Production of antibodies with peptide-CpG-DNA-liposome complex without carriers

Background The screening of peptide-based epitopes has been studied extensively for the purpose of developing therapeutic antibodies and prophylactic vaccines that can be potentially useful for treating cancer and infectious diseases such as influenza virus, malaria, hepatitis B, and HIV. To improve the efficacy of antibody production by epitope-based immunization, researchers evaluated liposomes as a means of delivering vaccines; they also formulated adjuvants such as flagella and CpG-DNA to enhance the magnitude of immune responses. Here, we provide a potent method for peptide-based epitope screening and antibody production without conventional carriers. Results We present that a particular form of natural phosphodiester bond CpG-DNA encapsulated in a specific liposome complex (Lipoplex(O)) induces potent immunomodulatory activity in humans as well as in mice. Additionally, Lipoplex(O) enhances the production of IgG2a specific to antigenic protein in mice. Most importantly, immunization of mice with several peptides co-encapsulated with Lipoplex(O) without carriers significantly induces each peptide-specific IgG2a production in a TLR9-dependent manner. A peptide-specific monoclonal antibody produced against hepatocellular carcinoma-associated antigen has functional effects on the cancer cells. Conclusions Our overall results show that Lipoplex(O) is a potent adjuvant and that complexes of peptide and Lipoplex(O) are extremely useful for B cell epitope screening and antibody production without carriers. Therefore, our strategy may be promptly used for the development of therapeutic antibodies by rapid screening of potent B cell epitopes.

Phosphorothioate-modified types of CpG-DNA (PS-ODN), which are resistant to nuclease activity and can be efficiently delivered into cells [10,11], have been utilized in clinical applications [9]. The immunomodulatory activities of PS-ODN are enhanced by liposome-encapsulation [12][13][14]. However, several studies have suggested that PS-ODN induces backbone-related side effects, such as transient splenomegaly [15], lymphoid follicle destruction [16], arthritis [17], and PS-ODN-specific IgM production [18] in PS-ODN-treated mice. Investigators consequently developed phosphodiester bond CpG-DNA (PO-ODN) as a natural counterpart of PS-ODN to induce optimal innate immune responses without severe side effects. In contrast to PS-ODN, the immunomodulatory effects of PO-ODN are found only in mouse cells and not in human cells [19]. However, induction of an effective immune response has been reported in human cells stimulated with PO-ODN and non-CpG-DNA encapsulated in cationic liposomes such as N- [1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methylsulfate (DOTAP) and lipofectin [20,21].
In previous studies, we screened natural PO-ODN with immunomodulatory activity from Mycobacterium bovis genomic DNA [22]. Our experimental analyses demonstrated that a potent PO-ODN, namely MB-ODN 4531 (O), which contains three CpG motifs, has functional effects as a powerful adjuvant for the induction of Agdriven Th1 responses without causing severe side effects in mice [18,22]. In this study, we compared the ability of MB-ODN 4531(O) encapsulated in several different liposomes to stimulate immune responses in human and mice cells, finding that MB-ODN 4531(O) encapsulated in a phosphatidyl-β-oleoyl-γ-palmitoyl ethanolamine (DOPE):cholesterol hemisuccinate (CHEMS) complex (Lipoplex(O)) was most potent in human as well as in mice. Furthermore, we extended the research to the selection of a synthetic peptide-based B cell epitope and revealed that complexes of several peptides and Lipoplex (O) without carriers significantly enhanced the each peptide-specific IgG production depending on TLR9. In this study, we identified a B cell epitope peptide from hepatocellular carcinoma (HCC)-specific transmembrane 4 superfamily member 5 (TM4SF5) protein [23] that potently induced epitope-specific antibodies. We also noticed that the monoclonal antibody produced by immunization with a complex consisting of antigenic peptide (TM4SF5R2-3) and Lipoplex(O) had functional effects on cells expressing the antigen. Our results suggest that the selection of B cell epitope can be facilitated by the delivery of the DOPE:CHEMS complex and by the adjuvant effect of MB-ODN 4531(O). Our strategy may be promptly used for the development of epitopebased peptide vaccines and production of therapeutic antibodies.

