Probes. These molecules have low photobleaching emission dipole orientation. However, the 1.25-NA (1977); J. Opt. Soc. Am. 67, 1607 (1977). rates(10-6 to 10-8 per excitation),near-unityquan- oil-immersion objective used here collected a calcu- 22. For a dipole on the air side of the interface,LlllLoois tum yield in a semirigid environment, and an absorp- lated 65% of the total light emitted by a molecule at 0.92, L~/Loo is 1.6,andthe radiativelifetimeIsde- tion cross section of 1.35 x 10-16 cm2 at 532 nm, the PMMA-alr interface [E. H. Helen and D. Axelrod, creased for a perpendicular emission dipole. estimated by us on the basis of the measured ab- J. Opt. Soc. Am. B 4, 337 (1987)], nearly indepen- 23. Aliphatic hydrocarbon immersion oil, type FF; sorption in methanol. dent of the emission-dipole orientation. The total col- Cargille Laboratories, Cedar Grove, NJ. 15. A quartz cover slip was first spin-coated with one lection or detection efficiency was 15%. 24. We deposited several micrometers of PMMA by spin- drop (0.2 ml) of 0.1 % by weight PMMA in chloroben- 18. A focused laser beam has a very small longitudinal coating several drops of a solution of 5% by weight zene and then spin-coated with one drop of a 1-nM field component along the z axis, on the order of PMMA in toluene and heating the sample to dry. solution of Dil'2C(12) in toluene. The nanomolar Elkw, whereEis the tangential laser field, k is 2'IT11.., 25. E. Akesson, V. Sundstrom, 1'. Gillbro, Chem. Phys. dye solution was freshly prepared from a micromo- and w is the focused spot size [M. Lax, W. H. Lett.121,513(1985). lar dye solution with each sample. Loulsell, W. B. McKnight, Phys. Rev. A 11, 1365 26. M. O. Scully, Appl. Phys. B 51,238 (1990). 16. Although the calculated saturation intensity is 1 MW (1975)]. 27. We obtained an ensemble measurement by raster- cm-2, in a separate study we found that the satura- 19. The emission dipole, which lies along the long axis of scanning a 36-fLm2 sample area prepared with 20 tion of the fluorescent transition was determined by the molecule that connects the two indole rings, is times higher coverage of molecules at a laser inten- transitions from the excited singlet state to a meta- oriented 28° from the absorption dipole [D. Axelrod, sity of 500 W cm-2, while continuously collecting stable state, probably the triplet state, with an inter- Biophys. J. 26, 557 (1979)]. fluorescence. It thus averaged over several hundred system crossing rate of about 0.15% and a triplet 20. The radiative lifetime for a broad molecular spectrum molecules. lifetime of 0.4 ms. is proportional to the spectrally averaged inverse 28. P. R. Bevington, Data Reduction and Error Analysis cube of the emission frequency [SoJ. Strickler and R. 17. The maximum fluorescence rate is Ro = TJ

dem repeat polymorphism (STRP) markers Positional Cloning of the Werner's at the glutathione reductase (GSR) gene Syndrome Gene and D8S339 were shown to be in linkage disequilibrium with WS in Japanese WS Chang-En Yu,* Junko Qshima,* Ying-Hui Fu,* Ellen M. Wijsman, patients (9, 10), indicating that these mark- Fuki Hisama, Reid Alisch, Shellie Matthews, Jun Nakura, ers are most likely close to WRN. Tetsuro Miki, Samir Quais, George M. Martin, To clone the WS gene, we generated a map from yeast artificial chromosomes John Mulligan, Gerard D. Schellenbergt (YACs), PI clones, and cosmid contigs (Fig. 1), starting at GSR and extended by Werner's syndrome 0NS) is an inherited disease with clinical symptoms resembling walking methods to cover approximately 3 premature aging. Early susceptibility to a number of major age-related diseases is a key Mb (11). Eighteen STRP markers (Fig. 1B) feature of this disorder. The gene responsible for WS (known as WRN) was identified by were identified in the contig; probable re- positional cloning. The predicted protein is 1432 amino acids in length and shows sig- combinants were detected at D8S2194 nificant similarity to DNA helicases. Four mutations in WS patients were identified. Two (which excluded the region telomeric to of the mutations are splice-junction mutations, with the predicted result being the ex- this marker) and at D8S2186 [which ex- clusion of exons from the final messenger RNA. One of these mutations, which results in cluded the region centromeric to this mark- a frameshift and a predicted truncated protein, was found in the homozygous state in 60 er (12)], making the 1.2 to 1.4 Mb interval p~ent of Japanese WS patients examlnea.lfle other two mutations are nonsense mutations. I ne Identification of a mutated putative helicase as the gene product of the C.-E. Yu and S. Matthews, Geriatric Research Education WS gene suggests that defective DNA metabolism is involved in the complex process of and Clinical Center (182B), Veterans Affairs Puget Sound Health Care System, Seattle Division, 1660 South Co- aging in WS patients. lumbian Way, Seattle, WA 98108, USA, and Department of Neurology, University of Washington, Seattle, WA 98195, USA. J. Oshima, Geriatric Research Education and Clinical Center (182B), Veterans Affairs Puget Sound Health Care Werner's syndrome is a rare autosomal prevalent geriatric disorders such as Alz- System, Seattle Division, 1660 South Columbian Way, recessive disorder that is considered a par- heimer's disease and hypertension are not Seattle, WA 98108, USA, and Department of Pathology, tial model of human aging (1-3). WS observed in WS. Moreover, there are sub- University of Washington, Seattle, WA 98195, USA. Y.-H. Fu, R. Alisch, J. Mulligan, Darwin Molecular Corpo- patients prematurely develop a variety of tle discordances between WS and normal ration, 1631 220th Street, S.E., Bothell, WA 98021, USA. the major age-related diseases, including aging, such as a disproportionately severe E. M. Wijsman, Division of Medical Genetics, Department several forms of arteriosclerosis, malignant osteoporosis of the limbs relative to the of Medicine, and Department of Biostatistics, University of Washington, Seattle, WA 98195, USA. neoplasms, type II diabetes mellitus, osteo- trunk and the high prevalence of nonepi- F. Hisama, Department of Neurology, Yale University porosis', and ocular cataracts; these indi- thelial neoplasms in WS. Finally, there are School of Medicine, New Haven, CT 06510, USA. viduals also manifest early graying and loss unusual clinical features unrelated to ag- J. Nakura and 1'. Miki, Department of Geriatric Medicine, Osaka University Medical School, 2-2 Yamadaoka Suita, of hair, skin atrophy, and a generally aged ing, including ulcerations around the an- Osaka 565, Japan. appearance. Growth retardation occurs, kles and soft tissue calcification (1, 2). S. Ouais, Section of Endocrinology, Damascus City Hos- typically around the time of puberty, but The WS locus (WRN) was initially lo- pital, Damascus, Syria. G. M. Martin, Department of Pathology, University of medical problems are rare during child- calized to 8p12 (5) by linkage analysis and Washington, Seattle, WA 98195, USA. hood. Cell culture studies also suggest a the genetic position refined by both meiotic G. D. Schellenberg, Geriatric Research Education and parallel between WS and aging; the repli- and homozygosity mapping (5-7). Initial Clinical Center (182B), Veterans Affairs Puget Sound mapping (6-8) placed WRN in an 8.3- Health Care System, Seattle Division, 1660 South Co- cative life-span of fibroblasts from WS lumbian Way, Seattle, WA 98108, USA, and the Depart- patients is reduced compared with age- centimorgan (cM) interval flanked by ments of Medicine, Neurology, and Pharmacology, Uni- matChed controls and is similar to the markers D8S137 and D8S87 (Fig. 1); versity of Washington, Seattle, WA 98195, USA. life-span of fibroblasts taken from more D8S339, located within this interval, was 'These authors contributed equally to this work. the closest marker. Subsequently, short tan- tTo whom correspondence should be addressed.

