The sensitivity of human fibroblasts to
N
-acetoxy-2-acetylaminofluorene is determined by the extent of
transcription-coupled repair, and/or their capability to counteract RNA synthesis
inhibition
The sensitivity of human fibroblasts to N -acetoxy-2-acetylaminofluorene is determined by the extent of transcription-coupled repair, and/or their capability to counteract RNA synthesis inhibition
Michiel F.
van Oosterwijk
1
,
Ronald
Filon
1,2
,
Wouter H. J.
Kalle
1
,
Leon H. F.
Mullenders
1,2,
*
and
Albert A.
van Zeeland
1,2
1
MGC-Department of Radiation Genetics and Chemical Mutagenesis, Leiden
University, Wassenaarseweg 72, 2333 AL
Leiden
,
The Netherlands
and
2
J. A. Cohen Institute, Interuniversity Research Institute for Radiopathology and
Radiation Protection,
Leiden
,
The Netherlands
Received September 5, 1996;
Revised and Accepted October 22, 1996
ABSTRACT
Nucleotide excision repair (NER) mechanism is the major pathway responsible for
the removal of a large variety of bulky lesions from the genome. Two different
NER subpathways have been identified, i.e. the transcription-coupled and the global genome repair pathways. For DNA-damage induced by ultraviolet light both transcription-coupled repair and global genome repair are essential to
confer resistance to cytotoxic effects. To gain further insight into the
contribution of NER subpathways in the repair of bulky lesions and in their
prevention of biological effects we measured the rate of repair of dG-C8-AF in active and inactive genes in normal human cells, XP-C cells (only transcription-coupled repair) and XP-A cells (completely NER-deficient) exposed to NA-AAF. XP-C cells are only slightly more
sensitive to NA-AAF than normal cells and, like normal cells, they are able to recover RNA
synthesis repressed by the treatment. In contrast, XP-A cells are sensitive to NA-AAF and unable to recover from RNA synthesis inhibition. Repair of
dG-C8-AF in the active ADA gene proceeds in a biphasic way and without
strand specificity, with a subclass of lesions quickly repaired during the
first 8 h after treatment. Repair in the inactive 754 gene occurs more slowly
than in the ADA gene. In XP-C cells, repair of dG-C8-AF in the ADA gene is confined to the transcribed strand and
occurs at about half the rate of repair seen in normal cells. Repair in the
inactive 754 gene in XP-C cells is virtually absent. Consistent with these results we found that
repair replication in XP-C is drastically reduced when compared with normal cells and abolished by
[alpha]
-amanitin indicating that the repair in XP-C cells is mediated by transcription-coupled repair only. Our data suggest that dG-C8-AF is a target for transcription-coupled repair and that this repair pathway
is the main pathway or recovery of RNA synthesis inhibition conferring
resistance to cytotoxic effects of NA-AAF. In spite of this, repair of dG-C8-AF in active genes in normal cells by transcription-coupled repair and global genome repair is not additive,
but dominated by global genome repair. This indicates that the subset of
lesions which are capable of stalling RNA polymerase II, and are, therefore, a
substrate for TCR, are also the lesions which are very efficiently recognized
by the global genome repair system.