Effects of CpG-DNA encapsulated in liposomes on IL-8 promoter activation
To identify the conditions driving the effective immunomodulatory activity of PO-ODN in humans, we encapsulated MB-ODN 4531(O) in several different liposomes and compared the abilities of the complexes to stimulate immune responses in human and mouse cells. First, we confirmed that IL-8 promoter was activated in human and mouse cells treated with CpG-DNA encapsulated in liposomes. When PO-ODN was encapsulated in a DOTAP, DOPE:CHEMS (1:1 ratio) complex, or 1,2-dioctadecanoylsn-glycero-3-phosphocholine (DSPC):CHEMS:phosphatidylethanolamine-poly(ethylene glycol) (PEG-PE) ( Depending on the delivery vehicles, CpG-DNA can elicit enhanced intracellular uptake, endosomal localization, and biological activity of CpG-DNA [20]. To chase the process, we used FACS to analyze the intracellular uptake of FITC-labeled MB-ODN 4531(O). As shown in Figure 2a, the intracellular uptake of MB-ODN 4531(O) is enhanced markedly when the MB-ODN 4531(O) is encapsulated in a DOPE:CHEMS complex (Lipoplex (O)). TLR9-mediated cytokine production by CpG-DNA requires acidification and maturation of endosomes [24]. As shown in Figure 2b, pretreatment with chloroquine, an inhibitor of endosomal processing, inhibits the IL-8 promoter activation induced by Lipoplex(O). These results confirm that the efficacy of PO-ODN can be enhanced by encapsulation in a DOPE:CHEMS complex via the improved intracellular uptake of CpG-DNA and endosomal localization.

Effects of CpG-DNA encapsulated in DOPE:CHEMS complex on cytokine production
To determine the effects of Lipoplex (O) on cytokine production, we treated human peripheral blood mononuclear cells (PBMCs) or mouse splenocytes with 50 μg/ml of nonencapsulated CpG-DNA or Lipoplex(O) for 12 h. The CpG-DNA failed to induce the release of IL-6, IL-12, and IFN-γ in human PBMCs when treated as a free form. In contrast, Lipoplex(O) induced the release of IL-6, IL-12, and IFN-γ in a CpG sequence-dependent manner from human PBMCs and mouse splenocytes (Figure 3a and 3b). When MB-ODN 4531(O) or Lipoplex(O) was applied in vivo, Lipoplex(O) induced secretion of IL-12 and IFN-γ for much longer periods than MB-ODN 4531(O) did ( Figure  3c). It may be related with the enhanced stability of MB-ODN 4531(O) in vivo when encapsulated in DOPE: CHEMS complex. As expected, the immunomodulatory activity of Lipoplex(O) was entirely TLR9-dependent; cytokine production was not found in TLR9-/-mice ( Figure  3d). These results confirm that cytokine secretion was increased by Lipoplex(O) in a CpG sequence-and TLR9dependent manner.

Adjuvant activity of CpG-DNA encapsulated in DOPE: CHEMS complex
To evaluate the adjuvant activity of CpG-DNA encapsulated in liposomes, we immunized BALB/c mice with different combinations of hen egg lysozyme (HEL), MB-ODN 4531(O), and DOPE:CHEMS complex ( Figure 4). Immunization of BALB/c mice with HEL only or with HEL encapsulated in DOPE:CHEMS complex did not induce HEL-specific IgG. Immunization with a mixture of HEL and MB-ODN 4531(O), there was increased production of HEL-specific total IgG, which can be resulted from the adjuvant effect of MB-ODN 4531(O). Most importantly, immunization with a complex consisting of HEL and Lipoplex(O) (Lipoplex(O) + HEL) resulted in a markedly greater abundance of HEL-specific total IgG than other mice. Amount of HEL-specific total IgG induced by a complex of HEL and Lipoplex (O) was approximately four times higher than that induced by a mixture of HEL and MB-ODN 4531(O) (Figure 4a). Titers of HEL-specific total IgG also show similar results ( Figure 4b). Next, we confirmed the IgG isotype of anti-HEL-specific antibodies. Remarkably, immunization with a complex of HEL and Lipoplex(O) or a mixture of HEL and MB-ODN 4531(O) induced more IgG2a than IgG1, indicating that MB-ODN 4531 (O) affects isotype profile independently of encapsulation ( Figure 4a). However, immunization with a mixture of HEL and incomplete Freund's adjuvant (IFA) induced higher level of IgG1 than IgG2a. Based on the result from the LipoplexGC(O) plus HEL, we confirmed that Lipoplex(O)-mediated HEL-specific antibody production is CpG sequence-dependent which is in accordance with Table 1 Abbreviations used in this study. the data in Figure 1. When we performed the same experiments with another antigen ovalbumin (OVA), Lipoplex(O) plus OVA appears to induce more IgG2a and hence a more balanced response than IFA plus OVA as shown in Figure 4c and 4d.
Th1-dominated humoral immune response induced by peptide epitope and CpG-DNA encapsulated in DOPE:

CHEMS complex
As the potent adjuvant effect of Lipoplex(O) was identified in Figure 4, we applied this system to the screening of B cell epitopes against tumor-specific antigen. Specifically, we aimed to screen B cell epitopes originated from human tetraspanin transmembrane 4 superfamily member 5 of human HCC (hTM4SF5) (GenBank: NP_003954) [23]. Six candidate peptide sequences were selected from extracellular domain of hTM4SF5 protein on the basis of their hydrophilicity, hydrophobicity, secondary structure, antigenicity index as described in "Methods" (Table 2).
First, we screened the B cell epitopes of hTM4SF5 protein. We analyzed the amounts of IgG in the serum of BALB/c mice injected i.p. with complexes of each predicted hTM4SF5 peptide and Lipoplex(O). As shown in Figure 5a, the production of peptide-specific IgG was most enhanced by a complex consisting of hTM4SF5R2-3 peptide and Lipoplex(O). Anti-hTM4SF5R2-3-specific antibodies were mostly composed of IgG2a (Figure 5b and 5c). We evaluated the kinetics of IgG production in immunized mice in response to the complex of hTM4SF5R2-3 peptide and Lipoplex(O). The mice produced larger amounts of peptide-specific IgG (IgG2a) during the secondary and tertiary responses ( Figure 5d). When the mice were immunized with hTM4SF5R2-3 peptide and MB-ODN 4531(O) co-encapsulated in other liposomes, the production of peptide-specific IgG was much lower, which demonstrates the superior effect of Lipoplex(O) (Figure 5e). However, MB-ODN 4531GC (O) containing a reversal sequence of the CG dinucleotide lost its ability to stimulate production of IgG (Figure 5f). IgG production by hTM4SF5R2-3 peptide and Lipoplex(O) was dependent on TLR9, whereas IgG production by antigen and IFA was not ( Figure 6). These results indicate that the complex of hTM4SF5R2-3 peptide and Lipoplex(O) without carriers significantly enhanced production of peptide-specific IgG2a in a CpG sequence-dependent and TLR9-dependent manner.

Functional effect of anti-hTM4SF5 monoclonal antibody on hepatocellular carcinoma cells in vitro
The expression of TM4SF5 induced uncontrolled growth of human HCC cells via loss of contact inhibition [23]. To evaluate the functional effects of IgG produced by immunization with the complex of hTM4SF5R2-3 peptide and Lipoplex(O), we screened monoclonal antibody against hTM4SF5R2-3 peptide and confirmed its specificity by competition ELISA using hTM4SF5R2-3 peptide (Figure 7a). We then analyzed the expression of hTM4SF5 in human HCC cell lines and evaluated the monoclonal antibody-mediated hTM4SF5 targeting in the cells. The expression of hTM4SF5 was detected in Huh-7 and SNU-761 cells at  the mRNA by RT-PCR ( Figure 7b) and protein levels by immunoblotting and FACS analysis with the hTM4SF5specific monoclonal antibody (Figure 7c and 7d). To investigate the effect of the anti-hTM4SF5R2-3 peptide monoclonal antibody on cell growth, we performed MTT assay. The growth of Huh-7 cells expressing hTM4SF5 was markedly delayed by antibody treatment. However, there was no significant reduction in the growth of SNU-739 cells, which did not express  hTM4SF5 (Figure 8a). The population of cells in S phase was also reduced in Huh-7 cells but not SNU-739 cells (Figure 8b) To further investigate the effects of the anti-hTM4SF5R2-3 peptide antibody on DNA synthesis, the BrdU incorporation assay was performed. As shown in Figure 8c, the rate of DNA synthesis in Huh-7 cells treated with anti-hTM4SF5R2-3 peptide monoclonal antibody was weakly affected one day after treatment and further suppressed in three days by approximately 40% compared to normal IgG treated-Huh-7 cells (Figure 8c). In contrast, there was no effect in SNU-739 cells. Next, we investigated whether the anti-hTM4SF5R2-3 peptide antibody induces apoptotic response of Huh-7 cells. As shown in Figure 8d, the treatment of anti-hTM4SF5R2-3 peptide antibody did not lead to an increase in the number of apoptotic cells containing double-stranded DNA breaks in Huh-7 cells or SNU-739 cells. Therefore, we conclude that the anti-hTM4SF5R2-3 peptide antibody has anti-proliferative effect rather than proapoptotic effect in Huh-7 cells. These results confirm that the monoclonal antibody produced by immunization with hTM4SF5R2-3 peptide and Lipoplex(O) complex was useful for hTM4SF5 protein detection and antibody-mediated hTM4SF5 targeting in HCC cells.

