This article is mentioned in:

Introduction

Colorectal cancer is among the leading 25 causes of global mortality (1). Age standardized rates for colorectal cancer incidence and mortality rank fourth as reported by the World Health Organization Globocan Project. Although early detection and/or treatment have improved the outcome substantially, a relatively high number of new cases and deaths are still expected for 2013 (2).

Currently used antineoplastic drugs against colorectal cancer include antimetabolites (3,4), alkylating agents (5) and camptothecin analogs (6). The antimetabolites, 5-fluorouracil (5-FU) and capecitabine, are utilized in various combination regimens. The response rate to 5-FU was shown to increase following leucovorin (LV) modulation but the survival rate remained unchanged. Addition of either oxaliplatin (FOLFOX) (7) or irinotecan (FOLFIRI) (8) to 5-FU and LV combination led to prolongation of progression-free survival (PFS) and higher response rates with an acceptable tolerability profile in patients with metastatic colorectal cancer when used as first-line therapy. Overall response and survival rates by FOLFOX and FOLFIRI regimens are similar (9). Regimens in which the alkylating agent oxaliplatin was added to fluorouracil (FUOX) or capecitabine (CAPOX) also yielded no significant difference between the two study arms, and the median survival rate remained between 18 and 21 months (10). Targeted therapies with monoclonal antibodies against the epidermal growth factor receptor (EGFR) were found to be correlated with prolonged survival rates; however, therapeutic efficacy decreased significantly in patients with K-RAS mutations (11,12). Anti-VEGF agents such as bevacizumab and aflibercept also demonstrated beneficial effects (13,14). Bevacizumab was evaluated in combination with either capecitabine plus oxaliplatin (XELOX) or with FOLFOX in patients with metastatic colorectal cancer. Addition of bevacizumab significantly prolonged PFS; however, overall survival and response rates were not improved (15). Today the prognosis of colorectal cancer is predicted by tumor staging. Overall survival rate in stage IV and recurrent colorectal cancer ranges between 15 and 25 months depending on the combination regimen employed. Therefore, new drugs or combinations are required to improve the prognosis of colorectal cancer.

Alkylphosphocholines (APCs) are a novel class of antineoplastic agents. APCs are structurally related to alkyllysophospholipids (edelfosine and ilmofosine) with the exception of the glycerol backbone, which was deemed unnecessary for cytotoxic activity (16,17). Unlike classical chemotherapeutics, APCs target the cell membrane instead of DNA. At clinically relevant doses, they interfere with phospholipid turnover; hence with cell signaling and survival pathways (18). Such distinctive features may facilitate their therapeutic efficacy, and render them potential candidates for combination therapies (19).

The prototype of APCs, hexadecylphosphocholine (miltefosine), did not meet expectations due to low response rates and high gastrointestinal toxicity. Today, miltefosine use is limited to topical treatment of skin metastases and oral leishmaniasis treatment (18). By structural modifications of APCs, many of these pending obstacles could be overcome. The most recent APCs, erucylphosphocholine (ErPC) and its homolog erufosine (erucylphospho-N,N,N-trimethylpropanolamine, ErPC3) are quite promising with favorable pharmacokinetic and pharmacodynamic properties. They have significant cytotoxic efficacy but cause less gastrointestinal toxicity. They possess the ability to cross the blood-brain barrier and to accumulate in the brain tissue, which indicates their potential for the treatment of brain tumors (19). Owing to their structure with a 22-carbon chain and a ω-9 cis-double bond, they have reduced myelotoxic and hemolytic effects (20–22) and because of the latter property, the agents are also the first intravenously applicable APCs.

In vitro, the antiproliferative effect of erufosine was demonstrated in various cell lines of human origin such as chronic myeloid leukemia (CML; alone and when combined with imatinib) (23), acute myeloid leukemia (AML) (24,25), chronic lymphocytic leukemia (CLL) (26), multiple myeloma (MM) (20,27), bladder carcinoma (28), breast carcinoma (29) and oral squamous cell carcinoma (30). Bladder carcinoma, AML and multiple myeloma cell lines were more sensitive to erufosine in vitro, and the IC50 values ranged between 4 and 14 μM (25,27,28). However, higher IC50 values (22–41 μM) were detected in CLL, breast and oral squamous carcinoma cell lines (26,29,30). Erufosine also reduced colony formation in human MM, breast and pancreatic carcinoma cells (22) and inhibited migration in human MM cells (20).

