Oncogene (2000) 19, 4764 ± 4772 ã 2000 Macmillan Publishers Ltd All rights reserved 0950 ± 9232/00 $15.00 www.nature.com/onc Di€erential regulation of human RecQ family helicases in cell transformation and cell cycle

Tamae Kawabe1, Naohiro Tsuyama2, Saori Kitao1, Kaori Nishikawa1, Akira Shimamoto1, Miwa Shiratori1, Takehisa Matsumoto1, Kumiko Anno2, Tatsuhiro Sato2, Youji Mitsui3, Masayuki Seki4, Takemi Enomoto4, Makoto Goto5, Nathan A Ellis6, Toshinori Ide2, Yasuhiro Furuichi*,1 and Masanobu Sugimoto1

1AGENE Research Institute, 200 Kajiwara, Kamakura, Kanagawa 247-0063, Japan; 2Department of Cellular and Molecular Biology, Hiroshima University School of Medicine, Minami-ku, Hiroshima 734-8551, Japan; 3National Institute of Bioscience & Human-Technology Higashi 1-1, Tsukuba, Ibaraki 305-8566, Japan; 4Molecular Cell Biology Laboratory, Graduate School of Pharmaceutical Sciences, Tohoku University, Sendai, Miyagi 980-8578, Japan; 5Department of Rheumatology, Tokyo Metropolitan Otsuka Hospital, Minami-Otsuka, Toshima-ku, Tokyo 170-0005 Japan; 6Laboratory of Cancer Susceptibility, Department of Human Genetics, Memorial Sloan-Kettering Cancer Center, New York, NY 10021, USA

