage DNA. This restriction digested DNA was passed
through a Nensorb column (Bio-Rad) for subsequent purification. A
fraction of the endonuclease-digested DNA was analyzed by
electrophoresis through a 1.4% agarose gel and visualized by ethidium
bromide staining. The distinctive greater mobility of linearized
single-stranded DNA relative to the corresponding circular DNA
confirmed that the reaction proceeded to completion. The linearized
M13mpL2 DNA was reconstituted in 10 mM Tris-HCl (pH 8.0), 1
mM EDTA at a concentration of 0.5 µg/µl and stored at
-20 °C.Determination of Ligation Efficiencies
The
efficiency of incorporation of the nonadducted and BPDE-adducted
N-ras 61 11-mer into the unique EcoRI site of
M13mp7L2 DNA was monitored by electrophoresis on a 1.4% agarose gel and
subsequent Southern blot analysis. A 
P-end labeled probe
was used that was complementary to the N-ras 61 11-mer with
the flanking M13mp7L2 sequences. Both singly and doubly ligated DNA
molecules were identified by autoradiography, but only doubly ligated
molecules were utilized to calculate ligation efficiencies.
Prehybridization and hybridization was performed at approximately 25
°C for 16 h. Densitometric scanning of autoradiographs (Hyperfilm
MP from Amersham Corp.) was performed with the VISAGE gel
electrophoresis analysis system (BioImage, Ann Arbor, MI).Bacterial Replication and Mutagenesis of N-ras
61-Ade
Repair-deficient AB2480 (uvrA-M13mp7L2
, recA) E. coli cells were transfected with 0.5 µg of the
N-ras 61-Ade-M13mp7L2 DNA by the calcium
chloride/rubidium chloride procedure(29) . The 51-mer scaffold
was displaced from the ligated template prior to transfection by the
addition of a 5-fold excess of a sequence complementary to the 51-mer
(c51-mer) followed by heat denaturation for 2 min at 95 °C and
rapid cooling on ice. After correcting for ligation efficiencies, the
plaque-forming abilities of different N-ras 61 adducts in
M13mp7L2 represented the frequency of replication bypass through the
BPDE lesions. The error rate of bypass and the spectrum of point
mutations resulting from in vivo replication was determined by
differential hybridization. Plaques were transferred onto
nitrocellulose filters by successively lifting each plate 4 times. Each
set of filters was hybridized with about 100 pmol of one of four P-labeled 17-mer oligonucleotides constituting an 11-base
complementary region to the N-ras 61 sequence. Each probe
differed by only 1 base at the site opposite the adducted position and
was represented by either A, T, C, or G. Hybridization was performed at
39 °C for 16 h, a stringency at which only perfect matches in
sequence could be scored as positives. Prehybridization and
hybridization solutions constituted 5 
saline/sodium
phosphate/EDTA (SSPE), 5 Denhart's solution, 50 µg/ml
fish milt DNA, 0.1% SDS, and 20% formamide. Excess radiolabel was
removed by washing the filters 3 times with 2 SSPE at 39 °C
for 10 min on each wash. Representative samples of wild-type and mutant
plaque DNA that had been identified by the technique of differential
hybridization were confirmed by dideoxy sequencing(32) .
