| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Institute of Cell and Molecular Biology, Edinburgh University, King’s Buildings, Edinburgh, Scotland, U.K.; and
Department of Pathology, Edinburgh University, Teviot Place, Edinburgh, Scotland, U.K.
1Correspondence: Institute of Cell and Molecular Biology, Edinburgh University, Darwin Building, King’s Buildings, Mayfield Road, Edinburgh EH9 3JR, Scotland, U.K. E-mail: David.Melton{at}ed.ac.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: DNA repair • cancer • knockout mice • cell cycle arrest
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
. NER is a multi-step
process that involves damage recognition, dual incision of the damaged
strand, followed by the removal of the lesion as part of an
oligonucleotide, gap filling, and strand ligation (1
, 2)
.
Defects in the proteins involved in this mechanism are associated with
three human disorders: xeroderma pigmentosum (XP), which is
characterized by a high incidence of skin cancer in sun-exposed areas
of the body; Cockayne’s syndrome (CS); and trichothiodystrophy (TTD).
ERCC1 (excision repair cross-complementing gene 1) was the first
mammalian NER gene to be cloned by virtue of its ability to complement
a UV-sensitive Chinese hamster ovary cell line (3)
. ERCC1
failed to correct the repair defect in any of the human NER-deficient
cell lines tested (4
, 5)
. Nevertheless, the ERCC1 gene
product is known to play a critical role in NER, where it acts in a
complex with XPF to make an incision 5' to the damage site
(6)
. This led to speculation that mutations in ERCC1 may
be developmentally lethal. Alternatively, ERCC1 deficiency may be
associated with a different human phenotype not represented among the
known NER deficiency diseases. Some NER proteins, such as XPB and XPD,
have additional non-NER-related functions, and defects in these have
been linked to the neurological and other not obviously NER-related
problems, such as retarded physical or sexual development and brittle
hair, seen in some XP, CS, and TTD patients (reviewed in ref
7
). There is evidence that ERCC1 may also have other key
functions. Based on homology studies (8
, 9)
, the ERCC1/XPF
complex has been proposed to play a role in mitotic recombination as
well as in NER. In addition, many ERCC1 and ERCC4(XPF) mutants are
extremely sensitive to agents, such as mitomycin C, which cause
interstrand cross-links and are repaired by a process involving
homologous recombination (7)
. ERCC1 has also been
implicated in the processing of heteroduplex recombination
intermediates in mammalian cells (10)
. However, recent
experiments suggest that in mouse cells ERCC1 is not essential for the
recombination-mediated repair of interstrand cross-links, but instead
may have a role in S-phase-dependent chromosome exchange
(11)
.
p53 has been studied extensively in relation to DNA repair and
its role in cell cycle regulation. On the detection of DNA damage, p53
levels are elevated through an increase in the protein half-life. p53
induces the transcription of downstream genes, such as the
cyclin-dependent kinase inhibitor p21 (12)
, resulting in
cell cycle arrest at both G1/S and
G2/M checkpoints; in some cell types it can also
lead to apoptotic cell death (13
, 14)
. p53 and p21 have
also been implicated in cellular aging: p53 levels accumulate as
diploid fibroblasts age, with a corresponding increase in p21 levels,
which in turn correlate with a slowing of cellular growth (15
, 16)
. Fibroblasts from p21 null mice cannot arrest in
G1 after DNA damage (17
, 18)
, and
human p21-deficient cells show similar characteristics (16
, 19
, 20)
. From studies of these human p21-deficient cells, another
role for p21 has emerged in p53-mediated G2
arrest and coupling of DNA synthesis and mitosis (19
, 20)
.
