Cisplatin Induction of ERCC-1 mRNA Expression in A2780/CP70 Human Ovarian Cancer Cells*

ERCC-1 is a critical gene within the nucleotide excision repair pathway, and cells without a functionalERCC-1 do not perform cisplatin-DNA adduct repair. We therefore investigated the cisplatin effect on ERCC-1mRNA expression in vitro. In response to a 1-h cisplatin exposure, A2780/CP70 human ovarian cancer cells showed a 6-fold increase in steady-state level of ERCC-1 mRNA. This rise was attributable to increased transcription as measured by nuclear run-on assays and a 60% increase in ERCC-1mRNA half-life. The increase in ERCC-1 mRNA was preceded by a 4–5-fold rise in mRNA expressions ofc-fos and c-jun, a 14-fold increase in c-Jun protein phosphorylation, and an increase in in vitronuclear extract binding activity to the AP-1-like site ofERCC-1. These data suggest that the induction ofERCC-1 expression in A2780/CP70 cells exposed to cisplatin results from two major factors: (a) an increase in the expression of transactivating factors that bind the AP-1-like site in the 5′-flanking region of ERCC-1 and (b) an increase in the level of c-Jun phosphorylation that enhances its transactivation property.

cis-Diamminedichloroplatinum (II) (cisplatin) 1 is one of the most widely used chemotherapeutic agents for the treatment of human ovarian cancer and other tumors (1)(2)(3)(4). However, the efficacy of cisplatin is hampered by intrinsic or acquired resistance of cancer cells to its cytotoxicity. Although the mechanism of cisplatin resistance in vivo is not clearly understood, laboratory studies on tumor tissues and cell lines suggest that resistance to cisplatin is nearly always multifactorial (5)(6)(7). These factors include impaired cellular uptake of cisplatin (5,7), enhanced intracellular detoxification by glutathione and metallothionein systems (5, 8 -10), altered patterns of DNA platination (11,12), increased tolerance of platinum-DNA damage (11,12), and enhanced repair of DNA damage (5)(6)(7)13).
The cellular toxicity of cisplatin occurs primarily through its ability to bind covalently to DNA and prevent DNA replication and transcription (3,14). Cisplatin reacts with DNA to form intrastrand and/or interstrand cross-links of platinum adducts (3). Cells exposed to cisplatin must either repair or tolerate the DNA damage if they are to survive. A percentage of platinum-DNA lesions formed in vivo are repaired by human cells.
The 414-bp sequence in the 5Ј-flanking region of the ERCC-1 gene has been studied in detail by Hoeijmakers and colleagues (27,28), who have shown that constructs of this region may drive transcription of ERCC-1. Classical promoter elements like CAAT, TATA, and GC boxes are absent from the Ϫ1 to Ϫ170-bp portion of this region (28), although an AP-1-like site exists further upstream (Fig. 1). We have previously demonstrated that there are elevated levels of ERCC-1 mRNA in ovarian cancer tissues of patients clinically resistant to platinum compounds (25). However, the fundamental molecular basis of transcriptional activation and regulation of ERCC-1 expression is not well elucidated.
The AP-1 (activator protein 1) family is a group of transcription factors responsible for the activation of a wide variety of genes in different cell types and tissues (29 -31). The AP-1 transcription factor consists of either heterodimers formed between Jun and Fos family members of proto-oncoproteins or homodimers of Jun proteins (29 -31). AP-1-binding sites (5Ј-TGAGTCA-3Ј) are frequently found in promoters or enhancers of genes that are inducible by a wide range of extracellular signals. Evidence showed that cisplatin induced expression of proto-oncogenes c-fos/c-jun (32,33) and activated c-Jun NH 2terminal kinase/stress-activated protein kinase (JNK/SAPK) (34,35) in ovarian cancer cells and other tumor cells. JNK/ SAPK is a subfamily of MAP kinases in the Ras pathway which is responsible for the phosphorylation of Jun protein. Phosphorylation of the c-Jun at serine residues 63 and 73 in its NH 2terminal domain greatly enhances the transcriptional activity of the AP-1-binding sites (36 -38) and AP-1-regulated genes (39 -41). Therefore, it is possible that the effect of cisplatin on ERCC-1 could be through AP-1 induction or c-Jun phosphorylation. Because methodological problems would limit investigations of mechanism in ovarian cancer tissues taken from patients, we have now conducted studies in the human ovarian cancer cell line, A2780/CP70, to investigate these possibilities.

