Oxaliplatin induces mitotic catastrophe and apoptosis in esophageal cancer cells

Abstract

The platinum‐based chemotherapeutic agent oxaliplatin displays a wide range of antitumor activities. However, the underlying molecular responses to oxaliplatin in esophageal cancer remain largely unknown. In the present study, we investigated the effect of oxaliplatin on two esophageal cancer cell lines, squamous cell carcinoma (TE3) and adenocarcinoma (TE7). Following cell‐cycle arrest at G₂ phase after oxaliplatin treatment, TE3 cells died via apoptosis and TE7 cells died via mitotic catastrophe. Survivin was inhibited more in TE7 cells compared with TE3 cells, but inhibition of survivin using small interfering RNA induced mitotic catastrophe in both cell lines. Further investigations indicated that survivin promoter activity was also inhibited by oxaliplatin. Among mitotic catastrophe‐associated proteins, 14–3‐3σ was decreased in TE7 cells; no evident changes were observed for aurora kinases. Oxaliplatin‐induced apoptosis in the TE3 cells was caspase dependent. However, downregulation of Bad, Bid, Puma, and Noxa, lack of cytochrome c release, and limited loss of mitochondrial membrane potential in early phase indicated possible initiation by pathways other than the mitochondrial pathway. Mechanistic studies showed that downregulation of survivin by oxaliplatin in TE7 cells was partially due to the proteasome‐mediated protein degradation pathway and partially due to the downregulation of Sp1 transcription factor. Similar results were obtained for another gastric adenocarcinoma cell line, MKN45, in which survivin was previously shown to be inhibited by oxaliplatin. These data indicate that survivin may be a key target for oxaliplatin. The ability of oxaliplatin to induce different modes of cell death may contribute to its efficacy in esophageal cancer. (Cancer Sci 2008; 99: 129–139)

Abbreviations: ADC

adenocarcinoma

IC₅₀

inhibitory concentration at 50%

PARP

polyADP‐ribose polymerase

SCC

squamous cell carcinoma

siRNA

small interfering RNA.

Oxaliplatin, a third‐generation platinum‐derived chemotherapy agent, displays a wide spectrum of antitumor activity.^{( 1 )} Similar to other platinum drugs, its biological activity is due to its ability to form lethal DNA lesions, including interstrand DNA crosslinks and DNA–protein crosslinks.^{( 2 )} Because it contains a 1,2‐diaminocyclohexane carrier ligand,^{( 3 )} oxaliplatin generates DNA adducts that are bulkier and more hydrophobic than the adducts formed by cisplatin and carboplatin, making it more effective at inhibiting DNA synthesis and thus more cytotoxic.^{( 4 )} In addition, oxaliplatin has fewer side‐effects compared to other platinum drugs in terms of nephrotoxicity and myelosuppression.^{( 5 )}

Clinically, oxaliplatin effectively treats colorectal cancer in combination with fluorouracil and leucovorin.^{( 6 )} In contrast to colorectal cancer, which is not very responsive to platinum‐based therapy, platinum‐derived chemotherapy agents such as cisplatin have become the standard therapy for gastroesophageal cancer. Esophageal cancer studies indicate that oxaliplatin is effective when used in combination with fluorouracil and radiotherapy, with or without cisplatin, with tolerable toxicity.^{( 7} , ^{8 )}

We previously demonstrated that oxaliplatin treatment induced downregulation of the cell‐cycle regulatory protein p21^waf1/cip1 in colon cancer cells, allowing abrupt S‐phase entry and resulting in G₂/M arrest.^{( 9 )} Other investigators have reported the involvement of pro‐apoptotic proteins, Bax and Puma, cytochrome c translocation, and caspase activation in oxaliplatin‐induced apoptosis in colon cancer cells.^{( 10} , ^{11 )} Inhibition of the anti‐apoptotic protein Bcl‐xL was also shown to enhance apoptosis in oxaliplatin‐treated cells.^{( 12 )}

The underlying molecular response to oxaliplatin in esophageal cancer remains largely unknown; this response may vary depending on the particular genetics and histological origins of the cancer. In the present study, we examined the cytotoxic effects of oxaliplatin in vitro, using two esophageal cancer cell lines: one derived from SCC (cell line TE3) and one from ADC (cell line TE7). Specifically, we investigated the underlying mechanism in oxaliplatin‐induced growth inhibition and cell death. Cisplatin was included as a control.

