Long-Patch Base Excision Repair of Apurinic/Apyrimidinic Site DNA Is

Long-patch base excision repair of apurinic/apyrimidinic site DNA is decreased in mouse embryonic ﬁbroblast cell lines treated with plumbagin: involvement of cyclin-dependent kinase inhibitor p21Waf-1/Cip-1

Aruna S Jaiswal1,2, Linda B Bloom3 and Satya Narayan*,1,2

1Department of Anatomy and Cell Biology, College of Medicine, The University of Florida, Gainesville, Florida, FL 32610, USA; 2UF Shands Cancer Center, College of Medicine, The University of Florida, Gainesville, Florida, FL 32610, USA; 3Department of Biochemistry and Molecular Biology, College of Medicine, The University of Florida, Gainesville, Florida, FL 32610, USA

Molecular interactions among cell cycle and DNA repair repair (BER) systems of both prokaryotes and eukar- proteins have been described, but the impact of many of yotes repair modified bases generated by ionizing these interactions on cell cycle control and DNA repair radiation, alkylating agents, and endogenous hydrolytic remains unclear. The cyclin-dependent kinase inhibitor, and oxidative processes (reviewed by Seeberg et al., p21, is known to be involved in DNA damage-induced 1995). Biochemical and genetic studies indicate that cell cycle arrest and blocking DNA replication and BER occurs through two sub-pathways that are repair. Participation of p21 has been implicated in differentiated by repair gap size and the enzymes nucleotide excision repair. However, the role of p21 in involved (reviewed by Wilson, 1998). These sub-path- the base excision repair (BER) pathway has not been ways are designed as ‘single-nucleotide BER’ also thoroughly studied. In the present investigation, we referred to as ‘short patch (SP) BER’, and ‘multi- treated isogenic mouse embryonic fibroblast (MEF) cell nucleotide BER’ also referred to as ‘long patch (LP) lines containing wild-type (MEF-polb)orDNA poly- BER’. In both sub-pathways, repair is initiated by merase b (polb) gene-knockout (MEFpolbKO) with excision of a damaged base by a DNA glycosylase oxidative DNA-damaging agent, plumbagin, and exam- leaving an abasic or AP site in DNA. The resulting AP ined its effect on p21 levels and BER activity. Plumbagin sites, which are also generated spontaneously or by treatment caused a S-G2/M phase arrest and cell death radiation and chemicals, are subsequently acted upon by of both MEF cell lines, induced p21 levels, and decreased an AP endonuclease (APE) to generate a 3’ hydroxyl p21-mediated long-patch (LP) BER by blocking DNA group and a 5’-deoxyribosephosphate (dPR) terminus. ligase activity in the polb-dependent pathway and by In SP-BER, DNA polymerase b (pol b) extends the 3’ blocking both FEN1 and DNA ligase activity in polb- terminus by a single nucleotide and removes the dRP independent pathway. These findings suggest that moiety with its dRP lyase activity. Finally, the nick is plumbagin induced p21 levels play a regulatory role in sealed by DNA ligase I or DNA ligase III/XRCC1 cell cycle arrest, apoptosis, and polb-dependent and (Srivastava et al., 1998). When AP sites are oxidized or -independent LP-BER pathways in MEF cells. reduced, they become resistant to b-elimination and Oncogene (2002) 21, 5912 – 5922. doi:10.1038/sj.onc. cannot be excised by the dRP lyase activity of pol b.In 1205789 these cases, the modified AP site is repaired via the LP- BER pathway where a DNA polymerase (b, d,ore) Keywords: DNA damage; cell cycle arrest; p21; long- incorporates 2 – 15 nucleotides displacing the strand patch BER containing the modified AP site. This DNA flap structure is cleaved by FEN1 and the nick is sealed by a DNA ligase (Biade et al., 1998; Dianov et al., 1999; Introduction Fortini et al., 1998; Frosina et al., 1996; Klungland and Lindahl, 1997; Prasad et al., 2000, 2001). The Exogenous and endogenous mutagenic agents attack the importance of BER in maintaining genomic integrity genomes of all living cells. DNA bases damaged by these and reducing cancer susceptibility is evidenced by the agents may be cytotoxic and/or mutagenic, and thus, fact that mutations and/or different levels of expression they are a major source of intermediates in carcinogen- of polb have been observed in many colon and lung esis (reviewed by Krokan et al., 1997). The base excision tumors (Bhattacharyya et al., 1999; Canitrot et al., 2000; Srivastava et al., 1999; Wang et al., 1992). Regulation of DNA replication and repair following *Correspondence: S Narayan, UF Shands Cancer Center, College of DNA damage is critical to mutation avoidance. The Medicine, Academic Research Building, Room R4-216, P.O. Box cyclin-dependent kinase inhibitor, p21Waf1-Cip1 (here- 100232, University of Florida, Gainesville, Florida FL32610, USA; E-mail: [email protected]fl.edu after p21), is one protein involved in regulating cell Received 14 March 2002; revised 6 June 2002; accepted 18 June cycle arrest and inhibition of DNA replication 2002 following DNA damage providing time for DNA to p21-mediated decrease in long-patch BER in plumbagin-treated cells AS Jaiswal et al 5913 be repaired. Interactions between p21 and proliferating husk surrounding the walnuts, a valuable commercial cell nuclear antigen (PCNA) are believed to be crop in the United States. In common with other responsible for blocking DNA synthesis (Waga and napthoquinones, it is an active redox agent. In Stillman, 1998). PCNA stimulates DNA synthesis by bacterial assay systems, plumbagin-induced oxidative acting as a processivity factor for pols d and e damage is associated with plumbagin-mediated genera- (reviewed by Jonsson and Hubscher, 1997). In vitro, tion of reactive oxygen species (ROS; Jamieson et al., binding of p21 to PCNA blocks the ability of PCNA to 1994; Kato et al., 1994; Prieto-Alamo et al., 1993). act as a processivity factor for pols d and e, modulates ROS can react with cellular components, including the primer-template recognition complex, and inhibits proteins, lipids, and nucleic acids, to induce DNA DNA replication. In vivo, the interaction of p21 with adducts such as 8-hydroxy-2’-deoxyguanosine (8-OH- PCNA is critical in regulation of DNA replication and dG) (reviewed by Ames, 1991; Beckman and Ames, cell cycle progression. Residues in the C-terminal 1997). It is believed that base excision repair (BER) domain, amino acids 141 – 160, of p21 bind to PCNA pathway predominantly repairs these oxidized bases to inhibit replication and residues in the N-terminal (Bjoras et al., 1997; reviewed by Brozmanova et al., domain bind to cyclin-Cdk complexes to inhibit cell 2001; Cadet et al., 2000). The DNA-damaging property cycle progression (reviewed by Dotto, 2000). PCNA is of plumbagin makes it an effective chemotherapeutic required for other pathways including LP-BER and has agent. Recently, plumbagin has also been shown to also been shown to interact with and stimulate a have a protective effect in azoxymethane-induced number of other proteins required for DNA replication intestinal carcinogenesis in rats (Sugie et al., 1998). and repair including FEN1, mismatch repair proteins These studies suggest that plumbagin may be a MSH2 and MLH1, the nucleotide excision repair potential chemopreventive agent for human intestinal endonuclease XP-G, human DNA-(cytosine-5) methyl neoplasia, especially, since minimal toxicity has been transferase, and DNA ligase I (reviewed by Dotto, observed in the animal model (Sugie et al., 1998). 2000; Kelman, 1997; Warbrick, 2000). Thus, the Elucidation of the mechanism(s) by which plumbagin possibility exists that p21 inhibits other pathways induces anti-tumorigenic and anti-carcinogenic activity including LP-BER by binding PCNA. is necessary for providing a solid foundation for its use DNA repair is often coupled with the arrest of cells as an agent for therapeutic strategies. In the present in G1 and S phase of the cell cycle (Holmquist, 1998). study, we hypothesize that ROS generated by plumba- However, prolonged arrest of cells in the G2/M phase gin affects the levels of p21, which in turn regulates of the cell cycle would be more advantageous by enzymes involved in the BER pathway. Depending providing more time to repair the damaged DNA and upon the extent of DNA damage, increase in the p21 posing less threat to genomic integrity than DNA level, and efficiency of repair, cells will be either damage present during G1 and S phase of the cell cycle arrested in the cell cycle to enable DNA repair or (Nasmyth, 1996). In any case, if the damage is un- undergo apoptosis. repairable, then cell death (apoptosis) provides a fall- back mechanism by which genomic integrity is Results maintained. In past years, the clinical and in vitro evidence of collateral sensitivity to multiple agents Growth arrest and protein levels of p21, polb, and PCNA suggested that it is the signaling of DNA damage (the of MEF cells treated with plumbagin downstream events) that differ in cancer cells, which resulted in chemosensitivity (Bedford et al., 1998; Hill In the present studies, to determine the role of polb- et al., 1994; Koberle et al., 1997; Sark et al., 1995). One dependent and -independent base excision repair (BER) of the highly studied DNA damage-induced down- pathways on DNA damage-induced cell cycle arrest, stream signaling molecules is p53, which by regulating we used a mouse embryonic fibroblast (MEF) cell line p21 gene expression causes cell cycle arrest and (MEF-polb) and the matched littermate of polb gene- apoptosis in target cells (Jaiswal and Narayan, 2002; knockout cell line (MEF-polbKO). These cell lines reviewed by Boulaire et al., 2000; Prosperi, 1997; were treated with different concentrations of plumbagin Taylor and Stark, 2001). In previous studies, although for 10 h. Cell cycle profiles were evaluated by controversial (Li et al., 1994a; McDonald et al., 1996; propidium iodide staining of nuclei, and the analysis Pan et al., 1995; Sheikh et al., 1997; Shivji et al., 1994), of their distribution in different phases of the cell cycle the role of p21 is reported in NER (reviewed by was done by FACScan flow cytometer. An increased Prosperi, 1997). However, whether DNA damage- arrest occurred in the S-G2/M phase of both MEF- induced levels of p21 collaborate in cell cycle arrest polb and MEF-polbKO cell lines after plumbagin through BER pathway is not well established. treatment (Figure 1a). Treatment of cells with 5 mM In the present study, the role of p21 in cell cycle plumbagin resulted in increased cell death (population arrest, apoptosis and DNA repair was examined using of cells in the sub-G1 phase). A decrease in the number plumbagin, 5-hydroxy-2-methyl-1,4-napthoquinone, a of cells in G2/M phase, an increase in cell death, and natural yellow pigment found in plants of the no change in the number of cells arrested in S phase Plumbaginaceae, Droseraceae, Ancistrocladaceae and was observed for MEF-polb cells treated with 5 mM Dioncophyllaceae families. Plumbagin is one of a series plumbagin. On the other hand, in the MEF-polbKO of well-established napthoquinones found in the fleshy cell line, the cell death was followed directly from the S