Erufosine induced apoptosis in CLL (26), AML (24,25), acute lymphocytic leukemia (ALL) (31), human glioblastoma (31–33), prostate (34) and oral squamous carcinoma (30) cell lines of human origin. Its cytotoxic effect was decreased considerably by caspase inhibitors (25,26). For this reason, part of its antineoplastic activity was associated with apoptosis (35). Activation of executive procaspase-3 and cleavage of its substrate poly(ADP-ribose) polymerase (PARP) are well documented time- and concentration-dependent effects of erufosine (24–26,31–34). Erufosine-induced apoptosis was modulated via the JNK 1/2, Raf/MEK/ERK and PI3K/Akt/mTOR signaling pathways (25,27,30,34).

The antineoplastic effects of APCs on colorectal cancer cell lines have been previously reported (36,37). Miltefosine was more effective in the colon adenocarcinoma cell line, HT29 (IC50, 3.1 μmol/l) as compared to mammary carcinoma cell lines (IC50, 29.4–69.9 μmol/l) (36). Against this colorectal cell line, miltefosine was more effective when compared to other APC congeners such as octadecenyl-(trans-9.10)-phosphocholine, octadecenyl-(cis-9.10)-phosphocholine and octadecylphosphocholine (IC50, 5.8, 17.8 and 4.4 μmol/l, respectively) (37). An antineoplastic effect of erufosine in colorectal cancer cell lines has not yet been reported. Therefore, the aim of the present study was to investigate and compare the antiproliferative, antimigratory and pro-apoptotic effects of erufosine in colorectal cell lines of human (SW480) and rat (CC531) origin.

Materials and methods

Cell culture

The colon adenocarcinoma cell lines, SW480 (human) and CC531 (rat), free of pathogenic contamination, were grown as monolayers in RPMI-1640 medium supplemented with 10% fetal calf serum (FCS) and L-glutamine (2 mM). The cell lines were maintained in an incubator with a humidified atmosphere (5% CO2 in air at 37°C). Cells were passaged two or three times a week to maintain them in a logarithmic growth phase. For isolation and propagation, the medium was discarded, and then the cells were washed with phosphate-buffered saline (PBS), trypsinized (0.25% trypsin/EDTA), pelleted at 1,500 rpm for 5 min and re-suspended at the desired concentration in RPMI-1640 medium.

Cell proliferation assay

Cell proliferation was assessed by MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide) dye reduction assay as described by Mosmann, with some modifications (38). In brief, MTT (Sigma, Munich, Germany) solution (10 mg/ml in PBS) was added (10 μl/well). Plates were further incubated for 3 h, and following removal of the medium, the formazan crystals were dissolved by the addition of 100 μl solvent (0.04 N HCl acid in 2-propanol) per well and then by thoroughly mixing. Optical density was measured at a 540-nm wavelength (690 nm reference wavelength) using an ELISA plate reader (Anthos Mikrosysteme GmbH, Krefeld, Germany). Cell doubling time (DT) was calculated by using the Patterson formula: Td = T × lg2/(lgN2 - lgN1) where Td is the doubling time (in hours), N is the number of cells and T is the time for cell growth from N1 to N2. Cell growth rates (in hours) were calculated by the following formula: Growth rate (μ) = ln (N2/N1)/T2 - T1.

The optimal cell number to be seeded was determined before assessment of the antiproliferative effects of erufosine in both cell lines. For the growth curves, SW480 and CC531 cells were seeded in 96-well microplates at final concentrations of 2×103, 4×103, 8×103 cells/well and 2×103, 4×103 cells/well, respectively. For exposure to erufosine, cells were seeded into 96-well microplates (2×103 cells/100 μl medium/well for SW480 and 4×103 cells/100 μl medium/well for CC531) and incubated with increasing concentrations (3.1, 6.3, 12.5, 17.7, 25, 35.4, 50, 70.7 and 100 μM) for 24, 48 and 72 h. Cell survival rates were expressed as the percentage of untreated controls at 24, 48 and 72 h. IC50 was calculated by the equation of logarithmic regression trendline.