, SCIENCE. VOL. 272 12 APRIL 1996 ~, ~ ;:'dYindivid"l; (4). How,v", wm, ~ Fig. 1. Genetic and physical maps of 088283 the WRN region. (A) The genetic map 088259 GSR 088278 088268 units are in centimorgans, assuming LPL 088136 088137 088131 088339 08887 FGFR1 088255 ANK1 sex-equal recombination rates (8). (B) The polymorphic loci are 8TRP mark- A ~~~~~ \ / , ers except for PPP2CB and 0882180 Telomere 8p I I I I I I I I I Centromere which are bi-allelic insertion-deletion 3.8 7.4 0.9 6.7 1.6 2.5 2.8 2.1 polymorph isms (11). The 8TRP loci at -IL... GSR and PPP2CB are 0882202 and 0882208, respectively. (C) The physi- r -, cal map has approximate distances in kilobases determined from sizes of 8 overlapping nonchimeric YACs, and from genomic ONA sequence from overlapping P1 clones 2233, 2253, 3833, 2236, and 3101. Marker and clone order was determined from the C 130 200 130 350 >200 sequence-tagged site (8T8) content of YACs, P1 clones, and cosmid clones and from genomic ONA sequence. (D) D 814E11:1080kb The YACs represent the minimal num- 896F4: 1200 kb

ber of YACs to cover the WRN region 763A7: 800 kb and are the YACs used for cONA selec- 780E6: 500 kb tion experiments. The small insert clones shown in (E)are the minimum number neededfor overlappingcover- 6738 - age of the WRN region and are P1 E 2934 - 2253 - 2932 - clones except for 8C11 and 176C6, 2233- 176C6- 3000 which are cosmids. A total of 20 P1 3833 - 2927- 2236 - 2294 - clones and 109 cosmid clones were / ~ identified from the 0882194 to 2237 ~ - 2246 - 8C11 - WRN gene 0882162 region. The complete 5469- genomic sequence (30) was deter- mined for eight P1 clones (2233, 2253, 3833, 2236, 2237, 2932, 2934, and three P1 clones (2234,2927, and 3000), and one mouse P1 clone (5469) was 6738) and one cosmid clone(176C6).Partial sequences were obtained for completely sequenced.

