In vivo sequencing of camptothecin-induced topoisomerase I cleavage sites in human colon carcinoma cells
In vivo sequencing of camptothecin-induced topoisomerase I cleavage sites in human colon carcinoma cellsCorinne Pondarré+, Dirk Strumberg, Akira Fujimori, Rosimar Torres-León and Yves Pommier*
Laboratory of Molecular Pharmacology, Division of Basic Sciences, Bldg 37, Rm 5C25, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892-4255, USA
Received May 29, 1997;Revised and Accepted August 28, 1997
Camptothecin (CPT) is a specific topoisomerase I (top1) poison which traps top1 cleavable complexes; e.g. top1-linked DNA single-strand breaks with 5'-hydroxyl and 3'-top1 linked termini. CPT is also a potent anticancer agent and several of its derivatives have recently shown activity in the chemotherapy of solid tumors. Our aim was to apply the ligation-mediated polymerase chain reaction (LM-PCR) method to DNA extracted from CPT-treated cells in order to: (i) evaluate LM-PCR as a sensitive technique to detect in vivo CPT-induced cleavable complexes; (ii) investigate the frequency and distribution of CPT-induced DNA damage in vivo; and (iii) compare the distribution and intensity of cleavage sites in vivo and in vitro. This report describes a protocol allowing the sequencing of top1-mediated DNA strand breaks induced by CPT in the coding strand of the 18S rRNA gene of human colon carcinoma cells. CPT or its clinical derivatives, topotecan, CPT-11, SN-38, and 9-aminocamptothecin differed in their potency and exhibited differences in their DNA cleavage pattern, which is consistent with our previous in vitro studies [Tanizawa et al. (1995) Biochemistry,43, 7200-7206]. CPT-induced DNA cleavages induced in the presence of purified top1 were induced at the same sites in the human 18S rDNA. However, the relative intensity of the cleavages were different in vivo and in vitro. Because mammalian cells contain ~300 copies of the rDNA gene per genome, rDNA could be used to monitor CPT-induced DNA cleavage in different cell lines and possibly in tumor samples.
INTRODUCTION
Topoisomerase 1 (top1) has been identified as the intracellular target of a new class of anticancer drugs, camptothecin (CPT) and its derivatives topotecan (HycamtinR), CPT-11 (IrinotecanR) and 9-aminocamptothecin (1 ,2 ). The catalytic cycle of top 1 can be divided in four steps: (i) binding of the enzyme to DNA; (ii) cleavage of one strand with covalent bond to the 3' phosphate terminus of the nicked intermediate leaving a 5' OH group; (iii) complementary strand passage (or swivel) through the break; (iv) resealing and release of the enzyme [for recent reviews see (3 ,4 )]. CPT and its derivatives inhibit the top1 catalytic cycle by preventing the resealing of the enzyme-mediated DNA single strand-breaks. Trapping of the top1 cleavable complexes represents the initial event of drug-induced cell death and plays a critical role in determining the cell sensitivity to top 1 inhibitors (5 ,6 ). This is illustrated by a recent study showing that in a panel of seven human carcinoma cell lines, drug cytotoxicity is better correlated with the frequency of CPT-induced cleavable complexes than top1 mRNA or protein levels (7 ). Thus, monitoring the amount of DNA breaks is important to determine the in vivo effects of camptothecins.
The alkaline elution and SDS-KCl techniques are commonly used to measure the frequency of cleavable complexes in cell culture systems (8 -11 ). However, they both require radioactive labeling of the cellular DNA and therefore cannot be applied to quiescent cells or tumor samples. Also, these methods do not localize the drug-induced DNA damage. Prior to this report, CPT-induced DNA breaks had been sequenced in vitro in DNA oligonucleotides, SV40 and HIV-1 DNA and specific genes (12 -18 ), and mapped in vivo but not at the nucleotide resolution (19 -23 ).
A technique known as ligation-mediated polymerase chain reaction (LM-PCR) allows the amplification of specific DNA fragments of different lengths simultaneously. This is achieved by ligation of a double-stranded oligonucleotide linker to blunt-ended DNA fragments generated by DNA polymerase extension of a gene specific primer using a template DNA that has been chemically or enzymatically cleaved. Thus, all fragments have the same end and can be amplified by PCR using one pair of primers (24 ,25 ). The goal of the present study was to apply this method to the sequencing of CPT-induced DNA breaks in drug-treated cancer cells.
