Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Alkylating chemotherapeutic agents, such as melphalan, carboplatin, and cisplatin, show variable efficacy against human tumors. To increase chemotherapy efficacy against human cancer it is desirable to increase dose intensity at the tumor without increasing the toxicity of chemotherapy side effects. Positive or negative modulation of intracellular and extracellular thiol levels, particularly glutathione, may provide the means to enhance chemotherapy activity at the tumor and protect against toxicity in normal tissues.

Glutathione is an endogenous cysteine-containing tripeptide important for chemotherapy detoxification through a number of mechanisms, including antioxidant and free-radical scavenging activity (Zhang et al., 1998), DNA repair (Chen and Zeller, 1991; Yen et al., 1995), conjugation of cellular toxins (Gamcsik et al., 1997), and pumping toxic chemotherapeutics out of cells via the multidrug resistance-associated proteins (Barrand et al., 1997). Elevated intracellular glutathione is associated with resistance to chemotherapeutic agents such as cisplatin and melphalan in tumors and tumor cell lines (Zhang et al., 1998; Vukovic and Osmak, 1999). Reducing glutathione levels may therefore be a means of enhancing the response of tumor cells to alkylating agents and improving the efficacy of chemotherapy.

Depletion of cellular glutathione can be achieved through inhibition of the rate-limiting enzyme in glutathione biosynthesis, -glutamylcysteine synthetase, by use of the highly specific agent L-buthionine-[S,R]-sulfoximine (BSO) (Griffith, 1982; Ali-Osman et al., 1996). BSO treatment reduces glutathione levels and enhances chemotherapeutic activity in cultured cells (Hamilton et al., 1985; Ali-Osman et al., 1996; Anderson et al., 1999b; Iida et al., 1999) and in animal models (Ozols et al., 1987; Vahrmeijer et al., 1999b). Clinical trials of BSO pretreatment followed by melphalan chemotherapy demonstrate that BSO has low toxicity alone, but in combination with melphalan can enhance toxicity, particularly myelosuppression (Bailey et al., 1994; O'Dwyer, 1996; Anderson et al., 1999a). In one clinical study, BSO enhanced melphalan efficacy in patients with neuroblastoma, even in patients who had recurred after bone marrow transplantation and high-dose melphalan (Anderson et al., 1999a). BSO cytoenhancement of other alkylating chemotherapeutics, such as cisplatin or carboplatin, has not been evaluated in the clinic.

It may be possible to reduce the toxicities of DNA-alkylating drugs such as carboplatin or melphalan by using sulfur-containing chemoprotective agents (thio, thiol, and thioether compounds) to mimic one or many of the activities of glutathione (Dedon and Borch, 1988; Links and Lewis, 1999; Safirstein et al., 2000). Such agents may also be effective against the enhanced toxicity due to BSO reduction of glutathione. Clinically relevant compounds that may provide chemoprotection include sodium thiosulfate, N-acetylcysteine, D-methionine, and glutathione ethyl ester. Proposed mechanisms of action for these reactive sulfur-containing agents include chemical modification (Dedon and Borch, 1988; Gamcsik et al., 1997), elevating intracellular glutathione levels (Anderson et al., 1990; Yim et al., 1994), multidrug resistance activation (Ishikawa and Ali-Osman, 1993; Zhang et al., 1998), or free-radical scavenging (Jarvinen et al., 2000; Safirstein et al., 2000; Sha and Schacht, 2000).

Despite the potential benefits of chemoprotection, such agents have relatively little clinical use due to concerns of impaired chemotherapeutic efficacy. Interactions of chemoprotectants with chemotherapy efficacy may be avoided by separating treatments in time or space. The two compartments created by the blood-brain barrier, intra-arterial versus intravenous administration, and separation by time have recently demonstrated the potential of this approach. In preclinical and clinical trials, spatial and temporal separation of platinum chemotherapy and the thiol sodium thiosulfate results in marked chemoprotection against ototoxicity without impairing tumor cytotoxicity (Neuwelt et al., 1996, 1998; Muldoon et al., 2000).