INTRODUCTION
Nucleotide excision repair (NER) is a major pathway by which a large variety of
bulky DNA lesions are removed from the genome. Two different NER subpathways
have been identified, i.e. the global genome repair pathway and the
transcription-coupled repair pathway. Both pathways have been well characterized for
their role in the repair of ultraviolet (UV) light-induced cyclobutane pyrimidine dimers (CPDs) in mammalian cells. In human
cells the global genome repair pathway processes CPDs with different
efficiencies across the genome: (potentially) active genes are repaired at a
much faster rate than non-expressed regions of the genome (
1
). In addition, CPDs are excised by the transcription-coupled repair pathway more rapidly from the transcribed strand of active
genes than from the non-transcribed strand (
2
,
3
). This accelerated repair of CPDs in the transcribed strand of active genes is
dependent on transcription (
4
). It has been proposed that stalled RNA polymerase may act as a signal to
target the NER apparatus to the site of a lesion. From the repair phenotype of
UV-sensitive human cells it is obvious that specific genetic factors are
involved in the NER subpathways. Xeroderma pigmentosum group C (XP-C) fibroblasts appear to be fully capable of performing transcription-coupled repair, but are defective in global genome repair, resulting
in the lack of CPD repair in non-transcribed DNA including the non-transcribed strand of active genes (
3
,
5
). In contrast with CPDs, pyrimidine 6-4 pyrimidone photoproducts (6-4 PP) are not preferentially removed from the transcribed strand of
active genes in normal cells. This lack of strand-specific repair of 6-4 PP in transcriptionally active genes has been shown to be the
result of a very efficient removal of these lesions by the global genome repair
pathway overruling the removal of 6-4 PP by transcription-coupled repair (
6
).
The molecular basis for the observed differences in repair of 6-4 PP and CPD by the global repair pathway most likely lies in local alterations of DNA conformation induced by the photolesions. Further information on the efficiency of processing of other bulky DNA
lesions by the global genome and by the transcription-coupled repair pathways will provide insights in the relationship between
the structure of a DNA lesion, its transcription inhibiting potency, and the
efficiency to be processed by both NER subpathways (with respect to repair of
lesions in active genes). In this study we focussed on the role of the global
genome and transcription-coupled repair pathways in the repair of
N
-acetoxy- 2-acetylaminofluorene (NA-AAF)-induced DNA adducts in active and inactive genes.
NA-AAF causes lesions in DNA which are substrate to NER (
7
-
10
). Although treatment of cells with NA-AAF may principally lead to the formation of two major lesions, i.e. the
deacetylated
N
-(deoxyguanosin-8-yl)-2-aminofluorene (dG-C8-AF) and the acetylated
N
-(deoxyguanosin-8-yl)-2-acetylaminofluorene (dG-C8-AAF), the main lesion induced in human
fibroblasts appears to be the deacetylated dG-C8-AF. Recent studies with normal human fibroblasts (
11
,
12
) and NA-AAF-sensitive Cockayne syndrome cells (CS) (
12
) revealed that transcription-coupled repair does not contribute significantly to the repair of dG-C8-AF in active genes. These observations, as well as the
reported marginal blockage of transcription by dG-C8-AF in
in vitro
transcription assays (
11
,
13
), raise the question whether dG-C8-AF lesions are actually targets for transcription-coupled repair. To assess the role of transcription-coupled repair in the repair of the dG-C8-AF lesion we compared the recovery of NA-AAF inhibited RNA synthesis, the
repair kinetics of dG-C8-AF in active genes and the cytotoxic effects in normal cells, XP-A fibroblasts (completely deficient in NER) and XP-C fibroblasts (only capable of performing transcription-coupled repair). In this study we demonstrate that
dG-C8-AF lesions are a target for both the global genome repair pathway
and the transcription-coupled repair pathway. Both pathways remove dG-C8-AF from active genes with biphasic kinetics, but the
efficiency of repair by global genome repair is ~2-fold higher than by transcription-coupled repair. The biological relevance of transcription-coupled repair of dG-C8-AF is illustrated by the recovery of RNA
synthesis and the mild cytotoxic effects of NA-AAF in XP-C cells when compared with XP cells belonging to complementation
group A and compared with CS cells. Our data indicate that repair of dG-C8-AF in active genes by the two NER subpathways is not simply
additive, and suggest that those lesions which are removed by the transcription-coupled repair process are also an efficient substrate for the global
genome repair.
