SEARCH

SEARCH BY CITATION

Keywords:

  • CD151;
  • intrahepatic cholangiocarcinoma;
  • c-Met;
  • prognosis;
  • biomarker

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

BACKGROUND:

Previous studies have indicated that CD151, a hydrophobic protein, forms a functional complex with the proto-oncogene that encodes an N-methyl-N′-nitro-N-nitroso-guanidine (MET) protein (c-Met), and CD151 overexpression reportedly is involved in metastasis/invasion of several tumors. The objective of the current study was to investigate the expression and role of CD151 and/or c-Met in intrahepatic cholangiocarcinoma (ICC).

METHODS:

Sixty ICC tissues with matched nontumorous tissues and 20 normal liver tissues were used to analyze CD151 expression at the level of messenger RNA (mRNA) and protein. Then, the expression of CD151 in an ICC cell line was interrupted using a specific lentiviral-mediated small hairpin RNA (shRNA)-CD151, and the role of CD151 in the proliferation, metastasis, and invasion of ICC cells was assessed. The expression of CD151/c-Met was examined further by immunohistochemistry in a tissue microarray (TMA) that included 140 samples of ICC, and the prognostic role of CD151 and/or c-Met in ICC was evaluated in Kaplan-Meier and Cox regression analyses.

RESULTS:

The expression of CD151 in ICC tissues was much higher than that in nontumorous samples and normal liver; and, after the down-regulation of CD151, HCCC-9810 cells had decreased capability for metastasis/invasion in vitro. CD151 overexpression was correlated significantly with larger tumors, poor differentiation, multiple nodular, microvascular/bile duct invasion, and lymphatic metastasis (P<.05). The postoperative 2-year and 5-year overall survival (OS) rates for patients with low CD151 expression (<50% tumor staining) and/or low c-Met expression (<20% tumor staining) were higher than the rates for patients with high CD151 expression (≥50% tumor staining) and/or high c-Met expression (≥20% tumor staining). Multivariate analysis revealed that CD151 overexpression and c-Met overexpression were independent prognostic markers for ICC.

CONCLUSIONS:

Overexpression of CD151 was implicated in metastasis/invasion of ICC, and both CD151 overexpression and c-Met overexpression may be potential molecular therapeutic targets for ICC. Cancer 2010. © 2010 American Cancer Society.

Intrahepatic cholangiocarcinoma (ICC) is the second most common primary liver cancer (PLC) and accounts for 4% to 6% of PLCs.1, 2 The incidence of ICC has experienced a marked increase in recent decades.3 Early detection of ICC because of advances in diagnostic modalities and clinical screening has made it possible to perform curative resection. However, the long-term survival of patients with ICC remains unsatisfactory because of high recurrence rates and early metastasis. Hence, a better understanding of the molecular mechanisms associated with progression of ICC would be beneficial for the development of effective therapeutic schemes.

Transmembrane 4 superfamily (TM4SF) proteins (also known as tetraspanins) are a family of hydrophobic proteins of approximately 25 kilodaltons (kD) to 50 kD with 4 transmembrane domains, 2 extracellular loops, and 2 short cytoplasmic tails, which are present in several cell types, including endothelial and epithelial cells, smooth muscle, cardiac muscle, and lymphocytes.4, 5 TM4SF proteins are involved in a multitude of biologic processes, such as fertilization, parasite and viral infection, synaptic contacts at neuromuscular junctions, platelet aggregation, maintenance of skin integrity, immune response induction, and tumor metastasis.4-6 The most prominent biochemical property of tetraspanins is that they form complexes by interacting with other tetraspanins and/or with a variety of transmembrane proteins, such as integrins and growth factor receptors, which are requisite for their functions.7-9 The hydrophobic protein CD151, as the core of the tetraspanins complexes (also called tetraspanin-enriched microdomains [TEMs]), has been studied extensively, especially in connection with the progression and prognosis of malignant tumors, including breast cancer, colon cancer, prostate cancer, and hepatocellular carcinoma (HCC).9-11 Recent studies have demonstrated that CD151 forms structural and functional associations with the proto-oncogene that encodes an N-methyl-N′-nitro-N-nitroso-guanidine (MET) protein (c-Met) and is involved in the regulation of downstream pathways of the c-Met/hepatocyte growth factor (HGF) system.12 Given its wide range of functions, especially in invasiveness and metastasis, it may have significance for investigating the expression and role of CD151 and/or c-Met in ICC.

