Oncogene (2007) 26, 5927–5938
& 2007 Nature Publishing Group All rights reserved 0950-9232/07 $30.00 www.nature.com/onc
ORIGINAL ARTICLE
Frequent silencing of RUNX3 in esophageal squamous cell carcinomas is associated with radioresistance and poor prognosis C Sakakura1, K Miyagawa1, K-I Fukuda1, S Nakashima1, T Yoshikawa1, S Kin1, Y Nakase1, H Ida2, S Yazumi2, H Yamagishi3, T Okanoue4, T Chiba2, K Ito5, A Hagiwara1 and Y Ito5 1 Department of Surgery and Regenerative Medicine, Division of Surgery and Physiology of the Digestive System, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kawaramachi, Kamigyo-ku, Kyoto, Japan; 2Division of Gastroenterology and Hepatology, Graduate School of Internal Medicine, Kyoto University, Kawahara-cho, Sakyo-ku, Kyoto, Japan; 3Department of Surgery and Regenerative Medicine, Division of Surgery and Oncology of Digestive System, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kamigyo-ku, Kawaramachi-dori, Kyoto, Japan; 4Division of Gastroenterology, Graduate School of Internal Medicine, Kyoto Prefectural University of Medicine, Kawaramachi, Kamigyo-ku, Kyoto, Japan and 5Institute of Molecular and Cell Biology, Proteos, Singapore, Singapore
Radiotherapy is an effective treatment for some esophageal cancers, but the molecular mechanisms of radiosensitivity remain unknown. RUNX3, a novel tumor suppressor of gastric cancer, functions in transforming growth factor (TGF)-b-dependent apoptosis. We obtained paired samples from 62 patients with advanced esophageal cancers diagnosed initially as T3 or T4 with image diagnosis; one sample was obtained from a biopsy before presurgical radiotherapy, and the other was resected in surgical specimens after radiotherapy. RUNX3 was repressed in 67.7% cases of the pretreatment biopsy samples and 96.7% cases of the irradiated, resected samples. The nuclear expression of RUNX3 was associated with radiosensitivity and a better prognosis than cytoplasmic or no RUNX3 expression (Po0.003); cytoplasmic RUNX3 expression was strictly associated with radioresistance. RUNX3 was downregulated and its promoter was hypermethylated in all radioresistant esophageal cancer cell lines examined. Stable transfection of esophageal cancer cells with RUNX3 slightly inhibited cell proliferation in vitro, enhanced the antiproliferative and apoptotic effects of TGF-b and increased radiosensitivity in conjunction with Bim induction. In contrast, transfection of RUNX3-expressing cells with a RUNX3 antisense construct or a Bim-specific small interfering RNA induced radioresistance. Treatment with 5-aza-20 deoxycytidine restored RUNX3 expression, increased radiosensitivity and induced Bim in both control and radioresistant cells. These results suggest that RUNX3 silencing promotes radioresistance in esophageal cancers. Examination of RUNX3 expression in pretreatment specimens may predict radiosensitivity, and induction of RUNX3 expression may increase tumor radiosensitivity. Correspondence: Dr C Sakakura, Department of Surgery and Regenerative Medicine, Division of Surgery and Oncology of Digestive System, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kamigyo-ku, Kawaramachi-dori, Kyoto 6028566, Japan. E-mail:
[email protected] Received 29 December 2006; revised 6 February 2007; accepted 8 February 2007; published online 26 March 2007
Oncogene (2007) 26, 5927–5938; doi:10.1038/sj.onc.1210403; published online 26 March 2007 Keywords: radioresistance; esophageal cancer; RUNX3; methylation
Introduction Although radiotherapy for esophageal squamous cell cancers is effective in selected patients, some patients show no response or encounter adverse effects such as immunosuppression. Overall, esophageal squamous cell carcinomas show an intermediate degree of radiosensitivity, but individual tumors exhibit widely different sensitivities, even within the same histologic type. Many factors influence tumor resistance in clinical radiotherapy, including tumor size, hypoxia and intrinsic radiosensitivity (Eifel et al., 1994). Because of heterogeneity of these factors among patients and within tumors, it is difficult to separate out the contributions of individual factors. A set of genes possibly related to radiosensitivity has been identified by retrospectively comparing the expression profiles of radiosensitive and radioresistant tumors in clinical samples obtained before treatment (Hanna et al., 2001; Kitahara et al., 2002; Fukuda et al., 2004) however, whether the radioresistance observed in these studies was related to patient factors, intrinsic tumor factors or characteristics acquired after the initiation of therapy has been difficult to determine. The transforming growth factor (TGF)-b signal transduction system is impaired in many types of cancer, including esophageal cancers (Hanahan and Weinberg, 2000; Massague et al., 2000), resulting in the loss of the growth-inhibiting response to TGF-b. This loss may occur through disruption of TGF-b type II receptors or Smad4 (Fukuchi et al., 2002; Natsugoe et al., 2002). Molecules involved in the TGF-b signaling
RUNX3 silencing in radioresistant esophageal cancers C Sakakura et al
5928
pathway, such as TGF-b1, mediate cellular responses to DNA damage, thus affecting radiosensitivity (Vodovotz et al., 2000; Ewan et al., 2002; Kim et al., 2003, 2005). Because RUNX3 interacts with Smads and is a downstream target of TGF-b signaling (Hanai et al., 1999), perturbation of RUNX3 function might influence the response phenotype of these cancers. In a previous study, we found that glandular epithelial cells of Runx3null mice are less sensitive to the growth-inhibitory effects of TGF-b and completely lack the ability to undergo apoptosis in response to TGF-b stimulation (Li et al., 2002). In conjunction with the importance of the TGF-b signaling pathway in irradiation-induced cell death (Alcock et al., 2002) and the difficulty of inducing TGF-b-dependent apoptosis in the gastric epithelium of Runx3-null mice (Li et al., 2002), these results led us to suspect that RUNX3 may play an important role in irradiation-induced apoptosis in esophageal squamous cell carcinomas. In the present study, we examined the relationship between RUNX3 and radiosensitivity in esophageal cancer cells and found that inactivation of RUNX3 enhances the radioresistance of cells.