Discussion
The application of CpG-DNA to therapeutics can be optimized via sequence rearrangement and backbone modification [25,26]. The recognition of CpG-DNA by TLR9 varies from mice to human immune cells due to a 24% difference at the amino acid level between the TLR9 molecules of mice and humans [8,19]. PS-ODN stimulates the human and mouse immune systems in several ways: efficient uptake into cells, endosomal acidification and maturation, and enhanced recognition by TLR9 in endosomes and lysosomes [24]. However, PO-ODN, which exhibits immunomodulatory activity in mice, shows poor activity in hPBMCs [8,27]. Recently, the cationic liposome DOTAP was used as a vehicle for delivery; it was found to improve the uptake of PO-ODN as well as the production of IFN-α in hPBMCs [20]. Here, we examined the effects of liposome-encapsulated PO-ODN on cytokine promoter activation and cytokine production in mouse and human cells. After stimulation with PO-ODN encapsulated in liposomes such as DOTAP, DOPE:CHEMS complex, and DSPC:CHEMS:PEG-PE complex, the IL-8 promoter was significantly activated in human and mouse cells and maximum activity was seen with PO-ODN encapsulated in DOPE:CHEMS complex (Lipoplex(O)) ( Figure 1). Recent studies by several investigators have shown that CpG-DNA has functional effects as a Th1-responsive adjuvant and that its potent adjuvant effects are enhanced by liposome encapsulation [12,14]. In this report, we found that Lipoplex(O) surpassed IFA in terms of the potency of antigen-specific IgG production and Th1-type immune responses in mice ( The screening of synthetic peptide-based epitopes has been studied extensively for the purpose of developing therapeutic antibodies and vaccines that have potential as prophylactics for cancer and infectious diseases such as influenza virus, malaria, hepatitis B, and HIV [28][29][30]. Although peptide vaccines have been actively studied in various animal models, their efficacy in the treatment of humans is limited. To improve the efficacy of peptide vaccines, researchers have evaluated liposomes as a means of delivering vaccines [13,[31][32][33]. Adjuvants such as flagella [34] and CpG-DNA [13] have also been formulated in order to enhance the magnitude of the immune responses. In this study, we performed peptidebased epitope screening and used complex of synthetic peptide and Lipoplex(O) without carriers to produce antibodies that act against hTM4SF5 in human HCC. We successfully screened a potent B cell epitope (hTM4SF5R2-3 peptide) from hTM4SF5 ( Figure 5). We found that Lipoplex(O) surpassed IFA and other forms of liposome-encapsulated CpG-DNA in terms of the potency of peptide-specific IgG production and Th1type immune responses in mice ( Figure 5 and 6). Therefore, our data offer a powerful tool for practical use: Lipoplex(O) is a potentially applicable universal adjuvant for peptide-based epitope screening and antibody production, and it obviates the need for other adjuvants and carriers. However, further study is needed to investigate whether or not IgG production induced by a Control IgG + 0.5 g hTM4SF5R2-3 Anti-hTM4SF5R2-3 Ab + 1 g hTM4SF5R2-3 + 5 g hTM4SF5R2-3 + 10 g hTM4SF5R2-3 Conc. (ng/ml)    complex of B cell epitope and Lipoplex(O) is dependent on MHC type, CD4 + cells and Th1 differentiation and if a novel mechanism is involved ( Figure 9). Examination of the IgG isotype levels demonstrates that the peptide and Lipoplex(O) complex induced higher levels of IgG2a than IgG1, indicating a predominant Th1 response. It is well-known that differential induction of Th1-type or Th2-type immune response by a vaccine is important for protective immunity in specified infectious diseases. For example, several studies have reported that formalin-inactivated human respiratory syncytial virus (RSV) vaccine was not effective because of its ability to induce an allergy-like Th2 immune response against the virus infection [35][36][37]. Therefore, our formulation may be applied as an effective adjuvant for peptide-based RSV B cell epitope screening and effective RSV vaccine: the formulation may decrease susceptibility to RSV and induce Th1biased or balanced Th1/Th2 adaptive immune response leading to the production of neutralizing antibodies.
This study suggest that Lipoplex(O) is a potent adjuvant and that complexes of epitope and Lipoplex(O) without carriers are extremely useful for B cell epitope screening and peptide-specific antibody production. As our novel vaccine strategy using Lipoplex(O) and synthetic peptide does not require protein antigen production and carrier conjugation, it is cost effective and enables rapid application. For example, TM4SF5 protein has gained attention as a target for HCC therapy since it is overexpressed in HCC [23]. TM4SF5 is also important in HCC formation by inducing morphological elongation, epithelial-mesenchymal transition, abnormal cell growth in multilayers in vitro, and tumor formation in vivo. But there was no report regarding epitope-specific monoclonal antibody against TM4SF5 yet. As indicated in a detailed study, we successfully produced monoclonal antibody without production of recombinant protein.
The anti-hTM4SF5R2-3 peptide-specific antibody can detect native protein and induce functional changes in cells expressing TM4SF5 protein, which suggests possible application in therapeutics (Figure 7 and 8). We are currently investigating whether or not anti-hTM4SF5R2-3 peptide-specific antibody can reduce HCC tumor formation in a xenograft mouse model after Huh-7 cells have been injected. We applied our strategy to monomeric peptide and produced antibody recognizing native protein in this study. Whether our technology can also be applied to conformationally dependent epitopes or poorly immunogenic peptides is to be determined in the future.