Oncogene p21-mediated decrease in long-patch BER in plumbagin-treated cells AS Jaiswal et al 5914

Figure 1 (a) Cell cycle proﬁle of MEF cell lines treated with plumbagin. The ranges for G0/G1,S,G2/M and sub-G1 phase cells were established based upon the corresponding DNA content of the histograms. At least 10 000 cells per sample were considered in the gated regions for calculations. The data are representative of two independent experiments. (b) Plumbagin-induced protein levels in MEF cell lines. The MEF cell lines were treated with different concentrations of plumbagin for 10 h. Cellular extracts were prepared and processed for Western blot analysis for determining the protein levels of p53, p21, polb and PCNA. Photographs of the autoradiograms are representatives of three independent experiments. A quantitative analysis of p21 protein levels is given in (c). The data are mean+s.e. of three experiments. * and ** are signiﬁcantly different than the untreated MEF-polb and MEF-polbKO cell lines, respectively

phase arrested cells treated with 5 mM plumbagin. clearly indicated that polb participates in SP- and L- These results suggest that plumbagin-treated growth BER pathways, and that both pathways occur arrest of MEF-polb and MEF-polbKO cell lines occur simultaneously when examined either with MEF-polb at the G2/M and S phases of the cell cycle, respectively, and MEF-polbKO cell extracts (Fortini et al., 1998) or which lead to apoptosis at higher concentrations of with purified BER proteins (Klungland and Lindahl, plumbagin treatment. 1997). We used an in vitro BER assay system based on To determine whether plumbagin treatment affected the extracts from MEF-polb and MEF-polbKO cell protein levels of p21 and BER proteins, polb and lines to determine the effect of plumbagin treatment on PCNA, a Western blot analysis was performed on SP- or LP-BER in these cells. In these experiments, extracts prepared from MEF cells treated with cells were either treated with 5 mM plumbagin for 10 h plumbagin for 10 h. The level of p21 significantly or untreated. increased in both MEF-polb and MEF-polbKO cell lines (Figure 1b,c), whereas the levels of PCNA and Pol b-dependent and -independent BER activity in polb were unchanged at these concentrations of extracts from plumbagin-treated MEF cell lines plumbagin (Figure 1b). In previous studies, we (Jaiswal and Narayan, 2002) and others (reviewed by Boulaire In previous studies, oxidative stress-induced expression et al., 2000; Prosperi, 1997; Taylor and Stark, 2001) of p21 at the mRNA and protein levels has been shown have established that p21 expression in cells is induced to be p53-independent but MAP kinase-dependent in a p53-dependent manner after treatment with DNA- (Esposito et al., 1998). Recently, using 7-ketocho- damaging agents. However, the induction of p21 after lesterol, an oxidizing agent, Agrawal et al. (2002) have plumbagin treatment in the MEF cell lines is p53- shown the role of Stat1 in p21-mediated apoptosis of independent. The p53 protein was not detectable in MEF cells lacking p53. Currently, it is unknown these MEF cell lines because they expressed the SV40 whether similar mechanisms are involved in the T-antigen, which is involved in the degradation of p53 plumbagin-induced p21 up-regulation in MEF-polb (Sobol et al., 2000). In previous studies, it has been and MEF-polbKO cell lines. Two separate double-

