Aurintricarboxylic Acid Inhibits Protein Synthesis Independent,Sanguinarine-Induced Apoptosis and Oncosis

Abstract

Sanguinarine, a benzophenanthridine alkaloid, has anticancer potential through induction of cell death. We previously demonstrated that sanguinarine treatment at a low concentration (1.5 μg/ml) induced apoptosis in K562 human erythroleukemia cells, and a high concentration (12.5 μg/ml) induced the morphology of blister formation or oncosis-blister cell death (BCD). Treatment of cells at an intermediate sanguinarine concentration (6.25 μg/ml) induced diffuse swelling or oncosis-diffuse cell swelling (DCS). To assess the underlying mechanism of sanguinarine-induced apoptosis and oncosis-BCD in K562 cells, we studied their response to pre-treatment with two chemical compounds: aurintricarboxylic acid (ATA) and cycloheximide (CHX). The pretreatment effects of both chemical compounds on apoptosis and oncosis-BCD were evaluated by measuring multiple parameters using quantitative morphology, electron microscopy, terminal deoxynucleotidyl transferase (TdT) end-labeling and annexin-V-binding. ATA, a DNA endonuclease inhibitor, efficiently prevented DNA nicking and inhibited apoptosis almost completely and oncosis-BCD by about 40%, while CHX, a protein synthesis inhibitor, failed to inhibit both apoptosis and oncosis-BCD. These results demonstrate, first, the importance of endonuclease in sanguinarine-induced apoptosis and to some extent in oncosis-BCD and, second, that this inhibition does not require de novo protein synthesis.

Introduction

Sanguinarine, an extract from the bloodroot plant Sanguinaria canadensis of the Papaveraceae family, is a quaternary benzo[c]phenanthridine alkaloid. Its principle medicinal use to date is in dental products based on its antibacterial, antifungal, and anti-inflammatory activities, which reduce both gingival inflammation and supragingival plaque formation (Kuftinec et al., 1990; Laster and Lobene, 1990; Godowski et al., 1995). Sanguinarine is also considered to be a potent anti-cancer as well as chemopreventative agent (Imanek, 1985; Dostal and Potácek, 1990; Vavreckova and Ulrichová 1994; Ahmad et al., 2000; Biswas and Khuda-Bukhsh, 2002; Adhami et al., 2003, 2004; Kemeny-Beke et al., 2005).

Previously, we have shown that K562 human erythroleukemia cells, when exposed to sanguinarine at concentrations of 1.5 μg/ml and 12.5 μg/ml for 2 hours, displayed the morphologies of 2 different modalities of cell death: at 1.5 μg/ml the morphology of apoptosis and at concentrations of 12.5 μg/ml, the morphology of single blister formation or oncosis/blister cell death (BCD) (Weerasinghe et al., 2001c). This dual cell death modality induced by sanguinarine was termed “bimodal cell death” (Weerasinghe et al., 2001a). At the intermediate sanguinarine concentration of 6.25 μg/ml cells displayed diffuse cell swelling (DCS), which might represent a variant of oncosis. The terms BCD and DCS are not accepted nomenclature but are used in this report to describe the primary morphological characteristic that were found in each of the two variants of oncosis: namely, the single blister formation to identify oncosis-BCD and the diffuse cell swelling to determine oncosis-DCS. Treatment of cells with concentrations of sanguinarine above 12.5 μg/ml resulted in ruptured blisters.

Furthermore, our previous studies have found sanguinarine to be an efficient chemotherapeutic agent. In this regard, we have shown that sanguinarine overcomes the multidrug resistant phenomenon in several cell lines. For example, we have shown that cisplatin-resistant human papillomavirus (HPV) type 16-immortalized endocervical cells are sensitive to sanguinarine (Ding et al., 2002). Additionally, sanguinarine was found to efficiently induce cell death in p53 null k562 cells (Weerasinghe et al., 2001b), which in several previous reports have been found to be rather resistant to the induction of apoptosis (Kobayashi et al., 1998). Thus, sanguinarine might be a valuable chemotherapeutic agent for most cancers, which carry p53 mutations. We have also found sanguinarine to overcome P-glycoprotein (Pgp)-mediated multidrug resistance (Weerasinghe et al., 2006). These findings emphasize the need to further study the sanguinarine-induced cell death pathways of apoptosis and oncosis. Several recent studies have also suggested that sanguinarine may also be an effective anti-genotoxic and chemopreventive agent (Biswas and Khuda-Bukhsh 2002; Adhami et al., 2003).

