TWIST1 and SNAI1 as markers of poor prognosis in human colorectal cancer are associated with the expression of ALDH1 and TGF-β1

  • Authors:
    • Yong Hun Kim
    • Gwangil Kim
    • Chang-Il Kwon
    • Jong Woo Kim
    • Pil Won Park
    • Ki-Baik Hahm
  • View Affiliations

  • Published online on: January 8, 2014     https://doi.org/10.3892/or.2014.2970
  • Pages: 1380-1388
Metrics: Total Views: 0 (Spandidos Publications: | PMC Statistics: )
Total PDF Downloads: 0 (Spandidos Publications: | PMC Statistics: )


Abstract

Epithelial-mesenchymal transition (EMT) is an important factor in cancer invasiveness and metastatic progression. During EMT, cancer cells acquire stem cell properties. The role of EMT and stemness in colon cancer has not been fully understood. We aimed to demonstrate the clinical significance of EMT and the stem cell phenotype in colorectal cancer. Two hundred and thirty-one surgically resected colon cancer cases were included in the present study. mRNAs of E-cadherin, TWIST1 and SNAI1 were analyzed by quantitative real-time polymerase chain reaction (qRT-PCR) (n=109). Immunohistochemical staining was performed for six markers (ALDH1, TGF-β1, E-cadherin, β-catenin, TWSIT1 and SNAI1) (n=231). We assessed clinicopathological characteristics according to the expression of the stem cell phenotype and EMT markers. Based on the results of qRT-PCR, TWIST1 and SNAI1 significantly influenced node metastasis (P=0.04 and P=0.02, respectively). High TWIST1 and SNAI1 mRNA expression was associated with poor overall survival according to the univariate analysis (P<0.01 and P=0.01, respectively) and the multivariate analysis (P=0.04 and P=0.04, respectively). ALDH1 expression as detected by immunohistochemical staining was associated with high nodal stage, advanced clinical stage, lymphatic invasion and poor survival (P=0.01, P=0.04, P<0.05 and P<0.01, respectively) and with the expression of TGF-β1 and β-catenin. In conclusion, in human colorectal cancer, the EMT markers TWIST1 and SNAI1 are suggested as important markers of poor prognosis. Their expression is associated with the expression of putative stem cell marker ALDH1, and ALDH1 is associated with the expression of TGF-β1.

Introduction

Colorectal cancer, one of the most common epithelial carcinomas originating from epithelial cells, is the major cause of cancer-related mortality in industrialized countries (1). Despite the development of numerous surgical techniques and new treatment methods, metastatic colorectal cancer is associated with a low (~10%) 5-year survival rate and accounts for 90% of colon cancer-related deaths (2,3). Primary tumors are known to metastasize through the process of epithelial-mesenchymal transition (EMT) (4), which is an important step by which epithelial cells transform into mesenchymal cells during embryonic development, but loss of cell polarity, decreased cell to cell adhesion, increased migratory ability and increased invasive property are caused by the aberrant activation of EMT in cancer cells (57). It is essential, therefore, to understand the process of EMT that occurs during the development of colon cancer, in addition to understanding the mechanisms involved in the progression of carcinoma, to be able to establish objectives for the prevention and treatment of metastasis. The process of EMT is accompanied by various changes, such as repression of epithelial markers, abnormal translocation of β-catenin, and upregulation of mesenchymal markers, as well as expression of and interaction between related proteins or so-called EMT regulators, such as TWIST1 and SNAI1 (SNAIL), but there is still a lack of understanding in individual carcinomas (6,8).

According to previous research, cancer cells appear to acquire a stem cell-like phenotype during the process of EMT (9). Cancer stem cells have features of self-renewal and tumor-initiating capacity, and the resultant tumors can be defined as cells having high heterogeneity of a primary tumor. It has been suggested that, unlike stem cells in the crypt of normal colonic mucosa, the microenvironment of cancer stem cells hosts variants of diverse self-renewable signaling pathways, such as the Wnt/β-catenin, TGF-β, Notch and Hedgehog signaling pathways (10). Cancer stemness is a concept introduced to describe cancer cell resistance to conventional chemoradiation therapies and has been suggested to have a close relationship with EMT in breast and prostate cancer (1113). ALDH1 has been found to be a valuable marker in previous studies for identifying breast and colon cancer stem cells (1316). It has been speculated that the extent of the expression of cancer stem cell phenotype markers may be associated with the clinical prognosis of colon cancer, but without sufficient clinical research to support it.

In the present study, we aimed to determine whether EMT markers and the extent of expression of stem cell phenotype markers are associated with the histological features and clinical prognosis of colon cancer patients.

Materials and methods

Patients and tissue samples

After obtaining approval from the Institutional Review Board of Bundang CHA Medical Center, 231 patients with primary colon cancer recruited between March 2002 and October 2006 were included in the present study. The patients underwent surgery for colon cancer at our center, and the collected samples were examined and classified by GI pathologists at the center. The 231 patients consisted of 130 males and 101 females, ranging in age from 31 to 95 years (mean age, 61.8). The patients were followed up for a median of 71 (1–106) months. Of the 231 cases, 23 cases were mucinous adenocarcinoma and 37 cases were right colon cancer. Other clinical parameters such as gender, clinical stage and survival time (months) are detailed in Tables I and II. Tissue microarray was used for the 231 cases of paraffin-embedded tissue blocks. For 109 cases with available normal tumor-paired samples, fresh snap-frozen samples were used for quantitative analysis of mRNA expression. In order to select cases containing at least 90% tumor cells, frozen tissues were subjected to histological evaluation with H&E staining. Clinical stages were determined according to the AJCC staging system, 7th edition by a clinician who reviewed the surgical pathology reports and clinical information (17).