Control assays were performed in vivo utilizing M13mp7L2 DNA
containing 11-mer inserts with N-ras codon 61, wherein the
adenine at position 2 (designated 61
) was modified to C, G,
and T successively and hybridized with their complementary 17-mer
probes. Parallel in vitro control assays involved dot blot
assays with 51-mers bearing the 11-mers, wherein the N-ras 61 adenine was replaced by C, G, and T and
correspondingly hybridized with the complementary probes.Enzymatic Construction of BPDE-adducted 33-mers for
Primer Extension Assays
To facilitate in vitro primer
extension of the 11-mer oligonucleotide bearing various
stereochemically-defined BPDE adducts, 33-mers were constructed for the
stable alignment of the primer. As detailed by Latham et al.(30) , a 5-fold molar excess of a 22-mer bridge was
utilized relative to the various N-ras 61 11-mers. The 5`
phosphorylated 22-mer bridge included a 17-base terminal region
complementary to the M13 sequencing(-40) primer and a portion of
the EcoRI site from the M13mp7L2 sequence. The N-ras 61 11-mers with the remaining sequence of the restriction site
were 5`-end labeled with 1:100 labeled/unlabeled ATP. A 27-mer scaffold
utilized in 3-fold excess to the two oligonucleotides was designed to
align the 5`-PO
of the 22-mer in proximity to the 3`-OH of
the N-ras 61 11-mer. Ligation reactions were performed in the
presence of 2000 units of T4 ligase at 16 °C for 48 h. Ligated
33-mer products were separated on 10% polyacrylamide gels and detected
by subsequent autoradiography. The corresponding bands were excised and
DNA eluted in water. Excess urea was eliminated, and the
oligonucleotides precipitated in absolute alcohol.In Vitro DNA Synthesis
M13 sequencing(-40)
primer was end-labeled with [
-P]ATP (3000
Ci/mmol) (DuPont NEN) and added to BPDE-adducted or nonadducted 33-mer
templates in an equimolar ratio of 0.5 pmol each. The primer was
annealed to various templates at 65 °C for 2 min and slow-cooled to
<30 °C. One unit (
0.5 pmol) of Klenow fragment was utilized
per reaction in a volume of 20 µl. Reactions were performed at 37
°C under conditions specified by Latham et
al.(30) . Aliquots of reactions were terminated at various
time points ranging from 2 to 30 min. Primer-extended products were
resolved on 15% denaturing polyacrylamide sequencing gels with 8.3 M urea (30 cm 35 cm .04 cm) and visualized by
autoradiography. Various exposures of the gel to Hyperfilm MP (Amersham
Corp.) enabled quantification of the percentage of extended primers by
densitometry using the VISAGE gel electrophoresis analysis system
(BioImage, Ann Arbor, MI). Band intensities were further confirmed by
the use of the PhosphorImager (Molecular Dynamics, Sunnyvale, CA).
Purity of Stereochemically-defined BPDE
Adducts
Oligodeoxynucleotides containing stereo-specifically and
site-specifically placed lesions are of great value in studying single
adduct mutagenesis in bacterial and mammalian
systems(17, 28) . DNA templates with defined adducts
can also be utilized in in vitro studies to evaluate the
replication capability of various DNA polymerases in the vicinity of
the lesion(16, 23) . To investigate the effect of
individual stereoisomers in vivo and in vitro, six
stereochemically-specific BPDE adducts on the N of adenine
at position 2 of N-ras codon 61 on an 11-mer sequence were
synthesized. The six modified N-ras codon 61
oligodeoxynucleotides contained (+)- or
(-)-anti-trans-, (+)-
or(-)-syn-trans-, or (+)-
or(-)-anti-cis-BPDE adducts (Fig. 1). Samples
were analyzed and purified by HPLC, polyacrylamide gel electrophoresis,
and capillary gel electrophoresis. In addition, the
oligodeoxynucleotides were digested with nuclease P1 followed by snake
venom phosphodiesterase and alkaline phosphatase, and the resulting
nucleoside mixture was analyzed by HPLC. Only the expected nucleosides
were observed (data not shown).
stereoisomers placed at the second position of N-ras codon 61 within an 11-mer
oligodeoxynucleotide.
-P]ATP.
Electrophoretic migration of BPDE-modified oligomers was distinctly
slower than that of unmodified 11-mers (Fig. 2). Slight
differences in mobility pattern were also observed among the various
adducted 11-mers. Gels exposed for longer periods revealed no traces of
unmodified 11-mer contaminants in the six dA-modified
oligodeoxynucleotides.
-BPDE adduct.
Oligodeoxynucleotides were labeled with P at the 5`
terminus and subjected to electrophoresis through 15% polyacrylamide
gels (33 44 0.04 cm). The BPDE-adducted oligomers are
shown in lanes2-7. Lane1,
unmodified 11-mer; lane2,
(-)-syn-trans-BPDE 11-mer; lane3,
(+)-syn-trans-BPDE 11-mer; lane4,
(+)-anti-cis-BPDE 11-mer; lane5,
(-)-anti-cis-BPDE 11-mer; lane6,
(+)- anti-trans-BPDE 11-mer; lane7,
(-)-anti-trans-BPDE 11-mer.
