Mutations in the XPC Gene in Families with Xeroderma Pigmentosum and Consequences at the Cell, Protein, and Transcript Levels1

[CANCER RESEARCH 60, 1974–1982, April 1, 2000] Mutations in the XPC Gene in Families with Xeroderma Pigmentosum and Consequences at the Cell, Protein, and Transcript Levels1

Franz Chavanne, Bernard C. Broughton, Daniela Pietra, Tiziana Nardo, Alison Browitt, Alan R. Lehmann, and Miria Stefanini2 Istituto di Genetica Biochimica ed Evoluzionistica CNR, 27100 Pavia, Italy [F. C., D. P., T. N., M. S.], and MRC Cell Mutation Unit, Sussex University, Falmer, Brighton BN1 9RR, United Kingdom [B. C. B., A. B., A. R. L.]

ABSTRACT (3). The patients from this group usually show only skin disorders and no neurological abnormalities. Cultured fibroblasts from XP-C pa- Xeroderma pigmentosum (XP)-C is one of the more common comple- tients exhibit very limited UV-induced DNA repair synthesis levels, mentation groups of XP, but causative mutations have thus far been ranging between 10 and 20% of normal, and are specifically defective reported for only six cases (S. G. Khan et al., J. Investig. Dermatol., 115: 791–796, 1998; L. Li et al., Nat. Genet., 5: 413–417, 1993). We have now in GGR. They are, however, capable of removing damage from the extended this analysis by investigating the genomic and coding sequence of transcribed strand of active genes at normal rates (4–6). the XPC gene, the level of expression of the XPC transcript and the status Phenotypic correction of XP-C cells by cDNA transfection resulted of the XPC protein in 12 unrelated patients, including all of the 8 Italian in the cloning of a partial but fully active XPC cDNA (7). The XP-C cases identified thus far and in 13 of their parents. Eighteen full-length cDNA, isolated by Masutani et al. (8), is 3558 nts long and mutations were detected in the open reading frame of the XPC gene, 13 of the encoded 940-amino-acid product shows limited homology with which are relevant for the pathological phenotype. The mutations are the Rad4 protein of Saccharomyces cerevisiae. The human XPC gene distributed across the gene, with no indication of any hotspots or founder spans about 24 kb, the transcribed sequence being divided into 15 effects. Only 1 of the 13 relevant changes is a missense mutation, the exons (9). remainder causing protein truncations as a result of nonsense mutations Masutani et al. (8) showed that the XPC gene encodes a M 125,000 (3), frameshifts (6), deletion (1) or splicing abnormalities (2). These find- r ings indicate that the XPC gene is not essential for cell proliferation and protein that is present in a tight complex with the Mr 58,000 protein viability and that mutations causing minor structural alterations may not encoded by hHR23B, one of the two human homologues of the yeast give an XP phenotype and may not, therefore, be identified clinically. RAD23 gene. Almost all of the XPC molecules appear to be com- XP13PV was the only patient carrying a missense mutation (Trp690Ser on plexed in vivo with hHR23B. Recent studies have shown that XPC- the paternal allele). This was also the only patient in which the XPC hHR23B binds to a variety of NER lesions and carries out the first transcript was present at a normal level and the XPC protein was detect- step in NER (damage recognition) in transcriptionally inactive DNA able, although at a lower than normal level. No quantitative alterations in (10, 11). the transcript or protein levels were detected in the XP-C heterozygous Characterization of the molecular defects in XP-C patients may parents. However, the expression of the normal allele predominated in all provide a tool to define further the biological role of the XPC protein, of them, except the father of XP13PV, which suggests the existence of a as well as the sites relevant for its activity. Thus far, six XP-C cell possible mechanism for monitoring the amount of the XPC protein. lines have been characterized at the cDNA level, and eight mutations including point mutations, deletions and insertions have been de- INTRODUCTION scribed (12, 13). In this report, we describe the clinical features and the cellular 3 NER is the principal pathway for removal of a broad spectrum of phenotype of 12 XP-C patients (8 from Italy, 1 from the United structurally unrelated lesions such as UV-induced cyclobutane pyrim- Kingdom, and 3 of Middle Eastern origin), as well as the mutations idine dimers and 6–4 photoproducts, and numerous chemical adducts. detected in the genomic and coding sequence of the XPC gene. For the The NER system has two distinct subpathways: (a) TCR, which Italian patients, the molecular analysis was extended to the parents to rapidly removes lesions from the transcribed strand of active genes; determine the allele inheritance and the linkage relationship of muta- and (b) GGR, which effects the slower repair of the rest of the genome tions. We have also investigated the level of expression of the XPC (recently reviewed in Ref. 1). Defects in NER have been found in transcript by Northern analysis and the occurrence of the XPC protein association with three rare human autosomal recessive syndromes, by Western analysis. which include XP. XP is clinically characterized by extreme sensitivity to sun-exposure, sunlight-induced pigmentation abnormalities, and MATERIALS AND METHODS a high incidence of skin cancer (2). Progressive neurological degen- eration is found in a proportion of patients. Case Reports. The study was performed on 12 patients showing clinical Complementation tests by cell fusion have provided evidence for symptoms typical of XP and classified by genetic analysis into the XP-C the existence of at least seven NER-deficient complementation group. The 8 patients coded with the suffix PV represent all of the XP-C cases groups: XP-A to XP-G. XP group C is one of the more common forms identified in Italy thus far. XP4BR is a typical XP-C patient of Middle Eastern origin (14). XP4RO is of historical interest as it was the first XP to be used in complementation analysis (15). XP6BR is a very unusual patient, who, at the Received 9/23/99; accepted 2/2/00. The costs of publication of this article were defrayed in part by the payment of page age of 67, had had multiple self-healing melanomas (16). XP14BR was charges. This article must therefore be hereby marked advertisement in accordance with unusual in that, apart from the expected sensitivity to UV light, both the 18 U.S.C. Section 1734 solely to indicate this fact. individual and her cells were sensitive to ionizing radiation. Clinical features 1 Supported by Associazione Italiana Ricerca sul Cancro Grant (to M. S.), by EC and related literature references for all of the patients are reported in Table 1. Human Capital and Mobility Grant CHRX-CT94-0443 (to M. S. and A. R. L.), and by EC contract QLG1-1999-00181. In common with most XP-C individuals described in the literature, none of the 2 To whom requests for reprints should be addressed, at Istituto di Genetica Biochimica analyzed cases showed neurological abnormalities. ed Evoluzionistica CNR, via Abbiategrasso 207, 27100 Pavia, Italy. Phone: 39-0382- Cells and Culture Conditions. Primary fibroblast cultures were estab- 546330; Fax: 39-0382-422286; E-mail: [email protected]. lished from biopsies of unaffected skin obtained from the 12 patients and 11 3 The abbreviations used are: NER, nucleotide excision repair; TCR, transcription- coupled repair; GGR, global genome repair; XP, xeroderma pigmentosum; UDS, unsched- parents. Fibroblasts were routinely grown in Ham’s F-10 medium (Life Tech- uled DNA synthesis; nt, nucleotide. nologies, Inc., Rockville, MD) supplemented with 12% FCS (Irvine, Santa 1974