Table I

Clinicopathological features and the mean values of the relative quantification of mRNA expression of EMT markers in the colorectal cancer cases.

Table I

Clinicopathological features and the mean values of the relative quantification of mRNA expression of EMT markers in the colorectal cancer cases.

nE-cadherin (mean ±SD)TWIST1 (mean ±SD)SNAI1 (mean ±SD)
Total no. of samples10972.73±293.628.13±31.73.47±6.06
Gender
 Male642.97±298.911.28±34.811.53±3.84
 Female453.64±281.930.66±26.261.48±8.17
 P-value0.7510.4750.929
Age (years)
 ≤5020105.63±363.013.53±6.223.67±4.02
 >508965.34±273.329.17±34.723.42±6.4
 P-value0.3620.6700.568
Location of colon cancer
 Right colon233.64±58.240.66±23.91.59±2.72
 Left colon863.22±326.530.78±33.311.48±6.4
 P-value0.8320.4690.741
Primary lesion
 Single1053.37±297.470.73±29.951.48±5.95
 Multiple41.46±4.191.12±54.684.47±7.17
 P-value0.2030.5730.375
Pathological diagnosis
 Adenocarcinoma933.16±312.530.8±31.731.45±6.21
 Mucinous adenocarcinoma165.91±118.630.53±30.581.99±4.81
 P-value0.1320.1370.177
Differentiationb
 Well to moderate863.23±324.340.81±32.451.38±3.52
 Poor to undifferentiated71.69±8.220.53±20.712.38±17.3
 P-value0.5310.4320.541
Tumor stage
 T1–2111.81±7.870.54±2.790.35±1.58
 T3–4983.62±307.30.78±33.191.48±6.4
 P-value0.0690.0730.024a
Nodal stage
 N0563.44±332.950.68±15.640.78±2.89
 N1–3533.27±241.820.89±41.841.6±7.94
 P-value0.7120.040a0.022a
Clinical stage
 Stage I–II543.23±258.890.68±15.921.14±2.92
 Stage III–IV553.37±320.050.89±42.121.59±7.84
 P-value0.3410.0570.064
Lymphatic invasion
 Negative843.1±232.070.66±35.281.5±3.8
 Positive251±0.482.06±12.531.48±10.35
 P-value0.1040.032a0.668
Vascular invasion
 Negative963.23±310.590.71±33.531.48±6.22
 Positive1316.52±23.150.89±3.241.6±4.34
 P-value0.7360.8700.483
Neural invasion
 Negative1023.28±301.610.69±32.181.51±3.94
 Positive77.62±7.10.88±17.713.52±20.14
 P-value0.4150.0610.781

{ label (or @symbol) needed for fn[@id='tfn1-or-31-03-1380'] } Data were analyzed using the Mann-Whitney U test.

a P<0.05.

b Except mucinous adenocarcinoma.

{ label (or @symbol) needed for fn[@id='tfn4-or-31-03-1380'] } EMT, epithelial-mesenchymal transition.

Table II

Relationship between the clinicopathological features and positive immunohistochemical expression of EMT and stemness markers (ALDH1, TGF-β1, E-cadherin, β-catenin, TWIST and SNAI1) in the 231 colorectal cancer patients.

Table II

Relationship between the clinicopathological features and positive immunohistochemical expression of EMT and stemness markers (ALDH1, TGF-β1, E-cadherin, β-catenin, TWIST and SNAI1) in the 231 colorectal cancer patients.

