ABSTRACT
The xeroderma pigmentosum syndrome complementation group C (XP-C) is due to a defect in the global genome repair subpathway of nucleotide
excision repair (NER). The XPC protein is complexed with HHR23B, one of the two
human homologs of the yeast NER protein, RAD23 [Masutani
et al
. (1994)
EMBO J.
8, 1831-1843]. Using heparin chromatography, gel filt- ration and native gel electrophoresis we demonstrate that the
majority of HHR23B is in a free, non-complexed form, and that a minor fraction is tightly associated with XPC.
In contrast, we cannot detect any bound HHR23A. Thus the HHR23 proteins may
have an additional function independent of XPC. The fractionation behaviour
suggests that the non-bound forms of the HHR23 proteins are not necessary for the core of the
NER reaction. Although both HHR23 proteins share a high level of overall
homology, they migrate very differently on native gels, pointing to a
difference in conformation. Gel filtration suggests the XPC-HHR23B heterodimer resides in a high MW complex. However, immunodepletion
studies starting from repair- competent Manley extracts fail to reveal a stable association of a
significant fraction of the HHR23 proteins or the XPC-HHR23B complex with the basal transcription/repair factor TFIIH, or with
the ERCC1 repair complex. Consistent with a function in repair or DNA/chromatin
metabolism, immunofluorescence studies show all XPC, HHR23B and (the free)
HHR23A to reside in the nucleus.
A complex network of DNA repair mechanisms protects the genetic information from
continuous genotoxic pressure caused by the DNA-damaging effect of exogenous and genotoxic agents. Such damage can lead to
inborn defects, cell death or neoplasia. Nucleotide excision repair (NER) is
one of the most important DNA damage repair pathways, since this process
recognizes a wide variety of lesions. Impaired NER activity has been
extensively investigated in cells from three human disorders: xeroderma
pigmentosum (XP), Cockayne syndrome (CS) and trichothiodystrophy (TTD) (
2
,
3
). These genetic diseases are characterized by sun (UV) hypersensitivity,
genetic instability and a marked clinical and genetic heterogeneity. Many genes
involved in XP, CS and TTD complementation groups have been cloned using cells
from human patients or from NER-deficient, UV-sensitive chinese hamster ovary mutants (
4
). Extensive sequence homology between the mammalian and yeast NER proteins has become apparent, indicating that the NER pathway is strongly conserved in
eukaryotic evolution (
5
).
At least two NER sub-pathways can be discerned: transcription-coupled repair (TCR) and global genome repair (GGR) (
6
). In contrast to all other xeroderma pigmentosum groups, group C patients are
only defective in GGR (
7
). Previously, we reported the identification and cloning of two human homologs
of the yeast NER gene
RAD23
:
HHR23A
and
HHR23B
(
8
). The yeast
Saccharomyces cerevisiae
rad23
null mutants display an intermediate UV-sensitive phenotype (
9
), suggesting that the affected protein is not required for NER. The HHR23B gene
product forms a tight complex with the XPC protein. This complex has a high
affinity for ssDNA and both subunits were found to be indispensible for
in vitro
NER (
10
). Like the human XPC-HHR23B complex, the
S.cerevisiae
RAD4 and RAD23 protein homologs were also determined to be complexed with each
other (
11
). XP-C cells harbour a specific defect in the repair of non-transcribed sequences of the genome, including the non-transcribed strand of active genes, whereas the NER subpathway
that accomplishes the preferential repair of the transcribed strand of active
genes is still operational (
7
,
12
). This implies a selective role for the XPC complex in the global genome NER
system (
13
). Additionally, XP-C cells are claimed to be defective in the repair of rDNA (
14
). Recently it has been shown that
in vitro
repair of a cholesterol-substituted oligonucleotide does not require XPC protein (
15
).
