View metadata, citation and similar papers at core.ac.uk brought to you by CORE

provided by Elsevier - Publisher Connector

Reinvestigation of the Classi®cation of Five Cell Strains of Xeroderma Pigmentosum Group E with Reclassi®cation of Three of Them

Toshiki Itoh,*² Stuart Linn,² Tomomichi Ono,§ and Masaru Yamaizumi* *Department of Cell Genetics, Institute of Molecular Embryology and Genetics; and §Dermatology, Kumamoto University School of Medicine, Kumamoto, Japan; ²Department of Molecular and Cell Biology, Division of Biochemistry and , University of California, Berkeley, U.S.A.

Xeroderma pigmentosum is a photosensitive syn- however, remained assigned to xeroderma pigmen- drome caused by a defect in nucleotide excision tosum group E. With the exception of the Ddb+ repair or . Individuals of xero- strain XP89TO, which demonstrated defective derma pigmentosum group E (xeroderma pigmento- nucleotide excision repair, both Ddb± and Ddb+ xer- sum E) have a mild clinical form of the disease and oderma pigmentosum E cells exhibited the same their cells exhibit a high level of nucleotide excision levels of variation in unscheduled DNA synthesis that repair as measured by unscheduled DNA synthesis, were seen in normal control cells. By genome DNA as well as biochemical heterogeneity. Cell strains sequencing, the two Ddb± xeroderma pigmentosum from one group of xeroderma pigmentosum E E strains were shown to have in the DDB2 patients have normal damage-speci®c DNA binding , con®rming previous reports for XP82TO and activity (Ddb+), whereas others do not (Ddb±). Using GM02415B, and validating the classi®cation of both a re®nement of a previously reported cell fusion cells. As only the Ddb± strains investigated remain complementation assay, the previously assigned classi®ed in the xeroderma pigmentosum E comple- Ddb+ xeroderma pigmentosum E strains, XP89TO, mentation group, it is feasible that only Ddb± cells XP43TO, and XP24KO, with various phenotypes in are xeroderma pigmentosum E and that mutations DNA repair markers, were reassigned to xeroderma in the DDB2 gene are solely responsible for the xero- pigmentosum group F, xeroderma pigmentosum derma pigmentosum E group. Key words: Cockayne variant, and -sensitive syndrome, respec- syndrome/damage-speci®c DNA binding /nucleotide tively. The Ddb± xeroderma pigmentosum E strains, excision repair/postreplication repair/ultraviolet±sensitive XP82TO, and GM02415B, which showed almost syndrome.J Invest Dermatol 114:1022±1029, 2000 normal cellular phenotypes in DNA repair markers,

eroderma pigmentosum (XP) is a rare autosomal (RDS) after UV irradiation (Lehmann et al, 1975, 1977; Rude recessive disease characterized bya clinical and and Friedberg, 1977; Cleaver and Kraemer, 1995). Recently, the cellular hypersensitivity to ultraviolet (UV) light XPV gene was cloned, and identi®ed as DNA polymerase X (Cleaver and Kraemer, 1995). Patients exhibit (Masutani et al, 1999; Johnson et al, 1999). In the seven NER dermatologic abnormalities, including thickening defective groups, the which are defective in XP with severe and hyperpigmentation of sun-exposed skin, and develop sunlight- clinical manifestations (XPA, B, and G) and with the classical type induced malignancies at an earlier age. Bycell fusion analyses,XP of XP (XPC, D, and F) have been cloned (Friedberg et al, 1995; has been classi®ed into seven genetic complementation groups (A± Sijbers et al, 1996). The XPE gene, however, has yet to be cloned G), which are defective in nucleotide excision repair (NER). A because of several dif®culties: (i) the clinical manifestations in XPE separate group, XP variant (XPV) has pro®cient NER but patients are verymild; (ii) as cells derived from XPE patients have exaggerated delayin recoveryof replicative DNA synthesis high levels of unscheduled DNA synthesis (UDS), it is very dif®cult to distinguish between XPE, XPV, and normal cells; and (iii) the Manuscript received July1, 1999; revised January13, 2000; accepted for high level of UDS makes analysis of a complementation assay using publication January24, 2000. cell fusion to determine the classi®cation of XPE strains Reprint requests to: Dr. Toshiki Itoh, Department of Molecular and problematic. Cell Biology, Division of Biochemistry and Molecular Biology, 401 Barker Biochemical heterogeneityin binding to damaged DNA has Hall, Universityof California, Berkeley,CA 94720-3202. Email: toshiki been reported for XPE cell free extracts (Kataoka and Fujiwara, @uclink4.berkeley.edu 1991; Keeney et al, 1992; Friedberg et al, 1995). Cell strains from Abbreviations: CS, ; DDB, damage-speci®c DNA binding; NER nucleotide-excision-repair; PRR, postreplication repair; two of 13 unrelated XPE individuals lack a damage-speci®c DNA- RDS, recoveryof replicative DNA synthesis;RRS, recoveryof RNA binding (DDB) activityin nuclear extracts (Chu and Chang, 1988; synthesis; TK, thymidine kinase; UVsS, ultraviolet-sensitive syndrome; Keeney et al, 1992) and are termed Ddb± XPE. Recently, three UDS, unscheduled DNA synthesis; XP, xeroderma pigmentosum. additional Ddb± XPE patients were reported (Otrin et al, 1998).

