ABSTRACT
In this study the role of nuclear architecture in nucleotide excision repair
(NER) was investigated by gentle dismantling of the cell and probing the
capability of chromatin to carry out repair
in vitro
. The rationale behind this approach is that compartmentalization of NER at
nuclear structures would make the enzymatic activities refractory to extraction
by buffers that solubilize cellular membranes. In order to obtain intact
chromatin primary human fibroblasts were encapsulated in agarose microbeads and
lysed in isotonic buffers containing the non-ionic detergent Triton X-100. Under these conditions the majority of cellular proteins
diffuse out of the beads, but the remaining chromatin is able to replicate and
to transcribe DNA in the presence of triphosphates and Mg
2+
. UV irradiation of confluent repair-proficient human fibroblasts prior to lysis stimulated the incorporation
of deoxynucleotide triphosphates
in Triton X-100-isolated chromatin, even under stringent lysis conditions. In
addition, experiments with UV-sensitive xeroderma pigmentosum (complementation groups A and C) and
Cockayne's syndrome fibroblasts (complementation group A) revealed that this
repair synthesis was due to global genome repair activity. Transcription-coupled repair was only detectable in cells permeabilized by streptolysin
O (SLO). Repair synthesis in Triton X-100-isolated chromatin amounted to 15% of the total repair synthesis as
measured in SLO-permeabilized cells. To allow the detection of these activities
in vitro
, presynthesis complexes have to be formed in intact cells, indicating that
chromatin from Triton X-100-lysed cells is unable to initiate NER
in vitro
. Our data indicate that the components involved in the resynthesis step of NER
are tightly associated with chromatin. A substantial fraction of total
proliferating cell nuclear antigen (PCNA), which is required for the
resynthesis step in NER, has been reported to become Triton X-100 non-extractable and tightly associated with nuclear structures after UV
irradiation of cells. We propose that Triton X-100-resistant repair synthesis might be mediated by this chromatin-bound fraction of total PCNA.
There is firm evidence that repair of UV-induced photolesions is heterogeneous across the genome and that several
hierarchies of DNA repair in mammalian cells exist (
1
,
2
). Based on the efficiency of repair of UV-induced cyclobutane pyrimidine dimers in various genomic regions,
different levels of nucleotide excision repair (NER) can be distinguished: (i)
slow repair of ribosomal genes and transcriptionally inactive tissue-specific genes; (ii) fast and efficient repair of (potentially) active
genes; (iii) accelerated repair of the transcribed strand of active genes which
involves a close coupling of repair to RNA polymerase II-driven transcription (transcription-coupled repair). It is obvious that transcription is not the sole
determinant governing the accelerated repair of active genes. This is indicated
by the fact that in the absence of transcription, (potentially) active genes
are still more efficiently repaired than inactive genes (
3
). This suggests that other factors, such as chromatin structure, play a role in
determining the efficiencies of NER in various parts of the mammalian genome.
It has become clear that nuclear functions such as DNA replication and
transcription are compartmentalized within the nucleus and that components
involved in these processes are associated with a nuclear skeleton termed the
nuclear matrix (
4
-
6
). One way to accomplish efficient repair of active genes is to concentrate
repair enzymes at the nuclear matrix in proximity to the attached transcription
machinery and active genes (
7
). Indeed, investigations into the localization of repair of UV-induced DNA damage have revealed evidence for the association of NER with
the nuclear matrix. Analysis of the distribution of repair sites in UV-irradiated human cells has indicated that DNA sequences at the attachment
sites of DNA loops to the nuclear matrix are the primary target for initial repair activity (
8
). Additionally, in cells expressing only transcription-coupled repair (i.e. xeroderma pigmentosum group C cells), repair is
confined to DNA sequences located proximal to the nuclear matrix and is
virtually absent in loop DNA (
9
). This preferential localization of UV-induced repair sites at the nuclear matrix in human cells correlates well
with the ability to perform preferential repair of cyclobutane pyrimidine
dimers (CPD) in (potentially) transcriptionally active DNA. Recently it was shown by Park
et al.
