Human Molecular Genetics, 2003, Vol. 12, No. 13 1507-1522
DOI: 10.1093/hmg/ddg174
© 2003 Oxford University Press
True XP group E patients have a defective UV-damaged DNA binding protein complex and mutations in DDB2 which reveal the functional domains of its p48 product
-Otrin1,
1Department of Molecular Genetics and Biochemistry, University of Pittsburgh School of Medicine, and University of Pittsburgh Cancer Institute, Pittsburgh, PA 15213, USA, 2Istituto di Genetica Molecolare CNR, Via Abbiategrasso, 207-27100 Pavia, Italy and 3Section on DNA Replication, Repair and Mutagenesis, National Institute of Child Health and Human Development, NIH, Bethesda, MD 20892-2725, USA
Received February 14, 2003; Revised May 2, 2003; Accepted May 7, 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Xeroderma pigmentosum (XP) is a skin cancer-prone autosomal recessive disease characterized by inability to repair UV-induced DNA damage. The major form of XP is defective in nucleotide excision repair (NER) and comprises seven complementation groups (A–G). The genes defective in all groups have been identified unambiguously with the exception of group E. The cells of some XP-E patients are deficient in a protein complex (consisting of two subunits: p127/DDBI and p48/DDB2) which binds to UV-damaged DNA (UV-DDB) and is specifically involved in the removal of photoproducts from the non-transcribed regions of the genome. However, other XP-E patients have been reported not to lack UV-damaged DNA binding activity (DDB+). Here we describe several genetically unrelated XP-E patients, not previously analyzed in depth, each carrying two mutated alleles for DDB2, causing either a single amino acid change or a protein truncation or internal deletion. These defects result in a severe decrease of detectable p48 protein, abolish interaction with the p127 subunit, and produce a deficiency in UV-DDB binding activity (DDB-). The role of p48 in the repair defect of these patients was demonstrated in vivo and in vitro. Investigation of four DDB+ cell strains from patients previously assigned to XP-E, allowed us to reclassify all of them into other groups and to show that they do not share the molecular and biochemical features typical for XP-E. Besides confirming that the true XP-E phenotype is DDB-, resulting from defects in a single gene, DDB2, our results identify the functional domains of the corresponding p48 protein.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Xeroderma pigmentosum (XP) is a rare autosomal recessive disease which is characterized by hypersensitivity to sun exposure, pigmentary alterations and premalignant lesions in the sun-exposed areas of the skin, an extremely high incidence of skin cancer, an increased incidence of tumors in other organs and tissues, and in some cases, neurological abnormalities. The clinical and cellular photosensitivity in XP is due to alterations in either nucleotide excision repair (NER) or trans-lesion synthesis (TLS), the two systems that in human cells remove and tolerate, respectively, damage induced in DNA by UV light. Defects in TLS account for a minority of XP patients, which constitute the XP variant group (XP-V), characterized in general by late onset and slow progression of clinical symptoms, usually without any sign of neurological involvement. In most XP patients the biochemical defect is in NER, the principal pathway for removal of a broad spectrum of structurally unrelated lesions such as UV-induced cyclobutane pyrimidine dimers (CPD), 6-4 photoproducts (6-4P) and numerous chemical adducts. The NER system has two distinct subpathways: transcription-coupled repair (TCR), which rapidly removes lesions from the transcribed strand of active genes, and global genome repair (GGR), which processes lesions in the non-transcribed DNA of the genome at a slower pace (1). XP patients, defective in NER, have varying degrees of deficient UV-induced DNA repair synthesis (unscheduled DNA synthesis, UDS). The deficiency in UDS has been used as a cellular parameter in classical genetic complementation tests based on somatic cell hybridization to demonstrate the presence of genetic heterogeneity in the disease. Seven complementation groups, designated XP-A to XP-G, have been identified. Each complementation group (except groups B, D and G) shows distinctive clinical and cellular features which are present in the majority of the patients classified in the group. A direct correlation is generally found between the degree of the UV hypersensitivity and the severity of clinical symptoms: the progression of the disease is usually faster and neurological abnormalities occur in the patients showing the highest photosensitivity (2,3).
In recent years, the genes and their encoded proteins which, when defective, yield the XP groups A–G, have been identified (4,5); these proteins have a primary function in the damage recognition, preincision or incision stages of NER (6–8). However, the molecular basis of the XP-E phenotype is still ambiguous (9). XP-E is the mildest and one of the least common forms of the disorder. The cells from XP-E patients show only a partial hypersensitivity to UV irradiation, with the UDS level reduced to about 50% of normal (10, and references therein) as a result of a partial deficiency in GGR (11). The cells from seven patients have been reported to lack a UV-damaged DNA binding activity (DDB-) associated with a protein complex, UV-DDB (10,12,13). However, another nine putative XP-E patients have been reported who retain this DNA binding activity (DDB+) (14,15). The UV-DDB complex comprises two subunits, p127 and p48 (16,17), also known as DDB1 and DDB2, and their genes are DDB1 and DDB2, respectively (5,18). Analysis of the cDNAs for both subunits of UV-DDB has revealed no mutations in either subunit in any DDB+ cell strains examined to date. Sequence changes have only been found in the DDB2 gene in cells from DDB- patients (13,19). Moreover, restoration of the UV-DDB activity in vivo was achieved when hamster cells, which express only p127 and have a DDB--like phenotype, were transfected with human normal, but not mutated, DDB2 cDNA (20).
A high affinity for UV-damaged DNA in vitro (16,21–23), and induction of this binding activity upon UV treatment of mammalian cells (11,17,24), originally suggested a role for UV-DDB in damage recognition. However, UV-DDB seems to have only an accessory role in NER in vitro when chromatin-free DNA is the substrate (25–27). Nevertheless, UV-DDB does complement the XP-E defect when microinjected into DDB- cells (10,28). More recent work has shed some light on this confusing state of affairs by indicating a role for UV-DDB in NER of chromatinized DNA (29–31), in particular the repair of CPD photoproducts (27).