Results Immunohistochemical assessment of RUNX3 in specimens obtained before radiotherapy We obtained paired specimens from 62 esophageal cancer patients before and after radiotherapy (diagnostic biopsies and surgical sections, respectively). We examined radiosensitivity in the samples surgically resected after radiotherapy: 21 specimens were grade 2 (partial response, PR) or 3 (complete response; Figure 1a), and 41 specimens were grades 0 or 1 (no change after radiotherapy; Figure 1b and c). We then examined RUNX3 expression in the preradiotherapy samples. RUNX3 protein was strongly expressed in the nuclei of basal layer of the normal esophageal mucosa (Figure 1a2) and in the nuclei of esophageal cancer cells whose paired sample was scored as having a grade-2 or -3 response to radiotherapy (19/20 cases; Figure 1a-3). In contrast, tissue pairs with cytoplasmic RUNX3 expression (Figure 1b-2; 4/4 cases) or no RUNX3 expression (Figure 1c) in the preradiotherapy sample showed no response to radiotherapy in the postradiotherapy sample. On the basis of these results, we divided the
Figure 1 Immunohistochemical analysis of RUNX3 and SMAD4 in biopsy samples of esophageal cancers obtained before radiotherapy. (a) Specimen from a patient whose postradiotherapy sample showed a complete response (CR, grade 3). (a-1) Whole mount view ( 40 magnification); boxes indicate the positions of the enlargements shown below. (a-2): RUNX3 ( 100). a-3: RUNX3 and SMAD4 ( 400). (b) Specimen from a patient whose postradiotherapy sample showed no change (NC, grade 0). (b-1) Whole mount ( 40); boxes indicate the positions of the enlargements shown below. (b-2) RUNX3 ( 100). b-3: RUNX3 and SMAD4 ( 400). (c) Specimen from a patient whose postradiotherapy sample showed NC. (c-1) Whole mount ( 40); boxes indicate the positions of the enlargements shown below. (c-2) RUNX3 ( 100). c-3: SMAD4 ( 400). Scale bars in a-1, b-1 and c-1 ¼ 2 mm; all others ¼ 200 mm. Oncogene
RUNX3 silencing in radioresistant esophageal cancers C Sakakura et al
5929
62 preradiotherapy samples of squamous cell carcinoma into two groups, those with nuclear RUNX3 expression (20 cases; RUNX3-positive) and those with cytoplasmic (four cases) or no (38 cases) RUNX3 expression (RUNX3-negative). The former showed significantly higher radiosensitivity than the latter. These results suggest that nuclear RUNX3 expression promotes radiosensitivity in esophageal cancers. Downregulation and promoter methylation of RUNX3 in primary esophageal cancers and their relationship with radioresistance and prognosis A significantly larger fraction of the cells in the biopsy specimens taken before irradiation was positive for RUNX3 expression in samples from patients whose
surgical samples were of grade 2 or 3 (81.6%, average rate of RUNX3-positive cells) than in those whose surgical samples were of grade 0 or 1 (average 20.8%; P ¼ 0.0002; Figure 2a. We obtained a similar result when we examined RUNX3 expression directly in the surgically resected specimens: RUNX3 expression in grade-2 or -3 samples was higher (51.0%, average rate of RUNX3-positive cells) than in the grade-0 and -1 samples (average 18.4%, P ¼ 0.005; Figure 2a. When we set the threshold for RUNX3 positivity at 70% of cells, RUNX3 was downregulated from above the threshold to below it in 67.8% (42 of 62 cases) of the pretreatment biopsy specimens (Figure 2a and in 96.7% (60 of 62 cases) of the specimens resected after radiotherapy (Figure 2a.
Figure 2 RUNX3 silencing and associated promoter hypermethylation in clinical samples of esophageal cancers: correlation with patient survival. (a-1) Immunohistochemical analysis of RUNX3 expression in biopsy samples obtained before radiotherapy. Grades 0, 1: cases with no change after radiotherapy; Grades 2, 3: cases sensitive to radiotherapy (P ¼ 0.0002). Dotted line, threshold for RUNX3 positivity (70%). IR, irradiation. (a-2) Immunohistochemical analysis of RUNX3 expression in specimens surgically resected after radiotherapy. Grades 0, 1: cases with no change following radiotherapy; Grades 2, 3: cases sensitive to radiotherapy (P ¼ 0.005). Dotted line, threshold for RUNX3 positivity (70%). (b) Survival rates of esophageal cancer patients with or without RUNX3 expression. Patients with RUNX3 expression (20 cases) had significantly higher survival rates than patients lacking RUNX3 expression (42 cases, P ¼ 0.0003). (c) Analysis of esophageal cancer biopsy specimens before radiotherapy using MSP. M, methylated sequencespecific PCR; U, unmethylated sequence-specific PCR; SM, size marker; N, normal mucosa; T, tumor tissue; DW, distilled water. MKN28, positive control for RUNX3 methylation, MKN1, negative control for RUNX3 methylation. Oncogene
RUNX3 silencing in radioresistant esophageal cancers C Sakakura et al
5930
Patients with nuclear RUNX3 expression (20 cases) had a higher survival rate than patients lacking RUNX3 expression (42 cases) (Figure 2b; P ¼ 0.0003). As noted above, grade-2 or -3 specimens obtained after irradiation have a lower proportion of RUNX3-positive cells (51.0%, average rate of RUNX3-positive cells) than their paired specimens obtained before irradiation (81.6%, average rate of RUNX3-positive cells). It is possible that many of the RUNX3-positive cells died by the time surgically resected samples were tested. Another possibility, however, is that the radiation induced DNA methylation, an epigenetic modification that downregulates gene expression. To test whether DNA methylation increased after radiation, we performed MSP on 51 resected specimens of esophageal cancer, as well as their surrounding non-neoplastic mucosa; representative MSP data are shown in Figure 2c. RUNX3 was methylated in 70.4% (36/51 cases) of esophageal cancers examined (Figure 2c, cases no. 2, 3, 4, 5, 6, 7, 8 and 10) and in the surrounding mucosa in 27.4% (14/51 cases) of the samples (cases no. 3, 7 and 10). A negative relationship between RUNX3 methylation and RUNX3 expression was evident. RUNX3 expression in esophageal cancers before radiotherapy and clinicopathological factors in 62 patients RUNX3 expression estimated in biopsy samples before radiotherapy and clinicopathological factors assessed in resected specimens after radiotherapy were examined. The prevalence of lymph node involvement, tumor depth and histological effect of radiotherapy was significantly greater in RUNX3-negative cases than in RUNX3-positive cases (Table 1). SMAD4 was expressed in 42% (26/62) of the cases. Five-year survival rate was 24 and 17% in SMAD4positive and SMAD4-negative cases, respectively. Our data suggests that patients with SMAD4 expression have a better prognosis than those without SMAD4 expression, but the difference was not significant. Furthermore, there was no statistically significant relationship between SMAD4 expression and radiosensitivity (data not shown). Downregulation and promoter methylation status of RUNX3 in radiation-resistant sublines obtained from esophageal cancer cell lines We further examined the correlation between the presence of RUNX3 and the radiosensitivity of esophageal cancer cells using a cell culture system. For this analysis, we derived radioresistant cell lines from esophageal cancer cell lines (see Materials and methods), using an ‘R’ to designate the radioresistant derivative; for example, the radioresistant derivative of the TE-2 cell line was designated TE-2R. The radioresistance of esophageal cancer cells was estimated using a clonogenic assay to assess survival after single-dose irradiation, as described previously (Fukuda et al., 2004). Survival curves for two radioresistant sublines, KYSE170R and TE-2R, confirmed that they are significantly more radioresistant than their parental cell lines, KYSE170 and TE-2 (Figure 3a). Oncogene
Table 1 RUNX3 expression in relation to clinicopathologic findings of 62 patients with esophageal squamous cell carcinoma before radiotherapy Variables
Age (years) Gender Male Female Histologic type Well Moderate Poor Tumor locationa Upper Middle Lower pM pM0 pM1 pN pN0
RUNX3 positive RUNX3 negative or in the nuclei cytoplasmic expres(n ¼ 20; 32.3%) sion (n ¼ 42; 67.7%)
P
66.3710.5
67.177.0
NS
12 8
22 20
NS
4 8 8
14 16 12
NS
4 10 6
6 21 15
NS
16 4
34 8
NS
4
19
o0.01 (0.0079)
pN1 pT pT1
16
23
2
0
pT2 pT3 pT4 Stage I IIA IIB III IV Lymphatic invasion Negative Positive Venous invasion Negative Positive Histological effect Grade 0
8 10 0
8 22 12
2 2 2 10 4
0 0 8 26 8
NS
9 11
22 20
NS
8 12
21 21
NS
0
30
o0.01 (0.0001)
1 17 2
10 2 0
Grade 1 Grade 2 Grade 3
o0.01 (0.0011)
Abbreviation: NS, not significant. aPosition in the esophagus. Classifications of the specimens were determined according to the International Union against Cancer tumor-node-metastasis (TNM) classification system. pT1, tumor invades lamina propria or submucosa; pT2, tumor invades muscularis propria; pT3, tumor invades adventitia; pT4, tumor invades surrounding tissue. pN0, no regional lymph node metastasis; pN1, regional lymph node metastasis. pM0, without distant metastasis, pM1; with distant metastasis.
