Liver-intestine cadherin predicts microvascular invasion and poor prognosis of hepatitis B virus-positive hepatocellular carcinoma

Authors

  • Zhen-Bin Ding,

    1. Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, China
    Search for more papers by this author
    • The first 4 authors contributed equally to this work.

  • Ying-Hong Shi,

    1. Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, China
    Search for more papers by this author
    • The first 4 authors contributed equally to this work.

  • Jian Zhou,

    1. Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, China
    2. Institutes of Biomedical Sciences, Fudan University, Shanghai, China
    Search for more papers by this author
    • The first 4 authors contributed equally to this work.

  • Guo-Ming Shi,

    1. Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, China
    Search for more papers by this author
    • The first 4 authors contributed equally to this work.

  • Ai-Wu Ke,

    1. Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, China
    Search for more papers by this author
  • Shuang-Jian Qiu,

    1. Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, China
    Search for more papers by this author
  • Xiao-Ying Wang,

    1. Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, China
    Search for more papers by this author
  • Zhi Dai,

    1. Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, China
    Search for more papers by this author
  • Yang Xu,

    1. Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, China
    Search for more papers by this author
  • Jia Fan

    Corresponding author
    1. Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, Fudan University, Shanghai, China
    2. Institutes of Biomedical Sciences, Fudan University, Shanghai, China
    • Department of Liver Surgery, Liver Cancer Institute, Zhongshan Hospital, Fudan University, 180 FengLin Road, Shanghai, 200032, China
    Search for more papers by this author
    • Fax: (011) 86-21-64037181


  • Patient samples were obtained after provision of informed consent according to an established protocol approved by the Ethics Committee of Fudan University.

Abstract

BACKGROUND:

Liver-intestine cadherin (LI-cadherin; CDH-17) is a new member of the cadherin superfamily with distinct structural and functional features. The study was designed to investigate the role of LI-cadherin in tumor invasion and prognosis of human hepatitis B virus (HBV)-positive hepatocellular carcinoma (HCC).

METHODS:

LI-cadherin expression in HBV-positive hepatocellular carcinoma cell lines with low- and high-invasive potentials was evaluated by Western-blot, immunofluorescence, and real-time polymerase chain reaction (PCR) analyses. The role of LI-cadherin in tumor invasion was also evaluated in vitro by a small-interfering ribonucleic acid (siRNA)-mediated approach. The prognostic significance of LI-cadherin was validated in a cohort of HBV-positive HCC patients by immunohistochemistry and Western-blot.

RESULTS:

Significant high levels of LI-cadherin mRNA and protein were found in the high-invasive HCCLM3 as compared with those in low-invasive PLC/PRF/5 and Hep3B cell line. Cell migration, adhesion to extracellular matrix, and matrigel invasion were significantly reduced after LI-cadherin knockdown in HCCLM3 cells. Immunohistochemical analysis of 255 HBV-positive HCC cases showed that overexpression of LI-cadherin was well correlated with microvascular invasion, which was confirmed by Western-blot in 32 tumor tissues, and its overexpression was strongly associated with shorter overall survival as well as higher incidence of tumor recurrence.

CONCLUSIONS:

LI-cadherin is predictive of microvascular invasion and poor prognosis of HBV-positive HCC, and would be a potential useful intervention target for HCC. Cancer 2009. © 2009 American Cancer Society.

Hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC) is 1 of the most common malignancies in the world, especially in China.1 Despite remarkable advances in diagnostic and therapeutic techniques, the prognosis for HBV-related HCC remains poor.2 The high mortality rate can be at least partly explained by the finding that most of these cancers show vascular invasion and intrahepatic metastasis by the time symptoms develop.3 Hence, investigating the mechanism of HCC spread is of great importance. A variety of data indicate that alterations in the adhesion properties of neoplastic cells have an important role in the metastasis of cancer.4-6 Multiple and diverse adhesion molecules have been found to be involved in intercellular and cell-extracellular matrix (ECM) interactions in cancer.7, 8