The objectives of the current study were to analyze the expression of CD151 in ICC tissues, matched nontumorous tissues, and normal liver tissues at the level of both protein and messenger RNA (mRNA); to assess the role of CD151 in the proliferation, metastasis, and invasion of an ICC cell line by down-regulating the expression of CD151 with a specific lentiviral-mediated small hairpin RNA (shRNA)-CD151; and to investigate correlations between the expression of CD151 and/or c-Met, clinicopathologic parameters, and survival in patients with ICC.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

Cell Line

A human intrahepatic cholangiocarcinoma cell line, HCCC-9810 (purchased from the Chinese Academy of Sciences Shanghai Branch Cell Bank, Shanghai, China), was maintained in RPMI-1640 supplemented with 10% fetal bovine serum at 37°C in a humidified incubator under 5% CO2 conditions.

Patients and Follow-Up

Fresh tumor samples that were taken from areas adjacent to the tumor margins were obtained from 140 consecutive patients with ICC who underwent curative resection between February 1999 and November 2006 at the Liver Cancer Institute, Zhongshan Hospital, Fudan University. Curative resection was defined as complete resection of tumor nodules, with the tumor margins rendered free of cancer on histologic examination, and resection of the regional lymph nodes, including the hilar, hepatoduodenal ligament lymph nodes and the caval lymph nodes, with no cancerous thrombus in the portal vein (main trunk or 2 major branches), hepatic veins, or bile duct. Patients with ICC who had lymph node involvement beyond these lymph nodes were categorized with distant metastasis and were excluded from the current study.2 The histopathologic diagnosis was based on World Health Organization criteria.13 Liver function was assessed by the Child-Pugh scoring system. The clinical classification of tumors was assigned according to the sixth edition of the TNM classification system published by the International Union Against Cancer. The histologic grade of tumor differentiation was determined according to the classification proposed by Edmondson and Steiner.14 Twenty normal liver tissues were collected from healthy living donors in Zhongshan Hospital. Ethical approval was obtained from the Zhongshan Hospital Research Ethics Committee, and informed consent was obtained from each patient. All patients with ICC in the study cohort had poor encapsulation, and their detailed clinicopathologic characteristics are listed in Table 1. Follow-up data were summarized as of February 2009, and the median follow-up was 25 months (range, 4-120 months). The follow-up procedures were described in detail in our previous report.9

Table 1. Correlations Between CD151 or c-Met and Clinicopathologic Features in 140 Patients With Intrahepatic Cholangiocarcinoma
VariableNo. of PatientsNo. of Patients
CD151HighCD151LowPac-MetHighc-MetLowPa
  • CD151 indicates cluster of differentiation 151 (a hydrophobic protein); CD151High, ≥50% tumor staining; CD151Low, <50% tumor staining; c-Met, a proto-oncogene that encodes the MET protein; c-MetHigh, ≥20% tumor staining; c-MetLow,<20% tumor staining; HBsAg, hepatitis B surface antigen; CA 19-9, carbohydrate antigen 19-9; ALT, alanine aminotransferase; AFP α-fetoprotein.

  • a

    Chi-square test.

  • b

    Fisher exact test.

Age, y
 ≥534030.3972446.429
 <533535 2248 
Sex
 Men3623.1322039.482
 Women3942 2655 
HBsAg
 Positive4740.8912958.515
 Negative2825 1736 
Cirrhosis
 Yes3225.6141443.060
 No4340 3251 
Serum CA 19-9, ng/mL
 ≥374639.8722461.104
 <372926 2233 
Serum ALT, U/L
 ≥75118.684316.070
 <756457 4378 
Child-Pugh score
 A7163.511b4589.356
 B42 15 
Serum AFP, ng/mL
 <206557.8574082.579
 ≥20108 612 
Greatest tumor dimension, cm
 ≤56445.0224069.052
 >51120 625 
Tumor differentiation
 Grade 3 and 44325.0263038.005
 Grade 1 and 23240 1656 
No. of tumors
 Multiple101.010b47.517
 Single6564 4287 
Microvascular/bile duct invasion
 Yes176.032716.497
 No5859 3978 
Lymphatic metastasis
 Yes2410.0221717.014
 No5155 2977 

