Portal de Periódicos da CAPES

Alterations of DNA Repair in Melanoma Cell Lines Resistant to Cisplatin, Fotemustine, or Etoposide

2000; Elsevier BV; Volume: 114; Issue: 1 Linguagem: Inglês

10.1046/j.1523-1747.2000.00844.x

1523-1747

Thomas M. Rünger, Steffen Emmert, Dirk Schadendorf, Caroline Diem, Bernd Epe, Doris Hellfritsch,

Cancer Genomics and Diagnostics

Resistance to chemotherapy is a common phenomenon in malignant melanoma. In order to assess the role of altered DNA repair in chemoresistant melanoma, we investigated different DNA repair pathways in one parental human melanoma line (MeWo) and in sublines of MeWo selected in vitro for drug resistance against four commonly used drugs (cisplatin, fotemustine, etoposide, and vindesine). Host cell reactivation assays with the plasmid pRSVcat were used to assess processing of different DNA lesions. With ultraviolet-irradiated plasmids, no significant differences were found, indicating a normal (nucleotide excision) repair of DNA photoproducts. With singlet oxygen-treated plasmid, the fotemustine- and cisplatin-resistant lines exhibited a significantly increased (base excision) repair of oxidative DNA damage. With fotemustine-treated plasmid, the fotemustine-resistant subline did not exhibit an increased repair of directly fotemustine-induced DNA damage. Similar results were obtained with cisplatin-induced DNA crosslinks in the cisplatin-resistant line. The fotemustine- and etoposide-resistant sublines have been shown to exhibit a reduced expression of genes involved in DNA mismatch repair. We used a ''host cell microsatellite stability assay'' with the plasmid pZCA29 and found a 2.0-fold to 2.5-fold increase of microsatellite frameshift mutations (p ≤0.002) in the two resistant sublines. This indicates microsatellite instability, the hallmark of an impaired DNA mismatch repair. The increased repair of oxidative DNA damage might mediate an increased chemoresistance through an improved repair of drug-induced DNA damage. In contrast, a reduced DNA mismatch repair might confer resistance by preventing futile degradation of newly synthesized DNA opposite alkylation damage, or by an inability to detect such damage and subsequent inability to undergo DNA-damage-induced apoptosis. Resistance to chemotherapy is a common phenomenon in malignant melanoma. In order to assess the role of altered DNA repair in chemoresistant melanoma, we investigated different DNA repair pathways in one parental human melanoma line (MeWo) and in sublines of MeWo selected in vitro for drug resistance against four commonly used drugs (cisplatin, fotemustine, etoposide, and vindesine). Host cell reactivation assays with the plasmid pRSVcat were used to assess processing of different DNA lesions. With ultraviolet-irradiated plasmids, no significant differences were found, indicating a normal (nucleotide excision) repair of DNA photoproducts. With singlet oxygen-treated plasmid, the fotemustine- and cisplatin-resistant lines exhibited a significantly increased (base excision) repair of oxidative DNA damage. With fotemustine-treated plasmid, the fotemustine-resistant subline did not exhibit an increased repair of directly fotemustine-induced DNA damage. Similar results were obtained with cisplatin-induced DNA crosslinks in the cisplatin-resistant line. The fotemustine- and etoposide-resistant sublines have been shown to exhibit a reduced expression of genes involved in DNA mismatch repair. We used a ''host cell microsatellite stability assay'' with the plasmid pZCA29 and found a 2.0-fold to 2.5-fold increase of microsatellite frameshift mutations (p ≤0.002) in the two resistant sublines. This indicates microsatellite instability, the hallmark of an impaired DNA mismatch repair. The increased repair of oxidative DNA damage might mediate an increased chemoresistance through an improved repair of drug-induced DNA damage. In contrast, a reduced DNA mismatch repair might confer resistance by preventing futile degradation of newly synthesized DNA opposite