In vitro wound healing (scratch) assay

SW480 and CC531 cells were seeded (10,000 cells/well) in 24-well plates and allowed to attach to the surface under standard incubation conditions for 24 h. After 24 h, the confluent cell monolayers were scratched in a straight line using a 200-μl sterile plastic pipette tip, as previously described (39). The cells were then carefully rinsed with culture medium to remove free-floating cells and debris. Then, erufosine was added at final concentrations of 1.56, 3.125 and 6.5 μM/well, and the effect on wound healing was monitored. Scratch zones representative for each cell line were photographed at 12, 24, 36 and 48 h by the Axio Observer.Z1 microscope (Carl Zeiss AG, Oberkochen, Germany). Each experiment was conducted in triplicate wells for each concentration of erufosine and the control. AxioVision Rel. 4.8 software was used for the measurements. For SW480 cells, cells that had migrated into the scratched area were counted within a 400×400 μm frame, which was created by the region of interest (ROI) function, enabling to select 3 random regions in the scratched area. For the CC531 cells, the distance between the wound edges was measured. For both cell lines, three random measurements were made per photographed sample at 12 h, which was used as baseline. Both cell lines were studied in parallel, and the duration of the microscopic procedure was kept the same to exclude environmental condition-related differences in wound healing responses.

Caspase assay

Caspase-3/−7 enzymatic activity was measured by Apo-ONE Homogeneous Caspase-3/7 assay (G7792 Promega, Germany), according to the manufacturer’s instructions. Briefly, SW480 and CC531 cells were seeded at a final concentration of 10,000 cells/well in a black 96-well plate, 24 h before drug treatment. Erufosine was added to the wells at final concentrations of 50 and 100 μM and incubated for 5 h in an incubator (5% CO2/95% O2 at 37°C). A standard assay (96-well, 200 μl final reaction volume) was conducted in triplicate involving three groups: blank [caspase reagent (CR) + cell culture medium without cells], negative control (CR + vehicle-treated cell culture) and assay (CR + treated cell culture). The contents of wells were gently mixed using a plate-shaker at 300–500 rpm for 5 h at room temperature (25°C). Immediately after this process, the fluorescence of each well was measured using a spectrofluorometer at an excitation wavelength range of 485±20 nm and an emission wavelength range of 530±25 nm. Blank values were subtracted from the experimental values to obtain the relative fluorescence units (RFUs).

Gene expression analysis

CC531 cells were seeded at a density of 200,000 cells in 25-ml flasks. After 24 h, the medium was changed and the cells were exposed to erufosine (12, 5 and 25 μM) for 48 h. For osteonectin expression, an additional experiment was conducted as follows. CC531 cells were seeded at a density of 400,000 cells/well of a 6-well plate. Medium was changed after 24 h, and the cells were incubated with erufosine (25 μM) for 24 and 48 h. Then, the cells were harvested, and the cell pellets were stored at −20°C until RNA isolation. Total RNA was isolated from CC531 cell pellets by the RNeasy Mini kit (Qiagen GmBH, Hilden, Germany) and then first strand cDNA was synthesized by Thermo Scientific Maxima reverse transcriptase kit (Thermo Scientific GmbH, Schwerte, Germany) according to the manufacturer’s protocols. Gene expression was studied by basic PCR protocol (Invitrogen GmbH, Karlsruhe, Germany). The following rat primers were used: ON, gagtttggcagctcagagga (left) and tctgct tctgagatgggtca (right); γ-tubulin, gatggcagtgacagcctagag (left) and gccgttccaagaggtagga (right); COL1A1, catgttcagctttgtgga cct (left) and gcagctgacttcagggatgt (right); COL1A2, cctggctctcgaggtgaac (left) and caatgcccagaggaccag (right). Each gene expression experiment was repeated twice. γ-tubulin served as the housekeeping gene. PCR products were analyzed by polyacrylamide gel electrophoresis (PAGE). Thermo Scientific pUC19 DNA/Mspl (Hpall) Marker was used for sizing and approximate quantification of the PCR products (Thermo Scientific). Lanes were framed automatically and corresponding band intensities were calculated by Quantity One 1-D Analysis software (Bio-Rad Laboratories Inc., Hercules, CA, USA). For normalization, the band intensity value of the target gene was divided by the band intensity of the housekeeping gene γ-tubulin.

