SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Concanavalin A (ConA), a lectin with mannose specificity that can induce acute hepatic inflammation, was tested for its therapeutic effect against hepatoma. ConA is cytotoxic or inhibitory to hepatoma cells, which is mediated by the autophagic pathway through mitochondria. Once it was bound to cell membrane glycoproteins, the ConA was internalized and preferentially localized onto the mitochondria. The mitochondria membrane permeability changed, and an autophagic pathway including LC3-II generation, double-layer vesicle, BNIP3 induction, and acidic vesicular organelle formation was induced. Either 3-MA or siRNA for BNIP3 and LC3, but neither beclin-1 nor ATG 5, partially inhibited the ConA-induced cell death. In addition to the autophagy induction, ConA is known to be a T cell mitogen. Using an in situ hepatoma model, ConA can exert an anti-hepatoma therapeutic effect, inhibiting tumor nodule formation in the liver and prolonging survival. Conclusion: ConA can be considered as an anti-hepatoma agent therapeutically because of its autophagic induction and immunomodulating activity. This dual function of ConA provides a novel mechanism for the biological effect of lectin. (HEPATOLOGY 2007;45:286–296.)

Lectins are carbohydrate-binding proteins that occur throughout the biosphere. They bind carbohydrates reversibly and possess the ability to agglutinate cells or precipitate polysaccharides and glycoconjugates. They have strong mitogenic activity to lymphocytes.1–3 Lectins have been used to differentiate malignant tumors from benign and the degree of glycosylation associated with metastasis.4, 5 Lectins have also been adopted for alternative tumor therapy in Europe.6–8

Hepatocellular carcinoma is the predominant cause of cancer mortality in males of Southern China and Taiwan. The current therapy for HCC is not satisfactory,9, 10 and more studies are needed to develop a more effective treatment of HCC. We are interested in developing new therapies for liver tumor and in testing whether lectins have any therapeutic effect for liver cancer. Concanavalin A (ConA) is a T-cell mitogen and has been used to induce hepatitis in mice through the triggering of natural killer (NK) T cells and subsequent activation of CD4+ T cells.11, 12 Autophagy is an evolutionarily conserved lysosomal pathway involved in cytoplasmic homeostasis to control the turnover of long-lived proteins and can be stimulated in response to different situations of stress, such as starvation, oxidative stress, irradiation, or even anti-cancer cytotoxic drugs.13 We reported here that ConA could also directly cause tumor cell death in an autophagic manner. The in vivo anti-hepatoma effect of ConA was further demonstrated in a recently established murine in situ hepatoma model.14

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Reagents and Antibodies.

ConA, ConA-FITC, ConA-beaded agarose and 3-methyladenine (3-MA) were purchased from Sigma (St. Louis, MO). The antiserum against microtubule-associated proteins light chain 3 (LC3), beclin-1, autophagy-related protein 5 (ATG 5), and phosphorylated AKT (p-AKT) were purchased from Santa Cruz (Santa Cruz, CA). Anti-AKT, anti-caspase 3 and anti-apoptosis-inducing factor (AIF) antibodies were purchased from Cell Signaling (Beverly, MA). The antibody against Bcl-2/adenovirus E1B 19 kd-interacting protein 3 (BNIP3), or endonuclease G (Endo G) was purchased from Sigma. Anti-poly(ADP-ribose) polymerase, COX IV, and GAPDH antibodies were purchased from Pharmingen (San Diego, CA), Abcam (Cambridge, UK), and Ambion (Austin, TX), respectively. RNA interferences of AIF, beclin-1, ATG 5, LC3 and unrelated control sequence were performed using small interfering RNA (siRNA) purchased from Santa Cruz. siRNA for BNIP3 was purchased from Invitrogen (Carlsbad, CA).

Cell Lines and Mice.

The BALB/c hepatoma cell line ML-1 was provided by Dr. C. P. Hu (Veterans General Hospital, Taipei, Taiwan).15 ML-14a cells were adapted from ML-1 cells in BALB/c mice for four generations. BALB/c mice (male, 8-10 weeks old) were purchased from National Laboratory Animal Center (Taipei, Taiwan); NOD/LtSz-PrkdcJ (SCID) mice were provided by the Animal Center of Tzu-Chi University (Hualien, Taiwan) and maintained in the pathogen-free facility of the Animal Laboratory of National Cheng Kung University. The animals were raised and cared for according to the guidelines set up by the National Science Council, ROC. The mouse experiments were approved by the institutional animal care and use committee.

