The tumor suppressor protein p53 stimulates the formation of ... - Nature

ª

Oncogene (2002) 21, 6614 – 6623 2002 Nature Publishing Group All rights reserved 0950 – 9232/02 $25.00 www.nature.com/onc

The tumor suppressor protein p53 stimulates the formation of the human topoisomerase I double cleavage complex in vitro Kent Søe1, Hella Hartmann1, Bernhard Schlott1, Tinna Stevnsner2 and Frank Grosse*,1 1

Institute of Molecular Biotechnology, Department of Biochemistry, Beutenbergstrasse 11, D-07745 Jena, Germany; 2Danish Center for Molecular Gerontology, Department of Molecular and Structural Biology, University of Aarhus, C.F. Møllers Alle´ Bldg. 130, DK-8000 Aarhus C, Denmark

Previous studies have shown that human topoisomerase I interacts directly with the tumor-suppressor protein p53. In the past few years it has repeatedly been suggested that topoisomerase I and p53 may play a joint role in the response to genotoxic stress. This led to the suggestion that p53 and human topoisomerase I may cooperate in the process of DNA repair and/or apoptosis. Recently we have demonstrated that a human topoisomerase I cleavage complex can be recognized by an additional topoisomerase I molecule and thereby form a so-called double cleavage complex. The double cleavage complex creates an about 13 nucleotides long single-stranded gap that may provide an entry site for recombinational repair events. Here we demonstrate that p53 stimulates both the DNA relaxation activity as well as the formation of the human topoisomerase I double cleavage complex by at least a factor of six. Stimulation of topoisomerase I activity by p53 is mediated via the central part of topoisomerase I. We also show that human, bovine, and murine p53 stimulate human topoisomerase I relaxation activity equally well. From these results it is conceivable that p53’s stimulatory activity on topoisomerase I may play a role in DNA recombination and repair as well as in apoptosis. Oncogene (2002) 21, 6614 – 6623. doi:10.1038/sj.onc. 1205912 Keywords: topoisomerase I; p53; DNA damage; repair; recombination; genomic instability Introduction Human topoisomerase I (htopoI) is a nuclear enzyme which is involved in several important pathways, such as transcription and replication (Yang et al., 1987; Stewart and Schutz, 1987; Snapka et al., 1988; Stewart et al., 1990), where the removal of positive supercoils is required for ongoing RNA and DNA synthesis. HtopoI can remove positive as well as negative supercoils by catalyzing a reversible transesterification reaction and thereby forming a covalent 3’-phospho-

*Correspondence: F Grosse; E-mail: [email protected] Received 27 February 2002; revised 24 July 2002; accepted 25 July 2002

tyrosyl bond between one DNA strand and an enzyme tyrosine residue while leaving a free 5’-hydroxyl terminus next to the cleavage site (Andersen et al., 1996). A family of alkaloids called camptothecins (CPT), which possess an antitumor effect, specifically inhibit the reversion of this reaction due to a specific interaction with htopoI. This leads to the stabilization of a reversible covalent complex between htopoI and the DNA backbone, a so-called ‘cleavage complex’, thereby introducing a long-lived nick into the DNA. The cytotoxic effect of these types of complexes is most likely due to a fragmentation of the genome that occurs when ongoing replication forks collide with covalently bound topoisomerases (Ryan et al., 1991; Squires et al., 1991). Different DNA lesions have been found to induce reversible htopoI cleavage complexes. These lesions include abasic sites, oxidative DNA damages, DNA alkylation or arylation events, gaps, base mismatches, and UV-photoproducts (Lanza et al., 1996; Pourquier et al., 1997a,b, 1999, 2001; Christiansen and Westergaard, 1999). Also, the nucleoside analog 1-b-Darabinofuranosylcytosine (Ara-C), a drug used for the treatment of acute leukemia, has been shown to stabilize the covalent complex (Pourquier et al., 2000). All the aforementioned lesions cause a local distortion of the DNA double helix and this is believed to promote the binding of these sites by htopoI and/or to slow down religation. UV lesions, DNA methylation, and Ara-C have also been shown to stabilize htopoI cleavage complexes in vivo (Subramanian et al., 1998; Mao et al., 2000; Pourquier et al., 2000, 2001). When such htopoI-DNA complexes are not repaired they may lead to genomic instability, e.g. through illegitimate recombination and block of transcription and/or replication (Ryan et al., 1991; Tsao et al., 1993; Megonigal et al., 1997; Wu and Liu, 1997). Mao et al. (2000) found that in a wt-p53 cell line htopoI cleavage complexes formed in vivo upon UV damage whereas these complexes did not form in a cell line carrying a p53 point mutation on amino acid 175. The tumor suppressor gene p53 has been found to contain mutations in more than 50% of all human tumors (Vogelstein and Kinzler, 1992). p53 is involved in regulatory processes of cell-cycle progression and in cellular responses following DNA damage (Lane, 1992). After exposure to stress signals p53 is activated