Statistical analysis

The data are presented as mean values ± SD. The survival rates between the treatment groups were compared by ANOVA Dunnett’s test. Statistical differences in wound healing and caspase assays were calculated by one way, single factor ANOVA test. P-values ≤0.05 were considered to indicate statistically significant results.

Results

Growth curves of SW480 and CC531 cells

SW480 cells had a doubling time of 22.5±2.8, 20.0±4.1 and 24.3±9.3 h and growth rates of 0.031, 0.035 and 0.031 h−1 when seeded at a density of 2×103, 4×103 and 8×103 cells/well, respectively. CC531 cells had a doubling time of 18.8±1.0 and 22.6±1.4 h and growth rates of 0.037 and 0.031 h−1 when seeded at a density of 2×103 and 4×103 cells/well, respectively.

Antiproliferative effect of erufosine on SW480 cells

Survival rates of the SW480 cells following exposure to erufosine are shown in Fig. 1A. Erufosine exerted a concentration- and time-dependent antiproliferative effect. This effect was observed after 24 h with concentrations ≥12.5 μM. However, a significant difference from the control group was found at all concentrations at 48 and 72 h. The antiproliferative effect of erufosine significantly increased with longer exposure times at all concentrations. The IC50 values of erufosine in the SW480 cell line were 30.3, 9.9 and 3.4 μM following 24, 48, and 72 h of incubation, respectively. The relevant regression equations and their corresponding R2 values are provided in Fig. 1B. In the dose-response curve, a steep decline in the survival rate was observed at high erufosine concentrations (70.7 and 100 μM), preceded by a gradual decrease at low concentrations (Fig. 1A). The shoulder width and the amplitude became smaller in a time-dependent manner.

Figure 1 Survival curves of SW480 cells following exposure to increasing erufosine concentrations (3.13–100 μM). The values represent the means ± SD. The survival rates between the treatment groups were compared by ANOVA Dunnett’s test, and P-values ≤0.05 were considered to indicate statistically significant results. (A) Survival rates of erufosine-treated cells are indicated as the percentage of untreated cells. The antiproliferative effect (treated vs. untreated) was significant at ≥12.5 μM concentrations at 24 h, but at all concentrations following 48 and 72 h (P<0.05). (B) Inhibition of proliferation in erufosine-treated cells is indicated as the percentage of the untreated control. Logarithmic regression equations and corresponding R2 values are provided in the boxes. IC50 values were calculated by logarithmic regression trendline analysis: IC50 (24 h) = 30.31 μM, IC50 (48 h) = 9.91 μM, IC50 (72 h) = 3.43 μM.

Antiproliferative effect of erufosine on CC531 cells

Survival rates of the CC531 cells following exposure to erufosine are shown in Fig. 2A. An antiproliferative effect was observed at concentrations ≥25 μM after 24 h (P=0.003) and 48 h (P<0.0001) of incubation. After 72 h, no significant cytotoxic effect was demonstrated at the concentrations of 3.125, 6.25 and 12.5 μM (P>0.05). At the effective concentrations, the survival rates decreased significantly as the incubation time was prolonged. The IC50 values of erufosine in the CC531 cell line were 34.9, 29.7 and 25.4 μM following 24, 48 and 72 h of incubation, respectively. The relevant regression equations and their corresponding R2 values are provided in Fig. 2B. In the dose-response curve, a steep decrease in the survival rate was observed at high erufosine concentrations (70.7 and 100 μM), preceded by a gradual decrease at low concentrations (Fig. 2A). This shoulder effect was concentration- and time-dependent. The amplitude and the width were higher at 48 and 72 h when compared to the survival curve of the SW480 cells. The shoulder width became smaller after 48 h. There was no significant difference between the survival rates for 70.7 and 100 μM after 48 and 72 h. The survival curves after 48 and 72 h nearly overlapped with each other.