Oncogene p21-mediated decrease in long-patch BER in plumbagin-treated cells AS Jaiswal et al 5915 stranded linear oligonucleotides containing either U MEF-polbKO cell extracts was slightly reduced by residue (U-DNA) or tetrahydrofuran (F) (F-DNA) at treatment with plumbagin, as there was a minor change the AP site were used as substrates in this assay. The in the 63-mer repaired products (Figure 2a, compare substrate containing U can be repaired by SP-BER, the lanes 1 with 2 and 3 with 4). predominant cellular pathway, whereas the substrate Since plumbagin treatment showed a minor effect on containing F can only be repaired by LP-BER. In our SP-BER, we focused our studies on LP-BER. The LP- assay system, the DNA repair activities associated with BER activity determined with the F-DNA substrate different BER proteins is measured by determining slightly decreased in extracts from plumbagin-treated in levels of 23-mer incision product formed by APE, 1 nt comparison with untreated MEF-polb cell lines (Figure gap-filling product formed by polb, strand-displace- 2a, compare lane 5 with 6). On the other hand, the LP- ment products formed by pold, e or b, and 63-mer BER activity significantly decreased in extracts from completely repaired product. The SP-BER activity as plumbagin-treated MEF-polbKO cell lines (Figure 2a, measured by repair of U-DNA in both MEF-polb or compare lane 7 with 8). Our results also show that the polb-dependent LP-BER activity is higher than polb- independent activity in MEF-polb cell line (Figure 2a, compare lane 5 with 7). The low efficiency of polb- independent LP-BER may be explained partly by the low efficiency of PCNA-dependent BER on the linear DNA (Biade et al., 1998), because ring-shaped PCNA trimer rapidly diffuses off ends of linear DNA. Interestingly, even though the overall level of SP- BER was unaffected by plumbagin treatment, the distribution of products did change. The polb- dependent (MEF-polb cell line) 1 nt gap-filling product increased with plumbagin treatment on both U- and F- DNA substrates (Figure 2a, compare lane 1 with 2 and 5 with 6). This accumulation of 1 nt gap-filling product in the MEF-polb cell extracts may be due to impaired dRP lyase or DNA ligase activity for the U-DNA substrate, and it may be due to the impaired activities of either FEN 1 or DNA ligase for the F-DNA substrate. The APE-mediated incision of the F-DNA substrate in MEF-polbKO cell extracts was not significantly different in untreated and plumbagin- treated systems (Figure 2a, compare lane 3 with 4 and 7 with 8 for 23-mer incision products). These results indicate that the decreased LP-BER in MEF- polbKO cells after plumbagin treatment affects steps following APE incision.

p21 mimics polb-dependent and -independent LP-BER activity in plumbagin-treated MEF cell lines The results from the Western blot analysis (Figure 1) show that p21 levels, not the polb or PCNA levels, in MEF-polb and MEF-polbKO cell lines significantly increased, suggesting that this increase is responsible for the decreased LP-BER activity (Figure 2). Furthermore, based on the known interaction of p21 Figure 2 Characterization of polb-dependent and -independent repair of U-DNA and F-DNA in MEF cell extracts. The cell ex- with PCNA (reviewed by Dotto, 2000), we hypothe- tracts from plumbagin untreated and treated MEF-polb and sized that the increased level of p21 is responsible for MEF-polbKO cell lines were prepared for BER assays as de- the decreased LP-BER through PCNA inhibition in scribed in Materials and methods. Photographs of the autoradio- both polb-dependent and polb-independent pathways. grams in (a) are representative of three independent experiments. If this is the case, then the addition of purified p21 Data shown in (b) are mean+s.e. of three experiments. The com- parative P-value of the student t-test is given in the graph. U- protein to extracts from untreated cells should mimic DNA, 63-mer double-stranded linear oligonucleotide containing the decreased LP-BER activity seen in plumbagin- U residue at position 24; F-DNA, 63-mer double-stranded linear treated cells. To test this hypothesis, LP-BER assays oligonucleotide containing tetrahydrofuran, an abasic analog of were supplemented with GST-fusion proteins contain- the natural AP site DNA at position 24. Fully ligated 63-mer products, 1 nt gap-filling product, and 23-mer-incision product are ing full-length p21 (p21F, amino acids 1 – 164) the N- shown with arrows. The strand-displacement products are shown terminal end of p21 (p21N, amino acids 1 – 82), and the within a bracket C-terminal end of p21 (p21C, amino acids 76 – 164).

