Liver cancer is one of the most malignant cancers in the world and has a high rate of metastasis. Therefore, development of a novel therapy for liver cancer is a critical issue. Indoleamine 2,3-dioxygenase (IDO) is known as a negative immune regulator in dendritic cells. Our previous study demonstrated that skin delivery of IDO short hairpin RNA (shRNA) induced antitumor immunity in subcutaneous bladder and colon tumor models. Because the immunological environment is quite different between skin and liver, it is essential to evaluate whether skin delivery of IDO shRNA is an effective treatment in metastatic and orthotopic animal tumor models. In the present study, IDO shRNA inhibited tumor growth in subcutaneous, metastatic and orthotopic liver tumor models. The cytotoxicity of splenocytes was significantly elevated in mice treated with IDO shRNA in the orthotopic and metastatic tumor models. Interleukin (IL)-12 and interferon (IFN)-gamma mRNA expression were upregulated while IL-10 was downregulated in the inguinal lymph nodes, which were collected from IDO shRNA-treated mice. Similar results were observed in the spleens of mice inoculated with IDO shRNA by gene gun. The results indicate that skin administration of IDO shRNA is an effective therapy in orthotopic and metastatic liver cancer animal models. Indoleamine 2,3-dioxygenase shRNA might be a potential new treatment for liver cancer in the future. (Cancer Sci 2011; 102: 2214–2220)
Liver cancer is the fifth most common cancer in the world and the third leading cause of cancer-related death.(1,2) The incidence of liver cancer is still rising in the United States and some areas of Asia.(3) The 5-year survival rate of patients is relatively low and many patients die because of recurrence and metastasis.(4) In addition, advanced liver cancer has a poor response to radiotherapy and conventional chemotherapy.(5,6) Treatment of 5-fluorouracil, adriamycin or doxorubicin provided little survival benefit.(7) Therefore, development of novel and effective therapeutic approaches is a critical issue in the management of liver cancer.
An inhibitor of multi-targeted tyrosine kinase, sorafenib, was shown to increase the survival rate.(8) Other new drugs targeting vascular epithelial growth factor (VEGF), epithelial growth factor receptor (EGFR) or the mammalian target of rapamycin (mTOR) pathway are also being evaluated in several clinical studies.(9–11) Immunotherapy provides another kind of potential treatment for liver cancer. Some antigens including alpha-fetoprotein, SSX-2, MAGE-A and telomerase reverse transcriptase have been used to activate cytotoxic T cells against liver cancer.(12–15) The patients in the clinical trials showed partial or complete responses to various immunotherapies.(16) However, activation of immune cells in these studies usually requires time-consuming procedures ex vivo.
Indoleamine 2,3-dioxygenase (IDO) is a rate-limiting enzyme of the tryptophan metabolic pathway(17) and plays a negative role in immune regulation. Indoleamine 2,3-dioxygenase-expressing dendritic cells are observed in tumor-draining lymph nodes.(18) IDO+ plasmacytoid dendritic cells promote differentiation from CD4+ T cells to regulatory T cells in tumor-draining lymph nodes.(19,20) In addition, IDO is demonstrated to be essential for human plasmacytoid dendritic cell-induced adaptive T regulatory cell generation.(21) Blockage of IDO by IDO-specific inhibitor is effective in inducing antitumor immunity.(22–24) Altogether, silencing IDO expression of dendritic cells is probably beneficial for treating liver cancer.
Previously, we have delivered short hairpin RNA (shRNA) of IDO by skin administration to delay tumor progression in multiple subcutaneous tumor models,(25) suggesting that the strategy might be useful in treating liver cancer. This gene gun approach provides a proof of concept that we can manipulate immune responses against cancer without ex vivo manipulation of dendritic cells. However, the immunological environment of tumor formed in subcutaneous tissue is quite different from that of tumors formed in orthotopic sites.(26) An orthotopic tumor animal model is more relevant to the cancer that develops in human.(27) Therefore, it is necessary to evaluate the therapeutic effect of IDO shRNA in an orthotopic liver tumor animal model. Because there is a high risk of metastasis observed in liver cancer,(28,29) we further examined whether skin administration of IDO shRNA can inhibit experimental metastasis.