Three human RecQ DNA helicases, WRN, BLM and helicase RecQ is the progenitor of this family and RTS, are involved in the genetic disorders associated presumably participates in double-strand break repair with genomic instability and a high incidence of cancer. and acts as a suppresser of illegitimate recombination RecQL1 and RecQL5 also belong to the human RecQ as well as in replication restart (Courcelle and helicase family, but their correlation with genetic Hanawalt, 1999; Nakayama et al., 1984; Umezu and disorders, if any, is unknown. We report here that in Nakayama, 1993; Umezu et al., 1990). Yeast S. human B cells transformed by Epstein-Barr virus (EBV), cerevisiae and S. pombe have SGS1 and rqh1+heli- human ®broblasts and umbilical endothelial cells trans- cases, respectively, that also belong to this family, and formed by simian virus 40, the expression of WRN, the lack of these helicases causes various disease BLM, RTS and RecQL1 was sharply up-regulated. In B phenotypes, such as slow growth, premature aging cells this expression was stimulated within 5 ± 40 h by the and aberrant assembly of chromosomes resulting tumor promoting agent phorbol myristic acetate (PMA). primarily from the hyper- and illegitimate recombina- Interestingly, RecQL5b, an alternative splicing product tion events (Sinclair et al., 1997; Sinclair and Guarente, of RecQL5 with a nuclear localization signal, is 1997; Stewart et al., 1997; Yamagata et al., 1998). In expressed in resting B cells without signi®cant modula- contrast to single-cell organisms that have only one tion of its synthesis by EBV or PMA, suggesting it has a species of RecQ helicase, higher eukaryotes contain role in resting cells. We also roughly determined the multiple RecQ helicases. In humans the ®ve members number of copies per cell for the ®ve RecQ helicase in B described to date are RecQL1 (Seki et al., 1994; cells. In addition, levels of the di€erent RecQ helicases Puranam and Blackshear, 1994), BLM (Ellis et al., are modulated in di€erent ways during the cell cycle of 1995), WRN (Yu et al., 1996), RecQL4 and RecQL5 actively proliferating ®broblasts and umbilical endothe- helicases (Kitao et al., 1998, 1999b). Recent studies lial cells. Our results support the view that the levels of showed that mutated BLM, WRN and RecQL4 WRN, BLM, RTS and RecQL1 are di€erentially up- helicases cause Bloom syndrome, Werner syndrome regulated to guarantee genomic stability in cells that are and a subset of Rothmund-Thomson syndrome, transformed or actively proliferating. Oncogene (2000) respectively, all of which have been known as recessive 19, 4764 ± 4772. genetic disorders that have in common genomic instability, and they present complex clinical pheno- Keywords: RecQ helicase; cell transformation; immu- types but all with a high risk of cancer. In this paper, noblotting; genomic stability we use RTS (Rothmund-Thomson syndrome) in referring to the RecQL4 helicase even though it remains to be clari®ed whether the mutation in Introduction RecQL4 helicase is responsible for the etiology of all patients with Rothmund-Thomson syndrome. Associa- DNA helicases have important roles in cellular tion of mutations in the RecQL1 and RecQL5 helicases processes, including DNA replication, recombination, with human disorders has so far not been shown. We repair, and transcription, by unwinding the duplex recently identi®ed a new RecQL5 species formed by genome strands (Lohman, 1993). Among many kinds alternative splicing of RecQL5 gene, called RecQL5b. of DNA helicases in living cells, the RecQ helicase Because RecQL5b, but not RecQL5, localized to the family has unique properties apparently involved in nucleus (Shimamoto et al., 2000), we have focused on maintaining genomic stability, although the exact RecQL5b in this study. biological pathways remain unknown. E. coli DNA Cumulated biochemical studies demonstrate that all RecQ helicases contain a helicase domain highly homologous to that of E. coli RecQ helicase, but that the N- and C- terminal regions di€er from each other *Correspondence: Y Furuichi Received 11 May 2000; revised 13 June 2000; accepted 3 August (Kitao et al., 1998). Cytochemical analyses indicate 2000 that the human RecQ helicases localize in the nucleus Expression of RecQ helicases in transformed cells T Kawabe et al 4765 (Kitao et al., 1999a; Marciniak et al., 1998; Ne€ et al., noblot analysis to measure the protein levels of 1999; Shiratori et al., 1999). In growing human individual RecQ helicases. The expression of BLM ®broblast and K562 tumor cell lines, mRNAs coding and RTS was highly dependent on cell cycle. On the for the ®ve RecQ helicases are expressed throughout other hand, RecQL5b was expressed at a relatively the cell cycle (Kitao et al., 1998), consistent with the high level in resting B cells, and its expression was presence of the Sp1-mediated house-keeping promoters not changed during the cell cycle in normal diploid characterized for WRN and RTS (Kitao et al., 1999a; cell lines or by cell transformation. These results Yamabe et al., 1998). Extensive studies on WRN suggest that WRN, BLM, RTS and RecQL1 are up- helicase showed that the N- terminal domain has regulated di€erentially by virus-mediated cell trans- 5'743' or 3'745' exonuclease activity (Huang et al., formations and at certain stages of the cell cycle 1997; Suzuki et al., 1999), and also DNA and RNA/ whereas RecQL5b plays a di€erent role such as DNA unwinding activity (Suzuki et al., 1997), transcription in resting cells. suggesting its involvement not only in replication (Hanaoka et al., 1985) but also in transcription. WRN binds to the nuclear proteins RPA, p53, PCNA, Results topoisomerase I and Ku proteins, and the lack of WRN in patients' cells induces an attenuation of Helicase up-regulation during cell transformation by EBV apoptosis (Brosh et al., 1999; Lebel et al., 1999; and stimulation by PMA in B cells Spillare et al., 1999; Cooper et al., 2000). Functional interaction between the WRN helicase and DNA EBV infects resting B cells in G0 and transforms them polymerase delta is also reported (Kamath-Loeb et into B-lymphoblastoid cell lines (LCLs) (Miller, 1990). al., 2000). BLM helicase is known to interact with Figure 1a shows that the expression of the four RecQ RPA and topoisomerase IIIalpha (Brosh et al., 2000; helicases WRN, BLM, RTS and RecQL1 was up- Wu et al., 2000). Despite this biochemical knowledge, regulated in LCLs after transformation by EBV: the biological role of each human helicase is still resting B cells (N0008R) expressed no WRN, BLM unclear, but the clinical and genetic data suggest that and RTS and a barely detectable level of RecQL1, but each helicase is involved in di€erent tasks. the EBV-transformed LCL N0008T expressed signi®- Cytogenetic studies indicate that the cells of patients cant levels of these four RecQ helicases. On the other with Werner, Bloom and Rothmund-Thomson syn- hand, resting B cells expressed a signi®cant level of dromes undergo di€erent types of genomic instability: RecQL5b, which was not increased after transforma- pseudodiploidy involving variable structural rearrange- tion by EBV. Two EBV-transformed LCLs from WS ments of chromosomes and abnormal dynamics of patients WS11001 and WS11301 expressed no WRN, telomeres in Werner syndrome (Salk et al., 1985; but expressed normal levels of other RecQ helicases Schulz et al., 1996; Tahara et al., 1997), an increase BLM, RTS, RecQL1 and RecQL5b, indicating that the in sister chromatid exchange in Bloom syndrome lack of normal WRN does not a€ect the expression of (German, 1993) and trisomy 8 mosaicism in Roth- other RecQ helicases. The e€ect of PMA, a B cell mund-Thomson syndrome accompanied by mutations mitogen (Bertoglio, 1983; Clevers et al., 1985), on the at RTS gene (Lindor et al., 1996). Thus, the RecQ expression of the ®ve RecQ helicases was also studied helicases relevant to these disorders are predicted to in resting B cells N0008R (Figure 1b). PMA was maintain genomic integrity by their divergent functions chosen because it stimulates the growth of human B that operate well in cells from healthy individuals but cells. Adding PMA at a concentration of 30 ng/ml to not in cells from patients lacking these RecQ helicases. the medium caused a marked increase in expression Indeed, in a surrogate experiment using yeast cells levels of WRN, BLM, RTS and RecQL1 although the lacking SGS1 helicase, Yamagata et al. (1998) showed level of RecQL5b is marginally decreased by the that BLM and WRN genes can suppress hyper- treatment. The levels of both RecQL1 and WRN recombination phenotypes of mutant yeast, implying markedly increased and reached near plateau levels that their role in human cells also is to suppress hyper- within 5 h after PMA treatment, while the BLM and recombination. Also, Ellis et al. (1999), Ne€ et al., RTS proteins remained at the same level until 16 h and (1999) showed that the intact BLM gene transfected increased thereafter up to 40 h after treatment (Figure into the cells of a Bloom syndrome cell line reduced the 1b). By PMA treatment, morphological changes of the increased sister chromatid exchange rate toward B cells with notable clump formation were observed normal. ®rst at 5 h, and the size of the clumps increased However, knowledge of the mechanisms regulating thereafter (Figure 1c). Resting B cells puri®ed from the expression of RecQ helicases is very limited PBLs are in G0 of the cell cycle, and they are reported (Yamabe et al., 1998). Related to this in an earlier by Spender et al. (1999) to begin to express an early study, we showed that the expression of WRN in marker of cell cycle entry, cyclin D2, within 24 h after human B cells of peripheral blood leukocytes (PBLs) treatment with PMA or EBV. Given these facts, our and ®broblasts is increased by transformation with results indicate (1) a sharp increase in WRN and Epstein-Barr virus (EBV) and simian virus 40 (SV40), RecQL1 occurs by PMA stimulation accompanied by respectively (Shiratori et al., 1999). In the present clump formation of the B cells, which precedes PMA- study, we have compared the expression kinetics of induced cell proliferation (Abb et al., 1979; Clevers et the RecQ helicases WRN, BLM, RTS, RecQL1 and al., 1985) and (2) up-regulation of BLM and RTS RecQL5b associated with the virus- or viral protein- occurs when B cells initiate entry into the cell division mediated cell transformations as well as with cycle. This apparent correlation of the increase in the stimulation by the tumor-promoting compound levels of BLM and RTS with entry into S phase of B phorbol-12-myristate-13 acetate (PMA), using immu- cells agrees with the fact that TIG-3 and tumor cells in