We selected the transcribed regions of ribosomal DNA (rDNA) because: (i) there are ~300 copies of human rRNA genes per genome (26 ); and (ii) top1 has been reported to be enriched within the transcribed region of the human rRNA genes (23 ,27 ,28 ). Figure 1 shows a human rRNA gene repeat with its 31 kb spacer and 13 kb transcribed region which contains the 18S, 5.8S and 28S ribosomal subunit genes. Such repeats are arranged in tandem in the nucleolar organizer regions of the five acrocentric chromosomes (26 ).
MATERIALS AND METHODS
Cell lines
The SW 620 and HT29 cell lines were provided by Dr Dominic Scudiero (NCI, Frederick, MD) and grown in monolayer cultures using RPMI 1640 supplemented with 5% heat inactivated fetal bovine serum, 2 mM glutamine, 100 U/ml penicillin and 100 µg/ml streptomycin. Both cell lines were cultured at 37oC in 5% CO2, trypsinized and passed 2-3 times per week.
Drugs, chemicals and enzymes
20-(S)-camptothecin lactone (camptothecin; CPT), 9-amino- 20(S)-camptothecin (9-aminocamptothecin; 9-AC) and topotecan (TPT) were obtained from the Drug Synthesis and Chemistry Branch, Division of Cancer Treatment, National Cancer Institute. CPT-11 (irinotecanR)and SN-38 (7-ethyl-10- hydroxy-20(S)-camptothecin) were kindly provided by Dr Kiyoshi Terada (Yakult Honsha Co., Tokyo, Japan). Drugs were dissolved in dimethyl sulfoxide at a concentration of 10 mM. Further dilutions were made in water immediately before use. VentR (exo-) DNA polymerase (2000 U/ml) and T4 polynucleotide kinase (10 U/µl) were purchased from New England Biolabs (Beverly, MA). T4 DNA ligase (4 U/µl) was obtained from Stratagene whereas Klenow enzyme (2 U/µl) was from Boehringer Mannheim. Recombinant human top 1 was purified from baculovirus-infected sf9 insect cells as described (29 ,30 ).
Drug treatments and genomic DNA preparation
Exponentially growing cells were treated with 0.1 µM or 10 µM concentration of each drug. Following 1 h drug treatments in culture medium at 37oC, cells were washed twice with ice-cold Hank's Balanced Salt Solution (HBSS) containing the same amount of drug to avoid reversal of top1 cleavable complexes (31 ). To examine reversal of drug-induced DNA cleavage, treated cells were washed in drug-free ice-cold HBSS twice and allowed to grow in drug-free RPMI 1640 medium for 30 min. Following treatment, cells were lysed in 10 mM Tris-HCl, 400 mM NaCl, 2 mM EDTA, 0.6% SDS, pH 8.2. After overnight incubation at 37oC in the presence of proteinase K (3 mg/ml), DNA was isolated by phenol-chloroform extractions (three times) followed by dialysis against a solution containing 50 mM Tris-HCl and 10 mM EDTA at pH 8.
Ligation-mediated PCR
Oligonucleotide primers and linker. The oligonucleotide primers and linker were purchased from The Midland Certified Reagent Company (Midland, TX). PCR primers were chosen from the published sequence of human ribosomal DNA complete repeating unit (DDBJ/EMBL/GenBank accession number U13369) with numbers referring to GenBank genomic positions. Three nested primers for the 18S rDNA upper strand were used (Figs 1 and 2 B): A1 for primer extension (corresponding to positions 5030-5050), A2 for PCR amplification (positions 5049-5074) and A3 for end-labeling extension (positions 5059-5080). The oligonucleotide linker consisted of a 25mer, L1 annealed to a 11mer L2 in 250 mM Tris-HCl, pH 7.7 (Fig 2 B).