Oncogene p21-mediated decrease in long-patch BER in plumbagin-treated cells AS Jaiswal et al 5916 Addition of either p21F or p21C proteins caused a protein is not always 100% biologically active. There- decrease in LP-BER of the F-DNA in a dose- fore, in our experiments, the effect of bacterially dependent manner in both MEF-polb and MEF- expressed p21 proteins on BER activity was determined polbKO cell lines (Figure 3). Because the C-terminal by the addition of different concentrations in the domain of p21 is known to bind PCNA (reviewed by reaction mixture. Based upon a semi-quantitative Dotto, 2000), our results suggest that p21 blocks LP- estimate, it was expected that approximately 2.5-fold BER through interaction with PCNA. A concentration higher concentrations of the bacterially expressed p21 of 500 ng of p21F or p21C proteins mimicked the proteins would be required to mimic the plumbagin- effects of plumbagin on LP-BER in both MEF-polb induced p21 protein levels on the BER activity in the and MEF-polbKO cell lines (Figure 3, compare lane 2 MEF cells (data not shown). The addition of p21F with 6 and 7, and lane 13 with 16 and 19, respectively). protein resulted in a decreased level of 63-mer ligation In these experiments, precise levels of endogenous and an increased level of 23-mer incision products. In plumbagin-induced p21 are difficulte to accurately contrast, the p21N protein had no effect on LP-BER in quantitate using the bacterially expressed p21 as a MEF-polb cells, but significantly decreased LP-BER in reference due to the fact that the bacterially expressed MEF-polbKO cells (Figure 3, compare lane 1 with 9 – 11, and lane 12 with 20 – 22, respectively). Given that the N-terminal domain of p21 is not known to interact with PCNA, its effect on LP-BER could be due to interaction with other BER proteins in MEF-polbKO cells. Currently, this effect is still not clear. Strand- displacement synthesis products accumulated in extracts from MEF-polb cells supplemented with p21F or p21C proteins. On the other hand, the level of strand-displacement synthesis products was decreased in MEF-polbKO cells supplemented with p21F or p21C proteins (Figure 3, compare lane 12 with 14 – 16 and with 17 – 19, respectively). There was no effect of the p21N protein on the strand-displacement synthesis with the extracts of the MEF-polbKO cells (Figure 3, compare lane 12 with 20 – 22). Taken together, these results suggest that p21/PCNA may block DNA ligase activity in polb-dependent pathway causing accumulation of the strand-displacement synthesis products. On the other hand, strand-displacement synthesis, FEN1 or DNA ligase activity may be blocked in polb-independent pathway. Since previous studies have shown that LP-BER is more efficient on covalently closed circular DNA (cccDNA) than on the linear DNA substrates (Biade et al., 1998; Li et al., 1995), we further validated our results of LP-BER of linear DNA with LP-BER of cccDNA. The pUC19 cccDNA containing random C Figure 3 Determining the role of p21 on LP-BER using F-DNA residues was deaminated by sodium bisulfite (Shortle as a substrate. The LP-BER reaction mixture with F-DNA was and Botstein, 1983). The resulting U-cccDNA was assembled with extracts of untreated and plumbagin-treated treated with uracil-DNA glycosylase (UDG) and then MEF-polb and MEF-polbKO cell lines. To examine the role of p21 in plumbagin-induced system, the purified p21F, p21C, or reduced with sodium borohydride to generate reduced p21N GST-fusion proteins were supplemented to the extracts AP sites on the cccDNA (R-cccDNA). This substrate is from untreated cells. The LP-BER activity was determined with suitable for LP-BER. To examine, whether p21 can 32 the repair of P-labeled F-DNA. (a) Shows a photograph of an inhibit LP-BER on R-cccDNA as on F-DNA, the LP- autoradiogram that is the representative of three independent experiments. Lanes 3 – 5, 6 – 8, and 9 – 11 were supplemented with BER assays were performed with R-cccDNA and increasing amounts (100, 250 and 500 ng) of purified p21F, extracts from MEF-polb and MEF-polbKO cell lines p21C, or p21N proteins, respectively, in the extracts of untreated supplemented with purified GST-p21 fusion proteins. MEF-polb cells. A similar addition of the purified p21 protein was In this assay system, the level of repaired R-cccDNA made to the extracts from untreated MEF-polbKO cells as shown decreased in extracts from both MEF-polb and MEF- in lanes 14 – 16, 17 – 19, and 20 – 22, respectively. Fully ligated 63- mer products, 1 nt gap-filling products, and 23-mer-incision pro- polbKO cell lines when treated with plumbagin (Figure ducts are shown with arrows. The strand-displacement products 4, compare lane 1 with 2 and lane 6 with 7). Addition are shown within a bracket. A quantitative analysis of the level of p21F and p21C proteins into the extract from of the 63-mer ligated products from (a) is given in (b). The data untreated cells blocked LP-BER on R-cccDNA in both are mean+s.e. of three experiments. * and § is significantly different than the extracts from untreated MEF-polb and MEF-polb MEF-polb and MEF-polbKO cell lines (Figure 4, KO cell lines, respectively. } is significantly different than compare lane 1 with 3 and 4 and lane 6 with 8 and 9, plumbagin-treated versus untreated MEF-polbKO cell extracts respectively). Addition of the p21N protein in the

Oncogene p21-mediated decrease in long-patch BER in plumbagin-treated cells AS Jaiswal et al 5917 both MEF-polb and MEF-polbKO cell lines (Figure 5a, compare lane 1 with 2 and 3, and lane 5 with 6 and 7, respectively). Strand-displacement products accumulated in extracts from MEF-polb cells supplemented with the p21F or the p21C proteins. Because pre- incision of F-DNA with purified APE did not restore LP-BER activity in cell extracts supplemented with p21F or p21C proteins, some step(s) following incision such as strand-displacement synthesis, flap cleavage (FEN1 activity), or DNA ligation must be affected.

DNA ligation step is impaired in p21-mediated LP-BER in MEF cell lines treated with plumbagin To examine whether DNA ligase activity is inhibited by p21 in MEF-polb and MEF-polbKO cell lines, LP- BER assays were performed with F-DNA substrate and extracts from untreated or plumbagin-treated cell lines. Before the addition of the F-DNA into the reaction mixture, different concentrations of T4 DNA ligase were added to cell extracts. The BER reaction was carried out for 30 min as before. The data demonstrate that supplementing plumbagin-treated cell extracts with purified T4 DNA ligase restored LP-BER Figure 4 Determining the role of p21 on LP-BER using R- cccDNA as a substrate. The LP-BER reaction conditions were activity (Figure 5b). There was no accumulation of same as described in Figure 3. A 500 ng of purified GST-fusion strand-displacement, 1 nt gap-filling or 23-mer incision proteins were added in the extracts from untreated and plumba- products that was seen in earlier experiments with gin-treated MEF-polb and MEF-polbKO cell lines. The LP- purified GST-p21 fusion proteins. Collectively, results BER activity was determined with the repair of R-cccDNA from these experiments suggest that plumbagin treat- (300 ng) by measuring the incorporation of [a-32P]-dCTP. Photo- graphs of the autoradiogram is representative of two independent ment induces p21 protein level that affects polb- experiments dependent and -independent LP-BER at the level of DNA ligation. extract showed no significant effect on blockage of LP- BER on R-cccDNA in MEF-polb cells but blocked it Discussion in MEF-polbKO cells (Figure 4, compare lane 1 with 5 and lane 6 with 10). Thus, in our assay system, the In the present study, we have examined the mechanism effect of p21 on LP-BER on F-DNA and R-cccDNA of p21 in plumbagin-induced cell cycle arrest and base are similar. These results suggest that both F-DNA and excision repair of AP site DNA in MEF cell lines with R-cccDNA are suitable substrates for determining the wild-type (MEF-polb)orpolb gene-knockout (MEF- role of p21 on LP-BER. polbKO) clones. Exposure of cells to plumbagin causes oxidative damage to DNA (Jamieson et al., 1994; Kato et al., 1994; Prieto-Alamo et al., 1993) that can be p21 inhibits LP-BER after APE step completion repaired by LP-BER pathway (reviewed by Ames, In the above experiments, accumulation of the 23-mer 1991; Beckman and Ames, 1997). Our results show an incision product was consistently observed in LP-BER induced level of p21, a S-G2/M phase arrest, and cell assays performed with extracts from either the death in both MEF-polb and MEF-polbKO cell lines plumbagin-treated or untreated MEF cells supplemen- exposed to plumbagin. In earlier studies, the role of ted with p21F or p21C proteins. To determine whether p21 is implicated in DNA damage-induced G1 and G2/ the accumulation of 23-mer incision products was the M phase arrest of cell cycle (Cayrol et al., 1998; Dulic result of a change in the APE incision step or a later et al., 1999; Li et al., 1994b). Recently, Ando et al. step in BER, the 63-mer F-DNA substrate was incised (2001) reported that p21 participates in G2/M phase by purified APE. LP-BER assays were then performed arrest in DLD1 human colon carcinoma cell line by on this incised F-DNA substrate using untreated or interacting with Cdc2/cyclin B1 complex and inhibiting plumbagin-treated MEF cell extracts with or without the Cdc2 kinase activity. These authors further report supplementation of p21F, p21C, or p21N proteins. that p21 plays a regulatory role in the maintenance of Results showed a decreased level of 63-mer ligated DNA damage-induced G2/M phase arrest by blocking products with p21F and p21C proteins in both MEF- the interaction of Cdc25C, a dual phosphatase, with polb and MEF-polbKO cell extracts. The decreased PCNA. Cdc25 is critical in the G2/M checkpoint of level of 63-mer ligated product correlated with the cells by controlling the phosphorylation status and increased level of 23-mer incised F-DNA substrate in subsequently, the kinase activity of Cdc2 (reviewed by