Oncogene Expression of RecQ helicases in transformed cells T Kawabe et al 4766

Figure 1 Augmented expression of RecQ helicases in B cells transformed by EBV or stimulated by PMA. (a) Comparison of the expression of the ®ve RecQ helicases among resting B cells (N0008R) and EBV-transformed LCL from a normal individual (N0008T) from whom the resting B cells were obtained, immortalized EBV-transformed LCL (N6803IM) as well as from two Werner syndrome patients (WS11001 and WS11301). (b) Cell lysates (20 mg protein for WRN, RTS, RecQL1 and RecQL5b samples and 60 mg protein for BLM samples) of resting B cells cultured with 30 ng/ml PMA for 0, 5, 16 and 40 h were separated by SDS ± PAGE, and were immunobloted. The bands of 180, 180, 160/140, 73 and 135 kDa indicate WRN, BLM, RTS, RecQL1 and RecQL5b, respectively. Actin was used to con®rm that same amounts of samples were loaded. (c) Phase contrast micrographs of the resting B cells cultured with 30 ng/ml for 0, 5, 16 and 40 h as shown in (a)

log phase expressed higher levels of BLM and RTS Table 1 Copy number of WRN helicase in resting B cells and EBV- than cells in a stationary phase as will be shown later. transformed LCL cells Table 1 shows the number of copies of the ®ve RecQ Cells WRN BLM RTS RecQL1 RecQL5b helicase proteins per cell of resting B cells (N0008R), Resting B cells 1200 330 120 2100 340 EBV-transformed LCL cells (N0008T) and immorta- (N0008R) lized EBV-transformed LCL cells (N6803IM). In both Pre-immortal LCL 94 000 27 000 1100 15 000 330 N0008T and N6803IM, the copy number was largest in cells (N0008T) WRN, followed by BLM, RecQ1, RTS and RecQL5b. Post-immortal LCL 200 000 33 000 1000 21 000 330 The number of WRN copies of a LCL (N0008T) cells (N6803IM) determined in this study (94 500 per cell) is near the The copy number of each RecQ heicase was determined by number of an LCL (about 60 000 per cell) reported by quantitative immunoblotting versus standard recombinant RecQ Moser et al. (2000). Resting B cells (N0008R) are helicases. Each ®gure represents a mean of duplicated samples characterized by virtual absence of WRN, BLM and RTS with a level of RecQL5b comparable to that of N0008T and N6803IM. The immortalized LCL N6803IM have an abnormal karyotype with trisomy of chromosome 8 where WRN gene locates (Kataoka Up-regulation during cell transformation by SVtsTin et al., 1997; Yu et al., 1996). Interestingly, N6803IM fibroblasts and endothelial cells cells have WRN copies about twice those of pre- We investigated the e€ect of cell transformation by immortal LCL N0008T. RecQL5b is considered to play SV40tsT on the expression of WRN, BLM, RTS, a role quite di€erent from other RecQ helicases RecQL1 and RecQL5b in human diploid ®broblast because the copy number of RecQL5b is much lower TIG-3 cells (Figure 2a). Proliferating normal diploid than that of other RecQ helicases and did not increase TIG-3 cells expressed a high level of RecQL1 but lower at all after transformation by EBV. levels of WRN, BLM, RTS and RecQL5b. By contrast,

Oncogene Expression of RecQ helicases in transformed cells T Kawabe et al 4767 the two transformed cell lines SVts9-3 and SVts9-4, as line SVts8; for the non-ts SVOD1-2 cell line, little well as the immortalized cell line SVts8, expressed di€erence was observed between permissive and non- markedly higher levels of WRN, BLM, RTS and permissive temperatures. In addition, in three immor- RecQL5b than the untransformed TIG-3 cells, the talized endothelial cell lines originated from HUE101-1 immortalized SVts8 cell line being the highest expresser that were transformed with SV40tsT and hTERT, the of RTS. No marked di€erence, however, was observed expression of WRN alone was again down-regulated in in the levels of RecQL1 helicase between pre- and post- a temperature-sensitive manner (Figure 3b). These transformed TIG-3 cells. Figure 2b shows the expres- results together indicate that the e€ect of SV40tsTin sion of WRN, BLM, RTS, RecQL1 and RecQL5b in increasing the expression of WRN is apparently various endothelial cell lines originated from HUE101- nulli®ed at a non-permissive temperature in both 1: original HUE101-1; hT-1 and hT-2 transformed with TIG-3 ®broblast and endothelial cell lines transformed hTERT gene, KA62, 31 and 31-2 transformed with by SV40tsT. These data strongly suggest that SV40tsT SV40tsT; 62-5 transformed with hTERT gene and is intimately involved in up-regulating the expression of SV40tsT. KA62, 62-5 and 31-2 are immortalized cell WRN helicase. In this regard, we earlier reported that lines. Transformation with hTERT or SV40tsT mod- p53 suppressed the promoter activity of the WRN gene erately increased the expression of WRN, BLM and (Yamabe et al., 1998). From this observation, together RTS but not of RecQL1. The expression of RTS was with the fact that SV40tsT binds to p53 at 348C but highly augmented in all the cell lines transformed with not at 38.58C resulting in release of free p53 protein SV40tsT. The expression of RecQL5b was not much (Tahara et al., 1995), suppression of WRN expression a€ected by transformation with hTERT gene nor at a non-permissive temperature may be mediated by SV40tsT. p53 binding to the WRN promoter. The fact that The SV40tsT gene used here originated from a WRN helicase binds to p53 (Spillare et al., 1999) raises mutant of SV40, tsA900, whose T-antigen was sensitive an alternative possibility that p53 changes the levels of to temperature for many functions, such as virus WRN by direct interaction. The possibility also replication, cell transformation, transcriptional repres- remains that SV40tsT directly binds temperature- sion of the SV40 early gene, and p53 binding (Kuchino sensitively to WRN, stabilizing WRN helicase. No and Yamaguchi, 1975; Tahara et al., 1995; Yamaguchi consistent e€ect of temperature, however, was observed and Kuchino, 1975). We therefore investigated if the for the expression of BLM, RTS, RecQL1 and modulation of the expression of the ®ve RecQ helicases RecQL5b. Although the expression of RTS in SVts9- by SV40tsT is sensitive to temperature. Figure 3a 3 at 84 PDLs was apparently lower in permissive than shows that expression of WRN at the non-permissive in non-permissive temperature, this phenomenon was temperature 38.58C was markedly lower than at the not observed in SVts9-3 at 51 PDLs nor SVts8, and is permissive temperature 348C for the SVts9-3 cell line considered to be not a consistent phenomenon (Figure 51 and 84 PDLs, as well as for the immortalized cell 3a). The fact that the expression levels of RecQL1,

