Screening Of A Novel Peptide Specifically Binding To High Metastatic Potential Human Hepatocellular Carcinoma Cells And Its Biological Functions

Hepatocellular carcinoma (HCC) is one of the most common and aggressive malignancies worldwide, ranking the sixth most important cancer in terms of numbers of cases and the third in terms of cancer mortality. And it has become the second cancer killer in China since 1990s. Unfortunately, despite efforts to improve its prognosis, the overall survival of patients with HCC is still unsatisfactory. Metastatic recurrence after HCC resection is one of the major obstacles to prolonged survival. Therefore, there is a tremendous interest and urgency to search for HCC metastasis related molecules that would provide new predictors for HCC metastasis as well as new targets for intervention.HCC metastasis is a multigene-involved, multistep, and changing process, and varies with different metastatic phenotypes. Tumor cell surface molecules play an important role in tumor invasion and metastasis. Tumor cell surface molecules, however, are highly complex, including, in addition to proteins, carbohydrate determinants and other post-translational modification products that are difficult to probe by gene expression-based approaches. Specific binding of phage from libraries that can display more than a billion random peptide sequences offers a high throughput approach to probe tumor cell surface. This approach is not only a promising tool to identifying cell-specific targeting peptides, but it may also be useful for identifying peptide ligands for unknown receptors, and ultimately the receptors themselves. So far, a large number of cancer cell surface binding peptides have been identified by screening peptide libraries.We have established a stepwise metastatic human HCC cell model system with similar genetic background to explore HCC metastasis related molecules. In the present study, to identify a novel peptide (AWYPLPP peptide) that could selectively bind highly metastatic HCC cell surface, we used the 'biopanning and rapid analysis of selective interacting ligands' (BRASIL) method to screen a phage-displayed random peptide library on human HCC cells with different metastatic potentials. We then showed that the peptide was able to promote in vitro invasion and in vivo lungmetastasis of highly metastatic HCC cells. Identification of the receptor responsible for AWYPLPP binding may contribute to the development of a new molecular marker for predicting metastasis and a therapeutic target for highly metastatic HCC.I . Subtractive selection of specific peptides binding to highly metastatic HCC cell surfaceWe have established a stepwise metastatic human HCC cell model system with similar genetic background to explore HCC metastasis related molecules. In the present study, to identify specific peptides that could bind highly metastatic HCC cell surface, we used the BRASIL method to screen a phage-displayed random peptide library on two human HCC cell lines with high and low metastatic potentials, HCCLM3 (with 100% of lung metastasis after orthotopic transplantation or subcutaneous transplantation) and MHCC97L (with 40% of lung metastasis after orthotopic transplantation, and lung metastasis was not found after subcutaneous transplantation). We designed a two-step subtractive panning strategy. The phage display library was per-absorbed against MHCC97L cell line with low metastatic potential to remove those phage that bind to the MHCC97L cell surface molecules. The depleted peptide library was then incubated with HCCLM3 cell line with high metastatic potential under conditions that minimized post-binding events such as receptor-mediated internalization. The ratio of output to input phage was calculated for each round used as a measure of the selection process. After three rounds of subtractive biopanning, sequencing the DNA corresponding to the inserts of individually selected phage clones led to the identification of 48 clones displaying peptides. One phage displaying AWYPLPP sequence accounted for more than 77% (37 of 48) of the phage analyzed. To further characterize the eleven candidate phage, we adapted the BRASIL method to evaluate the binding efficiency of them to the HCCLM3 cell line, except that the pre-clearing step was omitted. We set out the background recovery of insertless phage f1 -tet to 1. Each of the selected phage bound significantly more than the control insertless phage (range, 3.3-27.8-fold; median, 4.7-fold). In particular, the AWYPLPP phage showed up to 27.8-fold greater binding yields compared to the f1-tet phage. This suggests that the enrichment of AWYPLPP phage clone in the selected phage pool was due to its significant binding efficiency to the HCCLM3 cells. Analysis of AWYPLPP using a BLAST (NCBI) search for shortnearly exact matches did not reveal any significant homology with known membrane proteins/receptors. It is interesting to note that some biologically relevant matches with consensus motif were discovered. For example, a 71% match with human extracellular matrix protein 1 (ECM1) was found. ECM1 has been associated with angiogenesis; and the protein is known to be overexpressed in many epithelial tumors including 16.7% of HCC analyzed; moreover, ECM1 is preferentially expressed by metastatic epithelial tumors and correlated with the metastatic properties of breast carcinoma. Based on these results, the AWYPLPP phage was chosen as the lead candidate in the panel for further analysis.II. Verification of the specific binding activity of AWYPLPP peptide to highly metastatic human HCC cell surfaceWe initially attempted to identify peptides that selectively bind to highly metastatic human HCC cells. While AWYPLPP phage selected from our panning protocol displays good selectivity to HCCLM3 cell line, there is no assurance that the phage will recognize highly metastatic human HCC cells. We next determined such cell-specificity binding capacity amongst several human HCC cell lines with different metastatic potential, including