Figure 2 Survival curves of CC531 cells following exposure to increasing erufosine concentrations (3.13–100 μM). The values represent the means ± SD. The survival rates between the treatment groups were compared by ANOVA Dunnett’s test, and P-values ≤0.05 were considered to indicate statistically significant results. (A) Survival rates of the erufosine-treated cells are indicated as the percentage of untreated cells. The antiproliferative effect (treated vs. untreated) was significant only at ≥25 μM concentrations (P<0.05). (B) Inhibition of proliferation in erufosine-treated cells is indicated as the percentage of untreated control. Logarithmic regression equations and corresponding R2 values are provided in the boxes. IC50 values were calculated by logarithmic regression trendline analysis: IC50 (24 h) = 34.87 μM, IC50 (48 h) = 29.65 μM, IC50 (72 h) = 25.44 μM.

Wound healing assay in SW480 and CC531 cells

The wound healing assay was followed over 48 h for the low concentrations used in the cytotoxicity assay. Both cell lines displayed different migratory profiles during wound healing. SW480 cells migrated to the scratched area and closed the wound by forming colonies during the ‘healing’ process. In the CC531 cell culture, the wound edges approached each other to close the ‘wound.’ Untreated SW480 cells showed a time-dependent ability to close the gap and increasingly more cells migrated into the scratched area (Fig. 3A). This difference in migration was statistically significant at 24 (P=0.003), 36 and 48 h (P<0.0001). In SW480 cells, erufosine at a concentration of 1.56 μM inhibited wound healing, and this effect was significantly different from the control group after 12 (P=0.001), 24, (P=0.007), 36, (P=0.02) and 48 h (P<0.0001) (Fig. 3C). This effect observed following erufosine treatment (1.56 μM) indicated 53.8±13.3% inhibition of wound healing as compared to the control group at 48 h. In the SW480 cells, apoptotic morphological changes were noted at the 6.3 μM concentration even after 12 h of incubation and these signs were also observable, although less distinct, at the 3.1 μM concentration. Apoptotic cells lost contact with their neighbors, became rounded and detached from the surface displaying pronounced membrane blebbing. At a concentration of 1.56 μM, SW480 cells displayed no apparent signs of apoptosis. CC531 cells also had a good capacity to fill the gap (Fig. 3B). In the untreated group, the distance between the two edges of the wound progressively decreased. This decrease became statistically significant at 36 h (P<0.002), and was more pronounced at 48 h (P<0.0001), which corresponds to a wound healing of 62.2%. In the CC531 cells, erufosine inhibited wound healing capacity, which became evident by 48 h at all concentrations tested when compared with their corresponding baseline values (1.56, 3.125 and 6.25 μM; P<0.0001). As compared to the control following 48 h, the gap widths in the erufosine-treated cells were 183.5±75.2 (P=0.0025), 207.7±81.6 (P<0.0001) and 221.0±67.2% (P<0.0001) of the untreated cells following treatment with 1.56, 3.125 and 6.25 μM, respectively (Fig. 3D). Thus, erufosine-treated CC531 cells had an ~2-fold wider wound width after 48 h, as compared to the control group. A significant concentration-dependent effect was observed at erufosine concentrations between 1.56 and 6.25 μM at 24 and 36 h, respectively (P=0.0002). However, at 48 h this concentration-dependent effect could not be observed any longer between the treatment groups: 1 vs. 3 μM (P=0.277), 1 vs. 6 μM (P=0.064) and 3 vs. 6 μM (P=0.168). No morphological signs of apoptosis were detected in the CC531 cells at test concentrations.