CPT-induced DNA cleavage in 18S rDNA in the presence of purified top1
Primer A3 (Fig. 2 B) and B3 (5'-GCCGTGGGCCGACCCCGGCG, positions 5430-5410) were used to amplify a 372 bp A3-B3 fragment between position 5059 and 5430 (Fig. 1 ). All amplification reactions consisted of 0.1 µg of total control genomic DNA, appropriate primers (5 pmol each), the four deoxynucleotides (250 µM each), 1 U of Vent-DNA polymerase and its supplied buffer (New England Biolabs, Beverly, MA). Thirty PCR cycles were performed. Primers A3 and B3 were annealed at 68oC. The 5'-end labeling of the A3-B3 fragment was obtained by PCR using 5'-end-labeled primer B3. Substrate DNA was then incubated with or without CPT in 10 mM Tris-HCl, pH 7.5, 50 mM KCl, 5 mM MgCl2, 0.1 mM EDTA, and 15 µg/ml bovine serum albumin in the presence of recombinant human top1 at 37oC for 30 min. Reactions were stopped by adding SDS to a final concentration of 1% and digested with proteinase K (0.5 mg/ml for 1 h at 50oC).
DNA electrophoresis
DNA fragments were separated by electrophoresis in 7% polyacrylamide wedged gels. The gels were then autoradiographed or analysed using a PhosphoImager (Molecular Dynamics, Sunnyvale, CA). Cleavage sites were positioned by comparison with a purine ladder.
RESULTS
Detection of in vivo CPT-induced cleavable complexes using LM-PCR
Figure 2 describes the LM-PCR method and primers used to sequence the camptothecin-induced cleavage sites in the coding strand (transcribed; lower strand in GenBank) of the human 18S rDNA. A previous report indicated that top1 is preferentially associated with the coding strand of transcribed genes (22 ). LM-PCR was successfully applied to DNA extracted from both control and CPT-treated cells. Details are given in Materials and Methods for the LM-PCR protocol developed for the present experiments. Figure 3 illustrates the results of a typical experiment using 0.1 µg of genomic template DNA (equivalent to ~104 cells). No cleavage was detected in DNA from control cells indicating that amplification was specific for top1-induced DNA breaks. Treatment with 0.1 µM CPT induced several cleavage sites that were intensified at 10 µM CPT, indicating a dose-dependent effect. Also, 10 µM CPT induced additional sites that were not as intense as those that were already detectable at 0.1 µM. Overall, seven cleavage sites were identified within an 18S rRNA region encompassing ~400 nucleotides (from position 5059) with one site of high intensity (position 5110) (Fig. 3 B). The same pattern of distribution and intensity of CPT-induced cleavage sites was observed in another colon carcinoma cell line, HT-29 (data not shown).
Comparison of CPT derivatives
We next compared CPT derivatives presently used in the clinic: topotecan (HycamtinR), 9-aminocamptothecin, CPT-11 (IrinotecanR), and its active metabolite, SN-38. Figure 4 shows representative sequencing gel autoradiographies. Cleavages were generally induced by all drugs at similar sites with the exception of CPT-11. CPT-11 was inactive at 0.1 µM, and induced more diffuse and less intense cleavage at 10 µM (lane 9) compared with the other camptothecins. The cleavage patterns of the other CPT derivatives were comparable with some intensity differences. For instance, one cleavage site indicated with an asterisk (Fig. 4 ) was consistently observed with 9-aminocamptothecin and was weaker for CPT and the other derivatives. Finally, SN-38 was the most active compound at low drug concentration, consistent with previously published in vitro data (32 ).
In vitro CPT-induced DNA cleavages in human 18S rDNA
In order to study the location of top 1-mediated cleavage sites induced by CPT (10 µM) in vitro, the 5'-end labeled A3-B3 DNA fragment (Fig. 1 )was amplified and this enabled us to identify top1-mediated breaks in the genomic region between nucleotides 5059 and 5430 (Fig. 5 ).
DISCUSSION
To our knowledge, this is the first report of sequencing the DNA cleavage sites induced by camptothecins in genomic DNA. The presented cleavage sites correspond to top1-mediated DNA breaks since similar sites were induced in vitro and in vivo, and the cleavage was readily reversible upon CPT removal. Several cleavage sites (5094, 5110, 5275 and 5286) match the CPT preferential sequence 5'-T^G, e.g. with a T at the 3'-DNA terminus of the top1-induced DNA break, to which the enzyme is covalently linked and a G at the 5'-OH DNA terminus, which is specific for CPT (14 ,16 ). Two other sites (5142 and 5164) had a T at the 3'-DNA terminus of the top I-induced DNA break.