0022-202X/00/$15.00 ´ Copyright # 2000 byThe Societyfor Investigative Dermatology,Inc. 1022 VOL. 114, NO. 5 MAY 2000 RECLASSIFICATION OF XPE CELLS 1023

Mutations have been identi®ed in the DDB2 gene in the three and Ddb+ XPE strains actuallycomprise two separate XP XPE Ddb± strains examined, but not in XPE Ddb+ strains (Nichols complementation groups, with the DDB2 gene being responsible et al, 1996). The question has been proposed as to whether Ddb± for the NER defect in Ddb± XPE patients. To clarifythis question,

Table I. Cell strains used in this study

Cell strain Group Reference(s) Source

Mori Normal Itoh et al, 1994, 1995a, 1995b, 1996a, 1996b Established in our laboratory. Turu Normal Itoh et al, 1995a, 1996b Established in our laboratory. Sono Normal Itoh et al, 1996b Established in our laboratory. Goryo Normal Established in our laboratory. Umi Normal Established in our laboratory. Mura Normal Established in our laboratory. XP24KO XPEa Ddb+ Fujiwara et al, 1985 Gift from Dr Fujiwara. XP43TO XPEa Ddb+ Kondo et al, 1989 Gift from Dr Kondo. XP82TO XPEa Ddb± Kondo et al, 1988 Gift from Dr Kondo. XP89TO XPEa Ddb+ Kondo et al, 1989 Gift from Dr Kondo. GM02415B XPEa Ddb± Bootsma et al, 1970; Kleijer et al, 1973; de Weerd-Kastelein NIGMS Human Genetic Mutant Cell Repository. et al, 1974; Kraemer et al, 1975a, 1975b Kps3 UVsSb Itoh et al, 1994, 1995a, 1996c Established in our laboratory. Kps5 XPDc Itoh et al, 1994, 1995b Established in our laboratory. Kps6 XPFd Itoh et al, 1995b Established in our laboratory. Nps8 XPFd Itoh et al, 1995b Established in our laboratory. Mps1 CSAe Itoh et al, 1994, 1995a, 1996a Established in our laboratory. CS1MO CSBf Itoh et al, 1994, 1995a, 1996a Human Science Research Resources Bank. GM10905 CSBf Itoh et al, 1996a. NIGMS Human Genetic Mutant Cell Repository. XP2SA XPVg Itoh et al, 1994, 1995a, 1996b Human Science Research Resources Bank. Ops2 XPAh Established in our laboratory. Ops3 XPAh Established in our laboratory. Ops12 XPAh Established in our laboratory. Ops24 XPAh Established in our laboratory. Ops34 XPAh Established in our laboratory. Ops35 XPAh Established in our laboratory.

aXeroderma pigmentosum group E. bUVs syndrome. cXeroderma pigmentosum group D. dXeroderma pigmentosum group F. eCockayne syndrome group A. fCockayne syndrome group B. gXeroderma pigmentosum variant. hXeroderma pigmentousm group A.

Table II. Unscheduled DNA synthesis

Experiment Cell strain UDSa Range of relative rateb (%)

I XP43TO/Mori 16.0 6 0.3/18.1 6 0.3c 85±92 XP82TO/Mori 15.4 6 0.4/20.4 6 0.4c 75±80 XP89TO/Mori 6.8 6 0.2/24.0 6 0.5c 27±30 XP24KO/Mori 14.0 6 0.3/11.7 6 0.3c 114±125 GM02415B/Mori 23.9 6 0.4/26.6 6 0.4c 87±93 II XP82TO/Turu 26.6 6 0.3/22.0 6 0.3c 118±124 GM02415B/Turu 30.4 6 0.4/27.7 6 0.4c 120±125 III Turu/Mori 23.2 6 0.3/25.9 6 0.3c 87±92 Sono/Mori 18.8 6 0.3/15.6 6 0.2c 110±113 IV Goryo/Mori 11.6 6 0.2/11.8 6 0.3d 94±103 Umi/Mori 8.3 6 0.2/11.2 6 0.3c 70±78 Mura/Mori 12.5 6 0.3/17.6 6 0.2c 69±73 V XP24KO/Morie 15.3 6 0.3/13.6 6 0.2c 109±116 VIf Ops2 (XPA)/Mori 2.9 6 0.8/35.7 6 2.6c 5±11 Ops3 (XPA)/Mori 2.2 6 0.4/26.8 6 1.7c 6±10 Ops12 (XPA)/Mori 2.9 6 0.4/30.1 6 1.0c 8±11 Ops24 (XPA)/Mori 4.5 6 0.5/24.5 6 1.2c 16±21 Ops34 (XPA)/Mori 3.8 6 0.5/40.6 6 1.7c 8±11 Ops35 (XPA)/Mori 1.4 6 0.3/28.8 6 1.1c 4±6