Materials and Methods

Cell culture.  The esophageal cancer cell lines TE3 and TE7,^{( 13 )} and the gastric cancer cell lines MKN28 and MKN45, were obtained from the Cell Resource Center for Biomedical Research, Institute of Development, Aging and Cancer, Tohoku University (Sendai, Japan) and from the Japanese Cancer Research Bank (Tokyo, Japan). Cell lines were maintained in RPMI‐1640 medium supplemented with 10% fetal bovine serum and 100 U/mL penicillin and streptomycin in 5% CO₂ at 37°C.

Antibodies and chemicals.  The primary antibodies were as follows: actin polyclonal antibody (Sigma‐Aldrich, St Louis, MO, USA); cdc2 monoclonal antibodies, aurora A kinase, aurora B kinase, Bad, Bak, Bax, Bcl‐2, Bcl‐xL, caspase‐3, cleaved caspase‐3, caspase‐9, cleaved caspase‐9, PARP, Puma, Stat3, and phospho‐Stat3 polyclonal antibodies (Tyr705) (Cell Signaling Technology, Beverly, MA, USA); Bid monoclonal antibody (BD Transduction Laboratories); Bim polyclonal antibody (BD Pharmingen); p21^waf1/cip1 monoclonal antibody and Noxa polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA); survivin polyclonal antibody (Novus Biologicals, Littleton, CO, USA); and 14‐3‐3σ monoclonal antibody and Sp1 polyclonal antibody (Upstate Biotechnology, Lake Placid, NY, USA). Horseradish peroxidase‐labeled antimouse and antirabbit secondary antibodies were obtained from Amersham BioSciences (Piscataway, NJ, USA). Caspase‐3 (z‐DEVD‐fmk), caspase‐9 (z‐LEHD‐fmk), and pan‐caspase (z‐VAD‐fmk) inhibitors and Stat3 inhibitor peptide were obtained from Calbiochem (Darmstadt, Germany). The Sp1 inhibitor mithramycin A was purchased from Sigma‐Aldrich.

Drug treatment and cytotoxicity assays.  Oxaliplatin was obtained from Yakult (Tokyo, Japan) whereas cisplatin was purchased from Sigma‐Aldrich. The drug exposure time was 24 h. The WST‐1 assay (Dojindo Laboratories, Kumamoto, Japan) was carried out as reported previously.^{( 14 )} The results are reported as the mean of three independent experiments.

Clonogenic survival assay.  Cells were seeded at 1000 cells per dish and cultured for 10 days. Cells were fixed in ethanol and stained with Giemsa solution (Sigma‐Aldrich). Plate images were captured and the stained area (i.e. the cell colony) was measured using image analysis software (WinROOF; Mitani, Fukui, Japan). The survival of drug‐treated cells was compared with untreated control cells.

Flow cytometry.  Flow cytometry was carried out using FACSort, and data collection was carried out using CellQuest software (both from Becton Dickinson Immunocytometry Systems, San Jose, CA, USA). The data were derived from at least two independent experiments.

Cell‐cycle analysis.  Cell‐cycle analysis was carried out as described previously.^{( 9 )} The data were analyzed using ModFIT software version 3.0 (Becton Dickinson Immunocytometry Systems).

Annexin V apoptosis assay.  The annexin V–fluorescein isothiocyanate apoptosis detection kit was obtained from BioVision Research Products (Mountain View, CA, USA). Analysis was carried out as described previously.^{( 14 )} For experiments using caspase inhibitors, inhibitors were added simultaneously with drugs and continually after drug removal.

Mitochondrial membrane potential.  Changes in the mitochondrial membrane potential were assayed using a mitochondrial membrane potential detection kit (APO‐LOGIX JC‐1; Peninsula Laboratories, San Carlos, CA, USA). Experiments were carried out according to the manufacturer's instructions using flow cytometry. JC‐1 dye stains the mitochondria red in healthy cells (with intact mitochondrial membranes) and the cytosol green when the mitochondrial membrane collapses during apoptosis (loss of mitochondrial membrane potential).