Cadherins are transmembrane Ca2+-dependent homophilic adhesion receptors that have important roles in cell recognition and cell sorting during development.9 Classical cadherin genes like epithelial (E)-cadherin are considered tumor suppressors, and defects in their expression or function have been associated with tumor progression.6, 10 In highly invasive breast tumors, another important cadherin (neuronal [N]-cadherin) was shown to be involved in cell-cell contacts.11, 12 It has been proposed that N-cadherin–mediated carcinoma cell interactions with mammary stromal cells are involved in the promotion of breast cancer metastasis by facilitating carcinoma cell migration through the mammary stroma and reestablishing homophilic cell-cell adhesion in metastasis.13 Moreover, as a feature of aggressive tumors, epithelial to mesenchymal transition is characterized by reduced E-cadherin and increased N-cadherin expression levels, which contribute to a stroma-oriented cellular adhesion profile with increased tumor cell migration and invasion; this molecular profile has recently been reported in several tumors.14-16

Liver-intestine cadherin (LI-cadherin; CDH-17) is a new member of the cadherin superfamily with distinct structural and functional features. The exact adhesive function of LI-cadherin is unknown, but it is complementary to the coexpressed classical cadherins such as E-cadherin.17 The expression of LI-cadherin has been reported in colorectal carcinoma,18, 19 gastric adenocarcinoma,20, 21 pancreatic ductal adenocarcinoma,22 and hepatocellular carcinoma.23, 24 Interestingly, in human colorectal carcinoma, reduced expression of LI-cadherin is closely associated with tumor progression and lymph-node metastasis,19 but the opposite is true in gastric cancer.21, 25 Furthermore, the LI-cadherin haplotype is associated with increased risk of HCC,23 and messanger ribonucleic acid (mRNA) splicing of LI-cadherin results in poor prognosis.24 In another study on side population cells at our institute,26 increased expression of LI-cadherin was found in more invasive HCC cells. Therefore, LI-cadherin might have different roles in different tumors, and the mechanism of LI-cadherin in tumor promotion or suppression remains to be further determined.

In the present study, we found HCC cell lines with different invasive potentials have different LI-cadherin expression levels. By using specific RNA interference, the role of LI-cadherin in tumor-cell invasiveness was assessed. A cohort of HCC patients underwent curative resection at our surgical center, and the role of LI-cadherin was further validated. Taken together, these data indicate that LI-cadherin is a potential marker for tumor invasion and may have an important role in tumor microvascular metastasis, ultimately resulting in poor prognosis for HBV-positive HCC patients.

MATERIALS AND METHODS

Patient Samples

Patient samples were obtained after provision of informed consent according to an established protocol approved by the Ethics Committee of Fudan University. Data do not contain any information that could allow identification of the patients.

Tumor specimens used in tissue microarrays (TMA) studies were obtained from 255 consecutive HBV-positive patients with HCC who underwent curative resection without preoperative treatment at the Liver Cancer Institute, Zhongshan Hospital, Fudan University, between 1997 and 2000. Conventional clinicopathologic parameters were recorded and detailed in Table 1. Tumor stage was determined according to the 2002 AJCC/UICC tumor, node, metastasis (TNM) classification system. Tumor differentiation was graded by the Edmondson grading system. All the cases have complete follow-up data and the diagnosis of HCC was confirmed by pathological examination. Data were obtained at last follow-up for patients without relapse or death. Disease-free survival time (DFS; ie, time to recurrence) was defined as the time period from the date of surgery to confirmed tumor relapse date for relapsed patients or from the date of surgery to the date of last follow-up for nonrecurrent patients. Overall survival (OS) time was defined as the time period from the date of surgery to the confirmed death date for dead patients or from the date of surgery to the date of last follow-up for surviving patients.

Table 1. Correlation Between Liver-Intestine Cadherin Expression and Clinicopathological Characteristics
VariableCategoryLiver-Intestine Cadherin ExpressionP
Positive (n=121)Negative (n=134)
  1. TNM indicates tumor, node, metastasis.

SexFemale2218 
 Male99116.298
Age, y≤527272 
 >524962.353
Preoperative alpha-fetoprotein (ng/mL)≤204370 
 >207864.007
CirrhosisNo315 
 Yes118119.007
Tumor size, cm≤56177 
 >56057.259
No. of tumorsSingle102114 
 Multiple1920.863
Vascular invasionNo99123 
 Yes2211.018
TNM stageI83104 
 II/III3830.104
Tumor differentiationI-II3837 
 III-IV8397.507

By using a random number table, samples for Western-blot analysis were randomly collected from the patients undergoing curative resection at the Liver Cancer Institute, Zhongshan Hospital, Fudan University, in 2007, which were collected immediately after resection, transported in liquid nitrogen and stored at −80°C until use.