RAD23 is a ubiquitin-like fusion protein displaying significant homology to ubiquitin at the N-terminus (
8
,
16
). A second link has been identified between RAD23 and the ubiquitin pathway,
namely a twice repeated element, homologous to a C-terminal extension of a Class II ubiquitin-conjugating enzyme (E2) (
17
). The ubiquitin-conjugating pathway is involved in the proteolytic degradation of
proteins, and in additional cellular processes such as DNA repair, chromosome
condensation and decondensation, and cell cycle control (
18
,
19
). The link to these other DNA-metabolizing processes presumably comes from ubiquitin-mediated proteolytic degradation of key proteins involved in these
events.
Here we present data on the purification and stable association of HHR23A,
HHR23B and XPC proteins with other known NER factors. Additionally, the sub-cellular localization of HHR23A, HHR23B and XPC was determined by
immunofluorescence.
Purification of nucleic acids, restriction enzyme analysis, gel electrophoresis
of nucleic acids and proteins, transformation of
Escherichia coli
, etc. were performed according to standard procedures (
20
). RNA samples were separated on 1% agarose gels and transferred to Zeta probe
membrane (Bio-Rad) as described (
21
). Labelling of DNA probes was carried out using the random priming protocol (
22
). Immunoblotting was performed as described elsewhere (
23
). HHR23A and HHR23B proteins were translated
in vitro
using a rabbit reticulocyte lysate system as recommended by the manufacturer
(Promega) using 50 [mu]Ci of [
35
S]methionine (1 mCi/mmol). After polyacrylamide gel electrophoresis (PAGE) and
native gel electrophoresis, both labelled proteins were blotted and visualized
by autoradiography.
For non-denaturing gel electrophoresis, a 4-15% gradient polyacrylamide gel in TBE buffer (90 mM Tris, 80 mM
boric acid and 2.5 mM EDTA) and 12% glycerol was prepared. The gel was pre-run for 30 min at 70 V, loaded with samples and run for 2 h at 70 V
followed by 16-20 h at 150 V. Proteins included as molecular mass standards used for
estimation of the native molecular weight of HHR23A and HHR23B were ferritin
(440 kDa), catalase (240 kDa), lactate dehydrogenase (140 kDa), and albumin (67
kDa) (Boehringer Mannheim). The Western blot was stained with Ponceau-S to visualize the molecular weight markers for determination of the
apparent MW of the HHR23 proteins. HHR23A and HHR23B proteins were detected by
polyclonal antibodies and autoradiography.
The full-length
HHR23A
and
HHR23B
cDNAs were cloned into the pET11D vector (Novagen), transferred into
E.coli
strain BL21(DE3), and gene expression was induced over 4 h by IPTG. Cells were
homogenized in PBS and, after sonication, cleared by centrifugation.
Approximately 20 g cells were disintegrated by sonication, followed by
centrifugation to remove cell debris. Recombinant HHR23A and HHR23B proteins
were purified by chromatography on a Q-Sepharose Fast Flow column (1 * 12 cm, flow rate 18 ml/h). For HHR23A, the column was eluted with
a linear gradient 0 -> 0.5 M NaCl in 0.1 M NaCl, 10 mM K-phosphate pH 7, while for HHR23B a gradient 0 -> 0.4M NaCl in 10 mM NaCl, 10 mM K-phosphate pH 7 was used. HHR23A protein eluted at 0.3 M,
HHR23B protein at 0.11 M NaCl. For antibody production these proteins were
subjected to SDS-PAGE. Bands were cut from Coomassie stained gels, electroeluted and
concentrated with Centricon 30 concentrators (Amicon). The identity of the
eluted proteins was verified by amino acid sequencing. For large scale
purification, Q-Sepharose fractions were pooled, brought on to 20% ammoniumsulphate and
loaded on a butyl Sepharose Fast Flow column (1 * 12 cm, flow rate 18 ml/h). Columns were eluted with a linear gradient 20 -> 0% ammoniumsulphate in 0.1 M NaCl, 10 mM K-phosphate pH 7. Fractions containing HHR23A or HHR23B protein
(eluted at 4.5% ammoniumsulphate) were dialysed and kept frozen after addition
of 1/5 volume of glycerol. Polyclonal antibodies were raised in rabbits against
the
E.coli
-overproduced human HHR23 proteins, as described (
23
). Affinity-purified antibodies were derived from precise elution of antibodies
specifically bound to recombinant antigen immobilized on nitrocellulose after
transfer from SDS-polyacrylamide gels (
24
).