Cytological staining.  The cells were cultured on glass coverslips. They were then fixed in methanol and stained with hematoxylin solution (Sigma‐Aldrich). Cells in mitotic catastrophe were scored as reported previously^{( 15 )} (i.e. enlarged cells that contained two or more evenly stained nuclear fragments were scored as positive). One thousand cells were counted for each data point. Cells with nuclear enlargement were also analyzed. The diameters of cell nuclei were measured using MacScope software (Mitani). Nuclei in 500 cells were measured per data point. The results are reported as mean diameter ± SD from two independent experiments.

Western blot analysis.  Western blot analysis was carried out as described previously.^{( 16 )} Actin immunoblots were used to confirm near‐equivalent protein loading in all lanes.

Mitochondrial isolation.  Mitochondrial isolation was carried out using the Mitochondria Isolation Kit for Cultured Cells (Pierce, Rockford, IL, USA) according to the protocol recommended by the manufacturer. The lysate was then subjected to western blot analysis.

Quantitative polymerase chain reaction.  RNA extraction was carried out using Trizol reagent (Life Technologies, Gaithersburg, MD, USA). Complementary DNA was generated from 1 µg RNA with avian myeloblasotosis virus reverse transcriptase (Promega, Madison, WI, USA). Quantitative real‐time polymerase chain reaction was carried out using a LightCycler (Idaho Technology, Salt Lake City, UT, USA) as described previously.^{( 17 )} β‐Actin was used as a control. The primer sequences were as follows: survivin forward primer, 5′‐TCCGGTTGCGCTTTCCT‐3′; survivin reverse primer, 5′‐TCTTCTTATTGTTGGTTTCCTTTGC‐3′;^{( 18 )} β‐actin forward primer, 5′‐GAAAATCTGGCACCACACCT‐3′; and β‐actin reverse primer, 5′‐GTTGAAGGTAGTTTCGTGGAT‐3′.

Transfection.  Transfection was carried out using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) in Opti‐MEM I Reduced Serum Media (Invitrogen) according to the manufacturer's recommendations. siRNA for survivin (Hs_BIRC5–5_HP Validated siRNA) and non‐silencing control siRNA were obtained from Qiagen.

Luciferase reporter assay.  Survivin luciferase reporter plasmids were constructed as reported previously.^{( 19 )} Cells were transfected with 1 µg luciferase reporter plasmid and 0.02 µg control plasmid pRL‐SV40, a Renilla luciferase expression plasmid. At 6 h after transfection, the cells were treated with various drugs or inhibitors. The assays were carried out using a dual luciferase reporter assay system (Promega) according to the manufacturer's instructions. Firefly luciferase activity was normalized against Renilla luciferase activity.

Statistical analysis.  Statistical analyses were carried out using StatView 5.0 (SAS Institute, Cary, NC, USA). Differences between values were determined using the unpaired t‐test. Significance was indicated when the P‐value was less than 0.05.

Results

Growth inhibition of esophageal cancer cell lines by oxaliplatin.  Growth curves are shown in Fig. 1a. At 72 h after initial exposure, the IC₅₀ for oxaliplatin was determined to be 2.5 µM for both cell lines, whereas the IC₅₀ for cisplatin was 2.5 and 3 µM for TE3 and TE7, respectively. We confirmed that oxaliplatin inhibited the growth of TE3 and TE7 cells using a clonogenic survival assays. Oxaliplatin at 2.5 µM strongly inhibited cell growth (Fig. 1b). After 10 days, the oxaliplatin‐treated cells showed less than 20% survival for both cell lines compared with the untreated control groups (Fig. 1c). These concentrations were used in subsequent experiments unless otherwise noted.

Figure 1.

(a) Growth curves for TE3 and TE7 cells following treatment with oxaliplatin (OHP) and cisplatin (CDDP). (b) Clonogenic survival assay. Colony formation was strongly inhibited 10 days after OHP treatment (2.5 µM). (c) Image analysis of three independent experiments indicated that OHP significantly inhibited cell proliferation.