alkylation damage, or by an inability to detect such damage and subsequent inability to undergo DNA-damage-induced apoptosis. O6-methylguanine-DNA-methyltransferase Drug resistance is a major obstacle to successful treatment of metastatic malignant melanoma. Although moderate response rates (20% - 50%) can be obtained by mono or poly chemotherapy, relapses are almost certain and second line therapies are largely ineffective. In other malignancies, several different mechanisms of chemoresistance have been described, such as alterations in drug transport, an increase in drug detoxification, an induction of cellular protective agents, or an increased DNA repair (of drug-induced DNA damage) (Harris, 1985Harris A.L. DNA repair and resistance to chemotherapy.Cancer Surv. 1985; 4: 601-624PubMed Google Scholar). In malignant melanoma, the mechanisms of chemoresistance are largely unknown. In order to assess the role of altered DNA repair in chemoresistant melanoma, we investigated (i) the repair of ultraviolet-induced DNA photoproducts (through the nucleotide excision repair pathway), (ii) the repair of oxidative DNA base modifications (through the base excision repair pathway), (iii) the repair of fotemustine-induced DNA adducts (through direct reversal of DNA adducts), (iv) the repair of cisplatin-induced DNA crosslinks (through DNA recombination), and (v) microsatellite stability (to assess DNA mismatch repair) in one chemosensitive parental melanoma line (MeWo) and in fotemustine-, cisplatin-, etoposide-, and vindesine-resistant MeWo sublines. These stable sublines were selected and established by continuous drug exposure for over 2 y as described byKern et al., 1997Kern M.A. Helmbach H. Artuc M. Karmann D. Jurgowsky K. Schadendorf D. Human melanoma cell lines selected in vitro displaying various levels of drug resistance against cisplatin, fotemustine, vindesine or etoposide: modulation of proto-oncogene expression.Anticancer Res. 1997; 17: 4359-4370PubMed Google Scholar. The four agents are commonly used for therapy of metastatic melanoma and represent four different modes of action. Three of the agents target DNA: fotemustine (diethyl-1-[3,2-chloroethyl]-3-nitrosoureido ethyl phosphonate), a newer chloronitrosourea, is an alkylating agent; cisplatin (cis-dichlorodiamin-platin (II)) acts through binding to DNA, principally through the formation of guanine-guanine and adenine-guanine intra- and inter-strand crosslinks (McKeage et al., 1991McKeage M.J. Higgins J.D. Kelland L.R. Platinum and other metal coordination compounds in cancer chemotherapy. A commentary on the sixth international symposium: San Diego, California, 23–26th January 1991.Br J Cancer. 1991; 64: 788-792Crossref PubMed Scopus (26) Google Scholar); etoposide is an inhibitor of topoisomerase II and generates DNA strand breakage (Henwood et al., 1990Henwood J.M. Brogden R.N. Etoposide. A review of its pharmacodynamic and pharmacokinetic properties, and therapeutic potential in combination chemotherapy of cancer.Drugs. 1990; 39: 438-490Crossref PubMed Scopus (108) Google Scholar). Only the vinca alkaloid vindesine does not interact with DNA: it binds to tubulin and inhibits its polymerization to microtubules. We used four different ''host cell reactivation assays'' with the plasmid pRSVcat, damaged in vitro by either ultraviolet B (UVB), singlet oxygen, cisplatin, or fotemustine. This facilitated the separate functional analysis of four different DNA repair pathways. The assays in which the plasmids were damaged in vitro by cisplatin or fotemustine were newly established for this investigation. Microsatellite instability, the hallmark of a DNA mismatch repair deficiency, is usually identified by polymerase chain reaction amplification of microsatellites and size comparison of resulting DNA fragments from tumor tissue and normal tissue. Such comparison is hardly possible when investigating microsatellite stability in established cell lines. Recently, we described the new shuttle vector pZCA29 for the