EXPERIMENTAL PROCEDURES
Cell Line and Cell Culture Conditions-The human ovarian cancer cell line A2780/CP70 has been described previously (42) and were used in all experiments. Cells were cultured in monolayer using RPMI 1640 media supplemented with 10% (v/v) fetal calf serum, 2 mM L-glutamine, 0.2 units/ml human insulin, 50 units/ml penicillin, 50 g/ml streptomycin (Life Technologies, Inc., Gaithersburg, MD). Cells were grown in logarithmic growth at 37°C in a humidified atmosphere consisting of 5% CO 2 , 95% air. Cells were routinely tested for mycoplasmal infection using a commercial assay system (MycoTect; Life Technologies, Inc.), and new cultures were established monthly from frozen stocks. All media and reagents contained Ͻ0.1 ng/ml endotoxin as determined by Limulus polyphemus amebocyte lysate assay (Whittaker Bioproducts, Walkersville, MD). Cell viability was determined in triplicate by trypan blue dye exclusion. Before starting the experiments, the cells were grown to ϳ90% confluence after subculturing. Cisplatin (Drug Synthesis and Chemistry Branch, Division of Cancer Treatment, National Cancer Institute, Bethesda, MD) was initially dissolved in phosphatebuffered saline without Ca 2ϩ or Mg 2ϩ at 1.0 mg/ml (3.33 mM cisplatin), and dilutions from this solution were made in media to obtain the desired drug treatment concentrations. The cisplatin dose for A2780/ CP70 cells was 40 to 80 M unless otherwise indicated. Cisplatin treatments were for 1 h. After drug treatments, cells were washed twice with phosphate-buffered saline without Ca 2ϩ or Mg 2ϩ , given fresh drug-free media, and incubated for 24 to 48 h or the time indicated. Thereafter, the cells were harvested for use in the RNA and protein isolation assays. ␣-Amanitin (Calbiochem, San Diego, CA) and cycloheximide (Calbiochem) were dissolved in water. Cisplatin cytotoxicity was determined using the microculture tetrazolium assay, as described previously (43). The concentrations of other drugs used in the studies were not toxic to the cells as confirmed by cell recoveries, trypan blue dye exclusion, and cytotoxicity assay.
RNA Isolation and Northern Blot Analysis-Total RNA was isolated from cells by acid guanidinium thiocyanate-phenol-chloroform extraction (44), or using a commercial total RNA isolation reagent kit (Life Technologies, Inc.) according to the manufacturer's instructions. Thirty micrograms of denatured RNA per lane were separated by electrophoresis (Life Technologies, Inc.) through 1% agarose-formaldehyde and transferred to nylon membrane (Zeta-Probe GT; Bio-Rad) by electrophoretic transfer (Trans-Blot Cell; Bio-Rad). Membranes were prehybridized in Quik-Hyb (Stratagene, Menasha, WI) for 15-30 min at 68°C and then hybridized for 1 to 2 h at 68°C in Quik-Hyb containing 0.67 g/ml denatured salmon testes DNA (Stratagene) and 32 P-labeled cDNA probe. After washings of increasing stringency, the membranes were air dried, exposed to Kodak XAR-5 x-ray film with intensifying screens at Ϫ80°C, and then analyzed by Collage Analysis (Fotodyne Inc., New Berlin, WI) and quantitated by densitometrical scanning. Before hybridization with a second labeled cDNA probe, the first probe was removed by washing for 2 h at 75°C in 1 mM Tris-HCl (pH 8.0) containing 1 mM EDTA and 0.1 ϫ Denhardt's solution (45). The entire sequence of experiments (including growth of A2780/CP70 cells, drug treatment, and Northern blotting and hybridization) was performed and the results reproduced in two or more separate experiments. Equal RNA loading was determined by visualization of 18 S and 28 S riboso-mal RNA bands in ethidium bromide-stained gels and quantification of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) housekeeping gene transcript on Northern blots.