Cell Growth Assay.

ML-14a, CT-26, Huh-7 and HepG2 cells were seeded at 1 × 105 cells in 12-well plates. The cells were treated with ConA in the presence or absence of methyl-αD-mannopyranoside (0.1 M). After incubation, viable cell number was measured by counting viable cells with Eosin Y exclusion staining or3H-thymidine incorporation assay. In inhibition experiments, the cells were pre-treated with zVAD-fmk (100 μM) or various concentrations of 3-MA 1 hour before addition of ConA (40 μg/ml). After incubation, cell growth was measured by methyl-thiazol-tetrazolium (MTT) assay. Cell death was determined by staining with propidium iodide and analyzing by flow cytometry.

Flow Cytometry Assay.

Cells (1 × 105) were incubated with 5 μg/ml FITC-conjugated ConA for 30 minutes at 37°C. The binding of ConA to cells was detected by flow cytometry. Methyl-αD-mannopyranoside (0.5 M) was used to block the specific binding. The acidic vesicular organelles were detected by acridine orange (1 μg/ml). The development of acidic vesicular organelles was analyzed. The mitochondrial membrane potential (Δ Ψm) was measured by Rhodamine 123 (5 μM).

Microscopic Examination.

ML-14a cells (5 × 105) were transfected with pLC3-GFP (5 μg).16 using Lipofectamine 2000 (Invitrogen). The fluorescence of LC3-GFP was observed under a fluorescence microscope (Olympus IX 70, Melville, NY). To localize the cellular distribution of ConA, ML-14a cells treated with ConA-FITC were further stained with 1 μg/mL MitoTracker red or LysoTracker red (Molecular Probe, OR) and fixed with 3.7% formaldehyde. In the parallel experiments, the cells treated with ConA-FITC were stained with primary antibody against calnexin followed by secondary antibody conjugated with Alexa 594. The distribution of ConA within cells was analyzed by confocal microscopy (Olympus FV 300, Japan). To observe the autophagic vesicles, the ML-14a cells were treated with ConA, fixed with 4% glutaraldehyde, and postfixed in 1% OsO4. The cells were observed under the electron microscopy (Hitachi 7000, Japan).

Western Blot Analysis.

Cell lysates were prepared by extracting proteins with lysis buffer. Proteins were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred to PVDF membranes. The membranes were blocked and incubated with primary antibodies. After incubation with peroxidase-conjugated secondary antibodies, the blots were visualized by enhancing chemiluminescence reagents (PerkinElimer Life Sciences, Boston, MA). In some experiments, the nuclear and cytoplasmic fractions of cells were prepared using the kit “NE-PER nuclear and cytoplasmic extraction reagents” (Pierce, Rockford, IL).

Murine In Situ Hepatoma Model.

A murine in situ hepatoma model was set up by intrasplenic injection of 1 × 106 viable ML-14a cells in 0.1 ml DMEM into anesthetized mice (Pentobarbital, 50 mg/kg intraperitoneally).14 The ML-14a first colonized in the spleen and then migrated into the liver, forming liver tumor nodules of varied sizes beginning 1 week after injection. At 30 days after intrasplenic injection, the livers of hepatoma-bearing mice were removed to determine the numbers and sizes of the tumor nodules.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

ConA Binding to Hepatoma Cells Induced Growth Inhibition and Cell Death.

The binding of FITC-ConA on 4 tumor cell lines was tested using flow cytometry. ConA could bind to 3 hepatoma cell lines—ML-14a, Huh-7, HepG2—and colon cell line CT-26 in a mannose-specific manner, which was blocked by the ligand methyl-αD-mannopyranoside (Fig. 1A). The binding of ConA to ML-14a, CT-26, Huh-7, and HepG2 tumor cell lines inhibited their cell replication in a dose-dependent manner (Fig. 1B). The addition of the methyl-αD-mannopyranoside also reversed the ConA-induced growth inhibition. The sensitivity to ConA-induced cell growth inhibition varied among four tumor cell lines: HepG2 was the most sensitive, followed by CT-26, ML-14a, and Huh-7. The IC50 of ConA for HepG2, CT-26, ML-14a, and Huh-7 were 5, 10, 10, and 20 μg/ml, respectively. The growth inhibition at dose > 20 μg/ml was due to the induction of cell death as the cell viability decreased along the incubation time of ConA (Fig. 1C). However, at the low dose of 1 to 5 μg/ml, the cell proliferation was inhibited as well, shown by the significant inhibition of [3H]thymidine incorporation on ML-14a cells (Fig. 1D).