detection of microsatellite instability by a host cell replication error assay (Diem and Rünger, 1998Diem C. Rünger T.M. A novel plasmid shuttle vector for the detection and analysis of microsatellite instability in cell lines.Mutat Res. 1998; 407: 117-124Crossref PubMed Scopus (9) Google Scholar). Here, we used this assay to assess microsatellite instability in the melanoma lines. The establishment and testing of drug-resistant sublines from the drug-sensitive parental cell line MeWo, derived from a lymph node metastasis, was described earlier (Kern et al., 1997Kern M.A. Helmbach H. Artuc M. Karmann D. Jurgowsky K. Schadendorf D. Human melanoma cell lines selected in vitro displaying various levels of drug resistance against cisplatin, fotemustine, vindesine or etoposide: modulation of proto-oncogene expression.Anticancer Res. 1997; 17: 4359-4370PubMed Google Scholar). MeWoCIS1.0 exhibited a 6-fold-increased relative resistance against cisplatin, compared with the parental line, MeWoFOTE40 a 26-fold-increased resistance against fote- mustine, MeWoETO1.0 a 35.7-fold-increased resistance against etoposide, and MeWoVIND5 a 10.2-fold-increased resistance against vindesine. Cross-resistance has been shown for the three lines resistant to DNA damaging agents, but not with the one resistant to vindesine. Cells were cultured in Earl's modified essential medium supplemented with 10% fetal calf serum, 2 mM l-glutamine, nonessential amino acids, 1 mM pyruvate, 50 ng tylosin per ml, and 100 U penicillin/streptomycin per ml. In order to ensure maintenance of the chemoresistant phenotype, the media for the drug-sensitive sublines were supplemented with 0.25 μg cisplatin per ml, 4 μg fotemustine per ml, 0.025 μg etoposide per ml, or 0.25 ng vindesine per ml, respectively. Two sublines have been shown to have a reduced expression of proteins involved in DNA mismatch repair. Northern blot analysis revealed an 80% reduced expression of hMLH1 in MeWoETO1.0 and a 70% reduced expression of hMSH2 and 50% reduced expression of hMSH1 in MeWoFOTE40 (Lage et al., 1999Lage H. Christmann M. Kern M.A. Dietel M. Pick M. Kaina B. Schadendorf D. Expression of DNA repair proteins hMSH2, hMSH6, hMLH1, MGMT and MPT in melanoma cells with acquired drug resistance.Int J Cancer. 1999; 80: 744-750https://doi.org/10.1002/(sici)1097-0215(19990301)80:5 3.0.co;2-5Crossref PubMed Scopus (0) Google Scholar). Here, these two lines were tested for microsatellite instability, the hallmark of a deficient DNA mismatch repair. Only subconfluent cells at exponential growth phase were used. This assay was carried out as described previously (Seidman et al., 1985Seidman M.M. Dixon K. Razzaque A. Zagursky R.J. Berman M.L. A shuttle vector plasmid for studying carcinogen-induced point mutations in mammalian cells.Gene. 1985; 38: 233-237Crossref PubMed Scopus (215) Google Scholar;Rünger et al., 1995Rünger T.M. Epe B. Möller K. Repair of ultraviolet B and singlet oxygen-induced DNA damage in xeroderma pigmentosum cells.J Invest Dermatol. 1995; 104: 68-73Crossref PubMed Scopus (32) Google Scholar). The 5 kb, nonreplicating plasmid pRSVcat was irradiated with 1 kJ per m2, 5 kJ per m2, or 10 kJ per m2 UVB (Phillips, Hamburg, Germany; TL21 lamps, emission maximum at 315 nm, spectrum 275–365 nm) at 0°C and at a concentration of 30 μg per ml. Eight micrograms of plasmid were used to transfect 4 million cells using the electroporation procedure (Gene Pulser, BioRAD, Hercules, CA: 280 V, 960 μF capacitor, 400 μl of serum-free medium, room temperature, time constant between 20 and 30 ms). The cells were then transferred to 15 ml of prewarmed medium. At peak expression after 3 d, a cell extract was produced by three freeze-thaw cycles with ethanol in dry ice and subsequent rewarming to 37°C. Chloramphenicol acetyl transferase (CAT) activity in the cell extract, which depends on successful repair of the UVB-induced photoproducts prior to CAT expression, was determined using the one-vial procedure described byNeumann et al., 1987Neumann J.R. Morency C.A. Russian K.O. A novel rapid assay for chloramphenicol acetyltransferase gene expression.Biotechniques. 