ALDH1TGF-β1Nuclear E-cadherinβ-cateninTWIST1SNAI1
nn (%)n (%)n (%)n (%)n (%)n (%)
Gender
 Male13026 (20.0)40 (30.8)88 (67.7)66 (50.8)37 (28.5)33 (25.4)
 Female10116 (15.8)24 (23.8)68 (67.3)50 (49.5)28 (27.7)22 (21.8)
 P-value0.4160.2380.9530.8490.9010.524
Age (years)
 ≤50414 (9.8)9 (22.0)26 (63.4)15 (36.6)8 (19.5)11 (26.8)
 >5019038 (20.0)55 (28.9)130 (68.4)101 (53.2)57 (30.0)44 (23.2)
 P-value0.1790.3640.5350.0540.1760.617
Location of colon cancer
 Right colon378 (21.6)6 (16.2)27 (73.0)15 (40.5)10 (27.0)9 (24.3)
 Left colon19434 (17.5)58 (29.9)129 (66.5)101 (52.1)55 (28.4)46 (23.7)
 P-value0.5540.0880.4410.1990.8700.936
Primary lesion
 Single22239 (17.6)59 (26.6)149 (67.1)112 (50.5)61 (27.5)52 (23.4)
 Multiple93 (33.3)5 (55.6)7 (77.8)4 (44.4)4 (44.4)3 (33.4)
 P-value0.2120.1200.5030.7240.2670.448
Pathological diagnosis
 Adenocarcinoma20835 (16.8)57 (27.4)143 (68.8)109 (52.4)59 (28.4)50 (24.0)
 Mucinous adenocarcinoma237 (30.4)7 (30.4)13 (56.5)23 (44.4)6 (26.1)6 (21.7)
 P-value0.1080.7600.2350.046a0.8180.806
Differentiationb
 Well-moderate19534 (17.4)55 (28.2)137 (70.3)106 (54.4)59 (30.3)46 (23.6)
 Poor-undifferentiated131 (7.7)2 (15.4)6 (46.2)3 (30.4)0 (0.0)4 (30.8)
 P-value0.7000.5210.0640.042a0.022a0.517
Tumor stage
 T1–2244 (16.7)7 (29.2)15 (62.5)17 (70.8)7 (29.2)5 (20.8)
 T3–420738 (18.4)57 (27.5)141 (68.1)99 (47.8)58 (28.0)50 (24.2)
 P-value0.8390.8660.5780.033a0.9060.718
Nodal stage
 N011313 (11.5)24 (21.2)77 (68.1)62 (54.9)28 (24.8)26 (23.0)
 N1–311829 (24.6)40 (33.9)79 (79.0)54 (45.8)37 (31.4)29 (24.6)
 P-value0.010a0.032a0.8470.1670.2660.780
Clinical stage
 Stage I–II11114 (12.6)25 (22.5)76 (68.5)61 (55.0)29 (26.1)26 (23.4)
 Stage III–IV12028 (23.3)39 (32.5)80 (66.7)55 (45.8)36 (30.0)29 (24.2)
 P-value0.035a0.0900.810.1660.5130.895
Lymphatic invasion
 Negative17627 (15.3)48 (27.3)122 (69.3)94 (53.4)49 (27.8)41 (23.3)
 Positive5515 (27.3)16 (29.1)34 (61.8)22 (40.4)16 (29.1)14 (25.5)
 P-value0.045a0.7930.3000.0830.8570.743
Vascular invasion
 Negative18230 (16.5)51 (28.0)127 (69.8)89 (48.9)48 (26.4)37 (20.3)
 Positive4912 (24.5)13 (26.5)29 (59.2)27 (55.1)17 (34.7)18 (36.7)
 P-value0.1970.8360.1600.4410.2500.017a
Neural invasion
 Negative20937 (17.7)55 (26.3)144 (68.9)106 (50.7)58 (27.8)52 (24.9)
 Positive225 (22.7)9 (40.9)12 (54.5)10 (45.5)7 (31.8)3 (13.6)
 P-value0.5640.1460.1710.6390.6870.301
Total23142641561166555

{ label (or @symbol) needed for fn[@id='tfn5-or-31-03-1380'] } Data were analyzed using χ2 test or Fisher’s exact test.

a P<0.05.

b Except mucinous adenocarcinoma.

{ label (or @symbol) needed for fn[@id='tfn8-or-31-03-1380'] } EMT, epithelial-mesenchymal transition.

RNA extraction and quantitative real-time polymerase chain reaction (qRT-PCR)

RNA was extracted from 109 fresh snap-frozen samples of colon cancer tissues and from the paired normal tissues using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). cDNA was synthesized from the extracted RNA (1 μg) using the SuperScript III kit (Invitrogen). Each cDNA sample was run three times independently using the Bio-Rad CFX96 Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA). Gene-specific primers and probes for the TaqMan gene expression assay were purchased from Applied Biosystems (Paisley, UK). This included three genes: E-cadherin (Hs00170423_m1), TWIST1 (Hs00361186_m1) and SNAI1 (Hs00195591_m1). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Hs99999905_m1) was used for normalization. The PCR reaction mix consisted of a volume of 20 μl and was composed of 10 μl 2X TaqMan Universal PCR Master Mix (Applied Biosystems), 1 μl of primers and probe kit, 1 μl of cDNA and 8 μl of diethyl pyrocarbonate water. The reverse transcription was carried out for 15 sec at 90°C, repeated 40 times for 1 min at 60°C, and then was treated for 2 min at 50°C and for 10 min at 95°C. Initial copy number at real-time PCR analysis was quantified based on the threshold cycle (Ct), and target genes were normalized against GAPDH. Relative quantification of mRNA expression was calculated by the 2−ΔΔCt method.

Immunohistochemical analysis

For the tissue microarray of the 231 cases, tissue cores were obtained at a 2-mm diameter from expression regions on formalin-fixed paraffin-embedded blocks using a manual tissue microarray kit (Unitma, Quick-Ray; Unitech Science, Seoul, Korea). The obtained tissues consisted of three cancer tissue cores from different foci and one matched normal tissue core. Immunohistochemical staining was performed on 5-μm sections cut from formalin-fixed, paraffin-embedded tissue using mouse monoclonal anti-ALDH1 (1:50) (BD Biosciences, La Jolla, CA, USA), rabbit polyclonal anti-TGF-β1 (1:100) (Novus Biologicals, Littleton, CO, USA), mouse monoclonal anti-E-cadherin (1:100), mouse monoclonal anti-β-catenin (1:100) (both from BD Biosciences), rabbit polyclonal anti-TWIST1 (1:200) and rabbit polyclonal anti-SNAI1 (1:100) (both from Novus Biologicals). The visualization system used was the BenchMark XT with heat-induced epitope retrieval (CC1 solution) (both from Ventana). Sections were incubated with primary antibodies for 32 min at 37°C. Staining was detected with the ultraView Universal DAB detection kit (Ventana). Immunohistochemical content was scored depending on the extent and intensity of staining, as previously described (18). In brief, the intensity of staining was graded according to a 4-tiered scale of 0 to 3 (with 3 as the most intense staining). The extent of positive immunoreactivity was graded according to the percentage of stained cells in the region of interest: 0 points for 0%; 1 point for <20%; 2 points for 20–50% and 3 points for >50%. An overall score was obtained from the sum of the intensity and the extent of the positive-staining. Cases with a final score of >3 were defined as positive. Cases with cytoplasmic and/or nuclear localization of β-catenin were considered abnormal. All staining was separately scored in a blinded manner by two trained researchers (Y.K. and C.K.) and one pathologist (G.K.). In the case of disagreement in interpretation, the results were discussed by all three researchers and a consensus was reached for each case.

