Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Oxaliplatin [l-OHP, oxalato(trans-l-1,2-diaminocyclohexane)platinum(II)] is a third generation platinum (Pt) antitumor compound in which diaminocyclohexane (DACH) ligand replaces the amine groups present in cisplatin (Fig. 1) (Chaney, 1995; Raymond et al., 1998). Oxaliplatin has demonstrated a broad spectrum of antitumor activity (Rixe et al., 1996) with a partial or a non-cross-resistance with cisplatin in a wide range of human tumors in vitro and in vivo (Weiss and Christian, 1993; Kelland and McKeage, 1994; Chaney, 1995; Raymond et al., 1998). Ongoing clinical European phase II trials have reported encouraging activity and manageable toxicity in a variety of tumors usually resistant to cisplatin (for review, see Cvitkovic, 1998). Oxaliplatin is now approved, in combination with 5-fluouracil, for the treatment of advanced colorectal cancer in Europe.

View larger version (11K): [in this window] [in a new window]   Fig. 1.   The structures of oxaliplatin and cisplatin.

The critical role of DNA-Pt adducts in the antiproliferative effects, well documented for cisplatin, is generally accepted for all antitumor Pt drugs (for review, see Sanderson et al., 1996). The knowledge of oxaliplatin-induced DNA lesions is largely based on extrapolation of findings for cisplatin and DACH compounds other than oxaliplatin. However, the analogy between oxaliplatin and cisplatin should not be overinterpreted. Oxaliplatin is typically at least as potent as cisplatin in inhibiting the growth of cancer cells (Rixe et al., 1996). Thus, oxaliplatin would be expected to damage DNA to a similar extent as cisplatin. However, various methodologies suggested that oxaliplatin induced fewer lesions in naked and cellular DNA than did equimolar cisplatin (Saris et al., 1996; Woynarowski et al., 1998).

These paradoxical findings suggest that oxaliplatin-induced DNA damage may differ in various aspects from that of cisplatin. Although the structures of diaminocyclohexane (DACH)-Pt DNA adducts formed by oxaliplatin and cis-diammine-Pt DNA adducts (formed by cisplatin) are similar, the bulky DACH moiety that protrudes into the minor groove (Scheeff et al., 1999) may possibly lead to different biological properties of DACH-Pt-DNA adducts (Chaney, 1995; Scheeff et al., 1999). DACH-Pt adducts, for instance, appear to be more effective at inhibiting DNA synthesis (Gibbons et al., 1990, 1991; Schmidt and Chaney, 1993). On the other hand, the possibility that DACH moiety in oxaliplatin affects the localization of drug adducts in DNA has been ruled out because sequence and region specificities of oxaliplatin adducts are similar to those of cisplatin (Woynarowski et al., 1998).

Another unstudied possibility is that oxaliplatin might generate a greater proportion of highly lethal lesions, compensating in that way for the lower overall DNA adduction compared with cisplatin. Whereas intrastrand cross-links are probably the main type of oxaliplatin adducts, infrequent interstrand cross-links (ISC) were also detected with naked DNA (Woynarowski et al., 1998). ISC are regarded as lethal DNA lesions for cisplatin (Bedford et al., 1987; Roberts and Friedlos, 1987; Roberts et al., 1988). However, ISC induction by oxaliplatin in cellular DNA has never been reported. Also, the ability of oxaliplatin to form another likely lesion, DNA-protein cross-links (DPC), remains unknown.

Our current study explored the leads suggesting that oxaliplatin could differ from cisplatin in the ability to damage DNA. The results verify that oxaliplatin, like cisplatin, forms ISC and DPC in tumor cells. However, oxaliplatin-induced total Pt-adducts, ISC, and DPC are significantly lower than the respective lesions induced by equimolar concentrations of cisplatin, despite a similar or greater cytotoxicity of oxaliplatin.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Drugs, Cell Culture, and Cytotoxic Activity. Oxaliplatin was obtained from Sanofi-Synthelabo Research (Great Valley, PA). Cisplatin was purchased from Sigma Chemical Co. (St. Louis, MO). Drug stock solutions were made in water (oxaliplatin) or in saline (cisplatin) and stored at 20°C.