Control Assays for Detection of Specific Point Mutations
at Position 2 of N-ras Codon 61
Prior to presenting the data
concerning the biological fate of various BPDE-adducted 11-mers that
were introduced into the unique EcoRI site in M13mp7L2 DNA, it
was important to determine if all point mutations could be detected
with the same frequency at position 2 of N-ras codon 61. Given
that the ras gene is known to exhibit high secondary
structure, 11-mers bearing the N-ras codon 61, differing only
by AC or AG or AT at position 2, were inserted into
the phage DNA and detected by differential hybridization and then
plaque purified (Fig. 3). High titer phage stocks were readily
produced, indicating that there was no difficulty associated with
replication of each of these genomes. Using a reconstruction experiment
to ensure that we detect any point mutation in a background of wild
type plaques, the following experiments were performed. Mutant phage
particles were mixed with wild-type phage approximately at a ratio of
1:25 (mutant/wild-type, respectively) and transfected into
repair-deficient E. coli cells. As shown in Fig. 4A, differential hybridization was performed on
plaques transferred to nitrocellulose filters by four successive lifts
of each plate followed by incubation of each filter with one of the
four P-labeled 17-mers differing by only 1 base at
position 2 of N-ras 61. Hybridization was highly specific as
determined by a positive signal only with the complementary probes. Fig. 4B indicates the total number of phage that
hybridized with their corresponding complementary sequences, and these
values approximately equaled the expected ratio of 25:1,
wild-type/mutant phage, respectively.
C or AG or
AT at position 2, were inserted into the unique EcoRI
site of M13mp7L2 phage DNA. A shows a mixture of
mutant/wild-type plaques (1:25, approximately) that hybridized
specifically only to their complementary probes. B is a
tabular form of the actual number of wild-type and mutant phage that
give a positive signal, and these values approximately equal the
original number of plaques plated.
Ligation Efficiencies and Replication Competence of N-ras
61 BPDE-adducted DNA in Vivo
Plaque-forming abilities by the
various BPDE-adducted and nonadducted N-ras 61-M13mp7L2 DNA
were taken as a measure of their replication competence in
repair-deficient AB2480 E. coli cells (Fig. 3). For
each of the stereoisomers in this study, at least three separate
ligation reactions were performed followed by a minimum of two
transfections/ligation. To facilitate screening of plaques, experiments
were performed in three batches, each involving a pair of enantiomers.
The number of plaques screened for each of the adducted samples was
taken as a percentage of the total number of plaques obtained by
transfection with the nonadducted sample, the latter value being
specific to each individual set of stereoisomers. These results are
compiled in Table 1Table 2Table 3. Linearized
M13mp7L2 DNA was used as a control to detect the background levels of
plaque formation by these DNA molecules. The ligation efficiency with
each of the six stereochemically-defined BPDE-adducted DNAs was
markedly below that of the unmodified N-ras 61-11-mer
template. As tabulated, the percentage of ligation efficiency relative
to the nonadducted DNA ranged from 8.5% for the
(+)-anti-trans-BPDE adduct to 31.2% for the
(-)-syn-trans-BPDE adduct ( Table 1and Table 2). Besides doubly-ligated molecules, Southern
hybridization indicated the presence of singly-ligated molecules that
had a greater mobility than the covalently closed circular molecules on
a 1.4% agarose gel (data not shown). However, the contribution of these
linear molecules to the percentage of plaques formed could be no more
than 0.1-0.3% of the nonadducted sample based on the data derived
from the plaque-forming ability of the linearized nonadducted M13mp7L2
DNA (Table 1Table 2Table 3).
In Vivo Mutagenesis of Stereochemically Defined BPDE
Adducts
The plaques formed by the replication of various
BPDE-adducted and nonadducted N-ras 61-M13mp7L2 DNAs in AB2840 E. coli cells were screened by differential hybridization. To
determine the spectrum and frequency of point mutations as a result of
replication past the adducted site, four probes differing by only a
base opposite the modified site were utilized. As shown in Table 1-3, AG transitions were the sole mutations
observed with all six stereo-specific BPDE adducts analyzed. The
mutational frequency ranged from 0.26 to 1.20%, indicating subtle
changes in the incidence of point mutations. Similar mutation rates
were observed even on lowering the concentrations of the BPDE-adducted
11-mers in the ligation reactions such that their molar ratios relative
to the phage DNA was 10:1 and 2:1 respectively. This observation
strengthened the evidence for the fact that the frequency of single
base substitution was a true reflection of the impact of these bulky
adducts on the cell's replication machinery.Qualitative and Quantitative Analyses of in Vitro DNA
Replication of BPDE-adducted and Nonadducted Oligodeoxynucleotide
Templates
In vitro replication of a given template is
influenced by the polymerase utilized in the reaction. In the present
investigation, Klenow fragment was chosen due to its precedence of
widely being utilized for kinetic studies, although polymerase III is
normally the key enzyme involved in in vivo replication. To
understand the role of this polymerase in in vitro primer
extension assays for the bypass of BPDE lesions, 33-mers were
constructed from each of the adducted oligodeoxynucleotides. Using this
methodology, individual adducts were studied in isolation from all of
the others. Similar to the in vivo studies, six adducted
N-ras 61 templates were synthesized for in
vitro analysis; i.e. 33-mers containing (+)-
or(-) -anti-trans-, (+)-
or(-)-syn-trans-, and (+)-
or(-)-anti-cis-BPDE adducts. Since de novo synthesis of a molecule of this length with
stereochemically-defined BPDE adducts in the N-ras 61 codon is
technically difficult, the templates were constructed enzymatically as
schematically represented in Fig. 6and as detailed under
``Experimental Procedures.'' More than 50% of the ligated
material was recovered with each of the six BPDE adducted 33-mers.