Transfection of Lentiviral Vectors With shRNA for CD151

The lentiviral-mediated pGCSIL-GFP-shRNA-CD151 was constructed (Shanghai Genechem Company Ltd., Shanghai, China). In the current study, we constructed 3 shRNA-CD151 vectors (pGCSIL-GFP-shRNA-CD151) to silence the expression of CD151 in HCCC-9810 cells (shRNA-CD151-HCCC-9810). The 3 following shRNA targeting sequences for CD151 were used: 1) 5′-CATG TGGCACCGTTTGCCT-3′, 2) 5′-TACCTGCTGTTT ACCTACA-3′, and 3) 5′-CATACAGGTGCTCAATA AA-3′. Stable transfectant clones were characterized by quantitative real-time polymerase chain reaction (qRT-PCR), and immunoblotting was used to determine their expression levels of CD151 protein.

3-(4,5-Dimethylthiazol-2-yl)-2,5- Diphenyltetrazolium Bromide Assay, Cell Migration, and Matrigel Invasion Assays

Cells were aliquoted into a 96-well plate (2000 cells per 200 μL per well) and were incubated for 24 hours, 48 hours, and 72 hours. Twenty microliters of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide solution were added at the indicated time points, and the aliquots were incubated for 4 hours. One hundred fifty microliters of dimethyl sulfoxide were replaced with 200 μL Dulbecco modified Eagle medium (DMEM) that contained 10% fetal bovine serum (FBS), and the aliquots were shaken for 10 minutes. Absorbance at 490 nm was measured to determine the number of viable cells in each well. All experiments were performed in triplicate.

A wound-healing assay was used to evaluate the ability of cell migration. Cells grew to 80% to 90% confluence in 24-well plates, and a wound was made by dragging a plastic pipette tip across the cell surface. The remaining cells were washed 3 times to remove cell debris and incubated at 37°C with serum-free medium. At the indicated times, migrating cells at the wound front were photographed and compared. Three separate experiments were performed.

Cell invasion assays were performed using 24-well transwells (8 μm pore size; Minipore) precoated with Matrigel (Falcon354480; BD Biosciences, Franklin Lakes, NJ). In total, 1 × 105 cells were suspended in 500 μL DMEM that contained 1% FBS and were added to the upper chamber, and 750 μL DMEM that contained 10% FBS were placed in the lower chamber. After 48 hours of incubation, Matrigel and the cells remaining in the upper chamber were removed by cotton swabs. Cells on the lower surface of the membrane were fixed in 4% paraformaldehyde and stained with Giemsa. Cells in 5 microscopic fields (at ×200 magnification) were counted and photographed. All experiments were performed in triplicate.

RNA Extraction and qRT-PCR

Sixty ICC samples and matched nontumorous tissues and 20 normal liver tissue samples were analyzed by qRT-PCR, as described previously.9 Amplification and detection were performed using the ABI PRISM 7900 Sequence Detection System (Applied Biosystems, Foster City, Calif) starting with 1 μL combinational DNA and SYBR Green Realtime PCR Master Mix (Toyobo, Japan). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as an internal standard. The primers for CD151 were 5′-ACTTCATCCTGCTCCTCATCAT-3′ and 5′-TCCGTGTTCAGCTGCTGGT A-3′; and the primers for GAPDH were 5′-GGCATCCTGGGCTACACT GA-3′ and 5′-GTGGTCGTTGAG GGCAATG-3′. The qRT-PCR conditions were as follows: 10 minutes at 94°C, 40 cycles of denaturation at 94°C for 20 seconds, annealing at 59°C for 30 seconds, and extension at 72°C for 60 seconds. The relative expression of CD151 was analyzed by using the comparative cycle threshold (Ct) method according to the equation: 2−ΔCt(ΔCt = Ct [CD15]) − Ct [GAPDH]). All experiments were done in triplicate.