The purpose of these experiments was first to evaluate BSO cytoenhancement of the platinum-based alkylating agents carboplatin and cisplatin in comparison with melphalan, and second to test potentially important thiol agents for chemoprotection of enhanced chemotherapy toxicity. Additionally, we evaluated whether cytoenhancement and chemoprotection involved an apoptotic mechanism.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents. Carboplatin (Paraplatin), cisplatin (Platinol), and etoposide phosphate (Etopophos) were obtained from Bristol-Meyers Squibb (New York, NY), and melphalan (Alkeran) was obtained from Glaxo Wellcome (Research Triangle Park, NC), all via the Oregon Health Sciences University hospital pharmacy. Sodium thiosulfate, D-methionine, and glutathione ethyl ester were obtained from Sigma Chemical Co. (St. Louis, MO). Acetylcysteine sterile solution was obtained from American Reagent Laboratories (Shirley, NY), via the Oregon Health Sciences University hospital pharmacy. BSO was supplied by the National Cancer Institute, Bethesda, MD.

Tissue Culture. The cells used in these experiments were the B.5 LX-1 human small cell lung carcinoma (SCLC) cell line and the GM294 human fibroblast cell strain. The B.5 LX-1 cell line is a clonal line derived from the LX-1 parental cells (originally obtained from Mason Research Institute, Worcester, MA). These cells were maintained as a free-floating cell suspension in spinner flasks, in medium RPMI-1640 supplemented with 12% heat inactivated fetal bovine serum (Irvine Scientific, Santa Anna, CA) plus gentamicin, penicillin, and streptomycin. The GM294 fibroblasts were obtained from the NIGMS Human Mutant Cell Repository (Bethesda, MD), and were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and antibiotics.

Cytotoxicity Assay. Live cell number was evaluated with the Cell Proliferation Assay kit from Chemicon International (Temecula, CA). This colorometric assay is based on the cleavage of the tetrazolium salt WST-1 to formazan by cellular mitochondrial dehydrogenases. Cells were seeded to 96-well tissue culture plates at 1 × 10⁴ cells/well, four wells per condition. After growth for 20 to 24 h with or without BSO, chemotherapy and/or chemoprotective agents were added for an additional 44 to 48 h. The WST-1 reagent was then added for 2 h and absorbance at 450 nm was determined using a microplate reader. Because the thiol agents can interfere with the colorometric assay (this is particularly a problem with other tetrazolium reagents), blanks included experimental agents and cells dissolved by addition of sodium dodecyl sulfate to a concentration of 0.5%. This assay approximates linearity (absorbance versus cell number) in the range of 10³ to 10⁵ cells. Cell viability was also assessed by trypan blue exclusion.

Caspase-2 Enzymatic Assay. Apoptosis induction was evaluated by measuring caspase-2 protease activity using a kit from R&D Systems (Minneapolis, MN). This assay is based on the cleavage of a caspase-specific peptide that is conjugated to the color reporter molecule p-nitroanilide. Cells were seeded in 6-cm tissue culture plates at 2 × 10⁶ cells/plate, three plates per condition. After growth for 18 to 24 h with or without BSO, chemotherapy and/or chemoprotective agents were added for an additional 8 to 24 h. Protein content in cell lysates was determined with the bicinchoninic reagent kit (Pierce, Rockford, IL) and 100 to 200 µg was used for each assay. Caspase-2 reporter absorbance at 405 nm was determined using a microplate reader.

In Situ Apoptosis Detection. DNA fragmentation was detected with the terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) method. Studies were performed using the apo-TACS TUNEL assay kit from R&D Systems or the ApopTag Plus Peroxidase kit from Intergen (New York, NY). In both assays labeled nucleotides are incorporated onto DNA fragments using terminal deoxynucleotidyl transferase, followed by immunocytochemical detection of the labeled DNA (bromo-deoxyuridine for apo-TACS, digoxigenin for ApopTag). Cultured cells treated with chemotherapy for 12 to 24 h, with or without BSO, were suspended and fixed in formalin, and then 1 × 10⁵ cells were dried onto gelatin-treated slides for use in the TUNEL assays.