b
Relative levels of RUNX3 mRNA were determined as ratios of RUNX3 to GAPDH mRNA (Figure 3b). RUNX3 was significantly downregulated in the radioresistant cell lines TE-2R, TE-9R, TE-13R and KYSE170R compared with their respective parental cell lines (Po0.01–0.05). To test whether the downregulation of RUNX3 is correlated with the methylation status of RUNX3, we performed an MSP analysis for these cell lines (Figure 3c). Methylation of RUNX3 was evident in all radioresistant cell lines (TE-2R, TE-9R, TE-13R and KYSE170R). A negative relationship was observed
RUNX3 silencing in radioresistant esophageal cancers C Sakakura et al
5931
and KYSE170R cells relative to their respective parental (control) cell lines however, there was no significant difference in RUNX3 expression between TR-5 and TR-5N or KYSE160 and KYSE160N (Figure 3d). No significant difference in SMAD4 expression was detected between irradiation-resistant esophageal cancer cells and parental cells (data not shown).
Figure 3 RUNX3 silencing and promoter hypermethylation in radioresistant esophageal cancer cell lines established after fractionated irradiation. (a) Clonogenic assay for assessing survival after single-dose irradiation as an estimation of radiosensitivity. Survival curves for radioresistant KYSE170 and TE-2 cell lines show that these cells exhibit significantly greater radioresistance than their parental cell lines (*Po0.01). SF, survival fraction. (b) Relative RUNX3 mRNA expression in parental and radioresistant esophageal cancer cell lines, as measured by real-time RT–PCR. Expression of RUNX3 in radioresistant esophageal cancer cells was significantly lower than in control cells (*Po0.05, **Po0.01, NS, not significant). R, radioresistant cells; N, cells without acquired radioresistance. (c) Analysis of parental and radioresistant esophageal cancer cell lines using MSP. M, methylated sequencespecific PCR; U, unmethylated sequence-specific PCR; SM, size marker; DW, distilled water. R, radioresistant cells; N, cells without acquired radioresistance. Methylation-specific bands were observed in radioresistant sublines. MKN28, positive control for RUNX3 methylation, MKN1, negative control for RUNX3 methylation. (d) Western blot analysis of RUNX3 expression in parental, nonradioresistant and radioresistant TE-2, TE-13 and KYSE170 cells. R, radioresistant cells; N, cells without acquired radioresistance.
between RUNX3 methylation status and expression. In cell lines that did not acquire radioresistance during the procedure to isolate resistant cells (TE-5N and KYSE160N), neither RUNX3 expression nor the methylation status changed significantly after the procedure (Figure 3b–d). The gastric cancer cell lines MKN1 and MKN28 were used, respectively, as positive and negative control for RUNX3 methylation (Figure 3c). In a Western blot analysis, RUNX3 protein expression decreased in radioresistant TE-2R, TE-9R, TE-13R
Radiosensitivity of TE-2 and KYSE170 esophageal cancer cells stably transfected with RUNX3 or antisense RUNX3 To determine whether a lack of RUNX3 function leads to radioresistance in esophageal cancers, TE-2 and KYSE170 cells were transfected with RUNX3 or antisense RUNX3 DNA. We generated four independent TE-2 and KYSE170 clones stably expressing RUNX3, TE-2 (R3-1), TE-2 (R3-2), KYSE170 (R3-1) and KYSE170 (R3-2), and examined RUNX3 protein expression in these cells by western blot. Total protein levels for the analysis were normalized to GAPDH expression. TE-2 (R3-1), TE-2 (R3-2), KYSE170 (R3-1) and KYSE170 (R3-2) showed stronger expression of RUNX3 than did control cells transfected with the neomycin resistance gene alone (Figure 4a). In an RT– PCR analysis, RUNX3 was undetectable or was only weakly expressed in TE-2 and KYSE170 cells stably expressing antisense RUNX3 (Figure 4b). TE-2 (R3-1), TE-2 (R3-2), KYSE170 (R3-1) and KYSE170 (R3-2) cells grew slightly more slowly than TE-2 and KYSE170 cells. Expression of RUNX3 slightly but significantly prolonged the doubling times from 27 to 34 h. In contrast, transfection of antisense RUNX3 to create TE-2 (R3-AS-1), TE-2 (R3-AS-2), KYSE170 (R3AS-1) and KYSE170 (R3-AS-2) cells slightly enhanced the cell proliferation relative to the parental TE-2 and KYSE170 cells (doubling times ¼ 21–27 h). Cells in S phase tend to be more radiosensitive than cells in G1, whereas cells in G2/M are highly radiosensitive (Sinclair and Morton, 1966). Therefore, we examined cell cycle distributions by flow cytometry. In exponentially growing cell populations, all cell lines had very similar profiles (Table 2). Thus, effects on cell-cycle distribution cannot explain the altered radiosensitivity seen in resistant cells. We next measured the radiosensitivity of each stable transfectant with a clonogenic assay. Survival parameters for all cell lines are shown in Table 3. A significant increase in radiosensitivity was seen in TE-2 (R3-1), TE-2 (R3-2), KYSE170 (R3-1) and KYSE170 (R3-2) cells compared with TE-2 and KYSE170 cells (Figure 4c, Table 3). In contrast, transfection with antisense RUNX3 reduced the radiosensitivity in these cell lines. Cell growth, apoptosis and TGF-b-induced transcriptional activity in TE-2 and KYSE170 cells stably transfected with RUNX3 and antisense RUNX3 TGF-b did not inhibit cell proliferation in the parental TE-2 and KYSE170 cells transfected with a control (empty) vector. In contrast, treatment with TGF-b Oncogene