Oncogene p21-mediated decrease in long-patch BER in plumbagin-treated cells AS Jaiswal et al 5918

Figure 5 (a) APE does not affect p21-mediated LP-BER in MEF-polb and MEF-polbKO cell lines. The LP-BER assay was assembled with cell extracts and 32P-labeled F-DNA substrate pre-treated with APE for 60 min at 378C. The entire APE reaction mixture with DNA was incubated in the LP-BER assay. A 500 ng of the puriﬁed p21F, p21C, or p21N protein was added in the extracts from untreated MEF-polb and MEF-polbKO cell lines before the addition of the APE-treated F-DNA. Photographs of the autoradiograms are representative of three independent experiments. Lanes 9 and 10 show the APE-treated and untreated 32P-labeled F-DNA substrates, respectively. (b) Describes the role of DNA ligation on the p21-mediated LP-BER activity in MEF-polb and MEF-polbKO cell lines. The LP-BER assay was assembled with extracts from untreated and plumbagin-treated MEF cell lines supplemented with T4 DNA ligase (10, 25 and 50 units, respectively) and 32P-labeled F-DNA substrate. Photographs of the autoradiograms are representative of three independent experiments

Shivji et al., 1998). The inhibition of DNA synthesis While p21’s role in NER is unsettled, its role in other and cell cycle arrest at G1/S phase transition or in S DNA repair pathways including BER is also unclear. phase has been suggested due to inhibition of PCNA DNA re-synthesis during BER is mainly accomplished function and Cdk/cyclin kinase activity, respectively, by polb and to a lesser extent by pols d and e (reviewed by p21 (Rousseau et al., 1999). These possibilities will by Wilson, 1998). The repair of a U-DNA is completed be tested in our future studies. by the SP-BER and of a reduced or oxidized AP site Next, we examined how plumbagin-induced S-G2/M DNA (resistant to b-elimination) by the LP-BER. The checkpoint and how cell death is associated with the PCNA is involved in LP-BER through its interaction BER of the AP site DNA. Our target molecules were with FEN1 (Kim et al., 1998) and DNA ligase I (Levin p21 and PCNA, since both are involved in cell cycle et al., 2000). Thus, it is likely that p21 binds with arrest (Cayrol et al., 1998; Dulic et al., 1999; Li et al., PCNA, and in turn, impairs LP-BER either at the 1994b) and DNA repair (reviewed by Dotto, 2000; FEN1 or DNA ligase I step. We tested this possibility Jonsson and Hubscher, 1997; Kelman, 1997; Warbrick, in plumbagin-induced LP-BER using an AP site analog 2000; Wilson, 1998). The role of p21 in NER is containing the tetrahydrofuran (F)-modiﬁed linear reported to be through interaction with PCNA; DNA (F-DNA) or cccDNA containing reduced AP however, it is still unclear whether p21 inhibits (Li et site (R-cccDNA) as substrates. al., 1994a; Shivji et al., 1994, 1998) or creates resistance Plumbagin treatment increased p21 protein levels but to NER (Cooper et al., 1999; Pan et al., 1995). In a had no eﬀect on polb and PCNA levels in both MEF- recent study, it is suggested that p21 is required for polb (polb-dependent) and MEF-polbKO (polb-inde- NER at a step located downstream of the recruitment pendent) cell lines. LP-BER, not SP-BER, decreased in of PCNA to DNA repair sites (Stivala et al., 2001). both MEF-polb and MEF-polbKO cell lines treated