Immunoblot Analysis

Thirty micrograms of proteins extracted from 60 cases of ICC samples and matched nontumorous samples that were selected randomly from the same cohort and 20 normal liver tissues underwent immunoblotting as described previously.15 Monoclonal mouse antihuman CD151 (11G5a; 1:200 dilution; Serotec, Oxford, United Kingdom) was used to detect the expression of CD151. GAPDH (1:5000 dilution; Chemicon, Temecula, Calif) was used as an internal control. All experiments were performed in triplicate.

Construction of Tissue Microarrays and Immunohistochemistry

Tissue microarrays were constructed as described in our earlier study.9 Briefly, all ICC samples were reviewed histologically by hematoxylin and eosin staining, and representative areas away from necrotic and hemorrhagic materials were premarked in the paraffin blocks. Duplicate 1-mm-diameter punches from 2 different areas, from the tumor center and from the nearest noncancerous margin (designated as intratumor and peritumor, respectively; total, 4 punches), were included from each case, along with different controls, to ensure reproducibility and homogenous staining of the slides (Shanghai Biochip Company Ltd., Shanghai, China). Thus, 4 different tissue microarray blocks were constructed, each containing 140 cylinders. Sections (4 μm thick) were placed on slides that were coated with 3-aminopropyltriethoxysilane.

Monoclonal mouse antihuman CD151 (11G5a; 1:200 dilution; Serotec) and rabbit antihuman c-Met antibody (EP1454Y, 1:100; Abcam Plc., Cambridge, United Kingdom) were used to detect the expression of CD151 and c-Met, respectively. In each section, staining was captured by the Leica QWin Plus version 3 software (Leica Microsystems, Wetzlar, Germany). The intensity of positive staining was measured as described previously.9 The intensity of CD151 and c-Met staining was classified into 2 expression levels (the mean area of positive staining was used as the cutoff value): high CD151 expression (CD151high) indicated ≥50% CD151 staining, and c-Methigh indicated ≥20% c-Met staining; whereas CD151low indicated <50% CD151 staining, and c-Metlow indicated <20% c-Met staining)

Statistical Analysis

Statistical analyses were performed with the SPSS software package (version 12.0; SPSS Inc., Chicago, Ill). Values were expressed as the mean ± standard deviation. For immunohistochemical marker, the cutoff for defining subgroups was the median value. The chi-square test, the Fisher exact probability test, and the Student t test were used for comparison between groups. Overall survival (OS) and the time to recurrence were defined as described previously.16 The correlation between the expression of CD151 and c-Met was analyzed using Spearman rank correlation. Prognostic significance was assessed using Kaplan-Meier survival estimates and log-rank tests. A Cox proportional hazards regression model was used to analyze independent prognostic. All tests were 2-tailed, and P values <.05 were considered statistically significant.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

CD151 Was Overexpressed in ICC

The expression of CD151 was analyzed by qRT-PCR and immunoblotting in ICC tissues, matched nontumorous tissues, and normal liver tissues. Very low levels of CD151 expression were detected in nontumorous tissues and normal liver tissues compared with the levels detected in ICC tissues. The mean ΔCt values of CD151 expression were 3.73 ± 0.80 in ICC samples (range, 1.8-5.0), 1.81 ± 0.64 in nontumorous samples (range, 1.0-3.3), and 1.74 ± 0.63 in normal liver samples (range, 0.9-2.8)(Fig. 1A). There was a statistically significant difference in CD151 expression between ICC samples and nontumorous samples (P < .01). Moreover, the expression of CD151 mRNA varied greatly in tumor samples (range, 1.8-5.0). Of 60 ICC samples, we observed that the expression of CD151 mRNA was higher in 40 ICC samples from patients who had recurrent disease (>2 years after curative resection) than that in ICC samples from patients who did not have recurrent disease (20 tumor samples were from patients without recurrence), and the difference was significant (4.13 ± 0.45 vs 2.83 ± 0.35; P < .01)(Fig. 1B). When a combination of immunoblotting and semiquantitative analysis was used for analysis, the expression of CD151 protein in ICC samples, nontumorous samples, and normal liver samples was consistent with that observed with qRT-PCR analysis (ICC samples vs nontumorous samples vs normal liver tissues: 2.96 ± 0.90 vs 0.99 ± 0.37 vs 0.89 ± 0.56, respectively; P < .01; recurrent ICC group vs nonrecurrent group: 3.38 ± 0.77 vs 2.14 ± 0.52; P < .01)(Fig. 1C-E), indicating that CD151 was overexpressed in ICC and that its expression probably was related positively to recurrent ICC. A correlation analysis indicated that the expression of CD151 in 60 ICC samples using Western blot analysis was consistent with that observed on immunohistochemistry (r = 0.835; P < .001) Table 2.