Figure 2 Expression of WRN, BLM, RTS and RecQL1 in human TIG-3 cell lines and endothelial cell lines. (a) TIG-3, human embryonic ®broblasts; SVts9-3 and SVts9-4, TIG-3 cell lines transformed by SV40tsT; SVts8, SV40tsT antigen-transformed TIG-3 cell line that was immortalized. Cell lysates of 60 mg protein was loaded. (b) HUE101, original human endothelial cell line; hT-1 and hT-2, transformed with hTERT; KA62, transformed with SV40tsT; 62-5, 31, 31-2, transformed with hTERT and SV40tsT

Oncogene Expression of RecQ helicases in transformed cells T Kawabe et al 4768

Figure 3 The e€ect of temperature on the expression of RecQ helicases of TIG-3 cells and endothelial cells transformed by SV40tsT. The cells were cultured for 3 days at a permissive (P) (348C) or non-permissive (N) (38.58C) temperature. (a) Fibroblast cell lines. SVOD1-2, SVts9-3 and SVts8 were transformed by SV40T and SV40tsT, respectively; SVts8 was immortalized. (b) Endothelial cell lines established by transformation of HUE101-1 with hTERT and SV40tsT

BLM, RTS and RecQL5b were largely not sensitive to Discussion temperature suggests that their mechanisms of regula- tion are di€erent from WRN helicase. The results obtained in this protein-kinetic study contrasted the following di€erential modes of expres- sion of the ®ve human RecQ helicases. (1) RecQL1 was Changes during the cell cycle expressed at a maximal level in all the cells examined, We compared the expression levels of the ®ve RecQ except for resting B cells in G0, which agrees with our helicases in TIG-3 cell lines between log and stationary previous observation that RecQL1 mRNA was ex- phases. Figure 4a shows that in the TIG-3 cell line the pressed at similar levels in almost all tissues examined levels of WRN, BLM, RTS and RecQL1 decreased as (Kitao et al., 1998). These facts also suggest that the cells shifted from log (days 1 and 2) to stationary RecQL1 is present at all stages of the cell cycle except phases (days 3 and 4) whereas the level of RecQL5b G0 and that RecQL1 is involved in a housekeeping was una€ected; the decrease was much more pro- function. (2) The expression of WRN increased nounced for BLM and RTS than for WRN and together with cell transformation by EBV and RecQL1. The levels of WRN, BLM, RTS, RecQL1 SV40tsT, suggesting that WRN is involved in some and RecQL5b were also compared in the four tumor special function required to transform cells, such as to cell lines U251MG, T98G, A2182 and KMST-6 stabilize the genome in rapidly proliferating cells as (Figure 4c). In all the cell lines, the levels of BLM, discussed later. The increase in WRN at early stages RTS and RecQL5b apparently decreased from day 2 to after PMA stimulation agrees with the idea that WRN day 4, whereas the change in the levels of WRN and may be involved in a function in G1 of dividing cells, RecQL1 was less pronounced. Analyses by ¯ow such as in transcription mediated by RNA polymerase cytometry of TIG-3 and U251MG cell lines indicated I or II (Balajee et al., 1999). We and others reported that the proportions of S phase cells were far greater that WRN exists in the nucleoplasm, as well as in the when the cells were in log phase on day 2 than the cells nucleolus of rapidly proliferating cells (Marciniak et were in the stationary phase on day 4; the proportions al., 1998; Shiratori et al., 1999; Kiao et al., 1999a). of cells at S plus G2/M on day 2 were 33.1% for TIG-3 Also, WRN can unwind RNA/DNA heteroduplexes and 47.6% for U251MG, whereas on day 4 they were that would be formed during the transcription (Suzuki 6.6% for TIG-3 and 11.8% for U251MG (Figure 4b, et al., 1997). (3) Increased expression of BLM and RTS d). In these experiments the medium was not changed at later stages of PMA stimulation, as well as at the log during the four-day culture, but similar experiments phase of cell culture, is likely to be intimately were done by changing the medium every day for four associated with DNA replication. This hypothesis days (data not shown). These two series of experiments agrees with the observation in this study that the gave essentially the same results for immunoblot and expression of BLM and RTS was not sensitive to the ¯ow cytometry analyses. Thus, the decline in the levels temperature shift in TIG-3 cells transformed by of RecQ helicases at late phases of culture was unlikely SV40tsT, because induction of host DNA synthesis due to waste of nutrient or accumulation of metabolic by SV40tsT was similarly insensitive to temperature: products. The higher levels of BLM and RTS at the S, the SV40tsT used in this study can induce and G2 and M phases in actively dividing cells imply that maintain host DNA synthesis at both permissive and BLM and RTS are perhaps sepeci®cally synthesized or non-permissive temperatures in resting cells (Ide et al., maintained in the late G1 and/or S phases. 1983, 1984). Recently, Ellis et al. (1999); Ne€ et al.