Table 1 Clinical features and DNA repair data of the 12 XP-C patients analyzed in this study Ocular Age at last Skin Age at Repair synthesis Patient Sex examination (yr) tumors onset (yr) Lesions Tumors (% of normal) Reference XP5PV M 21 ϩ 12 ϩϪc 15 This article XP9PVa M35 ϩ 15 ϩϩ 15 37 XP10PV F 15b ϩ 4 ϩϩ 20 This article XP12PV M 5 ϪϪϪ20 This article XP13PV M 14 ϩ 4 ϩϪ 20 This article XP18PVa F9 ϩ 8 ϪϪ 18 This article XP19PV M 4 ϪϩϪ20 This article XP26PV M 3 ϪϪϪ10 This article XP4RO F 16 ϩ 10 15 XP4BR M 13 ϩ 20 14 XP6BR M 67 ϩ 28 10 16 XP14BR F 19b ϩ 11 Ϫ 10 38 a Parents are consanguineous. b Age at death. c Ϫ, symptoms not present.

Ana, CA) or Eagle’s MEM (Life Technologies, Inc.) supplemented with 15% Northern Blot Analysis. RNA was extracted by a cesium chloride-gradi- FCS (PAA Laboratories, Teddington, United Kingdom). Fibroblasts from eight ent centrifugation procedure from samples of 2 ϫ 107 fibroblasts or 1 ϫ 108 healthy donors (C1PV, C3PV, B119, CF, FB345, FB377, FB380, FB383) and lymphoblastoid cells resuspended in 1 ml of guanidinium thiocyanate buffer [4 from two XP patients previously assigned to group C were used as reference M guanidinium thiocyanate and 3 M sodium acetate (pH 6)]. strains in the study. Total RNA (5 ␮g) was electrophoresed on 1.2% agarose formaldehyde gel, Lymphoblastoid cell lines were established by EBV transformation of stained with ethidium bromide, and blotted onto Hybond-N membrane (Am- peripheral blood lymphocytes from a normal donor (352/96), XP26PV, and the ersham). Hybridization was carried out by overnight incubation with an XPC parents of the latter. These cell lines were cultured in RPMI 1640 (Sigma, St. probe corresponding to cDNA nts 286-1413. The probe was obtained by PCR

Louis, MO) supplemented with 10% FCS in a 3% CO2 atmosphere. amplification and was radiolabeled using the megaprime DNA labeling system DNA-Repair Investigations. The response to UV irradiation was analyzed (Amersham). The signals were normalized against the ethidium bromide- by measuring UDS, cell survival in proliferating and nonproliferating cultures, stained signals of 28S rRNA. and recovery of RNA synthesis after exposure to UV light. The definition of Sequence Analysis of the XPC Gene. RNA was extracted from approxi- the genetic defect responsible for the UV hypersensitivity was carried out by mately 2 ϫ 106 fibroblasts or 2 ϫ 107 lymphoblastoid cells using lysis with classical complementation assays. Procedures for cell survival, UDS, recovery guanidium isothiocyanate followed by phenol extraction and isopropanol pre- of RNA synthesis, and genetic analysis are routinely used in our laboratory and cipitation. cDNA synthesis was carried out using oligo d(T) primers, 2 ␮g have all been described previously (17, 18). RNA, and Mu-MLV reverse transcriptase (Life Technologies, Inc.) in a total Western Blot Analysis. Cells (2–5 ϫ 106) were sonicated on ice for 60 s volume of 40 ␮l. After incubation for1hat37°C, the mixture was diluted to in sample buffer [62.5 mM Tris-HCl (pH6.8), 4 M urea, 10% glycerol, 2% SDS, 50 ␮l. Ten ␮l of the cDNA synthesis reaction was used for PCR amplification 5% ␤-mercaptoethanol, and 0.003% bromophenol blue] and incubated at 65°C (Amplitaq, Perkin-Elmer, Norwalk, CT) in the buffer supplied by the manu- for 15 min before loading, as described by Shah et al. (19). Protein samples facturer. The whole XPC coding region was amplified in four overlapping were electrophoresed on 6% polyacrylamide-SDS gels and transferred onto fragments (Table 2). Amplification was performed under the conditions de- Hybond-C membrane (Amersham, Little Chalfont, United Kingdom) at 120V scribed by Li et al. (12), except for the primers F1 and F2, which were used for1hinice-cold transfer buffer (25 mM Tris, 192 mM glycine, and 20% with the following parameters: 35 cycles at 94°C for 1 min, 60°C for 1.5 min, methanol). The membranes were incubated in two blocking buffers succes- and 72°C for 2.5 min. sively: for1hin1%casein in PBS and for another hour in 5% skim milk in Genomic DNA was isolated from approximately 5 ϫ 106 fibroblasts or 50 mM Tris-HCl (pH7.5), 50 mM NaCl, and 0.15% Tween 20. The blots were 4 ϫ 107 lymphoblastoid cells using proteinase K digestion and phenol extrac- probed in a fresh solution of the second blocking buffer with the first antibody tion. Genomic DNA amplification was carried out on 0.5-␮g samples using the (anti-XPC directed against the last 19 amino acids of the XPC protein or the primers listed in Table 2. Fragments less than 5 kb in length were amplified antibody against the Mr 62,000 subunit of TFIIH as loading control), and then using AmpliTaq Gold (Perkin-Elmer) and the following conditions: 1 cycle at Ј with the second antibody (antirabbit F(ab )2 or antimouse, respectively) con- 95°C for 12 min and 35 cycles at 95°C for 1 min and at 63°C for 10 min. The jugated with horseradish peroxidase. Detection was carried out with the en- other fragments were amplified with XL-PCR (Perkin-Elmer) in the buffer hanced chemiluminescence system (Pierce, Rockford, IL) and Hyperfilm MP supplied by the manufacturer. PCR conditions were 1 cycle at 94°C for 4 min (Amersham). and 35 cycles at 94°C for 1 min and at 68°C for 10 min. PCR products were