Cell Lines

The human HCC cell lines Hep3B, PLC/PRF/5, and HCCLM3 (HCC cell lines with low- and high-invasion and that are positive for HBV DNA)27 were routinely maintained in high-glucose MEM or DMEM supplemented with 10% heat-inactivated fetal bovine serum, 100 U/mL penicillin and 100 mg/mL streptomycin. All cell lines were cultured at 37°C in a humidified incubator under 5% CO2.

Real-Time Polymerase Chain Reaction

Total RNA was extracted from cell lines using Trizol Reagent (Invitrogen, Carlsbad, Calif). Total RNA (2 μg) was reverse transcribed using a RevertAid first-strand cDNA synthesis kit (Fermentas, Burlington, ON, Canada). Reverse-transcription polymerase chain reaction (PCR) was performed before quantitative real-time PCR. LI-cadherin mRNA expression levels were determined by real-time PCR using SYBR Premix Ex Taq (TaKaRa, Dalian, China). PCR amplification cycles were programmed for 10 seconds at 95°C, followed by 40 cycles of 5 seconds at 95°C and 30 seconds at 60°C. Data were collected after each annealing step. Beta-actin was used as an endogenous control to normalize for differences in the amount of total RNA in each sample. Relative expression levels of genes were calculated and expressed as 2-ΔCt, as previously described.28 The following primers were used: Beta-actin 5′-CAACTGGGACGACATGGAGAAAAT-3′ and 5′-CCAGAGGCGTACAGGGATAGCAC-3′ (Sequence location: No.NM_001101 base304-510); LI-cadherin 5′-GGCCAATCCTCCTGCTGTGAC-3′ and 5′-AGATGGCTCCCGTTTTGTTGTTG-3′ (Sequence location: No. NM_004063 base276-688); N-cadherin 5′-TGAAACGCCGGGATAAAGAACG-3′ and 5′-TGCTGCAGCTGGCTCAAGTCAT-3′ (Sequence location: No. NM_001792 base2442-2580).

Western-Blot Analysis

Western-blot analysis was performed as previously described.29 Briefly, total cell lysates were generated and proteins were separated by standard SDS-PAGE, followed by transfer to polyvinylidene difluoride membranes. The membranes were washed and blocked before the specific primary anti-human LI-cadherin (1:250; R&D, Minneapolis, Minn) or GAPDH (1:5000; Millipore, Bedford, Mass) antibodies were applied, followed by the horseradish-peroxidase–conjugated secondary antibodies. The reactions were detected by enhanced chemiluminescence assay. In the semiquantitative analysis, gray values of bands in the images were quantified by densitometry (Bio-electrophoresis image analysis system Version 6.00.001). Relative potein expression of LI-cadherin equals to the ratio between gray values of LI-cadherin and GAPDH.

Immunofluorescence Assay

Briefly, cells cultured on glass slides were fixed by 4% paraformaldehyde for 10 minutes. Subsequently, the cells were permeabilized with 0.1% Triton X-100 for 15 minutes at room temperature, washed with phosphate-buffered saline (PBS), and blocked with PBS containing 0.5% (w/v) bovine serum albumin (BSA) and 0.15% (w/v) glycine (BSA buffer) for 1 hour at room temperature. Cells were treated with anti–LI-cadherin (1:50 dilution in BSA buffer) antibody for 1 hour at room temperature. Cells were then washed with BSA buffer and incubated with 2 ug/mL Alexa Fluor 488-conjugated goat antimouse antibody (Molecular Probes, Eugene, Ore) for 1 hour at room temperature. After rinsing in PBS, the slices were counterstained with DAPI (Diamidino phenylindole) and examined by fluorescence microscopy (Leica Microsystems Imaging Solutions, Cambridge, UK).