Contrary to some expectations, the absence of ERCC1 did not
result in an embryonic lethal phenotype (21
, 22)
. However,
unlike XPA and XPC knockout mice (23
24
25)
, animals
deficient in ERCC1 do not present the typical phenotype associated with
XP. Our ERCC1 knockout mice are severely runted at birth and die before
weaning, probably due to hepatic failure, with elevated levels of p53
in their liver, brain, and kidney (21)
. The most striking
pathology was seen in the liver, where there was a dramatic increase in
hepatocyte nuclear size. FACS analysis showed that this was due to
nuclei with greater than the normal diploid (2n) content, with both
polyploid (4n, 8n) and aneuploid nuclei being present
(21)
. Increased liver ploidy is normally only seen in
older mice (see below) and increased p53 levels have been found in
senescing fibroblasts (15)
, so the phenotype in our
ERCC1-deficient mice is reminiscent of premature aging. We wanted to
understand how such a phenotype can arise in the absence of exogenous
DNA-damaging agents. A clue to the involvement of cell cycle arrest in
the ERCC1-deficient phenotype came from a report that similar liver
abnormalities occurred in mice overexpressing p21 in their liver
(26)
. We also investigated the role of p53 in the
phenotype. If elevated p53 levels are responsible for the increased DNA
content, then the ERCC1-deficient liver phenotype might be rescued in
mice deficient in both ERCC1 and p53. We carried out studies on
3-wk-old mice of the following genotypes: wild-type (wt),
ERCC1-deficient, p53-deficient, and double null. The age of the mice
for this study was determined by the severe phenotype of the
ERCC1-deficient animals, which do not survive longer than 3 wk
(21)
. Incorporation of 5-bromo-2'-deoxyuridine (BrdU) (as
a marker for DNA synthesis) and morphological studies (nuclear area
measurement and binucleate cell count) of liver sections were used to
investigate the cell cycle status of the hepatocytes. The nuclear area
of the hepatocytes is directly related to DNA content (27
, 28)
, with the area increasing with increasing ploidy as the
animals age.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
and p53 knockout
(13)
mice has been described. Both strains were maintained
as heterozygotes and were crossed to produce double heterozygous mice.
These were intercrossed to obtain the mice used in this study, which
were genotyped for ERCC1 (21)
and p53 (13)
as
described. All the animals were maintained on an outbred background
segregating for several different genomes. To determine the DNA
replication index, mice were injected intraperitoneally with a 10 mM
BrdU solution (0.1 ml/10g body weight) and livers were removed 24 h later.
Northern analysis
Liver RNA samples (30 µg), prepared using the RNAzol method
(Biogenesis Ltd., Bournemouth, U.K.), were subjected to Northern
analysis as described (29)
. A 490 bp fragment of mouse p21
cDNA, produced by PCR amplification of total mouse embryonic cDNA with
the following p21 primers, was used as a probe:
p21F (5') ACCATGTCCAATCCTGGTGATGTCCG;
p21R (5') GGGCACTTCAGGGTTTTCTCTTGCAG.
Primers were derived from the complete mouse p21 cDNA sequence
(GenBank Acc. No. U09507). The filter was then stripped and reprobed
for glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA as a loading
control using a 1.2 kb cloned fragment of GAPDH cDNA (30)
.
Signals were visualized using a Molecular Dynamics (Sunnyvale, Calif.)
PhosphorImager and analyzed with ImageQuant software.
Immunohistochemistry
Livers were removed and immersion fixed for 24 h in
methacarn, followed by storage in 70% ethanol. For the morphometry
studies, 3 µm sections were stained with hematoxylin/eosin (H&E). For
immunohistochemistry, 3 µm sections were mounted on slides that had
been coated with poly-L-lysine. For the BrdU antibody, cells were
rendered permeable by incubating slides in 1N HCl at 60°C for 1 h. The slides were then washed in running water and rinsed in
phosphate-buffered saline (PBS) for 5 min. The sections were
subsequently blocked for endogenous peroxidase activity in 3% hydrogen
peroxide for 10 min at room temperature and incubated in a 1:5 dilution
of normal goat serum (Scottish Antibody Production Unit) for 10 min.
