XIAP-associated factor 1 (XAF1) negatively regulates the function of the X-linked inhibitor of apoptosis protein (XIAP), a member of the IAP family that exerts antiapoptotic effects. The extracellular signal-regulated kinase (ERK) pathway is thought to increase cell proliferation and to protect cells from apoptosis. The aim of the study was to investigate the correlation between the ERK1/2 signaling pathway and XAF1 in colon cancer.
Four human colon cancer cell lines, HCT1116 and Lovo (wildtype p53), DLD1 and SW1116 (mutant p53), were used. Lovo stable transfectants with XAF1 sense and antisense were established. The effects of dominant-negative MEK1 (DN-MEK1) and MEK-specific inhibitor U0126 on the ERK signaling pathway and expression of XAF1 and XIAP proteins were determined. The transcription activity of core XAF1 promoter was assessed by dual luciferase reporter assay. Cell proliferation was measured by MTT assay. Apoptosis was determined by Hoechst 33258 staining.
U0126 increased the expression of XAF1 in a time- and dose-dependent manner. A similar result was obtained in cells transfected with DN-MEK1 treatment. Conversely, the expression of XIAP was down-regulated. Activity of the putative promoter of the XAF1 gene was significantly increased by U0126 treatment and DN-MEK1 transient transfection. rhEGF-stimulated phosphorylation of ERK appeared to have little or no effect on XAF1 expression. Overexpression of XAF1 was more sensitive to U0126-induced apoptosis, whereas down-regulation of XAF1 by antisense reversed U0126-induced inhibition of cell proliferation.
Mitogen-activated protein kinases (MAPKs) or extracellular signal-regulated protein kinases (ERKs) comprise a family of serine/threonine protein kinases that mediate intracellular phosphorylation events linking receptor activation to the control of cell proliferation, chemotaxis, differentiation, and stress response.1 MAPKs are activated through phosphorylation of a specific threonine and tyrosine by dual specificity MAPK kinases referred to as MEKs. Homologous kinases in several sequential protein kinase cascades have been identified in yeast, mammalian cells, and plant, indicating conserved MAPK modules for signal transduction in eukaryotes.2–4 Recently, 2 new groups of protein kinases have been added to the family of mammalian MAPKs: the stress-activated protein kinases (SAPKs) or Jun kinases (JNKs) and the second family, the p38 kinases.5 Colon cancer displayed high levels of ERK1/2 phosphorylation.6 Inhibition of ERK1/2 phosphorylation in vitro by a synthetic MEK1/2 inhibitor, PD184352, decreased soft agar growth and inhibited the transformed phenotype of colon 26 cells. In vivo, PD184352 suppressed the growth of mouse and human colon tumor xenografts.7 Treatment of HCT116 human colon cancer cells with the specific mitogen-activated protein kinase inhibitor U0126 (5–50 μM) resulted in a time- and dose-dependent inhibition of ERK1/2 phosphorylation and induction of apoptosis.8 These results suggest that the inhibition of ERK1/2 phosphorylation is responsible for at least part of the induction of apoptosis.
The process of apoptosis is regulated at multiple levels by several regulatory mechanisms, including the inhibitors of apoptosis protein (IAPs) family. X chromosome-linked IAP (XIAP), the best investigated member of the IAP family, is differentially up-regulated in many human cancers, including colon cancer.9 Caspase inhibiting activity of XIAP is negatively regulated by some interacting proteins, including well-characterized Smac/DIABLO and HtrA2/Omi. XIAP-associated factor 1 (XAF1) is a novel antagonist of XIAP.10 The incubation of recombinant XIAP with caspase 3 in the absence or presence of XAF1 demonstrated that XAF1 blocked the inhibitory activity of XIAP for caspase 3.11
XAF1 was not only implicated as a tumor suppressor, we have also shown that it was involved in the cellular stress response as well.12 However, its role in apoptosis of colon cancer cells and its putative correlation with the MAPK pathway have not been investigated. In this study we report that inhibition of ERK1/2-induced XAF1 expression might be correlated with a different status of p53 in human colon cancer cells. XAF1 is essential for U0126-induced apoptosis through transcriptional regulation.