aTo measure UDS (unscheduled DNA synthesis), cells were irradiated with UV at a dose of 30 J per m2, and then immediatelylabeled with [ 3H]thymidine (50 mCi per ml) for 2.5 h. Exposure time was 24 h. UDS is given in grains per nucleus. Data are mean 6 SEM of 300 determinations except for experiment VI (50 determinations). bRange of relative rate is the range of UDS values (grain-numbers) of cell strains tested compared with those of the normal control (mori or Turu) cells. cp < 0.001. dp > 0.1. eThis experiment was performed at a dose of 10 J per m2. fThese experiments were control experiments using XPA cells. All cell strains were assigned to XPA bythe Kumamoto laboratory.Ops2, Ops3, and Ops24 ha ve homozygous G to C substitution at the 3¢ splice acceptor site in intron 3 of XPA gene (Itoh and Yamaizumi, unpublished data). 1024 ITOH ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY we have extensivelyassessed DNA repair markers in ®ve previously UV survival assay Colonyforming abilitywas performed as previously reported XPE strains using a re®nement of a cell fusion assay(Itoh described (Itoh et al, 1994, 1995a). Cells were trypsinized and were plated et al, 1994, 1995a, 1996b). We report here that three XPE Ddb+ on to 60 mm plates. After incubation for 10 h, theywere washed with 2 strains have been misclassi®ed, whereas two Ddb± strains are XPE. phosphate-buffered saline, irradiated with 3±10 J per m of UVC (254 nm) at a ¯uence rate of 0.7 J per m2 per s without medium present, and then Hence it is feasible that DDB2 gene is the true XPE gene and XPE fresh medium was applied. After 10±14 d, the plates were washed with Ddb+ does not exist. phosphate-buffered saline, ®xed with 80% (vol/vol) methanol, and stained with Giemsa, colonies were counted, and analyzed by survival curves. To MATERIALS AND METHODS determine the effect of caffeine on UV survival, cells were plated, washed, Cells and culture conditions The cell strains used in this studyare and irradiated as described above except that caffeine was added to a ®nal shown in Table I. All cells were cultured in Dulbecco's modi®ed Eagle's concentration of 1 mM just after UV irradiation and the plates were minimum essential medium (ICN, Irvine, CA) supplemented with 10% incubated for 2±3 wk. (vol/vol) fetal bovine serum (ICN), penicillin G (100 units per ml), and streptomycin (100 mg per ml) in a humidi®ed 5% CO incubator at 37°C. Post-replication repair (PRR) assay Semiconservative DNA 2 replication after UV irradiation was measured bya previouslydescribed method (Itoh et al, 1995a, 1996b). Cells were irradiated with UV light Table III. Recovery of RNA synthesis. (254 nm) at a dose of 5 J per m2, returned to fresh 10F-Dulbecco's modi®ed Eagle's minimum essential medium, and incubated for 60 min before being labeled for 30 min with 0.37 MBq per ml (10 mCi per ml) of [3H]thymidine. a b Cell strain RRS Range of relative rate (%) Cells were harvested, suspended in phosphate-buffered saline, then X- irradiated at a dose of 20 Gyon ice. The suspension was layeredon to 5 ml XP43TO/Mori 41.5 6 1.5/39.0 6 1.5 99±115 of a 5±20% (wt/vol) alkaline sucrose gradient containing 0.1 M NaOH, and XP82TO/Mori 44.0 6 2.0/41.0 6 1.0 100±115 centrifuged at 237,600 3 g for 90 min at 4°C in a RPS55T rotor (Hitachi, XP89TO/Mori 5.4 6 0.9/37.8 6 1.9 11±18 Tokyo, Japan). After centrifugation, drop fractions were collected on to XP24KO/Mori 12.0 6 2.4/37.2 6 3.3 24±42 Whatman Grade 17 paper strips, and the acid-insoluble radioactivitywas GM02415B/Mori 41.0 6 1.4/34.0 6 1.4 112±130 counted in a liquid scintillation counter.

aTo measure RRS (recoveryof RNA synthesis),cells were irradiated with UV Measurement of DNA synthesis after UV irradiation The rate of at a dose of 15 J per m2, incubated for 23 h, and labeled with [3H]uridine(40 mCi DNA synthesis after UV irradiation was measured as described (Itoh et al, per ml) for 1 h. RRS is given in grains/nucleus. The exposure time was 24 h. Data 1996b, 1999). Brie¯y, cells were irradiated or mock irradiated with UVC are mean 6 SEM of 50 determinations. (254 nm) at a dose of 5 J per m2 in fresh 10F-Dulbecco's modi®ed Eagle's bRange of relative rate is range of RRS values (grain numbers) of cell line tested minimum essential media. After incubation for 2±3 d, theywere pulse- compared with those of Mori (normal control) cells. labeled with [3H]thymidine (10 mCi per ml) for 30 min, and then lyzed with

Figure 1. UV survival. Appropriate numbers of cells were seeded on to 60 mm dishes. After UV irradiation at a ¯uence rate of 0.7 J per m2 per s, cells were incubated for 10±14 d, ®xed with 80% methanol, and stained with Giemsa. Caffeine was added as indicated at a ®nal concentration of 1 mM just after UV irradiation and the plates were incubated for 2±3 wk. Each point represents an average of three independent experiments; error bar, SEM. Cells utilized were: (a±d) Mori cells (h) a normal control. (a) GM02415B (n); XP82TO (d). (b) XP43TO (d); XP2SA (XPV) (s); XP43TO 1 mM caffeine (m); XP2SA 1 mM caffeine (n). (c) XP89TO (d); Nps8 (XPF) (s). (d) XP24KO (d); Kps3 (UVsS) (s); Mps1 (CSA) (n). VOL. 114, NO. 5 MAY 2000 RECLASSIFICATION OF XPE CELLS 1025