Growing evidence links mutations in the DDB2 gene with the DDB- XP-E phenotype, and reinvestigation of the cells from three patients originally assigned to XP-E has suggested an incorrect assignment of DDB+ patients to group E (32). However, a significant number of DDB+ cell strains remains unresolved, and a limited number of newly reported cases assigned to DDB- make the group E genotype–phenotype relation confusing. Here we have re-examined four DDB+ cell strains originally assigned to XP-E. In addition we describe an extensive analysis of the cellular, molecular and biochemical properties of three recently assigned DDB- patients (10) and one genetically unrelated XP-E patient. Our results confirm that the true XP-E phenotype is DDB-, associated with a mutant DDB2 gene, and that patients with the DDB+ phenotype have been incorrectly assigned to XP group E. Furthermore, in addition to clarifying the genotype–phenotype relationships in XP-E patients, our study offers new insight into the effects that specific mutations in the DDB2 gene have on the functionality of p48, on the cellular amounts of both subunits of UV-DDB, and on the activity of the entire complex in NER.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Investigation of four putative DDB+ cell strains
We re-examined the repair capacity of fibroblasts from four DDB+ XP-E patients available to us: XP80TO, XP81TO (33), XP93TO (34) and XP95TO (15). This analysis was performed in parallel with two DDB- XP-E strains, XP2RO (35) and XP82TO (33).
An altered response to UV irradiation was observed in two of the four analyzed DDB+ cell strains, XP81TO and XP80TO, in which UDS levels ranged between 20–25% and 31–39% of normal, respectively. The UDS levels in these two DDB+ strains were somewhat lower than those detected in the two DDB- strains (ranging between 41–46% and 45–53% of normal in XP82TO and XP2RO, respectively). Cells from these NER deficient strains were fused with XP cells representative of distinct complementation groups (Fig. 1). Complementation was observed with all fusion products except in heterodikaryons between XP80TO or XP81TO and XP-F cells and, as expected, in heterodikaryons between XP82TO or XP2RO and XP-E cells. As well as confirming that XP82TO and XP2RO belong to the XP-E group, these findings indicate that XP80TO and XP81TO patients belong to the XP-F group. In agreement with this finding, restoration of the repair defect was observed in XP80TO fibroblasts following microinjection of a construct expressing the wild-type XPF protein, with UDS levels dramatically increasing from 30% of normal in non-injected cells to 70–100% of normal in injected cells. The repair defect in XP80TO and XP81TO patients was further characterized by determining the quantity of XPF and ERCC1 proteins (these proteins constitute a heterodimeric complex which in NER effects the incision of damaged DNA 5' to the damaged site). In both cell extracts the XPF protein was undetectable and the level of ERCC1 was significantly decreased, as in the reference XP-F fibroblasts XP2YO. In the DDB- XP-E fibroblasts, XP82TO, both proteins were expressed at the levels seen in normal cells (Fig. 2A). Taken together, the results shown in Figures 1 and 2A clearly indicate that XP80TO and XP81TO patients must be re-assigned to the XP-F group.
|
|
In fibroblasts of XP93TO and XP95TO, the other two reinvestigated DDB+ cell strains, the capacity to perform UDS was similar to that observed in normal donor cells analysed in parallel. These results were reproducible in different experiments: UDS levels following irradiation with 20 J/m2 ranged between 98–109% and 91–98% of normal in XP93TO and XP95TO cells, respectively. These findings indicate that the patients XP93TO and XP95TO are characterized by a normal capacity to perform DNA repair following UV damage and therefore they do not have the form of XP characterized by defects in NER.
We recently demonstrated that UV-mediated degradation of p48 occurs independently of the cell's proficiency for NER, but recovery of p48 levels, and UV-DDB binding activity at later times (12 h and thereafter) seems to be dependent on the cell's capacity to repair non-transcribed DNA. The recovery of p48 is not detected 24 h post-UV in cells derived from most of the XP groups. The cells from XP-V and Cockayne's syndrome (CS) patients, which are deficient in TLS and in the TCR subpathway of NER, respectively, have kinetics of p48 recovery comparable to those found in normal human fibroblasts (31,36). Given the likelihood that the UV-induced degradation and recovery of p48 might serve as a diagnostic tool for discrimination between GGR-deficient and GGR-proficient XP cells, we measured the kinetics of post-UV expression of p48 in XP93TO and XP95TO fibroblasts. Consistent with the normal UDS activity in both cell strains, as in control fibroblasts, the p48 level was reduced 1 h after UV, and recovered by 24 h (Fig. 2B). In addition, the kinetics of post-UV recovery of the UV-DDB binding activity in XP95TO and XP93TO cells were indistinguishable from those of normal fibroblasts (data not shown). These data together support the notion that XP93TO and XP95TO do not belong to the NER-deficient form of XP.
Clinical and cellular characteristics of newly identified XP-E patients
After we reported three new DDB- XP-E cell strains, from two Italian patients (XP23PV and XP25PV) and one formerly classified as XP-V (GM01389) (10), we identified one new genetically unrelated XP case from Italy, XP27PV, that was assigned by genetic analysis to the XP-E group (Fig. 3A). The three Italian patients are characterized by the mild clinical and cellular features typical of the XP-E group. They all show only mild dermatological manifestations, no neurological abnormalities, and late onset of tumors (as detailed in the case reports). Fibroblasts from these patients show a partially reduced ability to perform UV-induced repair synthesis, with UDS levels ranging between 40 and 65% of normal (Fig. 3B), and survival levels significantly affected only at high UV doses (Fig. 3C).
|
To further characterize the new XP-E case, the UV-DDB-binding capacity of a whole-cell extract from XP27PV lymphoblastoid cells was compared in an electrophoretic mobility shift assay with those of the three previously identified DDB- XP-E cell strains, and the normal human lymphoblastoid cell line GM01953. As with cells from the XP-E patients, XP23PV, XP25PV and GM01646 (the lymphoblastoid cell line from the patient whose fibroblast strain is coded GM01389), the UV-DDB binding activity was undetectable in XP27PV cells (Fig. 3D).
Genomic structure and mutational analysis of DDB2
The genomic structure of the DDB2 gene is summarized in Table 1. The DDB2 gene, which is located on human chromosome 11p11.2, spans a region of approximately 24–26 kb and includes 10 exons. The sizes of the exons range from 46 to 413 nt; with the exception of intron 3, which is 15 768 nt long, the DDB2 introns have a size in the range of that for most human introns (37). We have sequenced all of the internal exons, the full introns 1, 2, 4, 5, 6, 8 and 9, and part of introns 3 (275 nt at the 5' region and 215 nt at the 3' region) and 7 (512 nt at the 5' region and 383 nt at the 3' region). Their sequences correspond to that reported in GenBank, accession AC090589.
|
To define the molecular defect, total RNA from the four XP-E patients, XP23PV, XP25PV, XP27PV and GM01389, was reverse-transcribed and the whole DDB2 cDNA was amplified in two overlapping fragments (corresponding to the cDNA regions 17–994 and 825–1766 nt, respectively), which were then directly sequenced. In the Italian patients the genomic DNA regions containing the mutations were also analyzed, and the pattern of inheritance of the alleles was established by sequencing of the relevant DNA regions of the parents. The results of our analysis are described below and are summarized in Table 2.