siRNA-Mediated LI-Cadherin Silencing

Two different sequences targeted to 2 different sites of LI-cadherin mRNA (GeneBank Accession No.NM_004063, base 417-435 and base 1649-1667) were designed without off-target effects. The sense and anti-sense strands of siRNAs were: LI-cadherin (sequence 1), 5′-GGACGCUAAUGGAAUUAUATT-3′ (sense), 5′-UAUAAUUCCAUUAGCGUCCAG-3′ (antisense); LI-cadherin (sequence 2), 5′-CCGGAUAUGUCAUAAUUAATT-3′ (sense), 5′-UUAAUUAUGACAUAUCCGGTG-3′ (anti-sense). For transfection of adherent HCCLM3 cells, LI-cadherin siRNAs and a negative-control mismatch sequence were transfected using Lipofectamine 2000 (Invitrogen). After 72 hours of transfection, cells were lysed and protein (20 μg) was assayed by Western-blot analysis.

Cell Migration, Adhesion, and Invasion Assay

Cell migration was evaluated using the scratch wound assay. Cells were cultured for 2 days to form a tight cell monolayer and then serum starved for 16 hours. After the serum starvation, the cell monolayer was wounded with a 10 μL plastic pipette tip. The remaining cells were washed twice with culture medium to remove cell debris and incubated at 37°C with normal serum-containing culture medium. At the indicated times, migrating cells at the wound front were photographed using an inverted microscope (Leica). A percentage of the cleared area at each time point compared with time zero was measured using Image-Pro Plus v6.2 software.

Fibronectin or BSA-coated 96-well plates were obtained from Chemicon and adhesion assays were performed following the manufacturer's instructions. Briefly, cells were detached using a cell-dissociation buffer, washed, and plated at a density of 1 × 104 cells per well in culture medium. Adhesion was allowed to occur for 1 hour. Nonadherent cells were removed by gentle washing with warm PBS. Cells were fixed, stained, and solubilized, and absorbance at 562 nm was measured using a BioRad 3550 microplate reader.

Matrigel invasion assays were performed using a Transwell. Filters coated with matrigel (BD Bioscience, Bedford, Mass) in the upper compartment were loaded with 100 μL containing 1 × 105 cells, and the lower compartment was filled with conditioned culture medium mixed with DMEM and supplemented with 10% FBS, NIH3T3, and HCCLM3 super supplements. After 36 hours, migrated cells on the bottom surface were fixed with 4% paraformaldehyde and counted after staining with Giemsa stain.

TMA and Immunohistochemistry

TMA were constructed as previously described.30 Briefly, all the HCC tissues were reviewed by 2 histopathologists, and representative areas free from necrotic and hemorrhagic materials were premarked in the paraffin blocks. Two core biopsies of 1-mm diameter were taken from the donor blocks and transferred to the recipient paraffin block at defined array positions. Three different TMA blocks were constructed. Consecutive sections of 4 μm thickness were taken on 3-aminopropyltriethoxysilane–coated slides (Shanghai Biochip, Shanghai, China).

Monoclonal mouse antibodies against human LI-cadherin (1:20, R&D) were used. Immunohistochemistry analyses were performed using a 2-step protocol (Novolink Polymer Detection System, Novocastra, Newcastle, UK) as previously described.30 Briefly, after microwave antigen retrieval, tissues were incubated with primary antibodies for 60 minutes at room temperature. After 30-minute incubation, with secondary antibody (RE7112, Novolink Polymer), the sections were developed in diaminobenzidine solution under microscopic observation and counterstained with hematoxylin. Negative control slides with the primary antibodies omitted were included in all assays.

Evaluation of Immunohistochemical Variables

Immunohistochemical staining results were assessed by 3 independent pathologists without knowledge of patient characteristics. The intensity and percentage of positively stained cytoplasm in the whole cylinder were recorded. Discrepancies were resolved by consensus between 3 pathologists using a multihead microscope. The criteria for achieving a positive LI-cadherin score was moderate or strong immunoreactivity present in >25% of the cells, as described previously.22, 31

Statistical Analyses

Quantitative data between groups were compared using the Student t test. Categorical data were analyzed by the chi-square test or Fisher exact test. The Kaplan-Meier method was used to determine survival probability and the differences were assessed by the log-rank test. Univariate and multivariate analyses were performed using the Cox proportional hazards regression model. Statistical significance was set at P < .05. All analyses were performed using the SPSS software (v.15.0, Chicago, Ill).