(
10
) that the XPG endonuclease, which is an enzyme essential for NER, is tightly
associated with nuclear structures identical to the nuclear matrix.
The finding that the products of several repair genes (xeroderma pigmentosum
groups B and D, trychothiodystrophy group A) are components of the human transcription factor TFIIH (required for the initiation
of class II gene transcription;
11
) may be additional evidence for nuclear architecture playing a role in DNA
repair. Recent studies have indicated that activities involved in transcription
are localized in discrete domains rather than diffusely distributed (
12
-
14
). Biochemical studies have provided evidence that RNA polymerase II activity as
well as nascent RNA are attached to the nuclear matrix (
15
,
16
) and that transcriptionally active genes are attached via specific nuclear matrix attachment regions (MARs)
(
17
,
18
). Thus the compartmentalization of transcription activity at the nuclear matrix
could have a major impact on the distribution of repair activity within the
nucleus and could influence repair rates across the genome.
One step forward to elucidate the role of the nuclear matrix in DNA repair is to
demonstrate that enzymatic activities involved in NER are associated with
chromatin. In order to address this question experimentally, it is essential to
measure repair in chromatin isolated under conditions which preserve structure
and function. Cook and co-workers have developed a method of chromatin preparation that fulfils
these requirements as much as possible (
19
). In their approach cells encapsulated in agarose microbeads are permeabilized
under ionic conditions that mimic the cytoplasm. Chromatin prepared in this way
contains intact DNA loops associated with the nuclear skeleton and is able to
replicate (in an S phase-dependent manner) and to transcribe.
In this study we have investigated whether NER activity is associated with
chromatin prepared from UV-irradiated normal human fibroblasts and UV-sensitive XP and CS fibroblasts. Our results indicate that enzyme
activities involved in the DNA resynthesis step of NER are tightly attached to
chromatin even when cellular and nuclear membranes are solubilized and the
majority of cellular proteins are removed. The Triton X-100 (Triton)-resistant repair synthesis amounted to 15% of total repair synthesis
when compared with cells mildly permeabilized by the bacterial cytotoxin
streptolysin O (SLO). This fraction of total repair activity closely parallels the fraction of total proliferating cell
nuclear antigen (PCNA), a protein required for the resynthesis step in NER (
20
), that has been reported to become Triton resistant and tightly associated with
nuclear structures after UV irradiation of cells (
21
,
22
). We propose that Triton-resistant repair synthesis might be mediated by this fraction of total
PCNA.
Primary human fibroblasts were grown in Ham's F10 medium supplemented with 15%
fetal calf serum and antibiotics in a 2.5% CO
2
atmosphere. To prelabel the DNA the cells were incubated for 3 days in the
presence of 0.1 [mu]Ci/ml [
3
H]thymidine (82 Ci/mmol). After 3 days the medium was replaced by fresh medium
and the cells were grown to confluence. To label proteins, cells were grown for
24 h in the presence of 1 [mu]Ci/ml [
35
S]methionine (1000 Ci/mmol) in methionine-depleted medium.
The following primary fibroblast cell lines were used: VH16 and VH25, derived
from two normal individuals; XP1TE and XP21RO, derived from xeroderma
pigmentosum complementation group C patients (XP-C); XP25RO and XP30RO derived from xeroderma pigmentosum group A (XP-A) and variant (XP-variant) patients respectively; CS3BE derived from a
Cockayne's syndrome patient belonging to complementation group A. In some of
the experiments [alpha]-amanitin (final concentration 20 [mu]g/ml) was added to confluent fibroblasts and the cells were
kept in this medium for 16 h. SV40-transformed normal human fibroblasts (MRC-5) were grown in Ham's F10 medium supplemented with 10% newborn calf
serum and antibiotics in a 5% CO
2
atmosphere and prelabelled under the same conditions as primary cells.