Statistical analysis

Quantitative values of qRT-PCR were presented as means ± standard deviation (SD) or median. The Mann-Whitney U test was used for quantitative comparison of mRNAs with respect to various clinical and histological parameters. The Chi-square test was used to compare immunohistochemical staining with clinical and pathological indices. For correlation of EMT and stemness markers, Spearman’s rank correlation was used. Patient overall survival was determined from official death records by the Korea Statistics Promotion Institute. Survival curves were calculated using the Kaplan-Meier method, and differences between the survival curves were analyzed using log-rank test. Multivariate analysis of independent prognostic factors of survival was performed using Cox’s proportional hazard regression model. All statistical analyses were performed with IBM® SPSS® Statistics (version 19.0.0), and P-value <0.05 was considered to indicate a statistically significant result.

Results

Correlation between the relative quantification (RQ) of mRNA expression of EMT markers and clinicopathologic characteristics of the colorectal cancer cases

The mean values of RQ for E-cadherin, TWIST1 and SNAI1 were 72.73±293.62 (median, 3.27), 8.13±31.7 (median, 0.73) and 3.47±6.06 (median, 1.48), respectively in the 109 cases. Table I summarizes the differences in the mean values of RQ of each marker according to the clinicopathological characteristics. The mean values of RQ of E-cadherin, TWIST1 and SNAI1 were not significantly different in regards to gender, age, location of the colon cancer, number of primary lesions and pathological diagnosis. When samples excluding mucinous adenocarcinoma were divided into a well to moderate differentiation group and a poor to undifferentiated group, the two subgroups were not significantly different in regards to the mean values of RQ for each marker. The mean values of RQ for TWIST1 and SNAI1 were significantly higher in the subgroup without nodal invasion (N0) when compared to that with nodal invasion (N1-3) (P=0.040 and P=0.022, respectively) (Fig. 1A and B, respectively). The mean value of RQ for TWIST1, but not for those of E-cadherin and SNAI1, was significantly high in the presence of lymphatic invasion (P=0.032) (Fig. 1C). SNAI1 showed a higher mean value of RQ in T3-4 stage than in T1-2 stage cases (P=0.024) (Fig. 1D), and the mean values of RQ for E-cadherin, TWIST1 and SNAI1 were not significantly different depending on whether vascular invasion or neural invasion was present or not.

Immunohistochemical staining for EMT and stem cell markers

qRT-PCR was performed for mRNAs of EMT and stem cell markers, and immunohistochemical staining was conducted to validate the possibility of EMT marker expression in the surrounding mesenchymal cells of the tumor cells. Representative images of the immunohistochemical staining for each marker are presented in Fig. 2. ALDH1 was positive in ~5% of cells at the crypt of the normal colon epithelium. E-cadherin and β-catenin were strongly expressed along the cell membranes of normal epithelium. TGF-β1, TWIST1 and SNAI1 were not expressed in normal epithelial cells. The relationship between immunohistochemical staining and the clinicopathological characteristics of our 231 cases is summarized in Table II. ALDH1, TGF-β1, E-cadherin, nuclear β-catenin, SNAI1 and TWSIT1 were not significantly different in terms of immunohistochemical staining according to gender, age, location of the lesion and number of primary lesions. The extent of nuclear β-catenin expression was significantly higher in the mucinous adenocarcinoma than in the adenocarcinoma cases (P=0.046). When carcinomas excluding mucinous adenocarcinoma were divided into a well to moderately differentiated group and a poor to undifferentiated group, the former was observed to express nuclear β-catenin and TWIST1 significantly at a higher extent (P=0.042 and P=0.022, respectively). Nuclear β-catenin expression was higher in T1-2 stage than in T3-4 stage cases (p=0.033), and ALDH1 and TGF-β1 were expressed at a higher extent in N1-3 stage than in N0 stage cases (P=0.010 and P=0.032, respectively). In terms of lymphatic invasion, vascular invasion and neural invasion analysis, ALDH1 was highly expressed in the presence of lymphatic invasion (P=0.045) and SNAI1 in the presence of vascular invasion (P=0.017).

Correlation analysis of EMT and stem cell markers

Spearman’s rank correlation was used to test the correlation between EMT and stemness markers (Table III). The mRNA expression levels of E-cadherin and TWIST1 were negatively correlated, while E-cadherin mRNA expression and immunohistochemical staining were positively correlated (P=0.045 and P=0.012, respectively). The mRNA expression of TWIST1 was positively correlated with the extent of expression of ALDH1, TGF-β1 and TWIST1 according to immunohistochemical staining (P=0.001, P=0.010 and P=0.004, respectively). Immunohistochemical staining of ALDH1 was significantly associated with that of TGF-β1 and nuclear β-catenin (P=0.018 and P=0.010, respectively), and immunochemical staining of TGF-β1 was positively correlated with that of TWIST1 (P=0.017).