RESULTS

LI-Cadherin Expression Correlates With Invasive Potentials of HCC Cell Lines

LI-cadherin expression was evaluated in 3 HBV-positive HCC cell lines (Hep3B, PLC/PRF/5, and HCCLM3) with low- and high-invasive potentials. Western-blot and immunofluorescence analyses showed a significant increase in LI-cadherin protein levels in high-invasive HCCLM3 cells compared with low-invasive Hep3B and PLC/PRF/5 cells (Fig. 1A,B). The same result was found at the transcriptional level with real-time PCR analysis (Fig. 1C).

Figure 1.

Liver-intestine cadherin (LI-cadherin) expression in 3 hepatitis B virus (HBV)-positive hepatocellular carcinoma (HCC) cell lines (Hep3B, PLC/PRF/5 and HCCLM3) with low- and high- invasive potentials. (A) Fluorescence microscopic analysis for LI-cadherin expression in Hep3B, PLC/PRF/5 and HCCLM3. HCCLM3 displayed a strong immunostaining with increased fluorescence in cytoplasm and cytolemma, compared with Hep3B and PLC/PRF/5. Isotype-matched IgG1 was used as negative controls (right panels). (B) Western-blot analysis showed a significant increase of LI-cadherin at the protein level in high-invasive HCCLM3 cell. (C) Real-time polymerase chain reaction (PCR) at the transcriptional level also validated different messanger ribonucleic acid (mRNA) expressions in HCC cell lines. Data shown are mean (±SD) from 3 independent experiments.

To further investigate the role of LI-cadherin in tumor invasion in vitro, we inhibited the expression of LI-cadherin using a siRNA-mediated approach in the high-invasive HCCLM3 cell line. Two pairs of siRNAs were used to evaluate the inhibitory efficiency, and results showed that mRNA and protein levels of LI-cadherin were markedly inhibited at 48 hours and 72 hours after transfection, respectively, which indicated that LI-cadherin expression could be silenced successfully by ribonucleic acid interference (RNAi) (Fig. 2A,C). Meanwhile, no obvious change of N-cadherin mRNA expression after transfection was detected, which confirmed the absence of off-target effect of these siRNAs (Fig. 2B). Cellular migration, adhesion, and invasion were then assessed.

Figure 2.

Knockdown of liver-intestine cadherin (LI-cadherin) attenuates HCCLM3 cell migration, ECM adhesion, and invasive potential in vitro. (A) Two pairs of small interfering ribonucleic acids (siRNA) were used to evaluate the inhibitory efficiency, with the result that protein levels of LI-cadherin were markedly inhibited at 72 hours after transfection. (B) Neuronal-cadherin (N-cadherin) messanger ribonucleic acid (mRNA) expression was detected at 72 hours after transfection. (C) The mRNA expression of CDH17 at dose-response levels and at different time points after LI-cadherin–1 siRNA transfection. (D) Monolayer of negative siRNA or LI-cadherin–siRNA HCCLM3 cells was abraded and then monitored at 0, 24, and 48 hours for wound channel closure. The cleared area was measured and plotted as the percentage of the original time point (0 hours); Data shown are mean (± SD) from 3 independent experiments. Magnification, × 200. (E) Negative siRNA or LI-cadherin-siRNA HCCLM3 cells were plated on fibronectin-coated plastic dishes and adhesion was measured by absorbance at 562 nm. Data shown are mean (± SD) from 3 independent experiments. (F) Migration of negative siRNA or LI-cadherin–siRNA HCCLM3 cells through a matrigel-coated 8um transwell. After 36 hours, migrated cells on the bottom surface were fixed and counted after staining with Giemsa. Data shown are mean (± SD) from 3 independent experiments. 12 random fields were counted. (Magnification × 200. *P < .05).

Cellular migration is a characteristic of metastatic tumors and the first step of invasion. To assess whether LI-cadherin down-regulation in HCCLM3 cells reduces cell migration, we performed a wound-healing migration assay. Microscopic examination at 24 hours and 48 hours revealed a significant delay in the wound closure rate of LI-cadherin-siRNA HCCLM3 compared with the control cell line, in which the wound was closed by 48 hours (Fig. 2D).