Preparation of cDNA Probes-A 1.05-kilobase cDNA probe for human ERCC-1 was obtained from Dr. Aziz Sancar (University of North Carolina, Chapel Hill, NC). A 0.8-kilobase cDNA for human GAPDH was obtained from Dr. Mitchell Olman (University of California, San Diego, CA). The c-fos and a c-jun probes were obtained commercially from Oncogene Research Products (Cambridge, MA). cDNA inserts were excised using appropriate restriction enzymes, isolated by electrophoresis through 1% agarose onto DEAE-membrane (NA-45; Schleicher & Scheull) (45), and purified by using the GeneClean II Kit (BIO 101 Inc., La Jolla, CA). cDNA was labeled with 32 P using a commercial random primer kit (Life Technologies, Inc.) according to the manufacturer's instructions.
Measurement of Transcription Rate-ERCC-1 transcription rate was measured using a modification of previously described nuclear run-on analysis (46). A2780/CP70 cells grown to ϳ90% confluence were lysed in 10 ml of lysis buffer, containing 10 mM Tris (pH 8.0), 2.5 mM MgCl 2 , 0.25% Triton X-100, 0.3 M sucrose, and 1 mM dithiothreitol, and nuclei were collected by centrifugation for 5 min at 500 ϫ g. Isolated nuclei were incubated with 250 Ci of [␣-32 P]UTP (NEN Life Science Products, Wilmington, DE) for 30 min at 37°C. cDNA for ERCC-1 and GAPDH, or vector DNA (the plasmid without ERCC-1 cDNA insert) (5 g of DNA per blot) used in the run-on assay was heat denatured and transferred to supported nitrocellulose (Life Technologies, Inc.) by vacuum filtration using a 24-well manifold (Hybri-Dot; Life Technologies, Inc.). The membrane was rinsed with 6 ϫ sodium chloride-sodium citrate (SSC), air dried, and baked at 80°C for 2 h in a vacuum oven. Membranes were prehybridized for 1 h at 42°C in 50% formamide, 6 ϫ SSC, 5 ϫ Denhardt's solution, 0.5% sodium dodecyl sulfate, and 100 g/ml denatured salmon testes DNA and then hybridized at 42°C for 3 days in prehybridization buffer containing run-on reaction mixtures adjusted to equalize radioactivity added in all reactions. Membranes were extensively washed with increasing stringency and then treated with 1 g/ml ribonuclease for 30 min at 37°C. Membranes were air dried, exposed to XAR-5 film, and quantitation of the results was achieved by densitometric scanning normalized to the signal for GAPDH.
Measurement of mRNA Stability-A standard technique for measuring stability of labile transcripts was used (47). Cells were incubated with fresh drug-free medium for 36 h following treatment with 40 M cisplatin for 1 h or left without treatment as control, after which ␣-amanitin (5 g/ml) or actinomycin D (5 g/ml) was added to the treated and control flasks. Total RNA was isolated at the time of ␣-amanitin or actinomycin D addition or at different times thereafter. Northern blot analysis was performed to determine mRNA levels.
Whole Cell Extract Preparation and Western Immunoblot Analysis-To prepare whole cell lysates, 2 ϫ 10 7 cells were washed 3 times in ice-cold phosphate-buffered saline, and resuspended in 500 l of buffer containing 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 0.25% sodium deoxycholate, 1% Nonidet P-40, 1 mM EDTA, 1 g/ml leupeptin, 1 g/ml pepstain, 0.5 mM phenylmethylsulfonyl fluoride, 1 mM sodium orthovanadate, 1 mM sodium fluoride at 4°C. Lysates were sheared through a 21-gauge needle and clarified at 4°C by microcentrifugation. Protein content in the supernatants was determined by means of the BCA protein assay (Pierce, Rockford, IL) using bovine serum albumin as the standard.