1987; 5: 444-447Google Scholar andEastman, 1999Eastman A. An improvement to the novel rapid assay for chloramphenical acetyltransferase gene expression.Biotechniques. 1999; 5: 730-732Google Scholar. The CAT activity was calibrated with several concentrations of CAT enzyme (Boehringer Mannheim, Germany). The values were corrected by subtracting the cell-specific background activity and were related to protein content, using the Coomassie blue method (BioRAD). The mean specific activity with undamaged plasmid was 0.2 U CAT per mg protein with MeWo. The relative CAT activity with the UV-irradiated samples, which reflects repair of DNA photoproducts on the plasmid by the host cells, was calculated as a percentage of the parallel control sample transfected with untreated plasmid. As described earlier (Rünger et al., 1995Rünger T.M. Epe B. Möller K. Repair of ultraviolet B and singlet oxygen-induced DNA damage in xeroderma pigmentosum cells.J Invest Dermatol. 1995; 104: 68-73Crossref PubMed Scopus (32) Google Scholar), the plasmid pRSVcat was treated with singlet oxygen, generated by the photosensitizer methylene blue (10 μg per ml), and exposure to visible light (1, 5, and 10 min illumination with a 1000 W Osram halogen lamp, 47 W per m2, emission between 400 and 800 nm). The plasmid concentration during treatment was 625 μg per ml. The DNA damage profile induced by this procedure comprises mainly formamidopyrimidine-DNA glycosylase (FPG)-sensitive lesions and very few single-strand breaks, pyrimidine dimers, or apurinic/apyrimidinic sites (Müller et al., 1990Müller E. Boiteux S. Cunningham R.P. Epe B. Enzymatic recognition of DNA modifications induced by singlet oxygen and photosensitizers.Nucl Acids Res. 1990; 18: 5969-5973Crossref PubMed Scopus (80) Google Scholar). It is identical to the DNA damage profile produced by NDPO2, a chemically clean source of singlet oxygen. One FPG-sensitive lesion was generated on 2000 bp per minute of illumination. Following this plasmid treatment, the host cell reactivation assay was identical to the one with UVB-irradiated plasmid. An almost identical assay has been described byDabholkar and Reed, 1992Dabholkar M. Reed E. Host cell reactivation of cisplatin-damaged plasmid DNA in human non-T leukocyte cell lines.Cancer Lett. 1992; 63: 143-150Abstract Full Text PDF PubMed Scopus (3) Google Scholar. The plasmid pRSVcat was treated with 1 μg per ml, 3 μg per ml, 5 μg per ml, or 10 μg per ml cisplatin for 2 h at 37°C in the dark and purified by alcohol precipitation. Neutral and alkali agarose gel electrophoresis of the treated plasmid did not detect cleavage. Following this plasmid treatment, the host cell reactivation assay was identical to the one with UVB-irradiated plasmid. The plasmid pRSVcat was treated with 25 μg per ml, 50 μg per ml, 100 μg per ml, or 200 μg per ml fotemustine for 2 h at 37°C and purified by alcohol precipitation. Neutral agarose gel electrophoresis of the treated plasmid detected modest cleavage with the two highest doses, as shown by an increasing, but not complete, loss of supercoiled plasmid, and an increase in the relaxed circular form. Correspondingly, alkali agarose gel electrophoresis showed a partial loss of circular forms and an increase of linear plasmid. Following this plasmid treatment, the host cell reactivation assay was identical to the one with UVB-irradiated plasmid. This assay with the plasmid pZCA29 has been described earlier (Diem and Rünger, 1998Diem C. Rünger T.M. A novel plasmid shuttle vector for the detection and analysis of microsatellite instability in cell lines.Mutat Res. 1998; 407: 117-124Crossref PubMed Scopus (9) Google Scholar). The plasmid carries the lacZ gene, interrupted by one (CA)14 and one (CA)15 repeat. These inserts inactivate lacZ by a + 1 frameshift. Four million melanoma cells were transfected with 4 μg of pZCA29 as described above for the plasmid