p53 stimulates htopoI double cleavage complex formation K Søe et al

6615

and stabilized through posttranscriptional mechanisms (reviewed by Jayaraman and Prives, 1999). Through its sequence-specific DNA binding and transactivation activity p53 modifies the expression of several proteins that are involved in cell cycle arrest or the onset of apoptosis. However, it is also known that the transactivator activity is not the only pathway in which p53 regulates apoptosis and cell cycle arrest. We have recently shown that a htopoI cleavage complex in vitro attracts a second htopoI molecule that introduces another cut in the close proximity upstream from the original cleavage complex (Søe et al., 2001). The formation of this double cleavage complex is mediated through direct protein – protein interaction between the two htopoI molecules. Frequently occurring double cleavage complexes may lead to apoptosis, however, as a rather rare event the second htopoI cleavage reaction could also function to initiate recombinational DNA repair (Søe et al., 2001). The data of Mao et al. (2000) suggest that htopoI cleavage complexes and maybe even double cleavage complexes form on UV irradiated DNA in a p53 dependent manner in the living cell. Thus, it was of particular interest for us to investigate if p53 is involved in the formation of htopoI double cleavage complexes. It has been previously shown that p53 and htopoI physically interact (Gobert et al., 1996, Albor et al., 1998, Mao et al., 2000). Here we confirm that p53 stimulates the relaxation activity of htopoI in vitro, and we show that the interaction is mediated through the central part of htopoI. We also demonstrate that p53 specifically and significantly stimulates the formation of htopoI double cleavage complexes in vitro. We conclude that double cleavage complex formation may function as a switch between apoptosis and DNA repair. Results p53 stimulates the relaxation activity of htopoI in vitro It has previously been shown that human and murine p53 stimulate the htopoI mediated relaxation of supercoiled plasmid DNA in vitro (Gobert et al., 1996, 1999; Albor et al., 1998). As shown in Figure 1a we confirm these results and show that the stimulation of htopoI by murine His-tagged p53 (mHis-p53) is concentrationdependent (compare lane 2 with lanes 3 – 6). BSA showed no stimulatory effect (lane 7), and the mHisp53 preparation contained no detectable topoisomerase I or endonuclease activity (lane 8). For this experiment we had to use low amounts (3 ng) of htopoI in order to see the stimulation clearly. If higher amounts of htopoI were used the turn-over rate of the substrate was so fast that no further stimulation could be detected (data not shown). When comparing lanes 3 – 5 (Figure 1b, upper and lower panel) it is seen that p53 from three different species, i.e. human, bovine and murine, was able to stimulate htopoI equally well. In the lower panel (Figure 1b) it is seen that this stimulation is not due

Figure 1 Human, bovine, and murine His-p53 stimulate the relaxation activity of htopoI equally well. (a) 3 ng htopoI were incubated with 1 mg pUC19 DNA in the presence or absence of mHis-p53 for 5 min at 378C. Lane 1, no htopoI, lane 2, htopoI, lanes 3 – 7, htopoI and 50, 75, 140, 210 ng mHis-p53, and 210 ng BSA respectively, lane 8, 140 ng mHis-p53. (b) 1 ng htopoI was incubated with or without 300 ng p53 for 5 min at 308C. Lane 1, no htopoI, 2, htopoI, lanes 3 – 5, htopoI and hp53, bp53, and mHis-p53 respectively. r., relaxed, sc., supercoiled, - etbr, electrophoresis performed in the absence of ethidium bromide, + etbr, electrophoresis performed in the presence of 0.5 mg/ml ethidium bromide. Plasmid DNA was separated on a 0.8% agarose gel and subsequently stained with ethidium bromide

to an increased appearance of nicked plasmid due to a possible contaminant in the p53 preparation. This shows that there is a high degree of conservation for the ability to stimulate topoisomerase I activity among p53 from the three species tested. The similar behavior of p53 from three different species may be due to the high degree of conservation of the sequence sufficient for htopoI activation, namely Oncogene


6616

amino acids 302 – 320 for the human p53 (Gobert et al., 1999). A sequence homology comparison revealed that this region is well conserved: The bovine p53 has a 90% sequence identity and 95% sequence similarity and the murine p53 still shows a 70% identity and 85% similarity compared to the human p53 sequence of this region (comparison was done using Clustal W). p53 interacts specifically with the central part of htopoI It has been published that a peptide comprising the amino acids 302 – 320 of p53 was sufficient to stimulate the relaxation activity of htopoI (Gobert et al., 1999). However, it was not yet known which part of htopoI bound to p53. We therefore performed a partial digest of htopoI using subtilisin (Figure 2) and by Far-Western analysis investigated with which domain p53 interacts. The partial digest led to the removal of the highly charged N-terminus, while the C-terminus and the core-domain were separated from each other (see also Stewart et al., 1996) (Figure 2a, lane 2 and Figure 2b, lane 1). The bands running just below the full-length htopoI represent degradation products with a missing N-terminus. When comparing the Coomassie stained gel with the Far-Western blot shown in Figure 2a a strong interaction between fulllength htopoI (band marked ‘a’) and p53 can be detected (lane 3), whereas no interaction with BSA was seen (lane 1). In Figure 2a, b, lanes 2 and 1, respectively, it is shown that this interaction was mediated solely through the core-domain of htopoI since the C-terminal domain (band marked ‘c’) showed no interaction and the N-terminal domain was removed (Figure 2b lane 1). This is also supported by the observation that htopoI degradation products without an N-terminus interact with p53 (Figure 2a, b lanes 3 and 2, respectively). N-terminal sequencing demonstrated that the N-terminus of the core-domain started with Asp156 (band marked ‘b’ seen in Figure 2a, lane 2 and Figure 2b, lane 1). FarWestern analysis using human or bovine p53 also showed a clear interaction with htopoI (data not shown). It was also verified that the p53 antibody did not bind unspecifically to htopoI (Figure 2c). Thus, p53 interacts with htopoI within amino acids 156 – 653 since Stewart et al. (1996) found that subtilisin cleaved the linker-domain after amino acid 653 under the same conditions used here. Human DNA topoisomerase I without an N-terminus can be stimulated by murine p53 The Far-Western analysis shown in Figure 2 revealed that p53 only interacted with the core-domain and the most proximal part of the N-terminal domain. However, since htopoI in the Far-Western blot may be partially denatured we wished to verify this result. To this end we subjected htopoI to a partial digest with subtilisin as it was done for the Far-Western analysis. The result of the digest is shown in Figure 3a. HtopoI was incubated at room temperature in the absence Oncogene