Table 2 PCR amplification of the XPC gene Amplified fragmenta 5Ј primer 3Ј primer

Size (bp)

Regionb cDNA gDNA Code Sequence Code Sequence Ϫ64–622 686 15,000c F1 5Ј-GGCGTTCTAGCGCATCGCGG-3Ј F2 5Ј-CCTTGTGTGTGTCCTCATGG-3Ј 286–1413 1,128 G1 5Ј-GGGGATGACCTCAGGGACTT-3Ј G2 5Ј-GGGGGCTTTCCTCTGCTTTG-3Ј 1167–2291 1,125 H1 5Ј-CAGAAAGAAACGGAGCAAGC-3Ј H2 5Ј-GGCAGGAAGAGGTACACATTC-3Ј 2083–3274 1,192 J1 5Ј-AGAGTGGTGAGGCTTGGAGA-3Ј J2 5Ј-GCAACCGAGGCGAGTGAA-3Ј 1180–1972 793 2,300c B1 5Ј-AGCAAGCCCTCCTCCAGCGA-3Ј B2 5Ј-CCTCATATTTCAGGAGATGCC-3Ј 1035–1413 379 A1 5Ј-CTCCTCAGAAACTTCCAGCC-3Ј G2 5Ј-GGGGGCTTTCCTCTGCTTTG-3Ј Intron 11–2447 ϳ800c C1 5Ј-ATTCCCGTCGTAGTCTTCCC-3Ј C2 5Ј-TCCTCGCAGACGATGTATCC-3Ј 2614–2884 271 D1 5Ј-GCAGCTCCCCACACAGATGC-3Ј D2 5Ј-GGTGTGGGGCCTGTAGTGG-3Ј 2004–2291 ϳ10,000c E1 5Ј-GTATTGTCGTGGAGAAGCGG-3Ј H2 5Ј-GGCAGGAAGAGGTACACATTC-3Ј a Fragments were amplified from cDNA and/or genomic DNA (gDNA). b The position ϩ1 corresponds to the A residue of the start codon at position 106 in the cDNA sequence reported in GenBank accession no. D21089 (8). c The estimate was made by eye from the rate of migration after electrophoresis in agarose gels. 1975

Fig. 1. Response to UV irradiation in fibroblast strains from eight Italian XP-C patients (----)and from four normal subjects (—–). The reported values are the mean of at least two independent exper- iments with SEs always lower than 10%. A, UV- induced DNA repair synthesis expressed as mean number of autoradiographic grains/nucleus. B, recovery of RNA synthesis after UV irradiation in cells labeled with [3H]uridine 24 h after irradiation; incorporation values in irradiated samples are expressed as percentages of those in unirradiated cells (%C). C-D, sensitivity to the lethal effects of UV light in proliferating (C) and nondividing (D) cells; gray-shaded areas, the range of survival in cells from normal subjects.

purified by agarose gel electrophoresis and manually sequenced using a patients XP4RO, XP4BR, XP14BR, and XP6BR (see references in Thermo Sequenase radiolabeled terminator cycle sequencing kit (Amersham). Table 1). This pattern of response to UV light is typically present in XP cells belonging to group C, and it reflects a specific defect in RESULTS GGR. Normal TCR in XP-C cells results in normal rates of recovery of RNA synthesis. In nondividing cells, the ability to carry out GGR The study was performed on 12 patients showing clinical symptoms typical of XP (Table 1) and classified by genetic analysis into the is of relatively minor importance because only the actively transcribed XP-C group. As shown in Fig. 1, the eight Italian patients showed regions of DNA are used, and nondividing XP-C cells consequently similar alterations in the cellular response to UV irradiation: drasti- have close-to-normal survival levels (4). cally reduced UV-induced DNA repair synthesis levels (with UDS To investigate the status of the XPC protein in the XP-C cells, levels ranging between 10 and 20% of normal); substantial sensitivity Western blot analysis was carried out using antibodies specifically to the killing effects of UV light in proliferating cultures, but normal recognizing the COOH-terminal region of the protein. The XPC recovery of RNA synthesis at late times after irradiation; and survival protein was clearly detectable in cell extracts from the 2 normal levels in nonproliferating cultures that were significantly affected only individuals and 13 XP parents, and in one patient (XP13PV), although at high UV doses. Similar alterations have been described in the at a lower level; it was not observed in any of the other XP-C patients

Fig. 2. Expression of the XPC protein. Total cell lysates from the different XP-C patients, their parents, and a normal individual were analyzed by Western blotting with anti-XPC polyclonal antibodies. Equal loading of proteins was visualized by using anti-p62 monoclonal antibodies. 1976

Table 3 Inactivating mutations and polymorphisms found in the 12 XP-C patients analyzed Patient codea