Figure 3 Wound healing responses of colorectal cancer cell lines following exposure to erufosine. AxioVision Rel. 4.8 software was used for the measurements. The values represent the means ± SD (n=9). Statistical significance was calculated by one way, single factor ANOVA test, and P-values ≤0.05 were considered to indicate statistically significant results. (A) SW480 and (B) CC531 cells were incubated with erufosine (1.56–6.25 μM) and photographed at 12, 24, 36 and 48 h by an Axio Observer.Z1 microscope. Magnification in the photomicrographs is ×10. Bars indicate 200 μm. (C) Erufosine (1.56 μM) significantly delayed wound healing in the SW480 cells at 12, 24, 36 and 48 h when compared to the untreated group (P<0.0001). (D) Erufosine (1.56, 3.125 and 6.25 μM) significantly delayed wound healing in CC531 cells at 48 h when compared to the untreated group (*P<0.0025 untreated vs. 1.56 μM, **P<0.0001 untreated vs. 3.125 and 6.25 μM).

Caspase assay

In both cell lines, erufosine induced caspase-3/−7 enzymatic activity (Fig. 4). In the SW480 cells, erufosine had a concentration-dependent effect on caspase release. The fluorescence readings (RFU) were 851±18, 1686±55 and 2093±144 for the untreated, 50 and 100 μM erufosine-treated groups, respectively. This concentration-dependent response was significantly different from the control group (P=0.035 for 50 μM and P<0.001 for 100 μM) and also between both concentrations (P=0.0001). In the CC531 cells, RFU values were 1746±274, 2232±471 and 15916±3973 for the untreated and erufosine (50 and 100 μM)-treated groups, respectively. In the CC531 cells, there was no significant difference between the 50 μM and the untreated control group, whereas the difference between 100 μM and the untreated group was statistically significant (P=0.0035).

Figure 4 Caspase-3/−7 enzymatic activity in the SW480 and CC531 cells following 5 h of incubation with erufosine (50 and 100 μM). Spectrofluorometric measurement was conducted at an excitation wavelength range of 485±20 nm and an emission wavelength range of 530±25 nm. Blank values were subtracted from the experimental values to obtain the relative fluorescence units (RFUs). The values represent the means ± SD (n=3). Statistical significance was calculated by one way, single factor ANOVA test, and P-values ≤0.05 were considered to indicate statistically significant results. *P<0.001, **P=0.0035, ***P<0.001; statistically significant difference for erufosine-treated cells vs. untreated (control) cells.

Osteonectin, COL1A1 and A2 expression in CC531 cells

Erufosine (25 μM) incubation attenuated osteonectin expression by 12 and 11% following 24 and 48 h of incubation (Fig. 5). Erufosine (25 μM) abolished COL1A2 expression after 48 h (Fig. 6) but no change was observed for COL1A1 (Fig. 6).

Figure 5 Osteonectin (ON) expression in CC531 cells following exposure to erufosine (0–25 μM). Expression patterns were analyzed by RT-PCR and PAGE. γ-tubulin (γ-TBL) served as the housekeeping gene. For normalization, the band intensity value of the target gene was divided by the band intensity of γ-TBL. (A) Osteonectin expression after incubation with erufosine (12.5 and 25 μM) for 48 h. (B) Osteonectin expression after incubation with erufosine (25 μM) for 24 and 48 h.

Figure 6 COL1A1 and COL1A2 expression in CC531 cells following exposure to erufosine (0–25 μM) for 48 h. Expression patterns were analyzed by RT-PCR and PAGE. γ-tubulin (γ-TBL) served as the housekeeping gene. For normalization, the band intensity value of the target gene was divided by the band intensity of γ-TBL.