The synthetic peptide (KTKREKKAAASHLFPFEKL), corresponding to the C-terminus of XPC, was used to produce a polyclonal antibody in rabbits.
Prior to injection, the peptide was cross-linked to KLH carrier protein. Affinity-purified antibodies were derived by eluting XPC-peptide from a column to which it had been coupled. As a
second antibody, alkaline phosphatase-conjugated goat anti-rabbit was used, the latter visualised by 5-bromo-4-chloro-3-indolyl phosphate. Immunoblots were
incubated with monoclonal antibodies (Mab3C9) against the p62 subunit of TFIIH
(generously provided by Dr J.-M. Egly, Strasbourg), as published earlier by Fischer
et al.
(
25
). A polyclonal antiserum raised against p89, a GST-ERCC3 fusion protein containing an internal part (amino acids 82-480) was used to detect the p89/ERCC3/XPB component of TFIIH.
HeLa cells and Chinese hamster ovary (CHO9) cells were grown in F10/DMEM medium
(1:1) supplemented with 10% fetal calf serum, penicillin 100 U/ml and
streptomycin 0.1 mg/ml. Cells were harvested and extracts were prepared from 2-5 ml of packed cell pellets by the method of Manley, as modified by Wood
(
26
,
27
) dialysed in buffer A and stored at -80oC until use. XP-A and XP-C patient cell lines used in these experiments were CW12
(XP-A) and XP4PA (XP-C) (
29
,
30
). [NB. Cell line CW12 was originally described as belonging to XP group C (
28
). During our work, however, we found it to carry an XP-A defect, but to harbour no mutation in the XPC gene.] COS-1 SV40-transformed African green monkey kidney fibroblasts were
seeded semi-confluent in 6-well plates, and grown on F10/DMEM medium (1:1) supplemented with 5%
fetal calf serum, penicillin and streptomycin. For DEAE-dextran/chloroquine transfection, SV40 promoter-driven constructs were used (pSVL derived pSLM vector; Pharmacia
biotech) containing full length
HHR23A
and
HHR23B
cDNAs. The empty pSLM vector was used as a negative transfection control in
parallel with
HHR23A
and
HHR23B
genes. A 10% DMSO shock for 1.5 min was given 4 h after transfection. Transient
expression of the corresponding HHR23A and HHR23B protein was analysed by
immunofluoresence 48 h after transfection.
Protein A-Sepharose CL-4B beads (Pharmacia-Biotech) (70 [mu]g) were washed three times with PBS, then incubated with
10 [mu]l anti-XPC antibodies or pre-immune serum for 15 min at 4oC. The beads were then washed three times in buffer A (25
mM HEPES-KOH pH 7.8, 0.1 M KCl, 12 mM MgCl
2
1 mM EDTA, 2 mM DTT and 17% glycerol) and added to a repair-competent HeLa extract for 30 min at 4oC. The supernatant obtained after spinning down the beads was used as
a depleted HeLa extract and tested on immunoblots for co-depletion. After boiling the protein A-Sepharose beads, the depleted `bound' fraction was analysed by
immunoblot analysis.
A HeLa cell free extract (14 [mu]g/[mu]l; 750 [mu]l) was applied to a heparin-Sepharose column. Proteins were eluted with a linear 30 ml
gradient from 0.15 to 1.15 M NaCl in PBS containing 10 mM 2-mercaptoethanol buffer.