The cell extract proteins (25 g) were loaded on a 12% Tris glycine gel (Novex, San Diego, CA), electrophoresed, and transblotted to a Protran (pure nitrocellulose) membrane (Schleicher & Schuell) by the procedure described by Towbin et al. (48). The blot was then rinsed with TBS, and incubated with a blocking buffer (1 ϫ TBS, 0.1% Tween 20 with 5% (w/v) nonfat dry milk) overnight at 4°C. After rinsing with TBS/Tween (0.1%) three times, the blot was incubated with phosho-c-Jun (Ser 63 and Ser 73 ) antibodies (1:1,000 in TBS/Tween with milk) overnight at 4°C on a shaker. The blot was washed three times with TBS/Tween, and incubated with a 1:2,000 dilution of secondary antibody coupled with horseradish peroxidase for 2 h at room temperature on a shaker. The blot was washed three times with TBS/Tween and then incubated with ECL Western blotting detection reagents (Amersham) for 1 min at room temperature. The blot was exposed to x-ray film to visualize the results.
Preparation of Nuclear Extracts-Nuclear extracts were prepared from resting or cisplatin-treated A2780/CP70 cells by a modification of the procedure described by Dignam et al. (49). Cells were harvested by scrapping and washed once with ice-cold phosphate-buffered saline. Preparation of Oligonucleotide Probes-The oligonucleotide sequence used in the following electrophoretic mobility shift assay was based on sequence analysis of the 5Ј-flanking region of ERCC-1 gene as described previously (28). Two duplex 21-bp oligonucleotides which encompassed a ERCC-1 AP-1-like site (5Ј-TCACTGCTGTGTCACCAGCAC-3Ϫ, within Ϫ355 to Ϫ375 from the transcriptional start site at ϩ1 in the ERCC-1 promoter region) (see Fig. 1) and an altered ERCC-1 AP-1-like site (5Ј-TCACTGCTGAGTCACCAGCAC-3Ϫ, Ϫ355 to Ϫ375) were synthesized by Lofstrand Labs Limited (Gaithersburg, MD) and purified by reverse-phase cartridge chromatography. The altered ERCC-1 AP-1like site contains a concensus AP-1 site produced by a 1-bp substitution from thymidine to adenosine as indicated by the underline. The doublestranded oligonucleotides were labeled with [␣-32 P]ATP by phosphorylation with bacteriophage T4 polynucleotide kinase and unincorporated precursors were removed using G-25 Sephadex columns (Boehringer Mannheim). 21-bp oligonucleotides that contained the accepted consensus sequence for AP-1, AP-2, CREB, TFIID, and NFB were obtained from Promega Corp. (Madison, WI) and used in binding or competition studies described below.
Electrophoretic Mobility Shift Assay (EMSA)-The nuclear extracts were analyzed for transcription factor binding activity by gel mobility shift assays. Briefly, nuclear extracts were incubated in a 20-l volume with 1 ϫ binding buffer (1 mM MgCl 2 , 0.5 mM EDTA, 0.5 mM 1,1,1trichloro-2,2-bis(p-chlorophenyl)ethane, 50 mM NaCl, 4% glycerol, 10 mM Tris-HCl, pH 7.5) and 2 g of poly(dI-dC)⅐poly(dI-dC) (Pharmacia, Piscataway, NJ) at room temperature for 10 min. The 32 P-labeled target DNA was then added and the mixture was incubated for 20 -30 min at room temperature. In some experiments, a 50-fold concentration of unlabeled competitor DNA was included in the sample prior to the addition of the radiolabeled probe. After the completion of the binding reaction, 2 l of 10 ϫ gel loading buffer (250 mM Tris-HCl, pH 7.5, 0.2% bromphenol blue, 0.2% xylene cyanol, and 40% glycerol) was added and samples were electrophoresed at room temperature through a 4% nondenaturing polyacrylamide gel in 0.5 ϫ TBE running buffer (0.045 M Tris borate, 0.001 M EDTA, pH 8.0) for 4 h at 100 V which had been pre-run at 100 V for 30 min prior the sample loading. The gels were dried under vacuum and visualized by autoradiography.