Fig. 2. Expression and ONA .$ ~ ~ from D8S2194 to D8S2186 (Fig. lC) the I:: OJ ;>, OJ minimal WRN region. Significant evidence sequence of the WRN gene. (A) A Northern blot (Clontech) 1;;.0; ~ OJ c;;~ ~ g for linkage disequilibrium (P < 0.03) be- 812345678 of multiple tissues, prepared A ~~£3S~~~ tween single markers and WS in a Japanese with 2 fLg of polyadenylated 2.0 patient population was obtained for 10 RNA per lane, was hybridized [D8S2198, D8S2196, D8S339, D8S540 with a 1970-bp probe gener- 9.5- 1.6 (GSR2), D8S2202 (GSRl), D8S2206, ated from the WRN helicase 7.5- D8S2134, D8S2162, D8S2174, and cONA clone by using primers 4.4- D8S2180] of the 18 STRP loci in the inter- K (5' -AGTGCAGTGGTGT- 2.4- val (12). CATCATAGC-3') and L (5'- CCT ATTT AATGGCACCCA- 1.0 , Potential expressed sequences in the in- terval between D8S2192 to D8S2186 were AAATGC-3'). The filter was 1.35- hybridized at 42°C in 50% identified exon trapping, hybridization by formamide for 16 hours and of complementary DNA (cDNA) libraries washed twice in 1x standard saline citrate (88G) and 0.1 % 808 at room temperature, followed by two to immobilized YACs, comparison of the washes in 0.1x 88C and 0.1% 808 at 65°C (30 min each wash). (B) RT-PCR products (16) were I genomic sequences to DNA.sequence data- produced with primers K and N (5' -ACCGCTTGGGATAAGTGCATGC-3') and the products resolved bases by means of BLAST, and analyzing by agarose gel electrophoresis. Lanes 1 and 8, 1-kb ladder size standard (Life Technologies); lane 2, i genomic sequence with the exon-finding Japanese normal control; lane 3, 8Y family, Japanese W8, mutation 2; lane 4, ZM family, Japanese , computer program GRAIL (13). BLAST W8, mutation 4; lane 5, Japanese normal control; lane 6, MCI family, Japanese W8, mutation 4; lane searches of the Eukaryotic Promotor Data- 7, Caucasian W8. RNA for RT-PCR was from Iymphoblastoid cell lines (lanes 2 to 4) or from fibroblast base (14) were performed to identify poten- cultures (lanes 5 to 7). Molecular sizes are indicated to the sides of (A) and (B) in kilobases. f tial regulatory elements. Each gene detec- tion method identifies short segments of expressed sequences, which can then be tor IIEj3 (GTF2E2), a j3-tubulin pseudo- The candidate WRN gene was detected used to screen an arrayed fibroblast cDNA gene 1 (TUBBPl), and six genes of un- by means of the genomic sequence of PI library (15) to identify longer cDNA known function were screened for muta- clone 2934, which was used to search the clones. Genes identified by this process tions and none were detected. WS expressed sequence tag (EST) database. A were screened for WRN mutations by se- patients were also screened for mutations 245-base pair (bp) EST (R58879) was quencing reverse transcriptase-polymerase in 14 GRAIL-predicted exons and three identified that was identical to three re- chain reaction (RT-PCR) products (16). putative promotor elements by sequencing gions of genomic sequence of 145, 94 (a Before identification of the WRN gene, PCR amplification fragments generated complete exon), and 6 bp in length, sep- GSR, protein phosphatase2 catalytic sub- from the genomic DNA from WS and arated in the genomic sequence by two normal individuals. introns (81 and 851 bp in length). An SCIENCE. VOL.272 . 12APRIL1996 259

~nit ~ (PPPlCR).g,n",! n=",iption fae- exon of the same gene was also identified 4-bp deletion spanning a splice junction is homozygous for this mutation. For 30 by exon trapping. The sequence from (mutation 3). In this kindred, three sibs Japanese kindreds examined, affected in- R58879 was used to identify longer cDNA with WS are homozygous for this muta- dividuals from 18 of the kindreds are ho- clones (17). Northern (RNA) blot analy- tion. A fourth sib, aged21, is homozygous mozygous for mutation 4. In three of these sis showed that R58879 is expressed; tran- for the same mutation but too young for a families two affected sibs were sampled, scriptsof ~ 5.8 and 8.1 kb were detected in definitive diagnosis of WS (1, 2). Al- and in each case both are homozygous. all tissues tested (Fig. 2A). The predomi- though these individuals are not from a Among mutation carriers, 12 of 16 have nant smaller transcript was present at consanguineous marriage, they do share the 141-bp allele at the GSR2 STRP, highest quantities in pancreas, followed by the same haplotype across the WS region which is overrepresented in WS cases (fre- placenta, muscle, and heart. Transcripts (12). This mutation was. not observed in quency = 0.40) and relatively rare in Jap- were also detected in RT-PCR products 96 Caucasian or 48 Japanese control indi- anese control individuals (frequency = from fibroblast and lymphoblastoid cell viduals (19, 20). 0.07) (10). The association of this allele line RNA (Fig. 2B). The completed se- A fourth mutation was first detected as with WS was responsible for the initial quence was assembled from 12 additional a RT-PCR product that was 95 bp shorter detection of linkage disequilibrium in this overlapping cDNA clones and by 5'- than products from other WS and control region (9). This mutation was not ob- RACE experiments (17) to yield a 5.2-kb individuals (Fig. 2B). Comparison of the served in 48 Caucasian WS patients. In final sequence (GenBank accession num- RT-PCR product sequence and the 187 Japanese control individuals, one het- ber L76937). genomic sequence from PI clone 2934 erozygote was observed for an estimated Four mutations in WS patients were revealed that the missing 95 bp corre- gene frequency of 0.003 (19, 20), which is detected in the gene corresponding to sponded to a single exon. The missing comparable with gene frequency estimates R58879 (Table 1). Two mutations were exon 'and flanking genomic segments were (0.001 to 0.005) based on WS prevalence nonsense mutations creating premature sequenced from the WS patient and con- rates and consanguinity estimates (2, 21). stop codons (mutations 1 and 2). Four trol individuals, and a single base change The protein predicted from the cDNA Japanese WS patients, the offspring of first of G-J>C was detected that changes a sequence is 1432 residues in length (Fig. cousin marriages, and one Caucasian, from splice donor sequence from ApG to ApC 3) and is highly similar to DNA helicases a second cousin marriage (18), were ho- (mutation 4) and results in a frame shift of from a wide range of organisms (Fig. 4). mozygous for the Arg1305TGA mutation codons 1078 to 1092. This WS patient is All seven helicase consensus domains are (mutation 1) (where the arginine codon at the offspring of a first cousin marriage and present, including the nucleotide binding position 1305 is mutated to stop codon TGA). The Glnl165T AG mutation (mu-