Statistical Analysis. Cytotoxicity assay data are expressed as mean ± standard deviation as a percentage of untreated control values, with n = 4 independent samples per condition, whereas caspase activity assays had n = 3 per group. Every cytotoxicity and caspase experiment was performed at least twice with similar results. For the cytotoxicity and caspase assays, Student's t test was used to determine differences between individual points, using Microsoft Excel software. Half-maximal effective concentrations (EC₅₀) and standard deviation were determined using Prism (GraphPad Software, San Diego, CA). TUNEL staining was performed on single samples, and each experiment was performed at least in triplicate. Staining with diaminobenzidine was evaluated visually. The number of stained versus unstained cells was counted in one visual field infour areas of each slide, one in each quadrant, with a minimum of 50 cells/visual field.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

BSO Enhancement of Chemotherapy Cytotoxicity. The dose response for enhanced cytotoxicity was evaluated in B.5 LX-1 human SCLC cells treated with increasing concentrations of BSO for 18 h before addition of chemotherapy. Cells then received approximately half-maximal cytotoxic doses of melphalan (10 µg/ml), carboplatin (100 µg/ml), or cisplatin (7.5 µg/ml) and cytotoxicity was evaluated after 48 h with the WST-1 colorometric assay. Figure 1 shows that melphalan was significantly more sensitive to BSO than were cisplatin and carboplatin. For melphalan, the half-maximal dose of BSO for enhancement was less than 10 µM, whereas for carboplatin the half-maximal dose of BSO was 37.7 ± 12.4 µM and for cisplatin it was 21.8 ± 6.2 µM. At 10 µM BSO the enhancement for melphalan cytotoxicity was 73.1 ± 4.9% of maximum, whereas cisplatin and carboplatin were only 38.3 ± 1.2 and 33.3 ± 8.5% of maximal enhancement, respectively (P < 0.001 compared with melphalan). BSO was significantly more enhancing for melphalan than for cisplatin at 10 µM (P < 0.001), 20 µM (P < 0.001), and 50 µM (P < 0.005). All three chemotherapeutics were near maximal enhancement at 100 µM BSO, and this concentration was used in all other experiments. At this dose, BSO alone did not decrease SCLC cell viability (mean 103.4 ± 6.7% of untreated control cell number, n = 20).

View larger version (26K): [in this window] [in a new window]   Fig. 1.   Dose response for BSO enhancement of cytotoxicity. Cytotoxicity was assessed in cultured LX-1 SCLC cells, 1 × 104 cells/well in 96-well plates, using the WST colorometric assay. BSO cytoenhancement consisted of preincubation with BSO at the indicated concentration for 18 h. Chemotherapeutic agents were then added at approximately half-maximal concentration (, melphalan = 10 µg/ml; , carboplatin = 100 µg/ml; , cisplatin = 7.5 µg/ml). Data are expressed as the percentage maximal enhancement, with zero corresponding to no BSO and 100% corresponding to the enhancement in cells treated with 500 µM BSO. Each point represents the mean ± standard deviation of four wells. Significance is indicated by **P < 0.01 and ***P < 0.001 comparing melphalan enhancement with cisplatin enhancement.

Chemoprotection against Cytotoxicity. The dose response for rescue from chemotherapy cytotoxicity was evaluated for four different small molecular weight sulfur-containing chemoprotectants. Each chemotherapeutic agent was used at a concentration affording approximately 90% lethality in the absence of BSO (20 µg/ml melphalan, 200 µg/ml carboplatin, 15 µg/ml cisplatin). Overall, N-acetylcysteine was the most effective of the thiol agents tested, on a microgram per milliliter basis. The concentration dependence for protection with N-acetylcysteine in comparison to D-methionine is shown in Fig. 2, and Table 1 shows the EC₅₀ for protection afforded by each protective agent. The cytotoxicity of each alkylator was reduced by 75 to 90% by concurrent administration of N-acetylcysteine (Fig. 2A), but N-acetylcysteine was more active against melphalan (EC₅₀ = 74 ± 18 µg/ml) than the platinum agents (Table 1). In contrast, D-methionine did not protect against melphalan toxicity at the doses tested (50-1000 µg/ml, Fig. 2B), although it was highly protective against cisplatin toxicity, with a half-maximal concentration of 140 ± 41 µg/ml (Table 1). The maximum magnitude of protection was variable between experiments, ranging from 70 to 100% protection, and protection was consistently less for carboplatin than for cisplatin or melphalan. All agents tested required a significantly higher dose to protect against carboplatin than against cisplatin or melphalan. On a microgram per milliliter basis, glutathione ethyl ester was the least effective protective agent.