Oncogene Expression of RecQ helicases in transformed cells T Kawabe et al 4769

Figure 4 Levels of RecQ helicases at log and stationary phases. (a) Comparison between log and stationary phases in the expression of RecQ helicases in TIG-3 cell line. On day 0, 106 TIG-3 cells/100 mm dish were seeded, and the cells were harvested on days 1, 2, 3 and 4, and the expression of RecQ helicases was determined by immunoblot. (b) Flow cytometry of TIG-3 cell line. After cells were harvested, they were ®xed with 70% ethanol in PBS, treated with RNase and were stained with propidium iodide. The cells were ®ltered using a mesh (FALCON 2235), and were analysed by EPICSR XL (Beckman Coulter). (c) Comparison between log and stationary phases in the expression of RecQ helicases in U251MG, T98G, A2182 and KMST-6 cell lines. On day 0, 106 cells/100 mm dish were seeded, the cells were harvested on days 2 and 4, and the expression of RecQ helicases was determined by immunoblot. (d) Flow cytometry of U251MG cell line

(1999) showed that increased sister chromatid exchange level even in resting B cells, and its expression was not in cells from Bloom syndrome patients is suppressed by much a€ected by transformation with EBV or by complementation with normal BLM. Our observation stimulation with PMA. Furthermore, the expression of that BLM are up-regulated during DNA replication this helicase was not cell cycle dependent in the diploid implies that BLM is perhaps involved in suppressing human ®broblast cell line (TIG-3). It is also noted that the sister chromatid exchange that is considered to the mRNA of RecQL5 was expressed in almost all the occur during DNA replication. (4) The mode of the tissues examined including the brain, liver, skin and expression RecQL5b was quite di€erent between muscle that are considered to contain a relatively small normal cells and tumor cells. RecQL5b was unique in number of replicating cells (Kitao et al., 1998). These that this helicase was expressed at a relatively high facts suggest that in normal tissues RecQL5b plays an