MATERIALS AND METHODS
Cell lines and culture conditions
Primary fibroblasts from a normal individual, VH25D, and primary xeroderma pigmentosum fibroblasts, XP21RO (complementation group C) and XP25RO (complementation group A) were cultured at 37oC in 2.5% CO
2
atmosphere using Ham's F10 medium (without hypoxanthine and thymidine)
supplemented with 15% Fetal Bovine Serum and antibiotics. In experiments aimed
to determine the initial frequency and the removal of adducts in defined genomic sequences, exponentially growing cells were prelabelled for 2 days with [
3
H]thymidine (0.06 [mu]Ci/ml, 82 Ci/mmol). For measurements of repair replication or RNA synthesis,
exponentially growing cells were prelabelled with [
32
P]orthophosphate (0.3 [mu]Ci/ml) or [
14
C]uridine (0.03 [mu]Ci/ml, 60 mCi/mmol), respectively.
Cell survival studies
Cell survival following treatment with NA-AAF was determined by measuring the colony forming ability of the treated cells relative to the untreated control. 500-1000 cells were seeded in 94 mm Petri dishes, allowed to attach for
16 h, and incubated with 0, 5, 10, 15, 20 or 30 [mu]M NA-AAF for 30 min at 37oC in complete medium. After incubation fresh medium was added to
the cells and 10-14 days after plating colonies were stained with methylene blue.
HPLC analysis of NA-AAF-induced DNA adducts
Confluent cells were incubated with 5 [mu]M of [
3
H]NA-AAF (649 mCi/mmol) for 30 min at 37oC, and immediately lysed in 150 mM NaCl, 10 mM Tris-HCl pH 8.0, 1 mM EDTA, 0.5% SDS and 100 [mu]g/ml proteinase K at 37oC overnight. The DNA was purified by phenol
extraction, RNase A-treated and the specific activity (d.p.m./[mu]g) of DNA was determined by scintillation counting in Plasmasol. HPLC analysis was performed as described previously (
12
).
Measurement of RNA synthesis, DNA repair replication and unscheduled DNA
synthesis
RNA synthesis measurements, DNA repair replication and unscheduled DNA synthesis
(UDS) were performed as described by van Oosterwijk
et al
. (
12
).
[alpha]
-Amanitine studies
Cell cultures were incubated for 5 h with [alpha]-amanitine (1 [mu]g/ml) prior to treatment with 90 [mu]M NA-AAF. After NA-AAF treatment, cells were washed twice with PBS
and medium containing [alpha]-amanitine was added to the cells. To allow repair, cells were
incubated for 24 h in this medium supplemented with [
3
H]thymidine, 5 [mu]Ci/ml (82 Ci/mmol) and BUdR/FUdR (10 [mu]M/1 [mu]M). To determine the amount of incorporated radioactivity the same
protocol as previously described for repair replication measurements was
followed (
12
). As a control, cells were treated with only [alpha]-amanitine or NA-AAF, or not treated at all.
DNA probes
Double-stranded DNA probes were radioactively labelled with [
32
P]dATP by random primer extension (
14
). Strand-specific single-stranded probes were radioactively labelled with [
32
P]dATP by a linear polymerase chain reaction, using a single primer recognizing
specifically one strand (
15
).
Determination of NA-AAF-induced DNA adducts in defined sequences
In order to measure DNA adduct frequencies in defined genomic sequences as a
function of dose, [
3
H]thymidine prelabelled confluent cells were incubated with 0, 5, 10 or 15 [mu]M NA-AAF in complete medium for 30 min at 37oC. After washing the cells twice with PBS, they were lysed. In repair experiments
3
H- prelabelled confluent cells were incubated with 5 [mu]M NA-AAF in complete medium for 30 min at 37oC, and either lysed immediately or incubated for up to 48 h
in complete medium supplemented with BUdR/FUdR (10 [mu]M/1 [mu]M).
DNA was isolated and purified by phenol-chloroform extractions and digested with restriction enzymes (
1
,
12
). The frequency of NA-AAF-induced DNA-adducts per restriction fragment was determined by incision at
the sites of adducts employing the UvrABC excinuclease complex of
Escherichia coli
as described previously (
6
,
12
).