Discussion

This is the first report on the activity of erufosine in colorectal cell lines, SW480 and CC531. Our in vitro results showed concentration- and time-dependent antiproliferative effects in both cell lines. After 72 h, similar growth inhibitory effects were observed at a concentration of 100 μM in both cell lines, but at lower concentrations, erufosine was significantly more effective in SW480 cells. The IC50 value decreased progressively with prolonged incubation times in the SW480 cells but such a decline was not observed in the CC531 cells. Following 72 h of exposure to erufosine, the IC50 value for CC531 cells was nearly 4-fold higher than that of the SW480 cells. An antiproliferative effect of erufosine has been demonstrated in various human-derived cell lines, as outlined in Fig. 7. When compared to other types of human cancers, erufosine exhibits marked antiproliferative activity in SW480 cells, similar to bladder carcinoma and several (OPM-2 and RPMI-8226) MM cell lines. However, erufosine failed to show the same degree of efficacy in CC531 cells, and its efficacy in SW480 cells was only comparable to oral squamous carcinoma and CLL cells. CC531 cells were also shown to be less sensitive to the alkylating agent melphalan, when compared to SW480 cells (40).

Figure 7 Comparison of IC50 values (μM) of erufosine in various cell lines, following incubation for 24, 48 or 72 h. Our results for SW480 and CC531 cells are represented under CRC (colorectal carcinoma). BLC, bladder carcinoma [Konstantinov et al (28)]; BRC, breast carcinoma [Dineva et al (29)]; OS, oral squamous carcinoma [Kapoor et al (30)]; MM, multiple myeloma [Yosifov et al (20)]; AML, acute myeloid leukemia [Martelli et al (25)]; CLL, chronic lymphocytic leukemia [Königs et al (26)].

For SW480 and CC531 cells, shoulder-type survival curves were demonstrated (Fig. 3A and 4A). Survival rates declined gradually between 3.1 and 50 μM but a steep decrease was observed at high concentrations (70.7 and 100 μM). Shoulders on survival curves have been explained by two models. In a multi-hit target model, shoulders are due to the need to hit more than one target for cell-killing and zero or non-zero initial slopes at low doses become exponential as higher concentrations are reached. In the repair model, shoulders indicate that cells can proficiently repair the damage (41). If the repair process is saturated with some damage left, the dying process is initiated (42). Our results demonstrated marked differences between shoulder-type survivals in both cell lines. The shoulder had a higher magnitude and a broader width in CC531 cells. The shoulder magnitudes and final slopes of the survival curves may be altered by dose-dependent DNA repair mechanisms and/or interdependence of lethal/mutational responses (41). Cells display differential capacities to accumulate and repair sublethal damage. For example, the survival curve of HeLa cells had a small initial shoulder and a modest dose-rate effect. However, Chinese hamster cells displayed a broad shoulder and a large dose-rate effect. Such a difference was suggested to indicate the dominance of apoptotic cell death in HeLa cells (43). The shoulders on the survival curves of CC531 and SW480 cells (Figs. 1A and 2A) may indicate the synthesis of anti-apoptotic or pro-survival factor(s), which can help to repair erufosine-induced cellular damage(s). For example, high expression of the inhibitor of apoptosis protein (IAP) family (survivin, XIAP and other members) has been associated with colon carcinogenesis and resistance to chemotherapeutic agents (44). Other pro-survival factors such as insulin-like growth factor binding proteins, IGFBP3 and 7, CXCL5 and sirtuin 1 were also reported for colorectal cancer cell lines (45–47). Anti-apoptotic or pro-survival factor(s) against erufosine-induced cytotoxicity remain to be identified.