thumbnail image

Figure 1. ConA–induced hepatoma cell growth inhibition through mannose binding moiety. (A) ConA bond to hepatoma cells (ML-14a, Huh-7, HepG2, and CT-26) through mannose moiety that is inhibited by methyl-α-D-mannopyranoside. (B, C) ConA inhibited hepatoma cell growth (B), induced cell death (C), and inhibited cell proliferation (D). *P < 0.05 versus PBS control.

Download figure to PowerPoint

ConA-Induced Autophagic-like Cell Death of Hepatoma Cells.

The mechanism of ConA-induced cell death was studied. ConA-treated cells were Annexin V-positive, but no DNA ladder or typical apoptosis were observed (data not shown). We went further to test whether ConA could induce another type of cell death, the autophagic cell death. As shown in Fig. 2A, ConA-treated LC3-GFP–transfected ML-14a cells showed the LC3-GFP aggregation as early as 6 hours posttreatment, suggesting that autophagy occurred after ConA treatment. The signal molecules involved in the process of autophagy are shown in Fig. 2B. The autophagic marker LC3-II was formed at 12 hours posttreatment. In addition, de novo synthesis of LC3-I was also found after ConA treatment. Another autophagic molecule, beclin-1, was not increased after ConA treatment, probably because of its abundance in ML-14a cells. Instead it was slightly decreased at the late 36th hour. The BNIP3 either monomer or dimer form were induced at 6 hours posttreatment. Phosphorylated AKT was also downregulated, indicating that the growth signal of AKT was altered after ConA treatment. The long stable COX-IV was decreased at 12 hours posttreatment. Under electronic microscopic observation on the ConA-treated ML-14a cells, the double-layer vesicle and many autophagosomes were detected at 12 hours posttreatment (Fig. 2C). The lysosomal activity as detected by acridine orange staining was also increased after ConA treatment (Fig. 2D). ConA-induced autophagic-like cell death was not only observed in ML-14a cell line; ConA-treated Huh-7, HepG2, and CT-26 cells also showed enhanced lysosomal activity and LC3-II formation (Fig. 2D,E). We further used the autophagy inhibitor 3-MA, a class III-PI3K inhibitor, to inhibit the pre-autophagosome formation. By MTT assay, different doses of 3-MA could partially block the ConA-induced growth inhibition (Fig. 3A). The induction of LC3-I and LC3-II formation and the autophagic cell death were also partially inhibited by 2 mM of 3-MA (Fig. 3B,C). The ConA-induced cell death was caspase-independent with no cleavage caspase-3 being detected, which is not attributable to the defect of caspase-3 in ML-14a cells because the etoposide can induce caspase-3 activation (Fig. 3D). The pan-caspase inhibitor zVAD-fmk had no effect on ConA-induced cell death even at a high dose of 100 μM (Fig. 3E). Based on these data, we concluded that ConA could induce hepatoma cells to undergo autophagic-like cell death.

thumbnail image

Figure 2. ConA-induced autophagy in hepatoma cell lines. (A) ConA-induced LC3 aggregation in ML-14a cells. The LC3-GFP aggregation of ConA-treated ML-14a cells was visualized at the indicated times by fluorescence microscopy. (B) Analysis of the signal molecules involved in autophagic pathway. ML-14a cell lysates were collected at the indicated times after incubation with ConA (40 μg/ml) and detected with anti-LC3, anti-beclin-1, anti-BNIP3, anti-phosphorylated AKT, anti-AKT, or anti-COX-IV antibodies. (C) Electron micrographs (×12,000, ×30,000) of ConA-treated ML-14a cells. The autophagic vesicles (arrows) with double-layer (low left-corner) were shown. (D) ConA-induced hepatoma cells to form acidic vesicular organelles stained with acridine orange. (E) ConA-induced four tumor cell lines to undergo LC3-II conversion.