In the present study, we demonstrated that skin administration of IDO shRNA not only delayed tumor growth in a subcutaneous tumor model, but also suppressed tumor growth in metastatic and orthotopic tumor models. Cytotoxicity of lymphocytes was induced after delivery of IDO shRNA. In addition, Th1 cytokines were upregulated in the mice vaccinated with IDO shRNA. These results suggest that this strategy might be a potential therapy for liver cancer in the future.
Materials and Methods
Animals. Inbred female BALB/c mice (8–10 weeks old) were obtained from the Laboratory Animal Center at National Cheng Kung University (Tainan, Taiwan). All mice study protocols were approved by the Animal Welfare Committee at National Cheng Kung University.
Cell culture. Mouse ML1-4a hepatoma cell line, which was a kind gift from Dr Huan-Yao Lei,(30) and mouse MBT-2 bladder cancer cell line(31) were maintained in low-glucose Dulbecco’s modified Eagle’s medium (DMEM) containing 10% fetal bovine serum, 100 units/mL penicillin and 100 μg/mL streptomycin (Invitrogen, Carlsbad, CA, USA). The ML1-4a-luciferase and MBT-2-luciferase stable clones were established by transfection with cytomegalovirus (CMV) promoter-driven luciferase plasmid (Promega, Madison, WI, USA) and selected with 400 μg/mL of G418 (Sigma-Aldrich, St. Louis, MO, USA). Cells were cultured at 37°C with 5% CO2 incubator.
DNA construction and plasmid preparation. Indoleamine 2,3-dioxygenase shRNA and scramble IDO shRNA have been described previously.(25) Plasmid DNA was purified using the Endofree Qiagen Plasmid Mega Kits (Qiagen, Chatsworth, CA, USA) and was resuspended in sterile water at a concentration of 1 μg/μL.
Animal tumor models. For the subcutaneous tumor model, 2 × 106 ML1-4a cells were injected into mice subcutaneously. For the metastatic hepatoma models, ML1-4a or ML1-4a-luciferase cells (5 × 105 cells in 200 μL DMEM) were intravenously injected into mice. For the orthotopic hepatoma model, mice were anesthetized with a mask using 1.5–2% isoflurane (Halocarbon Laboratories, North Augusta, SC, USA). ML1-4a or ML1-4a-luciferase cells (1 × 105 cells in 50 μL DMEM) were injected into the left liver lobe of mice and tumor cells were sealed using bipolar coagulation. To evaluate the therapeutic effect of IDO shRNA, all mice were bombarded with IDO shRNA at day 7 after tumor challenge.
Delivery of shRNA to abdominal skin by gene gun. The abdominal region of mice was shaved before gene gun delivery. Tumor-bearing mice were bombarded with 10 μg of plasmid DNA (20 μL of sterile water) through a low-pressure gene gun (BioWare Technologies Co. Ltd, Taipei, Taiwan) at 50 psi of helium gas pressure.(25) The subcutaneous and orthotopic tumor-bearing mice were bombarded once a week, whereas the metastatic tumor-bearing mice were bombarded twice a week.
Evaluation of therapeutic efficacy. The subcutaneous tumor was measured using a caliper two times a week. Tumor size was determined using the formula: volume = A2 × B × 0.5236, in which A and B represent the shortest and longest diameter. Mice were killed when the tumor volume was larger than 2500 mm3. In the metastatic and orthotopic tumor models, whole lung weight and liver weight were measured. The weight of the lung and liver was measured at day 28 post-injection.
