Index References

Five genes encoding RecQ helicase homologues have been identified in human cells, three of which are the causative genes for Bloom syndrome (BS), Werner syndrome (WS), and Rothmund-Thomson syndrome (RTS) (Ellis et al., 1995; Yu et al., 1996; Kitao et al., 1999). BS is characterized by retarded growth, sunlight sensitivity, immunodeficiency, male infertility, and predisposition to a wide variety of malignant tumors (German, 1993), and WS is characterized by premature aging associated with an early onset of age-related diseases including arteriosclerosis, malignant tumors, melituria, and cataracts (Martin, 1985). WS cells show subtle defects in DNA replication, such as an extended S phase and reduced frequency of replication origin firing (Fujiwara et al., 1977). RTS is an autosomal recessive inherited disease characterized by premature aging, developmental abnormalities, and predisposition to cancers such as osteogenic sarcoma (Kitao et al., 1999). Some features of RTS are reportedly caused by mutations of human RecQL4 (Kitao et al., 1999). Moreover, it has also been reported that distinct RecQL4 mutations result in RAPADILINO syndrome and Baller-Gerold syndrome, suggesting that RecQL4 has multiple functions (Dietschy et al., 2007).

A number of reports have shown that many cellular events, including DNA repair, recombination, transcription, telomere maintenance, and apoptosis, are affected in WS cells. The gene responsible for WS encodes a protein (WRN) that possesses DNA helicase and exonuclease activities. It has been reported that WRN physically interacts with BLM, the protein that is defective in BS, as well as DNA polymerase δ, replication protein A, PCNA, DNA topoisomerase I (Top1), the Ku 70/86 complex, p53, PARP, TRF1, TRF2, and RAD52 (Kusumoto et al., 2007). In addition, we found a novel WRN-interacting protein, WHIP (Werner helicase interacting protein), with the yeast two-hybrid method (Kawabe et al., 2001). We have renamed WHIP as WRNIP1 (Werner Interacting Protein 1) according to the recommendation of HUGO. Purified WRNIP1 indeed interacts directly with WRN in an ATP stimulated manner (Kawabe et al., 2006). Moreover, WRNIP1 co-immunoprecipitates with WRN, suggesting that WRNIP1 interacts with WRN in cells (Kawabe et al., 2006).

To study the function of WRNIP1, we used budding yeast as a model system, because they possess both a WRNIP1 homologue MGS1 (Maintenance of Genome Stability 1) (Hishida et al., 2001), and a RECQ homologue, SGS1. MGS1 was identified as a gene that encodes a protein whose central region is similar to Escherichia coli RuvB. An mgs1 deletion mutant shows an elevated rate of mitotic recombination, which causes genome instability, and is synthetic lethal with rad6 (Hishida et al., 2002). The Saccharomyces cerevisiae cells with disruptions in both mgs1 and sgs1 grow much more slowly than either single gene mutant, have a higher frequency of recombination, and tend to arrest in G₂/M phase, suggesting that Mgs1 and Sgs1 perform nonoverlapping functions to repair or otherwise accommodate DNA lesions (Branzei et al., 2002a). These results prompted us to generate vertebrate WRNIP1-WRN double-gene-knockout cells and examine their phenotypes. Although the frequency of spontaneously occurring sister chromatid exchange (SCE) in the wrnip1/wrn mutant is slightly increased compared to that of either single mutant, the double mutants grow at a rate similar to that of wild-type cells (Kawabe et al., 2006), clearly contrasting with budding yeast mgs1-sgs1 cells (Branzei et al., 2002a). One reason for this inconsistency between vertebrate cells and yeast may be the existence of multiple RecQs in vertebrate cells, whereas yeast has a single RecQ helicase, Sgs1. In addition, the human BLM gene, but not WRN, reportedly suppresses some of the phenotypes of budding yeast sgs1 cells (Heo et al., 1999) suggesting that Sgs1 is more functionally similar to BLM than WRN. In this context, we decided to examine the functional relationship between WRNIP1 and BLM in vertebrate cells.

To examine the functional relationship between BLM and WRNIP1, we generated a double-gene-disrupted mutant, wrnip1/blm, from chicken DT40 cells. Since we had already generated wrnip1 and blm DT40 cells (Kawabe et al., 2006; Wang et al., 2000), we generated wrnip1/blm DT40 cells by transfecting two BLM-targeting constructs successively into wrnip1 cells. The gene disruption of WRNIP1 and BLM was verified by RT-PCR (Fig. 1). A truncated WRNIP1 mRNA was expressed in wrnip1 and wrnip1/blm cells; our previous characterization indicated that the putative truncated WRNIP1 protein, if it exists in the cells, seems to have no biological activity (Kawabe et al., 2006).