Download figure to PowerPoint

thumbnail image

Figure 3. 3-MA partially blocked the ConA-induced LC3-II conversion and cell death. (A) ML-14a cells were pretreated with various concentration of 3-MA 1 hour before ConA (40 μg/ml) treatment. The cell growth was determined by MTT assay. (B, C) 3-MA (2 mM) inhibited LC3-II conversion and cell death. A representative Western blot is shown, and the mean ± SD of three experiments are indicated. (D, E) Caspase-3 was not activated during ConA-induced ML-14a cell death. zVAD-fmk was used as pan-caspase inhibitor. The means ± SD of 3 experiments are shown. *P < 0.05 versus control.

Download figure to PowerPoint

ConA Was Internalized to Induce Autophagy.

ConA binds to cell surface glycoprotein by its mannose-binding activity, but apparently the surface binding is not enough to induce autophagic-like cell death because the immobilized ConA to bead would not induce growth inhibition or LC3-II formation on ML-14a cells (Fig. 4A,B). We then used the FITC-ConA to track its cellular localization (Fig. 4C). The soluble ConA was first bound to cell membrane, then internalized and accumulated primarily onto the mitochondria as early as 1 hour posttreatment. Some of the labeled ConA was also found in the lysosome at 3 hours, but not in the endoplasmic reticulum. The internalized ConA preferentially bound to the mitochondria and would gradually increase mitochondria membrane permeability change as shown in Fig. 4D. The Endo G or AIF in the mitochondria would be released and translocated into the nucleus, but only at the late 24 hours after ConA treatment (Fig. 4E). The ConA after endocytosis would bind to mitochondria to cause mitochondria dysfunction and induce autophagy. However, the ConA-induced ML-14a cell death was only slightly inhibited by silencing AIF expression with siRNA (Fig. 5A). This indicates that AIF nuclear translocation only partially contributes to ConA-induced ML-14a cell death. Furthermore, the siRNA for beclin-1, ATG 5, or LC3 was tested in ConA-induced ML-14a cell death. The LC3-II conversion induced by ConA was slightly inhibited by beclin-1 siRNA, but not by ATG 5 siRNA. However, ConA-induced cell death was not affected by either beclin-1 or ATG 5 siRNA (Fig. 5B,C). Nevertheless, ConA-induced LC3-II conversion and cell death were blocked by LC3 siRNA, suggesting that LC3 plays a major role in ConA-induced cell death (Fig. 5D). Because the BNIP3 was formed at 6 hours, earlier than LC3-II conversion as shown previously, two different siRNA for BNIP3 were then used to knock-down the BNIP3 formation, and both the LC3-II conversion and cell death were inhibited (Fig. 5E). On the contrary, the LC3 siRNA would not affect the BNIP3 induction, indicating that the mitochondria BNIP3 was upstream of LC3-II. A mitochondria autophagic pathway was induced after ConA treatment.

thumbnail image

Figure 4. ConA effect was mediated by internalization and targeted to mitochondria. (A,B) Immobilized ConA failed to induce ML-14a cell death and LC3-II conversion. Various concentrations of soluble or beaded-agarose ConA were incubated with ML-14a cells for different times, and the cell proliferation was evaluated by MTT assay. (C) ConA was internalized and bound to cytoplasmic organelles. ML-14a cells were incubated with ConA–FITC (40 mg/ml) at 37°C for 1 or 3 hours. The cells were stained with MitoTracker red, LysoTracker red, or anti-calnexin antibody as indicators of mitochondria, lysosome, or endoplasmic reticulum (ER), respectively, then observed under a confocal microscope. ConA preferentially bound to the mitochondria. (D) ConA caused mitochondrial membrane potential (MMP) reduction in ML-14a cells. ConA-treated ML-14a cells were stained with Rhodamine 123 to detect the MMP by flow cytometry at the indicated times. (E) AIF and Endo G were translocated into nucleus at late 24 hours after ConA treatment. N, nuclear; C, cytoplasmic.