P-end-labeled primer were utilized for the polymerization
reactions. A time course study was performed that included primer
extension reactions for 2, 5, 10, and 30 min. Sequence analyses of
extended primers replicated on DNA templates were carried out by
subjecting the reaction mixtures to 15% polyacrylamide gel
electrophoresis. Following electrophoresis, positions of the oligomers
were established by autoradiography as shown in Fig. 7.
Sequences for both the template and primer are depicted above the
results of primer extension (Fig. 7). No qualitative differences
were observed within individual extension reactions for any of the
adducted templates, ranging from 2 to 30 min. The reactions appeared to
be complete within the first 2 min, indicating no further translesion
synthesis over longer incubation times. The P-end labeled
17-mer primer employed in each of the reactions was completely utilized
as indicated by the absence of a band at the 17-mer position (Fig. 8). The nonadducted 33-mer template exhibited a
full-length product at the end of 2 min of polymerization. In contrast,
none of the six adducted oligodeoxynucleotides accumulated full length
products even after 30 min of synthesis. All six BPDE-modified 33-mers
served as poor templates resulting in blockage of in vitro replication due to synthesis being stopped opposite and/or 1 base
3` to the adducted site. Replication of
(+)-anti-trans-, (+)-syn-trans-, and (-)-anti-cis-BPDE-adducted templates was
completely blocked at 1 base 3` to the adducted site.
With(-)-anti-trans-, (-)-syn-trans-, and
(+)-anti-cis-BPDE-modified 33-mers, a nucleotide was
placed opposite the adducted site in each case but no replication
occurred beyond that point (Fig. 7).
, uvrA genotype. This eliminated the effect of inducible responses
attributed to DNA repair. Furthermore, the choice of a single-stranded
vector was advantageous in two ways. First, single-stranded DNA are
poorer substrates than double-stranded genomes for repair, consequently
aiding in a better understanding of template-directed
mutagenesis(36, 37) . Second, the ease of introducing
nonadducted or adducted oligodeoxynucleotides into single-stranded
M13mp7L2 is far greater than inserting oligomers into a gapped
duplex(38, 39, 40, 41) . However it
is not uncommon to obtain poor ligation efficiencies with modified DNA
as observed in this study(40) . This could be attributed to the
interaction of the bulky BPDE adduct with the template, thus causing
structural distortions. The broad range of ligation efficiencies from
8.5 to 31.2% exhibited by the different stereoisomers is likely to be a
result of the chirality of the molecule. The presence of a distinct
population of singly- ligated molecules in the ligation reaction could
be either because of the formation of a secondary structure by the
vector DNA or because the physical presence of the adduct makes it
difficult to form a closed molecule. The recovery of relatively good
yields of the 33-mer constructs for in vitro analyses suggests
that the ligation at least at the 3`-OH of the 11-mer is reasonably
efficient. Therefore, the perturbations inhibiting ligation are more
likely to occur at the 5` end of the 11-mer.
(-)-anti-cis-BPDE-adducted templates (Table 1-III). It is possible that there is either direct
blockage of polymerase III activity or apparent loss of processivity by
this holoenzyme after bypassing the(-)-anti-
and(-)-syn-trans-BPDE adducts that lead to decreased
survival. Similar diminished levels of enzyme processivity were
observed in replicative bypass of an abasic DNA lesion(42) .