(lane 1) or presence of subtilisin (lanes 2 and 3), leading to fragments of approximately 80 and 60 kDa (lane 2). The 80 kDa fragment II represents htopoI where the major part of the N-terminal domain has been removed. The 60 kDa fragment III represents the core domain where a second cut took place in the linker domain and thereby separated the C-terminal domain (which runs out of the gel) from the core. In lane 3 htopoI is almost entirely converted to the 60 kDa fragment. It is known that although most of the N-terminus has been removed and the linker has been broken the enzyme still retains its activity (Stewart et al., 1996, Søe et al., 2001). Now we addressed the question whether or not the digested htopoI still could be stimulated by p53. As expected the full-length htopoI was strongly stimulated by mHis-p53 (Figure 3b, compare lanes 3 and 4). Interestingly, the partially digested htopoI also was strongly stimulated (compare lane 5 with 6 and lane 7 with 8). Since the degraded htopoI fraction used for the experiments shown in lanes 7 and 8 exclusively consisted of separated core and C-terminal domains (see Figure 3a, lane 3) we conclude that the core and C-terminal domains are sufficient for stimulation by p53 and that the linker is not essential. Thus, the results shown in Figure 3 support the results of the Far-Western analysis shown in Figure 2, namely that p53 interacts with htopoI within amino acids 156 – 653. Murine p53 stimulates the double cleavage of htopoI We have previously shown that a htopoI cleavage complex can be recognized by a second htopoI molecule and thereby introduce a second cut primarily about 13 nucleotides (nts), but also 15 or 17 nts upstream from the suicide cleavage site (see Figure 4a) (Søe et al., 2001). We called this phenomenon a htopoI (‘double cleavage’). Since p53 stimulates the activity of and forms a physical complex with htopoI (as shown above) we investigated whether p53 also was able to influence the htopoI double cleavage reaction. For this purpose we used the radioactively labeled ‘suicide substrate’ L193s depicted in Figure 4b (Søe et al., 2001). Cleavage of this substrate by htopoI results in an irreversible trapping of the htopoI-DNA cleavage complex. HtopoI was incubated with the suicide substrate and titrated with increasing concentrations of mHis-p53. From Figure 5a it can be seen that p53 had no significant effect on the first, i.e. the suicide cleavage step (compare lane 3 with lanes 4 – 8, see also Figure 5b). However, if lower concentrations of htopoI were used a small stimulation of the suicide cleavage reaction by mHis-p53 could be detected (data not shown). To simplify the interpretation of our investigations we used reaction conditions where the suicide cleavage efficiency was at a maximum and therefore no stimulation by p53 is detected as it can be seen in Figure 5a. Thus, p53 could not increase the binding or the turnover rate of htopoI on its specific substrate under the conditions used here. In striking contrast to


6617

Figure 2 p53 binds to the central part of htopoI. (a) A 7.5% SDS polyacrylamide gel was loaded with 11 mg (each) of BSA (lane 1), subtilisin-treated htopoI (lane 2), and htopoI (lane 3) and subjected to a Far-Western analysis as described in Materials and methods. In parallel a 7.5% SDS polyacrylamide gel was loaded with 4 mg (each) of BSA (lane 1), subtilisin-treated htopoI (lane 2), and htopoI (lane 3) as well as marker proteins (M). This gel was stained with Coomassie brilliant blue. (b) A 13.75% SDS – PAGE was casted and subjected to a Far-Western analysis. The gel was loaded with 11 mg of subtilisin-treated htopoI (lane 1) and 11 mg untreated htopoI. The Far-Western analysis was performed with murine His-p53 in solution. p53 bound to a target protein was detected with a polyclonal p53 antibody and an ECL assay as described in Materials and methods. An analogous gel was loaded with 4 mg (each) subtilisin-treated htopoI (lane 1), htopoI (lane 2), and a protein size marker (M). This gel was stained with Coomassie brilliant blue. Full-length htopoI is marked as (a), the core-domain as (b), and the C-terminal domain as (c). (c) 10% SDS – PAGE loaded with 4 mg BSA (lane 1) and 4 mg htopoI (lane 2) and stained with Coomassie brilliant blue or 11 mg BSA (lane 1) and 11 mg htopoI (lane 2) were blotted and incubated with p53 antibody alone Oncogene