View larger version (31K): [in this window] [in a new window]   Fig. 2.   Protection against chemotherapy cytotoxicity. A, dose response for N-acetylcysteine chemoprotection. B, dose response for D-methionine chemoprotection. Cytotoxicity was assessed in cultured LX-1 SCLC cells, 1 × 104 cells/well in 96-well plates, using the WST colorometric assay. Cells were treated with approximately 90% lethal dose of chemotherapy (, melphalan = 20 µg/ml; , carboplatin = 200 µg/ml; , cisplatin = 20 µg/ml). Chemoprotectant was added at the indicated concentration of N-acetylcysteine (A) or D-methionine (B) either alone () or immediately following chemotherapy. Data are expressed as the percentage of live cells compared with untreated control samples (without chemotherapy) and each point represents the mean ± standard deviation of four wells.

Cytoenhancement and Chemoprotection in Combination. The effects of BSO cytoenhancement and thiol chemoprotection on the dose-response relationships for cytotoxicity of the alkylating chemotherapeutics were evaluated in the B.5 LX-1 cells. BSO cytoenhancement consisted of preincubation with 100 µM BSO for 18 to 24 h before addition of chemotherapy, and rescue consisted of 1000 to 2000 µg/ml thiol chemoprotectant added immediately after chemotherapy. In this experimental paradigm, BSO consistently decreased the EC₅₀ for cytotoxicity (Fig. 3A; Table 2) and increased the maximum degree of toxicity. The specific case of carboplatin and N-acetylcysteine is shown in Fig. 3A. Glutathione depletion with BSO increased carboplatin cytotoxicity, reducing the EC₅₀ by 48% (P < 0.01). As detailed in Table 2, similar BSO cytoenhancement was found with melphalan (53% reduction of EC₅₀, P < 0.001), whereas the EC₅₀ for cisplatin was reduced only 29% (P < 0.05). Chemoprotection with N-acetylcysteine blocked carboplatin toxicity as well as BSO-enhanced cytotoxicity. Similar chemoprotection was found with additional thiol agents, sodium thiosulfate and glutathione-ethyl ester, but D-methionine was only effective against the platinum agents.

View larger version (30K): [in this window] [in a new window]   Fig. 3.   Cytoenhancement and chemoprotection. A, effect of BSO and N-acetylcysteine on carboplatin cytotoxicity. B, effect of BSO and N-acetylcysteine on etoposide phosphate cytotoxicity. Cytotoxicity was assessed in cultured LX-1 SCLC cells, 1 × 104 cells/well in 96-well plates, using the WST colorometric assay. BSO cytoenhancement consisted of preincubation at 100 µM BSO for 18 h. Chemoprotective agent was added immediately after chemotherapy. The experimental conditions were dose responses for chemotherapy (carboplatin or etoposide phosphate) alone (), BSO cytoenhancement (), N-acetylcysteine rescue (1000 µg/ml N-acetylcysteine, ), or BSO cytoenhancement and N-acetylcysteine rescue (). Data are expressed as the percentage of live cells compared with untreated control samples (without chemotherapy) and each point represents the mean ± standard deviation of four wells.