MATERIALS AND METHODS
Four human colon cancer cell lines including HCT1116 and Lovo (both with wild-type p53 gene), DLD1 and SW1116 (both with mutant p53 gene), were obtained from the American TypeCulture Collection (ATCC, Rockville, MD). The cells were cultured at 37°C with an atmosphere of 5% CO2 in RPMI 1640 medium containing 10% fetal bovine serum (Gibco BRL, Life Technologies, Grand Island, NY), 100 units/mL of penicillin, and 100 μg/mL of streptomycin.
The primers were designed according to the genomic sequence of XAF1 gene obtained from GenBank (Accession Nos. NT010718 and X99699) and we used the normal genomic DNA isolated from peripheral blood cells as a template. The XAF1 promoter region located at 5′-flanking of XAF1 gene13 (F291, from –109 to +164) was amplified by a polymerase chain reaction (PCR) method using the following primers: 5′-GGGGTACC AGA TCT CCT CCC TCC CTG AA-3′ and 5′-TCCGCTCGAG GTC TCC AGC TGC TTG TCC TC-3′. The fragment was digested by 2 restriction enzymes, KpnI and XhoI, and inserted into pGL3-basic plasmid (Promega, Madison, Wis) then confirmed by sequencing analysis.
Transient Transfection and Dual LuciferaseReporter Assay
Transient transfection was carried out using the LipofectAMINE 2000 reagent (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. Briefly, 1 × 105 cells were seeded in each well of a 24-well tissue culture plate. The cells were incubated until 70% confluent. Cells in each well were transfected with 0.8 μg of pGL-3 plasmid DNA by 1 μL LipofectAMINE 2000 reagent. The renilla luciferase reporter pRL-CMV plasmid (Promega) 0.01 μg/well was cotransfected as an internal control. After transfection for 6 hours, cells were transferred into normal medium. To inhibit ERK activation, HCT1116, Lovo, DLD1, and SW1116 cells were first transfected with pGL3-F291 plasmid for 6hours, then incubated with 30 μM of U0126 for 24 hours without antibiotics. In another experiment, the pGL3-F291 plasmid was cotransfected with DN-MEK1 or MEK1-active. Finally, the cells were collected and treated with passive lysis buffer (Promega). The firefly and renilla luciferase activities were measured using the Dual-luciferase Reporter Assay kit (Promega) according to the manufacturer's instructions with a luminometer (lumat LB 9507; Berthold, Bad Wildbad, Germany).
Establishment of Stable Transfectants of XAF1in Lovo Cells
Lovo cells were planted into 6-well plates (3 × 105 cells/ well) 18 hours before transfection and transfected with 4 μg/well of empty pcDNA3 plasmid (Lovo/vector), pcDNA3-XAF1 (Lovo/XAF1), and pcDNA3-XAF1 antisense (Lovo/XAF1-as). All of the plasmids were constructed in our previous study.14 LipofectAMINE 2000 reagent (Invitrogen) were used according to the manufacturer's instruction. After transfection for 48 hours the cells were passaged at 1:15 (volume/volume) and cultured in medium supplemented with geneticin (G418) at 1000 μg/mL for 8 weeks. The survival clones were selected, amplified with XAF1 expression detected by Western blot analysis, and maintained in medium containing 600 μg/mL G418 for the experiments.
Western Blot Analysis
Protein (30 μg) extracted from control and treated cells was resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to nitrocellular member. All target proteins were immunoblotted by the appropriate primary antibodies: anti-XAF1, anti-XIAP, anti-cleaved caspase-3 (Santa Cruz Biotechnology, Santa Cruz, Calif), anti-phospho-p44/p42 MAPK (Thr202/Tyr204), anti-p44/42 MAPK, anti-p-MEK1/2 and anti-MEK1/2 (Cell Signaling Technology, Danvers, Mass), respectively. Anti-β-actin antibody (Santa Cruz) probe was as an internal loading control. Antigen-antibody complexes were visualized by the ECL system (Amersham Biosciences, Little Chalfont, Buckinghamshire, UK).
Total RNA was extracted from the colorectal carcinoma cells using TRIzol reagent (Invitrogen). cDNA was synthesized using the ThermoScript reverse transcriptase (RT)-PCR System kit (Invitrogen). The PCR primer sequences were: XAF1 (forward) 5′-AAG TCC TCG CTG GAG TTT CA-3′; (reverse) 5′-TTT CAG CAG CTT GAC TTG GA-3′; XIAP (forward) 5′-CGG TGC TTT AGT TGT CAT GCA G-3′; (reverse) 5′-ACA AAA GCA CTG CAC TTG GTC A-3′; glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (forward) 5′-TGC CTC CTG CAC CAC CAA CT-3′; (reverse) 5′-CCC GTT CAG CTC AGG GAT GA-3′. XAF1 was amplified using 30 cycles and GAPDH using 28 cycles (94°C for 45 seconds, 57°C for 30 seconds, 72°C for 45 seconds) after normalization to the GAPDH internal control.