Oncogene p21-mediated decrease in long-patch BER in plumbagin-treated cells AS Jaiswal et al 5919 with plumbagin. However, the extent of inhibition of product formation to the same level of untreated cell LP-BER activity was more pronounced in MEF- extracts and caused no accumulation of strand- polbKO cell line. Nonetheless, the decreased LP-BER displacement and incision products. Thus, p21/PCNA correlated with the increased level of p21. Thus, we impairs polb-dependent LP-BER by blocking DNA postulate that p21 is involved in inhibition of LP-BER ligase activity. These findings support the earlier report in both polb-dependent and -independent pathways. in which PCNA is shown to stimulate DNA ligase To demonstrate that decreased LP-BER could result activity and play a key role in LP-BER (Levin et al., from plumbagin-induced levels of p21 in MEF cell 2000). Another explanation of p21/PCNA-mediated lines, our in vitro BER assay system was supplemented decrease in LP-BER activity in MEF-polb cell extracts with the purified p21 protein. We found that the can be given by the interaction of polb with DNA supplementation of the p21F or the p21C proteins into ligase I (Dimitriadis et al., 1998). Under these the extracts of untreated MEF-polb cells inhibited the circumstances, p21 binding to PCNA interferes with LP-BER of F-DNA or R-cccDNA substrates to similar DNA ligase I binding to PCNA. Since DNA ligase I level as observed with the extracts of plumbagin- also interacts with polb, p21 can interfere with polb- treated MEF-polb cell extracts. The purified p21 DNA ligase I interaction; thus, the completion of BER. protein-mediated decrease in the level of the 63-mer In a recent study, a similar view concerning the role of ligated product formation on F-DNA correlates with p21 in polb-dependent BER has been associated with the accumulation of the strand-displacement and DNA repair patch size of the natural AP site-mediated incision products. These results suggest that in polb- BER (Bennett et al., 2001). While this manuscript was dependent studies, the purified p21 (or plumbagin- in preparation, another report using double-stranded induced levels of endogenous p21) can impact multiple linear DNA substrate demonstrated that p21 inhibits steps of LP-BER pathway, i.e., APE, FEN1, or DNA LP-BER in vitro by disrupting PCNA-directed stimula- ligase. In the polb-independent LP-BER, the decreased tion of FEN1, DNA ligase I, and DNA polymerase d levels of 63-mer ligated products on F-DNA in (Tom et al., 2001). response to purified p21 supplementation or plumba- The plumbagin-induced activity of p21/PCNA on gin-induced levels of endogenous p21 were less effective polb-independent LP-BER is more favorable with the at the strand-displacement synthesis. Instead, the effect view reported by Klungland and Lindahl (1997). Since was more pronounced at the accumulation of the 23- the F-DNA is efficiently repaired in extracts from mer-incision product. These results suggest that in MEF-polb, as well as, in MEF-polbKO cell lines, a polb-independent LP-BER pathway, the p21 can have polb-independent pathway in MEF-polbKO cell line is its effect on the APE or the FEN1 activity. It can also more likely. This pathway is pols d and e-dependent have its effect on the pols d and e activity since there and requires PCNA for its activity (Klungland and are plenty of 23-mer-incision products but less strand- Lindahl, 1997). Based on our results, it appears that displacement products. Our findings ruled out the p21/PCNA interacts at the FEN1 and DNA ligase possible effect of p21/PCNA on the APE activity. In steps in the polb-independent LP-BER. No accumula- our experimental system, pre-treatment of F-DNA tion of the strand-displacement synthesis product and substrate with APE produced similar results on the the increased level of 23-mer incision products in the LP-BER as observed without APE treatment in both APE pre-treated BER system suggests that p21/PCNA MEF-polb and MEF-polbKO cell lines. This left three impairs FEN1 activity in plumbagin-treated MEF- possible steps of the LP-BER, i.e., strand-displacement polbKO cells. On the other hand, no accumulation of synthesis, FEN1 flap cleavage activity, and DNA the strand-displacement synthesis, the 23-mer incision ligation at which p21/PCNA may have delivered its product and the recovery of the 63-mer ligated product effect on the LP-BER after plumbagin treatment. after supplementation with T4 DNA ligase in extracts In earlier studies, the efficient excision of the F-DNA of both untreated and plumbagin-treated MEF-polb in polb-dependent LP-BER pathway is reported to be KO cells indicate that DNA ligase step is impaired in PCNA-dependent, but requires DNA synthesis (Kim et the MEF-polbKO cells. Thus, we conclude that both al., 1998). Contrary to this, other reports suggested a FEN1 and DNA ligase activity are affected by p21/ role of PCNA-stimulated FEN1 activity in polb- PCNA in response to plumbagin treatment in MEF- dependent LP-BER of the reduced AP site (generated polbKO cells. with sodium borohydride) or the oxidized AP site In summary, our results provide an evidence that (generated by ionizing radiation) (Klungland and plumbagin-induced levels of p21 arrests S-G2/M phase Lindahl, 1997). Our results support the first view of cell cycle followed by cell death, which is correlated proposed by Kim et al. (1998). In the polb-dependent with decreased activity of LP-BER. Our results suggest pathway of this proposal, the PCNA is involved at the that depending upon the polb level, p21/PCNA DNA ligation step and p21 can block PCNA- interaction with polb-dependent or -independent LP- dependent DNA ligase activity to impair the overall BER complexes is favored in the cells in response to LP-BER of the F-DNA. These results were further DNA damage. Furthermore, the sensitivity of cells to confirmed by supplementation of the T4 DNA ligase in DNA-damaging agents depends upon the type of the extracts of plumbagin-treated MEF-polb cell damage and efficiency of repair. For example, MEF- extracts. Our results show that the supplementation polb cells are less sensitive than MEF-polbKO cells in of T4 DNA ligase recovered the level of 63-mer ligated response to treatment with methylmethane sulphonate

Oncogene p21-mediated decrease in long-patch BER in plumbagin-treated cells AS Jaiswal et al 5920 (MMS) and methylnitosourea (MNU), which induce Preparation of R-cccDNA substrate Sn2-DNA alkylation that is repaired by polb-dependent LP-BER (Horton et al., 2000; Sobol et al., 1996). In To generate cccDNA with multiple U residues, we used our experimental system, treatment of these cell lines mutagenesis of pUC19 DNA with sodium bisulfite (Shortle with plumbagin show a similar sensitivity to cell cycle and Botstein, 1983). Gel-purified U-cccDNA will be treated with UDG. Then the AP site containing DNAs were treated arrest and cell death in MEF-polb than in MEF- with 0.1 M sodium borohydride for 30 min on ice. The R- polbKO cell lines. Since repair of oxidative DNA cccDNA was gel-purified and used in LP-BER. damage-induced by plumbagin is impaired in both cell lines, an accumulation of non-repaired or incompletely repaired DNA damage is acting as a trigger of BER assay apoptosis in MEF-polb and MEF-polbKO cell lines. To determine the ability of these substrates to undergo Thus, the DNA damage-induced p21 levels and DNA polb-dependent or -independent BER, the BER reaction repair capacity can be an important consideration for was assembled at 258C with MEF-polb and MEF-polbKO plumbagin, a chemotherapeutic agent-induced cell cycle cell line extracts with U-DNA or F-DNA substrates. The regulation and apoptosis of cancer cells. BER proficient cell extract was prepared by the procedure of Biade et al. (1998). The reaction mixture in a 20 ml volume contained: 50 mM HEPES, pH 7.9, 2 mM dithio- Materials and methods threitol (DTT), 10 mM MgCl2, 0.5 mM NAD, 5 mM diTris- phosphocreatine, 6 U of phosphocreatinekinase, 20 mM of Cell lines and treatment ATP, and 20 mM each of dNTPs. The BER was initiated by the addition of 5 – 10 ng of 32P-labeled U-DNA or F- The MEF-polb and MEF-polbKO cell lines were grown in DNA substrates (Biade et al., 1998; Prasad et al., 2001). Dulbecco’s modified Eagle’s medium (DMEM) supplemented When R-cccDNA was used as a substrate then the with 10% fetal bovine serum, 100 U/ml penicillin, 100 mg/ml reaction was initiated by the addition of [a-32P]dCTP. streptomycin and 80 mg/ml hygromycin B at 378C under a The BER reaction was stopped by adding 0.4% (w/v) SDS humidified atmosphere of 5% CO2 (Sobol et al., 2000). After (final concentration), 2 mg of proteinase K and 10 mgof cells reached 70% confluence, cells were treated with different carrier tRNA. They were incubated again for 30 min at concentrations of plumbagin (Sigma Chem. Co., St Louis, 378C. The DNA was recovered by chloroform/phenol MO, USA) for 10 h. extraction and ethanol precipitation and analysed on a 7 M urea-containing 12% (w/v) acrylamide gel. After FACS analysis electrophoresis, the gel was dried and the amount of radioactivity in the repaired DNA was quantified with A detergent and proteolytic enzyme-based technique was InstantImager (Packard Instrument Co., Meriden, CT, used for nuclear isolation and DNA content analysis of cells USA). in different phases of the cell cycle. After treatment with plumbagin, cells were harvested and processed for propidium iodide staining of the nuclei as described earlier (Vindelov Western blot analysis and Christensen, 1990). The cellular DNA content was To determine a change in the protein levels, a Western blot analysed by Becton-Dickinson FACScan flow cytometer. analysis was performed with proteins from extracts of The ranges for G0/G1,S,G2/M, and sub-G1 phase cells were untreated or plumbagin-treated cells as previously described established based upon their corresponding DNA content of (Narayan and Jaiswal, 1997). The following antibodies from histograms (Narayan et al., 2001). At least 10 000 cells per Santa Cruz Biotechnology Inc. (Santa Cruz, CA, USA) were sample were considered in the gated regions used for used in this study: p53, p21 and PCNA. Sam Wilson calculations. provided the anti-polb antibody (Sobol et al., 2000).