tation 2) was found in one Japanese pa- M SE KKLE TT AQQR K C PE WMNV Q NKR CAVE E RKACVRKSVFED D L P FLE FTG S IVY S YD AS 60 tient who is the offspring of a first cousin Fig. 3. Predicted protein DCS FLSED I SMSLSDGDVVGFDMEWP PL YNRGKLGKVAL I QLCVSES KCYLFHVSSMSVF 120 sequence of the WRN marriage and homozygous for the muta- gene product (22). Heli- tion. These two mutations were not ob- PQG LKMLLENKA VKKAGVG I EGDQWKLLRD FD I KLKNFVEL TDV ANKKL K CT E TW S LN S L 180 case domains I through VI I served in 48 Caucasian or 96 Japanese VKHLLGKQLLKDKS I RCSNWS KFPLTEDQKL YAATDAYAG FI I YRNLE I LDDTVQRF AIN 240 are in bold type, and their control individuals (19, 20). A third mu- KEEE ILLSDMNKQLT S I S EEVMDLAKHLPHAFS KLENPRRVS ILLKD I SENL YSLRRMI I 300 location was adapted tation, identified in a Syrian family, is a from previously described GS TN I ETE LRP S NNLNLLS FED S TTGGV QQ KQ I REHEVL I HVED ETWD PTLD HLAKHDG E 360 helicase domain align- DVLGNKVERKEDG FEDGVED NKLKENMERA CLM S LD I TEHELQ I LEQQS Q EEYL S D I A YK 420 ments (31). The amino ac- Table 1. WS mutations. Splice junctions are de- I. R:P,:~t1... .1.~:p:,:,t.2 - -.... ids corresponding to the noted by a double-headed arrow «-» and deleted S TEHLS PNDNENDT S YV I E S D E DLEMEML KHL S PNDNENDT S YVI E S D ED LEMEML KS LE 480 exon missing from RT- bases with a hyphen (-). Mutated or deleted bases PCR products from muta- NLN S GTVE PTH S K CL KME RNLGL PT KE E E EDD ENEANE GEED D D KD FL WP APNE EQ VTCL 540 are in bold. Intronic sequence is in small letters tion 4 individuals are over- and exonic sequence in capital letters. Mut., mu- I Ia 600 lined. The locations of tation; Ind., individual; Norm., normal; aa, amino KMYFG H S S F KPVQ WKV I HSVL EE RRDNVA VMA TGYGKS LCFQYP PVYVG K I GL V I S P LIS mutations 1 and 2 are acids. LMEDQVLQLKM S NI PACFLGSAQS ENVL TD I KLG KYRIVYVTPE YCSGNMGLLQQ LEAD I 660 II III noted with asterisks. The G I TLIA VDEAHCI SEWGHDFRDS FRKLGSLKTALPMVP IVALTATAS S SIRED IVRCLNL 720 exon presumed to be Pre- IV missing as a result of the dicted RN PQ I T CTG FD R PNL YLEVRRKTGNI LQD LQ PFL VKT S SHWEFE G PT I I YCP S RKMTQQV 780 mutation 3 splice junction pro- V Mut. Ind. Nucleotide sequence deletion is also overlined. tein TGELRKLNL SCG TYHAGMS F STRKD I HHRFVRD E I QCV IA T IAFGMG INKAD I RQVIHYG 840 VI Exon boundaries are from length AP KDMESYYQE I GRAGRDGLQS S CHVLWAP AD I NLNRHLL TE I RNE KF RL YKL KMMAKME 900 (aa) the genomic sequence of KYLHS SRCRRQ I ILSHFEDKQVQKASLG IMGTEKCCDNCRSRLDHCYSMDDS EDTSWDFG 960 P1 clone 2934. 1304 LeuGluArgAla PQAFKLLSA VDILGEKFG IGLPILFLRGSNSQRLADQYRRHSLFGTGKDQTESWWKAFSR 1020 Norm. TTGGAGCGAGCA t Missing exon, Mutation 4 WS TGA QLI TEG FL VEVSR YNKFMKI CAL TKKGRNWLHKANTESQSL I LQANEELCPKKFLL P S SK 108 0 2 AlaArgGlnLys 1164 Norm. GCTAGGCAGAAA * ~ Location of the premature stop codon for mutation 4 t TVSSGTKEHCYNQVPVELSTEKKSNLEKLYSYKPCDKISSGSNI SKKS IMVQSPEKAYSS 1140 WS TAG * ~ Mutation 2 (stop codon) 3 ThrAspLeuPhe 1392 SQPVI SAQEQETQIVL YGKLVEARQKHANKMDVPPAILATNKILVDMAKMRPTTVENVKR 1200 Norm. ctgtag<->ACAGACCTCTTT t Missing exon, Mutation 3 . IDGVSEGKAAMLAPLLEVIKHFCQTNSVQTDLFSSTKPQEEQKTSLVAKNKI CTLSQSMA 1260 WS ctgt--<->--AGACCTCTTT 4 GlyArgAsn 1060 Mutation 1 (stop codon) ~ * Norm. ttttaatag<->GGTAGAAAT ITYSLFQEKKMPLKS IAESRILPLMTIGMHLSQAVKAGCPLDLERAGLTPEVQKI IADVI 13 2 0

t RNPPVNSDMSKI SLIRMLVPENIDTYLIHMAIEILKHGPDSGLQPSCDVNKRRCFPGSEE 13 8 0 WS c I CS S SKRSKEEVG INTET S SAERKRRLPVWFAKGSDTS KKLMDKTKRGGLF S 1432