RUNX3 silencing in radioresistant esophageal cancers C Sakakura et al
5932
slightly but not significantly decreased the proliferation of TE-2 and KYSE170 cells stably transfected with RUNX3 compared with mock-transfected cells (Figure 5a). TGF-b treatment also strongly induced the apoptosis in RUNX3 transfectants (Figure 5b), but not in TE-2 and KYSE170 cells or stable transfectants of TE-2 and KYSE170 expressing antisense RUNX3 (TE-2 (R3-AS-1), TE-2 (R3-AS-2), KYSE170 (R3-AS-1),
KYSE170 (R3-AS-2)) (Figure 5b, Po0.05). RUNX3 overexpression alone, in the absence of TGF-b treatment, did not induce apoptosis. TE-2 and KYSE170 express endogenous RUNX3, but the level of expression is relatively low, whereas these cells stably expressing exogenous RUNX3 express RUNX3 at much higher levels. The difference in the expression level of RUNX3 must have caused the differential sensitivity of these cells to TGF-b in inducing apoptosis. We used the luciferase gene reporter assay to examine RUNX site-dependent transcriptional activity in TE-2 and KYSE170 cells stably transfected with RUNX3. Basal luciferase activity was modestly upregulated by RUNX3 in each cell line in the absence of stimulation (Figure 5c), but treatment with 5 ng/ml TGF-b for 16 h dramatically enhanced the transactivation activity in stable RUNX3 transfectants (Po0.01). Treatment with 5-aza-20 -deoxycytidine enhances radiosensitivity and induces RUNX3 and Bim Given that RUNX3 is frequently silenced by promoter methylation in esophageal cancers (Figures 2 and 3), we next examined whether treatment with the methyltransferase inhibitor 5-aza-20 -deoxycytidine would reactivate RUNX3. Indeed, treatment with 100 nM 5-aza-20 -deoxycytidine resulted in the reactivation of RUNX3 in TE-2, TE-9, TE-13 and KYSE170 cells and their radioresistant sublines (TE-2R, TE-9R, TE-13R and KYSE170R), at levels ranging from 9.7- to 81.0-fold (Figure 6a). The fold-increases in RUNX3 expression were higher in the
Figure 4 Radiosensitivity of TE-2 and KYSE170 esophageal cancer cells stably transfected with RUNX3 or antisense RUNX3. (a) Western blot analysis of RUNX3 expression in stable RUNX3 transfectants relative to parental TE-2 and KYSE170 cells. (b) RT– PCR analysis of RUNX3 expression in antisense RUNX3 transfectants relative to parental TE-2 and KYSE170 cells. (c) Clonogenic assay of survival after single-dose irradiation as an estimation of radiosensitivity in stable RUNX3 and antisenseRUNX3 transfectants. Survival curves of KYSE170 and TE-2 cell lines stably transfected with RUNX3 show that these cells are significantly more radiosensitive than their parental (control) cells (*Po0.01). In contrast, survival curves for KYSE170 and TE-2 cell lines stably transfected with antisense RUNX3 show that they are significantly more radioresistant than their parental (control) cells (*Po0.01). SF, survival fraction.
Table 3
Table 2 Cell-cycle distribution in parent cells and RUNX3 stable transfectants Cell lines TE2-control TE2-R1a TE2-R2a KYSE170-control KYSE170-R1a KYSE170-R2a
G0/G1 (%)
S (%)
G2/M (%)
50 52 54 50 48 55
28 26 23 28 30 23
22 22 23 22 22 22
a
P, not significant.
Clonological survival parameters fitting the data to multitarget models
Cell lines
D0
(95% CI)
n
(95% CI)
TE2-control TE2-R1a TE2-R2a TE2-AS-1 TE2-AS-2a KYSE170-control KYSE170-R1a KYSE170-R2a KYSE170-AS1 KYSE170-AS2a
0.75 0.52 0.60 0.80 0.82 1.31 1.12 1.05 1.37 1.41
0.68–0.84 0.45–0.71 0.55–0.72 0.71–0.94 0.72–0.96 1.17–1.49 1.09–1.27 0.95–1.12 1.27–1.50 1.29–1.55
2.63 2.85 2.94 3.63 3.73 2.46 2.13 2.21 3.16 3.32
1.49–4.63 1.27–4.43 1.18–4.37 2.69–5.63 3.69–4.63 1.55–3.90 1.24–3.58 1.33–3.69 2.55–3.80 2.65–4.20
a Significant difference (Po0.05). D0: mean lethal dose, n: extrapolation number, CI: confidence interval. D0 was a parameter to indicate the amount of irradiation required to reduce the survival fraction to approximately 0.37 from the survival curve. It is considered in the multitarget models that D0 is the dose of irradiation to reduce survival fraction up to 37% (1/e 100%). A regression analysis of clonogenic assay allowed the mean lethal dose (D0) and 95% confidence interval (95% CI) to be calculated.
Oncogene
RUNX3 silencing in radioresistant esophageal cancers C Sakakura et al
5933
Figure 5 TGF-b-induced growth inhibition and induction of apoptosis in TE-2 and KYSE170 cells expressing RUNX3. (a) TE-2 and KYSE170 cells stably transfected with RUNX3 showed a slightly but not significantly decreased proliferation rate compared with mock-transfected cells upon treatment with TGF-b. (b) TGF-b induces apoptosis in RUNX3 stable transfectants (*Po0.05), but not in mock-transfected or antisense RUNX3 stable transfectants. (c) Luciferase gene reporter assay. RUNX binding site-dependent transcriptional activation is observed in RUNX3 stable transfectants after TGF-b treatment (5 ng/ml) (*Po0.05). RLU, relative luciferase units (*Po0.01).