6618

Figure 3 Subtilisin treated htopoI is also stimulated by p53. (a) A 7.5% SDS polyacrylamide gel was loaded with htopoI (lane 1), htopoI treated with 60 ng subtilisin (lane 2), htopoI treated with 100 ng subtilisin (lane 3). After electrophoresis the gel was stained with Coomassie brilliant blue. (b) An 0.8% agarose gel. 1 mg supercoiled plasmid DNA was incubated with 10 ng htopoI and/or 300 ng mHis-p53 for 5 min at 308C as follows: 1 mg plasmid only (lane 1), mHis-p53 (lane 2), htopoI+BSA (lane 3), htopoI+mHis-p53 (lane 4), htopoI+60 ng subtilisin+BSA (lane 5), htopoI+60 ng subtilisin+mHis-p53 (lane 6), htopoI+100 ng subtilisin+BSA (lane 7), htopoI+100 ng subtilisin+mHis-p53 (lane 8). Full-length htopoI is marked as (I), htopoI lacking the distal part of the N-terminus is indicated as (II), the core-domain is denoted as (III), sc stands for supercoiled, and r for relaxed DNA. The gel was run in the presence of 0.5 mg/ml ethidium bromide

this, the second ‘rescue’ cut that was performed by another htopoI molecule next to the irreversibly bound htopoI was efficiently stimulated by p53. When the DNAs from the same samples as shown in Figure 5a were treated with proteinase and analysed on a 14% sequencing gel a dramatic increase in double cleavage by htopoI was detected with increasing p53 concentration (Figure 5c, compare lane 3 with lanes 4 – 8). Hence, p53 specifically stimulates a rescue or repair reaction mediated by htopoI, at least in vitro. Oncogene

Figure 4 Schematic representation of the htopoI double cleavage complex and the function of the suicide substrate L193s. (a) The htopoI double cleavage complex arises after a stable cleavage complex (suicide htopoI) has been formed. The initially bound htopoI is then recognized by a second htopoI molecule (A, b, or c), which cleaves the DNA at the indicated distances to the suicide cleavage site. Thereby a double cleavage complex is formed. (b) The DNA substrate consists of a 193 nts long DNA which carries a 32P label at position 15 counting from the 3’ end (indicated by a pentagon). This oligomer folds back on itself forming a 3 nts loop at either end and a 3 nts overhang at the center of the substrate. Here is a htopoI binding sequence located (bold lines) and the htopoI incision takes place at the indicated position (arrow). Upon cleavage a 3-nts oligomer is released and the 5’ overhang closes the gap. This 5’ end is phosphorylated and therefore prevents the religation of htopoI and the enzyme is irreversibly trapped

A quantification of p53’s stimulation of the double cleavage reaction ‘A’ revealed that p53 stimulated the double cleavage step up to sixfold whereas BSA had no effect (Figure 5d). Interestingly, an optimum was reached at around 800 ng p53, whereas the stimulation declined at higher concentrations of p53. The optimum was reached at a molar ratio of approximately 1 to 4 between htopoI and the p53 monomer. Since p53 preferentially forms a tetramer (Stu¨rzbecher et al., 1992; Friedman et al., 1993) and since an optimum was found when htopoI and tetrameric p53 were in a 1 : 1 ratio we conclude that tetrameric p53 is needed for an optimal stimulation. There are at least two possible ways for p53 to stimulate the htopoI double cleavage reaction. First, the stimulatory effect may be achieved by a comparable mechanism as that described for CPT, i.e. increasing


6619

Figure 5 Murine p53 stimulates the double cleavage of htopoI. (a) Autoradiogram of a 7.5% SDS polyacrylamide gel showing the extent of suicidally cleaved substrate in the presence or absence of p53 as described in Materials and methods. Lane 1, no htopoI, lane 2, 800 ng mHis-p53, lane 3, 400 ng htopoI, lanes 4 – 9, 400 ng htopoI and 100, 200, 400, 800, 1600 ng mHis-p53, and 1600 ng BSA respectively. (b) Comparison of the influence of mHis-p53 on the htopoI suicide cleavage as compiled from 3 to 5 experiments as shown in A. (c) Autoradiogram of a 14% denaturing polyacrylamide gel. Loading was as in (a). (d) Comparison of the influence of mHis-p53 on the htopoI double cleavage reaction as compiled from 3 to 5 experiments as shown in (c). (e) Effect of mHis-p53 on the double cleavage A in the presence of 10 mM CPT. Loading was as in (a) Oncogene