XP12PV XP18PV XP19PV XP5PV XP13PV XP4BR XP26PV XP9PV XP4RO XP10PV XP14BR XP6BR

Amino acid Mutation Exonb change 121212121212121212121212 Inactivating mutations ϪC128 2 fs43c378stop ϩϩϩϩ ϩ ϩAA321 3 fs1083113stop ϩϩ ϩ ϩT671 5 fs2573268stop ϩϩ ϪAA1103–1104 8 fs3683373stop ϩ ϪTG1643–1644 8 fs5483572stop ϩϩ ϪC2257 12 fs7533766stop ϩϩ C658T 5 Arg220opal ϩϩ C1735T 8 Arg579opal ϩϩ C2152T 11 Arg718opal ϩϩ G2069C 10 Trp690Ser ϩ del2421–2604 13 and 14 fs8073856stop ϩ del1627–1872d 3Ј third of exon 8 del543–624 ϩϩ ϩϩ del2251–2420e 12 del751–8063808stop ϩϩ Polymorphisms C303T 3 Asp-101 ϩ G1475A 8 Arg492His ϩϩϩ T1496C 8 Val499Ala ϩϩϩ ϩϩϩϩϩϩϩϩϩϩϩ ϩϩ G2061A 10 Arg-687 ϩϩ ϩ ϩϩ ϩϩ ϩ A2815C 15 Lys939Gln ϩϩϩϩϩ a ϩ, mutation present in the indicated allele. In the eight Italian patients (PV), allele 1 corresponds to the paternal allele, and allele 2 corresponds to the maternal allele. b Exon number refers to the XPC genomic structure reported in Ref. 9. c fs, frameshift; del, deletion. d Mutation detected on the cDNA. e Mutation detected only on the cDNA of the patient as result of an abnormal splicing due toaCtoAchange six bases upstream from the intron 11-exon 12 junction.

Fig. 3. Mutations in the XPC gene of patients XP5PV, XP9PV, XP12PV, XP13PV, XP18PV, and XP19PV. Autoradiographs of sequencing gels: A, the AA321 insertion in the patient XP5PV and the C2257 deletion and C to A transversion six bases upstream from the intron 11-exon 12 junction in XP9PV and his mother; B, the C128 deletion in the XP12PV and XP18PV families; C, the G2069C and G2061A changes and the AA321 insertion in the XP13PV family; D, the C128 deletion and AA 1103–1104 deletion in the XP19PV family.

1977

(Fig. 2 and data not shown), which indicated that the XPC protein in these patients either was not present or was lacking the COOH- terminal region that was used to raise the antibodies. To define the molecular defect, total RNA from the 12 XP-C patients was reverse transcribed and the whole XPC cDNA was amplified in four overlapping fragments, which were then directly sequenced. Results of our analysis are summarized in Table 3. Eight- een mutations were detected in the open reading frame of the XPC gene, 13 of which are relevant for the pathological phenotype, the other 5 being polymorphisms. However, when we analyzed the XPC cDNA from the parents using standard amplification conditions, we were unable to detect the mutations found in the patients in any of the cases except the father of XP13PV. This suggested that, with this one exception, the expression of the mutant RNA was much lower in heterozygotes than that of the normal RNA. Therefore, in the eight Italian patients, the genomic DNA regions containing the mutations were sequenced, and the pattern of inheritance of the alleles was established by analyzing the relevant genomic DNA regions of the parents. As shown in Table 3, the most common inactivating changes are frameshift mutations resulting from the insertion or deletion of one or two bases. The loss of the C residue at position 128 was detected in both alleles of patients XP12PV and XP18PV (Fig. 3B), and in the maternal allele of patient XP19PV—the paternal allele carrying an AA deletion at position 1103–1104 (Fig. 3D). An insertion of two A residues at position 321 was observed in both of the alleles of XP5PV Fig. 4. Southern analysis, using a 3Ј XPC cDNA probe that spanned nts 2400–3000 of (Fig. 3A) and in the maternal allele of XP13PV (Fig. 3C). Three the XP6BR (Lanes 1 and 3) and normal (Lanes 2 and 4) XPC genomic DNA after patients were homozygous for other frameshift mutations: insertion of digestion with BamHI (Lanes 1 and 2)orEcoRI (Lanes 3 and 4). Autoradiograph shows the presence in XP6BR of an extra band, in addition to the bands seen in normal cells. a single T residue at position 671 (XP4BR), deletion of the dinucleotide TG at position 1643–1644 (XP26PV), and deletion of C2257 (XP9PV; Fig. 3A). Three of the four deletions detected (namely, XPC mRNA was observed in three patients—XP9PV, XP10PV, and ϪC128, ϪTG1643–1644, and ϪC2257) and the AA321 insertion XP26PV. In XP9PV, the whole of exon 12 (nts 2251–2420) was occur in runs of identical bases and are likely to result from replication absent. Sequencing of the genomic region corresponding to exon 12 slippage. and of the intron sequences at the 5Ј and 3Ј end of exon 12 showed Three nonsense mutations consisting of C to T transitions at posi- that the patient was homozygous for two mutations: (a) deletion of the tions 658, 1735, and 2152, were observed in both alleles of the C residue at position 2257 in exon 12, as already described; and (b)a patients XP4RO, XP10PV, and XP14BR, respectively. All of the C to A change in the intron 11 acceptor site, six nts upstream from the transitions occur at CpG sites, probably resulting from demethylation intron 11–exon 12 junction (Fig. 3A). This mutation reduces of 5-methylcytosine to thymine. They all induce Arg to opal substi- the efficiency of the splice acceptor site of intron 11 by interrupting tutions. the polypyrimidine tract. These two mutations are 13 bp apart in the A G2069C transversion was found in the paternal allele of XP13PV genomic DNA and could conceivably have arisen from a single event. (Fig. 3C); this missense mutation causes the change of amino acid 690 The net result is the generation of two differentially spliced products, from Trp to Ser. Trp-690 is conserved in five homologues (human, one of normal size containing the C2257 deletion and the other mouse, Drosophila melanogaster, and the yeasts S. cerevisiae and lacking the whole of exon 12. Schizosaccharomyces pombe), and it is located in a sequence of five In the XPC coding sequences of patients XP10PV and XP26PV, amino acids that are predicted to be in an ␣-helical conformation by both of which contain mutations toward the 3Ј end of exon 8, we the PHD secondary structure prediction protocol (20–22). The alter- observed a deletion of nts 1627–1872, corresponding to the last 246 ation of Trp to Ser is predicted to destroy this ␣-helical conformation, nts of exon 8. Low levels of cDNAs with this deletion could also be so that this amino acid substitution probably induces some change in detected in the parents of both patients (Fig. 5). Amplification and the secondary structure of the XPC protein. The presence of a mis- sequencing of the regions around the mutations and around the splice sense mutation in this patient is consistent with the presence of donor site of intron 8 in the genomic DNA of these patients and their detectable XPC protein in the Western blots (Fig. 2). parents did not show the presence of any mutation other than the A large deletion of 184 nts, from position 2421 to 2604, was found C1735T transition in XP10PV and the TG1643–1644 deletion in in one allele of XP6BR cDNA. This deletion comprises exons 13 and XP26PV. The splice donor site of the alternative splice event is 14 (9) and could arise either as a splicing abnormality or as a genuine located in the 3Ј third of exon 8. This exon is unusually long (882 bp deletion in genomic DNA with the deletion break points in introns 12 in length), which could make it unstable. The new splice site scores 83 and 14. The latter is supported by Southern analysis using a 3Ј XPC using the system of Shapiro and Senapathy (23), well above the cDNA probe spanning nts 2400–3000, which showed—in addition to minimum necessary to form a splice donor site. The point mutation in the bands seen in normal cells—an extra band after the digestion of XP10PV and the two-base deletion in XP26PV are both located the genomic DNA with EcoRI or BamHI (Fig. 4). Because of a very toward the 3Ј end of exon 8 (nt 991-1872) and could induce some low level of expression of the second allele in XP6BR, the inactivat- change in the secondary structure of the mRNA, so that the cryptic ing mutation present on this allele was not identified. splice site at position 1627 is used instead of the normal one at the Besides the normally spliced product, abnormal splicing of the beginning of intron 8. Because the two patients are homozygous, the 1978