Synthesis of linear double-stranded DNA substrates Purification of GST-p21 The synthetic oligonucleotides were made on an ABI-392 The p21F, p21C, and p21N GST-fusion proteins were DNA synthesizer using standard b-cyanoethylphosphoramide purified as described by Fotedar et al. (1996). Briefly, the chemistry and reagents from Glen Research Corp. (Sterling, amplified bacterial cultures with over-expressed GST-p21 VA, USA). The nucleotide sequence of the 63-mer long proteins with isopropyl-1-thio-b-D-galactopyranoside (IPTG) oligonucleotide has a U residue at position 24 (5’-AGA- were pelleted and suspended in the ice-cold buffer containing: TGCCTGCAGCTGATGCGCUGTACGGATCCACGTGT- 20 mM Tris-HCl, pH 7.5, 1 mM EDTA, 100 mM KCl, 0.5% ACGGTACCGAGGGCGGGTCGACA-3’), which is re- NP-40, 1 mM DTT, 1 mM PMSF, 1.5% sarkosyl, 2 mg/ml of ferred as U-DNA. To examine the long-patch (LP) BER, benzamidine, 2 mg/ml of leupeptin and pepstatin A. Cells we synthesized a 63-mer AP site analog, 3-hydroxy-2- were sonicated for 30 s bursts, chilled on ice, and hydroxymethyltetrahydrofuran (F-DNA, 5’-TAGATGCCT- freeze – thawed for three times. Into the clarified lysates, GCAGCTGATGCGCFGTACGGATCCACGTGTACGGT- slurry of glutathione-Sepharose beads was added and ACCGAGGGCGGGTCGACA-3’), which is non-susceptible incubated for 1 h at 48C. After extensive washing of the to b-elimination. The gel-purified oligonucleotides were beads, the GST-p21 proteins were eluted in a buffer annealed to complimentary oligonucleotides with G residue containing: 50 mM Tris-HCl, pH 9.5, 20 mM reduced- opposite to the U residue. These oligonucleotides were glutathione and 120 mM NaCl. Samples were dialyzed for 32 labeled with g- P-ATP at the 5’ end and then annealed with 6 h with one change of the buffer containing: 20 mM Tris- the complementary oligonucleotides with G residue opposite HCl, pH 8, 1 mM EDTA, 10 mM NaCl, 1 mM DTT and 10% to the AP site. Gel purified double-stranded oligonucleotides glycerol. Purified proteins were quantified and stored at were used in BER assays. 7808C in small aliquots for future studies.

Oncogene p21-mediated decrease in long-patch BER in plumbagin-treated cells AS Jaiswal et al 5921 Abbreviations type and polb-gene-knockout clones; Dr Arun Fotedar AP, apurinic/apyrimidinic; APE, AP endonuclease; bp, base from the Cellular and Molecular Biology Program, Sidney pair; BER, base excision repair; cccDNA, covalently closed Kimmel Cancer Center (San Diego, CA, USA), for circular DNA; dRP, 2’-deoxyribose 5’-phosphate; dRPase, 2’- providing us the wild-type and mutant GST-p21 protein deoxyribose 5’-phosphatase; FEN1, flap endonuclease 1; LP, overexpression vectors. We thank Jessica A Salinas for long-patch; MEF, mouse embryonic fibroblast cell line; technical assistance and Nirupama Gupta for editorial PAGE, polyacrylamide gel electrophoresis; polb, DNA comments. The FACS analysis was performed in the FCC polymerase b; polbKO, polb-gene knocked out; PCNA, Laboratory of the ICBR of the University of Florida. proliferating cell nuclear antigen; SP, short-patch; UDG, These studies were supported in part by the grants awarded uracil-DNA glycosylase. to S Narayan from the National Cancer Institute, NIH (CA77721); the 2001-Research Opportunity Fund by the University of Florida (Gainesville, FL, USA); and the Acknowledgements 2001-Ralph E Powe Junior Faculty Enhancement Award We would like to thank Drs Rajendra Prasad and Samuel by the Oak Ridge Associated Universities (Oak Ridge, TN, H Wilson from NIEHS, Research Triangle Park, NC, for USA). The work of LB Bloom was supported by NSF providing the neutralizing anti-polb antibody and the grant MCB-9722356. mouse embryonic fibroblast (MEF) cell lines with wild-