Conclusions
Our overall results show that Lipoplex(O) is a potent adjuvant and that complexes of epitope and Lipoplex(O) without carriers are extremely useful for B cell epitope screening and peptide-specific antibody production. Therefore, our strategy may be promptly used for the development of therapeutic antibodies by rapid screening of potent B cell epitopes ( Figure 10). (4:6:0.06 ratio) were prepared as reported previously [38] with minor modifications. Briefly, DOPE and CHEMS were mixed in ethanol at a molar ratio of 1:1, or as indicated in the individual experiments, evaporated with nitrogen gas to make a solvent-free lipid film, and resuspended in a mixture containing equal volumes of watersoluble CpG-DNA and protein (or peptide), followed by vigorous stirring at room temperature for 30 min. After adjusting the pH to 7.0, the Lipoplex solution was slightly sonicated for 30 s with a sonicator. We then filtered the solution with a 0.22 μm filter and freeze-thawed it three times with liquid nitrogen.

Transfection and luciferase assays
Transfection and luciferase assays were performed as previously described [26].

Uptake of CpG-DNAs
RPMI 8226 cells and RAW 264.7 cells were plated in 24-well plates for 24 h and then treated with FITClabeled CpG-DNA or FITC-labeled CpG-DNA encapsulated in DOPE:CHEMS complex for 10 min. Intracellular uptake of CpG-DNA was measured by FACS analysis with a FACSCalibur cytometer (BD Biosciences, San Diego, CA, USA) as reported previously [39].

Measurement of cytokines
To measure cytokines, CpG-DNA (5 μg/ml) or CpG-DNA (5 μg/ml) encapsulated in DOPE:CHEMS complex was added to splenocytes or hPBMCs, followed by incubation at 37°C with 5% CO 2 for 24 h. The cytokine levels in the culture supernatant were then measured using commercially available ELISA kits (R&D Systems, Minneapolis, MN, USA). The amount of cytokines in the sera was quantified using a cytokine bead array kit for mouse inflammatory cytokines (CBA; BD Biosciences) and a FACSCalibur cytometer equipped with CellQuestPro and CBA software.

Selection and synthesis of B cell epitope peptides
The putative B cell epitope peptides of hTM4SF5 were selected on the basis of hydrophilicity values from Parker et al [40], surface accessibility values from Emini et al [41], β-turn region values from Chou and Fasman [42], and antigenicity index from Kolaskar and Tongaonkar [43]. The parameters were averaged over six amino acid residues and the regions above the threshold value 1.0 were chosen for each prediction factor (http:// tools.immuneepitpoe.org/main/index.html). To screen B cell epitope, we selected 6 candidates from the putative B cell epitope peptides, which are located in the extracellular domain of hTM4SF5 ( Table 2). The peptides were synthesized using an automated peptide synthesizer (Peptron III-R24, Peptron, Daejeon, Korea). The peptides were purified by reverse-phase HPLC (Prominence HPLC, Shimadzu Corp., Tokyo, Japan) to purity greater than 90%. The peptide was identified using a mass spectrometer (HP 1100 Series LC/MSD, Hewlett-Packard, Roseville, CA, USA).