0.2 M NaOH, harvested, and applied on to Whatman Grade 17 paper dose of 30 J per m2 (UDS) or 15 J per m2 (RRS). Then, UDS or RRS was strips. After washing with 5% trichloroacetic acid, the acid-insoluble measured as above. radioactivitywas counted byliquid scintillation. Each point represents an average of four experiments. DNA sequencing DNA sequencing was performed as described (Itoh et al, 1999). Genomic DNA was analyzed for mutations at nucleotide 730 Measurement of UDS, recovery of RNA synthesis (RRS) and RDS (XP82TO) and 818 (GM02415B). Polymerase chain reaction was used to after UV irradiation UDS, RRS, and RDS were measured as amplifythe DNA using the following two primers: 632: 5¢ -CTA- previouslydescribed (Itoh et al, 1994, 1995a, b, 1996a, b, 1999). Twenty GTAGCCGAATGGTGGTC and 920: 5¢-TCGGATCTCGCTCTTCT- microliter aliquots of the test and control cell suspension were plated with a GGT. The polymerase chain reaction fragments were sequenced with micropipette in two rows on a single coverslip (18 mm 3 18 mm) for each primers 632, 920, and 1000: 5¢-ATGGGTGTGAGGTGCTGGA using a experiment, as variations among coverslips has been observed to interfere Dye terminator cycle sequencing kit (PE Applied Biosystems, Foster City, with comparisons between cell strains. For UDS cells were irradiated with CA) with a Model 373S DNA sequencer (PE Applied Biosystems). UVC (254 nm) at a dose of 10 J per m2 or 30 J per m2, immediatelylabeled with [3H]thymidine (40 mCi per ml) for 2.5 h, and ®xed. Coverslips were RESULTS mounted on glass slides, dipped in Kodak NTB-3 emulsion, and exposed for 24 h at 4°C. Grains above nuclei were counted under a microscope. For Heterogeneity of UV sensitivity in cells previously classi®ed RRS, cells were irradiated with UVC (254 nm) at a dose of 15 J per m2. as XPE As patients with XPE show mild clinical manifestations After UV irradiation, theywere incubated for 23 h, and labeled with and heterogeneity(Kataoka and Fujiwara, 1991; Keeney et al, 1992; [3H]uridine (40 mCi per ml) for 1 h. Autoradiographywas performed as Cleaver and Kraemer, 1995; Friedberg et al, 1995), we ®rst above. For RDS cells were irradiated with UVC (254 nm) at a dose of 15 J examined the UV sensitivityof ®ve cell strains classi®ed as XPE by per m2. After UV irradiation, theywere incubated for 6 h and then labeled the colony-forming assay. We observed that the UV sensitivities of with [3H]thymidine (15 mCi per ml) for 1 h. Autoradiographywas GM02415B, XP82TO, and XP43TO cells were indistinguishable performed as above. from that of normal (Mori) cells, whereas XP89TO cells was slightlymore sensitive than normal ( Fig 1). XP24KO cells were Complementation assay Cell fusion was performed on a small scale as described previously(Itoh et al, 1994, 1995a, b, 1996a). Twentymicroliter signi®cantlymore sensitive to UV than the normal or the cells s aliquots (1±2 3 104 cells) of cell suspension fused with Sendai virus were classi®ed as XPE; theywere as sensitive as the UV S Kps3 cells, but plated in the center of coverslips (18 mm 3 18 mm) in 30 mm dishes and less so than the CSA Mps1 cells (Fig 1d). These results indicated 20 ml (1±2 3 104 cells) of parental cells was plated on each side of the fused that UV sensitivityof previouslyclassi®ed XPE cells showed cells. Cells were incubated for 20 h and then irradiated with UV light at a signi®cant heterogeneity.

Figure 2. RRS levels after UV irradiation in fused cells. The procedure for the complementa- tion assayis described in Materials and Methods.To measure RRS levels, cells were irradiated with 15 J per m2 UV light, incubated for 23 h, and then labeled with [3H]uridine (40 mCi per ml) for 1 h. Arrows show fused cells (a) between XP89TO and Kps5 (XPD); (b) XP89TO and Kps6 (XPF); (c) XP24KO and GM10905 (CSB); (d) and XP24KO and Kps3 (UVsS). Scale bar:5mm.

Table IV. Complementation of XP89TO cells by cell fusion.

Fused cells Cell type RRS (grains/nucleus 6 SEM)a Complementation

XP89TO/Kps5 (XPD)b Fused cells 21.3 6 1.7 + XP89TO 4.9 6 1.1 Kps5 5.8 6 0.8 XP89TO/Kps5 (XPD)b Fused cells 40.8 6 3.8 + XP89TO 16.3 6 2.8 Kps5 12.8 6 1.3 XP89TO/Kps6 (XPF) Fused cells 20.3 6 3.3 ± XP89TO 18.1 6 2.6 Kps6 19.8 6 3.0 XP89TO/Nps8 (XPF) Fused cells 7.7 6 1.6 ± XP89TO 12.6 6 1.2 Nps8 4.2 6 0.7

aCell fusion was performed as described in Materials and Methods. Data are mean 6 SEM of 50 determinations. bThe same experiment was performed twice to con®rm the result. 1026 ITOH ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Heterogeneity of UDS levels in cells previously classi®ed as XPE Although it has been reported that UDS levels of XPE cells were 40±60% those of normal cells (Bootsma et al, 1970; Kleijer et al, 1973; de Weerd-Kastelein et al, 1974; Kraemer et al, 1975b; Fujiwara et al, 1985; Okuno et al, 1994; Yamaizumi and Sugano, 1994; Cleaver and Kraemer, 1995), we have found that these relative levels varyamong measurements (Itoh and Yamaizumi, unpublished data), and we attributed this variation mainlyto differences among coverslips. To solve this problem, we have recentlydeveloped a simple method for measuring relative UDS, RRS, or RDS after UV irradiation bysimultaneouslyplating two cell lines on the same coverslip (Itoh et al, 1994, 1995a, 1996b).

Figure 4. Comparison of the sizes of DNA synthesized after UV irradiation in XP43TO, GM02415B, XP82TO, and XP2SA cells to Figure 3. Rate of DNA synthesis after UV irradiation. Cells were that with no irradiation. Cells were irradiated with UV light at a dose of irradiated with UV light at a dose of 5 J per m2. After 0, 2, 4, 6, or 8 h 5 J per m2, grown for 1 h, and then incubated for 30 min with incubation, theywere pulse-labeled with [ 3H]thymidine (10 mCi per ml) [3H]thymidine (10 mCi per ml). Centrifugation of the DNA through a 5± for 30 min. Cells utilized were: (a±c) Mori cells, normal control (h); (a) 20% (wt/vol) alkaline sucrose gradient was performed for 1.5 h at XP24KO (d); Kps3 (UVsS) (s); GM10905 (CSB) (n). (b) XP43TO (d); 237,600 3 g. The top of the gradient is to the right. (a) XP43TO; (b) XP2SA (XPV) (s). (c) GM02415B (n); XP82TO (d). Each value is the GM02415B; (c) XP82TO. h, unirradiated cells (control); j, UV irradiated mean for three or four experiments; error bar, SEM. cells. VOL. 114, NO. 5 MAY 2000 RECLASSIFICATION OF XPE CELLS 1027

