This article is mentioned in:

Introduction

Cell adhesion is crucial during the development and adult life of multicellular organisms. There are principally two types of adhesion: Cell-cell adhesion; and cell-extracellular matrix (ECM) adhesion. The canonical receptors for cell-cell adhesion are cadherins (1). Cadherins are a multi-member glycoprotein family of transmembrane Ca2+-dependent adhesion molecules, which maintain tissue structure in normal and pathological settings (2,3). They are major contributors to cell-cell adhesion. Numerous cadherin superfamily members have been previously identified and these are comprised of four different subfamilies (classical, desmosomal, atypical and protocadherins) (4). Distinct members of the cadherin family are important for morphogenesis during development, formation of junctional complexes, induction of the polarized cell type, development of cell-cell associations and the invasion of tumor cells (5).

Liver-intestine cadherin (LI-cadherin), also termed cadherin-17, is expressed in fetal livers and the gastrointestinal tract during embryogenesis as one member of the 7D-cadherin superfamily (6). LI-cadherin is often upregulated in certain types of cancer, including hepatocellular carcinoma, gastric, ductal pancreatic and colorectal cancer, however, it has not been reported to be expressed in brain tumors or blood malignancies (7). LI-cadherin has been observed to be expressed in 96% of tumor samples and is regarded as a useful diagnostic marker for adenocarcinomas of the digestive system (8). Compared with classical cadherins, including E-, N- and P-cadherin, LI-cadherin possesses several unique features (9). For example, it has seven extracellular cadherin domains, whereas classical cadherins have five cadherin repeats. Additionally, LI-cadherin has a short cytoplasmic domain composed of 20 amino acids, which shares no homology with the corresponding regions of classical cadherins, such as E-cadherin, which binds catenin proteins through their cytoplasmic domains to initiate signaling cascades. Although LI-cadherin is able to mediate Ca2+-dependent cell-cell adhesion (10), the difference in structure causes the adhesive function of LI-cadherin to be independent of any interaction with catenins, the actin cytoskeleton or other cytoplasmic components. The adhesive mechanism of LI-cadherin remains unclear.

Previous studies have demonstrated that LI-cadherin is associated with colorectal carcinoma (11), gastric cancer (12–16), ductal adenocarcinoma of the pancreas (17) and hepatocellular carcinoma (18–23). Furthermore, the expression level of LI-cadherin is associated with lymph node metastasis, advanced pTNM stage, early tumor recurrence and poor overall survival (15,16,24,25). Together, these previous studies indicate that LI-cadherin is associated with the ability of tumor cells to acquire an aggressive phenotype in several types of cancer. However, the exact mechanisms of LI-cadherin in the development of cancer remain unclear.

Galectin-3, a member of the β-galactoside-binding lectin family, is involved in several biological processes, including tumor cell proliferation, differentiation, angiogenesis, adhesion, motility, invasion, cancer progression and metastasis (26). Interaction of galectin-3 with adhesion and signaling receptors has been demonstrated to promote tumor cell migration. For example, galectin-3 binding to N-cadherin destabilizes cell-cell junctions by increasing the turnover of N-cadherin and other glycoconjugates, which may increase cell migration (27). Whether LI-cadherin exhibits a similar function via binding galectin-3 remains unclear. Additionally, a previous study demonstrated that galectin-3 is cleaved by members of the matrix metalloproteinase (MMP) family of enzymes. The 72-kDa (gelatinase A, MMP-2) and 92-kDa (gelatinase B, MMP-9) proteinases cleave galactin-3 to generate a 22-kDa fragment containing the carbohydrate recognition domain and a 9-kDa fragment comprising of the amino terminal of galectin-3. Thus, galectin-3 is considered as a substrate for human MMP-2 and -9 (28). It is possible that galectin-3, MMP-2 and MMP-3 are important regulating molecules in the LI-cadherin signaling pathway.

In the current study, to classify the role of LI-cadherin in regulating MMP-2, MMP3 and galectin-3 in colorectal cancer cells, an RNA interference strategy was employed to specifically knockdown LI-cadherin in LoVo cells. The results of the present study demonstrate that a reduction in the galectin-3 expression levels is associated with the increased expression of MMP-2 and MMP-9, which is mediated by LI-cadherin.

Materials and methods

Cell lines and cell culture

Derivation of the pU6-LI-cadherin-short hairpin RNA (shRNA)-transfected (Shanghai Genechem Co., Ltd., Shanghai, China) LoVo cell (American Type Culture Collection, Manassas, VA, USA) clone was performed as previously described (29). The pU6-LI-cadherin-shRNA-transfected and pU6-control-shRNA-transfected LoVo cell clones were maintained in Dulbecco's modified Eagle's medium (GE Healthcare Life Sciences, Logan, UT, USA) supplemented with 10% fetal calf serum (ExCell, Shanghai, China) and G418 (350 µg/ml) (Gibco; Thermo Fisher Scientific, Inc., Waltham, MA, USA).