radioresistant clones than in their parental cells, most likely because of their initial methylation status (Figure 6a). Reactivation of RUNX3 expression by 5-aza-20 -deoxycytidine resulted in significantly increased radiosensitivity (Po0.01) and an associated increase in Bim expression (Figure 6a–c). We used MSP analysis and nucleotide sequencing to confirm that the methylation status of RUNX3 was indeed changed by 5-aza-20 -deoxycytidine treatment (data not shown). The levels of RUNX3 protein in each cell line were upregulated after exposure to 5-aza-20 deoxycytidine (Figure 6b), and cell growth was inhibited
(data not shown), indicating that this treatment has a functional effect on esophageal cancer cells. Exogenous expression of RUNX3 increases Bim expression, and Bim knockdown inhibits radiationinduced cell death Exogenous expression of RUNX3 induced Bim expression (Figure 6b). Transfection of Bim siRNA, but not a control siRNA, into cells expressing exogenous RUNX3 dramatically suppressed Bim protein expression (Figure 6b) and significantly inhibited radiationinduced cell death (Po0.01) (Figure 6c). These results Oncogene
RUNX3 silencing in radioresistant esophageal cancers C Sakakura et al
5934
Figure 6 Expression of RUNX3 and Bim, and restoration of radiosensitivity by the methyltransferase inhibitor 5-aza-20 -deoxycytidine. (a) 5-Aza-20 -deoxycytidine significantly induces RUNX3 and Bim expression in the esophageal cancer cell lines TE-2 and KYSE170 and in radioresistant sublines compared with untreated cells, as determined with quantitative RT–PCR (*Po0.01). 5-AzadCyt, 5-aza-20 -deoxycytidine. (b) Transfection of a Bim-specific siRNA reduces Bim expression in the esophageal squamous cell carcinoma cell lines TE-2, KYSE170 and TE-2 RUNX3-1. 5-AzadCyt and 5-aza-20 -deoxycytidine. (c) Clonogenic survival curves of parental esophageal cancer cells and cells treated with 5-aza-20 deoxycytidine or Bim-specific siRNA. Curves were obtained by fitting the data to a multitarget model.
demonstrate that RUNX3 enhances radiation-induced cell death through the induction of Bim expression.
Discussion Cellular radiosensitivity involves multiple interacting factors, but several researchers have attributed the biological effectiveness of irradiation to the induction of apoptosis (Dewey et al., 1995; Meyn et al., 1996). Tumor cells are heterogeneous with respect to their sensitivity to therapeutic agents, and the radioresistant phenotype is often correlated with alterations in Oncogene
cell-cycle checkpoints and slowed growth as well as decreased apoptosis (Coleman and Stevenson, 1996; Maity et al., 1997). Radioresistance has been attributed to the expression of several genes, including p53 (Biard et al., 1994), ras (Sklar, 1988), raf-1 (Kasid et al., 1987), bcl-2 (Kitada et al., 1996), p16 (Matsumura et al., 1997), and p21 (Hsiao et al., 1997), but the molecular mechanisms involved are still unclear. We showed here that radiation-resistant sublines of esophageal carcinoma-derived cell lines express lower levels of RUNX3 than their parental, radiation-sensitive cells, and this lower level of expression correlates well with the promoter hypermethylation of RUNX3. Using biopsy specimens obtained before radiotherapy and samples surgically resected from the same patients after radiotherapy, we showed that esophageal cancer cells with a good response to radiation therapy expressed significantly higher levels of RUNX3 than those resistant to radiation therapy. Furthermore, esophageal cancer patients whose cells expressed RUNX3 had significantly higher rates of survival than those not expressing RUNX3. Together, these results indicate that RUNX3 is at least one of the proteins that mediate cell death after radiation therapy. Tissue samples in which RUNX3 was retained in the cytoplasm were resistant to irradiation. Cytoplasmic retention of RUNX3, as observed in many cases of esophageal cancer, may result from the absence or downregulation of one or more components of a signaling pathway required for RUNX3 nuclear translocation. RUNX3 localized in the cytoplasm cannot act as a transcription factor and does not elicit tumorsuppressive effects (Ito et al., 2005). The nuclear localization of RUNX3 is thus important, but its mechanisms require further examination. RUNX3 is a downstream target of the TGF-b signaling pathway. Most esophageal cancer-derived cell lines are resistant to stimulation by TGF-b and downregulation of RUNX3 in esophageal cancers frequently coincides with perturbations in TGF-b signaling. A previous study by Hanai et al. (1999) indicated that RUNX3 interacts with Smads, which transduce TGF-b signals, suggesting that transfection of RUNX3 could restore sensitivity to TGF-b. One study linked the downregulation of SMAD4 in esophageal cancers with poor prognosis (Natsugoe et al., 2002), and another identified TGF-b as an endogenous modulator of radiation effects (Kim et al., 2003). As shown by the luciferase assay, basal luciferase activity in stable RUNX3 transfectants was modestly upregulated by RUNX3 binding site-dependent transactivation in the absence of stimulation, and TGF-b treatment dramatically enhanced the activity in these cells. We detected no significant difference in SMAD2, 3 or 4 expression between parental cells and stable RUNX3 transfectants (data not shown), but increases in luciferase activity coincided with the restoration of responsiveness to TGF-b in the stable RUNX3 transfectants. RUNX3 is a transcription factor that regulates numerous downstream genes, for example, the cell-cycle regulator p21 (Chi et al., 2005). Although changes in
RUNX3 silencing in radioresistant esophageal cancers C Sakakura et al
5935
gene expression related to apoptosis, the cell cycle, drug metabolism, and cell adhesion are likely to contribute to radioresistance in a complex manner (Hanna et al., 2001; Kitahara et al., 2002; Fukuda et al., 2004), our results strongly suggest that RUNX3 is an important determinant of radiosensitivity in esophageal cancers and that it acts by affecting the expression of relevant genes. One such gene encodes the apoptosis-inducing factor Bim, which is positively regulated by RUNX3. Bim mediates many developmentally programmed and induced cytotoxic signals (Willis and Adams, 2005). Upregulation of Bim by RUNX3 results in apoptosis in gastric cancer cells (Yamamura et al., 2006; Yano et al., 2006). Here, we examined changes in Bim expression in esophageal squamous cell carcinoma cells and found that RUNX3 transfection induced Bim expression and enhanced sensitivity to radiation- and TGF-b-induced apoptosis. When RUNX3 expression was restored by a methyltransferase inhibitor, radiosensitivity increased in concert with the induction of Bim in both parental and radioresistant cells, suggesting that these inhibitors might also induce radiosensitivity in clinical esophageal cancers. Conversely, Bim knockdown by a Bim-specific siRNA effectively abolished the enhancement of radiation-induced apoptosis by methyltransferase inhibitor treatment or RUNX3 transfection. These findings indicate that the RUNX3 pathway participates in radiation-induced apoptosis in esophageal cancer cells, and they underscore the importance of Bim as a key mediator of radiation-induced apoptosis. In conclusion, our results indicate that inactivation of RUNX3 promotes radioresistance in esophageal cancers. Expression analysis of RUNX3 in pretreatment specimens may predict radiosensitivity, and RUNX3 induction could enhance radiosensitivity through the induction of Bim in esophageal squamous cell carcinomas.