Cell Viability Assay
Cells were seeded at 1 × 104 cells/well in 96-well plates and incubated overnight within 200 μL of culture medium. Cells were then treated with different concentrations of U0126 for indicated time intervals. Five mg per mL MTT (Sigma Chemical Company, St. Louis, Mo) (50 μL, final concentration 1 mg/mL) was added and incubated for an additional 4 hours at 37°C. The supernatant fluid was removed and 200 μL of dimethylsulfoxide (DMSO) (Sigma) was added to each well. A micro enzyme-linked immunoadsorbent assay (ELISA) reader (Bio-Rad, Hercules, Calif) measured the absorbency at the wavelength of 570 nanometers (nm). The results are expressed as percentage loss of cell viability compared with control.
Cells were seeded in 12-well plates at a density of 5 ×104 cells/well for 24 hours and treated with 30 μM of U0126 for 24, 48, and 72 hours. Then cells were washed with ice-cold phosphate-buffered saline (PBS) and stained with Hoechst 33258. The effect on nuclear morphology was determined by fluoroscopic microscopy. Apoptotic nuclei were identified by the condensation of nuclear heterochromatin and fragmentation of the nucleus. Greater than 150 cells per field were counted and the percentage of apoptotic nuclei was determined.
All statistical analyses were performed with the EXCEL program (Microsoft Corporation, Redmond, Wash). The Student t-test was used to compare parameters between the different study groups. Results were expressed as mean ± standard deviation (SD). P-values <.05 were considered statistically significant.
Inhibition of ERK1/2 Up-Regulated XAF1 Expression in Colon Cancer Cells
The effect of U0126 was detected in all 4 colon cancer cell lines (DLD1, HCT1116, Lovo, and SW1116). After treatment with U0126 (30 μM) for 48 hours the expression of XAF1 in HCT1116 and Lovo cells was significantly up-regulated. However, no change was observed in DLD1 and SW1116 cells (Fig. 1A). Figures 1B and 1C show that U0126 stimulated XAF1 expression in dose- and time-dependent manners in Lovo cells.
To exclude the possibility that U0126 increased XAF1 expression in an ERK-independent pathway, we examined the effect of dominant-negative MEK-1 (DN-MEK1) plasmid on XAF1 expression. As shown in Figure 1D, DN-MEK1 had the similar effect with U0126 after transient transfection. Inversely, both U0126 and DN-MEK1 suppressed XIAP expression slightly. This finding was consistent with the previous report.15 U0126 or DN-MEK1 significantly inhibited the phosphorylation of both ERK1/2 and MEK1/2, but did not alter the expression of total ERK1/2 and MEK1/2 proteins (Fig. 1D).
Inhibition of ERK1/2 Up-Regulated XAF1 Expression Through Transcription Regulation
To determine whether ERK1/2 inhibition induced XAF1 expression at the transcriptional level, we first checked the effect of U0126 on the transcription activity of a core promoter of XAF1 gene identified in our previous publication.13 We transiently transfected cells with the XAF1 promoter pGL3-F291 followed by treatment with U0126. The transcriptional activities were measured by dual luciferase assay. We showed that U0126 treatment increased the promoter activities by 3.56-fold in HCT1116 cells and 3.09-fold in Lovo cells (Fig. 2A). Similarly, when we cotransfected cells with pGL3-F291 and MEK1 constructs we showed that DN-MEK1 caused 1.14-fold and 0.82-fold increase of promoter activity in HCT1116 and Lovo cells, respectively. Consistently, MEK1-active caused 63% and 54% reductions of XAF1 promoter activities in these 2 cell lines, respectively (Fig. 2B). Taken together, these results suggested that inhibition of ERK1/2 pathway might regulate the transcriptional activity of XAF1 gene.