Cell survival depends on the balance between anti-apoptotic [Ras-Raf-MAPK/ERK and phosphatidylinositol 3′-kinase (PI3K)/Akt] and pro-apoptotic (SAPK/JNK) signaling mechanisms. Inhibition of survival pathways is a well-documented effect of APCs (19). Accordingly, the pro-apoptotic effect of erufosine has been demonstrated in various cell lines (24–27,30–35). Erufosine-induced apoptosis was correlated with hypo-phosphorylation (activation) of the retinoblastoma (Rb) protein, which inhibits Abl and JNK kinases as well as E2F transcription factors (35,48). The mitochondrial death pathway is another mechanism in erufosine-induced apoptosis (31–33,49). Caspase-3 was also suggested as one of the major determinants of erufosine-induced apoptosis (50). A concentration-dependent activation of caspase-3 and cleavage of the caspase-3 substrate PARP were detected in prostate cancer (PC3 and LNCaP) cells in response to erufosine treatment (12.5–25 μM) (34). In another study, erufosine (10 μM) treatment for 6–12 h resulted in depletion of procaspases and induced PARP cleavage (92%) in MM cells (OPM-2) (27). In freshly isolated CLL cells, erufosine (1–100 μM) cleaved PARP totally after 24 h in a concentration-dependent manner and a pan-caspase inhibitor completely abrogated apoptosis (26). Erufosine at concentrations of 30 and 50 μM enhanced caspase-3/−7 activity and cleavage of PARP in a concentration-dependent manner in oral squamous carcinoma cells (30). Our results showing enhanced caspase-3/−7 activity in both cell lines confirm these previous reports. Erufosine-induced caspase-3/−7 activity was concentration-dependent in SW480 cells, but was observed only at a high (100 μM) concentration in CC531 cells. Additionally, control caspase-3/−7 activity was higher (2-fold) in the CC531 than this activity in the SW480 cells, and this difference grew markedly (8-fold) following erufosine exposure (100 μM). However, erufosine (100 μM, 72 h) eventually exhibited the same antiproliferative effect in both cell lines. Caspase activation is not necessarily related to apoptosis in colorectal cancer cells (51). Beyond cell death, caspases are involved in many functions such as cellular formation and differentiation. Cytokines, tyr protein kinases, Ser/Thr protein kinases, protein phosphatases and G-proteins have been shown to be substrates of caspases. For this reason, caspase activation and cleavage of their substrates cannot be solely interpreted as irreversible processing of programmed cell death (52). Furthermore, caspase activation is not the unique determinant in programmed cell death. Caspase-independent cell death programs include autophagy, mitotic catastrophe, slow cell death and paraptosis (53). For example, APC-induced cell death was shown to be BCL-XL-sensitive and caspase-independent in human malignant glioma cells (54). According to our results, apoptosis is unlikely to be the major pathway for programmed cell death in CC531 cells at test concentrations, as supported by its characteristic survival curve (broad shoulder, high amplitude and steep slope) and lack of apoptotic signs during the wound healing assay (43).

APCs were shown to inhibit migration of human retinal pigment epithelial (RPE) cells by >90% in a concentration-dependent manner. Inhibition of migration was correlated with carbon chain length of the APC. For APCs longer than 20 carbon atoms (>C20), the IC50 value of RPE cell migration ranged between 0.1 and 1 μM, but increased to 10 μM for a chain length of 18C (55). Our wound healing assay showed that the IC50 of erufosine (22C) ranged from 1.56 to 3.1 μM in both cell lines. Retarding effects of antineoplastic drugs on wound healing are well known, but they vary in their efficacy to inhibit collagen synthesis. Only the DNA alkylating agent cisplatin strongly and specifically inhibited collagen synthesis in colon fibroblasts (56). Similarly, the alkylating agent ET-743 was shown to reduce COL1A1 mRNA levels up to 80% in scleroderma fibroblasts (57). We studied the effect of erufosine on COL1A1 and COL1A2 expression in CC531 cells, which was more prone to the retarding effect of erufosine on wound healing. Our results demonstrated that erufosine (25 μM) reduced the COL1A2 band intensity to an undetectable level but did not have any influence on COL1A1 expression. Osteonectin, a member of the ‘secreted protein acidic and rich in cysteine’ (SPARC) family of proteins is also involved in wound repair. SPARC silencing in human tendon fibroblasts did not interfere with cell proliferation but displayed an antifibrotic effect by reducing collagen I expression (58). Our results showed that erufosine (25 μM) caused a slight (12%) reduction in osteonectin expression.

In light of our findings, erufosine appears to be a promising chemotherapeutic agent in colorectal cancer. We observed marked differences in the antiproliferative, delayed wound healing and apoptotic effects between SW480 and CC531 cells. CC531, an immortalized cell line from a DMH-induced adenocarcinoma in rats, was less sensitive to erufosine than SW480 cells of human origin. Since a single human colorectal cancer cell line is inadequate to reflect the molecular heterogeneity of clinical colorectal tumors (59), future research should include various cell lines of different origins to further elucidate their survival mechanisms against erufosine.

References