Semi-quantitative reverse transcription-polymerase chain reaction (RT-PCR)

Total RNA was extracted using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.), then 1 µg RNA was reverse transcribed using the RevertAid First Strand cDNA Synthesis kit (Fermentas; Thermo Fisher Scientific, Inc., Pittsburgh, PA, USA) according to the manufacturer's instructions. RT-PCR was performed using standard methodology and Ex Taq DNA polymerase and dNTPs were purchased from Takara Biotechnology Co., Ltd., (Dalian, China). Primers were designed according to the sequences of MMP-2 (GenBank accession no. NM_004530), MMP-9 (GenBank accession no. NM_004994) and β-actin (GenBank accession no. NM_001101). The primers used for MMP-2, MMP-9 and β-actin were as follows: MMP-2, forward 5′-CCATCACTATGTGGGCTG-3′, reverse 5′-TGCTGGCTGCCTTAGAAC-3′ (168 bp); MMP-9, forward 5′-TTGACAGCGACAAGAAGT-3′, reverse 5′-AGTAGTGGCCGTAGAAGG-3′ (483 bp); and β-actin, forward 5′-AAAGACCTGTACGCCAACA-3′, reverse 5′-GGAGCAATGATCTTGATCTTC-3′ (125 bp). The PCR was conducted on an S1000™ Thermal Cycler (Bio-Rad, Hercules, California, USA) and the cycling conditions were as follows: For MMP-2 and β-actin, 94°C for 5 min, 30 cycles of 94°C for 30 sec, 58°C for 30 sec and 72°C for 30 sec, followed by an additional extension step of 10 min at 72°C; and for MMP-9, 94°C for 5 min, 30 cycles of 94°C for 30 sec, 50°C for 30 sec, and 72°C for 30 sec, followed by an additional extension step of 10 min at 72°C. The RT-PCR products were separated by electrophoresis in a 1.5% agarose gel and visualized using ethidium bromide (Tiangen, Beijing, China) and a 2UV Transilluminator (LM-26E; UVP Inc., Upland, CA, USA). The expression ratio (MMP-2/β-actin and MMP-9/β-actin) measured by densitometry (Gel-Pro Analyzer 4.0, Media Cybernetics, Inc., Rockville, MD, USA) was used to evaluate the mRNA levels of the genes. The mRNA level of each sample was detected following at least two independent experiments.

Western blot analysis

LoVo cells were lysed in radioimmunoprecipitation assay buffer (Beyotime Institute of Biotechnology, Beijing, China) containing 50 mM Tris-HCl, 150 mM NaCl, 1 mM ethylenediaminetetraacetic acid, 1 mM sodium orthovanadate, 1 mM NaF, 1% NP40, and 0.25% sodium deoxycholate supplemented with 1 mM phenylmethylsulfonyl fluoride and quantified using the DC Protein Assay kit (Bio-Rad Laboratories, Inc., Hercules, CA, USA). Equal amounts of proteins from each sample were applied to a 10% sodium dodecyl sulfate (SDS) polyacrylamide gel (Beijing Solarbio Science & Technology Co., Ltd., Beijing, China) and transferred to a polyvinylidene difluoride membrane (EMD Millipore, Billerica, MA, USA). The membranes were blocked in 5% non-fat milk and incubated at 4°C overnight with diluted rabbit polyclonal anti-galectin-3 (1:1,000; cat. no. sc-20157; Santa Cruz Biotechnology, Inc., Dallas, TX, USA) or mouse monoclonal anti-β-actin (1:3,000; cat. no. sc-47778; Santa Cruz Biotechnology, Inc.) primary antibodies, followed by incubation with horseradish peroxidase (HRP)-conjugated goat anti-rabbit IgG (cat. no. ZDR-5403; ZSGB-Bio, Beijing, China). Immunoreactive bands were visualized using Immobilon Western Chemiluminescent HRP Substrate (EMD Millipore) and signals were developed on X-ray film (Kodak, Rochester, NY, USA). The expression ratio (galectin-3/β-actin) as measured by densitometry (Gel-Pro Analyzer 4.0, Media Cybernetics, Inc.) was used to evaluate protein levels.