Apoptosis is a process of active cell death characterized by cell shrinkage with preservation of cell membrane permeability and subsequent formation of broad cell surface projections (budding) followed by cell fragmentation into apoptotic bodies; this process is also accompanied by nuclear DNA condensation and fragmentation (Arends et al., 1990; Majno and Joris, 1995). In contrast, oncosis is a process of passive cell death related to energy depletion leading to impairment of ionic pumps of the cell membrane, cell swelling and formation of cell surface blebs or blisters (Majno and Joris, 1995; Buja, 2005). It is known that various factors maintain inhibition of apoptosis. For example, our previous studies have shown Bcl-2 gene product to play the dual role of inhibiting apoptosis as well as oncosis-blister cell death (BCD) induced by sanguinarine (Weerasinghe et al., 2001a, 2001b). Apart from Bcl-2, growth and survival factors as well as chemicals that mimic the affects of these natural factors also inhibit the apoptotic process (Yanagihara and Tsumuraya, 1992; Liu et al., 1994). Chemical regulators of apoptosis include, aurintricarboxylic acid (ATA), a general endonuclease inhibitor (Hallick et al., 1977), and cycloheximide (CHX), a protein synthesis inhibitor (Liu et al., 1994), and these agents were utilized in the studies described in this report.

Understanding different cell death mechanisms of individual anti-cancer agents may lead to their effective administration, alone or in combination with other established therapies. Thus, the need to identify and characterize novel cell death modalities of potential anti-cancer agents remains. The objective of this study is to evaluate the role of endonuclease in sanguinarine-induced apoptosis and oncosis-BCD, and also to assess the importance of de novo protein synthesis in both these processes, by the use of ATA and CHX, respectively. By probing the processes of cell death using known inhibitors of endonuclease and de novo protein synthesis, we would be able to study the involvement of both these factors in apoptosis and oncosis. This, we believe, will shed more light on the mechanism of action of apoptosis and oncosis induced by sanguinarine.

Additionally, this would also facilitate the comparison of apoptosis and oncosis, based on the degree of involvement of endonuclease and protein synthesis in the execution of these cell death processes. The present study will be limited to comparing the characteristics and mechanisms of action of only 2 modes of cell death that we have identified here: apoptosis and oncosis (BCD), as these represent the 2 extremes in the spectrum of sanguinarine-induced cell death. Of the 2 chemicals, only ATA was found to prevent DNA nicking and almost completely inhibit apoptosis, and also to inhibit oncosis/BCD by about 40%.

Methods and Materials

Materials

The K562 erythroleukemia cell line was purchased from ATCC (Rockville, MD, USA). The drugs sanguinarine-HCl, ATA and CHX were purchased from Sigma Chemicals Co. (St. Louis, MO). Sanguinarine-HCl was maintained as a dry powder at room temperature. Annexin-V-Fluos staining kit for the detection and quantification of apoptosis and differentiation from necrosis at the single cell level was purchased from Boehringer Mannheim. The terminal deoxynucleotidyl transferase (TdT) end-labeling assay was purchased from Sigma Chemical Co. (St. Louis, MO). Both kits were stored and handled according to manufacturer’s instructions.

Cell Culture

The K562 erythroleukemia cells were routinely maintained as a cell suspension in RPMI-1640 medium supplemented with 10% fetal calf serum and 2 ml L-glutamine in a humidified atmosphere containing 5% CO₂ at 37°C. All experiments were performed on K562 cells during their exponential phase of growth.

Treatment of Cells with Sanguinarine

A working sanguinarine solution of 50 μg/ml was prepared from a stock solution of 1 mg/ml. Sanguinarine was serially diluted in RPMI + 10% FBS to give a concentration range of 25 μg/ml to 0.19 μg/ml (8 dilutions) in 96-well plates. Triplicate wells for each drug concentration was prepared and aliquots of 10,000 cells (ATA pretreated, CHX pretreated or chemically untreated) per well were added. These tissue culture plates were subsequently incubated at 37°C + 5% CO₂ for 2 hours. Cells thus prepared were used in all subsequent experiments.