Supershift Assay-The nuclear extracts were preincubated with antiserum at room temperature for 20 -30 min before analysis by EMSA as described above. The human anti-Fos and anti-AP-2 antibodies were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA); the affinity purified human anti-Jun antibody has been previously described (50). These sera specifically detect the presence of the corresponding transcription factor and do not interfere with nuclear factor binding.

Up-regulation of ERCC-1 Gene Expression by Cisplatin and
Modulation by Cycloheximide and ␣-Amanitin-The A2780/ CP70 cell line was treated with 40 M cisplatin for 1 h, and the expression of ERCC-1 mRNA was measured at various time points following drug exposure. Cisplatin caused time and dosedependent increases in ERCC-1 mRNA levels (51). Northern analysis showed that ERCC-1 mRNA accumulation was increased by more than 2-fold as early as 6 h after incubation with 40 M cisplatin and eventually attained a peak level of 5.5-fold increase at 24 -48 h after cisplatin administration. Dose-response experiments showed that the effect of cisplatin was maximal at 40 -80 M with about a 6-fold increase in the ERCC-1 mRNA level. The ERCC-1 increase was not associated with changes in the steady-state levels of GAPDH mRNA in A2780/CP70 cells.
We next examined the importance of protein synthesis in the observed increases in ERCC-1 mRNA. The effect of cycloheximide (a protein synthesis inhibitor) was investigated. When the cells were cultured in the presence of cycloheximide (10 M) after a 1-h exposure to cisplatin, the expected increase in ERCC-1 mRNA did not occur (Fig. 2, lanes 2 and 5). Cycloheximide alone had little effect on baseline ERCC-1 mRNA levels (Fig. 2, lanes 1 and 6), suggesting that on-going protein synthesis is required for cisplatin-dependent ERCC-1 mRNA induction. To determine whether the cisplatin effect on ERCC-1 expression was at the transcriptional level, cisplatin-induced cells were cultured with the transcriptional inhibitor ␣-amanitin at a concentration of 5 g/ml. Again the expected increase in ERCC-1 was not seen (Fig. 2, lane 3). The same results were observed when another transcriptional inhibitor (actinomycin D) was used at a concentration of 5 g/ml (data not shown). Moreover, ␣-amanitin repressed the basal expression of ERCC-1 mRNA (Fig. 2, lane 4). These data suggest that ERCC-1 mRNA induction by cisplatin is transcriptionally regulated. Since ␣-amanitin has been shown to be a specific inhibitor for RNA polymerase II (52), this suggests that these genes are normally transcribed by RNA polymerase II.
Transcriptional Regulation of the ERCC-1 Gene by Cisplatin-To determine more directly whether cisplatin can induce the transcription of the ERCC-1 gene, in vitro transcript elongation (nuclear run-on) assays were performed using purified A2780/CP70 cell nuclei. As seen in Fig. 3, the transcription rate more than doubled by 2 h after incubating A2780/CP70 cells with 40 M cisplatin. Transcription activity increased to about 4 times control (untreated cells) by 4 h and decreased gradually afterward, but still remained at a relatively higher level for up to 12 h of incubation after cisplatin. No signal was detectable when the "run-on" analysis was carried out using the plasmid vector, which is serving as a specific binding control for ERCC-1 cDNA in this experiment (Fig. 3). Transcription of the ERCC-1 gene in these assays was blocked by 5 g/ml ␣-amanitin (data not shown).