|
XP23PV
Amplification of the 3' region of the DDB2 cDNA (nt 825–1766) from XP23PV family members produced a fragment of a size shorter than normal in the patient and a fragment of normal size and a shorter one in his parents (Fig. 4A). Sequencing of the short fragment demonstrated that it was missing 321 bases, from positions 878 to 1198, corresponding to exons 6 and 7 (Fig. 4B). This deletion could arise either as a splicing abnormality or as a genuine deletion in genomic DNA. Analysis of the genomic DNA region including exons 5–10 identified a G to T transversion in the first base at the 5' end of intron 7 (Fig. 4C). The patient was homozygous and the parents were heterozygous for this mutation, which would destroy the splice donor site for intron 7. The loss of the 107 amino acids (residues 235–341) encoded by exons 6 and 7 is likely to destroy the activity of the protein completely.
|
XP27PV
Three different deletions were identified in the DDB2 cDNA of the patient XP27PV. Direct sequencing of the nucleotide fragment 17–994 revealed the occurrence of a 4 base deletion that could have arisen from the deletion of the quadruplet 901–904, 902–905, 903–906, 904–907 or 905–908 (Fig. 4D). This deletion causes in the transcript the loss of the bases AAAG at position 905–908, resulting in the change of the residue 244 in a stop codon (Lys244opal). The patient was homozygous and her parents were heterozygous for this deletion.
Amplification of the 3' region of the DDB2 cDNA (nt 825–1766) produced a fragment of normal size and two shorter fragments in the patient as well as in her parents (Fig. 4E). In the normal sized fragment, the 4 bp deletion at position 905–908 was present in the homozygous state in the patient and in the heterozygous state in her parents. In the two shorter fragments, two distinct large deletions were observed involving a 178 bp deletion of nucleotides 878–1055, corresponding to exon 6, that results in a frameshift from residue 235 and in a stop codon 10 codons downstream, and a 321 bp deletion of nucleotides 878–1198, corresponding to exons 6 and 7, that results in the loss of amino acids 235–341.
Amplification and sequencing of the genomic DNA region including exons 5–10 did not show the presence of any mutation other than the loss of the 4 bases at position 905–908. The patient was homozygous and her parents were heterozygous for this deletion (Fig. 4F). No other mutations were observed, suggesting that this alteration is responsible for the abnormally spliced products observed in the patient and her parents.
XP25PV
Complete sequencing of the DDB2 cDNA in the patient XP25PV revealed the presence, in the whole of the amplified population, of both a silent G to A transition at position 1093 and of a G to T transversion at position 1094, resulting in an asp307tyr substitution. The parents were heterozygous for these mutations, confirming that the patient is a homozygote (Fig. 4G).
GM01389
Complete sequencing of the DDB2 cDNA in the patient GM01389 confirmed the previous finding (17) that this patient is heterozygous for two changes: a T to C transversion at position 1224, resulting in a Leu350Pro substitution, and a 3 base deletion, resulting in loss of Asn349 (Fig. 4H). The latter was proved by sequencing individual clones, and could have arisen from the deletion of triplet 1218–1220, 1219–1221 or 1220–1222 (Fig. 4I).
DDB2 or the p48 subunit complement the XP-E defect in vivo and in vitro
To directly confirm that the partial repair deficiency in the three Italian XP-E patients is caused by mutations in the DDB2 gene, we measured UV-induced repair synthesis in fibroblasts following microinjection with a construct expressing the normal p48 protein. A 1.5–2-fold increase in the UDS level was observed in the injected cells, compared to non-injected cells (Table 3 and Fig. 5A). In contrast, no effect on the UDS levels was seen when the expression vector containing the wild-type DDB1 cDNA was injected into XP-E cells. Finally, addition of the normal p48 subunit, purified as a Flag-fusion p48 protein, restored UV-DDB binding activity in extracts of XP27PV lymphoblastoid cells tested in an electrophoretic mobility shift assay (Fig. 5B). The presence of Flag-p48 in the complex with UV-damaged DNA was confirmed by a partial or complete super-shift of the complex with anti-Flag and anti-p48 antibodies, respectively.
|
|
Mutations in DDB2 result in the loss of p48 protein
We examined the quantity of UV-DDB in the newly characterized DDB- and other XP-E cases under investigation in this study, including the four DDB+ patients, XP80TO, XP81TO, XP93TO and XP95TO, that we have found to be erroneously assigned to XP-E. The relative amounts of the subunits, p48 and p127, in whole cell extracts prepared from fibroblasts were determined by immunoblotting with anti-p127 and anti-p48 antibodies (Fig. 6A). The level of p127 in all of the patient cell strains was comparable to that found in normal cell extracts when polyclonal antibodies raised against either the N-terminus or the C-terminus of p127 were used. In contrast, notable differences in the amount of p48 were observed among the XP patients' cells. The endogenous level of p48 approximates that observed in normal cells in all four DDB+ patients, whereas different degrees of reduction were observed in the six XP-E DDB- patients. A nearly normal level of p48 was observed in XP82TO fibroblasts while reduced amounts, of about 90 and 80% of the normal level, were detected in XP25PV and GM01389, respectively. Almost undetectable levels of p48 were present in XP2RO, XP23PV and XP27PV fibroblasts. A polyclonal antibody prepared against residues 5–24 of human p48 failed to immunodetect the truncated p48 proteins predicted from the allelic sequences of the patients XP23PV and XP27PV, when whole cell extracts of fibroblasts or nuclear extracts from lymphoblastoid cells were probed. However, using different methods for preparation of the lymphoblastoid cell extracts (see Materials and methods), we detected a slightly faster migrating, presumably truncated, p48 at 2–3% of the normal level in the XP23PV and XP27PV cells. Immunoblotting revealed additional bands which we ascribe to degradation of p48 in extracts from these patients' cells (Fig. 6B). The migration of XP23PV and XP27PV mutated proteins on SDS–PAGE did not correspond to their deduced size of 35.5 and 27.9 kDa, respectively. Considering that intact p48 migrates on SDS–PAGE as a protein of 41 kDa, it is difficult to explain this observation without knowledge of its structure. Certainly, low levels of p48 proteins in these XP-E cell lines suggest that these mutated forms must be very unstable.