Effect of oxaliplatin on the cell cycle in esophageal cancer cell lines.  After oxaliplatin treatment, both cell lines showed an accumulation of cells in S phase, followed by G₂/M‐phase arrest (Fig. 2a). Later, TE3 cells started to accumulate at sub‐G₁ phase, indicating that cells were undergoing apoptosis. In contrast, the TE7 cell line showed an accumulation of a post‐G₂‐cell population, indicating the existence of cells with higher ploidy (>4 N). The cell cycle analysis results are summarized in Fig. 2b.

Figure 2.

Oxaliplatin (OHP) induces different modes of cell death. (a) Representative results of flow cytometric cell‐cycle analysis of TE3 and TE7 cells. A sub‐G₁ population appeared after treatment with 2.5 µM OHP in the TE3 cell line, whereas cells with higher DNA content (>4 N, post‐G₂) accumulated in the TE7 cell line after treatment. (b) A summary of the percentage of cells in G₂/M, S, and G₀/G₁ phase for TE3 and TE7 cells. Accumulation of cells at S phase and in G₂/M was evident in both cell lines.

Oxaliplatin induces different modes of cell death.  We then investigated changes in cell morphology after oxaliplatin treatment using light microscopy. We observed an increase in a population of shrunken cells with DNA condensation or nuclear fragmentation (Fig. 3a, indicated by arrows with dotted lines) in oxaliplatin‐ as well as cisplatin‐treated TE3 cells. In contrast, a large population of TE7 cells with two or more nuclei was observed, which is indicative of mitotic catastrophe (Fig. 3a, indicated by bold arrow). Mitotic catastrophe describes a unique mode of cell death, indicated by a giant multinucleated cell morphology.^{( 20 )}

Figure 3.

(a) Morphological changes in TE3 and TE7 cells 72 h after a 24‐h treatment with oxaliplatin (OHP; 2.5 µM for both) or cisplatin (CDDP; 2.5 µM for TE3, 3 µM for TE7). Apoptotic cells were observed frequently in treated TE3 cells, whereas multiple or enlarged nuclei were observed in TE7 cells after treatment. Cells with fragmented nuclei or condensed DNA are labeled with dotted arrows and giant multinucleated cells are labeled with bold arrows. Original magnification ×400. (b) Percentage of cells with two or more nuclei (undergoing mitotic catastrophe). TE3 and TE7 cells were treated as in (a), fixed at the time indicated, stained, and observed using light microscopy. A gradual increase in the percentage of cells undergoing mitotic catastrophe was observed in TE7 cells after OHP treatment. Changes in treated groups were compared with control cells. The statistical significance (P‐value) is indicated.

Next, we focused on the occurrence of mitotic catastrophe in oxaliplatin‐treated cells. Cells that displayed features indicative of mitotic catastrophe, namely cells with two or more evenly stained nuclear fragments, following oxaliplatin and cisplatin treatment were counted (Fig. 3b). Both drugs induced a higher rate of mitotic catastrophe in TE7 compared to TE3. Oxaliplatin induced a gradual increase in cells undergoing mitotic catastrophe in TE7 until 6 days after treatment, whereas cisplatin‐induced mitotic catastrophe decreased after 72 h.

Another distinct feature observed in the oxaliplatin‐treated TE7 cells was nuclear enlargement. Although these cells had single nuclei, the nuclei were gigantic compared with the control cells. Cells with nuclear enlargement were defined as containing a single nucleus with a diameter 1.5‐fold larger than the mean diameter of nuclei of control cells. After treatment with oxaliplatin, a gradual increase was observed in the percentage of TE7 cells with enlarged nuclei: 11.1% at 24 h, 21.3% at 48 h, and 35.6% at 72 h.