Second, we investigated if the induced XAF1 expression required de novo protein synthesis. Emetine is known to be able to block translation and its effect on protein synthesis is irreversible. In this study we detected XAF1 expression with treatment with U0126 alone or in combination with emetine. We showed that XAF1 expression stimulated by U0126 was completely abrogated by pretreatment with emetine (Fig. 2C).
These findings indicated that ERK1/2 inhibition induced XAF1 expression through the induction of novel protein synthesis and the subsequent transcription regulation.
EGF-Stimulated Phosphorylation of ERK1/2 With No Alteration in XAF1 Expression
It was reported that epidermal growth factor (EGF) strongly induced phosphorylation of ERK1/2 in HT29 colon cancer cells.16 To further solidify our hypothesis, we investigated whether ERK1/2 activation has effects on XAF1 expression or not. Lovo cells were pretreated with U0126 for 60 minutes and then treated with rhEGF (20 ng/mL). RT-PCR was performed after 24hours and Western blot analysis was performed after 48 hours. As shown in Figure 3, EGF alone had no effect on expressions of either XAF1 and XIAP at both the mRNA and protein levels (Fig. 3A,B). EGF activated ERK1/2 characterized by the increasing level of phosphorylated ERK1/2 protein (Fig. 3B). U0126 inhibited EGF-induced phosphorylation of ERK1/2 and simulated XAF1 expression as well, even in the presence of EGF (Fig. 3A and 3B).
Overexpression of XAF1 Rendered Lovo Cells More Susceptible to U0126-Induced Apoptosis
It has been well established that inhibition of ERK activity by U0126 will eventually induce apoptosis in human colon cancer cells.8 To check whether XAF1 was involved in this process or not, we first assessed the effect of U0126 on cell proliferation. Lovo cells stably expressing XAF1 (Lovo/XAF1) or vector control (Lovo/vector) were exposed to different doses of U0126 for different times, followed by evaluation of cell viabilities. We demonstrated that U0126 resulted in the loss of cell viability in a time- and dose-dependent manner, whereas the overexpression of XAF1 magnified this effect of U0126 (Fig. 4A). Thirty micromolar of U0126 inhibited cell proliferation of Lovo/XAF1 by 61.28% (24 hours), 26.88% (48 hours), and 21.43% (72 hours) compared with 74.61% (24 hours), 55.94% (48 hours), and 40.23% (72 hours) of that of Lovo/vector, respectively. A significant difference was found between Lovo/XAF1 and Lovo/vector (Fig. 4A) (P < .05). Conversely, no significant difference was found between 2 cell lines when a high concentration of U0126 (60 μM) was used (P = .265), indicating that even the Lovo/vector cells were susceptible enough to a high concentration of U0126-induced cell death.
Furthermore, we detected the effect of U0126 on the activation of a featured apoptotic protein, caspase 3. As shown in Figure 4B, U0126 could induce cleavage of caspase 3 and its effect was amplified by overexpression with XAF1. Also, we showed that U0126 induced typical morphological changes of apoptosis in Lovo cells characterized by a condensed nucleus, exhibited after staining with a DNA-binding agent, Hoechst 33258 (Fig. 4C). Similarly, overexpression with XAF1 increased U0126-induced apoptosis. The percentages of apoptotic cells were 27.24% (24 hours, P = .002), 64.62% (48 hours, P = .003), and 70.26% (72 hours, P = .016) in Lovo/XAF1 compared with 12.66% (24 hours), 31.58% (48 hours), and 48.25% (72 hours) of that of Lovo/vector.
These data suggested that overexpression of XAF1 rendered Lovo cells more susceptible to U0126-induced apoptosis.
Down-regulation of XAF1 Expression With XAF1 Antisense Rescued Lovo Cells From Apoptosis Induced by U0126
Finally, we examined if the inhibition of XAF1 could reverse U0126-induced apoptosis. We established a stable transfectant of Lovo cells expressing antisense (Lovo/XAF1-as) of XAF1 and identified by Western blot analysis (Fig. 5A). We showed that U0126 could not induce XAF1 expression in XAF1 antisense transfectant (Fig. 5A); in addition, its effect on suppressing cell proliferation and inducing apoptosis were also abrogated to some extent by XAF1 antisense, as expected (Fig. 5C).
In this study, we determined that a novel tumor suppressor, XAF1, is negatively correlated with the ERK1/2 signal pathway in colon cancer cells. Inhibition of ERK1/2 stimulated XAF1 expression through indirect transcription regulation. Furthermore, XAF1 was an effector in ERK inhibition-induced cell apoptosis. Our finding implicated XAF1 as a novel target for a combined chemo- and geno-therapy of colon cancer.