Oncogene Expression of RecQ helicases in transformed cells T Kawabe et al 4770 important role in resting cells in G0 and G1, perhaps in The results of this study strongly support the idea transcription. In the four tumor cell lines, however, the that up-regulation of WRN, BLM, RTS and expression of RecQL5b was cell cycle-dependent, and RecQL1 is ®rmly associated with actively dividing was upregulated in log phase (Figure 4c). These results cells, and perhaps contributes in various ways to the suggest that RecQL5b plays di€erent roles in normal genomic stability of cells that are expected to be diploid cells and tumor cells. faced with more errors in DNA replication, or Lastly, we speculate on the biological meaning of the entangled DNA or RNA/DNA molecules occurring up-regulation of RecQ helicases in general during during active gene expression. RecQL5b, however, transformation with DNA tumor viruses or stimulation may play other important function(s) in normal with PMA. Werner, Bloom and Rothmund-Thomson diploid cells in G0. syndromes are characterized by genomic instability in patient cells, accompanied by a high incidence of cancer (Ellis et al., 1995; German, 1993; Goto et al., 1996; Kitao et al., 1999b; Lindor et al., 1996; Yu et al., Materials and methods 1996). For instance, EBV-transformed LCLs from Werner syndrome patients show various increased Cells and cell culture chromosomal aberrations (Salk et al., 1985; Tahara et The puri®cation of PBLs, transformation of PBLs by EBV, al., 1997), suggesting that WRN helicase is required for and the culture of established B-lymphoblastoid cell lines genomic stability, notably detected in actively dividing (LCLs) from a normal individual (N0008T) and Werner transformed cells. The fact that B cells from the syndrome patients WS11001 and WS11301 with WRN gene mutations 6/6 and 1/4, respectively (Matsumoto et al., 1997), patients lacking WRN can be transformed by EBV have been described (Tahara et al., 1997). An immortalized into LCLs as eciently as those from normal EBV-transformed LCL with strong telomerase activity individuals (Tahara et al., 1997) indicates that these (N6803IM) was also used. CD19-positive resting B cells were RecQ helicases are not essential in the process of cell puri®ed from PBLs from normal individuals using the pan-B transformation. On the other hand, cells from Bloom Dyanabeads M450 kit (Dyanal, Oslo), and were resuspended syndrome patients often fail to give a lymphoblastoid at 106/ml in the culture medium: RPMI1640 (pH 6.5) cell line when exposed to EBV (JL German, personal supplemented with penicillin, streptomycin and 10% heat- communication), suggesting that BLM plays an inactivated fetal calf serum. In the experiments to examine important role in transformation of B cells. Natural the e€ect of PMA, B cells were incubated with 30 ng/ml infection with DNA tumor viruses is considered to PMA for 0, 5, 16 and 40 h. In the cell transformation experiment with SV40 temperature sensitive (ts) T-antigen, induce transformation of various cells in vivo, such as human fetal lung ®broblasts TIG-3 (Matsuo et al., 1982) were acute mononucleosis caused by EBV infection. In this transformed with pMT-1ODtsA that codes for an origin- regard, Burkitt lymphoma accompanied by character- defective SV40tsT gene from which were established cell lines istic tumor-speci®c chromosomal translations is con- SVts9-3, SVts9-4 and SVts8 (Tahara et al., 1995). SVts9-3 sidered to arise in boys with X-linked and SVts9-4 cells had a prolonged lifespan, but senesced after lymphoproliferative disease after infection by EBV serial passage around 100 population doubling levels (PDLs): (Egeler et al., 1992). SVts8 was immortalized (Tsuyama et al., 1991). SVOD1-2 Resting B cells become actively proliferating lym- ®broblasts were transformed by normal SV40 T-antigen. All phoblasts by antigen stimulation in immune responses. these cells were seeded at 56105 cells/100 mm dish at 348C, RecQ helicases may be up-regulated in EBV-trans- and the next day the temperature was changed to 38.58C for the cells of non-permissive temperature groups. Cells were formed cells and antigen-stimulated B-lymphoblasts in recovered by trypsinization 24 h after the temperature vivo, as observed in vitro in this study. We speculate change. Control cells were cultured continuously at permis- that the increased levels of RecQ helicases in sive temperature 348C. HUE101-1 (Amanuma and Mitsui, proliferating cells may be required to guarantee 1991), a human endothelial cell line isolated from human genomic stability of these cells. Previously, we showed umbilical cord veins, was cultured at 378C in MCDB151 that WRN helicase, when introduced into the yeast medium (Sigma) supplemented with 10% fetal bovine serum sgs1 mutant, could suppress the hyper-recombination (BioWhittaker), 5 ng/ml of acidic ®broblast growth factor of the mutant yeast cells which lack SGS1 protein (Wako Pure Chemicals, Osaka, Japan), 5 mg/ml of heparin belonging to the RecQ helicase family (Yamagata et sodium salt (GIBCO BRL). HUE101-1 cells ceased prolifera- al., 1998). In addition, we showed that LCLs from tion at approximately 65 PDLs. KA62, 7, 18, 27 and 31 cell lines were established by transforming HUE101-1 cells with Werner syndrome patients lacking WRN helicase show pMT1-ODtsA. Transformed cells showed an extended augmented genomic instability accompanied by abnor- proliferative life span, and some of them were immortalized. mal telomere dynamics (Tahara et al., 1997). These hT-1 and hT-2 cells were established by transformation of ®ndings further support the notion that WRN helicase HUE101-1 with pIRES-hTERT-hygr which carried telomer- is involved in genome stabilization. ase reverse transcriptase gene (hTERT) in pIRES-hygr vector As the e€ects of PMA on normal cellular functions with hygromycin resistance gene. 7 ± 3, 18 ± 11, 27 ± 13, 31 ± 2 are mostly mediated by activating the enzyme protein and 62 ± 5 cell lines were established by transforming 7, 18, kinase C (Castagna et al., 1982), the early increase in 27, 31 and KA62 cells with pIRES-hTERT-hygr, respectively, WRN and RecQL1 after PMA treatment could also and all were immortalized. The characteristics of these be mediated by this pathway. Alternatively, it is transformed cell lines will be published elsewhere. pIRES- hTERT-hygr was kindly donated by Professor Fuyuki possible that DNA tumor viruses use the same Ishikawa, Graduate School of Bioscience and Biotechnology, pathway or an unidenti®ed alternative strategy to Tokyo Institute of Technology. The glioblastoma cell lines augment host RecQ DNA helicases, eventually to U251MG and T98G, as well as ®broblast tumor cell line maintain their own genomic stability since SV40tsT A2182 and transformed ®broblast cell line KMST-6, were seems to be directly involved in up-regulating WRN kindly donated by Professor M Oshimura of the Faculty of helicase. Medicine of Tottori University.

Oncogene Expression of RecQ helicases in transformed cells T Kawabe et al 4771 Antibodies and immunoblotting 1999). Brie¯y, the cell pellets were lysed by RIPA bu€er containing 10 m Tris-HCl (pH 7.4), 1% NP-40, 0.1% Rabbit polyclonal antibodies against human RecQL1 heli- M sodium deoxycholate, 0.1% sodium dodecylsulfate, 0.15 case (Tada et al., 1996) and BLM (Ne€ et al., 1999) were M NaCl, 1 mM ethylenediaminetetraacetate and protease in- reported previously as was mouse monoclonal antibody hibitors (CompleteTM Protease Inhibitor Cocktail Set, Roche 4H12 against the C-terminal end of WRN helicase (Shiratori Diagnostics K.K., Kamakura, Japan), and the proteins were et al., 1999). Mouse monoclonal antibodies against human separated by sodium dodecylsulfate-polyacrylamide gel RTS and RecQL5b helicases were prepared by immunizing electrophoresis. The proteins were electrophoretically trans- mice with a puri®ed recombinant C-terminal fragment of ferred to the polyvinylidene di¯uoride membrane Immobilon RTS (amino acid residues 907 ± 1208) or of RecQL5b (amino (Millipore) and the membrane was blocked by incubating acid residues 848 ± 991), respectively (the details of these for 1 h in a bu€er containing 5% skim milk and 0.2% antibodies will be published elsewhere). Antibodies against Tween 20. RecQ helicases were detected using a speci®c BLM (C-18) (Santa Cruz Biotechnology Inc., California monoclonal or polyclonal antibody followed by rabbit anti- USA) was also used. The antibodies against WRN, BLM, mouse immunoglobulin, rabbit anti-goat immunoglobulin or RTS, RecQL1 and RecQL5b stained the bands of 180, 180, goat anti-rabbit immunoglobulin that were conjugated with 160/140, 73 and 135 kDa in immunoblot, respectively. The horseradish peroxidase (Dako, Denmark). The membrane antibodies against WRN, BLM and RTS were shown in was developed using enhanced chemiluminescence ECLTM other studies to stain the corresponding bands or cells for (Amersham Life Science, UK). samples from normal individuals, but they failed to stain the bands or cells of samples from Werner syndrome (Goto et al., 1999; Shiratori et al., 1999), Bloom syndrome (unpub- Determination of the number of RECQ helicase copies per cell lished observation) and Rothmund-Thomson syndrome patients (unpublished observation), respectively. Because The number of molecules per cell were roughly determined the antibody against RTS did not stain the major for the ®ve RecQ helicases essentially according to the (160 kDa) and minor bands (140 kDa) of the samples from method by Moser et al. (2000). Brie¯y, we used quantitative Rothmund-Thomson patients, we judged that both the immunoblotting versus puri®ed recombinant RecQ helicase major and minor bands represent speci®c RTS (unpublished standard proteins to determine the number of copies per cell observation). The antibody against RecQL5b mainly stained for resting B cells, EBV-transformed LCL (N0008) cells and nuclei (Shimamoto et al., 2000). Mouse monoclonal IgG1 immortalized LCL (N6803) cells. The recombinant standard against a-actin (ICN Costa Mesa, CA, USA) was used to proteins with 66HIS at the N teminus were as follows: con®rm that the same amounts of protein were applied to rWRN (877 ± 1432 a.a.), rBML (1248 ± 1415 a.a.), rRTS the sodium dodecylsulfate-polyacrylamide gels. The detailed (907 ± 1208 a.a.), rRecQL1 (339 ± 649 a.a.) and rRecQL5b procedure of the immunoblot was reported (Shiratori et al., (848 ± 991 a.a.).