View Details

Fig. 1 Generation of wrnip1/blm cells. DT40 cells were cultured in RPMI 1640 supplemented with penicillin, streptomycin, amphotericin B, 10% fetal bovine serum, and 1% chicken serum at 39.5°C. The wrnip1 and blm disruptants were generated as reported previously (Wang et al., 2000; Kawabe et al., 2006). To generate wrnip1/blm DT40 cells, we used a Gene Pulser apparatus (BioRad, Hercules, CA, USA) at 550 V and 25 μF to electroporate 107 wrnip1 cells with 30 μg of linearized BLM-targeting constructs (Wang et al., 2000) containing either a hygromycin or a puromycin selection marker. Drug-resistant colonies were selected in 96-well plates with medium containing 2.5 mg/ml hygromycin or 0.5 μg/ml puromycin. Gene disruption was confirmed by RT-PCR of total RNA from wild-type (lane 1), wrnip1 (lane 2), blm (lanes 3 and 4), and wrnip1/blm DT40 cells (lanes 5 [clone #1] and 6 [clone #18]). Primers used to detect WRNIP1 are as described previously (Kawabe et al., 2006). Open arrowhead and closed arrowhead indicated wild-type and the truncated form WRNIP1 mRNA, respectively, due to WRNIP1 disruption (Kawabe et al., 2006). The primers used to detect BLM and RECQL1 mRNA are as follows: BLM-sense: 5’-GGTGGTAAAAGTCTGTGCTAC-3’ BLM-antisense: 5’-CTACAGATTTTGGAAGGGAAGC-3’ RECQL1-sense: 5’-ATGACAGCTGTGGAAGTGCTAGAGGA-3’ RECQL1-antisense: 5’-TCAGTCAAGAACAACAGGTTGGTCATCTC-3’ RECQL1 was amplified as a loading control.

First, we examined the proliferation of wrnip1, blm, and wrnip1/blm cells. Consistent with our previous results, wrnip1 cells grew at a rate similar to that of wild-type cells, whereas blm cells proliferated more slowly than wild-type cells (Fig. 2A). wrnip1/blm cells proliferated almost the same as blm cells. Moreover, wrnip1 cells showed a high plating efficiency, similar to wild-type cells, and the plating efficiencies of wrnip1/blm cells were almost the same as blm cells (Fig. 2B).

View Details

Fig. 2 Growth curves and plating efficiency of wrnip1/blm DT40 cells. (A) Growth curves. Cells were inoculated into 6-well plates and counted after the indicated periods. Cells that were negative for trypan blue staining were counted as viable cells. The experiments were repeated three times, and representative data are shown. The bars indicate standard deviations. Often, the duplicate data were so close that the standard deviations are hidden by the symbols. (B) Plating efficiencies of cells with the indicated genotypes. Each value is the average from three independent experiments. Bars indicate standard deviations.

The most characteristic feature of BS cells is genomic instability, which is manifested as more frequent chromosome breaks, interchange between homologous chromosomes, and SCE. Elevated SCE, a hallmark of BS, is thought to be an excellent indicator of replication defects. Therefore, we next examined the frequency of SCE in wrnip1/blm cells. As reported previously, SCE was greatly and slightly elevated in blm cells and wrnip1 cells, respectively (Fig. 3). The SCE frequency of wrnip1/blm cells was additively increased compared to that of either single mutant. Since many proteins are involved in the formation of SCE, and a biochemical analysis has shown that BLM and TOP3α dissolve double Holliday junctions to prevent crossover (Wu and Hickson, 2003), the elevation of spontaneous SCE in blm cells can be explained by the failure of the dissolution of recombination intermediates. Previous studies of Mgs1 suggested that its function is somehow related to the function of DNA polymerase δ (Polδ), which is required for cellular DNA replication (Hishida et al., 2002; Branzei et al., 2002b). In this context, we have observed stimulation of Polδ activity by WRNIP1 at the initiation stage in vitro (Tsurimoto et al., 2005). Thus, the slight elevation of SCE in wrnip1 cells can be explained if lesions that remain on replicating DNA due to the defect of WRNIP1 are channeled toward a homologous recombination repair pathway that uses a sister chromatid as a template. It is likely, therefore, that the increase of spontaneous SCE in wrnip1/blm cells indicates that DNA replication is impaired in the cells lacking WRNIP1, resulting in the production of or a failure to remove lesions that are processed by homologous recombination; then the recombination intermediates are resolved to form a crossover, resulting in SCE in the absence of BLM. These results coincide well with those obtained in budding yeast, because a synergistic increase of spontaneous sister chromatid recombination is also observed in mgs1-sgs1 double mutants (Branzei et al., 2002a). Therefore, it seems likely that vertebrate WRNIP1 and BLM play roles similar to those of their budding yeast counterparts, Mgs1 and Sgs1, acting independently to deal with lesions during DNA replication and consequently to decrease sister chromatid recombination. In contrast, it was reported that the disruption of the WRN gene in blm cells resulted in 2/3 reduction of SCE frequency (Imamura et al., 2002; Dong et al., 2007), suggesting that WRNIP1 and WRN have opposing functions in blm cells in the prevention of SCE formation. Thus, although WRNIP1 binds to WRN, WRNIP1 may act independently of WRN in dealing with DNA lesions during DNA replication.