After this, the samples were incubated in the primary antibody to BrdU
(MAS250, Harlan SeraLab), diluted (1:50) in normal serum for 30 min at
room temperature, washed, and incubated for 30 min in biotinylated
secondary antibody diluted (1:50) in normal serum. After washing the
secondary antibody off, the slides were incubated in horseradish
peroxidase-conjugated AB Complex (streptavidin/biotin, DAKO Ltd.,
Bucks, U.K.) for 30 min at room temperature. AB Complex was washed off
and the positive nuclei were visualized in 3,3'-diaminobenzidine
tetrahydrochloride solution. Finally, the slides were counterstained
lightly in hematoxylin (Sigma Diagnostics, St. Louis, Mo.), rinsed in
Scott’s tap water (1% (w/v)
K2CO3, 10% (w/v)
MgSO4), dehydrated, and mounted. Controls were
carried out using preimmune serum in place of primary antibody.
Measurement of hepatocyte nuclear areas
The AxioHOME Morphometric System (Zeiss) was used to perform the
study of BrdU incorporation and nuclear area size distribution
(31)
. Typically, a 100x objective lens was used for the
measurements; 10 fields, each containing 50–70 nuclei, were measured
for the BrdU-stained sections. For the area measurements and binucleate
hepatocyte counts, all the nuclei in a field were measured and five
fields were assayed per animal.
Centromeric staining of hepatocytes
Fresh livers were gently pressed onto a glass coverslip. This
method allows single hepatocytes to detach in a nondisruptive manner so
that their nuclei remain intact. Liver cells were air-dried for 5 min
and fixed in 4% paraformaldehyde/PBS. After washing in PBS, the cells
were permeabilized in KB buffer (10 mM Tris-HCl pH 7.7, 0.15 M NaCl,
0.1% bovine serum albumin)/0.1% Triton, washed twice in PBS without
Triton, and incubated for 30 min at room temperature in the primary
antibody, NR antibody (32
; kindly donated by William
Earnshaw, Institute of Cell and Molecular Biology, Edinburgh
University) and diluted (1:1000) in PBS/0.1% Triton/0.1% sodium
azide. The coverslips were then washed twice in PBS/1% Triton/0.1%
sodium azide and subsequently incubated in a dilution (1:1000) of the
biotinylated secondary antibody (anti-human IgG) for 30 min at room
temperature. Then they were washed again and incubated for another 30
min in the tertiary antibody solution (streptavidin/Texas red complex,
1:1000 dilution in PBS/0.1% Triton/0.1% sodium azide) at room
temperature. Finally, the cells were washed and stained with
4,6-diamidino-2-phenylindole (DAPI; 0.5 µg/ml in PBS) for 5 min,
mounted on slides with Vectashield mounting medium, and sealed with
nail varnish. Red dots within the nucleus of the hepatocytes,
corresponding to the centromeres, were scored on a fluorescence
microscope (Axioplan 2, Zeiss) using a 100xNA 1.4 oil immersion
objective lens. The images presented here were acquired using a
Sedat/Agard deconvolution microscope (Applied Precision Inc., Issaquah,
Wash.) with an Olympus IX70 inverted microscope, a 100x/NA 1.4 oil
immersion objective lens, and a PXL CCD camera (Photometrics, Munich,
Germany). This process was performed using DeltaVision software running
on a Silicon Graphics computer. All pictures were processed using the
Adobe PhotoShop or Powerpoint packages.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. This imposed a technical limitation on the number of
double null samples analyzed at 3 wk of age.
To see whether the ERCC1-deficient liver phenotype was affected by p53
deficiency, studies of nuclear morphology and binucleate frequency were
carried out on H&E-stained liver sections. In normal mouse liver
development, the number of binucleate and polyploid cells increases
with age, concomitant with a decrease of mononucleate, diploid cells.
The appearance of both the wild-type (Fig. 1A
) and the p53-deficient (Fig. 1C
) livers was
normal, but this nuclear uniformity was lost in ERCC1-deficient (Fig. 1B
) and double null (Fig. 1D
) livers, with both
showing enlarged nuclei in addition to the smaller size characteristic
of wild-type. Thus, the lack of p53 also failed to rescue the
ERCC1-deficient liver phenotype. Despite overall similarities in the
appearance of the ERCC1 null and the double null livers, initial
observations suggested that the number of binucleate cells was elevated
in the double null compared to the ERCC1-deficient livers (see Fig. 1B, D
). To confirm these observations, the size of the
hepatocyte nuclei and the number of binucleate cells were determined.