View larger version (51K): [in this window] [in a new window]   Fig. 4.   Cytoenhancement and chemoprotection in fibroblasts. Cytotoxicity was assessed in GM294 human fibroblasts, 1 × 104 cells/well in 96-well plates, using the WST colorometric assay. Cells were pretreated with or without BSO, 100 µM for 18 h before addition of chemotherapeutics (melphalan = 10 µg/ml, carboplatin = 100 µg/ml, cisplatin = 7.5 µg/ml, etoposide phosphate = 100 µg/ml). The experimental conditions were chemotherapy either alone (), or with N-acetylcysteine rescue (1000 µg/ml N-acetylcysteine, ), BSO cytoenhancement (), or BSO cytoenhancement and N-acetylcysteine rescue (). Data are expressed as the percentage of live cells compared with control samples (without chemotherapy) and each point represents the mean ± standard deviation of four wells.

Time Dependence for Chemoprotectant Rescue from Chemotherapy Cytotoxicity. We evaluated how long the addition of chemoprotectant could be delayed after treatment with chemotherapy and remain effective. Cells were treated with doses of chemotherapy providing approximately 90% lethality, for melphalan (20 µg/ml), carboplatin (200 µg/ml), or cisplatin (15 µg/ml). The thiol chemoprotectants were added either concurrently with chemotherapy or up to 8 h after chemotherapy. For melphalan, chemoprotection was reduced if administration of sodium thiosulfate was delayed for 2 h, whereas sodium thiosulfate was still protective for the platinum chemotherapeutics if delayed up to 4 h after treatment (Fig. 5). Similarly, delayed administration of N-acetylcysteine and glutathione ethyl ester reduced their protective activity against melphalan cytotoxicity, whereas both agents maintained protective activity against platinum cytotoxicity (data not shown). In a separate experiment, we found that all three agents were completely protective if added within 1 h of melphalan, rather than 2 h as shown in Fig. 5. Chemoprotection was not effective against etoposide phosphate cytotoxicity at any time point.

View larger version (49K): [in this window] [in a new window]   Fig. 5.   Time dependence for rescue of chemotherapy cytotoxicity. Chemotherapy cytotoxicity was assessed in cultured LX-1 SCLC cells (1 × 104 cells/well in 96-well plates) using the WST colorometric assay. Cells were treated with melphalan at 20 µg/ml, carboplatin at 200 µg/ml, cisplatin at 10 µg/ml, or etoposide phosphate at 200 µg/ml. Cells then received either no protectant (), or sodium thiosulfate, 2000 µg/ml, added immediately (), 2 h (), or 4 h () after chemotherapy. Data are expressed as the percentage of live cells compared with control samples (without chemotherapy) and each point represents the mean ± standard deviation of four wells.

Effects of Cytoenhancement and Chemoprotection on Apoptosis. Apoptosis was evaluated by measuring caspase-2 enzymatic activity and by in situ TUNEL staining. Treatment of B.5 LX-1 cells with melphalan resulted in an increase in caspase-2 activity that was amplified by BSO pretreatment at low melphalan concentrations (Fig. 6A). The increase in caspase activity was variable between experiments and ranged from 50 to 100% at 7 to 8 h to 250 to 600% at 20 to 24 h after treatment with melphalan. TUNEL staining also demonstrated melphalan-induced apoptosis. In the experiment shown in Fig. 6B, TUNEL staining after melphalan treatment was positive in 29 of 3643 cells, compared with 7 of 4395 cells in the untreated control, and BSO treatment before melphalan increased the positive staining to 800 of 1699 cells. In both the caspase-2 assay (Fig. 6A) and the TUNEL staining assay (Fig. 6B), the effect of melphalan on apoptosis was reduced by the chemoprotectant N-acetylcysteine. In both assays, activity was maximal with low doses of melphalan, or with a 1-h pulse treatment with the doses used in the cytotoxicity assays. Continuous treatment with the cytotoxic dose of melphalan actually reduced caspase-2 activity and TUNEL staining.