thumbnail image

Figure 1. Expression of the hydrophobic protein CD151 is illustrated in intrahepatic cholangiocarcinoma (ICC) tissues, adjacent nontumorous tissues, and normal liver tissues. (A) Quantitative real-time polymerase chain reaction analysis demonstrated that CD151 messenger RNA (mRNA) expression in ICC tissues was greater than that in adjacent nontumorous and normal liver tissues, and the difference was statistically significant (P < .01).(B) The expression of CD151 mRNA was greater in recurrent ICC tissues than in nonrecurrent ICC tissues (P < .01). (C) Western blot analysis demonstrated CD151 protein expression in ICC tissues, adjacent nontumorous tissues, and normal liver tissues. (D) Semiquantitative analysis demonstrated that the expression of CD151 protein in ICC tissues was greater than that in adjacent nontumorous and normal liver tissues (P < .01).(E) CD151 protein in recurrent ICC tissues was greater than that in nonrecurrent ICC tissues (P < .01). GAPDH indicates glyceraldehyde 3-phosphate dehydrogenase.

Download figure to PowerPoint

Table 2. Correlation of CD151 Expression Detected by Western Blot Analysis and Immunohistochemistry in 60 Samples of Intrahepatic Cholangiocarcinoma
IHC of Total Area, %WB IntensityrP
  1. IHC indicates immunohistochemistry; WB, Western blot analysis; r, correlation coefficient.

54.2 ± 14.52.96 ± 0.900.835<.001

CD151 Promoted Invasion and Metastasis of ICC In Vitro

To examine the role of CD151 in ICC cells, HCCC-9810 was transfected successfully with pGCSIL-GFP-shRNA-CD151 (Fig. 2A,B). Of 3 shRNA-CD151s, the third was validated for the most efficient interference of CD151 by qRT-PCR (Fig. 2C) and immunoblotting (Fig. 2D) and was chosen for further study. Matrigel invasion assays revealed that decreased CD151 expression was accompanied by impairment in the invasiveness of HCCC-9810 cells (Fig. 2E). A wound-healing assay revealed an evident delay in the wound closure rate of shRNA-CD151-HCCC-9810 cells at 24 hours compared with mock HCCC-9810 cells (Fig. 2F). However, the down-regulation of CD151 exerted little influence on cell proliferation (P > .05)(Fig. 2G).

thumbnail image

Figure 2. Functional analysis is illustrated after transfection with small-hairpin RNA (shRNA)-CD151 in human intrahepatic cholangiocarcinoma (ICC) HCCC-9810 cells in vitro. (A,B) HCCC-9810 cells were transfected with lentiviral-mediated pGCSIL-GFP-shRNA-CD151. (C,D) The third shRNA targeting sequence was validated for the most efficient interference of CD151 by quantitative real-time polymerase chain reaction and immunoblotting. GAPDH indicates glyceraldehyde 3-phosphate dehydrogenase. (E) A transwell assay showed that the numbers of invading HCCC-9810-Mock cells were higher than that in shRNA-CD151-HCCC-9810 cells (original magnification, × 200). (F) HCCC-9810 cells transfected with shRNA-CD151 migrated slowly compared with mock-transfected cells at 24 hours for wound channel closure (original magnification, × 100). (G). Cell proliferation was detected by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay. OD indicates optical density.