Only cells that were unequivocally parenchymal in origin were counted,
taking care to avoid the areas surrounding the portal veins, where
there is an increased number of nonparenchymal cells. The distribution
of the nuclear areas is shown in Fig. 2
. The wild-type profile (top panel) shows a distribution of areas with a
mode of 20–30 µm2 and a maximum area of 70
µm2. The number of nuclei that fall within the
last two nuclear categories (from 50–70 µm2
was so low (1% of the total) that >50 µm2 was
the cut-off area chosen to define abnormal, enlarged nuclei. We have
previously demonstrated that this category comprises both polyploid and
aneuploid nuclei (21)
. The p53-deficient livers showed the
same size distribution as the wild-type (see Figs. 1
and 2
), suggesting
that p53 absence does not affect liver nuclear morphology, and this was
confirmed by the observations made in the double null livers. In both
the ERCC1-deficient livers and the double null livers the size
distribution profiles were flatter and broader, indicating a clear
shift to larger nuclear areas. We have previously reported this
distribution for ERCC1-deficient mice (21)
and it has also
been observed by Weeda et al. (22)
. The mode in these two
categories has increased from the wild-type value of 20–30 to 30–40
µm2, and the biggest nuclei were up to 200
µm2. Also, the percentage of nuclei >50
µm2 has increased for both the ERCC1 null and
the double null genotypes from the wild-type value of 1% to 34%. This
percentage remained low (5%) for the p53 null livers. Using the
Kolmogorov-Smirnov test, there was no significant difference in the
distribution of nuclear areas in ERCC1-deficient and the double
null livers, but both were significantly different (P=0.01)
from wild-type.
|
|
Binucleation is inhibited in ERCC1-deficient hepatocytes
The older the mouse, the higher the percentage of polyploid (4n,
8n) binucleate cells and the lower the number of diploid mononucleate
cells. Binucleation arises from consecutive rounds of replication in
the absence of cell division and is an essential step in the normal
development of liver polyploidization (27)
. Therefore, a
change in the pattern of binucleation in our mice would reflect changes
in the regulation of liver development. Binucleate cells were counted
and divided into two categories: normal (nuclear area <50
µm2) and enlarged (area>50 µm2). The
percentage of binucleate hepatocytes in wild-type (8%) and
p53-deficient livers (10%) was not significantly different (by the
Mann-Whitney U test). In these two genotypes, there was no
evidence of binucleate cells with enlarged nuclei (see Fig. 3
and Fig. 1
). In ERCC1-deficient livers, the percentage of binucleate
cells (3%) was significantly lower than in all other genotypes
(P=0.05 compared to wild-type), with 44% of the binucleates
having enlarged nuclei. In the double nulls the total number of
binucleate cells (7%) was significantly increased compared to
ERCC1-deficient livers (P=0.05), being similar to that of
the wild-type and p53-deficient tissues, with a high percentage (61%)
of those binucleate cells having enlarged nuclei. Thus, binucleation
seems to be impaired in the absence of ERCC1 and this block is rescued
by the absence of p53.
|
The absence of p53 rescues the replication block in ERCC1-deficient
liver
Sections of liver were stained with an anti-BrdU antibody to
establish the replication index for each genotype (see Fig. 4
). The DNA replication index is the percentage of cells labeled with
BrdU during the 24 h before collection of liver samples. Because
ERCC1-deficient and double null animals are always runted, it was
decided to use rare, runted wild-type and p53 nulls as controls for
this experiment. We have previously reported that the liver nuclear
abnormalities in ERCC1-deficient mice are not observed in runted
control animals (21)
. BrdU incorporation was also
determined for normal (nonrunted) wild-type animals. Overall, the
replication index for the runted wild-type, p53-deficient, and
ERCC1-deficient livers was very low (<2%, see Figs. 1
, 4
), although
these values were not significantly different (by the Mann-Whitney
U test) from the higher value (4%) in the normal-sized
wild-type animals. Lower values were expected since these groups all
suffered from a growth impairment. Although replication levels in
ERCC1-deficient livers were less than in nonrunted wild-type animals
when compared to runted wild-type animals, there was no detrimental
effect in terms of replication. In runted p53-deficient livers, the
levels of replication were not rescued to normal wild-type (nonrunted)
levels. However, in the double null livers the level of replication has
significantly increased compared to all other runted genotypic classes
(P=0.1 compared to ERCC1-deficient), reaching normal levels
and demonstrating a bypass of the replication arrest typical of runted
animals.