260 SCIENCE. VOL. 272 12 APRIL 1996

4 ~

motif in domain I and the DExH sequence tion of disrupted plasmids, decreased re- Sgslp helicase (Fig. 4), as part of a topo- (22) (where X is any amino acid) in do- pair rate of telomeres, rapid decrease in isomerase complex, functions to decat- main II. Across the seven helicase do- telomere length, and possibly altered enate intertwined chromosomes; muta- mains, the WRN predicted protein is 62 to DNA replication (24). In contrast, WS tions in the gene SOS 1 lead to hyper- 64% identical to the human RECQL and cells do not show an increased susceptibil- recombination between repeated sequenc- , Escherichia coli recQ gene products and ity to ultraviolet exposure or to other es with the deletion of intervening DNA, putative helicases from Caenorhabditis el- DNA damaging agents, do not appear to chromosome breakage, and nondisjunc- egans and yeast (Sgslp). In addition to the be defective in nucleotide excision repair, tion (28). seven-domain region, at the 3' end of the and do not show elevated sister chromatid The consequence of the WS defect in gene there are regions of limited similarity exchange (25). These characteristics set DNA metabolism may be the accumula- between the WRN gene product and the WS apart from other disorders (xeroderma tion of DNA mutations, leading to the C. elegans putative helicase F18C5.2 and pigmentosum, Cockayne's syndrome, and age-related diseases observed in WS. Dis- the E. coli RecQ protein (23). The WRN Bloom's syndrome) for which inherited ease susceptibility in WS could be the protein also contains a 98-amino acid defects have been identified in helicases result of mutations at specific genes. For acidic region that includes 13 aspartates or potentially involved in DNA repair or example, the increased incidence of neo- glutamates in a stretch of 17 amino acids chromosome exchange events (or both) plasia in WS may be caused by somatic (Fig. 3); the same region contains a 27- (26). However, a cryptic DNA repair de- mutations at oncogenes and tumor sup- amino acid repeat that is a perfect dupli- fect in WS cannot be ruled out, possibly pressor genes. Mutations in WRN may also cation at the nucleotide level. because the appropriate DNA-damaging playa role in tumorigenesis in non-WS As a putative helicase, the WRN pro- conditions have not been tested, or be- individuals because this gene is located in tein could be involved in DNA replica- cause other partially redundant repair sys- one of two chromosome 8p regions where tion, recombination, chromosome segre- tems may exist. For example, E. coli recQ loss-of-heterozygosity has been observed gation, DNA repair, transcription, or oth- mutants only show recombination defects in tumors (29). A more general mecha- er functions requiring DNA unwinding. and DNA damage sensitivity when other nism of disease susceptibility in WS may Indicators of defective DNA metabolism helicases (such as RecBC) are absent (27). be accumulated DNA damage that leads in WS include chromosomal instability, An alternate hypothesis to a DNA repair to premature replicative senescence and an elevated mutation rate at specific deficiency is that the WRN defect leads subsequent pathology. Whatever the spe- genes, elevated rates of nonhomologous directly to DNA damage and mutations. cific mechanisms involved in the WS phe- recombination, decreased accuracy of liga- For example, the Saccharomyces cerevisiae notype, identification of the WS gene now

I 1 la 100 WRN .DDENEANEGEEDDDKDFLWPAPNEEQVTCLKMYFGHSSFKPVQWKVIHSVLEERRDN KSLCFQYPpVYV... .GKIGLVISPLISL RECQ_ECOLIc .AQAEVLNLESGAKQV..LQETFGYQQFRPGQEEIrDTVLSG.RDC LcyQf:i>ALLL....NG SPLISL CELF18C5c .DRNVPQIDEATKMKWASMTSPPQEALNALNEFFGHKGFREKQWDVVRNVLGG.KDQ LPSLLL....NS SPLISL RECQ_HUMANc .DSDAGASNEYDSSPAAWNKED..FPWSGKVKDILQNVFKLEKFRPLQLETINVTMAG.KEV LPALC.. . .SDGF CP'LISL SGS1_YEASTc FDDDFSLSDIVSKSNLSSKTNGPTYPWSDEVLYRLHEVFKLPGFRPNQLEAVNATLQG.KDVp LPAVVKSGKTHGTTIVISPLISE Consensus L---F F---Q M-TG-GKS-C-Q-P V--PLISL ATPase 101 II 200 WRN ~YLQLKMSNIPACFLGSAQSE... .NVLTDI..KLGKYRIVYVTPBYCSGN...MGLLQQLEADIGITLIAVDBAHCXSEwGHD~RbSFRKLGSLKT RECQ_ECOLIc QLQANGVAAACLNSTQTREQQLEVMTGC..RTGQIRLLYIAPB. .RLM...LDNFLEHLAHWNPVLLAYDBAHCIS HDFRPEyAALGQLRQ CELF18C5c 'TLVSKGIDAVKLDGHSTQIEWDQVANN....MHRIRFIYMSPEMVTSQ...KGLELLTSCRKHISLLAIDEAH HDFRNSYRHLAEIRN RECQ_HUMANc VLKQLGISATMLNASSSKEHVKWVHAEMVNKNSELKLIYVTPEKIAKSKMFMSRLEKAYEARRFTRIAVDBVH HDFRPDYKALGILKR SGS1_YEASTc HLLNKNIKASHFSSRGTAEQRRQTFNLFI..NGLLDLVYISPEMISASEQCKRAISRLYADGKLARIVVpEAHC;V HDFRPDYKELKFFKR Consensus M-DQC--L A Y--PE DE-HC-S-WGHDFR L-----

201 III IV 300 WRN A. .LPMVPIV~~ATASS$IREDIVRCLNLRNPQITCTGFDRPNLYLEVRRKTGNILQDLQPFLVKTS.SHWEFEGPTIIYCPSRKMTQQVTGELRKLNL RECQ_ECOLIc R. .FPTLPF~TA~ADDTTRQDXVRLLGLNDPLIQISSFDRPNIRYMLMEK... .FKPLDQLMRYVQ.EQRGKSG..IIYCNSRAKyEDTAAALQSKGI CELF18C5c RSDLCNIPMIALTATA yIANLRLRKPLITTTSFDRKNLYISV.HSSKDMAEDLGLFMKTDEVKGRHFGGPTI IYCQTKQMVDDVNCVLRRIGV RECQ_HUMANc . . QFPNASLIGi, ;;QKILCIEKCFTFTASFNRPNLYYEVRQKPSNTEDFIEDI..VKLINGRYKGQSGrrYc;FSQKDSEQVTySLQNLGI SGS1_YEASTc . . EYPDIPMrALT IIHNLELKEPVFLKQsFNRTNLYYEVNKKTKNT.. . IFEI. .CDAVKSRFKNQTGIlycHSKK$C$QT$AQMQRNaI Consensus LTATA D L F-R-N IIYC------