Previous studies in genomic DNA were performed using the end-labeling technique and did not provide nucleotide level resolution (19 -23 ). Our in vivo sequencing is in agreement with the report of Culotta et al. (23 ) showing the high frequency of camptothecin-induced sites in transcribed regions of the rDNA genes. We find that within the 400 bp region examined the two strongest in vivo cleavage sites were ~200 bp apart (sites 5110 and 5328).
Because mammalian cells contain ~300 copies of the rDNA gene per genome, rDNA could be used to monitor CPT-induced DNA cleavage in different cell lines and possibly in tumor samples to monitor the cellular response to camptothecins.
Further studies are warranted to determine whether CPT-induced DNA cleavage sites can be mapped in single copy genes. The published in vitro cleavage sites sequenced in the human c-myc proto-oncogene (33 ,34 ) and top1 gene (16 ) may serve to guide the design of primers for LM-PCR based in vivo sequencing. A limitation of LM-PCR is that the method presently described cannot be used to sequence top2-induced DNA breaks because the 5'-DNA termini of such breaks are covalently linked to top2 and cannot be directly ligated.
REFERENCES
1 Hsiang, Y.H., Hertzberg, R., Hecht, S. and Liu, L.F. (1985) J. Biol. Chem., 260, 14873-14878.MEDLINE Abstract
2 Pommier, Y. and Tanizawa, A. (1993) In Hickman, J., and Tritton, T. (eds.), Cancer Chemotherapy. Blackwell Scientific Publications Ltd, Oxford, pp. 214-250.
3 Gupta, M., Fujimori, A. and Pommier, Y. (1995) Biochim. Biophys. Acta, 1262, 1-14.MEDLINE Abstract
5 Pommier, Y., Leteurtre, F., Fesen, M., Fujimori, A., Bertrand, R., Solary, E., Kohlhagen, G. and Kohn, K.W. (1994) Cancer Invest., 12, 530-542.MEDLINE Abstract
9 Bertrand, R. and Pommier, Y. (1995) In Studzinski, G. (ed.), Cell Growth and Apoptosis: A Practical Approach. IRL Press, Oxford University Press, Oxford, pp. 96-117.
10 Trask, D.K., DiDonato, J.A. and Muller, M.T. (1984) EMBO J., 3, 671-676.MEDLINE Abstract
11 Hsiang, Y.H. and Liu, L.F. (1988) Cancer Res., 48, 1722-1726.MEDLINE Abstract
12 Kjeldsen, E., Mollerup, S., Thomsen, B., Bonven, B.J., Bolund, L. and Westergaard, O. (1988) J. Mol. Biol., 202, 333-342.MEDLINE Abstract
13 Svejstrup, J.Q., Christiansen, K., Gromova, I.I., Andersen, A.H. and Westergaard, O. (1991) J. Mol. Biol., 222, 669-678.MEDLINE Abstract
14 Jaxel, C., Capranico, G., Kerrigan, D., Kohn, K.W. and Pommier, Y. (1991) J. Biol. Chem., 266, 20418-20423.MEDLINE Abstract
30 Pourquier, P., Pilon, A., Kohlhagen, G., Mazumder, A., Sharma, A. and Pommier, Y. (1997) J. Biol. Chem., In Press.
31 Covey, J.M., Jaxel, C., Kohn, K.W. and Pommier, Y. (1989) Cancer Res., 49, 5016-5022.MEDLINE Abstract
32 Tanizawa, A., Kohn, K.W., Kohlhagen, G., Leteurtre, F. and Pommier, Y. (1995) Biochemistry, 43, 7200-7206.
33 Leteurtre, F., Kohlhagen, G., Fesen, M.R., Tanizawa, A., Kohn, K.W. and Pommier, Y. (1994) J. Biol. Chem., 269, 7893-7900.MEDLINE Abstract
34 Leteurtre, F., Fujimori, A., Tanizawa, A., Chhabra, A., Mazumder, A., Kohlhagen, G., Nakano, H. and Pommier, Y. (1994) J. Biol. Chem., 269, 28702-28707.MEDLINE Abstract
*To whom correspondence should be addressed. Tel: +1 301 496 5944; Fax: +1 301 402 0752; Email: pommier@nih.gov +Present address: Hôpital Debrousse, 29 rue soeur Bouvier, 69322 Lyon cedex 05, France