Whole-cell extracts were prepared for fractionation on a phosphocellulose
column. This 5 ml column equilibrated with buffer A (25 mM HEPES-KOH pH 7.9, 1 mM EDTA, 10% glycerol, 0.01% Triton X-100, 1 mM DTT and 0.25 mM PMSF) contained 0.2 M KCl. The column was
washed with the same buffer and the adsorbed proteins were eluted with buffer A
containing 1 M KCl. The peak fractions from the flow-through (CFI) and the eluate (CFII) were concentrated by dialysis against
buffer A containing 0.1 M KCl and 20% sucrose, and stored at -80oC.
Size-fractionation of HeLa Manley extracts was performed on a Sephacryl S300-HR column. HeLa nuclear extracts were loaded on a 1 * 46.4 cm column with a flow rate of 0.92 ml/7.5 min and
eluted with PBS (7.4 ml/h). The resulting fractions were concentrated. Protein
profiles of HHR23A, HHR23B and XPC were visualized on immunoblots using
alkaline phosphatase-labeled secondary antibodies. Proteins included as molecular mass standards used for estimation of the
native molecular weight were thyroglobulin (669 kDa), ferritin (440 kDa),
catalase (232 kDa), aldolase (158 kDa), and albumin (67 kDa) (Boehringer
Mannheim).
HeLa, XP-A and XP-C cells were grown on slides in F10/DMEM medium supplemented with
10% fetal calf serum, penicillin and streptomycin, and washed prior to fixation
in PBS. Cells were fixed for 10 min in 2% paraformaldehyde-PBS, followed by incubation with methanol at room temperature for 20 min.
After extensive washing (3 * 5 min) with PBS supplemented with 0.15% glycine and 0.5% BSA (PBS
+
) the slides were incubated with affinity-purified primary antibodies (1:100 dilution in PBS) for 1.5 h in a moist
incubation chamber at room temperature.
Immunocytochemical controls were routinely included (omission of the primary
antibody incubation step and incubation with pre-immune serum). Background was negligible.
Slides were washed in PBS
+
and incubated with goat anti-rabbit-FITC-conjugated antiserum (1:80 dilution) for 1.5 h. Slides were
washed in PBS and preserved with vectashieldtm mounting medium (Brunschwig). The DNA was stained with 4'-6 diamino-2-phenylindole (DAPI) whereas the fluorescein-labeled second antibody visualized the
antigen of interest. Fluorescence microscopy was performed with an Aristoplan
laser beam microscope. Image modification for figures was performed by using
the Adobe Photoshop program on an IBM Compaq deskpro XE 560.
To characterize XPC, affinity-purified anti-XPC polyclonal antibodies were generated and tested by
immunoblotting, using
in vitro
translated XPC protein, XPC protein purified from HeLa cells, and XP-C cells in total cell extracts. Figure
1
shows their specificity on HeLa, XP-A and XP-C protein extracts. A clear band of the expected molecular weight of
125 kDa as determined by
in vitro
translation and purified XPC (
1
) was observed in total cell extracts of repair-proficient HeLa and XP-A cells. The XP-C extract from patient XP4PA is useful for testing the
specificity of the antibody. Due to a homozygous frameshift mutation that is
predicted to result in a premature termination of the protein (
30
), this patient lacks the C-terminal XPC region encoding the part used to raise the antibodies.
Further evidence for the specificity of the anti-XPC antibodies is derived from the immunofluoresence data depicted in
Figure
8
A.
Previously, we determined XPC to be complexed with HHR23B protein (
8
). An association of XPC with TFIIH has been claimed (
36
). In an attempt to identify stable associations with other repair components or
factors involved in the basal transcription machinery, systematically-purified protein fractions were tested for the presence or absence of
HHR23A, HHR23B and XPC proteins. Purification protocols were used which are
known to leave large stable protein complexes such as TFIIH intact (
37
).
To separate the components of the general transcription factors, a HeLa cell-free extract competent for
in vitro
repair and transcription was fractionated by heparin ultrogel column
chromatography. Figure
4
shows the load and the elution fractions analysed with the affinity-purified HHR23A, HHR23B, XPC and TFIIH antibodies.