|
Once the difference in the liver DNA replication indices was
established, we determined which subpopulation of hepatocytes (i.e.,
normal, or enlarged nuclei) was responsible for the increased BrdU
incorporation in double null livers. The size distribution of
BrdU-positive nuclei is shown in Fig. 5
. When the size distribution of BrdU-stained nuclei is compared to the
general nuclear size distribution, it is clear that although the
distributions of BrdU-stained nuclei in the wild-type, p53 null, and
ERCC1 null livers correspond with the general area distribution for the
same genotypic classes, this is not the case for the double null
livers. In this latter case, the distribution of sizes of BrdU-positive
nuclei is significantly shifted (P=0.01 by
Kolmogorov-Smirnov test) toward the enlarged nuclei (see also Fig. 1D
). The replicative potential of the double null livers is
concentrated in the subpopulation of hepatocytes with enlarged nuclei.
Taken together with the earlier demonstration that the percentage of
BrdU-stained nuclei in these livers is higher than in any of the other
runted animals and has reached levels equivalent to normal wild-type
livers, it is reasonable to think that those nuclei that have
‘gained’ a replicative status are the enlarged nuclei. This ‘gain of
replicative function’ effect is not seen in p53-deficient liver.
|
Enlarged ERCC1-deficient hepatocytes are arrested in G2
ERCC1-deficient liver has a reduced growth rate and DNA
replication index compared to normal control animals. Because of the
increased hepatocyte DNA content and the presence of many nonhepatocyte
cell types in young mouse livers, conventional FACS analysis of liver
nuclear DNA content could not be used to obtain a simple indication of
the stage of the cell cycle at which individual hepatocytes were;
therefore, centromeric staining of hepatocytes was used to aid the
determination (see Fig. 6
). The autoimmune serum used stains mammalian centromeres
(32)
. Each centromere appears as a single structure until
centromere duplication occurs in G2 and the
structures become double (33)
. Thus, the presence of
double centromere structures identifies cells in
G2. Fifty nuclei were scored from each 3-wk-old
liver: six wild-type, five ERCC1-deficient, one p53-deficient, and one
double null. The mean values (%) for hepatocytes in
G2 were wild-type 33 ± 2; p53-deficient 32;
ERCC1-deficient, normal nuclei 36 ± 4; enlarged nuclei 62 ±
3; double null, normal nuclei 20; enlarged nuclei 50. A typical
wild-type hepatocyte is shown in Fig. 6A
. Each red spot
represents a centromere; chromosomes appear blue. Since the majority of
the centromeres appear as single structures (single red dots), it is
concluded that this nucleus is not in G2, but is
instead in G1 (or G0), or
S-phase. A typical enlarged ERCC1-deficient hepatocyte in
G2 is shown in Fig. 6B
, identified by
a prominence of double centromeric structures. The values for wild-type
and normal ERCC1-deficient nuclei in G2 are not
significantly different (by Student’s t test), but
G2 levels are significantly higher in enlarged
ERCC1-deficient nuclei (P=0.01). We conclude there is a
G2 arrest in ERCC1-deficient and double null
hepatocytes that is restricted to the enlarged hepatocyte population.
|
p21 is a possible mediator of the G2 arrest and
increased DNA content in ERCC1-deficient liver
A role for p21 in G2 arrest and/or
increased liver ploidy was supported by the finding that mice
overexpressing p21 in their liver present a similar liver phenotype to
our ERCC1-deficient mice (26)
. To see whether elevated
levels of p21 were present in ERCC1 null mice, liver RNA was assayed
for p21 mRNA levels. Figure 7
corresponds to a sample Northern autoradiograph probed first with p21
and then reprobed with GAPDH as a loading control. A summary of the
standardized p21 mRNA levels is shown in Fig. 8
. The level of p21 mRNA in ERCC1-deficient and double null livers was 2-
and 4.5-fold higher than that of the wild-type livers, respectively.