Cellular Pt-DNA Adducts. Platination levels were monitored by atomic absorption as previously described (Gibbons et al., 1990; Schmidt and Chaney, 1993). Pilot studies had shown that the Pt-DNA levels were proportional to the drug concentration used over the range of 50 to 250 µM. For each experiment, cells (10⁶/10 ml medium in replicate 100-mm dishes) were exposed to 100 µM cisplatin or oxaliplatin as indicated in Fig. 2 and Table 2. DNA was purified from each dish separately using the Wizard Genomic DNA Purification Kit (Promega, Madison, WI) and analyzed for Pt levels by AA in triplicate on a Perkin Elmer (Norwalk, CT) Cetus model 560 atomic absorption spectrophotometer with an HGA 500 graphite furnace and an AS-1 autosampler. DNA preparations from different culture dishes and on different days were considered repeats and were as follows: n = 4 for 15/45 min, n = 6 for 1 h, and n = 3 for 4 h. The data for Pt adducts were normalized per 10⁶ base pairs (bp)/10 µM drug, to allow the comparison with other types of DNA lesions (Table 2).

View larger version (26K): [in this window] [in a new window]   Fig. 2.   Pt adducts formed by oxaliplatin (solid bars) and cisplatin (hatched bars) in CEM cells. Cells were treated with 100 µM drugs for 15 min, followed by a 45-min chase in drug-free medium (15/45 min), or for 1 or 4 h of continuous treatment as indicated, followed by DNA purification and determination of DNA platination levels by atomic absorption. The results are reported as means ± S.E.M. of data obtained from three to six replicate cell cultures/DNA preparations and are normalized per 106 bp/10 µM drug. The differences between cisplatin and oxaliplatin were significant (P < .01 for 15/45 min and 1 h and P < .05 for 4-h treatments).

DNA Repair and Host Cell Reactivation Assay. Plasmid reactivation assays were performed using a strategy similar to that described by others (Chao et al., 1991; Ali-Osman et al., 1994). Two plasmids from Promega were used: pGL3 control vector containing the luciferase construct was used as the drug damage reporter, whereas pSV -galactosidase control vector served as an internal control for transfection efficiency. Plasmids were prepared and purified as recommended by Promega. The pGL3 control vector (100 µg/ml) in TE buffer (10 mM Tris, 0.5 mM EDTA, pH 7.4) was treated with 0.25 to 5 µM cisplatin or 0.5 to 50 µM oxaliplatin for 22.5 h at 37°C. For oxaliplatin treatment, the reactions were supplemented with NaCl (100 mM final concentration) to accelerate formation of drug active species. After ethanol precipitation to remove unreacted drug, plasmid DNA was redissolved in TE buffer, and its concentration was quantitated by standard Hoechst 33258 fluorescence assay. Platination levels were also directly assessed in drug-treated plasmids by atomic absorption as described for cellular Pt adducts. One batch of drug-treated plasmids was used in all the transfection experiments.

ISC by Alkaline Sucrose Gradient Sedimentation. The procedure for sedimentation analysis of cellular DNA was used essentially as described elsewhere (Woynarowski et al., 1999). Briefly, cells prelabeled with [¹⁴C]thymidine were treated with drugs and, in some experiments, postincubated in drug-free medium as indicated. Aliquots of 10⁵ harvested cells in 50 mM Tris-HCl, pH 7.5, 10 mM EDTA, 0.2% Triton X-100 were loaded onto preformed alkaline sucrose gradients composed of 0.5 ml of 60% sucrose cushion, 10 ml of 5 to 20% sucrose, and 0.4 ml of lysing layer (1% Sarkosyl, 2.5% sucrose). All the solutions were in a gradient buffer (0.7 M NaCl, 0.3 M NaOH, 0.01 M EDTA). After the sample was applied, an additional volume (200 µl) of lysing solution was laid on the top, and lysis was allowed to proceed for 20 h. Sedimentation was carried out in an SW41 rotor (Beckman Instruments, Fullerton, CA) at 20°C for 20 h at 10,000 rpm (A2780 cells) and 9200 rpm (CEM cells). Gradients were fractionated and the fractions processed as described previously (Woynarowski et al., 1995).