Download figure to PowerPoint

Expression of CD151 or c-Met Was Positively Associated With Malignant Phenotypes of ICC by Immunohistochemistry

Weak staining of CD151 and c-Met in hepatocytes and small bile ducts was observed in normal liver samples (Fig. 3A-C) and in nontumorous samples (Fig. 3D-F). Immunoreactivity for CD151 protein was observed in the cell membranes (Fig. 3H,K). Positive staining for c-Met was observed in the cytoplasm and membrane of tumor cells (Fig. 3I,L,O,R,U). In the tumor tissues, CD151 and c-Met expression had considerable heterogeneity in the different samples. Representative samples are provided in Figure 3(CD151high/c-Met high[Fig. 3G-L], CD151high/c-Met low[Fig. 3M-O], CD151low/c-Methigh[Fig. 3P-R], and CD151low/c-Metlow[Fig. 3S-U]).

thumbnail image

Figure 3. The expression and correlation of CD151 and the proto-oncogene that encodes an N-methyl-N′-nitro-N-nitroso-guanidine (MET) protein (c-Met) are illustrated in samples of intrahepatic cholangiocarcinoma (ICC) and in normal liver tissues by immunohistochemistry. (A,D,G,J,M,P,S) Hematoxylin and eosin (H&E) staining is shown. (B,C,E,F) Weak staining of for CD151 (<50%) and c-Met protein are observed in hepatocytes and in small bile ducts of normal liver and peritumoral tissue, respectively. (H,I,K,L) High CD151 expression and high c-Met expression (≥20% staining) are shown. (N,O) High CD151 expression and low c-Met expression (<20%) are shown. (Q,R) Low CD151 expression (<50%) and high c-Met expression are shown. (T,U) Low CD151 expression and low c-Met expression are shown (original magnification, ×400 in J-L; ×100 in D-I and M-U). Scale bars = 50 μm.

Download figure to PowerPoint

CD151high accounted for 53.6% of tumors from patients with ICC (75 of 140 patients). CD151high expression was correlated significantly with microvascular/bile duct invasion (P = .032), lymphatic metastasis (P = .022), multiple tumors (P = .010), large tumors (P = .022), and poor differentiation (P = .026)(see Table 1). However, other clinical characteristics, including age, sex, hepatitis B surface antigen status, the presence or absence of cirrhosis, preoperative serum α-fetoprotein level, preoperative serum carbohydrate antigen 19-9 level, and Child-Pugh score, were not related significantly to the expression of CD151.

Forty-six of 140 ICC tissues expressed high levels of c-Met (32.9%). Consistent with a previous report,17 patients with high c-Met expression were more likely to exhibit aggressive disease features. Patients with high c-Met expression harbored more lymphatic metastasis (P = .014) and dedifferentiation (P = .005) than patients with low c-Met expression (Table 1). A correlation analysis revealed that the expression of CD151 was correlated modestly with the expression of c-Met (r = 0.410; P = .018).

Overexpression of CD151 and/or c-Met Were Independent Parameters for Predicting Prognosis in Patients With ICC

Up to the last follow-up, 108 patients had recurrent tumors, and 109 patients had died, including 7 patients who died of liver failure without evidence of disease recurrence. The 2-year and 5-year OS rates for the whole population were 35.7%, and 70%, respectively, and the 2-year and 5-year cumulative recurrence rates were 23.6% and 77.9%, respectively. Univariate analysis revealed that large tumors (>5 cm), poor tumor differentiation, multiple tumors, microvascular/bile duct invasion, and lymphatic metastasis were predictors of OS and cumulative recurrence. Other characteristics, including age, sex, and hepatitis background, had no prognostic significance for OS or cumulative recurrence (Table 3). The expression of CD151 was correlated with OS and cumulative recurrence rates (Table 3). The 2-year and 5-year OS rates in the CD151low group were significantly higher than the rates in the CD151high group (47.7% vs 25.3% and 35.4% vs 14.7%, respectively)(Fig. 4A). The 2-year and 5-year cumulative recurrence rates in the CD151low group were significantly lower than the rates in the CD151high group (52.3% vs 85.3% and 63.1% vs 90.7%, respectively)(Fig. 4B). The expression of c-Met also was correlated significantly with the OS and cumulative recurrence rates (Table 3). The postoperative 2-year and 5-year OS rates for patients with ICC were higher in the c-Metlow group than in the c-Methigh group (P < .01), and the corresponding cumulative recurrence rates in patients with ICC in the c-Methigh group were significantly higher than those in the c-Metlow group (P < .01).