Fig. 5. Mutations in the XPC gene of the XP10PV and XP26PV families. Top, agarose gel electrophoresis of PCR amplification products of the XPC cDNA region 1180–1972, showing a shorter fragment, in addition to the normal-size fragment, in the XP family members. Middle, autoradiographs of sequencing gels of the short fragment, showing the deletion of the XPC cDNA region 1627–1872. Bottom, autoradiographs of sequencing gels, showing the C1735T and the TG1643–1644 deletion on the XPC genomic DNA of the XP10PV and XP26PV families, respectively. presence of two different splice products indicates that both normal (12, 13) and in our work are summarized in Fig. 6. The mutations and cryptic splice donor sites are used. are distributed across the gene, with no indication of any hotspots In addition to the mutations relevant for the XP phenotype reported or founder effects. Excluding the five cases analyzed only at the above, three missense mutations, namely a G1475A transition cDNA level (12), 10 of the 13 remaining patients are homozygotes (Arg492His), a T1496C transition (Val499Ala), and an A2815C trans- for the mutated XPC alleles, which suggests that they were all born version (Lys939Gln), were observed either in the homozygous or from consanguineous marriages although consanguinity has been heterozygous state in some patients (Table 3). These mutations are reported in the family histories of only XP9PV and XP18PV. The likely to be polymorphisms because several phenotypically normal same inactivating mutations were found in XP12PV, XP18PV, and subjects who belonged either to the general population or to families XP19PV (the loss of the C residue at position 128); in XP13PV and with XP-affected members were homozygous for these changes. Fur- XP5PV (the insertion of two A residues at position 321); and in thermore, we observed two silent mutations—C303T (Asp-101) and XP4PA and XP26PV (deletion of the dinucleotide TG at position Ј G2061A (Arg-687)—and two G to C transversions located in the 5 1643–1644). However the analysis of the linkage relationship of and 3Ј untranslated regions at position Ϫ27 and 2919, respectively. inactivating mutations with polymorphisms showed that common Expression of XPC transcripts in the Italian patients and their alleles are shared only by XP12PV and XP18PV and by XP13PV parents were examined by Northern blot analysis (data not shown). and XP5PV. Compared with cells from normal donors, no significant differences Mutation Pattern in XP-C Patients. Different types and sites of were observed in the XP-C parents or in patient XP13PV. In the other changes in the XPC protein result in the XP phenotype. The majority patients, the levels of XPC transcript were slightly reduced with values ranging between 60 and 80% of normal. of the mutations relevant for the pathological phenotype (15 of 20) are predicted to cause premature termination of the protein as a result of frameshifts (nine), nonsense mutations (three), insertion (one), dele- DISCUSSION tions (one), or splicing abnormalities (one). In addition to these, it is Although XP-C is one of the more common XP complementation likely that other mutations also result in a null product. For instance, groups, up until now the causative mutations have been determined the patients XP10PV and XP26PV showed two different XPC tran- for only six patients. We have now extended this database by analyz- scripts at the cDNA level, one with a large in-frame deletion corre- ing an additional 12 patients including 8 who comprise all of the sponding to the 3Ј third of exon 8. This transcript with the internal known Italian XP-Cs. The results reported in previous publications deletion was not detected on Northern blots, nor did we observe any 1979

Fig. 6. XPC protein and amino acid changes caused by the inactivating mutations found in 18 XP-C patients. The diagram shows the XPC protein with the hHR23B binding domain u and the putative nuclear location signal f. The amino acid changes are shown boxed: black on white, XP-C cases reported in this study; white on gray, XP-C cases reported in Refs. 12 and 13. The numbers 1 and 2 after the patient code, the different alleles; in the Italian patients (PV), 1, the paternal allele; 2, the maternal allele.