301 V VI 400 WRN SCGTYHAGMSFSTRKDIHHRFVRDEIQCVIATIAFGMQINKAD,IRQVIHYGAPKDMBSYYQBIGRAGRDGLQSSCHVLWAPADINLNRHLLTBIRN.EK. RECQ_ECOLIc SAAAYHAGLENNVRADVQEKFQRDDLQIVVATyAFGMGINKPNVRFVVHFDIPRNIBSYYQBTGRAGRDGLPAEAMLFYDPADMAWLRRCLEBKPQ.GQ. CELF18C5c RSAHYHAGLTKNQREKAHTDFMRDKITTIVATyAFGMGIDKPDVRNVIHyGCPNIJIESyyQEIGRAGRDGSpSICRVFWAPKDLNTIKFKLRNSQQKEE. RECQ_HUMANc HAGAYHANLEPEDKTTVHRKWSANEIQVVVATY];)'GMGIDKPPYRFYIHHSMSKSMENyYQESGRAGB.DDMKAPCILYYGFGDIFRISSMVVME.NVGQQ SGS1_YEASTc KCAYYHAGMEPDERLSVQKAWQADEIQVICATVAFaMC;IPKPPVRFyyHFTVPRTLEGyyQETc;RAc;RDGl>TYSYC;ITYFSFRDIRTMQTMIQKDKNLDRE Consensus YHA AT-AFGMGI-K---R-V-H E-YYQE-GRAGRD D------