Pretreatment of Cells with ATA

A preliminary screening procedure was conducted in order to determine the concentration and incubation period of ATA pretreatment that produced maximum inhibition of sanguinarine-induced apoptosis and oncosis/BCD in K562 cells. The concentrations and time points for the screening process for endonuclease inhibitor activity was selected based on reports in the literature (Hallick et al., 1994). Concentrations of ATA used for the screening were 50, 100 and 200 μM solutions, which were dissolved in RPMI + 10% FBS; and the incubation periods were 2, 12, 24, and 48 hours. Maximum inhibition of the morphology of apoptosis and oncosis/BCD was observed when cells were pretreated with ATA concentrations of 100 μM for a period of 2 hours. All subsequent experiments were done under these conditions. Following ATA pretreatment, the cells were treated with serially diluted sanguinarine as described previously.

Pretreatment of Cells with Cycloheximide

A screening process similar to that with ATA was also carried out for CHX to determine the optimum pretreatment conditions that produced maximum inhibition of sanguinarine-induced apoptosis and oncosis/BCD. Again, concentrations and time periods for the screening process were obtained from the literature (Yanagihara and Tsumuraya, 1992; Liu et al., 1994). These concentrations and time points were found to coincide with the maximum protein synthesis inhibitory effect. The concentrations of CHX used were 0.1, 2.5, and 5.0 μg/ml; and the time periods: 2,12, 24, and 48 hours. Since CHX was not able to inhibit or change the apoptosis and oncosis/BCD induced by sanguinarine in any significant way, a random concentration of 2.5 μg/ml and an incubation of 2 hours were selected as pretreatment conditions with CHX for all subsequent experiments.

Quantitative Morphology

Cells were pretreated with ATA and CHX and exposed to sanguinarine as outlined here. The percentage of cell death, (i.e., apoptosis and oncosis/BCD) corresponding to each drug concentration was determined by light microscopy. The key morphological criterion for the identification of apoptosis was the formation of apoptotic bodies and the key criterion for oncosis/BCD was the formation of cell surface blisters. The number of apoptotic cells and blistering cells were counted using a hemocytometer.

Electron Microscopy

Upon exposure of cells to the appropriate concentrations of sanguinarine, ATA and CHX as described here, experimental and control cells (sanguinarine untreated, as well as ATA and CHX pretreated) were fixed with 2% glutaraldehyde, postfixed in 1% osmium tetroxide, dehydrated through a series of ethanols and embedded in TAAB 812 epoxy resin. Semithin (0.5 μm) sections were cut axially, through all layers of each pellet to select areas for ultramicrotomy. Ultrathin sections were stained with lead citrate and 30% alchoholic uranyl acetate and then examined in a JEOL 100-Cx electron microscope.

Terminal Deoxynucleotidyl Transferase (TdT) End-Labeling Assay

Specific 3′-hydroxyl ends of DNA fragments generated by endonuclease-mediated apoptosis are preferentially repaired by terminal deoxynucleotidyl transferase (TdT) (Gavrieli et al., 1992). The TdT mediated nick end labeling assay has been developed to label these strand-breaks with streptovidinhorseradish peroxidase conjugated nucleotides followed by the addition of a substrate (TBL).

After drug (sanguinarine, ATA and CHX) treatment, cells were removed from individual wells, washed in PBS and fixed in 10% neutral-buffered formaldehyde for 10 minutes. These cells were then resuspended in 80% ethanol at 1×10⁶ cells per ml. 50,000 cells were placed onto an electrostatically treated glass slide, air-dried at room temperature and permeabilized with proteinase K (Gavrieli et al., 1992). Thereafter, cell samples were incubated for 60 minutes at 37°C in the presence and absence of exogenous TdT and streptavidinhorseradish peroxidase conjugated dNTP (deoxynucleotide triphosphate), followed by the substrate TBL according to the manufacturer’s instructions (Sigma TACS In Situ Apoptosis Detection Kit, St. Louis, MO, USA). Cells were then examined and photographed under phase microscopy, and counted to determine the percentage of cells with DNA nicking to total cells.

Fluorescein-Conjugated Annexin V Binding Assay

After drug treatment, cells (1×10⁶ cells) were washed with PBS and incubated with Annexin-V Fluorescein Isothiocyanate (FITC) conjugate and propidium iodide (PI) utilizing the Annexin-V Fluos staining-kit (Boehringer Mannheim Corp.). After labeling, cells were resuspended in binding buffer and analyzed using flow cytometry. FITC-fluorescence was measured at 530–545 nm and fluorescence of DNA-PI complexes at 575–606 nm. Cell debris was excluded from analysis by appropriate forward light scatter threshold setting (Leist et al., 1997).