thumbnail image

Figure 4. Prognostic significance was assessed by Kaplan-Meier analysis and log-rank tests. (A,B) Patients with intrahepatic cholangiocarcinoma (ICC) who had high CD151 expression (≥50%)(CD151high) had a poorer prognosis in terms of overall survival and cumulative recurrence. c-Met indicates a proto-oncogene that encodes an N-methyl-N′-nitro-N-nitroso-guanidine (MET) protein. (C,D) Patients with ICC who had high c-Met expression (≥20% of tumor section stained)(c-Methigh) had poorer overall survival and a greater risk of recurrence. (E,F) The effects of combined CD151 and c-Met expression on prognosis in patients with ICC are illustrated.

Download figure to PowerPoint

Table 3. Univariate and Multivariate Analyses of Factors Associated With Survival and Recurrence
VariableOSCumulative Recurrence
 Multivariate Analysis Multivariate Analysis
Univariate PHR95%CIPUnivariate PHR95%CIP
  • OS indicates overall survival; HR, hazard ratio; CI, confidence interval, NA, not adopted; HBsAg, hepatitis B surface antigen; ALT, alanine aminotransferase; CA 19-9, carbohydrate antigen 19-9; AFP, α-fetoprotein; NS, not significant; CD151, a hydrophobic protein; c-Met, a proto-oncogene that encodes the MET protein.

  • a

    Group I, CD151low and c-Metlow; Group II, either high; Group III, CD151high and c-Methigh.

Age (<53 y vs ≥53 y).450  NA.308  NA
Sex (men vs women).959  NA.742  NA
HBsAg (negative vs positive).739  NA.366  NA
Cirrhosis (yes vs no).427  NA.652  NA
Child-Pugh score (A vs B).125  NA.073  NA
Serum ALT (<75 U/L vs ≥75 U/L).618  NA.782  NA
Serum CA 19-9(<37 ng/mL vs ≥37 ng/mL).387  NA.314  NA
Serum AFP(<20 ng/mL vs ≥20 ng/mL).241  NA.764  NA
Tumor size (≤5 cm vs >5 cm).025  NS.006  NS
Tumor differentiation (grade 1/2 vs grade 3/4).043  NS.032  NS
No. of tumors (multiple vs single).021  NS.002  NS
Microvascular/bile duct invasion (yes vs no).031  NS.030  NS
Lymphatic metastasis (no vs yes).0070.5890.383-0.904.016.0010.5780.377-0.884.012
CD151 density (<50% vs ≥50%).0010.5950.392-0.903.015<.0010.4850.319-0.738.001
c-Met density (<20% vs >20%).0010.6350.417-0.968.035<.0010.5960.393-0.904.015
Combine of CD151 and c-Met (Group I vs Group II/III)a<.0010.7240.627-0.837<.001<.0010.6690.602-0.813 

Next, we investigated the role of combined CD151 and c-Met expression in survival. We classified patients into 3 subgroups according to their CD151 and c-Met expression: Group I had low expression of both CD151 and c-Met (n = 52), Group II had high expression of either CD151 or c-Met (n = 55), and Group III had high expression of both CD151 and c-Met (n = 33). The 2-year and 5-year OS rates in Group I were significantly higher than those in Groups II and III (55.8% vs 22.7% and 44.2% vs 12.5%, respectively)(Fig. 4E,F). The 2-year and 5-year cumulative recurrence rates in Group I were significantly lower than those in Groups II and III (42.3% vs 86.4% and 55.7% vs 90.9%, respectively). The multivariate Cox proportional hazards model revealed that CD151, c-Met, lymphatic metastasis, and co-overexpression of CD151 and c-Met were independent prognostic indicators for OS and cumulative recurrence (Table 3).

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

The current report describes findings that are highly significant and relevant in the context of ICC. CD151 expression occurred at very low to almost undetectable levels in nontumorous samples and normal liver samples, whereas CD151 expression increased in samples of nonrecurrent ICC and was even higher in samples from the recurrent ICC group. In addition, CD151 and the co-overexpression of CD151 and c-Met have potential new tissue markers of a poor prognosis in patients with ICC and should be assessed in a large number of archival ICC samples.