Determination of BUdR-labelled repair patches in active and inactive genes
The distribution of BUdR-labelled repair sites in restriction fragments of genes was studied by an
immunochemical method previously described (
16
). The method is based on the separation of restriction fragments containing
BUdR-labelled repair patches from non-BUdR containing DNA using specific antibodies. Aliquots of 5 [mu]M NA-AAF treated cells were incubated for different post-incubation periods (0, 2, 4, 8 and 24 h) in medium
containing BUdR/FUdR (10 [mu]M/1 [mu]M). DNA was isolated, restricted, and caesium chloride gradients were
performed twice to separate parental DNA from replicated DNA. The parental DNA
was denaturated and subsequently incubated with a monoclonal antibody raised against BUdR in DNA, a biotinylated Goat-anti-Mouse antibody (G[alpha]Mbio) and polymeric magnetic particles coated with
streptavidin (Dynabeads M-280, Dynal). The bound DNA (repaired, BUdR containing) was separated from
the unbound DNA (unrepaired) using a magnetic particle concentrator (MPC-E, Dynal), and both fractions were applied to Hybond N
+
using a slotblot apparatus (Schleicher and Schuell). The blots were hybridized
with gene-specific
32
P-labelled probes and the amount of
32
P in the bound and unbound fractions was determined using a Betascope 603 blot
analyser (Betagen Corp.).
RESULTS
Cytotoxic effects of NA-AAF treatment
To determine the cytotoxic effects of NA-AAF on cells after a 30 min exposure, we measured the colony forming ability relative to untreated
cells. Figure
1
shows that the xeroderma pigmentosum cell line XP21RO (complementation group C)
is only slightly more sensitive to NA-AAF treatment than the normal human VH25D cell line, in contrast with
XP25RO cells (complementation group A). The survival of XP21RO cells following
NA-AAF treatment is also better when compared with the Cockayne syndrome cell
line CS3BE (complementation group A).
Formation of DNA adducts
HPLC analysis of DNA obtained from confluent cells treated with
3
H-labelled NA-AAF revealed that the distribution patterns of radioactivity in the
chromatograms were very similar for VH25D and XP21RO, and that the major lesion
induced in both cell lines after NA-AAF treatment is the deacetylated dG-C8-AF lesion with a frequency of 0.13 and 0.20 NA-AAF lesions/[mu]M/ 10 kb for normal and XP21RO cells respectively.
Effects of NA-AAF treatment on RNA synthesis
Since a stalled transcription complex is considered to be an essential
intermediate in transcription-coupled repair, we investigated the inhibition of RNA synthesis by NA-AAF treatment. Figure
2
A shows that RNA synthesis measured during a 30 min pulse immediately after
treatment with NA-AAF is inhibited in a dose-dependent way and the extent of inhibition is the same in normal and
XP-C cells. Next we determined the capabilities of these two cell lines as
well as XP-A cells to recover from NA-AAF-inhibited RNA synthesis. Figure
2
B shows that both normal and XP-C cells, after treatment with 15 [mu]M NA-AAF, are capable of recovering the inhibited RNA synthesis with
the same kinetics. This in contrast with XP-A cells, which are not able to recover RNA synthesis after NA-AAF treatment of 15 [mu]M.
Repair replication and UDS measurement
To assess the genome overall repair of NA-AAF induced DNA adducts in normal and XP-C cells, we measured the level of DNA repair replication.
32
P-prelabelled cells were treated with 30, 45 or 90 [mu]M NA-AAF and were allowed to carry out repair for 24 h in the
presence of [
3
H]thymidine, BUdR and FUdR. The results, shown in Figure
3
, indicate that the extent of repair synthesis measured over a period of 24 h
following NA-AAF treatment is substantially lower in XP-C cells when compared with normal cells.
DISCUSSION
Primary XP-C fibroblasts appear to be somewhat more sensitive to the cytotoxic
effects of NA-AAF than normal cells but less sensitive when compared with XP-A fibroblasts or CS-A and CS-B fibroblasts (
12
). This differential sensitivity of the various human fibroblast strains cannot
be attributed to variations in frequencies or types of DNA lesions induced by
NA-AAF exposure. The results of the current study and a previous
investigation by Amacher and Lieberman (
20
) demonstrate that in primary normal human fibroblasts, as well as in XP and CS
cells, NA-AAF induces dG-C8-AF as the major lesion.