Statistical Analysis

The results are presented as means ± SE. Statistical analyses were performed using the student’s t-test. Differences were considered significant when p < 0.05.

Results

The effect of sanguinarine-treatment in K562 cells was studied at different doses and time points. Treatment of K562 cells, when exposed to sanguinarine concentrations of 1.5 μg/ml, 6.25 μg/ml and 12.5 μg/ml for 2 hours showed maximum apoptosis, maximum oncosis/diffuse cell swelling and maximum oncosis/blister cell death, respectively (Figures 1 and 2). In Figure 3, we have summarized the percentages of apoptosis, oncosis/diffuse cell swelling, oncosis/blister cell death and ruptured cells/necrosis, when K562 cells were treated with ascending concentrations of sanguinarine for 2 hours (A), 12 hours (B), and 24 hours (C), respectively. A salient feature of this study is that the dose-response curves were found to shift to the left as the duration of exposure of cells to the drug gradually increased.

The percentages of trypan blue excluding cells, denoting cell viability (Majno and Joris, 1995; O’Brien et al., 1997), were measured at 2 hours and 12 hours of sanguinarine treatment. Treatment with sanguinarine for 2 hours at 1.5 μg/ml, 6.25 μg/ml and 12.5 μg/ml showed over 90% of cells excluding trypan blue at all 3 concentrations. However, after 12 hours of sanguinarine-exposure at concentrations of 1.5 μg/ml, 6.25 μg/ml, and 12.5 μg/ml, the percentage of trypan blue excluding cells were >85%, 50%, and 30%, respectively (results not shown). Pretreatment with ATA 100 μM for 2 hours on sanguinarine-induced apoptosis and oncosis/BCD resulted in an almost complete inhibition of apoptosis, and a reduction of oncosis/BCD by about 40% (Figures 4 and 5). Cells treated with ATA only (ATA controls) did not affect the morphology of K562 cells in any measurable way.

Light and electron microscopy of K562 cells exposed to concentrations of sanguinarine that induce apoptosis (1.5 μg/ml) showed the classic morphological changes: formation of apoptotic bodies containing organelles, chromatin condensation and nuclear fragmentation, and an absence of microvilli (Figures 2B and 4B). When treated with sanguinarine concentrations that induce oncosis/BCD (12.5 μg/ml), K562 cells displayed single blister formation devoid of organelles, patchy chromatin condensation, an increase in cytoplasmic vacuolization and an absence of microvilli (Figures 2D and 4C). When treated with the intermediate sanguinarine concentration of 6.25 μg/ml, cells displayed the morphology of diffuse cell swelling with patchy chromatin condensation and vacuolization (Figure 2C). ATA pretreatment resulted in a complete inhibition of the sanguinarine-induced apoptosis with the absence of both apoptotic bodies and nuclear fragmentation (Figures 4E and 5). ATA pretreatment of sanguinarine-induced oncosis/BCD, as shown in Figure 4F, resulted in an inhibition of blister formation (although the overall reduction of oncosis/BCD was by about 40%). However, both sanguinarine-induced apoptosis and oncosis/BCD pretreated with ATA showed an increase in chromatin condensation and cytoplasmic vacuolization as compared to control untreated cells; the presence of microvilli was also noted in both cases.

The TdT end-labeling method and the Annexin-V-assay was used to assess both apoptosis and oncosis. TdT end-labeling method showed DNA nicking in over 90% of apoptotic cells (Figure 6B) and an absence of DNA nicking during oncosis (Figure 6C). ATA pretreatment prevented DNA nicking during apoptosis (Figure 6E). The Annexin-V-assay used to detect the cell surface membrane phosphatidyl serine (PS) flip, usually associated with apoptosis (Leist et al., 1997), showed 51.8% ± 6.2 of Annexin-V-positive cells in apoptosis. ATA pretreatment of apoptotic cells prevented the PS flip and consequently the binding of Annexin-V to the cell membrane. However, cells that underwent oncosis did not show significant Annexin-V binding (18.2% ± 5.3) (Figure 7).