|
Mutations in DDB2 abolish the formation of the UV-DDB complex
We wondered if the newly characterized mutations in DDB2 affect the interaction of p48 with p127, as an explanation for the DDB- character of these cells. To test for the presence of the UV-DDB complex, we carried out a co-immunoprecipitation assay. The cellular extracts from normal donors (GM01953 and GM05757), and four XP-E patients (GM01646, XP25PV, XP3RO and XP82TO), were immunoprecipitated with rabbit anti-p48 antibody coupled to an AminoLink gel, and immunoprecipitates were tested for the presence of both subunits. Under these conditions, except in the extract from XP82TO cells, we could not detect p127 co-immunoprecipitated with p48 proteins in DDB- lines (Fig. 6C), which suggests that the molecular defects in these mutant lines prevent formation of a UV-DDB complex. Our results on the cells from XP82TO and XP3RO patients confirmed the presence and absence of a stable interaction between the UV-DDB subunits, respectively, as shown previously for ectopically expressed mutant proteins (38).
p127 is depleted in the nuclei of XP-E cells
As is true of other NER proteins, p48 is exclusively localized in the nucleus, while variable amounts of p127 are found in the cytosol as well as the nucleus (17,38–40). p48 seems to be a regulatory factor for the nuclear localization of p127. The amount of p48 and its functional status determine nuclear entry of p127, presumably for the purpose of UV-DDB complex formation (38,41). We analyzed the distribution of p127 between cytosol and nuclei in all four lymphoblastoid XP-E cell lines (Fig. 6D). As expected, mutations in DDB2 that result in the severe loss of p48 protein also affect subcellular localization of p127 such that the nuclear extract is depleted and more p127 is immunodetected in the cytosolic extracts of these cell lines. Further fractionation of the nuclei isolated from two mutant cell lines, XP25PV and GM01646, showed a large reduction of p127 in Triton-resistant high-salt nuclear fractions (not shown). The pattern of expression of p48 in the lymphoblastoid cells was the same as seen in primary fibroblasts, but we always detected small amounts of p48 in the cytosolic fractions of XP25PV and GM01646. It is unlikely that the cytosolic p48 results from leakage of p48 from the nucleus because the amount of p48 detected in the cytosolic fractions in the mutant cells was disproportionately high, compared to normal fibroblasts. In addition, antibody to cyclin B1, assayed as a control for proteins localized in nuclei, did not reveal any signal in cytosolic fractions.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Mutation pattern and genotype–phenotype relation in XP-E patients
Mutations in the DDB2 gene have been found in all five previously reported XP-E patients who are DDB- (13,17,19), whereas no mutations in either UV-DDB subunit have been detected in any of the DDB+ XP-E cell strains examined to date. Of all XP-E patients with known mutations, two (XP2RO and XP3RO) are genetically related (35). The three Italian patients reported here have mutations in the DDB2 gene that have not been described previously. The results reported in previous publications (13,17,19), and in our work are summarized in Figure 7. Various types and sites of changes in the p48 protein result in the XP phenotype and all of them are located in the C-terminal half of the protein. Any founder effect is indicated by the mutation pattern in the patients: the same inactivating mutation was found only in XP2RO and XP3RO, who are first cousins. Furthermore, mutations in the heterozygous state are present only in one case (GM01389), and by analysis of individual clones we demonstrated that the mutations found in this patient are on different alleles (Fig. 4I). All of the remaining patients are homozygotes for the causative mutation, suggesting that they were all born from consanguineous marriages. Although consanguinity has not been reported in the family histories, we confirmed the homozygosity for the mutated DDB2 alleles in the Italian cases by analysis of the patients' and their parents' genomic DNA.
|
The mutations observed in the patients result from different events. They include a deletion in a repetition of an identical sequence (XP27PV), presumably resulting from replication slippage; a C to T transition at a CpG site (Ops 1) as a consequence of demethylation of 5-methylcytosine; and mutations located in the splice sites (XP23PV) or affecting the splicing indirectly by interfering with the stability of the transcript. The latter would be the case for the transcripts with the deletion of either 878–1055 or 878–1198 nt found in association with a normal sized transcript containing the AAAG905–908 deletion (XP27PV). This was the only inactivating change observed at the genomic level in XP27PV family members. The 905–908 nt deletion is located towards the 5' end of exon VI and could induce some change in the secondary structure of the mRNA, such that the splice acceptor sites of introns 6 and 7 might be used instead of the intron 5 site. Since the patient is homozygous, the presence of a normally spliced product indicates that the splice acceptor site of intron 5 is also used. The region including the 4 nt deletion in XP27PV was also checked for exonic splicing enhancers (ESE, i.e. specific short oligonucleotide sequences that enhance pre-mRNA splicing when present in exons) by the RESCUE-ESE approach, a computational method that predicts which sequences have ESE activity (42). Four out of the six ESE hexamers predicted in the region 899–912 nt of the DDB2 cDNA are abrogated by the 4 nt deletion found in XP27PV.
Altered as well as truncated DDB2 proteins result from the mutations found in the eight XP-E patients so far analyzed. Three out of the four missense mutations result in the non-conservative lys244glu, asp307tyr and leu350pro substitutions. The former two are located in the third and fourth putative WD repeats, respectively, detected by the Pfam prediction method. The fourth missense mutation results in a conservative arg273his substitution that is very close to a putative WD repeat as predicted by Pfam (amino acids 233–272) as well as by Prosite (amino acids 258–272) methods.
Three of the nine mutations observed at the cDNA level are predicted to cause premature termination of the p48 protein as a result of a nonsense mutation (Ops 1) or frameshifts (XP27PV). A null product is likely to result also from the abnormal DDB2 transcript with a large in-frame deletion of 321 nucleotides (from position 878 to 1198, corresponding to the entire exons VI and VII) found in XP23PV and XP27PV. Although the corresponding protein should contain the DDB2 region used to raise antibodies, no product was observed on western blots (Fig. 6A), suggesting that the corresponding protein with the internal deletion is either unstable or is present in too small an amount to be detected (Fig. 6B).
Remarkably, the lack of the p48 protein and the presence of a mutated protein both result in similar clinical phenotypes and confer the same degree of cellular sensitivity to UV light in terms of survival and UDS. Furthermore, the newly diagnosed cases reported in this paper share the clinical and cellular features typically described in the XP-E group (i.e. mild dermatologic manifestations, no neurological abnormalities, late onset of tumors and a modest UDS reduction), thus confirming that the XP-E pathologic phenotype is homogeneous both at the clinical and cellular levels.