The UDS levels are shown in Table II. Among the cells previously investigate the remaining four cell strains in more detail, we classi®ed as XPE, onlyXP89TO cells had signi®cantlyreduced compared the rates of post-UV DNA synthesis (Fig 3). The results UDS. As Fujiwara et al (1985) measured UDS levels in XP24KO of these experiments were: (i) the post-UV DNA synthesis rates for cells at the UV dose of 10 J per m2, we also repeated the experiment XP24KO cells were similar to those for UVsS (Kps3) and CSB at that dose in experiment V of Table II. Although the UDS levels (GM10905) cells (Fig 3a); (ii) the pattern for XP43TO cells was in the cells previouslyclassi®ed XPE were statisticallyreduced similar to that of XPV (XP2SA) cells (Fig 3b); and (iii) the patterns compared with those in normal Mori cells in some cases (p < 0.001) for XP82TO and GM02415B cells were similar to that of normal (Table II), when we examined the normal range of UDS levels (Mori) cells (Fig 3c). among several normal control cells (Mori, Turu, Sono, Goryo, s Umi, and Mura (experiments III and IV of Table II), similar Reassignment of XP24KO cells as UV S XP24KO cells variations in UDS levels were observed. Therefore, with the showed normal UDS (Table II), reduced RRS (Table III), and a exception of XP89TO, the UDS levels of the strains previously similar pattern in the rate of post-UV DNA synthesis compared s classi®ed as XPE strains were within the range of normal cells with UV S and CS cells (Fig 3a), suggesting that theybelong to s (Table II). UV S or CS (Itoh et al, 1996b). We performed complementation assays (Itoh et al, 1994, 1995a,b) between XP24KO cells and CS or Heterogeneity of RRS levels in cells previously classi®ed as UVsS cells. These results reclassi®ed XP24KO cells as UVsS XPE RRS levels are summarized in Table III. XP43TO, (Fig 2c, d, Table V). XP82TO, and GM02415B cells showed normal RRS levels. Reassignment of XP43TO Cells as XPV XP43TO cells RRS levels of XP89TO and XP24KO cells, however, were clearly showed almost normal UDS (Table II), almost normal RRS reduced. (Table III), and a pattern for the rate of post-UV DNA synthesis Reassignment of XP89TO cells as XPF XP89TO cells that was similar to that of XPV (Fig 3b), properties that are each showed slight UV sensitivity( Fig 1c) and a clear reduction in characteristic XPV cells (Itoh et al, 1996b). Therefore, we measured both UDS and RRS levels (Tables II and III), reductions more RDS after UV irradiation with or without caffeine and conducted a characteristic of XPD or XPF cells (Itoh and Yamaizumi, PRR assay. The RDS level of XP43TO cells was reduced in the unpublished data). Therefore, we performed complementation absence of caffeine, with an additional reduction in the presence of assays (Itoh et al, 1994, 1995a, b, 1996c) between XP89TO cells 1 mM caffeine (Table VI). This combination of three repair and XPD (Kps5) or XPF (Kps6 and Nps8) cells. The results markers (normal UDS, normal RRS, and reduced RDS which is reclassi®ed XP89TO cells as XPF (Fig 2a, b, Table IV). enhanced bycaffeine) is characteristic speci®callyof XPV (Itoh et al, 1996b). Furthermore, by the PRR assay, the DNA synthesized Heterogeneity of the rate of DNA synthesis after UV after UV irradiation in XP43TO cells was found to be shorter than irradiation in cells previously classi®ed as XPE To that in unirradiated cells (Fig 4a). In normal cells DNA synthesized

Table V. Complementation of XP24KO cells by cell fusion.

Fused cells Cell type RRS (grains/nucleus 6 SEM)a Complementation

XP24KO/Kps3 (UVsS)b Fused cells 5.6 6 0.6 ± XP24KO 5.0 6 0.4 Kps3 7.3 6 1.3 XP24KO/Kps3 (UVsS)b Fused cells 12.2 6 2.3 ± XP24KO 10.1 6 2.1 Kps3 9.3 6 1.9 XP24KO/GM10905 (CSB) Fused cells 29.3 6 2.0 + XP24KO 7.5 6 1.1 GM10905 8.7 6 1.0 XP24KO/CS1MO (CSB) Fused cells 16.6 6 1.4 + XP24KO 9.1 6 1.0 CS1MO 8.7 6 1.4 XP24KO/Mps1 (CSA) Fused cells 33.2 6 4.0 + XP24KO 9.1 6 1.0 Mps1 10.1 6 2.5

aCell fusion was performed as described in Materials and Methods. Data are mean 6 SEM of 50 determinations. bThe same experiment was performed twice to con®rm the result.

Table VI. Recovery of replicative DNA synthesis.

RDSa Range of relative RDSb (%)

Cell strain With caffeine Without caffeine With caffeine Without caffeine

GM02415B/Moric 70.6 6 6.3/57.3 6 4.0 87.0 6 9.5/73.8 6 6.3 105±144 97±143 XP43TO/Mori 8.9 6 1.1/79.3 6 8.3 14.7 6 2.1/65.1 6 12.9 9±14 16±32 XP82TO/Mori 55.9 6 4.5/54.8 6 4.9 58.8 6 4.5/71.4 6 6.6 86±121 70±98