6620

the half-life of the htopoI cleavage complex. If so, p53 would not be expected to have a significant effect in the presence of CPT. Second, p53 may directly stimulate the formation of the htopoI double cleavage complex. In this case, it would be expected that incubation with p53 and CPT would result in an additive effect. Such an effect was visible when increasing amounts of p53 were incubated with htopoI in the presence of 10 mM CPT (compare Figure 5d, lane 3 with lanes 4 – 8) whereas CPT showed no effect on the htopoI suicide cleavage (data not shown). Moreover, p53’s stimulatory effect on the double cleavage reaction was comparable in the absence and presence of CPT (compare with Figure 5c). Thus, our data suggest that p53 stimulates the double cleavage reaction by inducing the formation of more double cleavage complexes, without increasing their half-lives. BSA showed no effect in the presence or absence of CPT (Figure 5e, compare lane 3 with lane 9). Discussion It has been known for quite a while that the DNA relaxing enzyme topoisomerase I is stimulated by p53. p53’s activation of specific genes may require a stronger DNA relaxation activity in order to minimize torsional strain, which elegantly might be achieved by an active recruitment and subsequent enzymatic stimulation of htopoI by p53. However, both p53 and htopoI seem also to be involved in DNA repair events and in the control of the genomic stability. Our recent finding of htopoI’s potential role as a repair enzyme (Søe et al., 2001) led us to address the question whether the known interaction of p53 and htopoI also can be extended on the repair function of htopoI. Here, we have shown that human, bovine, as well as murine p53 stimulate the relaxation activity of htopoI equally well (Figure 1a,b). This is most likely achieved by protein – protein interactions between the amino acids 302 – 320 of p53 (Gobert et al., 1999) and the central part of htopoI (Figure 2a,b). The latter contains the last 54 amino acids of the N-terminus (amino acids 156 – 210) and the entire core-domain. We have previously demonstrated that a htopoI cleavage complex stimulates the formation of an additional cleavage complex just immediately upstream from the original cleavage site (Søe et al., 2001). Here we show that p53 has a significant stimulatory effect on the htopoI double cleavage reaction (Figure 5c,d). The approximately sixfold stimulation is likely to be due to an increased formation of complexes and not to alterations of their half-lives (Figure 5e). HtopoI interacts with further proteins, such as the SV-40 T-antigen, SF2, and the so-called topors. All these proteins bind to the N-terminal domain (amino acids 1 – 210) of htopoI (Haluska et al., 1998, 1999; Labourier et al., 1998), whereas p53 binds to the coredomain plus the proximal 54 amino acids of the Nterminus. Upon writing of this manuscript Mao et al. (2002) published that by using a mammalian twoOncogene

hybrid system the main interaction site between p53 and htopoI was attributed to the first 170 amino acids of htopoI. Considering the data of Mao et al. (2002) we suggest that the region of interaction may be situated within the amino acids 156 – 170. However, this still needs further experimental clarification. The binding of p53 to this sequence may explain the intranuclear redistribution of htopoI caused by p53 (Mao et al., 2002), since binding of p53 to this region may hide the nucleolar localization signal, which is located within the amino acids 1 – 170 of htopoI. On the other hand, we have shown by Far-Western and stimulation assays (Figures 2 and 3) that the amino acids 1 – 155 of htopoI are not essential for the interaction with p53. Particularly our stimulation studies suggest that an additional binding site for p53 may also be located within the core-domain of htopoI. The direct stimulation of htopoI by p53 is to our knowledge the only known protein – protein interaction where p53 significantly increases the activity of another enzyme. There are at least three possible roles for the stimulation of htopoI by p53. Firstly, Mao et al. (2000) discovered that htopoI-containing cleavage complexes form after UV-damage in vivo in a p53-dependent manner. This suggested that htopoI is part of a p53dependent damage response pathway. Subramanian et al. (1998) suggested that damage-induced htopoI cleavage complexes may be involved in nucleotide excision repair (NER). This was based on the finding that the formation of cleavage complexes was reduced in NER-deficient cell lines. However, other data argue against a role of htopoI in NER since CPT hardly inhibits NER in vitro or in vivo (Jones et al., 1991; Frosina and Rossi, 1992; Stevnsner and Bohr, 1993; Thielmann et al., 1993). Moreover, htopoI cleavage complexes may even hamper the recognition of a proximally located lesion by the NER enzymatic apparatus (Frosina and Rossi, 1992). Secondly, p53 is known to be a transcription activator of several genes in DNA damage response pathways (reviewed by May and May, 1999). It is therefore possible that p53 recruits htopoI to the activated genes and increases its relaxing activity to ensure a rapid and full-length transcript formation. If during transcription htopoI comes close to a lesion, its catalytic cycle is blocked and the enzyme becomes covalently attached to the DNA. Such a complex must be removed to avoid blockage of ongoing transcription and replication. The formation of a double cleavage complex and the subsequent creation of a 13-nts long single-stranded gap could be parts of such a repair pathway. Thirdly, p53 has been shown to specifically recognize DNA lesions (Lee et al., 1995; Degtyareva et al., 2001). Damage recognition by p53 alone or in combination with htopoI may help to ‘sense’ the extent of damage. If too many complexes form these may function as a signal for apoptosis. It is known that CPT-induced cleavage complexes lead to apoptosis (Ryan et al., 1991). However, if apoptosis is not induced the htopoI cleavage complexes must be removed before the cell continues its cell cycle.