Table III

Correlation between mRNA expression, immunohistochemical staining of EMT and stemness markers in the colorectal cancer cases.

Table III

Correlation between mRNA expression, immunohistochemical staining of EMT and stemness markers in the colorectal cancer cases.

mRNA expressionImmunochemical staining


Rho valueE-cadherinTWIST1SNAI1ALDH1TGF-β1E-cadherinβ-cateninTWIST1SNAI1
E-cadherin1−1.930.050.01−0.130.240.010.050.16
P-value0.045a0.6360.9600.1810.012a0.9280.6360.101
TWIST110.140.310.250.030.140.27−0.01
P-value0.1530.001a0.010a0.7880.1520.004a0.940
SNAI110.180.07−0.060.12−0.70.16
P-value0.0580.5040.5090.2180.4500.101
ALDH110.230.050.250.070.13
P-value0.018a0.6280.010a0.4550.195
TGF-β11−0.040.090.23−0.010
P-value0.6550.3460.017a0.916
E-cadherin10.140.030.11
P-value0.1620.7390.260
β-catenin1−0.890.15
P-value0.3630.121
TWIST11−0.09
P-value0.377

{ label (or @symbol) needed for fn[@id='tfn9-or-31-03-1380'] } Data were analyzed using the Spearman’s rank statistical test.

a P<0.05.

{ label (or @symbol) needed for fn[@id='tfn11-or-31-03-1380'] } EMT, epithelial-mesenchymal transition.

Correlation between patient survival and EMT and stemness markers

EMT and stemness markers and overall survival were tested using univariate and multivariate analyses. Kaplan-Meier univariate analysis was performed according to the median values of RQ as cut-offs (i.e. high vs. low RQ) for E-cadherin, TWIST1 and SNAI1 mRNAs and overall survival. Cases that had high RQ for TWIST1 and SNAI1 mRNAs showed significantly lower overall survival. The mean survival of the group with high RQ of TWIST1 mRNAs was 45.1 months, which was significantly different from the mean survival of the group with a low mean RQ of TWIST1 mRNAs (53.5 months, P=0.014, log-rank test) (Fig. 3A); the mean survival of the group with a high mean RQ of SNAI1 mRNAs was 45.1 months, which was significantly different from 53.5 months in the lower group (P=0.002, log-rank test) (Fig. 3B). In the multivariate analysis, TWIST1 and SNAI1 were found to be independent significant prognostic factors for poor overall survival (P=0.041 and P=0.039, respectively; Cox’s proportional hazards regression model) (Table IV). In the Kaplan-Meier univariate analysis of overall survival and immunohistochemical staining of ALDH1, TGF-β1, E-cadherin, β-catenin, TWIST1 and SNAI1, overall survival was significantly different only for ALDH1 expression (P=0.003, log-rank test), but the significance was lost in the multivariate analysis (P=0.232, Cox’s proportional hazards regression model).

Table IV

Correlation of patient survival and EMT and stem cell markers in the colorectal cancer cases.

Table IV

Correlation of patient survival and EMT and stem cell markers in the colorectal cancer cases.

VariablesUnivariate analysisbMultivariate analysiscHazard ratio95% CI
E-cadherin0.8590.8970.950.46–1.97
TWIST10.002a0.039a2.291.04–5.00
SNAI10.014a0.041a2.111.03–4.33
Differentiation0.023a0.1932.120.68–6.59
Stage (I, II vs. III+IV)<0.0001a0.001a3.651.66–8.00

{ label (or @symbol) needed for fn[@id='tfn12-or-31-03-1380'] } CI, confidence interval.

a P<0.05.

b Log-rank test.

c Cox’s proportional hazards model.

{ label (or @symbol) needed for fn[@id='tfn16-or-31-03-1380'] } EMT, epithelial-mesenchymal transition.

Discussion

EMT and cancer stem cells are important pathophysiological concepts for tumorigenesis and metastasis of cancer. Mani et al confirmed the role of TWIST1 and SNAI1 as EMT regulators by demonstrating EMT induction and stem cell properties when TWIST1 and SNAI1 were overexpressed in immortalized human mammary epithelial cells (9). TWIST1 and SNAI1 are known to be controlled independently and to promote EMT through collaboration as well (19,20). In previous studies, TWIST1 overexpression was observed not only in breast, prostate, lung, uterine and skin cancers but also in upper gastrointestinal tract cancers of the esophagus and stomach, with significant outcomes in terms of invasiveness (2124). However, the role of TWIST1 overexpression is not yet clear in colorectal types of cancers, and there has been only a limited number of studies investigating the association between the degree of TWIST1 expression and survival in human colon cancer cases. Valdés-Mora et al suggested that TWIST1 overexpression in colon cancer is associated with nodal invasion, but unlike our study they did not confirm the association with overall survival (22). In addition, they concluded that TWIST1 expression was significantly higher in males and did not lead to a difference in lymphovascular invasion. They only determined the mRNA level of TWIST1 by using real-time PCR without validation of protein expression. In addition, they used a relatively small sample size of 54 colon cancer patients which included twice as many men as women. We found that higher expression of TWIST1 mRNA was associated with lymph node metastasis and poor overall survival in the 109 patients with colorectal carcinomas and their mRNA expression was validated by immunohistochemical staining in an expanded population of 231 colorectal carcinoma cases.