To detect possible complexes of HHR23 with other proteins, separation of repair-proficient HeLa Manley extracts under non-denaturing gel electrophoresis conditions was performed using wild-type cell extracts and anti-HHR23A and anti-HHR23B antibodies. The results are shown in Figure
5
. The HHR23A protein was detected as a single band migrating at ~70 kDa. For HHR23B, two forms with approximate sizes of 140 kDa were
distinguished in HeLa cell extracts. To determine the specificity of the
apparent molecular weights of both HHR23 proteins,
in vitro
translated protein and recombinant
E.coli
-overproduced HHR23 proteins were run in parallel. After immunoblotting and
autoradiography, all HHR23A protein samples were found to migrate at the same
size, suggesting that the recombinant polypeptide has a similar conformation as
the corresponding HeLa and
in vitro
translated proteins (Fig.
5
). The lower HHR23B band observed in the HeLa lane migrates at the same position
as the recombinant protein and the
in vitro
translated form. Thus, the upper band may represent a modified form of the
HHR23B protein or binding of another small polypeptide. The notion that both
bands are derived from HHR23B is supported by the observation that both signals
disappeared when the antibodies were competed with excess recombinant HHR23B
protein (data not shown).
Analysis of HHR23A, HHR23B and XPC protein sequences for the presence of a DNA
binding domain or a nuclear localization signal (
39
) revealed no clear matches conforming with the known consensus sequences.
Moreover, we failed to detect DNA binding activity of the isolated recombinant
HHR23A and HHR23B proteins (unpublished observations). To define the
subcellular distribution of the free HHR23A, HHR23B protein molecules and the
XPC-HHR23B complex, we performed indirect immunofluoresence in HeLa cells,
COS-1 transfected cells, normal repair-proficient fibroblasts and XP-C patient fibroblasts. No labeling was seen after treatment
with secondary antibodies alone or after competition with excess of the
recombinant HHR23 proteins or the XPC peptide used for immunization. The
specificity of the primary antibody was confirmed by the use of pre-immune sera in all experiments, included as a negative control. For both
human equivalents, HHR23A and HHR23B, a clear nuclear localization was
observed, and the protein appeared to be absent from the nucleoli (Fig.
7
A and D). However, this absence can be due to the fixation procedure.
This article describes the partial characterization of HHR23A, HHR23B and XPC
proteins. Both XPC and HHR23B proteins are known to be specifically involved in
global genome nucleotide excision repair. GGR deals with the repair of bulk
DNA, including the non-transcribed strand of active genes (
7
,
12
,
40
) and is important for preventing carcinogenesis. Evidence for this comes from
the lack of enhanced cancer risk in patients with the transcription-coupled NER disorder, Cockayne syndrome (
41
) and the high cancer predisposition when the GGR subpathway is defective as in
XP-A and XP-C. Purification of the XPC-correcting NER activity revealed a heterodimeric protein
complex consisting of XPC and HHR23B (
8
). However, the functional significance of the association of HHR23B with XPC is
not known, and could be, for instance, stabilization of the XPC protein. The
yeast RAD23 protein also has a role in NER, and was recently found to form a
protein complex with the yeast RAD4 protein (
11
), a structural homolog of XPC.
Except for potential phosphorylation sites, analysis of the primary amino acid
sequence of XPC gave no clues about a particular function. The primary amino
acid sequence of RAD23 protein and its mammalian homologues indicated that they
are N-terminal ubiquitin-like fusion proteins (
8
,
16
). In addition, a second link with the ubiquitin pathway was observed. Two repeated domains in the RAD23 amino acid sequence
shared homology to a C-terminal extension in a bovine ubiquitin-conjugating enzyme (E2-25kD) (
17
). This suggests that the RAD23 protein may have an involvement in the ubiquitin
system, within the context of NER or in another process, implying a dual functionality. Other NER proteins have also
been found to have dual functions. Examples include the XPB and XPD proteins in
the multisubunit TFIIH transcription repair factor (
42
) and the RAD1-RAD10 complex, additionally involved in mitotic recombination (
43
).