Exponentially growing MRC-5 cells were encapsulated in agarose microbeads to a final concentration
of 3-5 * 10
6
cells/ml beads and subsequently lysed essentially as described by Jackson
et al
. (
19
). In each experiment a part of the cells was not encapsulated, but lysed in
0.5% SDS and the specific activity of the cells (
3
H counts/10
6
cells) was determined by scintillation counting. Encapsulated cells were
incubated in complete medium for 1 h, washed with PBS and lysed in a
physiological buffer (PB) (
19
) containing 100 mM KH
2
PO
4
, 130 mM KCl, 10 mM Na
2
HPO
4
, 1 mM MgCl
2
, 1 mM Na
2
ATP (Sigma type II), 1 mM DTT, pH 7.4, supplemented with 0.5% Triton X-100. Lysis was performed by keeping the microbeads at 4oC for 15 min. In some experiments the lysis step was repeated twice.
Subsequently the beads were washed four times at 4oC with large volumes of PB without Triton.
Primary fibroblasts were grown to confluence and encapsulated at a concentration
of 3-5 * 10
6
cells/ml beads. The cells were grown for 1 h in medium and lysed in PB
supplemented with 0.5% Triton as described above. In several experiments SLO
(Wellcome) was used instead of Triton to mildly permeabilize the cells. The
cells were incubated in PB supplemented with 0.045 U/ml SLO for 30 min on ice.
Unbound SLO was removed by washing twice with large volumes of PB (4oC). Subsequently, the beads were incubated for 2 min at 37oC to permeabilize the cellular membranes and again washed twice with
PB (4oC).
To determine the amount of protein removed by cell lysis and subsequent washing
and centrifugation, aliquots of supernatant and beads (lysed completely by
incubation with 1% SDS for 90 min at 37oC) were taken and TCA was added to 10%. Precipitates were collected,
dissolved in 0.2 N NaOH and processed for liquid scintillation counting.
Prior to UV irradiation, encapsulated cells were kept in medium for 1 h, washed
twice with PBS and resuspended in 3 vol PBS. Irradiation was performed with a
Philips TUV lamp (predominantly 254 nm) at a dose rate of 0.2 J/m
2
/s; the effective dose of irradiation of encapsulated cells was ~50% as deduced from frequencies of cyclobutane pyrimidine dimers induced in
cells encapsulated in microbeads and in monolayer cells in Petri dishes. The
cells were post-UV incubated in PBS or in complete medium at 37oC for 30 min, washed with PB and lysed as described above. In every
experiment part of the cells were mock irradiated and processed like the
irradiated cells.
For repair and replication experiments a 10* concentrated nucleotide mix was prepared, which contained 2.5 mM dGTP,
dCTP, dTTP, 1.0 mM GTP, CTP, UTP, 10 mM ATP, 50 mM KPO
4
, pH 7.4, 25 mM MgCl
2
and 400 [mu]Ci/ml [
32
P]dATP (3000 Ci/mmol). For transcription experiments the 10* concentrated nucleotide mix (transcription mix) contained 2.5 mM GTP,
2.5 mM CTP, 20 [mu]M UTP, 10 mM ATP, 50 mM KPO
4
, pH 7.4, 25 mM MgCl
2
and 400 [mu]Ci/ml [
32
P]UTP (3000 Ci/mmol). In several of the experiments the transcription mix was
supplemented with dNTPs to a final concentration of 25 or 250 [mu]M for each nucleotide in the transcription reaction, thus resembling the
conditions for measurement of repair replication. The effect of [alpha]-amanitin (final concentration 20 [mu]g/ml) was investigated either by treatment of intact cells with
the inhibitor or by addition of [alpha]-amanitin to the transcription mix only. Aliquots of 0.9 vol lysed
and washed cells in microbeads and 0.1 vol nucleotide mix were kept at 37oC for 2 min and subsequently mixed. At various time intervals aliquots were
taken, washed four times with excess ice-cold physiological buffer and lysed completely by incubation with 1% SDS
for 90 min at 37oC. TCA was added to 10% and precipitates were collected on glass fibre
filters and processed for liquid scintillation counting.