However, the low to nonlethality of the remaining four BPDE isomers
examined could be due to very little or no blockage of DNA replication in vivo(43) . Thus, differences in spatial
configuration influence the template properties of lesions toward DNA
replication and survival of the cell(23) .G transitions. These findings are in contrast to earlier reports
both in prokaryotic and mammalian cells wherein AT transversions
were prevalent when adenine was the site of lesion for BPDE or other
bulky adducts such as 9,10-dimethy-1,2-benzanthracene and cis-diamminedichloroplatinum
II(4, 5, 11, 44, 45, 46, 47) .
Previous studies using a styrene oxide DNA adduct in the same sequence
context as this work, however, resulted in AG base
substitutions(30) . These mutations do not follow the
``A-rule'' put forth to explain the mutational behavior of
abasic and bulky, ``noninstructional'' lesions. DNA
polymerases that preferentially insert adenine opposite these sites of
lesions are believed to be subject to an A rule. Therefore, in
contrast, dA may be directly miscoding or
misinstructional rather than requiring a ``default'' mutation
mode. This misinstructional lesion effect could possibly be influenced
by local sequence context. In addition, it could be a consequence of
the structural distortion of the adducted base that preferentially
allows AG transitions alone to occur. Results presented in this
study exhibited a frequency of error spanning from 0.26 to 1.20%,
indicating a 5-fold difference. Furthermore, decreasing the molar
concentrations of the adducted 11-mers by 10-50 fold in the
ligation reaction caused no change in the percentage mutations, thus
attributing the mutagenecity to the BPDE adducts rather than to any
contaminants. The small yet significant changes in the mutation
frequency among the BPDE lesions studied, could be a consequence of
adduct conformational polymorphism resulting in varying interactions
with cellular enzyme systems. The adducts could directly be in contact
with the polymerase involved in replication, leading to stabilization
of a mispaired configuration, as proposed for one of the mitomycin
C/DNA lesions(48) . Alternatively, they could cause subtle
structural changes in the polymerase or the template such that optimal
base pairing with the incoming dNTP does not occur. The fact that more
than approximately 98% of the time these bulky adducts were not
mutagenic in this study, implies that DNA polymerases can be flexible
without completely compromising fidelity(17) . These enzymes
may have an additional ``sensor'' to bypass any structural
distortions the adducts make in the major groove, similar to the
observations made with the N-dG BPDE adducts(49) .
Furthermore, they may also identify bases in spite of improper hydrogen
bonding between base pairs, as observed with the DNA lesions induced by
vinyl chloride(50) . However, the impact of DNA repair enzymes
substantially altering our in vivo results in the present
study was curtailed by allowing replication to occur in a
repair-deficient environment. 3` strand polarity
indicates its influence on enzyme action(55) . Likewise it is
probable that polymerization by a specific enzyme is dictated by the
bulky adduct orientation relative to the site-specifically modified
single-stranded DNA. NMR studies on the duplex structure of adducted
oligonucleotides at N of dG determined the directionality
of various stereoisomers of
BPDE(49, 56, 57) . Similar observations on
spatial positioning were made with styrene oxide adducts that showed a
pattern of in vitro blockage parallel to those obtained in
this study(30, 58) . In spite of BPDE lesions forming
effective blocks to DNA synthesis by Klenow fragment in vitro,
it is clear that in vivo replication, which is necessary for
the survival of a cell, occurred beyond the adducted site, probably
through polymerase III.
)
We thank Dr. M. L. Dodson for helpful comments
throughout the course of this study and M. L. Augustine for technical
assistance.
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  L. Jia, N. E. Geacintov, and S. Broyde The N-clasp of human DNA polymerase {kappa} promotes blockage or error-free bypass of adenine- or guanine-benzo[a]pyrenyl lesions Nucleic Acids Res., November 1, 2008; 36(20): 6571 - 6584. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.-Y. Seo, A. Nagalingam, M. Tiffany, and E. L. Loechler Mutagenesis studies with four stereoisomeric N2-dG benzo[a]pyrene adducts in the identical 5'-CGC sequence used in NMR studies: G->T mutations dominate in each case Mutagenesis, November 1, 2005; 20(6): 441 - 448. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Nagatsugi, S. Sasaki, P. S. Miller, and M. M. Seidman Site-specific mutagenesis by triple helix-forming oligonucleotides containing a reactive nucleos |