Download figure to PowerPoint

thumbnail image

Figure 5. Autophagy-related proteins and BNIP3 regulated the ConA-induced LC3-II conversion and cell death. ML-14a cells were treated with siRNAs (100 nM) of (A) AIF, (B) beclin-1, (C) ATG 5, (D) LC3, (E) BNIP3, or unrelated sequences (control) for 16 hours and incubated with ConA (40 μg/ml) for another 24 hours. The expression of AIF, Beclin-1, ATG 5, LC3-I/II, and BNIP3 were analyzed and quantified by immunoblotting analysis, and normalization to β-actin control was expressed. The ConA-induced ML-14a cell death was also determined after silencing of autophagy-related proteins and BNIP3 expression by the PI stain on flow cytometry. A representative Western blot is shown, and the mean ± SD of 2 to 3 experiments are indicated.

Download figure to PowerPoint

ConA Inhibited Liver Tumor Nodule Formation In Vivo.

The direct in vivo effect of ConA was further tested on liver tumor nodule formation in severe combined immune deficiency (SCID) mice. The SCID mice were inoculated by intra-spleen injection of 1 × 106 ML-14a cells, which would colonize first in the spleen and then migrate to the liver to form tumor nodules in liver. At 1 week post inoculation, ConA was given to hepatoma-bearing mice intravenously twice at 3-day intervals. We found that ConA at a dose of 20 mg/kg inhibited liver tumor formation in SCID mice at 21 days after tumor inoculation (Fig. 6A), whereas the doses of 10 and 15 mg/kg had no effect. This indicates that ConA has a direct inhibitory effect on liver tumor nodule formation independent of lymphocyte activation. ConA is known to be a strong T-cell mitogen, so we tested its therapeutic effect on immunocompetent BALB/c mice. The intra-spleen injection of 1 × 106 ML-14a cells established the tumor nodule formation in the liver. One week after inoculation, the hepatoma-bearing mice were treated intravenously with ConA (7.5 mg/kg body weight) twice at 3-day intervals. Liver tumor nodule formation was inhibited significantly (Fig. 6B,C). The control mice had 150 tumor nodules of varying sizes, whereas the ConA-treated mice had only 40 tumor nodules 30 days after tumor injection. The numbers of large tumor nodules (1-4 mm or >4 mm) decreased dramatically. Approximately 30% to 40% of the mice were tumor-free. In the survival experiment, the survival of the hepatoma-bearing mice was prolonged from 40 to 70 days after ConA treatment, whereas 20% to 30% of the mice were cured (Fig. 6D). Because ConA (7.5 mg/kg) was given only twice at days 7 and 10, the residue tumors would continue growing and kill the non-tumor-free mice. When we increased the dose of ConA and the number of injections, for example, to 20 mg/kg and 4 times, the liver tumor nodules could be completely eradicated (data not shown). The histological tissue staining showed many lymphocytic infiltrations around the liver tumor nodule after ConA injection (Supplementary Fig. 1; available at: http://interscience.wiley.com/jpages/0270-9139/suppmat/index.html). The ConA-activated lymphocyte would infiltrate into the liver to kill the hepatoma cells. However, there was no apparent hepatocytes damage at the sub-optimal dose of 7.5 mg/kg used, and the serum ALT was not elevated during the treatment course (Supplementary Fig. 2). ConA seems to be less cytotoxic to normal hepatocytes in tumor-bearing mice than in naïve mice, probably because of more binding to tumor cells than to normal hepatocytes. The cells that participated in this inhibition were further tested with in vivo depletion of CD4+ or CD8+ T cells (Fig. 6E). Depletion of CD8+ T cells blocked the anti-tumor effect of ConA. More tumor nodules were present in CD8+ T-depleted mice than in non-depleted mice, suggesting that CD8+ cells played a major role in the ConA-mediated anti-hepatoma activity. Depletion of CD4+ T cells also partially affected the ConA anti-tumor activity. In the control group, however, CD4+ T depletion also partially inhibited the tumor formation in the PBS-treated group. The role of CD4+ T cells in anti-hepatoma activity needs further investigation. Based on the above data, we conclude that ConA, a mannose-specific binding lectin, can induce autophagic-like cell death of hepatoma cells in vitro and also can exert its direct cytotoxicity in vivo. Furthermore, because of its immunomodulating activity, ConA is more potent in immunocompetent mice and can have anti-hepatoma activity through the activation of lymphocytes. Therefore, it can be considered for use as an anti-hepatoma agent therapeutically.