The Snail family, a group of zinc-finger transcription repressors, is known to play an important role in EMT by inhibiting the same cell junction component as E-cadherin (25). Shioiri et al suggested that the expression of SLUG (SNAI2), one of the Snail family members, was an independent prognostic factor of colon cancer (26). SNAI1, with higher affinity to E-cadherin and other target genes than SNAI2, is a candidate prognostic factor of colon cancer and is known to be significantly associated with colon cancer invasiveness and metastasis (27,28). Although the association between SNAI1 and E-cadherin was not found in the present study, we found that higher expression of SNAI1 mRNA was associated with lymph node metastasis and higher T stages. Moreover, we confirmed that overexpression of SNAI1 mRNA is a poor prognostic factor in terms of patient overall survival.

Taken together, our results support the hypothesis that TWIST1 and SNAI1 overexpression in colon cancer stimulates the process of EMT, and the resultant lymphatic and nodal invasion affects the progression of the tumors and patient survival. Since previous studies suggest that EMT is induced by cancer stem cells (9), we analyzed the association of various EMT markers and putative colorectal cancer stem cell markers. The cancer stem cell hypothesis, first introduced by Rudolf Virchow, was confirmed by Al-Hajj et al (14) in breast cancer, which is an epithelium-derived tumor, and then by O’Brien et al (29) who confirmed the presence of colon cancer stem cells by transplanting CD133+ human colon cancer-initiating cells to immunodeficient mice. However, later reports of the non-specificity of CD133 expression in stem cells motivated us to focus on ALDH1 (30). Highly active ALDH1 increases the resistance to oxidative insults, thereby increasing the resistance to chemotherapies, and is suspected to play an important role in the recurrence and survival of carcinomas. In the present study, higher expression of ALDH1 immunostaining in colorectal cancer tissues was significantly associated with higher clinical and nodal stages, lymphatic invasion and worse overall survival. However, the association between ALDH1 expression and overall survival was significant only in the univariate analysis (P=0.003). One of the possible reason is that the expression of ALDH1 was highly correlated with and dependent on the expression of TWIST1 as shown in Table III. Our results showed that TWIST1 expression had a better association with expression of other EMT and stem cell markers than SNAI1. This may be because SNAI1 is less stable inside the cell nucleus when compared with TWIST1, although the exact half-life is unknown (31). Furthermore, in previous in vitro studies, TWIST1 was found to induce SNAI1 and other EMT regulators; Weinberg et al even suggested TWIST1 as the master transcription factor related to EMT (32,33).

In order to explain the association between EMT and cancer stem cells, we examined known EMT-related cytokines and the stem cell pathway. TGF-β is known as a cytokine that plays an important role in transcriptional activation of the Snail family, TWIST and others as well as in the regulation of cancer stem cells (34). We observed a positive correlation between TGF-β1 and ALDH1 in the immunohistochemical staining. Yet, TGF-β1 expression was not associated with the prognosis of the colorectal cancer. Increased nuclear β-catenin, the key effector of the Wnt pathway, affects EMT, cell to cell adhesion and stem cell phenotype (35). In the present study, nuclear β-catenin was expressed to a significantly higher extent in well to moderately differentiated carcinomas. Although nuclear β-catenin was identified to a greater extent in tumors with lower T-stage and higher ALDH1 expression, it was not associated with patient survival. Accordingly, our results suggest that TWIST1 is activated in human colon cancer by TGF-β, which is known to induce pro-EMT signaling stimulus and a stemness phenotype, and that activated TWIST1 regulates the initiation of EMT by suppressing E-cadherin, activates the EMT process by inducing SNAI1, and causes a stemness phenotype by inducing the expression of ALDH1. In addition, we propose that ALDH1 induces colon cancer proliferation by means of increasing nuclear translocation of β-catenin.

The present study had some limitation as functional experiments revealing causal relationship among factors were not performed. One more limitation was that the identification of EMT and stem cell markers in metastatic lesions was not shown. Nevertheless, the possibility of controlling TWIST1 and SNAI1 is an appropriate strategy in the treatment of colorectal cancers. In prostate cancer, downregulation of TWIST1 activates apoptosis and promotes chemosensitization to Taxol by downregulating the BCl-2/Bax ratio, while overexpression of TWIST1 inhibits paclitaxel-induced apoptosis in nasopharyngeal cancer (36,37). SNAI1 overexpression is also known to induce chemoresistance to oxaliplatin in colon cancer (38). TGF-β-mediated control of EMT markers and the stemness phenotype affect cancer metastasis and invasiveness. Clinical studies on ligand trap, antisense oligonucleotide and small molecule receptor kinase inhibitors, which inhibit TGF-β, are currently underway, and further studies are expected to determine the role of such agents in the inhibition of TWIST1 and SNAI1 expression and inhibition of EMT (39).

In conclusion, we demonstrated that overexpression of TWIST1 and SNAI1, both as EMT markers, is associated with poor overall survival in patients with colorectal cancers. Moreover, their expression was correlated with the expression of TGF-β1 and the cancer stem cell phenotype. Our results suggest that controlling the EMT process through TWIST1, SNAI1 or TGF-β1 in colorectal cancers can be a possible therapeutic target of cancer stem cells. Further studies concerning the regulation of the process of EMT with functional experiments are warranted.