In the present studies, we tried to find evidence for a stable association of
XPC and HHR23 proteins with each other and with previously identified protein
complexes which have defined enzymatic activity, involving endonuclease-mediated incision (ERCC1/ERCC4) or transcription initiation activity
(TFIIH). Heparin fractionation experiments revealed that HHR23A and a large
fraction of HHR23B resided in the flowthrough fraction (Fig.
4
A). Native gel electrophoresis indicated that the vast majority of both HHR23
proteins was present in the free, non-complexed form (Fig.
5
) a finding supported by the gel filtration experiments (Fig
6
A). The heparin, phosphocellulose and size fractionation experiments as well as
the immunodepletion studies all confirmed the complex formation of XPC protein with HHR23B protein (Figs
4
A and B and
6
A). No HHR23A protein could be detected in these purified heparin or
phosphocellulose fractions containing XPC. From these findings we conclude that
for HHR23B two forms exist: the majority is in a free form, whilst a small
fraction is complexed with XPC. For HHR23A, we can only detect a free form,
although it is not excluded that a fraction below our detection level is
complexed with XPC or another protein. The absence of detectable quantities of HHR23A in
the XPC-HHR23B containing high-salt fractions from the heparin (Fig.
4
A) and phosphocellulose chromatography suggests that HHR23A may not be
functionally fully equivalent to HHR23B. The small fraction of HHR23B that is
complexed with XPC is necessary for NER (
10
). This raises the question whether HHR23A and the free form of HHR23B are
involved in NER at all and/or whether they have an additional function. These
proteins resided in the flow-through of the phosphocellulose fractionation. Previously, Aboussekhra
et al.
(
44
) showed that only the RP-A complex and the PCNA protein from this fraction are necessary for
in vitro
NER. These data therefore suggest that HHR23A does not play a role in the core
NER reaction. However, the
in vitro
system might not reflect the step in which this protein plays a role
in vivo
. If only (the XPC-bound) HHR23B has a role in NER, one might wonder why no rodent or human
mutants for HHR23B were found (Fig.
3
). A possible explanation for the absence can be the dual function, that might
give rise to an unexpected phenotype. Alternatively, HHR23A may bind to XPC
when HHR23B is absent. A clear answer on what is the function of the free form
of both mammalian RAD23 equivalents and whether they are functionally redundant
should come from analysis of mutants generated by gene targeting and from
in vitro
reconstitution experiments (both experiments in progress).
Gel filtration studies suggested that the XPC protein can be part of protein
complexes of large size (Fig.
6
A). The purified XPC-HHR23B complex was previously determined to have a molecular weight of
500-550 kDa by gel filtration and a value of 110 kDa by glycerol gradients (
8
). Here we found in fractionated Manley-type cell extracts a molecular weight bigger than these previously
determined values. This suggests that the XPC-HHR23B proteins are part of a bigger complex, that can easily fall apart
during purification in a stable XPC-HHR23B subcomplex and other proteins. However, it cannot theoretically be
excluded that XPC protein selectively multimerizes or aggregates. Therefore, it
was investigated whether the large molecular weight XPC-containing complex also includes TFIIH and ERCC1 components, conforming
with the `repairosome' model reported by Svejstrup
et al.
(
45
) for yeast NER. Non-denaturing gel electrophoresis showed distinct bands for HHR23A (70 kDa)
and HHR23B (150 kDa), for both the HeLa proteins as well as the
E.coli
-produced recombinant polypeptides. In this context, it should be noted
that native molecular weight estimations themselves should be taken with
caution, since these may strongly depend on the conformation of the proteins or
protein complexes. However, bands migrating at a similar position can provide
evidence for complex formation. Therefore, it is evident that both native
molecular weights are different from the 280 kDa previously described for the
ERCC1 complex (
34
) and the minimal molecular weight calculated for TFIIH.