(1)

(2)

DPC. DPC were assayed using the K⁺/SDS precipitation technique. For DPC determinations in intact cells, [¹⁴C]thymidine-prelabeled cells were incubated with drugs as indicated. After drug treatment, cells were lysed, and DNA coprecipitable with proteins was determined as described previously (Woynarowski et al., 1989, 1997). For DPC determination in isolated nuclei, prelabeled cells were lysed in nuclei isolation buffer (2 mM KH₂PO₄, 5 mM MgCl₂, 150 mM NaCl, 1 mM EGTA, pH 6.9) supplemented with 0.3% (v/v) Triton X-100 for 20 min at 4°C, and then centrifuged for 13 min at 300g. The nuclear pellets were resuspended in the isolation buffer at 0.3 to 0.5 × 10⁶ nuclei/ml and incubated with drugs, followed by DPC determination as for intact cells (Woynarowski et al., 1989).

(3)

Chain Elongation Assay. The size distribution of radiolabeled nascent DNA was determined by velocity sedimentation analysis as described previously (Gibbons et al., 1991; Mamenta et al., 1994). Briefly, CEM or A2780 cells were plated on 60-mm dishes at an initial concentration of 5 × 10⁵ cells/dish and uniformly labeled with [¹⁴C]thymidine for 48 h. The old medium was then replaced with a medium containing various concentrations of oxaliplatin or cisplatin followed by a 15-min incubation. The drug-medium was then removed and replaced with a drug-free medium followed by a 30-min incubation. The cells were then pulsed for 15 min with 6 µCi/ml [³H]thymidine (specific activity, 85 Ci/mmol). The radioactive medium was removed, and the cells were harvested into 1 ml of ice-cold harvest buffer (0.1 M NaCl-0.01 M EDTA, pH 8.0). Cells were harvested and lysed and the lysates subjected to sedimentation in alkaline sucrose gradients. The gradients were fractionated, and acid-precipitable ³H counts were determined and normalized to the total amount of proteins in each sample. The percentage of chain elongation (relative to control) was calculated from the areas under the curve in the region of the profiles corresponding to the chain elongation.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Cytotoxic Activities. The cytotoxic activities (GI₅₀ values, drug concentration inhibiting cell growth by 50%) of oxaliplatin and cisplatin against various cell lines differing in cisplatin sensitivity, p53, and mismatch repair status are summarized in Table 1. Oxaliplatin and cisplatin showed similar levels of growth inhibition in CEM cells with mutated p53 with the GI₅₀ ~ 1.1/1.2 and 11/11 µM after continuous drug treatment and a 1-h pulse treatment, respectively. A2780 cells (with wild-type p53) were more sensitive to both drugs than CEM cells, with oxaliplatin being nearly 2-fold more cytotoxic than cisplatin. Furthermore, oxaliplatin was ~2-fold more cytotoxic than cisplatin against the cisplatin-resistant subline A2780PR and 6- to 7-fold more cytotoxic against a mismatch repair-deficient line, HCT116, and against a generally drug-resistant line, HT-29 (Table 1).

Total DNA Platination. Platination levels were used as an overall measure of adduct formation after various incubation conditions in CEM cells (Fig. 2). Oxaliplatin consistently formed significantly fewer adducts than cisplatin at equimolar external concentrations (Fig. 2 and Table 2). For instance, after a 1-h treatment, oxaliplatin and cisplatin formed 0.86 and 1.36 Pt adducts/10⁶bp/10 µM, respectively (Fig. 2 and Table 2). The difference between both drugs remained significant after a 4-h incubation, with the respective platination values of 2.4 versus 3.18 Pt adducts/10⁶ bp/10 µM. Similar differences in oxaliplatin and cisplatin adduct levels were observed in A2780 cells (data not shown).

Repair of Oxaliplatin Lesions-Host Cell Reactivation Assay. To address the possibility that oxaliplatin adducts are less prone to removal by repair processes, we analyzed the reactivation of drug-treated plasmids in several cell lines. The initial levels of plasmid platination showed a good dependence on drug concentration during plasmid treatment ranging from 1.3 ± 0.1 up to 180 ± 2.3 Pt atoms/kbp for 0.5 to 50 µM oxaliplatin and from 1.6 ± 0.1 up to 36.9 ± 0.7 Pt atoms/kbp for 0.25 to 5 µM cisplatin. The detected cisplatin platination levels are in reasonable agreement with literature data for similar treatment conditions (Chao et al., 1991).