Expression of cell cycle and mitotic catastrophe‐associated proteins.  To gain insights into the molecular mechanisms responsible for the growth inhibition and mitotic catastrophe events observed following oxaliplatin treatment, we analyzed the expression of cell‐cycle regulatory proteins, namely p21^{waf1 / cip1}, cdc2 (Fig. 4a), and several proteins that are shown to be associated with mitotic catastrophe, namely, 14‐3‐3σ, aurora kinases, and survivin (Fig. 4b,c). The expression of p21^waf1/cip1 was downregulated strongly in both cell lines following oxaliplatin treatment. A decrease in the expression of 14‐3‐3σ was evident in TE7 cells following oxaliplatin treatment. Downregulation of aurora A kinase and a slight decrease in aurora B kinase were also noted in both cell lines. Of interest was the expression of survivin (Fig. 4c), which was more strongly suppressed in TE7 cells compared with TE3 cells after oxaliplatin treatment. Cisplatin also inhibited survivin expression in both cell lines but to a lesser extent in TE7 cells. We further examined expression at the mRNA level (Fig. 4d). Oxaliplatin significantly inhibited the level of survivin mRNA. Next, we tested whether this was due to the inhibition of survivin promoter activity (Fig. 4e). Both drugs demonstrated significant inhibition of survivin promoter activity at 48 h after treatment.

Figure 4.

Expression of (a) cell cycle‐regulatory proteins and (b) mitotic catastrophe‐associated proteins in TE3 and TE7 cells following 24‐h treatment with 2.5 µM oxaliplatin (OHP). Cells were harvested after 24 or 48 h and subjected to western blot analysis. Survivin (c) protein (by western blot), (d) mRNA expression (by quantitative polymerase chain reaction), and (e) promoter activity (by reporter assay) in TE3 and TE7 cells after 24‐h treatment with OHP (2.5 µM for both cell lines) and cisplatin (CDDP) (2.5 µM for TE3; 3 µM for TE7). Changes in treated groups were compared with control cells. The statistical significance (P‐value) is indicated.

Survivin inhibition induces mitotic catastrophe.  Transfection of siRNA targeted against survivin strongly inhibited survivin protein expression in TE3 and TE7 cells (Fig. 5a). This was accompanied by morphological changes indicative of cells undergoing mitotic catastrophe (Fig. 5b). A significant increase (20.3% in TE3 and 44.6% in TE7) in multinucleated cells was observed at 72 h compared with control cells transfected with non‐silencing control siRNA in both cell lines (Fig. 5c).

Figure 5.

Survivin knockout using small interfering RNA (siRNA). (a) Survivin protein expression was downregulated in cells transfected with survivin‐targeted siRNA (400 nM siRNA for TE3 cells and 200 nM for TE7 cells). Treated cells were harvested at the indicated time interval and subjected to western blot analysis. (b) At 72 h after transfection, mitotic catastrophe was observed in cells transfected as in (a) with survivin‐targeted siRNA. Original magnification ×400. (c) After 72 h, the percentage of multinucleated cells was greatly increased in cells transfected with survivin‐targeted siRNA. The statistical significance (P‐value) is indicated.

Oxaliplatin induces apoptosis in TE3 cells.  We next confirmed that oxaliplatin induces apoptosis in TE3 cells. The annexin V assay (Fig. 6a), which detects early changes in membranes during apoptosis, indicated that there was an increase in cells in early apoptosis in TE3 cells after oxaliplatin treatment. Cisplatin, however, induced less apoptosis in TE3 but a gradual increase was noted in TE7 cells. Moreover, over time there was a progressive increase in the fraction of cleaved PARP protein in oxaliplatin‐treated TE3 cells (Fig. 6b), which is a biochemical change specific to apoptosis. We further investigated caspase expression (Fig. 6c). We found an increase in the fraction of cleaved caspase‐3 in oxaliplatin‐treated TE3 cells, which indicates that caspase‐3 is activated in oxaliplatin‐induced apoptosis. However, there was no increase in the fractions of cleaved caspase‐9 (Fig. 6c), which is involved in apoptosome formation, cleaved caspase‐8, which is often indicative of death‐receptor pathway activation, or cleaved caspase‐7, another downstream effector caspase (data not shown). Experiments using caspase inhibitors (Fig. 6d) confirmed that caspase‐3 inhibitor was able to block oxaliplatin‐induced apoptosis. However, significant inhibition by caspase‐9 inhibitor was also noted, indicating the caspase‐9 was also active during oxaliplatin‐induced apoptosis.