pRSVcat and incubated for 3 d. Replicated plasmids were recovered and nonreplicated plasmids were removed by digestion with the restriction enzyme DpnI. Recovered plasmid was introduced into Escherichia coli DH10B by electrotransformation (Gene Pulser, BioRAD: 2.5 kV, 25 μF capacitor, 600 Ω, total volume 80 μl, time constant between 11 and 14 ms). The bacteria were plated on selective LB agar dishes containing 100 μg per ml ampicillin for the selection of plasmid-containing bacteria (the plasmid carries the ampicillin-resistance gene) as well as 9mgisopropyl-β-D-thiogalactopyranoside and 3.6 mg 5-bromo-4-chloro-3-indolyl-β- D-galactoside per plate for the detection of β-galactosidase activity, and incubated at 37°C overnight. Blue bacterial colonies could be distinguished from white ones. At least three independent bacterial transformations were performed with every sample of recovered plasmid DNA. Based on our previous sequence analysis of recovered pZCA29, we know that mutations in plasmids leading to a normal reading frame or to a - 1 frameshift cause blue coloration. Eighty-nine percent of blue colonies were shown to contain deletions or insertions of one or two CA dinucleotides. Deletions or insertions of three CA dinucleotides do not change the reading frame and therefore are not detectable. The total number of bacterial colonies and the number of blue colonies were counted. The mutation frequency, which reflects the instability of the CA repeats in the plasmid during replication by the host cells, was calculated as the total number of blue colonies divided by the total number of all bacterial colonies. Student's t test was used to test for differences. Table 1 lists the host cell reactivation of UVB-induced DNA damage during passage of UVB-irradiated plasmid pRSVcat through the drug-sensitive parental MeWo line and the four drug-resistant sublines of MeWo. No significant differences were noted between the melanoma lines, indicating that there are no pronounced deficiencies or increases of nucleotide excision repair of DNA photoproducts in these lines. The data are based on 10 independent samples with the parental MeWo line and on two to four independent samples with the drug-resistant lines.Table 1Repair of UVB-induced DNA damage on the plasmid pRSVcat by a drug-sensitive melanoma line (MeWo) and drug-resistant sublines of MeWo, as measured with a host cell reactivation assayIrradiation with UVB0kJ per m21kJ per m25kJ per m210kJ per m2MeWo (drug-sensitive, parental)100.097.9±4.5% aRelative CAT activity (as percentage of unirradiated control) in cell extracts after transfection of the plasmid pRSVcat irradiated with UVB (mean ± SD). The reactivation of the plasmid reflects the capacity of these host cells to repair UVB-induced DNA photoproducts.80.8±20.4%57.7±23.2%MeWoFOTE40 (fotemustine-resistant)100.095.4±6.6%66.5±18.5%41.0±10.0%MeWoCIS1.0 (cisplatin-resistant)100.087.2±12.0%61.3±2.1%51.1±9.5%MeWoETO1.0 (etoposide-resistant)100.096.9±4.4%96.1±5.5%68.9±3.6%MeWoVIND5 (vindesine-resistant)100.091.8±16.5%95.5±5.3%66.6±22.6%a Relative CAT activity (as percentage of unirradiated control) in cell extracts after transfection of the plasmid pRSVcat irradiated with UVB (mean ± SD). The reactivation of the plasmid reflects the capacity of these host cells to repair UVB-induced DNA photoproducts. Open table in a new tab A markedly and significantly increased repair of oxidative DNA damage was found in the fotemustine-resistant and the cisplatin-resistant MeWo sublines (Figure 1); e.g., with 12.5 FPG-sensitive lesions/plasmid, the host cell reactivation in MeWoFOTE40 was 2.5-fold higher and in MeWoCIS1.0 2.2-fold higher than in the parental MeWo line (p 3.0.co;2-5Crossref PubMed Scopus (0) Google Scholar in the same fotemustine-resistant subline does not lead to an improved repair of directly fotemustine-induced DNA damage and should not be the mechanism of fotemustine resistance. This exemplifies the power of host cell reactivation assays as used here, because they