References

Abb J, Bayliss GJ and Deinhardt F. (1979). J. Immunol., 122, HanaokaF,YamadaM,TakeuchiF,GotoM,MiyamotoT 1639 ± 1642. and Hori T. (1985). Adv. Exp. Med. Biol., 190, 439 ± 457. Amanuma K and Mitsui Y. (1991). Biochim. Biophys. Acta., HuangRP,FanY,NiZ,MercolaDandAdamsonED. 1092, 336 ± 340. (1997). J. Cell. Biochem., 66, 489 ± 499. Balajee AS, Machwe A, May A, Gray MD, Oshima J, Martin Ide T, Tsuji Y, Ishibashi S and Mitsui Y. (1983). Exp. Cell. GM, Nehlin JO, Brosh R, Orren DK and Bohr VA. (1999). Res., 143, 343 ± 349. Mol. Biol. Cell., 10, 2655 ± 2668. Ide T, Tsuji Y, Nakashima T and Ishibashi S. (1984). Exp. Bertoglio JH. (1983). J. Immunol., 131, 2279 ± 2281. Cell. Res., 321 ± 328. Brosh Jr. RM, Orren DK, Nehlin JO, Ravn PH, Kenny MK, Kamath-Loeb AS, Johansson E, Burgers PM and Loeb LA. Machwe A and Bohr VA. (1999). J. Biol. Chem., 274, (2000). Proc. Natl. Acad. Sci. USA, 97, 4603 ± 4608. 18341 ± 18350. Kataoka H, Tahara H, Watanabe T, Sugawara M, Ide T, BroshJ,LiJL,KennyMK,KarowJK,CooperMP, Goto M, Furuichi Y and Sugimoto M. (1997). Di€erentia- Kureekattil RP, Hickson ID and Bohr VA. (2000). J. tion, 62, 203 ± 211. Biol. Chem. 275, 23500 ± 23508. KitaoS,LindorNM,ShiratoriM,FuruichiYand CastagnaM,TakaiY,KaibuchiK,SanoK,KikkawaUand Shimamoto A. (1999a). Genomics, 61, 268 ± 276. Nishizuka Y. (1982). J. Biol. Chem., 257, 7847 ± 7851. KitaoS,OhsugiI,IchikawaK,GotoM,FuruichiYand Clevers HC, Versteegen JM, Logtenberg T, Gmelig-Meyling Shimamoto A. (1998). Genomics, 54, 443 ± 452. FH and Ballieux RE. (1985). J. Immunol., 135, 3827 ± Kitao S, Shimamoto A, Goto M, Miller RW, Smithson WA, 3830. Lindor NM and Furuichi Y. (1999b). Nat. Genet., 22, 82 ± Cooper MP, Machwe A, Orren DK, Brosh RM, Ramsden D 84. and Bohr VA. (2000). Genes Dev., 14, 907 ± 912. Kuchino T and Yamaguchi N. (1975). J. Virol., 15, 1302 ± Courcelle J and Hanawalt PC. (1999). Mol. Gen. Genet., 262, 1307. 543 ± 551. Lebel M, Spillare EA, Harris CC and Leder P. (1999). J. Biol. EgelerRM,deKrakerJ,SlaterRandPurtiloDT.(1992). Chem., 274, 37795 ± 37799. Cancer, 70, 683 ± 687. Lindor NM, Devries EM, Michels VV, Schad CR, Jalal SM, Ellis NA, Groden J, Ye TZ, Straughen J, Lennon DJ, Ciocci Donovan KM, Smithson WA, Kvols LK, Thibodeau SN S, Proytcheva M and German J. (1995). Cell, 83, 655 ± 666. and Dewald GW. (1996). Clin. Genet., 49, 124 ± 129. Ellis NA, Proytcheva M, Sanz MM, Ye TZ and German J. Lohman TM. (1993). J. Biol. Chem., 268, 2269 ± 2272. (1999). Am.J.Hum.Genet.,65, 1368 ± 1374. MarciniakRA,LombardDB,JohnsonFBandGuarenteL. German J. (1993). Medicine (Baltimore), 72, 393 ± 406. (1998). Proc. Natl. Acad. Sci. USA, 95, 6887 ± 6892. Goto M, Miller RW, Ishikawa Y and Sugano H. (1996). Cancer Epidemiol Biomarkers Prev, 5, 239 ± 246. Goto M, Yamabe M, Shiratori M, Okada M, Kawabe T, Matsumoto T, Sugimoto M and Furuichi Y. (1999). Human Genetics, 105, 301 ± 307.