product on Western blots although the corresponding protein should to a lesser extent, in CSB (28, 29). It raises the possibility that some contain the XPC region used to raise the antibodies. These findings missense mutations that cause minor structural XPC alterations might indicate that the transcript with the internal deletion and the corre- result in a milder clinical phenotype that would not be diagnosed as sponding protein are either present in too low an amount to be XP. Conversely, in the diagnosed cases, the lack of the XPC protein detected or are unstable. and the presence of a mutated protein both result in similar clinical As well as indicating that XPC is not essential for cell proliferation phenotypes and confer the same degree of cellular sensitivity to UV and viability, the mutation pattern in the patients has enabled us to light in terms of survival and UDS. Ten of the 12 XP-C cases reported identify a few positions in the XPC protein that are important for its in this study (the 8 Italian cases, XP4BR, and XP4RO) show the functionality, namely the amino acid residue 334, mutated in the clinical features typically described in the XP-C group. As already patient XP1MI, and the region around the amino acid 690–699, mentioned, XP-C is a large group, but its pathological phenotype is containing the amino acid 690 changed in XP13PV, and the insertion rather homogeneous. The patients usually show skin and ocular symp- of an amino acid at position 698 in XP8BE (our analysis indicates that toms, whereas mild mental retardation has been reported for only one the A2815C change, resulting in Lys939Gln, described as a second case (XP1MI). Differences in the severity of skin disorders depend on putative causative mutation in XP8BE by Li et al. (12) is in fact a age, climate, and life-style (essentially the protection from the sun). polymorphism). Accordingly, in the Italian cases, no skin tumors have been reported The mutations observed in the patients result from different events in the three youngest patients. and include: (a) deletions and insertions in runs of identical bases XP6BR is an unusual patient in that he survived to the age of 66 and presumably resulting from replication slippage; (b) C to T transition at his multiple melanomas regressed spontaneously. He was the only CpG sites, as a consequence of the demethylation of 5-methylcy- patient with a large deletion in the XPC gene, but this is unlikely to be tosine; and (c) mutations located in the splice sites or affecting the related to his clinical features because, like most of the other patients, splicing indirectly by interfering with the stability of the transcript. he did not express any XPC protein in his cells (the possibility that a This would be the case for the transcript with the deletion of nts 1627–1872, found in association with a normal-sized transcript con- small amount of partially functional XPC protein, below the limit of taining the C1735T transition (XP10PV) or the TG1643–1644 dele- detection in our assays, can account for these features is not exclud- tion (XP26PV). These were the only inactivating changes observed at ed). XP14BR is unique in that both the patient and her cells were the genomic level in XP10PV and XP26PV family members. The extremely sensitive to ionizing radiation, but this feature is unrelated presence of these mutations in the 3Ј end of the exon 8 may interfere to the defect in the XPC gene, because transfection with the XPC gene 4 with the normal splicing of this unusually long (882-nt) exon, leading corrects the UV sensitivity but not the ionizing radiation sensitivity. to the partial activation of a cryptic donor site at position 1627 and to All of the fourteen XP-C patients examined show varying degrees the appearance of a transcript in which the last 246 nts of exon 8 are of reduction in the XPC transcript level (Refs. 12, 13 and “Results”). lost [exon 8 ends at nt 1872, as indicated by analysis in our patients The only exception is represented by patient XP13PV, who carries the and in XP22BE (13)]. missense mutation Trp690Ser on his paternal allele. Accordingly, Genotype-Phenotype Relation. A preponderance of protein trun- XP13PV was the only case of the 12 analyzed by us in which the XPC cation mutations, as seen in XPC, has also been found in other nonessential DNA repair genes such as XPA (24–26), ATM (27), and, 4 C. Arlett et al., unpublished results. 1980