facilitate a functional analysis of DNA repair processes. Oxidative DNA damage, however, was repaired more efficiently in the fotemustine-resistant subline. Oxidative guanine base modifications are processed by the base excision repair pathway, with the initial action of DNA glycosylases, such as hOGG1 (Radicella et al., 1997Radicella J.P. Dherin C. Desmaze C. Fox M.S. Boiteux S. Cloning and characterization of hOGG1, a human homolog of the OGG1 gene of Saccharomyces cerevisiae.Proc Natl Acad Sci USA. 1997; 94: 8010-8015Crossref PubMed Scopus (557) Google Scholar). Therefore, we would like to hypothesize that fotemustine resistance could be conferred by an increased repair of a subset of fotemustine-induced DNA lesions, one that is processed by the base excision repair pathway. The weakness of host cell reactivation assays is that they do not identify which repair protein or which repair pathway is responsible for detected abnormalities. Both the pathway of direct repair of alkylation damage and the base excision repair pathway are known to be inducible (Friedberg et al., 1995Friedberg E.C. Walker G.C. Siede W. DNA Repair and Mutagenesis. eds. ASM Press, Washington1995Google Scholar;Grösch et al., 1998Grösch S. Fritz G. Kaina B. Apurinic endonuclease (ref-1) is induced in mammalian cells by oxidative stress and involved in clastogenic adaptation.Cancer Res. 1998; 58: 4410-4416PubMed Google Scholar). A key enzyme of the latter is the AP-endonuclease, a plausible candidate for an increased repair of oxidative DNA damage; however, the list of plausible mechanisms could be extended considerably. Hill et al., 1994Hill B.T. Shellard S.A. Fichtinger Schepman A.M. Schmoll H.J. Harstrick A. Differential formation and enhanced removal of specific cisplatin-DNA adducts in two cisplatin-selected resistant human testicular teratoma sublines.Anticancer Drugs. 1994; 5: 321-328Crossref PubMed Scopus (25) Google Scholar demonstrated that cisplatin resistance in testicular teratoma was associated with an elevated removal of platinum-DNA adducts and crosslinks. A cisplatin-resistant subclone of a cisplatin-sensitive ovarian cancer line exhibited a near 2-fold increased ability to repair cisplatin-induced DNA damage (Masuda et al., 1988Masuda H. Ozols R.F. Lai G.M. Fojo A. Rothenberg M. Hamilton T.C. Increased DNA repair as a mechanism of acquired resistance to cis-diaminedichloroplatinum (II) in human ovarian cancer cell lines.Cancer Res. 1988; 48: 5713-5716PubMed Google Scholar). Inhibition of DNA repair by aphidicoline restored cisplatin sensitivity in the resistant line. This was also confirmed byJohnson et al., 1994Johnson S.W. Swiggard P.A. Handel L.M. Brennan J.M. Godwin A.K. Ozols R.F. Hamilton T.C. Relationship between platinum-DNA adduct formation and removal and cisplatin cytotoxicity in cisplatin-sensitive and -resistant human ovarian cancer cells.Cancer Res. 1994; 54: 5911-5916PubMed Google Scholar) who found an up to 2.5 times higher removal of cisplatin-induced inter-strand crosslinks in cisplatin-resistant ovarian cancer lines.Maynard et al., 1989Maynard K.R. Hosking L.K. Hill B.T. Use of host cell reactivation of cisplatin-treated adenovirus 5 in human cell lines to detect repair of drug-treated DNA.Chem Biol Interact. 1989; 71: 353-365Crossref PubMed Scopus (8) Google Scholar used a host cell reactivation assay with a cisplatin-treated adenovirus in human cells. A nucleotide-excision-repair-deficient xeroderma pigmento- sum cell line was found to be defective, as were some cisplatin-resistant ovarian cancer lines.Zeng Rong et al., 1995Zeng Rong N. Paterson J. Alpert L. Tsao M.S. Viallet J. Alaoui-Jamali M.A. Elevated DNA repair capacity is associated with intrinsic resistance of lung cancer to chemotherapy.Cancer Res. 1995; 55: 4760-4764PubMed Google Scholar used a very similar assay with pRSVcat (the same plasmid we used) and found a correlation between repair of cisplatin-induced DNA damage in the host cell reactivation assay and the intrinisic resistance of lung cancer