Oncogene Expression of RecQ helicases in transformed cells T Kawabe et al 4772 Matsumoto T, Imamura O, Yamabe Y, Kuromitsu J, Spillare EA, Robles AI, Wang XW, Shen JC, Yu CE, Tokutake Y, Shimamoto A, Suzuki N, Satoh M, Kitao Schellenberg GD and Harris CC. (1999). Genes Dev., 13, S, Ichikawa K, Kataoka H, Sugawara K, Thomas W, 1355 ± 1360. Mason B, Tsuchihashi Z, Drayna D, Sugawara M, Stewart E, Chapman CR, Al-Khodairy F, Carr AM and Sugimoto M, Furuichi Y and Goto M. (1997). Hum. Enoch T. (1997). Embo. J., 16, 2682 ± 2692. Genet., 100, 123 ± 130. Suzuki N, Shimamoto A, Imamura O, Kuromitsu J, Kitao S, Matsuo M, Kaji K, Utakoji T and Hosoda K. (1982). J. Goto M and Furuichi Y. (1997). Nucleic. Acids Res., 25, Gerontol., 37, 33 ± 37. 2973 ± 2978. Miller G. (1990). Virology. Fields BN aK, KM (ed.).Raven Suzuki N, Shiratori M, Goto M and Furuichi Y. (1999). Press: New York, pp. 1921 ± 1958. Nucleic. Acids Res., 27, 2361 ± 2368. Moser MJ, Kamath-Loeb AS, Jacob JE, Bennett SE, Oshima Tada S, Yanagisawa J, Sonoyama T, Miyajima A, Seki M, Ui J and Monnat Jr. RJ. (2000). Nucleic. Acids Res., 28, 648 ± M and Enomoto T. (1996). Cell. Struct. Funct., 21, 123 ± 654. 132. Nakayama H, Nakayama K, Nakayama R, Irino N, Tahara H, Sato E, Noda A and Ide T. (1995). Oncogene, 10, Nakayama Y and Hanawalt PC. (1984). Mol. Gen. Genet., 835 ± 840. 195, 474 ± 480. Tahara H, Tokutake Y, Maeda S, Kataoka H, Watanabe T, Ne€ NF, Ellis NA, Ye TZ, Noonan J, Huang K, Sanz M and Satoh M, Matsumoto T, Sugawara M, Ide T, Goto M, Proytcheva M. (1999). Mol. Biol. Cell., 10, 665 ± 676. Furuichi Y and Sugimoto M. (1997). Oncogene, 15, 1911 ± Puranam KL and Blackshear PJ. (1994). J. Biol. Chem., 269, 1920. 29838 ± 29845. Tsuyama N, Miura M, Kitahira M, Ishibashi S and Ide T. Salk D, Au K, Hoehn H and Martin GM. (1985). Ad. Exp. (1991). Cell. Struct. Funct., 16, 55 ± 62. Med. Biol., 190, 541 ± 546. Umezu K and Nakayama H. (1993). J. Mol. Biol., 230, Schulz VP, Zakian VA, Ogburn CE, McKay J, Jarzebowicz 1145 ± 1150. AA, Edland SD and Martin GM. (1996). Hum. Genet., 97, Umezu K, Nakayama K and Nakayama H. (1990). Proc. 750 ± 754. Natl. Acad. Sci. USA, 87, 5363 ± 5367. Seki M, Miyazawa H, Tada S, Yanagisawa J, Yamaoka T, Wu L, Davies SL, North PS, Goulaouic H, Riou JF, Turley HoshinoS,OzawaK,EkiT,NogamiM,OkumuraK, H, Gatter KC and Hickson ID. (2000). J. Biol. Chem., 275, Taguchi H, Hanaoka F and Enomoto T. (1994). Nucleic. 9636 ± 9644. Acids Res., 22, 4566 ± 4573. Yamabe Y, Shimamoto A, Goto M, Yokota J, Sugawara M Shimamoto A, Nishikawa K, Kitao S and Furuichi Y. and Furuichi Y. (1998). Mol. Cell. Biol., 18, 6191 ± 6200. (2000). Nucleic. Acids Res., 28, 1647 ± 1655. Yamagata K, Kato J, Shimamoto A, Goto M, Furuichi Y ShiratoriM,SakamotoS,SuzukiN,TokutakeY,Kawabe and Ikeda H. (1998). Proc. Natl. Acad. Sci. USA, 95, Y, Enomoto T, Sugimoto M, Goto M, Matsumoto T and 8733 ± 8738. Furuichi Y. (1999). J. Cell. Biol., 144, 1±9. Yamaguchi N and Kuchino T. (1975). J. Virol., 15, 1297 ± Sinclair DA and Guarente L. (1997). Cell, 91, 1033 ± 1042. 1301. Sinclair DA, Millis K and Guarente L. (1997). Science, 277, Yu CE, Oshima J, Fu YH, Wijsman EM, Hisama F, Alisch 1313 ± 1316. R,MatthewsS,NakuraJ,MikiT,OuaisS,MartinGM, Spender LC, Cannell EJ, Hollyoake M, Wensing B, Gawn Mulligan J and Schellenberg GD. (1996). Science, 272, JM, Brimmell M, Packham G and Farrell PJ. (1999). J. 258 ± 262. Virol., 73, 4678 ± 4688.

Oncogene