Prior to encapsulation, exponentially growing
3
H-labelled MRC-5 cells were incubated for 1 h in medium containing 10 [mu]M bromodeoxyuridine (BrdU) and 1 [mu]M fluorodeoxyuridine (FdU). Addition of BrdU and FdU to the
medium enables the replicated DNA to be separated from the parental DNA in CsCl
density gradients. The cells were encapsulated, lysed and incubated with
nucleotides as described above, but with 2.5 [mu]M BrdUTP instead of dTTP. After the reaction the beads were incubated with
Hae
III (500 U/ml) for 1 h at 37oC in the appropriate buffer, followed by complete lysis of the chromatin
with 0.1% sarcosyl and 100 [mu]g/ml proteinase K at 37oC and isolation of DNA by electroelution. The purified DNA was
centrifuged to equilibrium in CsCl density gradients. The gradients were
fractionated, TCA precipitated and radioactivity of acid-insoluble material present in each fraction was determined.
For size analysis of newly synthesized DNA, permeabilized encapsulated cells
were incubated with nucleotides, lysed with 1% SDS and subjected to
electrophoresis in 0.6% agarose gels equilibrated with a buffer containing 0.03
N NaOH and 1 mM EDTA. The lanes containing DNA were sliced, the gel slices were
melted in 1 N HCl at 90oC, mixed with scintillation solution and the radioactivity measured.
Encapsulated cells were UV irradiated (15 J/m
2
), permeabilized in Triton-containing PB as described above and resuspended in PB supplemented with 5
mM EDTA. The chromatin was treated or mock treated with the CPD-specific enzyme T4 endonuclease V for 15 min at 37oC (
23
) and the beads were extensively washed with cold PB containing 5 mM EDTA.
Subsequently, the treated and mock-treated samples were incubated with nucleotide mix at 37oC and incorporation of labelled nucleotides was measured in samples
taken at various incubation times up to 30 min. Aliquots were taken to confirm
the presence of T4 endonuclease V-induced incisions by alkaline gel electrophoresis.
In order to isolate intact and functionally active chromatin we employed cells
encapsulated in agarose microbeads as described by Jackson and Cook (
19
,
24
). Basically the cells are mixed with low gelling agarose to achieve
encapsulation and lysed under isotonic conditions to obtain chromatin. The
chromatin is protected by the agarose coat from shearing and aggregation, but
remains fully accessible to macromolecules up to 1.5 * 10
2
kDa. Moreover, the chromatin has been reported to be able to perform DNA
metabolic activities, such as replication and transcription.
The aim of our study was to address the question of whether this chromatin is
capable of performing NER and which of the NER repair subpathways, i.e.
transcription-coupled repair or/and global genome repair, are involved. Therefore, we
utilized confluent primary normal human fibroblasts as well as fibroblasts from
patients suffering from xeroderma pigmentosum (XP) and Cockayne's syndrome
(CS). First we established lysis conditions and replication activities in
exponentially growing immortalized cells as previously described (
24
) and subsequently applied the methodology to confluent or exponentially growing
primary cells. After encapsulation in agarose microbeads, immortalized as well
as primary cells remained viable and were able to grow in the presence of
culture medium. The latter was tested by growing the cells in medium containing
BrdU for 24 h and immunostaining with antibodies against BrdU-containing DNA. Lysis of immortalized or primary cells in PB supplemented
with 0.5% Triton resulted in release of 80% of the [
35
S]methionine-labelled cellular proteins. Repeated incubations with 0.5% Triton-containing buffer did not lead to further release of proteins. Based
on trypan blue staining, all cells became permeable. Permeabilization of
encapsulated cells by SLO (
25
) resulted in release of ~5% of the cellular proteins leaving the cellular membrane fully intact.