In three patients (XP23PV, XP27PV and Ops1), both alleles appeared to be affected by mutations that are likely to abolish the p48 function completely. This suggests that DDB2 is not essential for cell proliferation and viability. The presence of protein truncation mutations, as seen in DDB2, has also been found in other genes involved in NER, such as CSB (43,44), XPG and to an even greater extent, XPA and XPC (45 and references therein).
XP-E is a homogeneous group
The high level of residual NER activity, defined as UDS, in the cells from XP-E patients makes it very difficult to assign these cells by complementation tests and especially to distinguish them from two other XP groups with normal NER proficiency: XP-V and ‘UV-sensitive syndrome’ (UVsS). Consequently, the XP-E group until recently has been considered to be heterogeneous with DDB+ (nine cases) and DDB- (seven cases) subphenotypes (9 and references therein). We first found that the GM01389 cell line, initially classified as XP-V, had cellular and biochemical characteristics of the DDB- phenotype, and suggested the reassignment of GM01389 to XP group E (10). Further study revealed a molecular defect in the DDB2 gene of GM01389 fibroblasts (Fig. 3H and I) (17), confirming the XP-E genotype. A significant advance in resolving the heterogeneity issue in XP-E came from the study in which authors showed that three DDB+ patients, XP24KO, XP43TO and XP89TO, were wrongly classified in group E, and reassigned them to UVsS, XP-V and XP-F groups, respectively (32).
In this report we demonstrate that two DDB+ XP-E patients, XP80TO and XP81TO, belong to XP group F (Figs 1 and 2). Interestingly, cell lines from three Japanese patients, XP80TO, XP81TO and XP82TO, were originally classified as XP-E after fusion with XP24KO in a complementing cell assay (33). This finding is rather surprising considering that XP24KO has been reclassified as UVsS (32), and furthermore that XP82TO is a true XP-E patient (19,32 and this study). However, patients XP80TO and XP81TO had developed skin malignancies at the ages of 46 and 41 years, respectively, while patient XP82TO had not developed any skin cancer by 41 years (33).
We re-examined two more DDB+ XP-E patients, XP93TO and XP95TO, with no mutations either in the DDB2 or DDB1 genes (19). In both cell lines the ability to perform UV-induced DNA repair synthesis, the level of expressed p48 and p127 proteins (Fig. 6A) and the kinetics of post-UV expression of p48 (Fig. 2B) were similar to that in normal fibroblasts. With only limited information about their clinical and cellular features (15,34) we presume that these patients might belong either to the XP-V or to the UVsS group.
The data reported by us previously (10) and in this study increase the total number of true XP-E cases (DDB-) from four to eight (Fig. 7), and reduce the putative DDB+ XP-E group from six to two remaining cases (XP70TO and XP26KO, which are unavailable to us), greatly weakening the notion of XP-E as a heterogeneous group. Moreover, this observation, together with our extensive characterization of three newly identified XP-E cases, enables us to conclude that the mutations in the DDB2 gene which result in UV-DDB deficiency are the only defects responsible for the pathologic phenotype in patients who were proven to have been correctly assigned to the XP-E group.
Functional domains of the p48 protein
This study on naturally occurring mutations in XP-E patients provides an initial understanding of the functional domains of the p48 protein. Little was known of the p48 domains previously due to the limited number of characterized DDB2 mutants (4) and limited deletion mapping (46). In contrast to the DDB1 gene, which is well conserved from S. pombe to humans, the DDB2 homolog has only been described in the mouse (47). This circumstance precludes identification of the evolutionarily conserved sequences or putative domains in DDB2. Moreover, besides forming a heterodimer with p127, very little is known about the other proteins with which p48 directly interacts (46,48).
Our data show that mutations of DDB2 affect the interaction of p48 with p127. All substitutions and a small deletion of p48 responsible for the loss of binding to p127 mapped to the exons VI, VII and VIII (XP2RO, XP25PV and GMO1389 mutants), strongly suggesting that the p127-binding domain is confined to these three exons (Figs 6C and 7). We assume that the deletions in XP23PV and XP27PV cells result in a disrupted interaction with p127 as well, although we could not perform reliable co-immunoprecipitation assays due to the almost undetectable level of mutated p48 (Fig. 6A, B and D). Our results (Fig. 6C) are consistent with the previously published report on deletion analysis of p48 that implicated the C-terminal region of p48 as essential for binding to p127 (46). Disruption of the interacting interface of p48 with p127 abolishes UV-DDB binding activity (Fig. 3D), giving rise to the DDB- phenotype.
Interestingly, the XP82TO mutant retains the interaction of p48 with p127, but is deficient in DNA binding activity. So far this is the only DDB- mutant with such biochemical properties. The fact that a single substitution (lys244glu) (19) has such a profound effect on binding of the UV-DDB complex highlights the significance of this amino acid residue, and implies that the UV-damaged DNA binding domain resides upstream from the beginning of exon VI.
Remarkably, no mutations were identified in the N-terminus region of the p48 protein among eight XP-E patients analyzed so far. This could simply be circumstantial, or it may be that modification in this region will cause minor structural p48 alterations and result in a phenotype so mild as to elude a clinical diagnosis of XP.
The p48 protein contains two types of structural domains/motifs: WD40 repeats and nuclear localization signals (Fig. 7). Three putative nuclear localization signals (NLS) have been identified in exons I and VI of p48: Two overlap between residues 3–7 (PKKRP) and 241–244 (HKKK) (38), which correctly predicts the nuclear localization of wild-type p48 (17,38). The substitution in XP82TO, and large deletions in XP23PV and XP27PV, result in the loss of the third putative NLS. However, nuclear localization of the XP82TO mutant (17) implies that the HKKK sequence is dispensable for the nuclear entry of p48. In two other mutants, the low level of p48 detected in whole cell extracts was below immunosensitivity when cytosolic and nuclear fractions were probed (Fig. 6B and D). The substitution (asp307tyr) in XP25PV or deletion 349 and substitution (leu350pro) in GM01389 map outside of the NLS which weakens the possibility that the small amounts of p48 detected in the cytosolic fractions of these mutant cells (Fig. 6D) are due to impaired nuclear entry of defective protein. Rather, these small amounts are likely due to instability of a UV-DDB complex containing the mutated p48 subunit.