Gelatin zymography

The culture supernatant was collected and the total protein concentration of the supernatants of each sample was determined using a Bradford Protein Assay kit (Beyotime Institute of Biotechnology). Culture supernatants containing equal amounts of total protein were mixed with SDS loading buffer (Beyotime Insititute of Biotechnology) and electrophoresed on 10% denaturing SDS polyacrylamide gels containing 1 mg/ml gelatin. Following electrophoresis, the gels were soaked in 2.5% Triton X-100 on a shaker for 1 h, the solution was changed after 30 min to eliminate SDS. Following incubation in zymogen activation buffer (50 mM Tris-HCl pH 7.5, 0.1% Triton X-100, 0.02% NaN3, 5 mM CaCl2 and 1 µM ZnCl2) at 37°C for 12 h, the gels were rinsed in distilled water and stained for 5 h with Coomassie blue R250 (Beyotime Institute of Biotechnology). Gelatinolytic bands were observed as clear zones against the blue background and the intensity of the bands was estimated using ImageJ software, version 1.45 (imagej.nih.gov/ij/). The gelatinase expression level of each sample was determined following a minimum of three independent experiments.

Statistical analysis

All analyses were performed using SPSS software, version 11.5 (SPSS, Inc., Chicago, IL, USA). The mRNA and protein levels were analyzed by one-way analysis of variance followed by a post-hoc least-significant difference test. P<0.05 was considered to indicate a statistically significant difference.

Results

Silencing of LI-cadherin increases the mRNA level of MMP-2 and MMP-9 in LoVo cells

To investigate the effect of silencing LI-cadherin on the mRNA levels of MMP-2 and MMP-9 in LoVo cells, RT-PCR analysis was performed. The results demonstrated that there was no significant difference between the mRNA levels of MMP-2 in untreated LoVo cells and in LoVo cells stably expressing pU6-control-shRNA (P>0.05). By contrast, the mRNA level of MMP-2 in LoVo cells expressing pU6-LI-cadherin-shRNA was significantly increased compared with untreated (P<0.05) and pU6-control-shRNA-expressing LoVo cells (P<0.05; Fig. 1A). Similarly, the mRNA level of MMP-9 did not differ between untreated and LoVo cells stably expressing pU6-control-shRNA (P>0.05). However, the mRNA level of MMP-9 was significantly increased in LoVo cells following silencing LI-cadherin compared with untreated and control shRNA-transfected cells (P<0.05; Fig. 1B).

Figure 1 Effects of pU6-LI-cadherin-shRNA on mRNA transcript levels of MMP-2 and MMP-9 in LoVo cells. Semi-quantitative reverse transcription-polymerase chain reaction was performed to assess the mRNA levels of (A) MMP-2 and (B) MMP-9. β-actin was used as a reference gene to normalize the samples. Statistical analysis was performed using analysis of variance followed by the post-hoc least significant difference test. *P<0.05, comparison indicated by brackets. Values are presented as the mean ± standard deviation of 3 independent samples performed in triplicate. LI-cadherin, liver-intestine cadherin; MP, matrix metalloproteinase; shRNA, short hairpin RNA.

Silencing of LI-cadherin increases the protein levels and activity of MMP-2 and MMP-9 in LoVo cells

To explore whether the protein expression levels and activity of MMP-2 and MMP-9 were altered following knockdown of LI-cadherin in LoVo cells, a gelatin zymography experiment was performed (Fig. 2). The results of the current study demonstrated that the protein levels of proMMP-2 and -9, and their active forms were significantly increased in LoVo cells expressing pU6-LI-cadherin-shRNA compared with untreated and control shRNA-transfected LoVo cells (P<0.05).

Figure 2 Effects of silencing LI-cadherin on the protein levels and activities of MMP-2 and MMP-9. (A) Gelatin zymographic analysis. (B) Quantification of zymography. Statistical analysis was performed using analysis of variance followed by post-hoc least significant difference test. *P<0.05, comparisons indicated by brackets. Values are presented as the mean ± standard deviation of 3 independent samples performed in triplicate. LI-cadherin, liver-intestine cadherin; MMP, matrix metalloproteinase; shRNA, short hairpin RNA.

Silencing LI-cadherin reduces the protein expression level of galectin-3

Western blot analysis was performed to detect the effect of of LI silencing via transfection with pU6-LI-cadherin-shRNA on the protein levels of galectin-3 in LoVo cells. The protein level of galectin-3 exhibited no significant difference between the untreated cells and cells stably expressing pU6-control-shRNA. By contrast, following knockdown of LI-cadherin, LoVo cells exhibited significantly reduced protein levels of galectin-3 compared with untreated and control shRNA-transfected cells (P<0.05; Fig. 3).