The capability of performing DNA replication was determined in chromatin
isolated from exponentially growing MRC-5 and primary cells by incubation in PB supplemented with nucleotides
(including [
32
P]dATP), Mg
2+
and ATP at 37oC. The replication activity, measured as the incorporation of [
32
P]dATP into acid-precipitable counts, was fully dependent on the presence of Mg
2+
and ATP in the reaction mix and was inhibited by aphidicolin. Both primary and
immortalized cells exhibited increased incorporation of [
32
P]dATP with time. Figure
1
A shows the results obtained with immortalized cells. However, the efficiency of
the reaction was strongly dependent on the dATP concentration of the reaction
mixture: the incorporation of nucleotides was strongly enhanced by an increase
in the dATP concentration in the reaction mix, indicating that submicromolar
concentrations of dATP are rate limiting for the replication reaction (data not
shown). To prove the validity of the system with regard to replicative
synthesis, chromatin derived from exponentially growing primary fibroblasts was
incubated with replication mix containing BrdUTP, cut with restriction enzymes
and lysed completely with sarcosyl and proteinase K. Subsequently, the newly
synthesized DNA was analysed by neutral CsCl density gradient centrifugation.
As shown in Figure
1
B, the
32
P-labelled newly synthesized DNA strands were shifted to higher density in
the gradient compared with the non-replicated DNA fraction, suggesting that efficient replication takes place
in chromatin. To investigate the efficiency of DNA synthesis in more detail,
samples were taken at various times after addition of the replication mix to
the chromatin and run in an alkaline agarose gel; lanes containing DNA were
sliced and the amount of radioactivity in each slice was determined by liquid
scintillation counting. The results indicated that the newly synthesized
Okazaki fragments are very rapidly ligated into large DNA molecules during the
course of the reaction (data not shown). When encapsulated, immortalized cells
were irradiated with UV (effective dose 15 J/m
2
) prior to permeabilization, a clear inhibition of replication activity was
observed (Fig.
1
A).
To study repair of DNA damage by NER in intact chromatin, stationary primary
fibroblasts prelabelled with [
3
H]thymidine were encapsulated and irradiated with an effective dose of 15 J/m
2
UV in PBS or mock irradiated. The cells were left at room temperature for
variable time periods, permeabilized with either Triton or SLO, washed with PB
and incubated with nucleotides (including [
32
P]dATP), Mg
2+
and ATP. At various times after the start of the incubation, aliquots were
taken and the amount of incorporated label per 10
6
cells was determined.
When stationary primary normal fibroblasts were encapsulated, UV irradiated and
lysed immediately in Triton-containing PB, no effect of UV on incorporation of
32
P-labelled nucleotides was observed. However, when the cells were irradiated
and incubated either at room temperature or at 37oC in PBS prior to lysis, incorporation of [
32
P]dATP was clearly enhanced in irradiated cells compared with non-irradiated cells. This stimulation of incorporation reached a maximum when
chromatin was prepared from encapsulated cells 30 min after UV irradiation
(Fig.
4
). Incubations of encapsulated and UV-irradiated cells for longer periods than 30 min did not lead to a further
stimulation of incorporation. Based on these observations, all subsequent
experiments were performed with UV-irradiated cells kept at 37oC for 30 min prior to lysis to prime the repair reaction. Under these
conditions, chromatin from UV-irradiated and Triton-lysed cells is able to incorporate [
32
P]dATP up to 60 min following the start of incubation. However, the majority of
incorporation takes place during the first 20 min of the reaction. After the
initial 20 min incorporation in chromatin from UV-irradiated cells continues, albeit at a slower rate, but the rate of
incorporation clearly exceeds that in chromatin prepared from unirradiated
cells. To confirm that UV-stimulated incorporation of [
32
P]dATP was due to repair synthesis, we performed
in vitro
repair reactions in the presence of BrdU and analysed the DNA by CsCl density
gradient centrifugation. All label incorporated in chromatin from UV-irradiated cells appeared to be present in the parental DNA fraction (Fig.
4
B). Since the absolute levels of incorporation varied in a series of
experiments, the extent of repair incorporation in the various experiments was
quantified by expressing the incorporated label in UV-irradiated cells relative to that in non-irradiated cells. This ratio is termed the UV stimulation factor
(UVst).
In mammalian cells UV-induced photolesions are processed by two NER subpathways, i.e. the global
genome repair pathway and the transcription-coupled repair pathway. The latter removes DNA photolesions from the
transcribed strand of active genes and depends on active transcription. We
addressed the question whether
in vitro
repair synthesis was due to global genome repair or to transcription-coupled repair by investigating repair synthesis in XP-C and CS fibroblast strains that can only carry out transcription-coupled repair or global genome repair respectively (
7
,
26
).