to cisplatin. With our cisplatin-resistant MeWo subclone we found results similar to those with the fotemustine-resistant subline: no difference in the assay where we investigated drug-induced DNA lesions directly, but an increased repair of oxidative DNA lesions. As with fotemustine, this might point to an increased repair of a subset of cisplatin-induced DNA lesions, e.g., of monofunctional DNA adducts, but not of bifunctional adducts (crosslinks). The hereditary disorder xeroderma pigmentosum exemplifies that a reduced nucleotide excision repair of DNA photoproducts increases the risk for nonmelanoma skin cancer and melanoma. Therefore, an increased repair of DNA photoproducts would not be expected in melanoma. Some of the chemotherapy-induced DNA lesions, however, are also processed by nucleotide excision repair, known to process a wide range of DNA lesions including, for example, intra-strand crosslink. For instance, cisplatin sensitivity has been linked to a defective nucleotide excision repair, with low levels of xeroderma pigmentosum group A protein in testicular germ cell tumors (Koberle et al., 1999Koberle B. Masters J.R. Hartley J.A. Wood R.D. Defective repair of cisplatin-induced, DNA, damage caused by reduced,XPA protein in testicular germ cell tumours.Curr Biol. 1999; 9: 273-276Abstract Full Text Full Text PDF PubMed Scopus (260) Google Scholar) and of XPG protein in the mouse leukemia line L1220 (Vilpo et al., 1995Vilpo J.A. Vilpo L.M. Szymkowski D.E. O'Donovan A. Wood R.D. An XPG DNA repair defect causing mutagen hypersensitivity in mouse leukemia L1210 cells.Mol Cell Biol. 1995; 15: 290-297Crossref PubMed Scopus (31) Google Scholar). This is why a decreased nucleotide excision repair should not be excluded as a mechanism of chemoresistance. This is in line with the findings ofHatton et al., 1995Hatton D.H. Mitchell D.L. Strickland P.T. Johnson R.T. Enhanced photoproduct repair: its role in the DNA damage-resistance phenotype of human malignant melanoma cells.Cancer Res. 1995; 55: 181-189PubMed Google Scholar who demonstrated UVC-resistant subclones of metastatic melanoma with an enhanced rate of postreplication recovery, coresistance to cisplatin, and an increased repair of DNA photoproducts. In murine melanoma cells, chemosensitivity was also found to correlate with the capacity to repair UVC-induced DNA damage (and with metastatic potential) (Wei et al., 1997Wei Q. Cheng L. Xie K. Bucana C.D. Dong Z. Direct correlation between DNA repair capacity and metastatic potential of K-1735 murine melanoma cells.J Invest Dermatol. 1997; 108: 3-6Crossref PubMed Scopus (18) Google Scholar). None of our examined MeWo sublines, however, showed a reduced nucleotide excision repair of UVB-induced DNA photoproducts. Unlike the DNA repair pathways mentioned above, where an increased repair has been shown to confer chemoresistance, a decreased DNA mismatch repair has been implicated in tolerance to alkylating agents (Friedberg et al., 1995Friedberg E.C. Walker G.C. Siede W. DNA Repair and Mutagenesis. eds. ASM Press, Washington1995Google Scholar;Karran and Hampson, 1996Karran P. Hampson R. Genomic instability and tolerance to alkylating agents.Cancer Surv. 1996; 28: 69-85PubMed Google Scholar;Fink et al., 1998aFink D. Aebi S. Howell S.B. The role of DNA mismatch repair in drug resistance.Clin Cancer Res. 1998; 4: 1-6PubMed Google Scholar). Attempts to process DNA mismatches at persistent O6-methylguanine are thought to be made by the DNA mismatch repair pathway. In view of the strand specificity of mismatch repair for the newly synthesized strand, mismatch repair opposite the modified base would be futile, because the intact mismatch repair would continue to excise the newly synthesized DNA. This continuing futile degradation of DNA with its attendant persistence of strand breaks in DNA has been suggested to increase the potential of cell killing in mismatch-repair-proficient cells and of cell survival in mismatch-repair-deficient cells. In addition, the inability