The initial evidence that CD151 promotes metastasis came from a study in which an antibody with unknown specificity inhibited metastasis formation by a human epidermoid carcinoma line in vivo. The antibody reportedly recognized CD151 and inhibited cell migration without affecting adhesion or proliferation.18 Overexpression of CD151 has been identified in many tumor types. High CD151 expression has been associated with a poor prognosis in breast cancer,11 pancreatic cancer,19 nonsmall cell lung cancer,20 and HCC9; and CD151 reportedly is a better parameter than histologic grading for predicting the prognosis of patients with prostate cancer.21 In the current study, ICC tissues from patients who had early recurrences had higher CD151 expression at both the mRNA and protein levels than ICC tissues from patients without recurrent disease, indicating that the overexpression of CD151 may promote metastasis and recurrence. This possibility was supported further by the finding that aggressive histopathologic characteristics, such as microvascular/bile duct invasion, lymphatic metastasis, multiple tumors, larger tumors, and poor differentiation, were significantly more frequent in patients with high CD151 expression than in patients with low CD151 expression. In the current study, the knockdown of CD151 expression in HCCC-9810 cells impaired the invasion and metastasis of ICC, consistent with previous research.9-11 Therefore, we conclude that the overexpression of CD151 does promote the metastasis and invasion of ICC and may be a prognostic indicator for patients with ICC.

The hydrophobic nature of tetraspanins clearly indicates that they are likely to be associated with one another or with other membrane proteins. Indeed, it is well established that tetraspanins act as a signaling platform in the plasma membrane by the formation of TEMs. To date, different types of membrane proteins (including growth factor receptors; integrins; immunoglobulin domains; and glutamic acid, tryptophan, isoleucine [EWI] motif-containing protein F [a subfamily of immunoglobulin proteins]) have been identified in TEMs.5, 9, 12, 22 Moreover, functions of these membrane proteins reportedly are associated intimately with TEMs.5 For example, a combination of TEMs by different tetraspanins affected the function of growth factor receptors and integrins.5 More important, the expression levels of individual tetraspanins in the TEMs also had a vital effect on the associated proteins.4-6 Recently, CD151 was proposed as the core of many TEMs based on findings that knock-down of CD151 may impair the completion of TEMs and that deletion of CD151 may inhibit the function of many membrane proteins, whereas other tetraspanins do not.4, 5 CD151 blocking markedly impaired the invasiveness and metastatic potential of tumor cells, as reported previously.18 Moreover, targeting the tetraspan protein CD151 or TEMs has become a promising therapeutic strategy.6, 23, 24 Recent studies have demonstrated that CD151 forms structural and functional associations with c-Met and is involved in the regulation of downstream pathways of the c-Met/HGF system. Previous studies indicated that c-Met expression was correlated inversely with the survival of patients with ICC.16 In the current, CD151 expression was identified as an independent predictor of OS and cumulative recurrence, but the predictive ability of combined CD151 and c-Met overexpression was more sensitive than the that for the overexpression of either CD151 or c-Met alone, lending support to the idea that CD151 plays a role in the function of c-Met. Thus, we believe that our current findings are significant and that CD151 may be an important therapeutic target in ICC. In conclusion, CD151 overexpression is implicated in the metastasis and invasion of ICC, and CD151 and/or c-Met overexpression may have potential as molecular therapeutic targets in ICC.

CONFLICT OF INTEREST DISCLOSURES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES

Supported by the 11th 5 Years Key Programs for Science and Technology Development of China (2006BAI02A04), the National Key Sci-Tech Special Project of China (2008ZX10002-022 and 2008ZX10002-025), the National Hi-Tech Research and Development Program of China (2007AA02Z479), the Ph.D. Programs Foundation of Ministry of Education of China (13-192), and the Program for Excellent Disciplinary Leaders of Shanghai Health Bureau (LJ06004).

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. CONFLICT OF INTEREST DISCLOSURES
  7. REFERENCES