Both values are significantly higher than wild-type (by the
Mann-Whitney U test, P=0.1 for ERCC1-deficient,
P=0.05 for double null), but not significantly different
from each other. Since we have not measured p21 levels in individual
hepatocytes and ERCC1-deficient liver contains both normal (diploid)
and enlarged cells (arrested in G2 with increased
DNA content), we are unable to determine whether the elevated p21
levels are associated with the G2 arrest,
increased DNA content, or both. In p53-deficient livers (in the
presence of proficient NER) the levels of p21 were equivalent to
wild-type. Thus, in mouse liver, the expression of p21, both in the
presence and absence of NER, is p53-independent.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 35)
. Tight regulation of DNA repair and cell cycle progression
is necessary for the cell to cope with genotoxic stress
(36
37
38)
, and p21 has been shown to play an important role
in this process via both p53-dependent and independent pathways
(39
, 40)
. The liver phenotype in the ERCC1-deficient mouse
reflects a profound misregulation of the cell cycle. p53 levels had
already been found to be elevated in ERCC1-deficient livers
(21)
, making this a good model for the in vivo
study of the relationship between p53, DNA repair and cell cycle
control. We have previously reported that 3-wk-old ERCC1-deficient mice
have a distribution of hepatocyte nuclear areas that is reminiscent of
an aged wild-type (21)
, suggesting that the phenotype
presented by the young ERCC1-deficient mice may also be related to
premature aging. We have now investigated this further by carrying out
an analysis of aspects of the hepatocyte’s life cycle, such as nuclear
size and binucleation rates, which change during liver development and
differentiate young from old livers, in addition to studying the cell
cycle regulation.
Binucleation
The normal binucleation process was impaired in ERCC1-deficient
animals, but wild-type levels of binucleation were observed in p53 and
ERCC1 double null livers, indicating a role for p53 in preventing
binucleation in the absence of ERCC1. However, in repair-proficient
normal circumstances, p53 absence did not affect binucleation. This
p53-dependent suppression of binucleation in ERCC1-deficient cells can
be understood in the context of the protective role of p53. If cells
that cannot repair their damaged DNA proceed through either replication
or binucleation (which involves a round of replication followed by an
acytokinetic cell division), this could lead to the fixation of
mutations and the creation of chromosomal aberrations.
DNA replication index
The protective effect of p53 in suppressing DNA replication in
ERCC1-deficient liver can be seen from the BrdU labeling data. Except
for the double null mice, all runted animals have lower replication
indices than nonrunted wild-type animals, consistent with their
intrinsic growth retardation. The values obtained for normal wild-type
animals agree with results obtained by other groups (41
, 42)
. These low runt-related levels in the ERCC1-deficient
animals are rescued back to normal, higher levels in the absence of
p53.
p21 in liver
p21 has a well-established role in cell cycle control and the
regulation of cell growth; overexpression of p21 in liver has also been
linked to an ERCC1 null-like phenotype (26)
. Elevated
levels of p21 have been reported in primary fibroblasts derived from
ERCC1-deficient mice (22)
, although no data were provided
to support this affirmation. Our data demonstrate that increased liver
p21 mRNA levels are associated with either the G2
arrest or the premature increase in nuclear DNA content seen in
ERCC1-deficient mice; furthermore, the p21 response is p53 independent.
Another line of evidence comes from studies performed in
vitro using a human colorectal carcinoma cell line (DLD1), where a
p53-independent role for p21 in cell growth and DNA repair has been
proposed (43)
. p21 has been shown to play a role in the
control of the initiation of mitosis (G2/M
transition) through its inhibitory effect on cdk2 and cdk1-cyclin B
(44)
. p21 has also been shown to bind and inhibit
cdk1-cyclin A complexes, which play an important regulatory role in
early mitosis (45)
. The ERCC1 null mouse may be a useful
tool to understand the effect of p21-mediated growth arrest and its
relationship with the DNA repair machinery in vivo.