Current studies demonstrate that XAF1 is 1 of the intracellular XIAP-interacting proteins that directly antagonizes the anti-caspase activity of XIAP. The tissue-expressing profile reveals that XAF1 is expressed substantially by most of the normal but not malignant tissues. These findings implied that it was a novel tumor suppressor. Although our previous data indicated that it was conversely correlated with the stress response of cancer cells,12 the direct role of XAF1 with MAPKs has not been reported.
ERK1/2 activation is essential to maintain the growth and malignant phenotype of cancer cells, including colon cancer.17, 18 It has been reported that inhibition of ERK1/2 phosphorylation decreased soft agar growth and inhibited the transformed phenotype of colon 26 cells in vitro and suppressed the growth of mouse and human colon tumor xenografts.7 In this study, our finding that XAF1 negatively correlated with ERK1/2 activity provided further evidence supporting the notion of XAF1 as a tumor suppressor.
ERK1/2 is the downstream kinase of MEK1 and its activation requires dominant active MEK1. In our study, we found that U0126 induced XAF1 expression significantly. However, to exclude the putative side effect of this chemical, we specifically suppressed ERK1/2 by transfection with a dominant-negative MEK1 construct. This dominant site has been reported to be crucial in maintenance of MEK1 activity.19 Its effect on XAF1 expression further substantiates the direct correlation between XAF1 and ERK1/2.
In our previous study, we identified the transcription starting site of the XAF1 gene to located at −26 nt by 5′-RACE assay12 and further identified a putative promoter segment13 that contained −107 base pairs (bp) (upstream ATG codon) 5′-flanking region of the XAF1 gene. To determine whether inhibition of ERK1/2 induced XAF1 expression through transcription regulation or posttranslation modification, we examined the effect of U0126 on the transcription activity of this construct (pGL3-F291). The dual luciferase assay and the use of the protein synthesis inhibitor emetine determined that U0126 might induce the synthesis of some novel transcription factors that up-regulated XAF1 transcription. However, the direct mechanism remains to be clarified.
In our study we found that not all colon cancer cell lines responded to U0126 in the induction of XAF1 expression. One candidate interpretation was that the p53 status of these cell lines was different. The 2 cell lines, HCT1116 and Lovo, which responded well to U0126 treatment, were both expressing wildtype p53 protein, whereas both DLD1 and SW1116 expressed mutant p53. Our result was consistent with other reports that U0126 induced cell death only in p53 wild-type cells, wherever the p53-deficient cells were not affected by blocking the MEK/MAPK pathway.20 In addition, studies using isogenic colon cancer cell lines and RNA interference assay revealed that loss of p53 significantly reduced MAPK phosphorylation and renders cells resistant to U0126 treatment.20 Lee et al.21 reported that the apoptosis-sensitizing and growth-suppression function of XAF1 was markedly impeded by blockade of p53 function. However, the direct evidence concerning the interaction between XAF1 and p53 remains to be collected.
Inhibition of ERK1/2 has been implicated in inducing cancer cell differentiation.22, 23 In agreement, we and other investigators have identified XAF1 as an interferon (IFN)-stimulated gene11 and a mediator of ATRA-induced differentiation of colon cancer.13 Although we did not evaluate the direct role of XAF1 in U0126-induced cell differentiation, we did find that XAF1 mediated U0126-induced cell growth suppression and apoptosis. It is well known that differentiation is accompanied by a loss of the capacity to proliferate.
In conclusion, the results of the current study demonstrate that inhibition of the ERK1/2 pathway induced the expression of XAF1 through transcription regulation in colon cancer cells. XAF1 is essential in U0126-induced apoptosis. Our findings reveal a novel element in the signal pathway of ERK1/2 in regulating cancer cell behaviors and provide a potential target for the combined chemo- and geno-therapy of colon cancer.
Supported by grants from the Key Medical Subjects of Shanghai (No. 05 III 005) and Gastrointestinal Carcinoma Research Fund, Shanghai Jiao Tong University School of Medicine, China, and Kwok Sau Po Gastroenterology Research Fund and Gordon YH Chiu Stomach Cancer Research Fund, University of Hong Kong, Hong Kong.