The nuclear localization of p48 and the absence of a nuclear localization signal in the DDB1 sequence implicate p48 in the nuclear import of p127 through an unknown mechanism. The changes in the DDB2 sequences in all four mutants (XP23PV, XP25PV, XP27PV and GM01646; Fig. 4D and Table 2) which result in the loss of most p48 protein (Fig. 6A) and the disruption of its interaction with p127 (Fig. 6C) can account for the depletion of p127 in their nuclei. This explains our earlier observation that the extracts from XP23PV and GM01646 cells, prepared for an in vitro DNA repair assay, contained reduced p127 levels compared with that of repair-proficient 705ori lymphoblasts (10). Although Shiyanov et al. (38) reported that the XP82TO mutant failed to stimulate nuclear entry of p127, we found that the lys244gly transition had no effect on the p48 level nor interaction with p127 compared with normal fibroblasts (31, and Fig. 6A and C). This discrepancy is most likely due to the ectopically versus endogenously expressed UV-DDB subunit, as utilized in Shiyanov et al. (38) and our study, respectively.
The p48 protein contains five putative WD40 repeats positioned downstream from the second exon (Fig. 7). The WD motif identified in p48 is characteristic of proteins involved in the recognition of chromatin proteins (49), which is in agreement with the role proposed for UV-DDB in NER in a chromatin context (29–31).
XP-E and NER
The connection between the UV-DDB and the XP-E group was first made after the discovery that DNA binding activity was missing in the cell extract of the first diagnosed XP-E patient, XP2RO (12). Further evidence that UV-DDB is functionally involved in the XP-E phenotype came from microinjection of the purified UV-DDB complex into DDB- XP-E cells; the correction of UDS activity correlated with the abundance of the p48 subunit in the microinjected complex (10,28). Here we demonstrated directly that microinjection of the expression vector for DDB2 but not for DDB1 into fibroblasts from the three Italian XP-E patients complements NER in vivo, measured by UDS (Table 3 and Fig. 5A). Furthermore, complementation of the UV-DDB activity in vitro was obtained when the wild-type p48 subunit was mixed with an extract from the DDB- cell strain, XP27PV, used in the microinjection assay (Fig. 5B). The rescue of binding activity in vitro was shown previously for the GM01389 mutant (17). This study confirms that the underlying defect of XP-E is in DDB2, although the precise function of the p48 protein and the UV–DDB complex in NER is still uncertain.
Data obtained in vitro favors the binding of UV-DDB to a DNA duplex containing the (6–4) photoproduct over the weaker recognition of CPD, strongly suggesting that the UV-DDB protein recognizes a kink caused by the damage on naked DNA (16). Surprisingly, in a defined system with purified repair proteins, UV-DDB enhances the incision of CPD (27), but has very little stimulatory effect on the excision of (6–4) photoproducts (25,27). An in vivo study of two DDB- XP-E cell strains, XP2RO and Ops 1, showed a partial deficiency of GGR of CPD and (6–4) photoproducts, respectively (11,13). This finding is in agreement with the observation that overexpressed p48 in human cell lines with a low level of UV-DDB activity enhanced the in vivo removal of both photoproducts (27). The availability of the four new XP-E cell lines defined in this study should certainly help in resolving questions about the true target(s) of UV-DDB, and the molecular functions of this complex.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Case reports
This study was performed on 11 patients showing clinical symptoms typical of XP and classified by genetic analysis in the XP-E group. Detailed clinical and DNA repair data on several of these cases have been published and related literature references are reported in Table 4. We have reassigned four of these patients either to group XP-F (XP80TO and XP81TO) or to the XP form with normal NER proficiency (XP93TO and XP95TO). The three patients coded with the suffix PV represent all of the XP-E cases identified in Italy so far. In common with XP-E individuals described in the literature, none of the Italian cases show neurological abnormalities.
|
XP23PV, a male born in 1976, is the second child of healthy unrelated parents who come from the same small village in southern Italy. At 4 years of age, he experienced sunburn after exposure to sunlight. During childhood, he developed a photosensitive butterfly erythema on the face. On the same photoexposed area of the face, he developed the first basal cell carcinoma (BCC) at 16 years of age and, in the following 2 years, seven other BCCs that were surgically removed. His response to the skin erythema test following UVA and UVB was in the normal range.
XP25PV is a female, born in 1966 to healthy unrelated parents who have another phenotypically normal daughter and are living in a small village in northern Italy. She experienced sunburn after exposure to sunlight at 21 years of age. At age 22, she developed the first BCC. Subsequently, she developed four BCC and one squamous cell carcinoma (SCC) at 28 years of age. On clinical examination at age 29 years, she showed photophobia.
XP27PV is a female, born in 1962 to healthy unrelated parents who are living in Tuscany (Central Italy). She was referred to us when she was 35 years old, when she had already developed in the photoexposed areas of the skin several cutaneous tumours of different types (BCC, SCC and melanomas). She has a phenotypically normal daughter.
Cells and culture conditions
All cell strains and lines used in this study are summarized in Table 4. Primary fibroblast cultures were established from biopsies of unaffected skin obtained from the three Italian XP patients XP23PV, XP25PV and XP27PV. Cell strains from one healthy donor (C3PV), from the trichothiodystrophy (TTD) patient TTD8PV, representative of complementation group XP-D (50), and from the two XP patients XP20PV and XP30PV representative of complementation groups XP-A and XP-F, respectively (our unpublished observations), were used as reference strains in genetic analyses. Primary human fibroblasts were grown in Dulbecco's modified Eagle's medium (DMEM) supplemented with 15% fetal calf serum (FCS) and subcultured by trypsinization. Lymphoblastoid cell lines were established by Epstein–Barr virus transformation of peripheral blood lymphocytes from the patients and their parents. These cell lines were cultured in RPMI 1640 medium supplemented with 15% FCS in a 3% CO2 atmosphere.
Cellular response to UV irradiation
The response to UV light was analyzed in fibroblasts by measuring UDS and cell survival. Procedures for the evaluation of these cellular parameters are routinely used in our laboratories and have all been described previously. Briefly, UV treatment was performed with a Philips TUV 15 W lamp (predominantly 254 nm), giving a dose rate of 2 J/m/s. UDS was determined by counting the number of grains on at least 50 non-S-phase cells in autoradiographic preparations of cultures incubated for 3 h after UV in medium containing 3H-thymidine (3H-TdR, specific activity 25 Ci/mmol, Perkin Elmer Life Sciences/NEN, Boston, MA, USA) (51). Cell survival was analysed by measuring the viability in stationary phase fibroblasts (52).