LI-cadherin has been shown to be capable of mediating cell–cell adhesion. However, it is largely unknown whether LI-cadherin is involved in tumor cell–ECM adhesion. To assess the role of LI-cadherin in cell adhesion to ECM substrates, LI-cadherin-siRNA HCCLM3 cells or control cells were plated on plastic plates coated with fibronectin. The LI-cadherin-siRNA cells showed a significant reduction in their ability to adhere to fibronectin compared with control cells (Fig. 2E).

Matrigel invasion assay was used to validate the invasive potentials of HCC cell lines. As shown in Figure 2F, silencing LI-cadherin expression was associated with a 54% reduction in invaded cells in the LI-cadherin–siRNA group compared with control cells. This result showed the matrigel invasive ability of HCCLM3 cells was markedly suppressed after inhibition of LI-cadherin expression.

Collectively, these results indicate that reduction of LI-cadherin in HCCLM3 cells attenuates cellular migration, ECM adhesion, and invasive potential. As a consequence, LI-cadherin should be considered as an invasive marker for HCC cells.

LI-Cadherin Mediates Microvascular Metastasis of HBV-Related HCC

To further confirm the relationship between LI-cadherin and HBV-related HCC, immunohistochemical analysis of patient HCC samples was used. By testing against 50 pairs of HCC and adjacent nontumor tissues, positive expression of LI-cadherin was identified in 28 (56%) of HCC. Only 12 (24%) of peritumoral tissues showed positive staining. Among these 50 HCC patients, 36 cases (72%) showed overexpression of LI-cadherin in HCC tissues compared with nontumor tissues and 10 cases (20%) were similar. Furthermore, none of the 5 normal liver specimens tested was positive with LI-cadherin.

The expression of LI-cadherin in HCC tissues was also investigated in by TMA analysis of 255 HBV-positive HCC cases. No correlation was found between LI-cadherin expression and clinical factors such as age, sex, tumor size, tumor number, tumor differentiation and TNM stage. Patients with high alpha-fetoprotein level or cirrhosis possessed more positive LI-cadherin expression. Moreover, there was a significant association between overexpression of LI-cadherin and vascular invasion (P = .018) (Table 1). LI-cadherin-positive staining was also observed in portal vein thrombosis by immunohistochemical analysis (Fig. 3A).

Figure 3.

Liver-intestine cadherin (LI-cadherin) mediates microvascular metastasis of hepatitis B virus (HBV)-related hepatocellular carcinoma (HCC). (A) LI-cadherin-positive tumor cells in portal vein thrombosis by immunohistochemical staining combining CD34 and LI-cadherin in consecutive sections. (Magnification × 400.) (B) LI-cadherin protein at significantly higher levels was observed in 16 HCC cases with microvascular invasion by Western-blot. In the semiquantitative analysis, bands of LI-cadherin and GAPDH were quantified by densitometry. GAPDH protein level served as internal control. Mean values (± SD) of independent experiments measured and repeated thrice.

To further confirm changes in LI-cadherin expression levels with microvascular metastasis in HBV-positive HCC cells, we performed Western-blot analysis of 32 clinical samples undergoing curative resection at our surgical center (half the cases were HCC with microvascular invasion and half were HCC without microvascular invasion, according to pathological examination). By semiquantitative analysis, LI-cadherin protein expression was found at significantly higher levels in HCC cases with microvascular invasion (P = .01).

LI-Cadherin Overexpression Is Associated With Poor Prognosis of HBV-Positive HCC

As microvascular invasion always predicts tumor recurrence or poor prognosis of HCC patients undergoing curative resection, we investigated whether the observed expression of LI-cadherin in HCC cells (Fig. 4A) is correlated with clinical outcome.

Figure 4.

The expression patterns of liver-intestine cadherin (LI-cadherin) in normal liver, adjacent nontumor and tumor tissues, and survival analysis are shown. (A) Representative immunohistochemical staining of LI-cadherin in normal liver tissue (top left), adjacent nontumor tissue (top right), and tumor tissue (negative and positive case, bottom left and right). (Magnification × 200.) (B) Kaplan-Meier analysis of disease-free survival and overall survival for LI-cadherin expression in whole study population.