Materials and methods Cell culture of esophageal cancer cell lines and RNA/DNA preparation The human esophageal squamous cancer cell lines TE-2, TE-5, TE-9 and TE-13 were obtained from the Cell Resource Center for the Biomedical Research Institute of Development, Aging and Cancer (Tohoku University, Sendai, Japan) (Nishihira et al., 1993). The cell lines KYSE160 and KYSE170 were kindly provided by Dr Shimada (Kyoto University, Kyoto, Japan) (Shimada et al., 1992). Cells were cultured in RPMI 1640 (Life Technologies, Grand Island, NY, USA) with 10% fetal bovine serum and incubated at 371C in 5% CO2 and 95% air. The establishment of radioresistant cell lines was reported by (Fukuda et al, 2004). Briefly, the cells were grown to 50% confluency, exposed to 2 Gy of X-ray irradiation and further incubated until the cells became 90% confluent. The cells were subcultured and irradiated again with the same dose of X-rays when they reached 50% confluency. The fractionated irradiation was continued until the total dosage reached 60 Gy. The parental cells were treated in the same way except that they were not exposed to X-ray irradiation. For all assays on irradiated cells, at least 2 weeks were allowed between the last 2 Gy irradiation and the experiments. Total RNA, mRNA
and genomic DNA were extracted as previously described (Sakakura et al., 1994, 1996). Clinical specimens The study population consisted of 62 patients with advanced esophageal squamous cell carcinomas diagnosed initially as T3 or T4 with image diagnosis, treated with preoperative radiotherapy and esophagectomy at the Kyoto Prefectural University of Medicine from 1999 to 2003. Written informed consent was obtained from each patient before tissue acquisition. All experiments with clinical samples were conducted under institutional guidelines from the Ministry of Health and Welfare. The specimens were classified according to the International Union Against Cancer tumor-node-metastasis classification system (Sobin and Fleming, 1997). Pretreatment samples were obtained from initial diagnostic biopsies of the tumor sites of each of the 62 patients. Tissue samples were also obtained from surgical specimens of each patient, following preoperative irradiation with 2 Gy per fraction to a total dosage of 40 Gy. The histological criteria for responses to radiotherapy were as follows (Japanese Society for Esophageal Disease, 1999): Grade 0, no radiotherapeutic effect; grade 1, slightly effective, with cancer cells accounting for more than one-third of viable cells; grade 2, moderately effective, with cancer cells accounting for less than one-third of viable cells; and grade 3, markedly effective, with no viable cancer cells. We compared the expression rates of RUNX3 for patients whose histological grade after radiotherapy was 0 or 1 with those for patients whose histological grade after radiotherapy was 2 or 3. Radiosensitivity assay Cell survival after X-irradiation was measured by a clonogenic assay as described previously (Fukuda et al., 2004). All survival curves represent at least three independent experiments. The data were fitted to the single-hit, multitarget formula, where survival (S) is related to dose (D) by the expression S ¼ 1 ð1 eDD0 Þn neDD0 . D0 was used as a parameter to indicate the amount of irradiation required to reduce the survival fraction to approximately 0.37 on the survival curve. Real-time quantitative RT–PCR cDNA was produced from total RNA with a SuperScript preamplification system (BRL, Bethesda, MD, USA) following the procedures suggested by the manufacturer, as described previously (Sakakura et al., 2002; Nakase et al., 2005). Quantitative PCR was performed using real-time Taqman TM technology, and products were analysed on a Model 5700 Sequence Detector (Applied Biosystems Corp., Foster City, CA, USA) as described previously (Sakakura et al., 2002; Nakase et al., 2005). To minimize errors arising from variation in the amount of starting RNA among samples, GAPDH mRNA was amplified as an internal reference against which other RNA values were normalized. Normalized results were expressed as the ratio of copies of each gene to copies of the GAPDH gene. Immunohistochemistry Paraffin-embedded tissue sections were routinely stained with hematoxylin-eosin. Specimens were stained immunohistochemically with the ABC method using monoclonal antibodies for RUNX3 (R3-6E9) (Ito et al., 2005) and SMAD4 (Santa Cruz Biotechnology, Santa Cruz, CA, USA). RUNX3 expression can be observed immunohistochemically in cells of normal esophageal epithelium, mainly in the Oncogene
RUNX3 silencing in radioresistant esophageal cancers C Sakakura et al
5936 basal layer. In the present study, we determined the percentage of cells showing RUNX3 expression in both the tumor and adjacent noncancerous epithelium among 1000 cells. More specifically, the fraction of RUNX3-positive tumor cells in five different areas was calculated and then averaged. Cases in which the majority of cells (>70%) showed nuclear RUNX3 expression were counted as positive. Cases with no expression, or with only minimal and equivocal expression in a minority of cells, were counted as negative, as described previously (Ito et al., 2005). Western blot analysis Western blot analysis was performed as described previously (Sakakura et al., 1994, 1996). An anti-RUNX3 antibody (mouse monoclonal antibody R3-5G4) (Ito et al., 2005), antiBim antibody (Santa Cruz Biotech, Santa Cruz, CA, USA), and anti-GAPDH antibody (Santa Cruz Biotech, Santa Cruz, CA, USA) were used for immunodetection. A horseradish peroxidase-conjugated anti-mouse IgG antibody was incubated with the blot at a 1:1000 dilution in PBS with 1% nonfat dry milk. The blot was visualized with an ECL Western blotting detection reagents (Amersham, Pittsburgh, PA, USA). Flow cytometry Cells were analysed during exponential growth. They were trypsinized, rinsed in PBS and fixed in 70% ethanol. Before analysis, cells were spun out of ethanol and resuspended in 1.6 ml PBS, 0.2 ml propidium iodide (PI) 400 mg/ml (Sigma Chemical Co., Poole, UK) and 0.2 ml RNase 1 mg/ml (Sigma, Poole, UK). After incubation for 30 min at 371C, the cells were spun down and resuspended in PBS. The cells were analyzed by a FACScan flow cytometer (Becton Dickinson, San Jose, CA, USA) for DNA content using PI and a photomultiplier masked with a 610-nm long-pass filter. The data were analysed with a software program (Cell Quest; Becton Dickinson) with dead cells gated out using pulse processing. Two independent experiments were performed for each data set. Treatment with TGF-b1 or 5-aza-20 -deoxycytidine TGF-b1 from porcine platelets (R&D, Minneapolis, MN, USA) was added to cell cultures at concentrations of 0.01– 5 ng/ml 24 h after cell seeding without a change of medium. 5-Aza-20 -deoxycytidine (Sigma, St Louis, MO, USA) at concentrations of 1–5 mM was added to cell cultures 24 h after cell seeding, and analysis was performed after 72 h of incubation. Luciferase assay The effect of RUNX3 on RUNX3-mediated transcriptional activity was evaluated with a dual luciferase assay as described previously (Sakakura et al., 2005). Cells were plated at 60 000 cells/well in six-well plates in complete growth medium, incubated for 24 h and transfected with pEF-BOS-RUNX3 or a neomycin construct according to the manufacturer’s protocol. The next day, the cells were treated for 16 h with TGF-b (5 ng/ml). Firefly luciferase (Luc) activities in cell lysates were determined with the Dual Luciferase Reporter Assay system (Promega, Pittsburgh, PA, USA) according to the manufacturer’s protocol. All assays were carried out in triplicate; each experiment was performed four times. Measurement of apoptosis To determine the apoptotic effect of RUNX3/TGF-b, cells were treated with 5 ng/ml TGF-b for 24 h in serum-free Oncogene