In contrast to ATA, pretreatment by CHX failed to show significant change in the parameters of sanguinarine-induced apoptosis and oncosis as measured by any of the aforementioned methods.

Discussion

Our results show that pretreatment with the endonuclease inhibitor ATA prevented DNA fragmentation and also completely inhibited the morphology of sanguinarine-induced apoptosis. The inhibition of apoptosis is also clearly shown from results of microscopy, quantitative morphology and Annexin-V-binding. Contrary to expectations, ATA pretreatment consistently show a reduction of sanguinarine-induced oncosis/BCD by about 40%. This finding is puzzling, as oncosis is not associated with DNA nicking.

On the other hand, apoptosis has generally been found to correlate with DNA fragmentation (Liepins and Younghusband, 1987), and is considered to be one of its hallmarks (Arends et al., 1990; Compton, 1992). DNA fragmentation is thought to occur at the internucleosomal regions due to the activation of a specific endonuclease (Arends et al., 1990; Compton, 1992; Wyllie et al., 1992; Barry and Eastman, 1993; Majno and Joris, 1995). Not all cells, however, manifest a strict correlation between the morphology of apoptosis and internucleosomal DNA fragmentation (Ucker et al., 1992). Initially, DNA is cleaved at the sites of attachment of chromatin loops to the nuclear matrix, which results in the appearance of discrete 300–350 kb size fragments (Oberhammer et al., 1993).

Subsequently, DNA is preferentially cleaved between nucleosomes. The products are discontinuous nucleosomal and oligonucleosomal sized DNA fragments. They generate a characteristic “ladder” pattern during agarose gel electrophoresis. However, in many cell types, DNA cleavage during apoptosis does not proceed to inter-nucleosomal sized sections but rather proceeds only to 300–350 kb size DNA fragments (Cohen et al., 1992; Collins et al., 1992; Oberhammer et al., 1993; Zakeri et al., 1993; Ormerod et al., 1994; Zamai et al., 1996).

Sanguinarine-induced apoptosis showed no evidence of laddering in agarose gel electrophoresis (Liepins et al., 1996); however, as detected by the TdT end-labeling assay, showed DNA nicking. Thus, DNA fragmentation in sanguinarine-induced apoptosis may not progress to internucleosomal segments, but rather stop at the 300–350 kb-fragment range. Activation of endonulease(s) is considered necessary for DNA nicking to occur (Oberhammer et al., 1993), and our results show that pretreatment of cells with the endonuclease inhibitor ATA (Hallick et al., 1977), efficiently inhibited DNA nicking as well as the sanguinarine-induced apoptotic process; thus increasing cell survival.

Therefore, the involvement of endonucleases associated with DNA fragmentation might be a determining factor in the biochemical pathway of sanguinarine-induced apoptosis. However, several studies have reported on some endonuclease-independent actions of ATA. These include the inhibition of topoisomerases (Catchpoole and Stewart, 1994; Benchokroun et al., 1995), interferon-α receptors and NMDA receptors (Roberts-Lewis et al., 1993; Zeevalk et al., 1993). ATA is also known to activate the mitogen-activated protein kinase (MAPK) receptor (Okada and Koizumi, 1995; Rui et al., 1998) and the erbB4 receptor (Okada and Koizumi, 1997). Thus, confirmation of these results should follow testing with the use of additional inhibitors and other methods as appropriate.

In contrast, pretreatment of cells with the protein synthesis inhibitor cycloheximide (CHX) failed to inhibit apoptosis or oncosis induced by sanguinarine. This indicates the lack of de novo protein synthesis in both cell death processes and points to the importance of posttranslational modification of proteins in sanguinarine-induced cell death. This notion was corroborated by recent findings in our laboratory using Western blot and cDNA expression array (Weerasinghe et al., 2001c). Apoptotic cells analyzed by Western blot showed an increase in the pro-apoptotic protein Bax, but cDNA expression studies showed no changes in Bax at the gene transcript level (Weerasinghe et al., 2001c). Our findings agree with several reports in the literature that have also found protein synthesis inhibition to be ineffective in limiting the apoptotic process (Liu et al., 1994); however, several others have found it effective (Williams et al., 1990; Yanagihara and Tsumuraya, 1992). Furthermore, it has been proposed that the effects of CHX on apoptosis may depend on many factors including the nature of the death-inducing agent, its dose and the type of cell line (Yanagihara and Tsumuraya, 1992). Thus, it may be more appropriate to suggest that CHX failed to inhibit sanguinarine-apoptosis in K562 cells at the concentrations and periods of exposure used in this study. Evidence of CHX acting as a stimulant of apoptosis is also found in literature (Thomas and Hersey, 1998).