The 3-year, 5-year, and 7-year DFS and OS rates of all 255 HBV-positive HCC patients were 55.5% and 70.2%, 45.5% and 55.2%, 40.4% and 42.6%, respectively. The patients with positive LI-cadherin expression had a significantly poorer prognosis (DFS, P < .0001 and OS, P < .0001, by Kaplan-Meier survival analysis) than those without LI-cadherin expression. The 3-year, 5-year, and 7-year DFS and OS rates in LI-cadherin–positive patients were 40.9% and 59.5%, 30.8% and 44.5%, 25.7% and 28.2%, respectively, while the 3-year, 5-year, and 7-year DFS and OS rates in LI-cadherin–negative patients were 68.5% and 79.9%, 58.7% and 64.9%, 53.4% and 55.6%, respectively.

In further investigations of the prognostic value of LI-cadherin expression in HBV-related HCC patients, univariate analysis (Table 2) showed that age, sex, cirrhosis, alpha-fetoprotein levels, tumor size, and tumor differentiation had no prognostic significance for DFS and OS. For OS and DFS, both tumor node TNM stage and tumor number were predictors of prognosis; vascular invasion was also a prognostic factor for DFS. LI-cadherin expression was also a significant predictor for tumor recurrence and OS (P < .0001 and P < .0001, respectively) in the study population.

Table 2. Univariate Analyses of Factors Associated With Recurrence and Survival*
VariableDisease-Free SurvivalOverall Survival
Hazard Ratio (95% CI)Median Survival (mo)PHazard Ratio (95% CI)Median Survival, moP
  • Abbreviations: CI indcates confidence interval; TNM, tumor, node, metastasis; LI-cadhein, liver-intestine cadherin.

  • *

    Univariate analysis, Cox proportional hazards regression model.

  • Median survival, time from the date of surgery to the date when half the patients are confirmed to be alive.

Sex (female vs male)1.260 (0.786-2.017)5167.3370.892 (0.581-1.370)6267.602
Age, y, ≤52 vs >520.897 (0.649-1.241)3958.5130.879 (0.637-1.212)6172.431
Alpha-fetoprotein, ng/mL, ≤20 vs >201.161 (0.840-1.604)5943.3651.295 (0.937-1.788)7362.117
Cirrhosis, no vs yes1.482 (0.755-2.909)9649.2531.894 (0.887-4.045)67.099
Tumor size, cm, ≤5 vs >51.355 (0.983-1.867)6033.0631.348 (0.980-1.854)7354.066
No. of tumors, single vs multiple1.726 (1.149-2.594)6032.0091.640 (1.099-2.447)7152.015
Vascular invasion, no vs yes1.934 (1.258-2.975)5912.0031.442 (0.925-2.248)6850.106
TNM stage, I vs II/III1.963 (1.398-2.757)6619<.00011.628 (1.159-2.285)7552.005
Tumor differentiation, I-II vs III-IV1.352 (0.959-1.907)6032.0851.396 (.995-1.959)7255.054
LI-cadherin, negative vs positive2.247 (1.618-3.121)28<.00012.117 (1.530-2.929)n.a51<.0001

Multivariate analysis (Table 3) investigated the 3 variables that had shown prognostic significance in univariate analysis and no obvious correlation between each other (LI-cadherin expression, vascular invasion, tumor number). LI-cadherin expression was shown to be the promising independent variable for predicting poor DFS and OS (P < .0001 and P < .0001, respectively).

Table 3. Multivariate Analyses of Factors Associated With Recurrence and Survival*
 Median Survival, moHazard Ratio (95% CI)P
  • CI indicates confidence interval; LI-cadherin, liver-intestine cadherin.

  • *

    The Cox proportional hazards regression model was used for the multivariate analysis. Variables were adopted for their prognostic significance by univariate analysis and no obvious correlation between each other.

  • Median survival was the time from the date of surgery to the date when half the patients were confirmed to be alive.

Disease-free survival    
 No. of tumors, single vs multiple60321.745 (1.161-2.622).007
 Vascular invasion, no vs yes59121.750 (1.133-2.702).012
 LI-cadherin, negative vs positive282.146 (1.541-2.989)<.0001
Overall survival    
 No. of tumors, single vs multiple71521.689 (1.130-2.522).011
 LI-cadherin, negative vs positive512.141 (1.547-2.023)<.0001

In conclusion, these results demonstrate that in HBV-positive HCC patients, LI-cadherin expression is significantly correlated with poor survival.