Liver ploidy
The most striking feature of ERCC1-deficient liver is the
increased hepatocyte DNA content at such an early age. Our previous
studies have shown that the population of cells with enlarged nuclei
comprises both polyploid and aneuploid cells (21)
. In
normal mammals, polyploidization of hepatocytes is strongly correlated
with age and the extent of genotoxic damage, but aneuploidy is not
usually observed (28)
. Because hepatocytes absorb toxic
products as part of their function, liver cells may accumulate genetic
damage more rapidly than cells from other tissues. The absence of a
special pool of nondifferentiated stem cells may make polyploidization
a safer way to compensate for cellular losses than mitosis, which may
result in a high incidence of chromosomal abnormality
(28)
. Apart from its role in protecting against the
deleterious cellular effects of endogenous DNA damage, polyploidization
allows the liver to preserve its normal size and functional capability
without the risk of accumulating through mitosis a large proportion of
aberrant, aneuploid cells, which may increase the level of cellular
transformation. Several hypotheses seek to explain the possible
mechanism by which polyploidization exerts its protective role: by
reiteration of vital genes, by making the cell more resistant to toxins
and carcinogens, and by providing a morphogenetic strategy against
aging in specialized diploid systems. However, the polyploidization of
hepatocytes in later periods of life, particularly the progressive
polyploidization observed in mouse liver during senescence, by
formation of nuclei up to 32n in DNA content is apparently an integral
part of the aging process; more than representing a real protection
against aging, it could well be a tissue-specific adaptation to
age-related cellular loss (28)
.
Endoreduplication
Polyploidization comes about by continuous rounds of DNA
replication in the absence of mitosis (a process known as
endoreduplication). p21 has also been implicated in the regulation of
the checkpoints that prevent endoreduplication (20
, 46)
.
We have observed endoreduplication in the presence of elevated levels
of p21, as opposed to its absence, so we propose that misregulation of
p21 expression (either its absence or its overexpression) results in
the uncoupling of the normal link between S-phase and mitosis, leading
to cells in G2 undergoing repeated rounds of
replication in the absence of mitosis. Alternatively, elevated levels
of p21, in response to nonrepaired, accumulating damage, may lead the
hepatocytes down a terminal (and also protective) differentiation
pathway. p21 has been reported to play an important part in the
differentiation process of several cell types (47
, 48)
.
Since polyploidization is the natural state of ‘aged’ hepatocytes,
the ERCC1 null phenotype could indeed be equivalent to a premature
aging syndrome.
Greater levels of DNA replication were observed in enlarged G2-arrested hepatocytes from animals deficient in both ERCC1 and p53. This observation would also predict higher DNA content (increased nuclear size) in double nulls compared to ERCC1-deficient, which was not observed. Perhaps the replication is only partial and not sufficient to result in higher levels of endoreduplication and a detectable increase in nuclear volume. However, it again illustrates the protective role of p53 in preventing DNA replication in ERCC1-deficient liver.