References

Agrawal S, Agarwal ML, Chatterjee-Kishore M, Stark GR Frosina G, Fortini P, Rossi O, Carrozzino F, Raspaglio G, and Chisolm GM. (2002). Mol. Cell. Biol., 22, 1981 – 1992. Cox LS, Lane DP, Abbondandolo A and Dogliotti E. Ames BN. (1991). Jap. J. Cancer Res., 82, 1460 – 1461. (1996). J. Biol. Chem., 271, 9573 – 9578. Ando T, Kawabe T, Ohara H, Ducommun B, Itoh M and Hill BT, Scanlon KJ, Hansson J, Harstrick A, Pera M, Okamoto T. (2001). J. Biol. Chem., 276, 42971 – 42977. Fichtinger-Schepman AM and Shellard SA. (1994). Eur. J. Beckman KB and Ames BN. (1997). J. Biol. Chem., 272, Cancer, 30A, 832 – 837. 19633 – 19636. Holmquist GP. (1998). Mutat. Res., 400, 59 – 68. Bennett SE, Sung JS and Mosbaugh DW. (2001). J. Biol. Horton JK, Prasad R, Hou E and Wilson SH. (2000). J. Biol. Chem., 276, 42588 – 42600. Chem., 275, 2211 – 2218. Bjoras M, Luna L, Johnsen B, Hoff E, Haug T, Rognes T Jaiswal AS and Narayan S. (2002). Mutation Res., 500, 17 – and Seeberg E. (1997). EMBO J., 16, 6314 – 6322. 30. Bedford P, Fichtinger-Schepman AM, Shellard SA, Walker Jamieson DJ, Rivers SL and Stephen DW. (1994). Micro- MC, Masters JR and Hill BT. (1998). Cancer Res., 48, biology, 140, 3277 – 3283. 3019 – 3024. Jonsson ZO and Hubscher U. (1997). Bioessays, 19, 967 – Bhattacharyya N, Chen HC, Comhair S, Erzurum SC and 975. Banerjee S. (1999). DNA Cell Biol., 18, 549 – 554. Kato T, Watanabe M and Ohta T. (1994). Mutagenesis, 9, Biade S, Sobol RW, Wilson SH and Matsumoto Y. (1998). J. 245 – 251. Biol. Chem., 273, 898 – 902. Kelman Z. (1997). Oncogene, 14, 629 – 640. Boulaire J, Fotedar A and Fotedar R. (2000). Pathol. Biol. Kim K, Biade S and Matsumoto Y. (1998). J. Biol. Chem., (Paris), 48, 190 – 202. 273, 8842 – 8848. Brozmanova J, Dudas A and Henriques JA. (2001). Klungland A and Lindahl T. (1997). EMBO J., 16, 3341 – Neoplasma, 48, 85 – 93. 3348. Cadet J, Bourdat AG, D’Ham C, Duarte V, Gasparutto D, KoberleB,GrimaldiKA,SuntersA,HartleyJA,Kelland Romieu A and Ravanat JL. (2000). Mut. Res., 462, 121 – LR and Masters JR. (1997). Int. J. Cancer, 70, 551 – 555. 128. Krokan HE, Standal R and Slupphaug G. (1997). Biochem Canitrot Y, Hoffmann JS, Calsou P, Hayakawa H, Salles B J., 325, 1–16. and Cazaux C. (2000). FASEB J., 14, 1765 – 1774. Levin DS, McKenna AE, Motycka TA, Matsumoto Y and Cayrol C, Knibiehler M and Ducommun B. (1998). Tomkinson AE. (2000). Curr. Biol., 10, 919 – 922. Oncogene, 16, 311 – 320. Li R, Waga S, Hannon GJ, Beach D and Stillman B. (1994a). Cooper MP, Balajee AS and Bohr VA. (1999). Mol. Biol. Nature, 371, 534 – 537. Cell., 10, 2119 – 2129. Li Y, Jenkins CW, Nichols MA and Xiong Y. (1994b). Dianov GL, Prasad R, Wilson SH and Bohr VA. (1999). J. Oncogene, 9, 2261 – 2268. Biol. Chem., 274, 13741 – 13743. Li X, Li J, Harrington J, Lieber MR and Burgers PMJ. Dimitriadis EK, Prasad R, Vaske MK, Chen L, Tomkinson (1995). J. Biol. Chem., 270, 22109 – 22112. AE, Lewis MS and Wilson SH. (1998). J. Biol. Chem., 273, McDonald 3rd ER, Wu GS, Waldman T and El-Deiry WS. 20540 – 20550. (1996). Cancer Res., 56, 2250 – 2255. Dotto GP. (2000). Biochim. Biophys. Acta, 1471, M43 – M56. Nasmyth K. (1996). Science, 274, 1643 – 1645. Dulic V, Stein GH, Far DF and Reed SI. (1999). Mol. Cell. Narayan S and Jaiswal AS. (1997). J. Biol. Chem., 272, Biol., 19, 205 – 215. 30619 – 30622. Esposito F, Cuccovillo F, Russo L, Casella F, Russo T and Narayan S, Jaiswal AS, Multani AS and Pathak S. (2001). Cimino F. (1998). Cell Death Differ., 5, 940 – 945. Br. J. Cancer, 85, 898 – 901. FortiniP,PascucciB,ParlantiE,SobolRW,WilsonSHand Pan ZQ, Reardon JT, Li L, Flores-Rozas H, Legerski R, Dogliotti E. (1998). Biochemistry, 37, 3575 – 3580. Sancar A and Hurwitz J. (1995). J. Biol. Chem., 270, Fotedar R, Fitzgerald P, Rousselle T, Cannella D, Doree M, 22008 – 22016. Messier H and Fotedar A. (1996). Oncogene, 12, 2155 – Prasad R, Dianov GL, Bohr VA and Wilson SH. (2000). J. 2164. Biol. Chem., 275, 4460 – 4466.

Oncogene p21-mediated decrease in long-patch BER in plumbagin-treated cells AS Jaiswal et al 5922 Prasad R, Lavrik OI, Kim SJ, Kedar P, Yang XP, Vande Sobol RW, Prasad R, Evenski A, Baker A, Yang XP, Horton Berg BJ and Wilson SH. (2001). J. Biol. Chem., 276, JK and Wilson SH. (2000). Nature, 405, 807 – 810. 32411 – 32414. Srivastava DK, Berg BJ, Prasad R, Molina JT, Beard WA, Prieto-Alamo MJ, Abril N and Pueyo C. (1993). Carcinogen- Tomkinson AE and Wilson SH. (1998). J. Biol. Chem., esis, 14, 237 – 244. 273, 21203 – 21209. Prosperi E. (1997). Prog. Cell Cycle Res., 3, 193 – 210. Srivastava DK, Husain I, Arteaga CL and Wilson SH. Rousseau D, Cannella D, Boulaire J, Fitzgerald P, Fotedar (1999). Carcinogenesis, 20, 1049 – 1054. A and Fotedar R. (1999). Oncogene, 18, 4313 – 4325. Stivala LA, Riva F, Cazzalini O, Savio M and Prosperi E. Sark MW, Timmer-Bosscha H, Meijer C, Uges DR, Sluiter (2001). Oncogene, 20, 563 – 570. WJ, Peters WH, Mulder NH and de Vries EG. (1995). Br. Sugie S, Okamoto K, Rahman KM, Tanaka T, Kawai K, J. Cancer, 71, 684 – 690. Yamahara J and Mori H. (1998). Cancer Lett., 127, 177 – Seeberg E, Eide L and Bjoras M. (1995). Trends Biochem. 183. Sci., 20, 391 – 397. Taylor WR and Stark GR. (2001). Oncogene, 20, 1803 – 1815. Sheikh MS, Chen YQ, Smith ML and Fornace Jr AJ. (1997). Tom S, Ranalli TA, Podust VN and Bambara RA. (2001). J. Oncogene, 14, 1875 – 1882. Biol. Chem., 276, 48781 – 48789. Shivji MK, Ferrari E, Ball K, Hubscher U and Wood RD. Vindelov LL and Christensen IJ. (1990). Cytometry, 11, (1998). Oncogene, 17, 2827 – 2838. 753 – 770. Shivji MK, Grey SJ, Strausfeld UP, Wood RD and Blow JJ. Waga S and Stillman B. (1998). Mol. Cell. Biol., 18, 4177 – (1994). Curr. Biol., 4, 1062 – 1068. 4187. Shortle D and Botstein D. (1983). Methods Enzymol., 100, Wang L, Patel U, Ghosh L and Banerjee S. (1992). Cancer 457 – 468. Res., 52, 4824 – 4827. SobolRW,HortonJK,KuhnR,GuH,SinghalRK,Prasad Warbrick E. (2000). Bioessays, 22, 997 – 1006. R, Rajewsky K and Wilson SH. (1996). Nature, 379, 183 – Wilson SH. (1998). Mutat. Res., 407, 203 – 215. 186.

Oncogene