The relatively mild cytotoxic effect of NA-AAF in XP-C cells is in contrast with the low level of genome overall repair
of NA-AAF induced DNA lesions in these cells when compared with normal human
cells. This reduced repair of NA-AAF adducts in XP-C cells resembles the poor genome overall repair of UV-induced photolesions in these cells (
1
,
5
). Analogous to repair of UV-damage, the residual repair of NA-AAF induced DNA damage in XP-C cells is confined to transcriptionally active genes, because
treatment of XP-C cells with [alpha]-amanitin, an inhibitor of RNA polymerase II, severely
inhibited NA-AAF induced repair replication, whereas the inhibitor had no inhibitory
effect on repair replication in normal cells. A similar inhibitory effect of [alpha]-amanitine on repair synthesis in UV-irradiated confluent XP-C cells has been reported by Carreau and Hunting (
17
). Thus our data indicate that removal of NA-AAF induced DNA lesions in XP-C cells depends on transcription and that dG-C8-AF lesions are actually targets for transcription-coupled repair.
ACKNOWLEDGEMENTS
The UvrABC proteins used in this study were obtained from Dr P. van de Putte,
Department of Molecular Genetics, Leiden Institute of Chemical Research, Leiden
University, The Netherlands. The UDS measurements were performed by Dr K.
Jaspers, MGC-Department of Cell Biology and Genetics, Erasmus University, Rotterdam,
The Netherlands. We are grateful to Drs S. Bol and J. H. N. Meerman for help in
performing the HPLC analysis. This study was supported by a grant of the
Medical Genetics Centre South-West Netherlands and the European Commission (contract EV5V-CT94-0397).
REFERENCES
1 Venema,J., van Hoffen,A., Natarajan,A.T., van Zeeland,A.A. and Mullenders,L.H.F. (1990) Nucleic Acids Res.18, 443-448.MEDLINE Abstract
2 Mellon,I., Spivak.G. and Hanawalt,P.C. (1987) Cell 51, 241-249.MEDLINE Abstract
3 Venema,J., van Hoffen,A., Karcagi,V., Natarajan,A.T. van Zeeland,A.A. and Mullenders,L.H.F. (1991) Mol. Cell. Biol.11, 4128-4134.MEDLINE Abstract
4 Venema,J.A., Bartosova,Z., Natarajan,A.T., van Zeeland,A.A. and Mullenders,L.H.F. (1992) J. Biol. Chem.267, 8852-8856.
5 Kantor,G.J., Barsalou,L.S. and Hanawalt,P.C. (1990) Mutat. Res.235, 171-180.MEDLINE Abstract
6 Van Hoffen,A., Venema,J., Meschini,R., van Zeeland,A.A. and Mullenders,L.H.F. (1995) EMBO J.14, 360-367.MEDLINE Abstract
7 Regan,J.D. and Setlow,R.B. (1974) Cancer Res.34, 3318-3325.MEDLINE Abstract
8 Maher,V.M., Curren,R.D., Ouelette,L.M. and McGormick,J.J. (1976) In de Serres,F.J., Fouts,J.R, Bend,J.R and Philpot,R.M. (eds), In Vitro Metabolic Activation in Mutagenesis Testing. Elsevier, North-Holland, pp. 313-336.
9 Brown,A.J., Fickel,T.H., Cleaver,J.E., Lohman,P.H.M., Wade,M.H. and Waters,R. (1979) Cancer Res.39, 2522-2527.MEDLINE Abstract
*To whom correspondence should be addressed at: MGC-Department of Radiation Genetics and Chemical Mutagenesis, Leiden
University, Wassenaarseweg 72, 2333 AL Leiden, The Netherlands. Tel: +31 71
5276126; Fax: +31 71 5221615; Email: mullenders@rullf2.medfac.leidenuniv.nl