View Details

Fig. 3 Frequency of spontaneous SCE seen in wrnip1/blm cells. Preparation of the cells on the slides for counting spontaneously occurring sister chromatid exchanges (SCE) was described previously (Wang et al., 2000). SCE levels are shown for wild-type (A), wrnip1 (B), blm (C), and wrnip1/blm (D) cells. Numbers represent means and standard deviations of scores from 150 metaphase cells.

We next examined the sensitivity of wrnip1/blm cells to the DNA-damaging agents methyl methanesulfonate (MMS) and camptothecin (CPT) to investigate the functional relationship between WRNIP1 and BLM in the presence of DNA damage. blm cells were moderately more sensitive to MMS than wild-type cells (Fig. 4A), in agreement with the results of a previous study (Imamura et al., 2002). In contrast, the sensitivity of wrnip1 cells to MMS was similar to that of wild-type cells (Fig. 4A), consistent with previous results (Kawabe et al., 2006; Yoshimura et al., 2006), and there was little or no enhancement of sensitivity to MMS in wrnip1/blm cells over that seen in blm cells. WRNIP1 thus plays no apparent role in DNA repair or damage tolerance for lesions generated by MMS, even in the absence of BLM. In contrast, wrnip1 cells were moderately sensitive to CPT, to a degree comparable to that seen in blm cells, and there was an additive increase in the sensitivity to CPT in the wrnip1/blm cells (Fig. 4B). CPT is a Top1 inhibitor and stabilizes so called “cleavage complexes,” which are formed by reversible transfer of a phosphodiester bond to a tyrosine residue in the catalytic site of the enzyme, causing single-stranded nicks (Pommier, 2004). The cytotoxicity of CPT is thought to be due to the DNA double-stranded breaks (DSBs) that are formed when replication forks encounter the single-strand nicks. Thus, both WRNIP1 and BLM seem to respond to DNA lesions generated by CPT, such as DSBs, but through independent pathways.

View Details

Fig. 4 Sensitivity of wrnip1/blm DT40 cells to MMS and CPT. Cells were treated with the indicated concentrations of MMS (A) or CPT (B). Briefly, cells (2~4 × 102) were plated into dishes containing various concentrations of methyl methanesulfonate (MMS) or camptothecin (CPT) in D-MEM/F-12 medium supplemented with 1.5% (w/v) methylcellulose, 1.5% chicken serum, and 15% fetal bovine serum. After 10 to 14 days, visible colonies were counted. The data represent the percent survival compared to untreated cells. The experiments were repeated three times, and representative data are shown. The bars indicate standard deviations.

Genetic studies using yeast revealed that the simultaneous mutation of MGS1 and RAD6 (Hishida et al., 2002), RAD18 or RAD5 causes a synthetic growth defect, suggesting that Mgs1 and Rad18/Rad5 act in a similar but parallel pathway in processing stalled replication forks, perhaps by promoting a template switching mode of DNA synthesis (Branzei et al., 2004; Vijeh Motlagh et al., 2006). We previously found that human WRNIP1 binds three of four subunits of human Polδ and stimulates the activity of Polδ in vitro (Tsurimoto et al., 2005). When the possible function of Mgs1 in yeast, the in vitro activities of WRNIP1, and the results obtained in the current study are taken into account, it seems likely that WRNIP1 functions in processing stalled replication forks by interacting with Polδ. Sgs1 is reported to be involved in the reestablishment of replication forks or replication restart (Cobb et al., 2003), and very recently it was reported that BLM is involved in efficient replication fork restart (Davies et al., 2007). In addition, our recent study using various gene knockout cells, including xrcc3/blm cells, suggested that BLM functions downstream of XRCC3, in a pathway involving a template switching mode of DNA synthesis (Otsuki et al., 2007). In conclusion, WRNIP1 and BLM function in similar but independent pathways to repair or accommodate DNA lesions.

This work was supported by Grants-in-Aid for Scientific Research and for Scientific Research on Priority Areas from the Ministry of Education, Science, Sports, and Culture of Japan.