DISCUSSION

Microvascular invasion is a key step in HCC metastasis and is an important predictor of poor prognosis.32 Therefore, it is important to study the molecular mechanism of HCC invasion and investigate clinically available markers. In our ongoing research on side population of human HCC cancer cell lines, we have noticed increased LI-cadherin expression levels in more invasive side population HCC cells by gene-expression profiling analysis (data not shown). This observation prompted us to investigate the physiologic/clinical relevance and the role of LI-cadherin in HCC invasion.

In general, loss or inactivation of adhesion molecules is correlated with inhibition of cell aggregation and promotion of tumor invasion.6, 33 Interestingly, LI-cadherin–mediated cell-cell contacts do not require cytoplasmic interactions because the cytoplasmic domain of LI-cadherin is relatively short.34 LI-cadherin comprises 7 extracellular cadherin repeats and a short cytoplasmic region that does not interact with catenins. In contrast, E-cadherin comprises 5 cadherin repeats and a large cytoplasmic domain that is linked via catenins to the actin cytoskeleton. Although E-cadherin is concentrated in adherent junctions, LI-cadherin is evenly distributed along the lateral contact area of intestinal epithelial cells.17, 35 Hence, LI-cadherin has markedly different roles in different gastroenterological tumors owing to its unique biological characteristics.

An important observation in our study is that the more invasive HCC cell line expressed increased LI-cadherin, indicating a role for LI-cadherin in the invasiveness of HCC cells. The clinical relevance of this observation is supported by the finding that LI-cadherin was overexpressed in a significant fraction of HCC cells exhibiting microvascular invasion. As the invasiveness of cancer cells depends on increased migration, adhesion and invasion, we hypothesized that LI-cadherin overexpression in HCC cells and cell lines will promote cell migration, adhesion to the ECM and invasion. Thus, loss of function (by siRNA-mediated knockdown approaches) was used to determine the important role of LI-cadherin in the migration, adhesion and invasion of HCC cells. SiRNA-mediated knockdown of LI-cadherin expression in HCCLM3 cells significantly reduced cell migration, ECM adhesion, and invasion through the matrigel. These results indicate that LI-cadherin is a positive regulator for migratory, adhesive, and invasive behaviors. It is conceivable that, consistent with N-cadherin,11, 12, 16 in HCC LI-cadherin mediates carcinoma cell interaction with hepatic stroma and is involved in the promotion of HCC microvascular metastasis by facilitating carcinoma cell migration through the hepatic stroma and reestablishing homophilic cell–cell adhesion in metastasis. Therefore, we suggest that LI-cadherin contributes to a stroma-oriented cellular adhesion profile with increased tumor cell migration and invasion, and that this molecular profile represents an invasive HCC cellular phenotype.

The extremely poor prognosis of HCC is largely the result of a high rate of recurrence after surgery due to de novo tumors in cirrhotic liver or intrahepatic metastases; extrahepatic metastases are less common.36 As an invasive marker, LI-cadherin may have a role in tumor microvascular metastasis, ultimately resulting in poor prognosis. Immunohistochemical and western blot analyses of HBV-positive HCC tissue samples confirmed the strong association between LI-cadherin expression and vascular invasion. 3-year, 5-year, and 7-year OS and DFS rates were significantly lower in LI-cadherin–positive patients than in LI-cadherin–negative patients. Moreover, LI-cadherin was the most promising independent predictor for OS and DFS in HBV-related HCC patients by both multivariate and univariate analyses. To our knowledge, this is the first report to describe LI-cadherin-mediated microvascular metastasis in human HBV-positive HCC. However, it remains to be determined whether these results can be applied to other patient populations, including those with hepatitis C virus-related HCC.

In conclusion, LI-cadherin could be considered as a potential HBV-related HCC marker for more aggressive treatment. Particularly, because of its sensitivity to tumor invasion, LI-cadherin could potentially serve as an indicator predicting microvascular invasion and the treatment outcome of HCC. Furthermore, LI-cadherin might be a potential therapeutic target for HBV-positive HCC.

Conflict of Interest Disclosures

This work is supported by China 863 project (2007AA02Z479) and National TCM project (2006BAI02A04).

Ancillary