As shown in Figure
6
A, in normal human cell lines (VH16 and VH25) the UVst after Triton lysis
amounted to ~4.0-4.5 (mean of five and six experiments respectively). UV irradiation
did not have any effect on incorporation of label in XP-A cells (completely defective in NER), providing additional proof that
incorporation of [
32
P]dATP into chromatin is due to DNA repair synthesis. Also, chromatin from
Triton-lysed XP-C cells did not exhibit any stimulation of repair by UV, suggesting
that the repair activity detected in chromatin from normal cells cannot be
attributed to transcription-coupled repair. This conclusion is in line with the observation that the
extent of repair synthesis in chromatin derived from CS-A cells was very similar to normal human cells. Taken together, the
results indicate that the vast majority of repair synthesis in chromatin
prepared from UV-irradiated human fibroblasts is due to global genome repair.
In this study we approached the role of nuclear architecture in NER by gently
dismantling the cell and probing the capability of chromatin to carry out repair
in vitro
. The rationale behind this approach is that compartmentalization of NER at the
nuclear skeleton would, by analogy with replication and transcription, make the
enzymatic activities refractory to extraction by buffers that solubilize
cellular membranes. In order to obtain chromatin with preserved cellular
functions, we adapted the cell-free system developed by Jackson
et al
. (
6
), which has previously been used to study replication and transcription in HeLa
cells (
19
). Chromatin from these cells isolated with Triton in a physiological buffer
exhibited efficient replication and transcription approaching, under optimal
conditions, the rate
in vivo
; however, UV-induced nucleotide excision repair could not be measured (
25
).
Primary human cells resemble HeLa cells with regard to replication and
transcription capacities of chromatin. After Triton extraction 80% of the
cellular proteins diffuse out of the beads but chromatin prepared from
exponentially growing fibroblasts still synthesizes DNA and RNA in the presence
of magnesium, triphosphates and ATP. To allow a direct comparison between
replication and repair synthesis, replication assays were performed under
suboptimal conditions, i.e. in the presence of micromolar concentrations of
labelled deoxynucleotide triphosphate. Obviously, at this low triphosphate
concentration replicative synthesis in Triton-lysed cells is much less efficient than in more intact cells, as
demonstrated by the 5-fold difference in incorporation between the SLO- and Triton-permeabilized cells. This difference in replication efficiency
has also been observed for HeLa cells when replication was compared in cells
either lysed by Triton or permeabilized by complement and assayed under
conditions of suboptimal deoxyribonucleotide triphosphate concentration (
19
).
When exponentially growing cells were UV irradiated prior to Triton lysis,
replicative synthesis in chromatin was almost completely inhibited, reflecting
the situation in intact cells. However, the remaining replication activity was
still too high to reveal any repair synthesis. In contrast, UV irradiation of
confluent fibroblasts clearly stimulated incorporation of deoxynucleotide
triphophates. Similarly to replicative synthesis, this UV-stimulated incorporation depended on magnesium and triphosphates,
including ATP. Like replicative synthesis, UV-stimulated incorporation was reduced in Triton-isolated chromatin when compared with SLO-permeabilized cells, but clearly detectable. This is in
contrast to HeLa cells, in which UV-stimulated DNA synthesis could be detected in SLO-permeabilized cells, but not in Triton-isolated chromatin (
25
). By several criteria the UV-stimulated incorporation in primary fibroblasts was shown to be DNA repair
synthesis: firstly, XP-A cells, which are completely deficient in NER, do not show this
stimulation; secondly, repair label is recovered in parental DNA in CsCl density gradients in contrast to label incorporated by replicative synthesis. A
substantial fraction of the repair synthesis (~15% of the repair synthesis in SLO-lysed cells) is associated with chromatin even under stringent lysis
conditions. Repair synthesis in Triton-isolated chromatin cannot be stimulated by incubation with the CPD-specific enzyme T4 endonuclease V, in contrast to several permeable