Oncosis and apoptosis are two different manners of cell death. Oncosis is by far the older term but apoptosis has long been known by such terms as single cell necrosis or shrinkage necrosis (Majno and Joris, 1995). Oncosis refers to the pre-lethal phase that follows a lethal cell injury such as complete ischemia or the effects of many chemical toxins (Phelps et al., 1989; Trump and Berezesky, 1992). Although there has been considerable research on both types of cell death, apoptosis has been more studied from the standpoint of molecular genetics. Both types of cell death are “programmed” in the sense that the genetic information and many of the enzymes and other factors pre-exist in the cell (Trump et al., 1997; Levin et al., 1999). Recent advances in studies pertaining to the relationship between cell injury and death indicate that apoptotic and oncotic mechanisms can proceed together with oncotic mechanisms and morphology dominating the end stage of irreversible injury (Buja, 2005).

Sanguinarine-induced oncosis/BCD was found to exclude trypan blue (Weerasinghe et al., 2001b), and thus may not represent necrosis, as necrosis is associated with trypan blue permeability (Majno and Joris, 1995; O’Brien et al., 1997). Moreover, necrosis does not represent a form of cell death, but refers only to changes secondary to cell death by any mechanism (Levin et al., 1999; Majno and Joris, 2004). Majno and Joris (1995) describe oncosis as a form of accidental cell death accompanied by cellular swelling, organelle swelling, blebbing, and increased membrane permeability caused by the failure of the ionic pumps of the plasma membrane. Trump et al. (1997) and Trump and Berezesky (1992) associate oncosis (blister formation) with increases in concentration of cytosolic calcium and rearrangement of cytoskeletal proteins. Also, oncosis, like apoptosis, is reported to be triggered by the activation of cell surface receptors such as PORIMIN (pre-oncosis receptor induced membrane injury) (Ma et al., 2001). Recent reports indicate that a modest increase in the expression level of uncoupling protein 2 (UCP-2) leads to a rapid and dramatic fall in mitochondrial membrane potential and a reduction in intracellular ATP, resulting in oncosis (Mills et al., 2002).

Strangely, the present study shows that pretreatment with ATA partially inhibits sanguinarine-induced oncosis/BCD; as was evident from the results of quantitative morphology. This partial inhibition of oncosis was despite the apparent lack of endonuclease involvement, as shown by the absence of DNA fragmentation by the TdT end labeling method. Recently, however, it has been suggested that apoptosis may share common molecular pathways with other types of cell death at an early stage (Shirai, 1999; Igney and Krammer, 2002). These pathways are thought to share some, but not all, the characteristics of classical apoptotic pathways (Igney and Krammer, 2002). For example, it has been reported that there are molecularly less-well defined cell-death pathways that do not require caspase activation (Borner and Monney, 1999; Sperandio et al., 2000). However, recent reports also indicate that the cell death process of oncosis is caspase 1-dependent (Sun et al., 2005; Thumbikat et al., 2005), and also that the transcription factor NF-kappa B protects cells against oncosis (Franek et al., 2004). The partial inhibition of sanguinarine-induced oncosis by ATA-pretreatment indicates that apoptosis and oncosis may share common molecular pathways, especially at the early stages; however, activation of endonuclease and DNA fragmentation typically also occurs in the late stages of apoptosis. It is also possible that ATA inhibited low levels of random DNA fragmentation mediated by various endonucleases that occurred in K562 cells undergoing oncosis but that this process was not detected by the TdT nick end-labeling method. Further studies are needed to better explain these findings.

Thus, our findings show that sanguinarine at low levels induces tumor cell death by apoptosis, and at higher levels, by oncosis. Also, we have shown that the sanguinarine-induced apoptotic process is independent of de novo protein synthesis, but requires endonuclease activity. In contrast, the cell death process of oncosis, although independent of de novo protein synthesis, may require only partial endonuclease activity. We believe that the identification and characterization of different modalities of cell death induced by potential anti-cancer drugs, an emerging field in toxicogenomics, might be important in the war against cancer.