medium. The APO Percentage Apoptosis Assay (Biocolor, Belfast, Ireland) was used to assess the effect of RUNX3 overexpression on apoptosis. Vector or RUNX3-expressing esophageal cancer cells were seeded in a 96-well plate at a density of 2 103 cells/well. Apoptosis was induced by the addition of 5 ng/ml TGF-b in RPMI 1640 medium supplemented with 10% fetal bovine serum. Cellular dye uptake was analysed according to the manufacturer’s protocol. Methylation-specific PCR Methylation-specific PCR (MSP) was performed as reported previously (Herman et al., 1993). Briefly, 2 mg DNA was diluted in 50 ml dH2O, and 5.5 ml 2 M NaOH was added. After incubation of the mixture at 371C for 10 min, 30 ml freshly prepared 10 mM hydroquinone (Sigma, St Louis, MO, USA) and 520 ml 3 M sodium bisulfite (pH 5.0) were added. The mixture was incubated at 501C for 16 h, and the DNA was purified with Wizard DNA Purification Resin (Promega, Pittsburgh, PA, USA). The DNA pellet was dissolved in 20 ml H2O, and 2 ml was used for PCR using previously described primers (Sakakura et al., 2005). Methylated nucleotides between nucleotides –218 and –65 in the RUNX3 promoter were identified by sequencing the PCR products. Plasmids and stable transfection To obtain RUNX3-expressing TE-2 or KYSE170 cells, pEFBOS-RUNX3 was stably transfected into TE2 or KYSE170 cells by the Lipofectamine method as specified by the manufacturer (Gibco/BRL). Colonies resistant to 600 mg/ml G418 were selected and subcultured as described previously (Sakakura et al., 2005). Independent clones that expressed RUNX3 strongly were screened by Northern blotting. TE-2 or KYSE170 cells stably transfected with pEF-BOS-neo were used as controls. Independent clones that stably expressed RUNX3 or reduced levels of RUNX3 mRNA were identified, respectively, by Western blotting and by RT–PCR using the primers 50 -ctacggacatcctctggctcc-30 (1008–1029) and 50 catctctgccagcagcgtgctg-30 (1825–1846) to amplify 839 bp of RUNX3 cDNA not included in pEF-BOS-RUNX3. Small interfering RNA (siRNA) siRNA duplexes specifically targeting human Bim and a control nontargeting siRNA were purchased from Applied Biosystems Corp. (Foster City, CA, USA). Cells were transfected with siRNA duplexes using the transfection agent siPORT XP-1 (Applied Biosystems Corp., Foster City, CA, USA) and subjected to Western blotting and radiosensitivity assays 72 h after transfection. Statistical analysis Group differences were analysed using the Fisher’s exact test and t-test. The Kaplan–Meier method was used for survival analysis, and differences in survival were estimated using the log rank test. Po0.05 was considered to indicate statistical significance. Analyses were performed using JMP5 for Windows software (SAS Institute, Inc., Cary, NC, USA). Abbreviations 5-Aza-dCyt, 5-aza-20 -deoxycytidine; CR, complete response; MSP, methylation-specific PCR; PR, partial response; NC, no change; siRNA, small interfering RNA; TGF, transforming growth factor.
RUNX3 silencing in radioresistant esophageal cancers C Sakakura et al
5937 References Alcock RA, Dey S, Chendil D, Inayat MS, Mohiuddin M, Hartman G et al. (2002). Farnesyltransferase inhibitor (L-744, 832) restores TGF-beta type II receptor expression and enhances radiation sensitivity in K-ras mutant pancreatic cancer cell line MIA PaCa-2. Oncogene. 2002: 21:7883–21:7890. Biard DS, Martin M, Rhun YL, Duthu A, Leaix JL, May P. (1994). Concomitant p53 gene mutation and increased radiosensitivity in rat lung embryo epithelial cells during neoplastic development. Cancer Res 54: 3361–3364. Chi XZ, Yang JO, Lee KY, Ito K, Sakakura C, Li QL et al. (2005). RUNX3 suppresses gastric epithelial cell growth by inducing p21(WAF1/Cip1) expression in cooperation with transforming growth f. Mol Cell Biol 25: 8097–8107. Coleman CN, Stevenson MA. (1996). Biologic basis for radiation oncology. Oncology 10: 399–411. Dewey WC, Ling CC, Meyn RE. (1995). Radiation-induced apoptosis: relevance to radiotherapy. Int J Radiat Oncol Biol Phys 33: 781–796. Eifel PJ, Morris M, Wharton JT, Oswald MJ. (1994). The influence of tumor size and morphology on the outcome of patients with FIGO stage IB squamous cell carcinoma of the uterine cervix. Int J Radiant Oncol Biol Phys 29: 9–16. Ewan KB, Henshall-Powell RL, Ravani SA, Pajares MJ, Arteaga C, Warters R et al. (2002). Transforming growth factor-beta1 mediates cellular response to DNA damage in situ. Cancer Res 62: 5627–5631. Fukuchi M, Masuda N, Miyazaki T, Nakajima M, Osawa H, Kato H et al. (2002). Decreased Smad4 expression in the transforming growth factor-beta signaling pathway during progression of esophageal squamous cell carcinoma. Cancer 95: 737–743. Fukuda K, Sakakura C, Miyagawa K, Kuriu Y, Kin S, Nakase Y et al. (2004). Differential gene expression profiles of radioresistant oesophageal cancer cell lines established by continuous fractionated irradiation. Br J Cancer 91: 1543– 1550. Japanese Society for Esophageal Disease. (1999). General Rules for Clinical and Pathological Studies on Cancer of the Esophagus, 5th edn. Hanahan D, Weinberg RA. (2000). The hallmarks of cancer. Cell 100: 57–70. Hanai J, Chen LF, Kanno T, Ohtani-Fujita N, Kim WY, Guo WH et al. (1999). Interaction and functional cooperation of PEBP2/CBF with Smads. Synergistic induction of the immunoglobulin germline C alpha promoter. J Biol Chem 274: 31577–31582. Hanna E, Shrieve DC, Ratanatharathorn V, Xia X, Breau R, Suen J et al. (2001). A novel alternative approach for prediction of response of squamous cell carcinoma of head and neck. Cancer Res 61: 2376–2380. Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB. (1993). Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands. Proc Natl Acad Sci USA 93: 9821–9826. Hsiao M, Tse V, Carmel J, Costanzi E, Strauss B, Haas M et al. (1997). Functional expression of human p21 (WAFI/ CIPI) gene in rat glioma cells suppresses tumor growth in vivo and induces radiosensitivity. Biochem Biophys Res Commun 233: 329–335. Ito K, Liu Q, Salto-Tellez M, Yanao T, Tada K, Ida H et al. (2005). RUNX3, a novel tumor suppressor, is frequently inactivated in gastric cancer by protein mislocalization. Cancer Res 65: 7743–7750.