401 495 WRN .FRLYKLKMMAKMEKYLHSSRCRRQIILSHFEDKQVQKASLGIMGTEKCCDNCRSRLDHCYSMDDSEDTSWDFGPQAFKLLSAV .. RECQ_ECOLIc .LQDIERHKLNAMGAFAEAQTCRRLVLLNYFGE .GRQEPCGNCDICLDPPKQYDGSTD .AQIALSTIGRV ...... CELF18C5c .VVENLTMMLRQLELVLTTVGCRRYQLLKHFDPSYAKPPTM....QADCCDRCTEMLNGNQDSSSSIVDVTTESKWLFQVINEMYNGKTGIGKPI RECQ_HUMANc ...KLYEMVSYCQNISKCRRVLMAQHFDEVWNSEAC .NKMCDNCCKDSAF..ERKNITEYCRDLIKILKQA.. ... SGS1_YEASTc NKEKHLNKLQQVMAYCDNVTDCRRKLVLSYFNEDFDSKLC ..HKNCDNCRNSANVINEERDVTEPAKKIVKLVESI ...... Consensus CRR F C--C------Fig. 4. Protein alignment of genes showing homology to predicted protein genomic sequence between exons 6 and 7. Manual inspection of the WRN (22). Helicasedomains are marked as shaded regions (31). F18C5.2 alignment led to the identification of a missing exon (nucleotides 10240 (GenBank entry CElF18C5) is a predicted gene from a C. elegans cosmid through 10427), which was included in the F18C5.2 sequence shown. , that has been sequenced (32). When the predicted WRN protein was Sequences for human RECOl and E. coli RecO proteins and yeast Sgs1p aligned to F18C5.2, one region, highly conserved in other helicases, was are from GenBank. Numerous other helicases (not shown here) also missing.The missing region was identified by aligningthe WAN gene to the showed significant homology. SCIENCE. VOL.272 . 12APRIL1996 261 ~ '-'-'~""'" . ~.- "'~'-"-'-' ..~_..'.",_,~~----- II provides evidence that at least some com- 5EA to synthesize first-strand cDNA by reverse cy (0.0027), giving a 95% upper confidence limit of transcriptase (SuperScript II RT, Life Technolo- 0.008. ponents of "normal" aging and disease gies). RNA was removed by ribonuclease H, and 21. D. Cerimele et al., Hum. Genet. 62,25 (1982). susceptibility in late life may be related to single-strand DNA was purified with a GlassMAX 22. Abbreviations for the amino acid residues are as aberrations in DNA metabolism. spin cartridge (LifeTechnologies). A deoxycytidine follows:A,Ala; C, Cys; D,Asp; E, Glu; F, Phe;G,Gly; tail was added to the 3' end of the DNA by using H, His; I, lie; K, Lys; L, Leu; M, Met; N, Asn; P, Pro; Q, terminal deoxynucleotide transferase. After heat in- Gin; R, Arg; S, Ser; T, Thr; V, Val; W, Trp; and Y, Tyr. REFERENCES AND NOTES activation of the terminal transferase (70°C, 10 23. J, Mulligan, unpublished data. min), the resulting DNA (5 ILl)was amplified with a 24. H. Hoehn et al., Cytogenet. Cell Genet. 15, 282 nested primer 5EM (5'-TCGATCAAAACCAGTACA- 1. M. Goto, K.Tanimoto, Y. Horiuchi, T. Sasazuki, Clin. {1975}; D. Salk, K. Au, H. Hoehn, G. M. Martin, ibid. GGTG-3') and the 5'-RACE anchor primer (5'-CUA- 30,92 (1981); S. Scappaticci, D. Cerimele, M. Frac- Genet. 19,8 (198'1\ M. Goto, Y. Horiuchi, K. Tani- CUACUACUAGGCCACGCGTCGACTAGTACG- moto, T. ishii,H. Nakashima, J. Am. Geriatr.Soc. 26, caro, Hum. Genet. 62, 16 (1982); K. Fukuchi et al., GGIIGGGIIGGGIIG-3',Life Technologies). A 2.5-kb 341 (1978); S. J. Thannhauser, Ann. Intern. Med. 23, Somatic Cell Mol. Genet. 11, 303 (1985); K. Fukuchi, 559 (1945). product was obtained that contained an additional G. M. Martin, R. J. Monnat, Proc, Nat!. Acad. Sct. 1.4 kb of unique 5' sequence. 2. C. J. Epstein, G. M. Martin, A. L. Schultz, A. G. U.S.A. 86,5897 (1989); K. Fukuchi et al., Hum. Gen- Motu/sky, Medicine 45,177 (1966). 18. The WS patients were from an Intemationai Registry et. 84, 252 (1990); T. M. Runger et al., J. Invest. 3. G. M. Martin, Birth Defects 14, 5 (1978). of Werner's Syndrome (G.M.M.).The diagnostic cri- Dennatol. 102, 45 (1992); V. Schulz et al., Hum. 4. -, C. A.Sprague, C. J. Epstein, Lab. Invest. 23, teria and the ethnic origins of the individuals are as Genet., in press; R. Z. Cheng, S. Murano, B. Kurz, R. 86 (1970); T. O. Tollefsbol and H. J. Cohen, Age 7, previously described (7, 12). Contributors to the reg- J. Shmookler Reis, Mutat. Res. 237, 259 (1990); P. 75 (1984). istry are S. S, Agarwal, F. Amato, p, Amblard, J. A. Kruk, N. J. Rampino, V. A. Bohr, Proe. Natl. Acad. 5. M. Goto, J. Weber, K. Woods, D. Drayna, Nature Anders, R. Anschuetz, M. Y. Apak, S. Balci, M. Bi- Sci. u.S.A. 92,258 (1995); F. Takeuchi et al., Hum. ancalana, T. D. Bird, J. Bonar, A. Braganza, I. 355,735 (1992). Genet. 60, 365 (1982). 6. G. D. Schellenberg et a/., Lancet 339, 1002 (1992); Boumerias, J. Chevrant Breton, W. T. Brown, G. 25. Y. Fujiwara, T. Higashikawa, M. Tatsumi, J. Cell. W. Winston, M. Rubenstein, M. Goto, D. Drayna, Burg, D. P. Cavalcanti, C. Danesino, V. Dumas, J. Physiol. 92, 365 (1977); M. Stefanini et al., Mutat. Genomics 16, 685 (1993); J. Nakuraet a/., Gerontol- Ellis,C. J. Epstein, W. Fischer, M. Fraccaro, K. Fuku- Res. 219, 179 (1989); E. Gebhartetal., Hum. Genet. ogy 39 (suppl.), 11 (1993). chi, Y. Fujiwara, P. Gamier, S. Gilgenkranz, E. Ha- 80,135 (1988). chulla, T. J. Harrison, H. Hoehn, Y. Hosokawa, A. F. 26. C. A. Weber, E. P. Salazar, S. A. Stewart, L. H. 7. J. Nakura et a/., Genomics 23, 600 (1994). Hurlimann, J. Jabkowski, K. Kamino, H. Kayserili,S. 8. J. Oshimaeta/., ibid., p. 100. Kiso, G. Klein, J. Lamit, D. Lewis, H. Little, M. C. Thompson, EMBOJ. 9, 1437 (1990); W. L. Frejteret 9. L.Yeeta/.,ibid.28,566(1995);K.Kiharaetal.,Jpn. Martin, P. Martinet, K. Manumo, M. I. Melagrano, W. al., Proc. Natl. Acad. Sei. u.S.A. 89,261 (1992); P. Sung et al., Nature 365, 852 (1993); G. Weeda et al., J. Hum. Genet. 39,403 (1994). Mills,A. Mohan, A. Motulsky, M. Mumenthaler, S. 10. C. E. Yu et a/., Am. J. Hum. Genet. 55,356 (1994). Murano, N. Marakami, J. Matthews, p, Modiano, Cell 62,777 (1990); C. Troelstra et al., ibid. 71,939 (1992); N. A. Ellis et al., ibid. 83, 655 (1995). 11. C. E. Yu et al., in preparation. Tsukasa Murakami, O. Nikaido, G. Natchiar, T. Ogi- 12. P values for linkage disequilibrium were obtained by hara, S. Quais, A: Partalci, F. Pasquali, N. Philip, M. 27. V. R. Mendonca, H. D. Klepin, S. W. Matson, J. using a Fisher exact test on a k x 2 contingency table Poot, C. Puissan, J. Revuz, M.W. Rizzo, C. Rubin, T. Bacterial. 