Figure 6.

Oxaliplatin (OHP)‐induced apoptosis. (a) Annexin V apoptosis assay. Following 24‐h treatment with OHP (2.5 µM for both) and cisplatin (CDDP; 2.5 µM for TE3, 3 µM for TE7) treatment, cells were labeled with annexin V–fluorescein isothiocyanate antibody and detected using flow cytometry. Positive cells were predominantly in early apoptosis. With OHP treatment, an increase in apoptotic cells was detected in the TE3 cell line, but was limited in TE7 cells. CDDP induced apoptosis to a lesser extent in TE3 cells but a higher extent in TE7 compared with OHP treatment. Statistical significance is indicated by the P‐value. (b) Western blotting analysis of polyADP‐ribose polymerase (PARP) cleavage confirmed that apoptosis was occurring in TE3 cells after OHP treatment (an increase in the fraction of cleaved PARP), whereas PARP remained uncleaved in TE7 cells until 72 h. (c) Caspase expression. Cleaved caspase‐3 was evident at 72 h in TE3 cells. Actin was included as a loading control. (d) Apoptosis assessed by annexin V assay with flow cytometry. Caspase‐3 (z‐DEVD‐fmk), caspase‐9 (z‐LEHD‐fmk), and pan‐caspase (z‐VAD‐fmk) inhibitors significantly blocked OHP‐induced apoptosis in TE3 cells.

We also examined the expression of bcl‐2 family proteins, which are associated with apoptosis signaling. The anti‐apoptotic protein Bcl‐2, as well as members of the pro‐apoptotic BH3‐only proteins, Bad, Bid, Puma, and Noxa, were similarly downregulated in both cell lines (Fig. 7a). Translocation of cytochrome c from the mitochondria into the cytoplasm can activate downstream effector caspases, leading to apoptosis. However, in this case, the observed increase in the mitochondrial fraction of cytochrome c was not accompanied by its release into the cytosol (Fig. 7b). Mitochondrial membrane potential depolarization (Fig. 7c), as measured by JC‐1 dye, indicated an increase in the number of cells with loss of mitochondrial membrane potential after oxaliplatin treatment in TE3 cells, but a substantial increase was evident in late phase at 72 h.

Figure 7.

(a) Expression of Bcl‐2 family members. (b) Cytochrome c expression in the mitochondrial and cytosolic fractions at 48 h after oxaliplatin (OHP) treatment. Citrate synthase and glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH) were used as loading controls for the mitochondrial and cytosolic fractions, respectively. (c) Mitochondrial membrane potential was measured by flow cytometry following JC‐1 dye staining. The graph shows the fraction of cells with loss of mitochondrial membrane potential (detected by green dye in the cytosol). OHP treatment led to an increase in the number of cells with depolarized mitochondrial membrane in TE3, with a substantial increase evident in late phase (72 h).

Survivin regulation following oxaliplatin treatment.  

 Proteasome‐mediated degradation 

The regulation of survivin inhibition after oxaliplatin treatment was further investigated in TE7 and another gastric adenocarcinoma cell line, MKN45, which downregulates survivin following oxaliplatin treatment.^{( 21 )} Downregulation of survivin expression after oxaliplatin treatment was partly rescued in both cell lines by the addition of the proteasome inhibitor lactacystin (Fig. 7a).

Transcription factors: Stat3 and Sp1  In search of possible regulators of survivin downregulation by oxaliplatin, we found that phosphorylated Stat3 (i.e. activated Stat3) and Sp1 expression decreased in oxaliplatin‐treated cells compared with untreated controls in both the TE7 and MKN45 cell lines (Fig. 8b,c). However, further experiments using Stat3 and Sp1 inhibitors showed that inhibition of Stat3 phosphorylation upregulated survivin promoter activity in TE7 and did not show significant inhibitory effects on other cells (Fig. 8d). Meanwhile, the Sp1 inhibitor mithramycin A consistently inhibited survivin promoter activity in these cell lines (Fig. 8e).

Figure 8.