6621

The repair of htopoI cleavage complexes may be achieved by the action of a tyrosyl-DNA phosphodiesterase (Yang et al., 1996; Pouliot et al., 1999, 2001). However, it seems that this pathway only removes a fraction of these complexes. Based on the presented here data we suggest that the induction of htopoI double cleavage complexes by p53 may represent an alternative repair pathway. We have found that approximately 40% of the double cleavages were not reversible (Søe et al., 2001; Stephan et al., submitted). These 40% may represent double cleavages where the original cleavage complex (suicide complex) has been released with the attached oligomer and thereby made a reversal of the second cleavage impossible. In this situation a gap is formed which may serve as an entry site for a 5’-OH group-containing DNA strand. After hybridization such a DNA strand may start a recombination event that uses the remaining htopoI molecule as a ligase (Stephan et al., submitted). Such a repair event could lead to a complete repair of the DNA lesion and the htopoI double cleavage complex. In agreement with this view is the recent finding of Mao et al. (2000) that damage-induced htopoI cleavage complexes may consist of two htopoI molecules in vivo. Gobert et al. (1999) showed that a mutant form of p53 (R273H) was constantly associated with and stimulated htopoI in vivo whereas wt p53 only stimulated htopoI during a brief period following cellular stress. In addition, Albor et al. (1998) found that two other p53 mutants N239S and R245S stimulated htopoI in vitro. Therefore, high concentrations of mutated p53, as observed in many tumor cell lines, may strongly increase legitimate and illegitimate recombination events by htopoI, particularly after DNA damage. This in turn could lead to the genomic instability observed in several p53 mutant cell lines. Due to the loss of the transcriptional function of p53 (Cao et al., 1997; Sigal and Rotter, 2000) these cells are no longer able to commit apoptosis. The loss of the apoptotic function and the strong increase of recombination events eventually may lead to cancer. A similar scenario was also recently suggested by Albor and Kulesz-Martin (2000). Materials and methods Recombinant baculoviruses The recombinant baculovirus encoding htopoI was a generous gift from Dr James Champoux, University of Washington, Seattle, USA. The recombinant baculoviruses encoding human wt p53 was as described in Bischoff et al. (1990), the murine His-tagged p53 was as described in Wang et al. (1993), and the bovine p53 was created using the Bacto-Bac System (Gibco). The bovine p53-encoding vector was a generous gift from F Dequiedt, Faculty of Agronomy, Gembloux, Belgium (Dequiedt et al., 1995).

Human wt, bovine wt, and murine His-tagged p53 were purified as follows. HiV insect cells (ITC Biotechnology, Heidelberg, Germany) were infected with either one of the above mentioned baculovirus clones for 46 h at 278C. Approximately 66108 cells were harvested by centrifugation. For the murine His-p53 the cells were resuspended in four pellet volumes lysis buffer I (20 mM Tris-HCl pH 8.0, 100 mM NaCl, 20 mM imidazole, 0.5% NP-40, 1 mM DTT, 1 mM PMSF, 1 mM aprotinin, and 1 mM leupeptin). For the purification of human and bovine wt p53 lysis buffer II was used (100 mM Tris-HCl pH 7.5, 100 mM NaCl, 5 mM KCl, 0.5 mM MgCl2, 1 mM EDTA, 0.1% NP-40, 1 mM DTT, 1 mM PMSF, 1 mM aprotinin). The cells were lysed using a Dounce Teflon homogenizor, spun at 48 000 g, and the resulting crude extract was loaded on a pre-equilibrated column. Crude extract from murine His-p53 infected cells was loaded onto a Ni-NTA agarose column (Qiagen) that was pre-equilibrated in wash buffer I (20 mM Tris-HCl pH 8, 100 mM KCl, 20 mM imidazole, 2.5 mM b-mercaptoethanol). The column was washed with the same buffer and eluted with elution buffer I (20 mM Tris-HCl pH 8, 100 mM KCl, 250 mM imidazole, 2.5 mM b-mercaptoethanol, 1 mM PMSF). The protein-rich fractions were identified according to Bradford (1976), pooled, and loaded on a phosphocellulose column (P11, Whatmann) pre-equilibrated with wash buffer II (100 mM potassium phosphate pH 7.8, 1 mM EDTA, and 2 mM b-mercaptoethanol) using FPLC (AmershamPharmacia). The column was washed, and the protein was eluted with elution buffer II (300 mM potassium phosphate pH 7.8, 1 mM EDTA, and 2 mM b-mercaptoethanol). The fractions were analysed according to Bradford (1976) and additionally by 10% SDS – PAGE. p53-containing fractions were dialyzed overnight against storage buffer (20 mM HEPES-KOH pH 7.8, 50 mM NaCl, 0.1 mM EDTA, 0.5 mM DTT, and 10% glycerol) and stored at 48C until use. For the purification of human and bovine wt p53 the crude extract was loaded on a phosphocellulose column. The column was washed with wash buffer II, and then eluted with elution buffer II; the protein-containing fractions were analysed as described. p53-containing fractions were pooled, adjusted to 0.5 M ammonium sulfate, and loaded on a HiLoad phenyl-Sepharose column (AmershamPharmacia) pre-equilibrated in Wash Buffer III (50 mM KPi pH 7.8, 0.5 M (NH4)2SO4, 1 mM EDTA, 2 mM b-mercaptoethanol). The column was washed and eluted with elution buffer III (50 mM potassium phosphate pH 7.8, 1 mM EDTA, 2 mM bmercaptoethanol), and the fractions were examined as above. p53-containing fractions were diluted 1 to 3 in wash buffer IV (25 mM potassium phosphate pH 7.8, 1 mM EDTA, 2 mM bmercaptoethanol) and loaded on a HiTrap heparin-Sepharose column (AmershamPharmacia) pre-equilibrated in wash buffer IV. The column was washed and eluted with a linear gradient from 0 – 50% elution buffer IV (800 mM potassium phosphate pH 7.8, 1 mM EDTA, 2 mM b-mercaptoethanol). The eluate was analysed as above and diluted with wash buffer II to about 100 mM potassium phosphate. Thereafter, the protein was loaded on a phosphocellulose column as described above, and eluted with a linear gradient from 0 – 100% elution buffer II. The final samples were dialyzed overnight against storage buffer and stored at 48C until use. Relaxation assay

Preparation of recombinant protein Recombinant wt htopoI was expressed in the baculovirus system and purified as described previously (Søe et al., 2001).