Genetic analysis of the repair defect
Complementation analysis was performed by measuring the UDS level in hybrids obtained by fusing the patient's cells with XP reference strains, as previously described (52). Briefly, fibroblast strains used as partners in the fusion were grown for three days in medium containing latex beads of different sizes (0.8 and 1.7 µm) that were incorporated into the cytoplasm as a marker. The cells were fused using polyethylene glycol (PEG-4000, Merck, Hohenbrunn, Germany), incubated for 48 h at 37°C, UV irradiated (20 J/m2), labelled for 3 h in medium containing 3H-TdR, and processed for autoradiography. UDS was measured by counting the number of grains over nuclei in at least 25 homodikaryons (identified as binuclear cells containing beads of one size) and in 25 heterodikaryons (identified as binuclear cells containing beads of different sizes). Two cell strains were classified in the same group if no restoration of the UDS level was observed in the heterodikaryons.
Microneedle injection
Nuclear microinjection of cDNA was performed according to a method described elsewhere (53). Briefly, DDB1 and DDB2 were cloned into the mammalian expression vectors pCEP and pTaT (Invitrogen, Carlsbad, CA, USA), respectively, yielding the plasmids pCEP127 and pTaT48. Cells were seeded onto a coverslip containing a grid and cultured for 24 h at 37°C. Plasmid cDNA at a concentration of 100 µg/ml in phosphate-buffered saline (PBS) was injected into the nucleus of cells present in specific areas of the coverslip using a glass microcapillary and an Olympus inverted microscope equipped with a Narishige 1 MB microinjection apparatus (Narishige, Tokyo, Japan). For each experiment at least 100 cells were injected. Coverslips with injected and un-injected cells were cultivated for 24 h to allow expression of the injected cDNA before they were assessed for their capacity to perform UV-induced DNA repair synthesis by means of UDS. To identify p48 or p127 overexpressing cells, the cultures were fixed in 4% paraformaldeyde, incubated with polyclonal antibodies against either p48 (GPp48) or p127 (anti-p127), and then with an FITC-conjugated anti-rabbit immunoglobulin (Jackson ImmunoResearch Laboratories Inc., West Grove, PA, USA; diluted 1 : 100). Following identification of p48 or p127 overexpressing cells by fluorescence microscopy, slides were fixed in cold methanol and processed for autoradiography. XP-F cells were microinjected with the plasmid pEF6 (XPF) and assessed for their capacity to perform UV-induced DNA repair synthesis by means of UDS 24 h after microinjection, as described above.
Sequence analysis of the DDB2 gene
Fibroblasts, 2x107, or 108 lymphoblastoid cells were resuspended in 1 ml of guanidinium thiocyanate buffer (4 M guanidinium thiocyanate, 25 mM Na acetate, pH 6) and RNA and genomic DNA were simultaneously extracted by a cesium chloride gradient centrifugation procedure. RNA was reverse-transcribed into cDNA with Moloney Murine Leukemia Virus Reverse Transcriptase (M-MLV RT). Briefly, 2 µg RNA in a total volume of 10 µl were heated to 90°C for 2 min and then 30 µl of a mix containing 1x first-strand cDNA buffer (Invitrogen, Carlsbad, CA, USA), 10 mM DTT, 1 mM dNTPs, 100 ng oligo(dT)15 (Promega, Madison, WI, USA) 1 U RNAsin (Promega), and 200 U M-MLV RT (Invitrogen) were added. After incubation at 37°C for 1 h the samples were heated to 95°C for 5 min and then stored at -20°C. The whole DDB2 coding region was amplified in two overlapping fragments. PCR amplification was carried out as follows: 20 µl cDNA were used in 100 µl reactions containing 1x GeneAmp buffer II (Applied Biosystems Roche, Branchburg, NY, USA), 2 mM MgCl2, 0.2 mM dNTPs, 100 pmol each of the required primers (Table 5) and 2.5 U AmpliTaq (Applied Biosystems Roche). Amplification was performed with one cycle at 94°C for 4 min and 35 cycles at 94°C for 1 min, annealing (at temperatures indicated in Table 5) for 1 min, and elongation at 72°C for 2 min.
|
Genomic DNA amplification was carried out on 1 µg DNA samples using the Gene-Amp XL-PCR kit (Applied Biosystems, Roche) in a reaction mix containing 1x XL buffer, 1 mM Mg(OAc)2, 0.2 mM dNTPs, 100 pmol each of the required primers (Table 5) and 4 U rTth DNA polymerase XL. PCR conditions were one cycle at 94°C for 4 min and 40 cycles each of 94°C for 1 min, and 60°C for 10 min. PCR products were separated and excised from 1 to 1.5% low-melting-point (LMP) agarose gels and then were purified by means of the Wizard PCR Preps Purification System (Promega, Madison, WI, USA).
The products from the primer pair P4 and P2 were cloned into the pMOSBlue T-vector by use of the pMOSBlue T-vector kit (Amersham Pharmacia Biotech, Buckinghamshire, UK) and double-stranded plasmid DNA was prepared by a mini alkaline lysis method.
Purified PCR fragments and plasmid DNA with appropriate inserts were directly sequenced by means of the Thermo Sequenase radiolabeled terminator cycle sequencing kit (USB Corporation, Cleveland, OH, USA) according to the manufacturer's instructions.
DDB2 sequencing was carried out by using the specific oligonucleotides listed in Table 6 and the primers P1, AP1, P4, P3, AP4, AP6, AP7 and P2 listed in Table 5.
|
Protein expression and purification
To overproduce FLAG-tagged-p48 proteins in insect cells, we designed a set of primers complementary to the N- and C-terminal portions of the protein sequence with FLAG epitope incorporated into 5' primers. The tagged full-length cDNAs of the human DDB2 genes was produced by PCR and inserted into pFastBack1 donor plasmid. The transformation of DH10Bac competent cells with the recombinant plasmid and production of recombinant Bacmid DNA and recombinant baculovirus particles were carried out according to the manufacturer's protocol (Invitrogen, Carlsbad, CA, USA).
Protein was expressed in Sf9 cells, maintained as an adherent culture in HyQSFX-Insect media (HyClone) supplemented with 0.5%FCS (HyClone). The infected cells were incubated at 27°C for 24, 48 or 72 h, and harvested at the indicated times by suspending in SDS–PAGE sample buffer. The optimum protein expression level, determined by immunoblotting with antibodies against p48 protein and Flag (Sigma, St Louis, MO, USA) peptide, was 48 h post-infection. The p48 protein, soluble in the cell-free extract, was purified with DEAE-Sepharose FF (AKTA design system, Amersham Biosciences Inc.) and anti-Flag affinity M2 gel (Sigma).