References

1 

McDermott U, Longley DB and Johnston PG: Molecular and biochemical markers in colorectal cancer. Ann Oncol. 13(Suppl 4): 235–245. 2002. View Article : Google Scholar

2 

Jemal A, Murray T, Ward E, et al: Cancer statistics, 2005. CA Cancer J Clin. 55:10–30. 2005. View Article : Google Scholar

3 

Sporn MB: The war on cancer. Lancet. 347:1377–1381. 1996. View Article : Google Scholar : PubMed/NCBI

4 

Voulgari A and Pintzas A: Epithelial-mesenchymal transition in cancer metastasis: mechanisms, markers and strategies to overcome drug resistance in the clinic. Biochim Biophys Acta. 1796:75–90. 2009.PubMed/NCBI

5 

Kalluri R and Weinberg RA: The basics of epithelial-mesenchymal transition. J Clin Invest. 119:1420–1428. 2009. View Article : Google Scholar : PubMed/NCBI

6 

Thiery JP, Acloque H, Huang RY and Nieto MA: Epithelial-mesenchymal transitions in development and disease. Cell. 139:871–890. 2009. View Article : Google Scholar : PubMed/NCBI

7 

Yang J and Weinberg RA: Epithelial-mesenchymal transition: at the crossroads of development and tumor metastasis. Dev Cell. 14:818–829. 2008. View Article : Google Scholar : PubMed/NCBI

8 

Rosivatz E, Becker I, Specht K, et al: Differential expression of the epithelial-mesenchymal transition regulators snail, SIP1, and twist in gastric cancer. Am J Pathol. 161:1881–1891. 2002. View Article : Google Scholar : PubMed/NCBI

9 

Mani SA, Guo W, Liao MJ, et al: The epithelial-mesenchymal transition generates cells with properties of stem cells. Cell. 133:704–715. 2008. View Article : Google Scholar : PubMed/NCBI

10 

Roy S and Majumdar AP: Signaling in colon cancer stem cells. J Mol Signal. 7:112012. View Article : Google Scholar

11 

Reya T, Morrison SJ, Clarke MF and Weissman IL: Stem cell, cancer, and cancer stem cells. Nature. 414:105–111. 2001. View Article : Google Scholar : PubMed/NCBI

12 

Bhat-Nakshatri P, Appaiah H, Ballas C, et al: SLUG/SNAI2 and tumor necrosis factor generate breast cells with CD44+/CD24 phenotype. BMC Cancer. 10:4112010. View Article : Google Scholar : PubMed/NCBI

13 

Giannoni E, Bianchini F, Masieri L, et al: Reciprocal activation of prostate cancer cells and cancer-associated fibroblasts stimulates epithelial-mesenchymal transition and cancer stemness. Cancer Res. 70:6945–6956. 2010. View Article : Google Scholar

14 

Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ and Clarke MF: Prospective identification of tumorigenic breast cancer cells. Proc Natl Acad Sci USA. 100:3983–3988. 2003. View Article : Google Scholar : PubMed/NCBI

15 

Ginestier C, Hur MH, Charafe-Jauffret E, et al: ALDH1 is a marker of normal and malignant human mammary stem cells and a predictor of poor clinical outcome. Cell Stem Cell. 1:555–567. 2007. View Article : Google Scholar : PubMed/NCBI

16 

Huang EH, Hynes MJ, Zhang T, et al: Aldehyde dehydrogenase 1 is a marker for normal and malignant human colonic stem cells (SC) and tracks SC overpopulation during colon tumorigenesis. Cancer Res. 69:3382–3389. 2009. View Article : Google Scholar : PubMed/NCBI

17 

Edge SB, Byrd DR, Compton CC, Fritz AG, Greene FL and Trotti A III: AJCC Cancer Staging Manual. 7th edition. Springer; Chicago, IL: 2010

18 

Qiao Q, Ramadani M, Gansauge S, Gansauge F, Leder G and Beger HG: Reduced membranous and ectopic cytoplasmic expression of β-catenin correlate with cyclin D1 overexpression and poor prognosis in pancreatic cancer. Int J Cancer. 95:194–197. 2001.

19 

Yang MH, Chen CL, Chau GY, et al: Comprehensive analysis of the independent effect of twist and snail in promoting metastasis of hepatocellular carcinoma. Hepatology. 50:1464–1474. 2009. View Article : Google Scholar : PubMed/NCBI

20 

Schwock J, Bradley G, Ho JC, et al: SNAI1 expression and the mesenchymal phenotype: an immunohistochemical study performed on 46 cases of oral squamous cell carcinoma. BMC Clin Pathol. 10:12010. View Article : Google Scholar : PubMed/NCBI

21 

Lee TK, Poon RT, Yuen AP, et al: Twist overexpression correlates with hepatocellular carcinoma metastasis through induction of epithelial-mesenchymal transition. Clin Cancer Res. 12:5369–5376. 2006. View Article : Google Scholar

22 

Valdés-Mora F, Gómez del Pulgar T, Bandrés E, et al: TWIST1 overexpression is associated with nodal invasion and male sex in primary colorectal cancer. Ann Surg Oncol. 16:78–87. 2009.PubMed/NCBI

23 

Gomez I, Peña C, Herrera M, et al: TWIST1 is expressed in colorectal carcinomas and predicts patient survival. PLoS One. 6:e180232011. View Article : Google Scholar

24 

Natalwala A, Spychal R and Tselepis C: Epithelial-mesenchymal transition mediated tumourigenesis in the gastrointestinal tract. World J Gastroenterol. 14:3792–3797. 2008. View Article : Google Scholar : PubMed/NCBI

25 

Nieto MA: The snail superfamily of zinc-finger transcription factors. Nat Rev Mol Cell Biol. 3:155–166. 2002. View Article : Google Scholar : PubMed/NCBI

26 

Shioiri M, Shida T, Koda K, et al: Slug expression is an independent prognostic parameter for poor survival in colorectal carcinoma patients. Br J Cancer. 94:1816–1822. 2006. View Article : Google Scholar : PubMed/NCBI

27 

Bolós V, Peinado H, Pérez-Moreno MA, Fraga MF, Esteller M and Cano A: The transcription factor Slug represses E-cadherin expression and induces epithelial to mesenchymal transitions: a comparison with Snail and E47 repressors. J Cell Sci. 116:499–511. 2003.