Survivin regulation. (a) Effects of a proteasome inhibitor (lactacystin) on survivin expression following oxaliplatin (OHP) treatment. (b) Total Stat3 and phosphorylated (Tyr705) Stat3 protein expression. (c) Sp1 expression at 48 h after OHP (2.5 µM) treatment. Actin was used as a loading control. (d) Effect of Stat3 inhibitor peptide and (e) Sp1 inhibitor (mithramycin A) on survivin promoter activity at 24 h after transfection in two esophageal (TE7 and TE3) and two gastric (MKN45 and MKN28) cancer cell lines. Luciferase activity was normalized against Renilla luciferase activity, and the control (without treatment) was adjusted to 100%. The statistical significance (P‐value) is indicated.

Discussion

Epidemiological studies have indicated that there is geographic variation in the incidence of esophageal cancer: SCC occurs at high incidence in Asia and Africa, whereas ADC shows increased incidence in several Western countries.^{( 22 )} Targeting these two major types of esophageal cancer, we used an SCC‐derived cell line (TE3) and an ADC‐derived cell line (TE7) to explore the efficacy and mechanism of oxaliplatin in the treatment of esophageal cancer. DNA damage induced by oxaliplatin could activate various checkpoint proteins, leading to cell‐cycle arrest.^{( 2} , ³ , ⁴ , ²³ , ^{24 )} Among various factors, a decrease in p21^waf1/cip1 protein levels in these two cancer cell lines after oxaliplatin treatment could partly contribute to the abrupt S‐phase entry, resulting in G₂/M arrest.^{( 9 )} Following this, however, the two cell lines displayed different responses to oxaliplatin. The TE3 cell line showed a marked accumulation of cells in sub‐G₁, consistent with the shrunken morphology of apoptotic cells; the TE7 cell line showed an increase in cells in the post‐G₂ (>4 N) population, along with the appearance of giant multinucleated cells, compatible with mitotic catastrophe.

Mitotic catastrophe is a defined cell‐death process that occurs in addition to or as an alternative to apoptosis.^{( 20 )} Studies have shown that bleomycin, etoposide, and radiation can all induce mitotic catastrophe.^{( 25} , ²⁶ , ^{27 )} Here, we showed that TE7 cell death following oxaliplatin treatment occurred primarily through mitotic catastrophe (although some apoptotic cells were present), whereas SCC TE3 cell death occurred via apoptosis (but also had some cells undergoing mitotic catastrophe). Cisplatin treatment showed a similar tendency but was able to induce a higher rate of apoptosis in TE7. The primary mechanism of cell death of a particular cell line in response to oxaliplatin thus appeared to depend on cell type. However, it has also been suggested that mitotic catastrophe is an alternative death pathway that overcomes or bypasses resistance to apoptosis. A cytotoxic agent, curcumin, can successfully induce cell death through mitotic catastrophe in an apoptotic‐resistant cell line.^{( 28 )}

There are several mitosis‐related proteins that cause mitotic catastrophe. These include 14‐3‐3σ, aurora A kinase, aurora B kinase, p21^waf1/cip1, and survivin.^{( 29} , ³⁰ , ³¹ , ³² , ^{33 )} Among these molecules, we found that oxaliplatin largely decreased survivin expression in TE7 cells that underwent drastic mitotic catastrophe, whereas downregulation of survivin was limited to the early phase response in TE3 cells (at 24 h after oxaliplatin treatment). Another SCC cell line, TE8, also demonstrated only modest downregulation of survivin and did not undergo mitotic catastrophe (data not shown). Cisplatin also decreased survivin in a similar pattern, but showed a weaker inhibition in TE7 cells. Both drugs were able to repress survivin promoter activity but it remained unclear why despite inhibition of survivin promoter activity in TE3, its expression was upregulated at 48 h. Survivin is essential for cell survival and has been found to be upregulated in some human cancers, including esophageal cancer.^{( 34} , ^{35 )} It has been reported that lack of survivin causes a defect in cell division leading to binucleation or to bilobed nuclei and polyploidy.^{( 36} , ^{37 )} To address whether survivin plays a central role in protecting cells from mitotic catastrophe, we used siRNA to decrease survivin expression in the cell lines. We found that sufficient suppression of survivin resulted in mitotic catastrophe in TE7 and TE3 cells (Fig. 5), and also in TE8 cells (data not shown). These findings suggest that oxaliplatin‐induced mitotic catastrophe may be mediated by survivin inhibition. Downregulation of 14‐3‐3σ protein and p21^waf1/cip1 protein by oxaliplatin might have contributed to the formation of multinucleated cells in the TE7 cell line, as proposed by other investigators.^{( 29} , ^{31 )}