Preparation of anti-p127 antibody
A new p127 antibody was prepared against a synthetic peptide encompassing residues 408–425 of human p127 (NH2-KGLWPLRSDPNRETDDTL-COOH) in rabbits by Covance Research Products (VA, USA). p127N antibody was affinity purified by using a column of synthetic peptide coupled to AminoLink Plus Gel (Pierce) and concentrated with a Microsept 10 apparatus (Filtron Technology Corp.). Other UV-DDB antibodies used in this study were previously described: anti-p127 (54), GPp48 and p48N (31).
Immunoblotting and immunoprecipitation
Whole cell extract was made by resuspending cells, and harvested by scraping (fibroblast cells) or spinning (lymphoblasts), in Triton lysis buffer (TL buffer): 100 mM Tris–HCl, pH 7.5–500 mM NaCl–0.1% Triton X-100–2 mM EDTA–1 mM Na2VO4–1 mM NaF and a protease inhibitor cocktail (CompleteTM, Boehringer), and incubated on ice for 30 min. The supernatant, recovered after centrifugation at 16 000g for 10 min at 4°C, was used as the whole cell extract (wce). Working extensively with XP-E lymphoblastoid cells, we found that substitution of TL buffer with CHAPS lysis buffer (CL buffer: 1xPBS–10 mM CHAPS–1 mM Na2VO4–1 mM NaF–protease inhibitor cocktail) led to the recovery of slightly more p48 protein from these mutant cells. CL buffer was used to prepare wce from the XP23PV and XP27PV lymphoblastoid cells for immunodetection.
For co-immunoprecipitation of the UV-DDB subunits, the lymphoblastoid and fibroblast cells were lysed in NP-40 buffer (50 mM Tris–HCl, pH 8–100 mM NaCl–1% Nonident P-40–5 mM MgCl2–1 mM Na2VO4–1 mM NaF–protease inhibitor cocktail) for 1 h on ice, and centrifuged at 16 000g for 10 min at 4°C. Cell extracts were incubated with 15 µl rabbit anti-p48N antibody coupled to the AminoLink gel for 2–3 h at 4°C with gentle agitation. The gels were washed four times in lysis buffer and resuspended in gel-loading buffer. The co-immunoprecipitates were separated by 10% SDS–PAGE. The same percentage of SDS–PAGE was used to separate proteins for all immunoblotting analyses.
Western blotting analysis was performed using a chemiluminescent detection system (Tropix, MA, USA) as described (54) and the following antibodies: anti-p127, p127N, p48N, anti-cyclin B (Santa Cruz Biotechnology Inc.), anti-actin (Oncogen Science), and anti-FLAG (Sigma). Primary fibroblast cells were UV-irradiated, and the levels of p48 within 24 h after treatment were quantified, as described previously (31).
|
ACKNOWLEDGEMENTS |
|---|
We are grateful to Drs Gambini (Ospedale Galliera, Genova), L. Miori (Clinica Dermatologica, University of Pavia, Italy), and G. Vezzoni (Ospedale S. Giacomo, Massa, Italy) for providing us with cells and clinical details of the Italian XP patients; Dr R. Wood (University of Pittsburgh, PA, USA) for providing us with antibodies against XPF, ERCC1 and RPA/p70; Dr Beate Koeberle (University of Pittsburgh, PA, USA) for helping us with the immunodetection of the XPF and ERCC1 proteins; and Dr M. Takao (Tohoku University, Japan) for providing the plasmid pTaT48. This work was supported by Health Target Projects 2002 grant C161 and FIRB grant RBNE01RNN7 to M.S. and a University of Pittsburgh School of Medicine start-up grant to V.R.O. We gratefully acknowledge the support of the Italian Association for Cancer Research for making this collaborative study possible.
|
FOOTNOTES |
|---|
* To whom correspondence should be addressed at: University of Pittsburgh, Scaife Hall, Suite 401, 3550 Terrace Street, Pittsburgh, PA 15261, USA. Tel: +1 4126488975; Fax: +1 4126481236; Email: alevine{at}hs.pitt.edu
The authors wish it to be known that, in their opinion, the first two authors should be regarded as joint First Authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
-
de Laat, W.L., Jaspers, N.G. and Hoeijmakers, J.H. (1999) Molecular mechanism of nucleotide excision repair. Genes Dev., 13, 768–785.
[Free Full Text] - Cleaver, J.E. and Kraemer, K.H. (1995) Xeroderma pigmentosum and Cockayne syndrome. In Scriver, C.R., Beaudet, A.L., Sly, W.S. and Valle, D. (eds), The Metabolic and Molecular Basis of Inherited Disease. McGraw-Hill, New York, Vol. 3, pp. 4393–4419.
- Bootsma, D., Kraemer, K.H., Cleaver, J.E. and Hoeijmakers, J.H.J. (2002) Nucleotide excision repair syndromes: xeroderma pigmentosum, Cockayne syndrome, and trichothiodystrophy. In Vogelstein, B. and Kinzler, K.W. (eds), The Genetic Basis of Human Cancer. McGraw-Hill, New York, pp. 211–237.
- Cleaver, J.E., Thompson, L.H., Richardson, A.S. and States, J.C. (1999) A summary of mutations in the UV-sensitive disorders: xeroderma pigmentosum, Cockayne syndrome, and trichothiodystrophy. Hum. Mutat., 14, 9–22.[CrossRef][ISI][Medline]
-
Wood, R.D., Mitchell, M., Sgouros, J. and Lindahl, T. (2001) Human DNA repair genes. Science, 291, 1284–1289.
[Abstract/Free Full Text] - Batty, D. and Wood, R.D. (2000) Damage recognition in nucleotide excision repair of DNA. Gene, 241, 193–204.[CrossRef][ISI][Medline]
- Berneburg, M. and Lehmann, A.R. (2001) Xeroderma pigmentosum and related disorders: defects in DNA repair and transcription. Adv. Genet., 43, 71–102.[Medline]
-
de Boer, J. and Hoeijmakers, J.H. (2000) Nucleotide excision repair and human syndromes. Carcinogenesis, 21, 453–460.
[Abstract/Free Full Text] - Tang, J. and Chu, G. (2002) Xeroderma pigmentosum complementation group E and UV-damaged DNA-binding protein. DNA Repair, 1, 601–616.[Medline]
-
Rapi Otrin, V., Kuraoka, I., Nardo, T., McLenigan, M., Eker, A.P., Stefanini, M., Levine, A.S. and Wood, R.D. (1998) Relationship of the xeroderma pigmentosum group E DNA repair defect to the chromatin and DNA binding proteins UV–DDB and replication protein A. Mol. Cell. Biol., 18, 3182–3190.