Model for the ERCC1-deficient liver phenotype
It is important to remember that the ERCC1-deficient liver
phenotype arises spontaneously in the absence of exogenous DNA-damaging
agents. The liver is required for a large number of metabolical
processes such as lipid peroxidation, intermediate metabolism of
carbohydrates and proteins, synthesis of proteins and detoxification
and removal of foreign material such as bacteria, drugs and other
noxious substances. All this results in the liver cells being exposed
to and producing large quantities of genotoxic agents, such as
oxygen-free radicals and toxic metabolic byproducts. Many of these will
result in bulky DNA adducts that are normally repaired by NER. However,
given that the same liver phenotype is not observed in XPA (23
, 25)
, XPC (24)
, or the recently reported XPG
knockout mice (49)
, it has been suggested that other forms
of damage, which are dealt with by the recombinational repair pathway,
may be more significant (22)
. Probable involvement in this
recombinational pathway distinguishes ERCC1 from most other NER
proteins. The contribution that different forms of endogenous damage
make to the ERCC1-deficient phenotype will be resolved when the DNA
lesions are identified directly in ERCC1-deficient liver. The
accumulation of unrepaired damage does not result in the normal cell
cycle arrest at G1/S. Instead, cells replicate
and a population of cells arrested in G2 with
increased DNA content accumulates. This is associated with a
p53-independent accumulation of p21. The premature polyploidization may
be a protective response to the accumulation of damaged DNA. The
development of polyploidy in young ERCC1-deficient mice is different
from the normal situation in aging wild-type animals, where aneuploidy
is not usually observed. ERCC1-deficient hepatocytes with increased,
but aneuploid, DNA content may represent incomplete rounds of
endoreduplication where replication has been physically prevented by
unrepaired DNA lesions. p53 accumulates in ERCC1-deficient liver, but
the main features of the phenotype are independent of p53. However, p53
does act to prevent binucleation and reduce the amount of DNA
replication.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Handra-Luca, J. Hernandez, G. Mountzios, E. Taranchon, J. Lacau-St-Guily, J.-C. Soria, and P. Fouret Excision Repair Cross Complementation Group 1 Immunohistochemical Expression Predicts Objective Response and Cancer-Specific Survival in Patients Treated by Cisplatin-Based Induction Chemotherapy for Locally Advanced Head and Neck Squamous Cell Carcinoma Clin. Cancer Res., July 1, 2007; 13(13): 3855 - 3859. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. O. Andrieux, A. Fautrel, A. Bessard, A. Guillouzo, G. Baffet, and S. Langouet GATA-1 Is Essential in EGF-Mediated Induction of Nucleotide Excision Repair Activity and ERCC1 Expression through ERK2 in Human Hepatoma Cells Cancer Res., March 1, 2007; 67(5): 2114 - 2123. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Dikovskaya, D. Schiffmann, I. P. Newton, A. Oakley, K. Kroboth, O. Sansom, T. J. Jamieson, V. Meniel, A. Clarke, and I. S. Nathke Loss of APC induces polyploidy as a result of a combination of defects in mitosis and apoptosis J. Cell Biol., January 16, 2007; 176(2): 183 - 195. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. R. Jones and K. Ravid Vascular Smooth Muscle Polyploidization as a Biomarker for Aging and Its Impact on Differential Gene Expression J. Biol. Chem., February 13, 2004; 279(7): 5306 - 5313. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Tian, R. Shinkura, N. Shinkura, and F. W. Alt Growth Retardation, Early Death, and DNA Repair Defects in Mice Deficient for the Nucleotide Excision Repair Enzyme XPF Mol. Cell. Biol., February 1, 2004; 24(3): 1200 - 1205. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-E. Guidotti, O. Bregerie, A. Robert, P. Debey, C. Brechot, and C. Desdouets Liver Cell Polyploidization: A Pivotal Role for Binuclear Hepatocytes J. Biol. Chem., May 23, 2003; 278(21): 19095 - 19101. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.-T. Hsia, M. R. Millar, S. King, J. Selfridge, N. J. Redhead, D. W. Melton, and P. T. K. Saunders DNA repair gene Ercc1 is essential for normal spermatogenesis and oogenesis and for functional integrity of germ cell DNA in the mouse Development, March 2, 2003; 130(2): 369 - 378. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-H. Ling, L. Liebes, J.-D. Jiang, J. F. Holland, P. J. Elliott, J. Adams, F. M. Muggia, and R. Perez-Soler Mechanisms of Proteasome Inhibitor PS-341-induced G2-M-Phase Arrest and Apoptosis in Human Non-Small Cell Lung Cancer Cell Lines Clin. Cancer Res., March 1, 2003; 9(3): 1145 - 1154. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Selfridge, K.-T. Hsia, N. J. Redhead, and D. W. Melton Correction of liver dysfunction in DNA repair-deficient mice with an ERCC1 transgene Nucleic Acids Res., November 15, 2001; 29(22): 4541 - 4550. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||
