Misincorporation rate and type on the leading and lagging strands of UV-damaged DNA
Misincorporation rate and type on the leading and lagging strands of UV-damaged DNA
A.
Calcagnile
,
T.
Basic-Zaninovic
+
,
F.
Palombo
w
and
E.
Dogliotti*
Laboratory of Comparative Toxicology and Ecotoxicology, Istituto Superiore di
Sanita',
Roma
,
Italy
Received March 25, 1996;
Revised and Accepted June 10, 1996
ABSTRACT
We have examined the fidelity of replication of the leading and lagging strands
of UV-irradiated DNA by using an EBV-derived shuttle vector system which contains as marker gene for
mutation analysis the bacterial
gpt
gene in both orientations relative to the EBV oriP. Human cells stably
transformed with this vector were UV irradiated and
gpt
mutation rate and type were analysed. An increased mutagenicity associated with UV irradiation was
observed, but the average error frequency was unaffected by the direction of
replication of the target gene. Some variability by position and sequence
context of leading and lagging strand errors was detected, suggesting that the
different architecture of the replication complex for the two strands might, to
some extent, affect mutation spectra. The comparable fidelity of translesion
replication on the leading and lagging strands is in agreement with the current
model for eukaryotic replication that postulates the simultaneous synthesis of
both strands by a DNA polymerase with a proof-reading exonuclease.
INTRODUCTION
The irradiation of cells with UV produces lesions in their DNA which have been
shown to transiently inhibit replication. The primary lesions produced are
cyclobutane pyrimidine dimers (CPD) and less frequent lesions are pyrimidine-pyrimidone (6-4) photoproducts (6-4 PP) (for a review see
1
). Following UV irradiation of
Escherichia coli
cells, DNA replication is transiently arrested at DNA lesions. The arrest of DNA replication is followed by
resumption of DNA synthesis past the blocking lesion. This gap-filling reaction of single-stranded DNA gaps present opposite UV lesions in the irradiated
chromosome is mutagenic (for a review see
2
). Several studies suggest that in eukaryotic cells pyrimidine dimers are also
bypassed and that this bypass is mutagenic. CPD interfere with DNA replication
in vitro
, but the mammalian cell replication complex is capable of carrying out
replication of UV-damaged templates and mutation fixation was observed opposite dipyrimidine
sites (
3
,
4
). More recently, differential
in vitro
replication of a single
cis
,
syn
-thymine dimer in the leading or lagging strand was reported using human cell
extracts (
5
), suggesting that uncoupling of concerted synthesis of the two strands might
occur at lesion-blocked replication forks.
In this study we have addressed the question of whether the asymmetric nature of
the replication complex might lead to differences in error rates for leading and lagging strand replication of UV-irradiated DNA. This same issue has been previously (
6
) investigated by using a cell-free SV40 origin-dependent replication mutagenesis assay. In this system human cell extracts replicate a
SV40-derived plasmid in the presence of T antigen. In our study we have used an
EBV-derived shuttle vector system stably maintained in human cells which
contain as target gene for mutagenicity studies the
gpt
gene in both orientations relative to the EBV oriP. In this system cellular DNA
polymerases replicate the plasmid molecules in synchrony with host cell DNA (
7
) and cellular proteins regulate DNA replication from EBV oriP (
8
).
MATERIALS AND METHODS
Plasmids and cell lines
The shuttle vector used in this study has been previously described (
9
). Briefly, pTF-EBV is a derivative of pF1-EBV (
10
,
11
) which contains as elements required for episomal replication in human cells EBV
oriP and EBNA and as target for mutagenicity studies the bacterial
gpt
gene under the control of the inducible promoter of the mouse metallothionein I
gene (MT-I). As a result of the construction scheme, in pTF-EBV the transcribed strand of the
gpt
gene is the leading strand, while in pF1-EBV it is the lagging strand (Fig.
1
A).
Cell treatment and analysis of mutant plasmid DNA
Exponentially growing cells were UV irradiated after removal of the growth medium. Cells were exposed to 7-21 J/m
2
UVC (15 W 15T8 low pressure mercury lamp) and immediately incubated in growth medium.
After three to four cell doublings plasmid DNA was isolated by an alkaline
extraction method as previously described (
10
,
11
). Plasmid DNA was transformed into
E.coli
strain DT2 (
gpt
-
,
pur
+
) by electroporation (BioRad electroporation system). Transformants were plated on minimal salt plates supplemented with
ampicillin and 6-thioguanine (54 [mu]M final concentration) (MATG medium). Mutant
gpt
genes were amplified by PCR with
Taq
polymerase and a 5'-biotinylated primer. Sequencing reactions were performed on single-stranded DNA bound to magnetic beads (Dynabeads M280
Streptavidin).
RESULTS
UV-induced mutation frequency
Clone 7 cells were exposed to increasing UV doses and
gpt
-
mutation frequency was analysed in pTF-EBV plasmid progeny. In this shuttle vector the transcribed strand of the
gpt
gene is the leading strand (Fig.
1
A). As shown in Figure
1
B, a dose-response curve was obtained with a 20-fold increase over background at the dose of 21 J/m
2
. Figure
1
B also displays the UV mutation frequency data previously obtained with the
vector pF1-EBV (
12
), which contains the
gpt
gene in the inverted orientation (Fig.
1
A). The similarity between the two dose-response curves indicates that UV-induced mutation frequency is not significantly affected by the
directionality of replication of the target gene.
Molecular analysis of UV-induced mutations
DISCUSSION
In a previous study (
9
), using the same shuttle vector system, we have shown that the base
substitution error rate and type on an alkylated substrate is similar for
replication of the leading and lagging strands. In this study we have also
shown that in the case of UV-induced lesions neither mode of replication (leading or lagging DNA
synthesis) is more error prone during the bypass reaction. In eukaryotic cells,
as inferred from the SV40 DNA replication model, there is a DNA polymerase switch from [alpha] to [delta] during initiation at the replication origin and for synthesis of
each Okazaki fragment (
14
). This polymerase switching mechanism allows two molecules of DNA polymerase [delta] to replicate both strands of the double helix conjointly. The high
fidelity of synthesis across lesions on both strands that we have observed
would confirm that the contribution to lagging strand synthesis of inaccurate
DNA polymerase [alpha] activity is very limited. The simultaneous synthesis of leading and
lagging strand DNA by polymerase [delta] would indeed position a proof-reading exonuclease on both sides of the replication fork and
therefore would prevent replication-driven mutation strand bias. Whether polymerase [epsilon] plays a role in leading and/or lagging strand synthesis of the
eukaryotic genome is still an open question.
ACKNOWLEDGEMENTS
We thank L.Gargano for technical assistance and Dr M.Bignami for critically
reading the manuscript. This work has been partially supported by EC grant EV5V-CT92-0223.
Present addresses:
+
Faculty of Food Technology and Biotechnology, University of Zagreb, Croatia and
[sect]
IRBM, Laboratory of Biochemistry, Pomezia, Italy