Kasid U, Pfeifer A, Weichselbaum RR, Dritschilo A, Mark GE. (1987). The raf oncogene is associated with a radiation-resistant human laryngeal cancer. Science 237: 1039–1041. Kim AH, Lebman DA, Dietz CM, Snyder SR, Eley KW, Chung TD et al. (2003). Transforming growth factor-beta is an endogenous radioresistance factor in the esophageal adenocarcinoma cell line OE-33. Int J Oncol 23: 1593–1599. Kim WJ, Kim EJ, Jeong P, Quan C, Kim J, Li QL et al. (2005). RUNX3 inactivation by point mutations and aberrant DNA methylation in bladder tumors. Cancer Res 65: 9347–9354. Kitada S, Krajewski S, Miyashita T, Krajewska M, Reed JC. (1996). Gamma-radiation induces upregulation of Bax protein and apoptosis in radiosensitive cells in vivo. Oncogene 12: 187–192. Kitahara O, Katagiri T, Tsunoda T, Harima Y, Nakamura Y. (2002). Classification of sensitivity or resistance of cervical cancers to ionizing radiation according to expression profiles of 62 genes selected by cDNA microarray analysis. Neoplasia 4: 295–303. Li QL, Ito K, Sakakura C, Fukamachi H, Inoue K, Chi XZ et al. (2002). Causal relationship between the loss of RUNX3 expression and gastric cancer. Cell 109: 113–124. Maity A, Kao GD, Muschel RJ, McKenna WG. (1997). Potential molecular targets for manipulating the radiation response. Int Radiat Oncol Biol Phys 3: 639–653. Massague J, Blain SW, Lo RS. (2000). TGFbeta signaling in growth control, cancer, and heritable disorders. Cell 103: 295–309. Matsumura Y, Yamagishi N, Miyakoshi J, Imamura S, Takebe H. (1997). Increase in radiation sensitivity of human malignant melanoma cells by expression of wild-type p16 gene. Cancer Lett 115: 91–96. Meyn RE, Stephens LC, Milas L. (1996). Programmed cell death and radioresistance. Cancer Metast Rev 15: 119–131. Nakase Y, Sakakura C, Miyagawa K, Kin S, Fukuda K, Yanagisawa A et al. (2005). Frequent loss of RUNX3 gene expression in remnant stomach cancer and adjacent mucosa with special reference to topography. Br J Cancer 92: 562–569. Natsugoe S, Xiangming C, Matsumoto M, Okumura H, Nakashima S, Sakita H et al. (2002). Smad4 and transforming growth factor beta1 expression in patients with squamous cell carcinoma of the esophagus. Clin Cancer Res 8: 1838–1842. Nishihira T, Hashimoto Y, Katayama M, Mori S, Kuroki T. (1993). Molecular and cellular features of esophageal cancer cell lines. J Cancer Res Clin Oncol 119: 441–449. Sakakura C, Hagiwara A, Nakanishi M, Shimomura K, Takagi T, Yasuoka R et al. (2002). Differential gene expression profiles of gastric cancer cells established from primary tumour and malignant ascites. Br J Cancer 87: 1153–1161. Sakakura C, Hasegawa K, Miyagawa K, Nakashima S, Yoshikawa T, Kin S et al. (2005). Possible involvement of RUNX3 silencing in the peritoneal metastases of gastric cancers. Clin Cancer Res 11: 6479–6488. Sakakura C, Sweeney EA, Shirahama T, Igarashi Y, Hakomori S, Nakatani H et al. (1996). Overexpression of bax sensitizes human breast cancer MCF-7 cells to radiation-induced apoptosis. Int J Cancer 67: 101–105. Sakakura C, Yamaguchi-Iwai Y, Satake M, Bae SC, Takahashi A et al. (1994). Growth inhibition and induction Oncogene
RUNX3 silencing in radioresistant esophageal cancers C Sakakura et al
5938 of differentiation of t(8;21) acute myeloid leukemia cells by the DNA-binding domain of PEBP2 and the AML1/ MTG8(ETO)-specific antisense oligonucleotide. Proc Natl Acad Sci USA 91: 11723–11727. Shimada Y, Imamura M, Wagata T, Yamaguchi N, Tobe T. (1992). Characterization of 21 newly established esophageal cancer cell lines. Cancer 69: 277–284. Sinclair WK, Morton RA. (1966). X-ray sensitivity during the cell generation cycle of cultured Chinese hamster cells. Radiat Res 29: 450–474. Sklar MD. (1988). The ras oncogenes increase the intrinsic resistance of NIH 3T3 cells to ionizing radiation. Science 239: 645–647. Sobin LH, Fleming ID. (1997). TNM Classification of malignant tumors. 5th ed. Union Internationale Contre le
Oncogene
Cancer and the American Joint Committee on Cancer. Cancer 80: 1803–1804. Vodovotz Y, Lucia MS, DeLucca AM, Mitchell JB, Kopp JB. (2000). Hematopoietic function and enhanced radiosensitivity of transforming growth factor-beta1 transgenic mice. Int J Cancer 90: 13–21. Willis SN, Adams JM. (2005). Life in the balance: how BH3-only proteins induce apoptosis. Curr Opin Cell Biol 17: 617–625. Yamamura Y, Lee WL, Inoue KI, Ida H, Ito Y. (2006). RUNX3 cooperates with FoxO3a to induce apoptosis in gastric cancer cells. J Biol Chem 281: 5267–5276. Yano T, Ito K, Fukamachi H, Chi XZ, Wee HJ, Inoue K et al. (2006). The RUNX3 tumor suppressor upregulates Bim in gastric epithelial cells undergoing TGF-b-induced apoptosis. Mol Cell Biol. 26: 4474–4488.