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 2000 American Association for Cancer Research. GENOTYPE-PHENOTYPE RELATIONSHIPS IN XP-C protein was present, although at lower than normal levels. As already us with biopsy and clinical details on the Italian XP patients; to Drs. P. van der mentioned, this does not result either in milder clinical features or in Spek (Erasmus Medical Centre of Rotterdam, Rotterdam, the Netherlands) and a less severe cellular response to UV. J. M. Egly (Institut de Genetique et de Biologie Moleculaire et Cellulaire, Predominant Expression of the Normal XPC Allele in Heterozy- Strasbourg, France) for providing us with anti-XPC and anti-p62 antibodies; gotes. In the 13 cell strains from heterozygous XP-C parents, the and to Dr. R. Legerski for information concerning the genomic structure of the XPC gene. transcript and protein levels were in the normal range. However, when we analyzed the XPC cDNA from the parents by using standard amplification conditions, we were unable to detect the mutations REFERENCES found in the patients in any of the cases except the father of XP13PV. 1. de Laat, W. L., Jaspers, N. G. J., and Hoeijmakers, J. H. J. Molecular mechanism of This indicates that, with this one exception, the expression of the nucleotide excision repair. Genes Dev., 13: 768–785, 1999. mutant RNA in heterozygotes is much lower than that of the normal 2. Kraemer, K. H., Lee, M. M., and Scotto, J. Xeroderma pigmentosum. Cutaneous, ocular and neurologic abnormalities in 830 published cases. Arch. Dermatol., 123: RNA. The finding that noncoding mRNAs that carry nonsense codons 241–250, 1987. are unstable is not unprecedented (for a recent review, see Ref. 30). 3. Bootsma, D., Kraemer, K. H., Cleaver, J., and Hoeijmakers, J. H. J. Nucleotide excision repair syndromes: xeroderma pigmentosum, Cockayne syndrome and tricho- However the detection of normal levels of XPC transcript on Northern thiodystrophy. In: B. Vogelstein and K. W. Kinzler (eds.), The Genetic Basis of blots suggests that, in heterozygous carriers of XPC mutations: (a) the Human Cancer, pp 245–274. New York: McGraw-Hill, 1998. level of expression of the mutated allele is lower than that of the same 4. Kantor, G. J., Barsalou, L. S., and Hanawalt, P. C. Selective repair of specific chromatin domains in UV-irradiated cells from xeroderma pigmentosum complemen- mutated allele in the affected progeny; and (b) the expression of the tation group C. Mutat. Res., 235: 171–180, 1990. normal allele predominates and compensates for that of the mutated 5. Venema, J., van Hoffen, A., Dorcagi, V., Natarajan, A. T., van Zeeland, A. A., and XPC allele. Mullenders, L. H. F. XP complementation group C cells remove pyrimidine dimers selectively from the transcribed strand of active genes. Mol. Cell. Biol., 11: 4128– This may imply the existence of a possible mechanism of regulation 4134, 1991. of the transcript level, which is perhaps not so unexpected considering 6. van Hoffen, A., Venema, J., Meschini, R., van Zeeland, A. A., and Mullenders, L. H. F. Transcription-coupled repair removes both cyclobutane pyrimidine dimers the biological role of the XPC protein (for recent reviews, see Refs. 1, and 6–4 photoproducts with equal efficiency and in a sequential way from tran- 31). The XPC protein is indispensable during the initial phases of scribed DNA in xeroderma pigmentosum group C fibroblast. EMBO J., 14: 360–367, GGR—the NER subpathway that repairs the damage on the nontran- 1995. 7. Legerski, R. J., and Peterson, C. A. Expression cloning of a human DNA repair gene: scribed strand of active genes and in the inactive regions of the xeroderma pigmentosum complementation group C. Nature (Lond.), 359: 70–73, genome. Sugasawa et al. (10) have recently termed XPC “the initiator 1992. of GGR.” In association with hHR23B, XPC initiates GGR by sensing 8. Masutani, C., Sugasawa, K., Yanagisawa, J., Sonoyama, M. U., Enomoto, T., Takio, K., Tanaka, K., van der Spek, P., Bootsma, D., Hoeijmakers, J. H., and Hanaoka, F. and binding to lesions and recruiting the other components of the Purification and cloning of a nucleotide excision repair complex involving the repair apparatus. Several lines of evidence indicate that XPC is xeroderma pigmentosum group C protein and a human homologue of yeast RAD23. EMBO J., 13: 1831–1843, 1994. dispensable only for the repair of lesions that induce a large distortion 9. Li, L., Peterson, C. A., and Legerski, R. J. Sequence of the mouse XPC cDNA and in the DNA structure, such as an artificial cholesterol DNA adduct or genomic structure of the human XPC gene. Nucleic Acids Res., 24: 1026–1028, damage in an “open structure” (32). It is also not required for TCR, in 1996. 10. Sugasawa, K., Ng, J. M. Y., Masutani, C., Iwai, S., van der Speck, P. J., Eker, which the RNA polymerase, blocked at the lesion site, serves as a A. P. M., Hanaoka, F., Bootsma, D., and Hoeijmakers, J. H. J. Xeroderma pigmen- damage-recognition signal. Besides being absolutely required early in tosum group C protein complex is the initiator of global genome nucleotide excision the DNA damage recognition step (i.e., the limiting step of the overall repair. Mol. Cell, 2: 223–232, 1998. 11. Wakasugi, M., and Sancar, A. Assembly, subunit composition, and footprint of process), XPC is the limiting subunit of the XPC-hHR23B complex. human DNA repair excision nuclease. Proc. Natl. Acad. Sci. USA, 95: 6669–6674, In vitro, hHR23B can be replaced by substituting hHR23A, the second 1998. 12. Li, L., Bales, E. S., Peterson, C. A., and Legerski, R. J. Characterization of molecular human homologue of the yeast NER factor Rad 23, in binding and defect in xeroderma pigmentosum group C. Nat. Genet., 5: 413–417, 1993. stimulating XPC activity (33–35). In vivo, hHR23B and hHR23A are 13. Khan, S. G., Levy, H. L., Legerski, R., Quackenbush, E., Reardon, J. T., Emmert, S., both much more abundant than XPC and mostly exist in a free form Sancar, A., Li, L., Schneider, T. D., Cleaver, J. E., and Kraemer, K. H. Xeroderma pigmentosum group C splice mutation associated with autism and hypoglycinemia. in the cells (8, 36). These observations suggest that it may well be that, J. Investig. Dermatol., 111: 791–796, 1998. in the presence of only one normal XPC allele, the rate of XPC 14. Lehmann, A. R., and Stevens, S. A rapid procedure for measurement of DNA repair transcription or the stability of the transcript is somehow up-regulated in human fibroblasts and for complementation analysis of xeroderma pigmentosum cells. Mutat. Res., 69: 177–190, 1980. to ensure the proper and efficient functioning of GGR. 15. de Weerd-Kastelein, E. A., Keijzer, W., and Bootsma, D. Genetic heterogeneity of In conclusion, as well as identifying a few positions in the XPC xeroderma pigmentosum demonstrated by somatic cell hybridation. Nature (Lond.), 238: 80–83, 1972. protein that are important for its function, the results of our analysis 16. Anstey, A. V., Arlett, C. F., Cole J., Norris, P. G., Hamblin, A. S., Limb G. A., indicate that the inactivating mutations in the XP-C patients are Lehmann, A. R., Wilkinson, J. D., and Turner, M. Long-term survival and preserva- distributed across the gene, with no indication of any hotspots or tion of natural killer cell activity in a xeroderma pigmentosum patient with sponta- neous regression and multiple deposits of malignant melanoma. Br. J. Dermatol., 125: founder effects, and mainly result in truncated proteins. This implies 272–287, 1991. that XPC is not essential for cell proliferation and viabilility. In 17. Stefanini, M., Giliani, S., Nardo, T., Marinoni, S., Nazzaro, V., Rizzo, R., and addition, we have demonstrated that, in the heterozygous carriers of Trevisan, G. DNA repair investigations in nine Italian patients affected by trichothiodystrophy. Mutat. Res., 273: 119–125, 1992. XPC mutations, the expression of the normal allele predominates and 18. Stefanini, M., Lagomarsini, P., Giliani, S., Nardo, T., Botta, E., Peserico, A., Kleijer, compensates for that of the mutated XPC allele, which suggests the W. J., Lehmann, A. R., and Sarasin, A. Genetic heterogeneity of excision repair defect associated with trichothiodystrophy. Carcinogenesis (Lond.), 14: 1101–1105, 1993. existence of a possible mechanism for monitoring the amount of the 19. Shah, G. M., Poirier, D., Duchaine, C., Brochu, G., Desnoyers, S., Lagueux, J., protein. Verreault, A., Hoflack, J. C., Kirkland, J. B., and Poirier, G. G. Methods for biochemical study of poly(ADP-ribose) metabolism in vitro and in vivo. Anal. Biochem., 227: 1–13, 1995. 20. Rost, B., and Sander, C. Prediction of protein structure at better than 70% accuracy. ACKNOWLEDGMENTS J. Mol. Biol., 232: 584–599, 1993. 21. Rost, B., and Sander, C. Combining evolutionary information and neural networks to We are grateful to Drs. E. Berardesca (Clinica Dermatologica, University of predict protein secondary structure. Proteins, 19: 55–72, 1994. Pavia), P. Getti (Clinica Dermatologica, University of Bologna, Bologna, 22. Rost, B., Sander, C., and Schneider, R. PHD-an automatic mail server for protein Italy), G. Grosso (Clinica Dermatologica, University of Padova, Padova, Italy), secondary structure prediction. Comput. Appl. Biosci., 10: 53–60, 1994. 23. Shapiro, M. B., and Senapthy, P. RNA splice junctions of different classes of V. Nazzaro (Clinica Dermatologica, University of Milan, Milan, Italy), and G. eukaryotes: sequence statistics and functional implication in gene expression. Nucleic Zambruno (Istituto Dermopatico dell’Immacolata, Rome, Italy) for providing Acids Res., 15: 7155–7174, 1987. 1981