Effect of Cisplatin on the Turnover of ERCC-1 mRNA-It is possible that the increase in the mRNA level reflects a decreased rate of mRNA degradation. This possibility is supported by the experiment shown in Fig. 4, in which the half-life of ERCC-1 mRNA, as determined by ␣-amanitin chase (5 g/ ml), is seen to be prolonged by cisplatin. Thus, it appears that both increased ERCC-1 gene transcription and increased ERCC-1 mRNA stability underlie the increased steady-state ERCC-1 mRNA levels in cisplatin-treated A2780/CP70 cells.
Induction of Nuclear Factor AP-1 Binding Activity by Cisplatin Exposure-As a step toward determining how the ERCC-1 gene is regulated, we examined the effect of cisplatin on the expression of AP-1 in A2780/CP70 cells. As shown in Fig. 5, cisplatin exposure induced transient expression of c-fos and c-jun mRNA in a time-dependent manner with peak levels of 5-fold increase at 1-2 h in this system. We next determined whether cisplatin exposure induced an increase in c-Jun protein content or its level of protein phosphorylation in these cells, by Western blot (immunoblot) (Fig.  6). Consistent with the Northern (RNA) analysis (Fig. 5), expression of c-Jun protein was evident at 1 h and peaked at 3 to 5 h with an approximate 3.5-fold increase after the cisplatin exposure (Fig. 6). Similarly, levels of protein phosphorylation of c-Jun at serine residues 63 and 73 increased also under the same conditions (Fig. 6). However, the change was much more pronounced. Cisplatin caused a 3.5-fold increase in c-Jun protein level, but a 14-fold increase in c-Jun phosphorylation (Fig.  6, panels B and C).
On the basis of these findings, it was of interest to examine whether cisplatin treatment may result in altered amounts of transcription factors capable of binding the AP-1 site. We analyzed the 5Ј-flanking region of the ERCC-1 gene (28) and found that it contained one region with high homology to AP-1-binding sites. An AP-1-like site located between Ϫ361 and Ϫ367 bp in the ERCC-1 promoter consisted of TGTGTCA, which is shown schematically in Fig. 1. No concensus sequence for AP-2, CREB, or NFB was found in this region. We initially evaluated the AP-1 element and its transcription factor. We constructed two oligonucleotides of 21 bp (Ϫ355 to Ϫ375), as described above, encompassing this potential binding region and performed EMSA with these oligonucleotides. Induction of AP-1 nuclear factor binding to ERCC-1 AP-1-like site (TGT-GTCA) was found as an early event immediately after cisplatin exposure, and binding activity peaked at 3 h and decreased to the baseline level at about 8 h (Fig. 7). The same results were observed with a ERCC-1 AP-1-like site was altered to contain a consensus AP-1 site (TGAGTCA) and an oligonucleotide which contained the consensus AP-1 sequence (5Ј-CGCTTGAT-GAGTCAGCCGGAA-3Ј) of the collagenase promoter (Promega) were utilized in this experiment (data not shown). To ascertain whether the binding observed was consistent with transcription factor binding to a classical AP-1-binding sequence, we used an oligonucleotide containing the accepted concensus sequence for AP-1 as a competitor in the EMSA. This oligonucleotide at 50-fold excess abolished binding to the experimental 32 P-labeled AP-1-like site (Fig. 7). Same results were observed when a 21-bp oligonucleotide containing the ERCC-1 AP-1-like site was used as a competitor at identical concentrations (Fig. 7). In contrast, TFIID (Fig. 7) or other oligonucleotides used included AP-2, CREB, and NFB did not (data not shown).