Mice and immunization
Mice were maintained under specific-pathogen-free conditions. Production of mouse anti-hTM4SF5 monoclonal antibody On three occasions at 10 day intervals, BALB/c mice were injected i.p. with hTM4SF5R2-3 peptide (50 μg) and MB-ODN 4531(O) (50 μg) co-encapsulated in DOPE: CHEMS complex. In accordance with standard hybridoma technique, we then screened hybridoma cells that produce anti-hTM4SF5R2-3 peptide-specific monoclonal antibody [44]. Anti-hTM4SF5R2-3 peptide monoclonal antibody (IgG2a) was purified from the ascites fluid by protein A column chromatography.
Reactivity of mouse anti-hTM4SF5 monoclonal antibody: Competitive ELISA assay To measure the titers of anti-TM4SF5R2-3 peptide monoclonal antibody, we coated 96-well immunoplates with 10 μg/ml of human TM4SF5R2-3 peptides and then blocked them with PBST containing 1% BSA. The antibodies were added to the top row of each plate, and serial 1:4 dilutions in PBST were then placed into subsequent rows. For competition assay, diluted antibodies were preincubated with the indicated amounts of each peptide for 30 min and then added to the wells of each plate. The plates were incubated for 2 h at room temperature, washed with PBST, and then incubated with anti-IgG antibody conjugated with horseradish peroxidase for 2 h. The colorimetric assay was developed with a TMB substrate solution (Kirkegaard and Perry Laboratories), and we used a Spectra Max 250 microplate reader (Molecular Devices, Sunnyvale, CA, USA) to measure the absorbance at 450 nm.

Cell culture
The human B cell line RPMI 8226, mouse macrophage cell line RAW 264.7, and the human hepatoma cell lines Huh-7 and HepG2 were obtained from ATCC. The human hepatocellular cell lines SNU-398, SNU-423, SNU-739, and SNU-761 were obtained from the Korean Cell Line Bank (Seoul, Korea). The cells were maintained in RPMI 1640 medium containing 10% FBS. Human normal hepatocyte cells (Promo Cell, Heidelberg, Germany) were maintained as the vendor suggested. All cells were cultured at 37°C in an atmosphere of 95% air and 5% CO 2 .

BrdU proliferation assay
To assess the cell proliferation after treatment with anti-hTM4SF5 monoclonal antibody, Huh-7 and SNU-739 cells were seeded into a 96-well microplate at a density of 10 5 cells/well. The cells were treated with anti-hTM4SF5 monoclonal antibody (5 μg/ml) for the indicated time periods. Subsequently, cells were fixed, washed, and incubated with primary and secondary antibodies according to the instruction of CycLex BrdU Cellular ELISA Kit (MBL International, Woburn, USA). The immune complexes were developed with a substrate solution, and we used a Spectra Max 250 microplate reader (Molecular Devices, Sunnyvale, CA, USA) to measure the absorbance at 450 nm.

Cell cycle analysis
To analyze DNA content, we incubated the cells in a solution of propidium iodide (20 μg/ml) in PBS containing RNase (200 μg/ml). The cells were stained at room temperature for 30 min and analyzed using FACSCalibur cytometer (BD Biosciences).

TUNEL assay
We performed terminal deoxynucleotidyl transferasemediated dUTP nick-end labeling (TUNEL) assay in accordance with the manufacturer's specifications (Promega, Madison, WI, USA). Briefly, Huh-7 and SNU-739 cells were placed in a complete medium for 24 h before being treated with anti-hTM4SF5 monoclonal antibody (5 μg/ml) for 3 days. The cells were fixed with 4% paraformaldehyde and permeabilized with 0.2% Triton X-100 for 5 min at room temperature, and end-labeled with TdT using the DeadEnd™ Fluorometric TUNEL system. The DNase I treatment served as a control of TUNEL-positive cells. After washing with PBS-T, cells were observed with FACSCalibur cytometer (BD Biosciences).