thumbnail image

Figure 6. ConA inhibited liver tumor nodule formation in vivo. (A) ConA inhibited liver tumor nodule formation in SCID mice. Intraspleen inoculation of 1 × 106 ML-14a established tumor nodule formation in the liver of SCID mice. ConA given twice at 3-day intervals beginning at day 7 inhibited tumor nodule formation (n = 7). At day 21, the number and size of the tumor nodules in the liver were determined. The experiment was repeated twice with the same results. (B-D) ConA inhibited liver tumor nodule formation in BALB/c mice. Intra-spleen inoculation of 1 × 106 ML-14a established tumor nodule formation in the liver of BALB/c mice. ConA (7.5 mg/kg) was given twice at 3-day intervals beginning at day 7. At day 30 or 32, the number and sizes of the tumor nodules in the liver were determined. ConA caused hepatoma regression in the liver of 1-week-long hepatoma-bearing mice (B,C, n = 7), and prolonged the survival of the tumor-bearing mice (D, n = 10). The experiments were repeated three times with the same results. *P < 0.05 versus PBS control. (E) ConA-mediated liver tumor nodule inhibition is CD8+ T cell dependent. The CD4+ or CD8+ lymphocytes were depleted by anti-CD4 (GK1.5) or anti-CD8 (2.43) MAb 7 days before tumor inoculation and every 7 days thereafter during the experimental period. One week after intraspleen inoculation, the hepatoma-bearing mice were treated intravenously with ConA (7.5 mg/kg body weight) twice at 3-day intervals. At day 30, the number and size of the tumor nodules in the liver were determined (n = 8).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Lectins such as ConA are known to be T-cell mitogens and are cytotoxic to cells.17 Using its unique characteristics of the sugar-binding specificity, we found that ConA can induce differential effects on hepatoma cells and lymphocytes; the induction of autophagy on hepatoma cells, and activation of lymphocytes. Lymphocytes are more sensitive to ConA than hepatoma tumor cell lines, probably because of their high content of mannose- or glucose-containing moieties on the cell membrane. ConA at doses of 1 to 10 μg/ml is mitogenic but becomes toxic at doses above 50 μg/ml. Autophagic-like cell death induction was observed at doses higher than 20 μg/ml, and its sensitivity varied on different cell lines.

The mechanism of ConA-induced growth inhibition or cytotoxicity is delineated as the autophagic pathway, primarily mediated through the mitochondria. ConA after binding to cell membrane would be internalized and accumulated preferentially in to the mitochondria. The mitochondria membrane permeability was changed. Although the Endo G or AIF in the mitochondria would release and translocate into the nucleus later, the typical caspase-dependent apoptosis was not observed. However, the siRNA for the AIF slightly inhibited ConA-induced cell death, indicating that AIF translocation into the nucleus only partially contributes to the death process. After onset of mitochondria permeability transition induced by stimulus, autophagy would be activated to lead to lysosomal degradation of the affected mitochondria and cell repair.18 The autophagic pathway molecules including LC3-II formation, double-layer vesicle, BNIP3 induction, and acidic vesicular organelle formation were demonstrated in this study. The class III-PI3K inhibitor 3-MA partially inhibited the ConA-induced autophagic-like cell death. The BNIP3, an adenovirus E1B 19-kDa interacting protein that is a cell death–inducing factor and a member of the BH3-only subfamily of Bcl-2 family protein, was detected at 6 hours, earlier than the formation of LC3-II at 12 hours. A major function of this class of protein is to determine the on/off state of the mitochondria permeability transition pore.19 Recently, a short mitochondria form of p19ARF that localized in the compartment of the mitochondria could reduce mitochondria membrane potential and then caused autophagy and caspase-independent cell death.20 We found that the siRNA for BNIP3 and LC3, but neither beclin-1 nor ATG 5, could inhibit the ConA-induced cell death. The siRNA of BNIP3 could knock down the BNIP3 induction as well as the LC3-II conversion, whereas the siRNA of LC3 inhibited only the LC3-II, but not BNIP3 induction. We did not observe the inhibition of the ConA–induced ML-14a cell death by the siRNA for beclin-1 or ATG 5, indicating that it is not a typical or global autophagic cell death. Instead, a BNIP3-mediated mitochondria autophagy was induced after ConA treatment. The autophagy induced by the ConA would provide a novel mechanism for the lectin.