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Repair of DNA lesions in oxaliplatin-treated () or cisplatin-treated () plasmids in A2780 (A) and HT29 (B) cells. The luciferase reporter plasmid was treated with various concentrations of drugs as described under Materials and Methods, followed by the determinations of platination levels (abscissa). Cells were cotransfected with the drug-damaged luciferase plasmids and undamaged -galactosidase plasmid. After 48 h, luciferase levels were measured and normalized to their respective -galactosidase levels. Relative luciferase activity is expressed as a percentage of activity of the undamaged control plasmid. Each point represents the mean ± S.E.M. of two to four different transfections of the damaged plasmids.

ISC in Intact Cells. Induction of ISC by oxaliplatin and cisplatin was examined to assess whether the lower total Pt adducts formed by oxaliplatin might be compensated by a greater proportion of highly lethal lesions such as ISC. Sedimentation analysis after a 4-h incubation of CEM cells with 25 µM cisplatin produced the characteristic pattern for interstrand cross-links indicated as shifts in DNA sedimentation profiles toward the bottom of the gradients (fractions 19-24, brackets in Fig. 4). DNA from oxaliplatin-treated cells also revealed fast-sedimenting material, although less prominent than for cisplatin. A similar pattern was observed in A2780 cells (data not shown). These results confirm that oxaliplatin forms interstrand cross-links but at markedly lower levels than equimolar cisplatin.

View larger version (25K): [in this window] [in a new window]   Fig. 4.   ISC induced by cisplatin (A) and oxaliplatin (B) in CEM cells treated continuously with drugs for 4 h. The figure depicts examples of sedimentation on alkaline sucrose gradients of DNA from control cells () and cells treated with oxaliplatin at 25 µM () and cisplatin at 25 µM (). The arrows show the direction of sedimentation, and the brackets show the area of the sedimentation profiles that is indicative of interstrand cross-links. The sedimentation profiles shown were averaged from three to five independent determinations. Average S.E.M. values, omitted for clarity, ranged from 1 to 1.5% of recovered radioactivity.

DPC in Intact Cells. DPC are another type of previously unstudied DNA lesion, in which oxaliplatin might possibly compensate for its lower overall reactivity with DNA, compared with cisplatin. The induction of DPC was analyzed based on direct physical separation of protein-bound and protein-free DNA by the K⁺/SDS precipitation technique (Woynarowski et al., 1989, 1997). After a continuous 4-h treatment, oxaliplatin formed significantly fewer DPC than did cisplatin (Fig. 5A). Normalized per 10 µM drug, oxaliplatin and cisplatin formed an estimated 0.8 ± 0.1 and 1.5 ± 0.3 DPC/10⁶ bp, respectively (Table 2). The difference between two drugs was diminished but still noticeable in A2780 cells (data not shown).

View larger version (21K): [in this window] [in a new window]   Fig. 5.   DPC induced by oxaliplatin and cisplatin in intact CEM cells. A, cells were treated for 4 h with oxaliplatin () or cisplatin () at the indicated concentrations. B, CEM cells were treated with drugs for 1 h, followed by a post-treatment incubation as indicated. The data shown in B are normalized per 10 µM drug based on determinations for 79 and 158 µM oxaliplatin () and 34 and 68 µM cisplatin (). DPC were determined by the K+/SDS precipitation method. DPC frequency was estimated as described under Materials and Methods. Data represent means ± S.E.M. from two to three determinations.

ISC and DPC in Isolated Nuclei. To unequivocally establish whether oxaliplatin is inherently less efficient than cisplatin in the formation of both ISC and DPC, these lesions were measured in isolated nuclei from CEM cells. In this system, the drugs react with intact nuclear chromatin but the induction of secondary effects, including strand breaks, which affect the determinations in intact cells, was unlikely. Also, potential differences in drug uptake can be ruled out in the nuclear system. The results provide a clear-cut answer that there are markedly fewer ISC and DPC induced by oxaliplatin than cisplatin (Fig. 6). Sedimentation profiles observed at 25 µM drugs (Fig. 6A) correspond to 5.5 and 15.6 ISC/10⁶ bp, for oxaliplatin and cisplatin, respectively. Similarly, oxaliplatin remained less efficient in DPC induction than did cisplatin (Fig. 6B). For instance, 10 µM oxaliplatin and cisplatin produced ~8 and ~20 DPC/10⁶ bp, respectively, in the nuclear system. Moreover, the DPC induction reached a plateau for both drugs after ~4 h (data not shown), but the levels of oxaliplatin-induced DPC were significantly lower than cisplatin-induced DPC in their steady state.