to detect DNA damage (through mismatches) in mismatch- repair-deficient cells would also impair their ability to activate apoptosis, thereby increasing cell survival, but also mutability. This model explains how cells that are incapable of mismatch repair could become tolerant to alkylating agents; however, the possibly resulting hypermutability would contribute to tumor diversification, an undesirable effect in tumor therapy. In the fotemustine-resistant MeWo subline, we found an increased microsatellite instability. This points to a disturbance of the DNA mismatch repair pathway and is in accordance with the reduced expression of DNA mismatch repair genes described in this line earlier (Lage et al., 1999Lage H. Christmann M. Kern M.A. Dietel M. Pick M. Kaina B. Schadendorf D. Expression of DNA repair proteins hMSH2, hMSH6, hMLH1, MGMT and MPT in melanoma cells with acquired drug resistance.Int J Cancer. 1999; 80: 744-750https://doi.org/10.1002/(sici)1097-0215(19990301)80:5<744::aid-ijc19>3.0.co;2-5Crossref PubMed Scopus (0) Google Scholar): MeWoFOTE40 showed a 70% - 50% reduction of hMSH2 and hMSH1 RNA levels. In accordance with the above-mentioned model, the reduced mismatch repair is suggested to be responsible for the tolerance to the alkylating agent fotemustine in this subline. In addition to alkylating agents, topoisomerase II inhibitors have also been identified as agents for which loss of DNA mismatch repair causes drug resistance (Fink et al., 1998bFink D. Nebel S. Norris P.S. Aebi S. Kim H.K. Haas M. Howell S.B. The effect of different chemotherapeutic agents on the enrichment of DNA mismatch repair-deficient tumour cells.Br J Cancer. 1998; 77: 703-708Crossref PubMed Scopus (67) Google Scholar). Therefore the impaired mismatch repair in MeWoETO1.0, shown here by microsatellite instability and byLage et al., 1999Lage H. Christmann M. Kern M.A. Dietel M. Pick M. Kaina B. Schadendorf D. Expression of DNA repair proteins hMSH2, hMSH6, hMLH1, MGMT and MPT in melanoma cells with acquired drug resistance.Int J Cancer. 1999; 80: 744-750https://doi.org/10.1002/(sici)1097-0215(19990301)80:5<744::aid-ijc19>3.0.co;2-5Crossref PubMed Scopus (0) Google Scholar by an 80% reduction of hMLH1 RNA levels, is suggested to be responsible for the tolerance to etoposide. Recognition of an increased repair of chemotherapy-induced DNA damage (or of a decreased DNA mismatch repair) as a mechanism of chemoresistance in melanoma offers the outlook to overcome resistance by blocking the corresponding repair pathway. This approach has already been taken by blocking MGMT with a suicide substrate (Maze et al., 1996Maze R. Carney J.P. Kelley M.R. Glassner B.J. Williams D.A. Samson L. Increasing DNA repair methyltransferase levels via bone marrow stem cell transduction rescues mice from the toxic effects of 1,3-bis (2-chloroethyl) -1-nitrosourea, a chemotherapeutic alkylating agent.Proc Natl Acad Sci USA. 1996; 93: 206-210Crossref PubMed Scopus (110) Google Scholar;Kurpad et al., 1997Kurpad S.N. Dolan M.E. McLendon R.E. Archer G.E. Moschel R.C. Pegg A.E. Bigner D.D. Friedman H.S. Intraarterial O6-benzylguanine enables the specific therapy of nitrosourea-resistant intracranial human glioma xenografts in athymic rats with 1,3-bis (2-chloroethyl) -1-nitrosourea.Cancer Chemother Pharmacol. 1997; 39: 307-316https://doi.org/10.1007/s002800050577Crossref PubMed Scopus (28) Google Scholar). In melanoma, this would require the exact identification of the involved pathway first. Therefore, the four different host cell reactivation assays presented here, which assess the processing of relatively well defined DNA lesions introduced in vitro, and the host cell microsatellite instability assay are suggested to be suitable tools toward that goal. This work was supported by the Deutsche Forschungsgemeinschaft within the Sonderforschungsbereich (Collaborative Research Center) 172 ''Molecular Mechanisms of Carcinogenic Primary Lesions'' and by individual grants to TMR (Ru 377/4–1) and DS (Scha 422/7–2).