24. Sato, M., Nishigori, C., Yagi, T., and Takebe, H. Aberrant splicing and truncated- 32. Mu, D., and Sancar, A. Model for XPC-indipendent transcription-coupled repair of protein expression due to a newly identified XPA gene mutation. Mutat. Res., 362: pyrimidine dimers in human. J. Biol. Chem., 272: 7570–7573, 1997. 199–220, 81996. 33. Li, L., Lu, X., Peterson, C. A., and Legerski, R. J. XPC interacts with both HHR23B 25. Kuraoka, I., Morita, E. H., Saijo, M., Matsuda, T., Morikawa, K., Shirakawa, M., and and HHR23A in vivo. Mutat. Res., 383: 197–203, 1997. Tanaka, K. Identification of a damaged-DNA binding domain of the XPA protein. 34. Sugasawa, K., Masutani, C., Uchida, A., Maekawa, T., van der Spek, P. J., Bootsma, Mutat. Res., 362: 87–95, 1996. D., Hoeijmakers, J. H. J., and Hanaoka, F. HHR23B, a human Rad23 homolog, 26. States, J. C., McDuffie, E. R., Myrand, S. P., McDowell, M., and Cleaver, J. E. stimulates XPC protein in nucleotide excision repair in vitro. Mol. Cell. Biol., 16: Distribution of mutations in the human xeroderma pigmentosum group A gene and 4852–4861, 1996. their relationships to the functional regions of the DNA damage recognition protein. 35. Sugasawa, K., Ng, J. M. Y., Masutani, C., Maekawa, T., Uchida, A., van der Spek, Hum. Mutat., 12: 103–113, 1998. P. J., Eker, A. P. M., Rademakers, S., Visser, C., Aboussekhra, A., Wood, R. D., 27. Rotman, G., and Shiloh, Y. ATM: from gene to function. Hum. Mol. Genet., 7: Hanaoka, F., Bootsma, D., and Hoeijmakers, J. H. J. Two human homologs of Rad23 1555–1563, 1998. are functionally interchangeable in complex formation and stimulation of XPC repair 28. Mallery, D. L., Tanganelli, B., Colella, S., Steingrimsdottir, H., van Gool, A. J., activity. Mol. Cell. Biol., 17: 6924–6931, 1997. Troelstra, C., Stefanini, M., and Lehmann, A. R. Molecular analysis of mutations in the CSB (ERCC6) gene in patients with Cockayne syndrome. Am. J. Hum. Genet., 62: 36. van der Spek, P. J., Eker, A. P. M., Rademakers, S., Visser, C., Sugasawa, K., 77–85, 1998. Masutani, C., Hanaoka, F., Bootsma, D., and Hoeijmakers, J. H. J. XPC and human 29. Colella, S., Nardo, T., Mallery, D., Borrone, C., Ricci, R., Ruffa, G., Lehmann, A. R., homologs of RAD23: intracellular localization and relationship to other nucleotide and Stefanini, M. Alterations in the CSB gene in three Italian patients with the severe excision repair complexes. Nucleic Acids Res., 24: 2551–2559, 1996. form of Cockayne Syndrome (CS) but without clinical photosensitivity. Hum. Mol. 37. Nuzzo, F., Lagomarsini, P., Casati, A., Giorgi, R., Berardesca, E., and Stefanini, M. Genet., 8: 935–941, 1999. Clonal chromosome rearrangements in a fibroblast strain from a patient affected by 30. Culbertson, M. R. RNA surveillance. Unforeseen consequences for gene expression, xeroderma pigmentosum (complementation group C). Mutat. Res., 219: 209–215, inherited genetic disorders and cancer. Trends Genet., 15: 74–79, 1999. 1989. 31. Wood, R. D. DNA damage recognition during nucleotide excision repair in mamma- 38. Salob, S. P., Webb, D. K. H., and Atherton, D. J. A child with xeroderma pigmen- lian cells. Biochimie (Paris), 81: 39–44, 1999. tosum and bone marrow failure. Br. J. Dermatol., 126: 372–374, 1992.

1982

Downloaded from cancerres.aacrjournals.org on September 26, 2021. © 2000 American Association for Cancer Research. Mutations in the XPC Gene in Families with Xeroderma Pigmentosum and Consequences at the Cell, Protein, and Transcript Levels

Franz Chavanne, Bernard C. Broughton, Daniela Pietra, et al.

Cancer Res 2000;60:1974-1982.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/60/7/1974

Cited articles This article cites 36 articles, 6 of which you can access for free at: http://cancerres.aacrjournals.org/content/60/7/1974.full#ref-list-1

Citing articles This article has been cited by 14 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/60/7/1974.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/60/7/1974. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.