high metastatic potential HCC cell lines (MHCC97, MHCC97H, HCCLM3, and HCCLM6), low metastatic potential HCC cell lines (MHCC97L, PLC/PRE/5, and Hep3B), and nontransformed liver cell line (CCL13). The BRASIL method as above was also used to test the binding efficiency of AWYPLPP phage to these cell lines. The phage preferentially bound to the MHCC97, MHCC97H, HCCLM3, and HCCLM6 cells as compared with the MHCC97L, PLC/PRE/5, Hep3B, and CCL13 cells. The ability of AWYPLPP phage to identify highly metastatic human HCC cells was further confirmed using flow cytometry analyses and immunocytochemistry studies. The AWYPLPP phage bound specifically to the high metastatic potential cell lines - MHCC97, MHCC97H, HCCLM3, and HCCLM6 cells, but no phage binding was observed on the low or non-metastatic potential cell lines-MHCC97L, PLC/PRE/5, Hep3B, and CCL13 cells. Immunofluorescence studies were performed in an attempt to verify the localization of AWYPLPP phage in HCCLM3 cells. Adherent phage were visualized by confocal microscopy after incubation with a rabbit anti-bacteriophage primary antibody followed by fluorescently labeled goat anti-rabbit antibody. Specific staining was seenwith the AWYPLPP phage on the surface of HCCLM3 cells and was not found inside the cells, whereas the control insertless phage failed to label. Moreover, to further confirm the specific targeting mediated by phage-displayed peptide, competition inhibition assay were performed with synthetic cognate or scrambled control peptide. The results showed that the binding activity of HCCLM3 cells with the AWYPLPP phage was inhibited by synthetic cognate peptide in a dose-dependent manner. Half maximal inhibition (IC50) was observed at concentrations as low as 0.5μM, whereas the scrambled control peptide had no such effect. These results show that the AWYPLPP peptide specifically binding to high metastatic potential human HCC cell surface.III. AWYPLPP peptide promotes in vitro invasion and in vivo lung metastasis of highly metastatic HCC cellsTo examine whether AWYPLPP peptide has any impact on HCC cell invasiveness, we performed an in vitro Matrigel invasion assay. Incubation with the AWYPLPP peptide, not the scrambled control peptide, resulted in a concentration-dependent increase in HCCLM3 cell invasion. At a 100μM concentration of AWYPLPP peptide, 250% promotion was observed. Similar results were obtained with other highly metastatic HCC cells such as MHCC97, MHCC97H and HCCLM6, whereas Hep3B, PLC/PRE/5, and MHCC97L cells were not affected, implying that this effect is cell-specific because cells that do not bind AWYPLPP peptide were not affected. Since invasion has been shown to be directly dependent upon migration for some systems, it is expected that the promotion of cell invasion be related to cell motility. Surprisingly, the peptide at any concentrations used for the invasion assay did not change transport of cells through the membrane in the absence of Matrigel barrier. Moreover, the promotion of cell invasion was not due to cell viability as there was no effect on cell proliferation when the cells were cultured for 48 hr with the peptide at the same concentrations. Because matrix-degrading proteinases have been implicated as playing an important role in HCC invasion and metastasis, we speculated that the increased invasiveness might be related to MMP activity. To test this hypothesis, we performed a gelatin zymography assay. Increased activity of MMP9 was observed after incubation of HCCLM3 cells with the AWYPLPP peptide, but not with the scrambled control peptide. The peptide had littleeffect on cell adhesion. Given that AWYPLPP peptide promoted invasion of highly metastatic HCC cells in vitro, we next examined its effect on lung metastasis of HCC cells in vivo. We gave HCCLM3 tumor mice s.c. injection of the AWYPLPP peptide, starting after the mice had established palpable tumors. There was no significant difference among the weight of primary tumor in PBS, scrambled control peptide, and AWYPLPP peptide groups, which is consistent with our in vitro observations that the AWYPLPP peptide does not affect HCC cell growth. Eight of nine mice (89%) in the AWYPLPP peptide group showed lung metastasis. Incidence of lung metastasis was markedly increased in the AWYPLPP peptide group compared with the PBS group (two of nine, 23%). No difference in the incidence of lung metastasis was observed between the scrambled control peptide and PBS groups.Conclusions1. The BRASIL method combined with the subtractive selection strategy allows the selection of AWYPLPP peptide as a specific peptide ligand binding highly metastatic human HCC cell surface.2. The AWYPLPP peptide can promote in vitro invasion of highly metastatic HCC cells and in vivo lung metastasis of HCC tumors.3. The mechanism of AWYPLPP peptide on promoting HCC invasion and metastasis may be associated with activating matrix metalloproteinase -9.4. This work further develops the tumor heterogeneity theory. Tumor cells with metastatic potential may already exist early during tumor evolution, and metastatic heterogeneity is a dynamic ever-changing process.The potential application of this study1. The BRASIL method combined with the subtractive selection strategy allows investigation on HCC cell surfaces.2. The ability of AWYPLPP peptide to identify highly metastatic HCC cells may help to define HCC patients with metastatic potential.3. The AWYPLPP peptide correlates with HCC invasion and metastasis, which may provide some insight into the molecular mechanism of HCC metastasis.4. Identification of the receptor for AWYPLPP peptide binding may contribute to thedevelopment of a new molecular marker for predicting metastasis and a therapeutic target for highly metastatic HCC.The novelty of this project1. The BRASIL method, for the first time, is applied to screen peptide ligands specifically binding to tumor cell surface.2. To our knowledge, this is the first time that a HCC metastasis related peptide is discovered using phage display technology.