View larger version (20K): [in this window] [in a new window]   Fig. 6.   ISC and DPC in isolated nuclei from CEM cells. A, sedimentation profiles for DNA from nuclei incubated for 4 h with no drugs ( ), oxaliplatin (25 µM, ), and cisplatin (25 µM, ). The arrow shows the direction of sedimentation, and the bracket shows the areas of sedimentation profiles, which are indicative of interstrand cross-links. The profiles shown correspond to 5.5 and 15.6 ISC/106 bp/25 µM, for oxaliplatin and cisplatin, respectively. B, DPC in nuclei treated for 4 h with oxaliplatin () or cisplatin () and analyzed by the K+/SDS precipitation method. The ordinate gives the frequency of DPC estimated as described under Materials and Methods. Data represent means ± S.E.M. from two determinations carried out in duplicate.

Inhibition of DNA Chain Elongation. To assess the potential consequences of DNA damage, additional experiments explored the relationship between drug adducts and DNA chain elongation. The short time of drug treatment needed for these determinations necessitates the use of relatively high drug concentrations (Gibbons et al., 1991; Mamenta et al., 1994). However, the parallel determinations of the platination levels (cf. Fig. 2) allowed us to express drug effects on chain elongation as a function of adduct frequency (Fig. 7). Oxaliplatin adducts seemed to be approximately 7-fold more effective in inhibiting DNA chain elongation than cisplatin adducts in CEM cells, considering the number of Pt-DNA adducts per 10⁶ bp (Fig. 7). Somewhat less profound (approximately 2-fold) difference in the adduct inhibitory efficiency was found in A2780 cells (data not shown).

View larger version (14K): [in this window] [in a new window]   Fig. 7.   Inhibition of chain elongation by oxaliplatin-treated () or cisplatin-treated () CEM cells as a function of Pt adducts. Cells were treated with drugs for 15 min, followed by a 30-min chase in drug-free medium and a 15-min pulse of [3H]thymidine. The percentages of chain elongation (relative to control) were obtained from sedimentation profiles of pulse-labeled nascent DNA as described under Materials and Methods and plotted as a function of DNA-Pt adducts using data from Fig. 2. Each point represents the mean ± S.E.M. of three different experiments.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In contrast to extensive studies of DNA damage by cisplatin, little is known about damage to cellular DNA by oxaliplatin. It is commonly accepted that DNA damage by antitumor Pt drugs is responsible for their cytotoxic properties. According to this central paradigm, oxaliplatin should be more proficient in damaging DNA than cisplatin, given that oxaliplatin is at least equally cytotoxic or frequently more cytotoxic than cisplatin (Table 1) (Rixe et al., 1996). Systematic characterization of oxaliplatin-induced DNA lesions in this study suggests that oxaliplatin declines from the predictions of the central paradigm. In addition to total platination, we quantified oxaliplatin-induced ISC and DPC in comparison to analogous effects of cisplatin. The results demonstrate that oxaliplatin is undoubtedly a DNA-reactive drug in cellular systems and resembles closely cisplatin with regard to the types and proportions of specific DNA lesions. Yet, oxaliplatin consistently forms markedly fewer total adducts and specific type of cross-links than cisplatin, despite at least equal or greater cytotoxicity.

The determinations of the total levels of DNA platination demonstrate that oxaliplatin formed significantly fewer adducts with cellular DNA than did cisplatin (Fig. 2 and Table 2). This observation that platination levels produced by oxaliplatin are consistently lower than those induced by cisplatin is in agreement with the determinations by other researchers (Saris et al., 1996), who found a 10-fold lower level of oxaliplatin adducts, compared with cisplatin adducts, in A2780 cells by immunochemical detection. Also, our recent studies showed that the level of Pt adducts induced by oxaliplatin in specific regions of DNA from drug-treated A2780 cells was 2 to 6 times lower than that of cisplatin (Woynarowski et al., 1998). Clearly, in various systems, oxaliplatin needs to form fewer adducts than cisplatin for comparable cytotoxicity.