The liver can trap blood-borne foreign substances, and its anatomic location makes it a good site to concentrate ConA. ConA trapped in the liver binds to hepatocytes through its specific binding activity with mannose residue on cell membrane glycoproteins. ConA accumulated in the liver then directly induces hepatic inflammation to cause liver tumor regression. ConA not only activates immune cells, but also agglutinates then to hepatoma cells in the liver. The inflammatory response initiated by the ConA would subsequently induce the adaptive immune response against the transplantation tumor. The activated CD8+ T cells became the major effector cells to kill the tumor cells. This process would lead to the establishment of the tumor-specific memory immunity to protect from further challenge with the same tumor (unpublished observation).

Recently, Miyagi et al.21 reported the anti-tumor effect of ConA via NK cell- and IFN-γ–dependent manners in the CT-26 hepatic metastasis model. ConA has been reported to be hepatocyte-cytotoxic.17, 22 In this study, we further reported that ConA can induce hepatoma cells to undergo autophagy, and the direct anti-tumor effect was observed on hepatoma-bearing SCID mice. However, the use of ConA as an anti-hepatoma agent is dependent on the time, dose, and frequency of administration. ConA can induce acute hepatitis in a naive mouse at a dose of 15 to 20 mg/kg by activating NK T and CD4+ T cells.11, 12 The hepatotoxic dose is usually higher than 15 to 20 mg/kg. At the dose of 7.5 mg/kg used in this study, no elevation of serum ALT in hepatoma-bearing mice was observed (Supplementary Fig. 2); however, when the ConA dose was increased to 20 mg/kg, the serum level of ALT was slightly elevated, but it was still far less than that in the naïve mice. There seems to be a therapeutic window for tumor-bearing mice compared with the naïve mice. The infiltrating T cells are found spread around the whole liver section, but more T cells deposited on the tumor parts (Supplementary Fig. 1). Many reports have suggested that malignant transformation is associated with various and complex alterations in the glycosylation process.4, 5 This carbohydrate-differential expression can be used to develop tumor-preferential killing for a more invasive and metastatic phenotype with some particular lectins.23, 24 Although this result seems promising, there is a caveat for the continuing injection of ConA in vivo. Mice would become insensitive to ConA after repetitive challenge of ConA, the mice would develop T-cell anergy via production of IL-10 after multiple administrations of ConA.25 We also detected anti-ConA antibodies after five or six injections of ConA (unpublished observation). The generation of anti-ConA antibody, which is a common problem encountered in the use of protein drugs, will restrict the continuing use of ConA. To avoid this phenomenon, sequential usage of various lectins that have similar effects could be adopted.

Cancer immunotherapy with tumor-associated antigens has been debated recently.26, 27 Our result showing that mannose-specific lectin, which bears both immunomodulating and autophagy-inducing activity, can have anti-cancer activities would open up new possibilities for natural anti-cancer compounds.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

The authors thank Drs. N. Mizushima and T. Yoshimori for providing the GFP-LC3 plasmid, and Michel Theron for editorial assistance.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Supplementary material for this article can be found on the H EPATOLOGY website ( http://interscience.wiley.com/jpages/0270-9139/suppmat/index.html ).

FilenameFormatSizeDescription
jws-hep.21509.fig1.pdf212KSupplementary fig 1. Histopathology of the liver of hepatoma-bearing-mice treated with Con A. Intra-spleen inoculation of 1106 ML-14a established tumor nodule formation in the liver of BALB/c mice. Con A (7.5 mg/kg) was administrated intravenously at day 7, and the liver was removed at 2 days post Con A treatment and stained with H&E. (a) Na?ve. (b) PBS-treated. (c&d) Con A-treated. The arrow points to lymphocyte infiltrations around the tumor nodule.
jws-hep.21509.fig2.pdf212KSupplementary fig 2. Con A is less cytotoxic to hepatocytes in tumor-bearing mice than in na?ve mice. Intra-spleen inoculation of 1106 ML-14a established tumor nodule formation in the liver of BALB/c mice. On day 7, these hepatoma-bearing mice (n=5) were received with PBS or Con A intravenously twice at 3-day intervals. The liver injury was determined by serum level of alanine transaminase (ALT) at indicated time post Con A/PBS injection. (a) Kinetic changes of ALT post 7.5 mg/kg Con A treatment. (b) ALT level at 24 h post different doses of Con A treatment. The na?ve mouse treated with Con A was used as a positive control.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.