aTo measure RDS, cells were irradiated with UV light at a dose of 15 J per m2, incubated for 6 h, and then labeled with [3H]thymidine (15 mCi per ml) for 1 h. Exposure times were 24 h (with caffeine) or 13 h (without caffeine), respectively. Data are mean 6 SEM of 50 determinations. bRange of Relative RDS is the range of RDS values of cell strain tested compared with those of Mori (normal control) cells. cItoh et al, 1996b. 1028 ITOH ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY after UV irradiation is the same size as that of unirradiated normal DISCUSSION cells (Itoh et al, 1995a, 1996b, 1999). Moreover, a diminution by XPE was designated as the ®fth complementation group based on a caffeine on UV survival was also found (Fig 1b). These results taken cell fusion technique; however, XPE strains exhibit the highest together indicate that XP43TO cells ought to be reassigned as UDS level of all the NER defective XP groups (Kraemer et al, XPV. 1975a,b). The patients, XP2RO (GM02415B) and XP3RO were Identi®cation of mutations in the DDB2 Gene in XP82TO ®rst assigned to XPE and were clinicallyclassi®ed as ``classical XP, and GM02415B cells As both XP82TO and GM02415B cells light to moderatelysevere'' (Bootsma et al, 1970; Kleijer et al, 1973; had similar phenotypes for DNA repair markers (Figs 1a and 3c, de Weerd-Kastelein et al, 1974). Next, XP24KO cells were Tables II and III), we examined whether the two strains belonged assigned to XPE using GM02415B cells as the complementing cells to the same complementation group. Although it seemed that both in the cell fusion assay(Fujiwara et al, 1985). After that, XP80TO, GM02415B and XP82TO cells complemented NER-defective XP81TO, and XP82TO cells were assigned to XPE using XPA, XPC, XPD, XPF, and XPG cells (data not shown), the grain XP24KO cells (Kondo et al, 1988). The patient XP24KO was a numbers in fused cells were almost the same as those of unfused 15 yold Japanese female who showed acute sun sensitivityreaction GM02415B or XP82TO cells, because these cells showed almost without blistering at age 1. Small pigmented freckles and normal ranges of UDS (Table II) and RRS (Table III). Indeed telangiectasia on her face were observed at about age 11 or 12. similar ranges of results were found in complementation assays She had no benign or malignant neoplasias (Fujiwara et al, 1985). between NER-defective XPA, XPC, XPD, XPF, and XPG cells Judging bythese clinical manifestations, XP24KO was clinically and normal or XPV cells. Furthermore, UDS levels in fused cells similar to patients with UVsS (Itoh et al, 1995a, 1996d), which has were sometimes slightlyhigher than those of individual XPV or been con®rmed bythe reassignment of XP24KO cells to UV sSin normal cells (Itoh and Yamaizumi, unpublished data). Therefore, this study. bycomplementation we could not de®nitivelyclassifyXP82TO or XP80TO, XP81TO, and XP82TO cells were previously GM02415B cells. assigned to XPE using XP24KO cells as the complementing cells To rule out whether these cells belong to XPV, we examined in the cell fusion assay. From the current study, it does appear that RDS levels with or without caffeine (Table VI) and conducted a XP82TO is indeed XPE. As we determined that XP24KO cells are PRR assay( Fig 4b,c). Both XP82TO and GM02415B cells in fact UVsS and show almost normal UDS levels, however, the showed almost normal RDS levels, and no effect of caffeine was classi®cation of XP80TO and XP81TO remains to be re-evaluated. found (Table VI). Furthermore, the synthesized DNA after UV Currently, six XPE patients have been reported to have defects irradiation was the same size as that of unirradiated cells (Fig 4b, c). in DDB activity(Chu and Chang, 1988; Keeney et al, 1992; Otrin The results of these experiments indicate that XP82TO and et al, 1998). All of these patients exhibit a mild clinical form of XP GM02415B cells do not belong to XPV. and UDS levels of approximately50% of those of normal cells have The absence (Ddb±) of a damage-speci®c DNA binding (DDB) been reported. UDS is usuallymeasured on single cell-seeded activityhas been identi®ed in six of the 16 cell strains previously coverslips (Bootsma et al, 1970; Kleijer et al, 1973; de Weerd- classi®ed XPE (Kataoka and Fujiwara, 1991; Keeney et al, 1992; Kastelein et al, 1974; 1975b; Kraemer et al, 1975a; Okuno et al, Otrin et al, 1998; Itoh et al, 1999). DDB protein is a heterodimer of 1994; Yamaizumi and Sugano, 1994). Variations among coverslips, 127 kDa (DDB1) and 48 kDa (DDB2) subunits (Keeney et al, 1993; however, are recognized as a limitation in measuring UDS Dualan et al, 1995) and mutations in DDB2 have been reported in accurately. Grain numbers on nuclei, which show UDS levels XP2/3RO (GM02415B) and XP82TO Ddb± XPE cells by upon autoradiography, are affected by various factors, such as UV sequencing analysis of both DDB2 cDNA and genomic DNA dose, labeling time, amount of label, kind of emulsions, thickness of (Nichols et al, 1996). We have characterized the genomic emulsions, exposure time, and the temperature at time of organization of the DDB2 gene and found it to be comprised of development. Although most of these factors are kept uniform, it 10 exons (Itoh et al, 1999). We have recon®rmed the previously is dif®cult to control differences in the thickness of emulsions on report that XP82TO carried a homozygous A®G transversion at different coverslips. Therefore, it is necessaryto compare cells on nucleotide position 730 which is now assigned to exon 6 (data not the same coverslip and to con®rm the results using other criteria. In shown). In the cell strain GM02415B, we located a homozygous particular, variations among preparations make the counting of G®A transversion at nucleotide position 818, also in exon 6 (data grain numbers dif®cult especiallyin the case of high UDS. To not shown). Therefore, both of these cell strains have mutations in resolve this problem, we adopted an alternative assaysystemthat the DDB2 gene, in addition to being Ddb± and XPE. includes both XP and normal control cells on the same coverslip,

Table VII. Summary of characteristics of cells previously classi®ed as XPE.