To characterize further the proteins binding to the AP-1-like site in cisplatin-treated cells, we used supershift assays with antibodies specific to Fos, Jun, and AP-2. The results (Fig. 8) show that antibodies to Fos or Jun, respectively, cause an altered gel migration pattern for the AP-1⅐DNA complex (the two far right lanes). Antibody to AP-2 protein does not cause such a supershift. Data presented in Fig. 8 confirm the data above, suggesting that AP-1 is the relevant DNA binding factor which is composed of Fos/Jun heterodimers. Collectively, these results suggest that both increased AP-1 levels and increased levels of Jun phosphorylation contribute to the increased transactivation of ERCC-1 expression in cisplatin-treated A2780/ CP70 human ovarian carcinoma cells. DISCUSSION ERCC-1 is a single-stranded DNA endonuclease (53) which forms a tight heterodimer with xeroderma pigmentosum complementation group F (54,55). Its role in NER is to incise DNA on the 5Ј side of a lesion such as platinum-DNA adduct (53-55). Therefore, overexpression of ERCC-1 and other NER enzymes during ovarian cancer chemotherapy with cisplatin appears to be implicated in the formation of cellular and clinical drug resistance (25). By contrast, repair of cisplatin-DNA adduct does not occur in the absence of a functional ERCC-1 (15). Thus, understanding the mechanism of regulation and control of ERCC-1 expression in ovarian tumors is of pathophysiological importance. In the present study, we show that cisplatin induces ERCC-1 up-regulation in the A2780/CP70 human ovarian carcinoma cell line and that this induction by cisplatin is regulated at both the transcriptional and posttranscriptional levels. Our results also indicate that the binding activity for transcription factor AP-1 is induced in response to cisplatin in our system.
Protein kinases regulate signaling pathways for a broad spectrum of cellular processes, including differentiation, oncogenesis, and damage response (56). There is considerable evidence that one or more protein kinases are involved in the Fresh medium was then replaced, and RNA was isolated immediately (0 h) or at different times after 1 h exposure to the drug and analyzed by Northern blotting as described under "Experimental Procedures." Cells without treatment of cisplatin were controls (ctrl). Total cellular RNA (30 g) was blotted and hybridized with a 32 P-labeled c-fos or c-jun probe. Loading of RNA was monitored by rehybridization to labeled GAPDH probe. c-fos and c-jun band densities were quantified by densitometry and expressed as a ratio to GAPDH, respectively, and these values are shown graphically in panel B. Ctrl, untreated cells. activation of gene expression in response to DNA damage, and JNK appears to be one such kinase.
JNK is part of a protein kinase cascade in the Ras pathway that eventually leads to increased phosphorylation of the transcription factor c-Jun. Recent work has shown that cellular damage induced by DNA damaging agents, including UV-C radiation (39), ionizing radiation (57), alkylating agents (39,40,58), and cisplatin (34,35), results in activation of the JNK/SAPK pathway involving the transcription factor AP-1. This signal transduction pathway has been reported to protect against cisplatin-induced DNA damage and that this response is required for DNA repair and survival following cisplatin treatment (34). Furthermore, inhibition of this pathway in cells modified by overexpression of a dominant negative mutant of c-Jun, blocks DNA repair, and leads to decrease in viability following treatment with cisplatin (34). These reports suggest that the Ras/JNK pathway may mediate a physiological response to DNA damage such as induction of one or more DNA repair enzymes. However, numerous protein constituents of a complex enzyme system are likely involved in NER, and which of these constituents might be the downstream targets of Ras/ JNK cascade-dependent events and mediate protection against cisplatin-induced damage are currently not known.