The stimulation of htopoI’s relaxation activity by p53 (Figures 1 and 3) was assayed as follows. 1 mg supercoiled pUC19 plasmid DNA (purified according to Wood et al., Oncogene


6622

1995) was incubated with the indicated amounts of htopoI and the indicated amounts of recombinant human, bovine, or murine His-p53. The relaxation reaction was performed in 20 ml buffer A (19 mM HEPES-KOH pH 7.9, 50 mM KCl, 30 mM NaCl, 6 mM MgCl2, 0.5 mM EDTA, 1 mM DTT, 16% glycerol, and 15 mg/ml BSA) for 5 min at 378C or 308C (as indicated). The reaction was stopped by adding agarose loading buffer and heating for 5 min at 958C. Then the plasmids were analysed on a 0.8% agarose gel in the presence or absence of 0.5 mg/ml ethidium bromide (as indicated). Protease digestion of htopoI 200 mg/ml htopoI was incubated with 4 or 6 mg/ml subtilisin (Roche) as indicated. The digestion reaction was performed in 10 mM Tris-HCl pH 7.5, 1 mM EDTA, for 20 min at room temperature (258C), and stopped with 5 mM PMSF. Far-Western analysis 11 mg htopoI and 11 mg subtilisin-treated htopoI (as described above) were incubated in sample buffer (2.5% SDS, 2.5 mM Tris-HCl, pH 8.0, 100 mM DTT, 10% glycerol, 0.05% Pyronin T1) for 5 min at room temperature. 11 mg of BSA were treated the same way. Proteins were then separated by SDS – PAGE as indicated. Far-Western analysis was performed as described in Weisshart et al. (2000) using a murine His-p53 concentration of 12 mg/ml. Both the p53 primary polyclonal sheep antibody Ab-7 (Oncogene) and the HRP-conjugated anti-sheep secondary antibody (Sigma) were used in a 1 : 3000 dilution. Reactions with suicide substrate L193s The substrate L193s was produced by ligation of three oligomers in such a way that the 32P radioactive label was

placed at position 15 from the 3’ end as described previously (Søe et al., 2001). The stoichiometry between L193s and htopoI was estimated from Western blotting and Coomassie staining of suicide cleavage reactions to be about 1 – 10. 400 ng of htopoI was incubated with L193s (about 1500 Bq) and murine His-p53 as indicated for 10 min at 378C in buffer A. 10 mM CPT or 2 ml DMSO were added 2 min after the reactions were started. The reactions were stopped by adding SDS to a final concentration of 0.7%. An aliquot of 2 ml was taken from each sample and analysed on a 7.5% SDS – PAGE. The bands were quantified using a phosphorimager and the ImageQuant software (Storm 860, Molecular Dynamics, AmershamPharmacia). The remainder of each sample was incubated with 1 mg/ml proteinase K for 30 min at 378C. The DNA was precipitated with three sample volumes 0.6 M LiCl in absolute ethanol for at least 30 min at 48C. The pellet was redissolved in sequence gel loading buffer (40% formamide, 25 mM Tris-borate (pH. 8.3), 0.5 mM EDTA, 0.05% (w/v) xylene cyanol, 0.05% (w/v) bromophenol blue, heated to 958C for 5 min, and analysed on a 14% denaturing polyacrylamide gel (SequaGel, National Diagnostics). The gel was dried and analysed using a phosphorimager as above.

Acknowledgments We wish to thank the undergraduate students Michael Stilmann and Falk Remane for purification of htopoI used in the Far-Western analysis. This work was sponsored by the Deutsche Krebshilfe.

References Albor A and Kulesz-Martin M. (2000). DNA Alterations in Cancer Ehrlich M (ed). Natrick MA: Eaton Publishing, pp. 409 – 422. Albor A, Kaku S and Kulesz-Martin M. (1998). Cancer Res., 58, 2091 – 2094. Andersen AH, Bendixen C and Westergaard O. (1996). DNA Replication in Eukaryotic Cells DePamphilis ML (ed). Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press, pp. 587 – 617. Bischoff JR, Friedman PN, Marshak DR, Prives C and Beach D. (1990). Proc. Natl. Acad. Sci. USA, 87, 4766 – 4770. Bradford MM. (1976). Anal. Biochem., 72, 248 – 254. Cao Y, Gao Q, Wazer DE and Band V. (1997). Cancer Res., 57, 5584 – 5589. Christiansen K and Westergaard O. (1999). Biochim. Biophys. Acta., 1489, 249 – 262. Degtyareva N, Subramanian D and Griffith JD. (2001). J. Biol. Chem., 276, 8778 – 8784. Dequiedt F, Willems L, Burny A and Kettmann R. (1995). DNA Seq., 5, 261 – 264. Friedman PN, Chen X, Bargonetti J and Prives C. (1993). Proc. Natl. Acad. Sci. USA, 90, 3319 – 3323. Frosina G and Rossi O. (1992). Carcinogenesis, 13, 1371 – 1377.