cell systems in which a clear stimulation was found (
27
). This suggests that repair components retained in chromatin are tightly
associated with sites of repair halted at lysis and are incapable of starting
DNA repair synthesis by diffusion to sites incised
in vitro
. Interestingly, other studies have shown that proteins involved in the
resynthesis step of NER are tightly bound to chromatin. PCNA is the auxiliary
protein of polymerases [delta] and [epsilon] and required for DNA replication and NER (
20
,
29
-
31
). UV irradiation of quiescent human fibroblasts triggers the appearance of PCNA
in chromatin, visualized by immunofluorescent staining (
21
,
30
-
32
) and part of the total PCNA becomes resistant to extraction by Triton (
22
) or high salt (
21
). Consistent with these observations we found, by Western blotting analysis,
that a major fraction of PCNA associates with Triton-isolated chromatin in UV-irradiated normal fibroblasts, whereas PCNA is virtually absent in
chromatin prepared from unirradiated cells. In contrast to this, in HeLa cells
UV-induced binding of PCNA to chromatin could not be demonstrated
(unpublished results), which might underlie the absence of Triton-resistant repair synthesis in these cells (
25
). Thus, the possibility emerges that Triton-resistant repair synthesis is mediated by a fraction of total PCNA that
becomes tightly associated with the nucleus after UV irradiation. In this
respect the requisite to leave the irradiated cells intact for a period of time
before permeabilization in order to detect
in vitro
repair synthesis might be related to damage-dependent relocation of PCNA to the nucleus. Since different types of PCNA
complexes might exist (
31
), a specific form of PCNA may associate with a subfraction of chromatin or with
nuclear structures, possibly the nuclear matrix (
33
).
By employing repair replication as a measure of repair we were unable to detect
initiation of repair in Triton-lysed chromatin. The most plausible explanation is that factors involved
in the initiation of NER are free in the nucleoplasm and extracted by Triton.
However, even in cells mildly permeabilized by SLO, initiation of repair was
virtually absent, suggesting that the damage recognition and incision steps of
NER are very sensitive to small perturbations of the cellular membrane. In
contrast to Triton-lysed XP-C cells, SLO-permeabilized XP-C cells showed a 2-fold increase in incorporation following UV
irradiation. However, it is evident that this repair depends also on halted
repair events
in vivo
, as inhibition of
in vitro
transcription by [alpha]-amanitin did not affect repair synthesis. If Triton affects the
efficiency of transcription-coupled repair to a similar extent as global genome repair (i.e. 80% loss
of activity after lysis with 0.5% Triton), repair synthesis mediated by the
transcription-coupled repair pathway would be below the detection level.
Taken together, our results suggest that a substantial fraction of the enzymatic
activities involved in the resynthesis step of NER is tightly attached to
chromatin. However, presynthesis complexes have to be formed in intact cells to
allow detection of these activities
in vitro
. So far, repair of UV-induced lesions in chromatin under cell-free conditions has only been achieved with SV40 chromosomes (
34
), but in this system repair was only efficient in the presence of naked plasmid
DNA. Combining nuclear extracts with the chromatin template described in this
study provides the opportunity at the biochemical and immunochemical level to
characterize the role of chromatin structure in NER and to isolate proteins
essential for processing DNA damage in chromatin.
We are grateful to Dr P.R.Cook for his help in preparing chromatin. We thank Dr
R.M.A.Mason for critical reading of the manuscript. This research was supported
by the association of the Leiden University with Euratom, the European Science
Foundation by a fellowship to K.Bouayadi and a NATO Collaborative Research
Grant to P.R.Cook.
*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
Present addresses:
+
Centre National de la Recherche Scientifique, Laboratoire de Pharmacologie et de
Toxicologie Fondamentales du CNRS, 205 Route de Narbonne, 31077 Toulouse Cedex,
France and
[sect]
Laboratory of Molecular Genetics, National Institute of Aging, NIH, 4940 Eastern
Avenue, Baltimore, MD, USA
REFERENCES
Return