Fractions containing highly purified ERCC1 complex described by van Vuuren
et al
. (
38
) were also checked for the presence of HHR23 proteins, and were found to be
negative. This makes a tight association of these proteins with the ERCC1
complex highly unlikely.
From the data presented here one can also conclude that there is no stable
complex of a significant fraction of XPC-HHR23B and TFIIH, under these conditions and in our (Manley) extracts.
This is in conflict with the findings of Drapkin
et al.
(
36
), who after six purification steps for TFIIH components, still detected XPC
protein in the purified fractions. However, their TFIIH complex is not
completely pure, and no data are provided with respect to the yield and the
fraction (percentage) of XPC present in the TFIIH preparation. Therefore, it is
hard to identify whether this is a significant amount of XPC and whether cross-contamination is excluded. From our unpublished results, it appeared that
the XPC complex by coincidence behaved in a similar way during several
purification steps as TFIIH. Moreover, no physical interaction (e.g.
immunodepletion) was shown by Drapkin
et al.
(
36
). On the other hand it should be stressed that our data on XPC/TFIIH interaction do not exclude a transient association, as
reported by Bardwell
et al.
(
46
) for the
S.cerevisiae
system. The absence of any detectable interactions between these factors is of
relevance in the context of the evidence for a `repairosome' in
S.cerevisiae
(
45
), in which (almost) all NER components are represented in one super complex. A
difference might exist between yeast and mammals. Alternatively, a `NERosome'
in mammals may be more delicate, and might dissociate sooner than its yeast
counterpart. Therefore, the extract preparation procedure can be of crucial
relevance. The Manley extracts are made from non-UV irradiated cells. Possible interactions of repair proteins might not be
visible in extracts of non-damaged cells.
The subcellular localization of a protein can provide possible clues about its
function. Immunofluoresence data displayed a clear nuclear localization of the
XPC protein on interphase cells (Fig.
7
). This finding supports a function in DNA metabolism and is in agreement with
the previously described ssDNA binding activity (
8
). The observation that during anaphase and telophase, XPC specifically
associates with chromatin, suggests a role for XPC after the metaphase/anaphase
transition, and is consistent with the DNA-binding activity of the XPC-HHR23B protein complex. However, such a function remains to be
clarified.
Nuclear localization was also found for both HHR23 proteins, consistent with a
role for these proteins in DNA or chromatin metabolism. These results are in
accordance with previous
S.cerevisiae
data, in which the RAD23 protein was described to be nuclear (
15
). Both recombinant HHR23 proteins were found not to have specific affinity for
ss or dsDNA, and, therefore, a direct role for these proteins in DNA damage
recognition can be regarded as unlikely. On the other hand, recent analyses of
dimer and 6-4 photoproduct repair in
rad23
deletion mutants in
S.cerevisiae
which are only intermediately UV-sensitive indicated that removal of both lesions is strongly impaired for
both TCR and GGR (
46
). This implies an important role of this protein in both NER subpathways in
yeast. However, it is unknown at present whether RAD23 is in excess over RAD4
and consequently whether a free form of RAD23 exists in
S.cerevisiae.
Additional functions for both HHR23 proteins seem likely however, based on the
excess of both HHR23 proteins in the cell compared to XPC, as deduced from the
experiments presented in this paper.
We are grateful to Dr Michael McKay for helpful suggestions and critical reading
of the manuscript. Sigrid Swagemakers is acknowledged for her help with the COS-1 transfections and immunofluoresence experiments and Dr J.-M. Egly (Strasbourg) is acknowledged for some of the antibodies used
in this work. We thank Mirko Kuit for photography. The work at the Department
of Cell Biology and Genetics of the Erasmus University was financially
supported by the Medical Genetics Centre South-West Netherlands and in part by a Human Frontier Science Program Research
grant. KS was supported by a grant of the Biodesign Research Program of the
Institute of Physical and Chemical Research (RIKEN).
&form=6&uid=96178000&Dopt=r">MEDLINE Abstract
&form=6&uid=96178000&Dopt=r">MEDLINE Abstract
REFERENCES
Return