The lower overall DNA adduction by oxaliplatin might still be reconciled with the equal or higher cytotoxic activities, if oxaliplatin-DNA adducts were considerably more difficult to repair compared with cisplatin adducts. Adduct repair is an important factor in Pt drug action (Petersen et al., 1996; Damia et al., 1998; Hibino et al., 1999; Koeberle et al., 1999). However, oxaliplatin-induced DNA damage appears to be no more difficult to repair than cisplatin-induced damage as judged by plasmid reactivation in several cell lines (Fig. 3), which reflects mainly excision repair processes. This result is consistent with the observation that both cisplatin and oxaliplatin adducts are similarly removed in a cell-free excision repair system (Reardon et al., 1999). Thus, the lower total platination in the case of oxaliplatin is unlikely to be compensated, compared with cisplatin, by impeded adduct removal.

Oxaliplatin cytotoxicity might possibly result from proportionately more highly lethal lesions, such as interstrand cross-links, or other previously uncharacterized lesions, such as DPC. Oxaliplatin forms both types of lesions in cellular DNA (Figs. 4 and 5 and Table 2). These results document for the first time that oxaliplatin is able to induce ISC and DPC in cellular DNA. Although ISC and DPC constitute a minor fraction of total adducts for either oxaliplatin or cisplatin, these lesions may contribute to oxaliplatin effects. However, the absolute levels of ISC and DPC induced by oxaliplatin were markedly lower than those induced by cisplatin, after the reduced total platination levels (Table 2). Thus, it seems unlikely that these lesion types may compensate for the overall reduction of oxaliplatin adducts.

Inherently lower reactivity of oxaliplatin compared with cisplatin is corroborated in a clear-cut way in isolated nuclei (Fig. 6). Thus, potential differences in cellular uptake of the two drugs can be ruled out as a simple explanation for the lower levels of oxaliplatin adducts in intact cells. Markedly lower reactivity of oxaliplatin was shown also with naked DNA (Woynarowski et al., 1998). Although the details of intracellular oxaliplatin activation remain unknown, the slow dissociation of the oxalate ligand may be the bottleneck in oxaliplatin-DNA reactivity (Luo et al., 1998, 1999a,b).

Our data suggest that oxaliplatin-DNA adducts may be more lethal than cisplatin adducts. In CEM cells, GI₅₀ values for oxaliplatin and cisplatin are virtually identical, even though oxaliplatin treatment results in significantly fewer lesions than cisplatin treatment at equimolar concentrations. The greater lethality of oxaliplatin adducts is consistent with their greater inhibition of DNA chain elongation (Fig. 7). Similar differences in replicative bypass have been noted between DACH adducts and cisplatin adducts in other cell lines (Gibbons et al., 1991; Mamenta et al., 1994). Processing of oxaliplatin and cisplatin adducts may also elicit different downstream responses. For instance, cisplatin depends on intact mismatch repair for its maximal cytotoxicity (Aebi et al., 1996; Fink et al., 1996; Vaisman et al., 1998; Ferry et al., 1999). In contrast, oxaliplatin adducts are poorly recognized by mismatch repair proteins (Fink et al., 1996) and do not activate JNK and c-Abl (Nehme et al., 1999), and oxaliplatin retains a high cytotoxicity in mismatch repair-deficient cells (Fink et al., 1997; Vaisman et al., 1998).

Finally, it is important to recognize that DNA damage, although probably crucial, represents only one aspect of the pleiotropic effects of Pt drugs. Only 5 to 10% of covalently bound cell-associated cisplatin is found in the DNA fraction, whereas cisplatin binding to proteins is 1 order of magnitude greater (~75-85%; Akaboshi et al., 1992, 1994). An intriguing possibility is that functional protein damage (interference with enzymatic, receptor, and/or structural functions) may play a greater role in the effects of oxaliplatin than other Pt drugs. The hydrophobic DACH moiety in oxaliplatin may shift drug reactivity toward a subset of cellular proteins with hydrophobic binding pockets (Chaney, 1995). These proteins might well be different from those that react with cisplatin. Thus, enhanced and/or different protein binding might also be a factor in the disproportionately potent apoptosis induction by oxaliplatin (Faivre and Woynarowski, 1998), given the drug's modest DNA reactivity. Studies are under way to elucidate whether protein damage, in addition to DNA damage, contributes to the cytotoxic and proapoptotic properties of oxaliplatin.