Figure 3 Effect of LI-cadherin knockdown on the protein levels of galectin-3 in LoVo cells. (A) Western blot analysis was performed to detect galectin-3. β-actin was used as a loading control. (B) Statistical analysis was performed using analysis of variance followed by the post-hoc least significant difference test. *P<0.05, comparisons indicated by brackets. Values are expressed as the ratio of galectin-3 to β-actin level and presented as the mean ± standard deviation of 3 independent samples performed in triplicate. LI-cadherin, liver-intestine cadherin; MMP, matrix metalloproteinase; shRNA, short hairpin RNA.

Discussion

A previous study demonstrated that knockdown of LI-cadherin promotes cell migration, invasion and adhesion (29). However, the mechanisms that mediate these changes remain unclear. The present study demonstrated that silencing of LI-cadherin increases the expression levels (mRNA and protein) and activation of MMP-2 and -9, and downregulates the protein level of galectin-3, which is a substrate for human MMP-2 and MMP-9. Based on the present data, it is proposed that knockdown of LI-cadherin expression facilitates the invasion of cancer cells by degrading ECM components via enhanced expression and activation of MMP-2 and -9, and increases cancer cell adhesion and migration via altered expression of galectin-3.

The MMPs are a tightly regulated family of enzymes that degrade ECM and basement membrane components, thus allowing cancer cells access to subepithelial structures (30). MMP-2 and -9, in particular, are important for cleaving major components of the basement membrane, such as type IV collagen. They interact with αvβ3 integrin, which contributes to collagen degradation, cell migration and cell invasion (31,32). A previous study demonstrated that silencing LI-cadherin promotes cell invasion (29). The present study demonstrated that knockdown of LI-cadherin increased the expression levels and activation of MMP-2 and -9. Thus, it is concluded that LI-cadherin-associated invasion may contribute to LI-cadherin-induced alteration and activation of MMP-2 and -9

A previous study demonstrated that galectin-3 (31 kDa) is cleaved by MMP-2 and -9 to generate 22-kDa and 9-kDa fragments (28,33). In the current study, silencing LI-cadherin in LoVo cells significantly increased the mRNA levels of MMP-2 and -9, whereas the protein level of galectin-3 was downregulated under the same conditions. Together these studies suggest that the reduction in the protein levels of galectin-3 (31 kDa) is induced via enhanced cleavage resulting from increased expression levels of MMP-2 and -9 meditated by LI-cadherin.

In the present study, silencing LI-cadherin reduced the protein levels of galectin-3, a protein that is closely involved in tumor cell transformation, migration, invasion and metastasis. A previous study indicates that Ca2+-dependent cell-cell adhesion mediated by LI-cadherin is independent of interaction with cytoplasmic components (10). Thus, it is possible that LI-cadherin inhibits LoVo cell migration and adhesion via galectin-3 in an indirect manner. However, galectin-3 predominantly promotes tumor development in cancer, however, it acts as a tumor-suppressor in certain types of cancer (34). Additionally, Bartolomé et al (7) reported a different mechanism when they investigated the association between LI-cadherin and cell proliferation and adhesion. They demonstrated that cell adhesion, proliferation and liver metastasis were reduced following knockdown of LI-cadherin in KM12 metastatic colorectal cancer cells, and the effects are regulated by interaction between LI-cadherin and α2β1 integrin. The previous study also observed a significant correlation between LI-cadherin overexpression and poor survival in colorectal cancer, whereas, other studies had previously demonstrated that reduced expression of LI-cadherin is associated with lymph node metastasis (11), or tumor dedifferentiation and poor overall survival (35) based on immunohistochemical analysis. The conflicting results are possibly caused by differences in samples number, the ratio of cancer stage and cell types. The association of LI-cadherin expression with cell adhesion requires further elucidation.

In summary, previous studies have demonstrated that LI-cadherin has an important function in migration, invasion and adhesion (13,19,29,36). LI-cadherin acts via various molecular mechanisms depending on the cancer cell type. In the current study, it was identified that in colorectal cancer cells, there was an association between LI-cadherin, gelatinases and galectin-3, providing insight into the functional regulation of LI-cadherin and a better understanding of the molecular mechanisms of the LI-cadherin. Further study of LI-cadherin may be important to advance the understanding of human cancer progression and developing novel diagnostic markers.

Acknowledgments

The present study was supported by grants from the National Natural Science Foundation of China (grant nos. 81460462 and 31460696), the Technology Pedestal and Society Development Project of Jiangxi Province (grant no. 20141BBG70040) and the Science Foundation of Educational Department of Jiangxi Province (grant no. 86283702).

References