Cell strain NERa PRRb UV sensitivityc DDB activityd Mutations in Proposed complementation DDB2 gene group

XP24KO Normal NDe ++g NDe UVsS XP43TO Normal Abnormal +f NDe NDe XPV XP82TO Near normal Normal ± ±h,i A730Gk XPE XP89TO Signi®cantlyreduced ND e ++h Nonek XPF GM02415B Near normal Normal ± ±j G818Ak XPE

aNER (nucleotide-excision-repair) determined byUDS (unscheduled DNA synthesis)levels. bPRR determined by PRR and RDS (recovery of replicative DNA synthesis) assays. cUV sensitivitymeasured bya colony-formingassaywith/without caffeine. dDDB (damage-speci®c DNA binding) activity assayed by an electrophoretic mobility shift assay. eNot determined. fUV sensitivitywas observed in the presence of caffeine. gKataoka and Fujiwara, 1991. hKeeney et al, 1992. iItoh et al, 1999. jChu and Chang, 1988. kNichols et al, 1996. VOL. 114, NO. 5 MAY 2000 RECLASSIFICATION OF XPE CELLS 1029 and con®rmed the results byan additional experiment (Itoh et al, Hayes S, Shiyanov P, Chen X, Raychaudhuri P: DDB, a putative DNA repair 1994, 1995a, 1996a, b, c). Moreover, we used NTB-3 emulsion, as protein, can function as a transcriptional partner of E2F1. Mol Cell Biol 18:240± 249, 1998 this is an emulsion with the highest sensitivity, allowing even Itoh T, Ono T, Yamaizumi M: A new UV±sensitive syndrome not belonging to any slightlyreduced DNA repair synthesisto be detected as a reduced complementation groups of xeroderma pigmentosum or Cockayne syndrome: number of grains. siblings showing biochemical characteristics of Cockayne syndrome without Other factors, such as cell senescence, thymidine kinase (TK) typical manifestations. Mutat Res 314:233±248, 1994 Itoh T, Fujiwara Y, Ono T, Yamaizumi M: UVs syndrome, a new general category levels, and intracellular nucleotide pools, also affect the incorpora- of photosensitive disorder with defective DNA repair, is distinct from tion of label in UDS experiments. Recently, it was reported that xeroderma pigmentosum variant and rodent complementation group 1. Am J DDB protein binds to E2F1, thus modulating the activityof E2F1 Hum Genet 56:1267±1276, 1995a as a regulator (Hayes et al, 1998). As TK is one of the Itoh T, Watanabe H, Yamaizumi M, Ono T: A young woman with xeroderma genes that are controlled byE2F1, TK activitymaybe lower in pigmentosum complementation group F and a morphoeic basal cell carcinoma. Br J Dermatol 32:122±127, 1995b DDB-defective cells. Taken together, we feel that it is verydif®cult Itoh T, Cleaver JE, Yamaizumi M: Cockayne syndrome complementation group B solelyto use UDS complementation criteria for cells with near- associated with xeroderma pigmentosum phenotype. Hum Genet 97:176±179, normal NER activity, such as XPE cells to assign groups. 1996a A summaryof the results of this studyare shown in Table VII. Itoh T, Ono T, Yamaizumi M: A simple method for diagnosing xeroderma pigmentosum variant. J Invest Dermatol 107:349±353, 1996b The three cell strains with defective DNA repair markers, each of Itoh T, Shiomi T, Shiomi N, Harada et al: Rodent complementation group 8 + which is of the Ddb phenotype, have been reassigned to XPF, (ERCC8) corresponds to Cockayne syndrome complementation group A. XPV, or UVsS. The remaining two cell strains, which have a Ddb± Mutat Res 362:167±174, 1996c phenotype, showed slightly reduced UDS levels, but almost normal Itoh T, Yamaizumi M, Ichihashi M, Hiro-oka M, Matsui T, Matsuno M, Ono T: phenotypes of other DNA repair markers. Together these results Clinical characteristics of three patients with UVs syndrome, a photosensitive + disorder with defective DNA repair. Br J Dermatol 134:1147±1150, 1996d show that (i) at least three XPE Ddb strains were misclassi®ed, and Itoh T, Mori T, Ohkubo H, Yamaizumi M: A newlyidenti®ed patient with clinical (ii) the clinical photosensitive syndrome, XP, has three subgroups xeroderma pigmentosum phenotype has a nonsense in the DDB2 based upon UDS: a low UDS subgroup (NER-defective XP), a gene and incomplete repair in (6±4) photoproducts. J Invest Dermatol 113:251± near normal UDS subgroup [DDB2-defective XP (Ddb± XPE)], 257, 1999 Johnson RE, Kondratick CM, Prakash S, Prakash L: hRAD30 mutations in the and a normal UDS subgroup [DNA polymerase eta-defective XP variant form of xeroderma pigmentosum. Science 285:263±265, 1999 (XPV)]. Because of this heterogeneity, to diagnose photosensitive Kataoka H, Fujiwara Y: UV damage-speci®c DNA-binding protein in xeroderma patients with/without skin cancer(s) and/or neurologic abnormal- pigmentosum complementation group E. Biochem Biophys Res Commun ities, it is necessaryto consider whether theysuffer from XP, UV sS, 175:1139±1143, 1991 or CS. Indeed, we previouslyreported that two CS patients KeeneyS, Wein H, Linn S: Biochemical heterogeneityin xeroderma pigmentosum complementation group E. Mutat Res 273:49±56, 1992 (GM10903 and GM10905) have apparent XP phenotypes KeeneyS, Chang GJ, Linn S: Characterization of a human DNA damage binding (Greenhaw et al, 1992; Itoh et al, 1996a). Of the ®ve XPE strains protein