View larger version (32K): [in this window] [in a new window]   Fig. 6.   Effect of cytoenhancement and chemoprotection on apoptosis. A, caspase-2 enzymatic activity. B, TUNEL staining for DNA fragmentation. Apoptosis was assessed in cultured LX-1 SCLC cells pretreated for 18 h with or without 100 µM BSO. The experimental conditions were no addition (), melphalan (10 µg/ml, ), or melphalan (10 µg/ml) plus N-acetylcysteine (1000 µg/ml) () for 20 h before harvest. Caspase-2 activity is expressed as percentage activity in untreated control samples (mean ± standard deviation, n = 3). TUNEL staining is expressed as the percentage of cells showing positive staining.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Our studies have demonstrated that BSO treatment enhances the cytotoxicity of carboplatin and cisplatin, in addition to melphalan. Thiol agents protect against the cytotoxicity of alkylating chemotherapeutic agents, and chemoprotection against carboplatin or cisplatin can be delayed for at least 4 h without reduced protective activity. These findings may be clinically applicable.

Cytoenhancement. BSO blocks glutathione synthesis (Griffith, 1982), leading to greater than 80% reduction in glutathione levels in vitro within 8 to 24 h after treatment with 100 µM BSO (Ali-Osman et al., 1996; Pendyala et al., 1997; Vahrmeijer et al., 1999a). Glutathione depletion itself is cytotoxic in some cultured cells, particularly neuroblastoma cells (Anderson et al., 1999b). We found no change in baseline cell viability or apoptosis in response to BSO in SCLC cells.

Chemoprotection. Glutathione ethyl ester, sodium thiosulfate, and N-acetylcysteine all reduced or prevented the cytotoxicity induced by melphalan, cisplatin, and carboplatin, independent of BSO treatment. The cysteine analog N-acetylcysteine was the most effective of the chemoprotectants tested, on a microgram per milliliter basis, in both fibroblasts and tumor cells. This result contrasts with previous reports indicating that N-acetylcysteine was not protective against BSO-enhanced melphalan cytotoxicity (Vahrmeijer et al., 1999a) or cisplatin cytotoxicity (Iida et al., 1999). In vivo, N-acetylcysteine has been shown to be protective against a number of toxic insults (Safirstein et al., 2000), including ifosfamide-induced urotoxicity (Holoye et al., 1983) and contrast-induced nephrotoxicity (Tepel et al., 2000). Thus, N-acetylcysteine may be a safe and effective agent for reducing some of the side effects of alkylating chemotherapy in patients.

Mechanism of Cytoenhancement and Chemoprotection. Glutathione is known to have multiple detoxifying activities, so determination of the mechanism(s) involved in enhancing or protecting different chemotherapeutics is difficult. Glutathione ethyl ester activates all glutathione pathways because it is readily taken up and converted to glutathione intracellularly (Anderson et al., 1990). In mice, glutathione diethyl ester is effective at protecting against the toxicity of cisplatin, and reverses the potentiation by BSO (Anderson et al., 1990). N-Acetylcysteine induces de novo synthesis of glutathione over a period of hours to days (Yim et al., 1994). The limited time frame for N-acetylcysteine chemoprotection against melphalan argues against glutathione biosynthesis as the mechanism of protection. Glutathione can conjugate toxins to either directly inactivate them or direct them to glutathione-dependent transporters such as the multidrug resistance-associated proteins (Barrand et al., 1997). Active pumping of conjugated or unconjugated cisplatin from cells has been shown (Ishikawa and Ali-Osman, 1993; Zhang et al., 1998). Sodium thiosulfate chemoprotection may be primarily through conjugation and inactivation. A high molar ratio of thiosulfate to platinum results in drug neutralization (Dedon and Borch, 1988), and conjugation also occurs between thiosulfate and melphalan (Gamcsik et al., 1997).

Clinical Potential. The possibility of reduced anticancer effect due to chemoprotective agents is a major concern limiting their use (Links and Lewis, 1999). To minimize interactions between chemoprotectants and chemotherapy, the agents should be separated either in time or in space. Two-route administration of sodium thiosulfate with cisplatin, such as intra-arterial cisplatin with i.v. sodium thiosulfate for head and neck cancer, has been used in to provide local chemoprotection while sparing antitumor activity (Robbins et al., 1994). The two-route paradigm actually increased cisplatin antitumor effects against mouse tumors (Iwamato et al., 1984).