177, 1326 (1993). where k is the number of alleles. K. A. B. Goddard et Saida, K. R. Sathish, S. Scappaticci, J. Schmidtke, 28. S. Gangloffet a/., Mol. Cell. BioI. 14,8391 (1994); P. al., Am. J. Hum. Genet., in press. K. Singh, R. Singh, M. W. Steele, V. P. Sybert, C. M. Watt, E. J. Louis, R. H. Barts, I. D. Hickson, Cell 13. A. J. Buckler et al., Proc. Nat!.Acad. Sci. u.S.A. 88, Tannock, J. Taziri, B. Uyeno, A. Verloes, A. 81, 253 (1995); J. C. Wang, J. Bioi. Chem. 266, 4005 (1991); D. M.Church et al., Nature Genet. 6, 98 Wakayama,and M. Yuksel. 6659 (1991). (1994); S. Parimoo, S. R. Patanjali, H. Shukla, D. 19. Control individuals were screened for mutations 1,2, 29. Y. Fujiwara et al., Cancer Res. 53, 1172 (1993); M. Chaplin, S. M. Weissman, Proc. Nat!. Acad. Sci. and 4 (Table 1) by DNA sequence analysis of PCR Chang et al., Am. J. Pathol. 144, 1 (1994); M. J. U.S.A. 88,9623 (1991); S. F. Altschul, W. Gish, W. products amplified from genomic DNA.For mutation Pykett et al., Cancer Genet. Cytogenet. 76, 23 Miller,E. W. Myers, D. J. Lipman, J. Mol. Bioi. 215, 1, primers E8A (5'-GATGTGACAGTGGAAGCTAT- (1994); J. Kagan et al., Oncogene 11, 2121 (1995); 403 (1990); E. C. Uberbacherand R.J. Mural,Proc. GG-3') and E8B (5'-GGAAAAATGTGGTATCTGAA- F. Kerangueven et al., ibid. 10, 1023 (1995). Nat!. Acad. Sci. u.S.A. 88, 11261 (1991). GCTC-3') were used to produce a 267-bp product. 30. P1 DNA for PCR analysis was prepared by the 14. The Eukaryotic Promotor Database was accessed Both strands were sequenced by using the same standard alkaline lysis method followed by phenol- through the National Center for Biotechnology Infor- primers. The same procedure was used for mutation chloroform extraction and ethanol precipitation [J. mation Basic Local Alignment Search Tool (BLAST) 2 [primers E11A (5'-TAAAGGATTAATGCTGTTAA- Sam brook, E. F. Fritsch, T. Maniatis, Molecular Network service at ftp://[email protected] CAGTG-3') and E11B (5'-TCACACTGAGCATTTA- Cloning, a Laboratory Manual (Cold Spring Harbor 15. D. J. Munroe et al., Proc. Nat!. Acad. Sci. u.S.A. 92, CTACCTG-3')] and mutation 4 [primers E4A (5'-CT- Laboratory, Cold Spring Harbor, NY, ed. 2, 1989), 2209 (1995). , TGTGAGAGGCCTATAAACTGG-3') and E4B (5'- pp. 1.25-1.28] and then by CsCI density gradient 16. RT-PCR products synthesized from RNA (Qiagen GGTAAACAGTGTAGGAGTCTGC-3')]. which pro- centrifugation in the presence of ethidium bromide. Oligotex, Qiagen, Chatsworth, CA) prepared from duced 360- and 267-bp fragments, respectively. For sequencing, P1 clones were randomly frag- affected WS and control individuals were amplified Some control individuals were also tested for mu- mented and subcloned into an M13 phage-de- with a variety of primers and the PCR products cy- tations by a PCR mismatch amplification assay. rived vector. Single-stranded template DNA from cle-sequenced with the dye terminator cycle se- Exons were peR-amplified by exon primers in 20- recombinant M1'3 clones was prepared by a stan- quencing kit (Perkin-Elmer). In other experiments, ILlreaction. The entire mixture was then diluted into dard mini prep procedure, DNA quality and yields RT-PCR products were gel purified, reamplified with 200 ILlof distilled water. From the dilution mixture 2 were determined by agarose gel electrophoresis the same primers, and sequenced with a U.S. Bio- ILl were used for mismatch PCR. For each point and by measuring absorbance at 260 nm. chemical Sequenase PCR Product sequencing kit. mutation, two mismatched primers were designed Dideoxynucleotide sequencing was performed 17. WRN clones were from a normal fibroblast cell line in which the 3' end of the primer corresponds to with Taq DNA polymerase and ftuoreseently la- cDNA library that was primed with polyadenylate either wild-type (WT) or mutant (MT) nucleotide. beled dye primers. DNA sequences were deter- and cloned in Lambda ZAP pBK-CMV (Strat- Mismatch PCR is prepared in two separate tubes mined with a fluorescent DNA sequencer (ABI agene). The library was arrayed for PCR screening with the combination of one mismatched primer 373A). (15) and could be screened either by PCR amplifi- and one exon primer. PCR was then carried out 31. A. E. Gorbalenya, E. V. Koonin, A. P. Donchenko, V. cation or by conventional plaque hybridization. The under high-stringency conditions for 30 cycles. The M. Blinov, Nucleic Acids Res. 17,4713 (1989). initial 2.1-kb cDNA clone containing the EST mismatched primer sequencesand PCR conditions 32. R. Wilson et al., Nature 368, 32 (1994). R58879 and corresponding to the 3' end of the were asfollows: Exon4WT primerE4D(5'-CTTTAT- 33. Supported by National Institute on Aging grants R01 gene was obtained by PCR screening with primers GAAGCCAATTTCTACCCT) together with primer AG12019 (GoO.S.), P01 AG01751 (G,M.M.), T32 A (5'-ACTGGCAAGGATCAAACAGAGAG-3')and B E4B amplifya 106-bp fragment under conditions of AGOO057 (C.E.Y.), R37 AG08303 (G.M.M.), and K11 (5'-CTTTATGAAGCCAATTTCTACCC-3'), which annealingtemperature(TA)= 63°C, 1.0 mM Mg and AGO0671-01 (F.M.H.); a Yale-L. P. Markey Trust were designed from the DNAsequence of R58879 pH 9.5. Exon 4 MT primer E4C (5'-TAAAAGATC- Physician-Scientist Training Fellowship (F.M.H.); and and produce a 145-bp fragment from WRN cDNA CTTTTTGCTTTTAATAg together with primer E4A a grant from Darwin Molecular (GoO.S.). We thank E. clones. To obtain longer clones, we designed primers amplifya 212-bp fragment underconditions of TA= Loomis, L. Anderson, M. Sullivan, M. Tsuru, S. Fre- 5EA (5'-GAACTTTGAAGTCCATCACGACC-3')and 63°C, 1,5 mM Mg, and pH 10.0. dell, A Smith, C. Ogborn, A. Jarzebowicz, E. Nem. 5EB (5'-GCATTAATAAAGCTGACATTCGCC-3') 20. An upper confidence bound on the allele frequen- ens, and A. Sarthy for technical assistance; the Dar- from a GRAIL-predicted exon located 5' in the P1 cy, P, when no mutations are found, can be calcu- win Molecular DNA sequencing group; R. Monnat, clone 2934 genomic sequence to exons containing lated by means of the equation (1 - PiN=

262 SCIENCE. VOL. 272 12 APRIL 1996