implicated in xeroderma pigmentosum E. J Biol Chem 268:21293± tested in this study, only the Ddb± strains remain classi®ed as XPE. 21300, 1993 This suggests that XPE is possiblya single group with DDB2 Kleijer WJ, de Weerd-Kastelein EA, Sluyter ML, Keijzer W, de Wit J, Bootsma D: implicated as the defective gene. As DDB has not been UV-induced DNA repair synthesis in cells of patients with different forms of xeroderma pigmentosum and of heterozygotes. Mutat Res 20:417±428, 1973 demonstrated to be involved directlyin NER, we are currently Kondo S, Fukuro S, Mamada A, Kawada A, Satoh Y, Fujiwara Y: Assignment of investigating possible alternative role(s) for it in response to DNA three patients with xeroderma pigmentosum to complementation group E and damage. their characteristics. J Invest Dermatol 90:152±157, 1988 Kondo S, Mamada A, Miyamoto C, Keong CH, Satoh Y, Fujiwara Y: Late onset of skin cancers in 2 xeroderma pigmentosum group F siblings and a review of 30 Japanese xeroderma pigmentosum patients in group D, E and F. We are grateful to Dr. Yoshisada Fujiwara and Dr. Seiji Kondo for providing Photodermatology 6:89±95, 1989 XP24KO, XP43TO, XP82TO, and XP89TO cells. We also thank Dr. Kraemer KH, Coon HG, Petinga RA, Barrett SF, Rahe AE, Robbins JH: Genetic Hironori Niki and Mrs. Yuka Itoh for preparation of the manuscript. This work was heterogeneityin xeroderma pigmentosum: complementation groups and their relationship to DNA repair rates. Proc Natl Acad Sci USA 72:59±63, 1975a supported in part by Grands-in-Aid from the Ministry of Education, Science, Kraemer KH, de Weerd-Kastelein EA, Robbins JH, Keijzer W, Barrett SF, Petinga Sports, and Culture of Japan (to T.I. and M.Y.), by the Naito Foundation (to RA, Bootsma D: Five complementation groups in xeroderma pigmentosum. T.I.), by the Kao Foundation for Arts and Sciences (to T.I.), and by the USPHS Mutat Res 33:327±340, 1975b (P30ES08196; to S.L.). Lehmann AR, Kirk-Bell S, Arlett CF, Paterson MC, Lohman PH, de Weerd- Kastelein EA, Bootsma D: Xeroderma pigmentosum cells with normal levels of excision repair have a defect in DNA synthesis after UV-irradiation. Proc Natl Acad Sci USA 72:219±223, 1975 REFERENCES Lehmann AR, Kirk-Bell S, Arlett CF, Harcourt SA, de Weerd-Kastelein EA, Keijzer W, Hall-Smith P: Repair of ultraviolet light damage in a varietyof human Bootsma D, Mulder MP, Pot F, Cohen JA: Different inherited levels of DNA repair ®broblast cell strains. Cancer Res 37:904±910, 1977 replication in xeroderma pigmentosum cell strains after exposure to ultraviolet irradiation. Mutat Res 9:507±516, 1970 Masutani C, Kusumoto R, Yamada A, et al: The XPV (xeroderma pigmentosum Chu G, Chang E: Xeroderma pigmentosum group E cells lack a nuclear factor that variant) gene encodes human DNA polymerase eta. Nature 399:700±704, 1999 binds to damaged DNA. Science 1242:564±567, 1988 Nichols AF, Ong P, Linn S: Mutations speci®c to the xeroderma pigmentosum Cleaver JE, Kraemer KH. Xeroderma pigmentosum. In: Scriver CR, Beaudet AL, group E ddb± phenotype. J Biol Chem 271:24317±24320, 1996 SlyWS, Valle D (eds). The Metabolic and Molecular Basis of Inherited Disease, 7th Okuno Y, Tateishi S, Yamaizumi M: Complementation of xeroderma pigmentosum edn. New York: McGraw-Hill, 1995, pp. 4393±4419 cells bymicroinjection of mRNA fractionated under denaturing conditions: an Dualan R, BrodyT, KeeneyS, Nichols AF, Admon A, Linn S: Chromosomal estimation of sizes of XP-E and XP-G mRNA. Mutat Res 314:11±19, 1994 localization and cDNA cloning of the genes (DDB1 and DDB2) for the p127 Otrin VR, Kuraoka I, Nardo T, et al: Relationship of xeroderma pigmentosum group and p48 subunits of a human damage-speci®c DNA binding protein. Genomics E DNA repair defect to the chromatin and DNA binding UV-DDB 29:62±69, 1995 and . Mol Cell Biol 18:3182±3190, 1998 Friedberg EC, Walker GC, Siede W: Nucleotide excision repair in mammalian cells. Rude J, Friedberg EC: Semiconservative deoxyribonucleic acid synthesis in In: DNA Repair and Mutagenesis. American Society for Microbiology, Washington, unirradiated and ultraviolet-irradiated xeroderma pigmentosum and normal 1995, pp. 283±365 human skin ®broblasts. Mutat Res 42:433±442, 1977 Fujiwara Y, Uehara Y, Ichihashi M, Yamamoto Y, Nishioka K: Assignment of 2 Sijbers AM, de Laat WL, Ariza RR, et al: Xeroderma pigmentosum group F caused a patients with xeroderma pigmentosum to complementation group E. Mutat Res defect in a structure-speci®c DNA repair endonuclease. Cell 86:811±822, 1996 145:55±61, 1985 de Weerd-Kastelein EA, Keijzer W, Bootsma D: A third complementation group in Greenhaw GA, Hebert A, Duke-Woodside ME, Bulter IJ, Hecht JT, Cleaver JE, xeroderma pigmentosum. Mutat Res 22:87±91, 1974 Thomas GH, Horton WA: Xeroderma pigmentosum and Cockayne Yamaizumi M, Sugano T: UV-induced nuclear accumulation of p53 is evoked syndrome: overlapping clinical and biochemical phenotypes. Am J Hum through DNA damage of activelytranscribed genes independent of the cell Genet 50:677±689, 1992 cycle. Oncogene 9:2775±2784, 1994