Oncogene

Gobert C, Bracco L, Rossi F, Olivier M, Tazi J, Lavelle F, Larsen AK and Riou JF. (1996). Biochemistry, 35, 5778 – 5786. Gobert C, Skladanowski A and Larsen AK. (1999). Proc. Natl. Acad. Sci. USA, 96, 10355 – 10360. Haluska Jr P, Saleem A, Edwards TK and Rubin EH. (1998). Nucleic. Acids Res., 26, 1841 – 1847. Haluska Jr P, Saleem A, Rasheed Z, Ahmed F, Su EW, Liu LF and Rubin EH. (1999). Nucleic. Acids Res., 27, 2538 – 2544. Jayaraman L and Prives C. (1999). Cell. Mol. Life Sci., 55, 76 – 87. Jones JC, Stevnsner T, Mattern MR and Bohr VA. (1991). Mutat. Res., 255, 155 – 162. Labourier E, Rossi F, Gallouzi IE, Allemand E, Divita G and Tazi J. (1998). Nucleic. Acids Res., 26, 2955 – 2962. Lane DP. (1992). Nature, 358, 15 – 16. Lanza A, Tornaletti S, Rodolfo C, Scanavini MC and Pedrini AM. (1996). J. Biol. Chem., 271, 6978 – 6986. Lee S, Elenbaas B, Levine A and Griffith J. (1995). Cell, 81, 1013 – 1020. Mao Y, Mehl IR and Muller MT. (2002). Proc. Natl. Acad. Sci. USA, 99, 1235 – 1240. Mao Y, Okada S, Chang LS and Muller MT. (2000). Cancer Res., 60, 4538 – 4543.


6623

May P and May E. (1999). Oncogene, 18, 7621 – 7636. Megonigal MD, Fertala J and Bjornsti M-A. (1997). J. Biol. Chem., 272, 12801 – 12808. Pouliot JJ, Robertson CA and Nash HA. (2001). Genes Cells., 6, 677 – 687. Pouliot JJ, Yao KC, Robertson CA and Nash HA. (1999). Science, 286, 552 – 555. Pourquier P, Ueng L-M, Fertala J, Wang D, Park H-J, Essigman JM, Bjornsti M-A and Pommier Y. (1999). J. Biol. Chem., 274, 8516 – 8523. Pourquier P, Ueng L-M, Kohlhagen G, Mazumder A, Gupta M, Kohn KW and Pommier Y. (1997a). J. Biol. Chem., 272, 7792 – 7796. Pourquier P, Pilon AA, Kohlhagen G, Mazumder A, Sharma A and Pommier Y. (1997b). J. Biol. Chem., 272, 26441 – 26447. Pourquier P, Takebayashi Y, Urasaki Y, Gioffre C, Kohlhagen G and Pommier Y. (2000). Proc. Natl. Acad. Sci. USA, 97, 1885 – 1890. Pourquier P, Waltman JL, Urasaki Y, Loktionova NA, Pegg AE, Nitiss JL and Pommier Y. (2001). Cancer Res., 61, 53 – 58. Ryan AJ, Squires S, Strutt HL and Johnson RT. (1991). Nucleic. Acids Res., 19, 3295 – 3300. Sigal A and Rotter V. (2000). Cancer Res., 60, 6788 – 6793. Snapka RM, Powelson MA and Strayer JM. (1988). Mol. Cell. Biol., 8, 515 – 521. Søe K, Dianov G, Nasheuer H-P, Bohr VA, Grosse F and Stevnsner T. (2001). Nucleic. Acids Res., 29, 3195 – 3203. Squires S, Ryan AJ, Strutt HL, Smith PJ and Johnson RT. (1991). J. Cell. Sci., 100, 883 – 893. Stevnsner T and Bohr VA. (1993). Carcinogenesis, 14, 1841 – 1850.

Stewart AF, Herrera RE and Nordheim A. (1990). Cell, 60, 141 – 149. Stewart AF and Schutz G. (1987). Cell, 50, 1109 – 1117. Stewart L, Ireton GC and Champoux JJ. (1996). J. Biol. Chem., 271, 7602 – 7608. Stu¨rzbecher HW, Brain R, Addison C, Rudge K, Remm M, Grimaldi M, Keenan E and Jenkins JR. (1992). Oncogene, 7, 1513 – 1523. Subramanian D, Rosenstein BS and Muller MT. (1998). Cancer Res., 58, 976 – 984. Thielmann HW, Popanda O, Gersbach H and Gilberg F. (1993). Carcinogenesis, 14, 2341 – 2351. Tsao Y-P, Russo A, Nyamuswa G, Silber R and Liu LF. (1993). Cancer Res., 53, 5908 – 5914. Vogelstein B and Kinzler KW. (1992). Cell, 70, 523 – 526. Wang Y, Reed M, Wang P, Stenger JE, Mayr G, Anderson ME, Schwedes JF and Tegtmeyer P. (1993). Genes Dev., 7, 2575 – 2586. Weisshart K, Fo¨rster H, Kremmer E, Schlott B, Grosse F and Nasheuer HP. (2000). J. Biol. Chem., 275, 17328 – 17337. Wood RD, Biggerstaff M and Shivji MKK. (1995). Methods: A Companion to Methods Enzymol., 7, 163 – 175. Wu J and Liu LF. (1997). Nucleic. Acids Res., 25, 4181 – 4186. Yang SW, Burgin Jr AB, Huizenga BN, Robertson CA, Yao KC and Nash HA. (1996). Proc. Natl. Acad. Sci. USA, 93, 11534 – 11539. Yang L, Wold MS, Li JJ, Kelly TJ and Liu LF. (1987). Proc. Natl. Acad. Sci., 84, 950 – 954.

Oncogene