The oxaliplatin‐induced apoptosis could be partially caused by the early downregulation of survivin. Survivin was reported to inhibit apoptosis by binding to caspase‐3 in early studies.^{( 30} , ³⁸ , ^{39 )} Inhibition of survivin could thus activate apoptosis. Caspase‐3 was indeed found to be activated along with PARP cleavage in TE3 after oxaliplatin treatment. However, considering that the caspase‐9 inhibitor demonstrated an inhibitory effect on oxaliplatin‐induced apoptosis without detectable protein cleavage, we could not rule out the involvement of other caspases in oxaliplatin‐induced apoptosis. We also investigated Bcl‐2 family members involved in the mitochondrial (intrinsic) pathway to determine whether they were functioning in upstream initiator signaling for apoptosis. Involvement of the Bcl‐xL and Bax proteins in oxaliplatin‐mediated apoptosis was reported previously for colon cancer cells.^{( 10} , ^{12 )} Here, in esophageal cancer cells, Bcl‐xL and Bax showed no changes in expression level but Bcl‐2 was downregulated following oxaliplatin treatment. Furthermore, the lack of cytochrome c release from mitochondria and limited loss of mitochondrial membrane potential in early phase indicated that the apoptosis occurring in TE3 cells was not initiated by an intrinsic pathway. Thus, these findings indicate that oxaliplatin‐induced apoptosis in SCC was caspase‐dependent. In addition, oxaliplatin specifically downregulated BH3‐only proteins, with the exception of Bim, in esophageal cancer cells (Fig. 6b).

Finally, we investigated the underlying mechanism of how oxaliplatin reduced survivin levels using both TE7 cells and another gastric cancer cell line, MKN45, both of which demonstrated significant survivin downregulation and drastic mitotic catastrophe after oxaliplatin treatment.^{( 21 )} Similar to proteasome‐mediated degradation of p21^waf1/cip1 by oxaliplatin,^{( 9 )} we found that survivin was subject to oxaliplatin‐induced degradation at the protein level. On the transcriptional level, it was reported that survivin is a downstream target of Stat3 in breast cancer cells;^{( 40} , ^{41 )} we explored this possibility and found inhibition of Stat3 phosphorylation in both cell lines by oxaliplatin treatment (TE7 and MKN45; Fig. 8b). However, when we attempted to verify the Stat3–survivin axis, we found that inactivation of Stat3 phosphorylation did not significantly downregulate survivin promoter activity in these cells (Fig. 8d). Next, we examined Sp1, a transcription factor that was also reported to be involved in the transcriptional activation of survivin.^{( 42 )} Sp1 protein was similarly downregulated by oxaliplatin treatment. We then confirmed that inhibition of Sp1 binding could actually downregulate survivin promoter activity in the cell lines used (Fig. 8e). Survivin is regulated through various pathways^{( 43 )} and other regulatory mechanisms may therefore come into play; however, it appears that these two mechanisms (i.e. proteasome degradation and Sp1 downregulation) can mediate survivin inhibition after oxaliplatin treatment, at least in certain gastrointestinal cell lines.

In vitro data have indicated a clinically achievable dosage for oxaliplatin in treating esophageal cancer.^{( 44 )} The ability of oxaliplatin to downregulate survivin expression in cancer cells also suggests that it may have a wide range of clinical applications. In conclusion, the ability of oxaliplatin to induce different modes of cell death in different types of esophageal cancer at relatively low concentrations indicates that it may be highly efficacious in killing tumor cells, and warrants further study of clinical applications for oxaliplatin.