In this study, we found that cisplatin treatment of A2780/ CP70 cells led to an increase in ERCC-1 mRNA expression and in nuclear AP-1 binding activity in a time-and dose-dependent manner. In vitro run-on assays in nuclei isolated from control and cisplatin-stimulated A2780/CP70 cells revealed that the appearance of ERCC-1 mRNA occurred in large part because of an increase in the level of de novo transcription. The induction of AP-1 binding activity was selective, insofar as basal DNA binding activities for the inducible transcription factors AP-2, CREB, and NFB were not affected by cisplatin at the doses used and the time points examined in this study. 2 Although our analysis of trans-acting DNA-binding proteins is limited, these findings suggest that cisplatin has a differential effect on the inducible transcription factor activity, and AP-1 is involved, at least partially, in the transactivation of the ERCC-1 gene expression in our system. Evidence for the presence of Jun and Fos proteins in cisplatin-induced AP-1 binding activity was demonstrated by highly specific anti-c-Jun and anti-c-Fos antibodies in supershift analysis of the DNA binding activity in nuclear extracts from A2780/CP70 cells exposed to cisplatin. The ability of Jun/AP-1 to activate transcription is modulated through phosphorylation of the NH 2 -terminal transactivating domain by the JNK. We suspected, therefore, that JNK activity might be affected by cisplatin. In support for this hypothesis, phosphorylation of the JNK substrate c-Jun at serine residues 63 and 73 was rapidly increased following cisplatin exposure. Thus, Jun protein activity is both transcriptionally and posttranscriptionally activated in cisplatin-treated cells.
Overexpression of Ras proteins has been associated with resistance to chemotherapeutic agents and radiation (59 -61). Sklar (59) initially reported that NIH3T3 cells transfected by either the normal or mutant c-Ha-ras oncogene were significantly more resistant to cisplatin than control cells. This study has been confirmed in cisplatin-resistant human cells in vitro and from patients (60,62,63). Moreover, protein kinase inhibitors are able to enhance the cytotoxicity of cisplatin. Although the detailed mechanisms responsible for these resistance phenomena are not entirely clear, evidence is accumulating that Ras transactivates a set of genes coding for enzymes or proteins involved in DNA repair and drug detoxification which affect the sensitivity of cells to cisplatin. These proteins may include AP-1, the products of c-jun and c-fos genes (64), and possibly some proteins in NER. This is supported by the evidence that overexpression of wild-type c-Jun is associated with cisplatin resistance (34), whereas dominant negative Jun expressing cells were inhibited in repair of cisplatin adducts and are sensitized to the cytotoxic effects of cisplatin (34). In addition, increased levels of xeroderma pigmentosum complementation group A and ERCC-1 have been correlated to cis-platinum 2 Q. Li and E. Reed, unpublished observations. FIG. 7. EMSA of AP-1 binding activity in nuclear extracts from cisplatin-treated and untreated A2780/CP70 cells. Cells were incubated with 80 M cisplatin for 1 h. Fresh medium was then replaced, and nuclear extract was prepared from cells harvested at different times after drug removal. EMSAs were performed by using a 32 Plabeled synthetic double-stranded oligonucleotide containing the ERCC-1 promoter sequence of AP-1-like binding site as probe. The protein-DNA complexes formed are indicated (arrow). The unbound (free) probe in the gel is indicated at the bottom. Cisplatin-inducible DNA binding activity at 1 to 3 h following the drug exposure was abolished by competition with a 50-fold molar excess of unlabeled natural AP-1 oligonucleotide (lanes labeled AP-1 competitor) and by an unlabeled oligonucleotide containing the ERCC-1 promoter sequence of AP-1-like binding site (lanes labeled AP-1-like competitor), but not by an identical concentration of TFIID oligonucleotide (lanes labeled TFIID competitor).
FIG. 8. Supershift assay of binding activity in cisplatin-treated A2780/CP70 cell nuclear extracts to experimental ERCC-1 AP-1like binding site. The use of anti-Fos or anti-Jun before EMSA is seen to result in a super-retardation in band mobility consistent with the formation of protein-antibody (Ab)-DNA complexes (arrow with AP-1-Ab/DNA complex), whereas a supershift in band mobility is not seen when anti-AP-2 antibody was used before EMSA (lanes labeled anti-AP-2). resistance (25). Our data in this work suggest that AP-1 may directly modulate ERCC-1 gene expression through some specific targets of its cis-regulatory elements. That would place this critical NER protein under the influence of the Ras/JNK pathway of signal transduction. This potentially opens the door for molecular based strategies to control or modulate DNA repair activity through direct influence on this specific signal pathway, or through direct influence on AP-1.