28 

Fan F, Samuel S, Evans KW, Lu J, Xia L, Zhou Y, Sceusi E, Tozzi F, Ye XC, Mani SA and Ellis LM: Overexpression of Snail induces epithelial-mesenchymal transition and a cancer stem cell-like phenotype in human colorectal cancer cells. Cancer Med. 1:5–16. 2012. View Article : Google Scholar : PubMed/NCBI

29 

O’Brien CA, Pollett A, Gallinger S and Dick JE: A human colon cancer cell capable of initiating tumour growth in immunodeficient mice. Nature. 445:106–110. 2007.PubMed/NCBI

30 

Shmelkov SV, Butler JM, Hooper AT, et al: CD133 expression is not restricted to stem cells, and both CD133+ and CD133 metastatic colon cancer cells initiate tumors. J Clin Invest. 118:2111–2120. 2008.PubMed/NCBI

31 

Zhou BP, Deng J, Xia W, et al: Dual regulation of Snail by GSK-3β-mediated phosphorylation in control of epithelial-mesenchymal transition. Nat Cell Biol. 6:931–940. 2004.PubMed/NCBI

32 

Smit MA, Geiger TR, Song JY, Gitelman I and Peeper DS: A Twist-Snail axis critical for TrkB-induced epithelial-mesenchymal transition-like transformation, anoikis resistance, and metastasis. Mol Cell Biol. 29:3722–3737. 2009. View Article : Google Scholar : PubMed/NCBI

33 

Yang J, Mani SA, Donaher JL, et al: Twist, a master regulator of morphogenesis, plays an essential role in tumor metastasis. Cell. 117:927–939. 2004. View Article : Google Scholar : PubMed/NCBI

34 

Imammura T, Hikita A and Inoue Y: The roles of the TGF-β signaling in carcinogenesis and breast cancer metastasis. Breast Cancer. 19:118–124. 2012.

35 

Huang D and Du X: Crosstalk between tumor cells and microenviroment via Wnt pathway in colorectal cancer dissemination. World J Gastroenterol. 14:1823–1827. 2008. View Article : Google Scholar : PubMed/NCBI

36 

Kwok WK, Ling MT, Lee TW, et al: Up-regulation of TWIST in prostate cancer and its implication as a therapeutic target. Cancer Res. 65:5153–5162. 2005. View Article : Google Scholar : PubMed/NCBI

37 

Yu L, Li HZ, Lu SM, et al: Alteration in TWIST expression: possible role in paclitaxel-induced apoptosis in human laryngeal carcinoma Hep-2 cell line. Croat Med J. 50:536–542. 2009. View Article : Google Scholar : PubMed/NCBI

38 

Yang AD, Fan F, Camp ER, et al: Chronic oxaliplatin resistance induces epithelial-to-mesenchymal transition in colorectal cancer cell lines. Clin Cancer Res. 12:4147–4153. 2006. View Article : Google Scholar : PubMed/NCBI

39 

Connolly EC, Freimuth J and Akhurst RJ: Complexities of TGF-β targeted cancer therapy. Int J Biol Sci. 8:964–978. 2012.

Related Articles

Journal Cover

2014-March
Volume 31 Issue 3

Print ISSN: 1021-335X
Online ISSN:1791-2431

Sign up for eToc alerts

Recommend to Library

Copy and paste a formatted citation
x
Spandidos Publications style
Kim YH, Kim G, Kwon C, Kim JW, Park PW and Hahm K: TWIST1 and SNAI1 as markers of poor prognosis in human colorectal cancer are associated with the expression of ALDH1 and TGF-β1. Oncol Rep 31: 1380-1388, 2014
APA
Kim, Y.H., Kim, G., Kwon, C., Kim, J.W., Park, P.W., & Hahm, K. (2014). TWIST1 and SNAI1 as markers of poor prognosis in human colorectal cancer are associated with the expression of ALDH1 and TGF-β1. Oncology Reports, 31, 1380-1388. https://doi.org/10.3892/or.2014.2970
MLA
Kim, Y. H., Kim, G., Kwon, C., Kim, J. W., Park, P. W., Hahm, K."TWIST1 and SNAI1 as markers of poor prognosis in human colorectal cancer are associated with the expression of ALDH1 and TGF-β1". Oncology Reports 31.3 (2014): 1380-1388.
Chicago
Kim, Y. H., Kim, G., Kwon, C., Kim, J. W., Park, P. W., Hahm, K."TWIST1 and SNAI1 as markers of poor prognosis in human colorectal cancer are associated with the expression of ALDH1 and TGF-β1". Oncology Reports 31, no. 3 (2014): 1380-1388. https://doi.org/10.3892/or.2014.2970