In order to measure the growth of ML1-4a-luciferase cells in vivo, mice were intraperitonally injected with 100 μL d-luciferin in saline at a dose of 100 mg/kg (Synchem OHG, Altenburg, Germany). The tumor size was indicated by luciferase, which was visualized using a Night Owl imaging unit (Berthold Technologies, Bad Wildbad, Germany) at day 21 for metastatic tumor-bearing mice and day 28 for orthotopic tumor mice.
India ink staining. Mice were killed at 21 days after ML1-4a injection. Then, 2 mL of 15% India ink dye was intratracheally injected into the lung. The lung was washed in water for 5 min and then bleached in Fekete’s solution (70% ethanol, 3.7% paraformaldehyde and 0.75 M glacial acetic acid) overnight.
Immune cell infiltration of tumor. Tumor samples were collected 1 week after the third vaccination in the orthotopic tumor model. Immunohistochemistry was performed on 5 μm cryostat sections. Anti-CD4 and CD8 antibodies (BD Pharmingen, San Diego, CA, USA) were used to detect infiltrated T cells, which were counted randomly in five fields of each sample from at least three independent samples.
In vitro cytotoxicity assay. Spleen cells were collected at day 28 after tumor injection in both tumor models. Then, 2 × 107 spleen cells were incubated with 10 μg tumor lysate of ML1-4a in 1 mL of RPMI 1640 with 25 mM HEPES and l-glutamate, penicillin (100 U/mL), streptomycin (100 μg/mL), 10% FBS (Invitrogen) and 50 mM 2-mercaptoethanol (Sigma-Aldrich) for 48 h. Next, 1 × 104 cells ML1-4a-luciferase, which served as target cells, were co-incubated for 6 h with serial dilution (50:1, 25:1, 12.5:1) of effector spleen cells in 200 μL medium. For in vitro depletion of CD8 T cells or natural killer (NK) cells, 1 × 107 spleen cells were incubated with anti-CD8 monoclonal antibody (10 mg/mL, clone 2.43), anti-Asialo GM1 antibody (1:100 dilution; Wako Pure Chemicals, Osaka, Japan) or control antibody (10 mg/mL, purified rat IgG) at 37°C for 30 min before cytotoxicity assay. The lysis of target ML1-4a-luciferase or MBT-2 luciferase cells were measured in the supernatant using the luciferase detection system (Promega) in the luminometer (Lumat LB9506; Berthold Technologies).
Sorting of plasmacytoid dendritic cells. Plasmacytoid dendritic cells were isolated from spleens in orthotopic tumor-bearing mice at day 28 post-injection using anti-mouse PDCA-1 magnetic microbeads (Miltenyi Biotec, Auburn, CA, USA) according to the manufacturer’s instruction.
Reverse transcription PCR and real-time PCR. At day 28 post-tumor injection, tumor, spleen and inguinal lymph nodes were collected from the IDO shRNA-vaccinated mice. Total RNA of lymphocytes was extracted from cells using TRIZOL (Invitrogen). cDNA synthesis was performed by MMLV-Reverse Transcriptase (Promega) according to the manufacturer’s directions. Primer sequences are in Supporting Information Table S1. Real-time PCR was performed on an Applied Biosystems 7900HT real-time PCR instrument (Applied Biosystems, Foster City, CA, USA) using FastStart Universal SYBR Green I Master (Roche Diagnostics, Mannheim, Germany). The cycling conditions were 10 min of 95°C, 45 cycles at 95°C for 15 s and 60°C for 60 s.
Graph and statistical analysis. All numerical data and graphs were analyzed with GraphPad Prism version 4.00 for Windows (GraphPad Software, San Diego, CA, USA). Student’s t-test was used for analysis of the difference between experimental groups.
Indoleamine 2,3-dioxygenase shRNA delayed liver tumor growth in the subcutaneous, orthotopic and metastatic tumor models. Skin administration of IDO shRNA was shown to have an antitumor effect on the murine colon tumor model and bladder tumor model,(25) but its therapeutic effect has not been evaluated in the murine liver tumor model. We first determined the therapeutic effect of IDO shRNA in the subcutaneous liver tumor model. Mice were implanted with ML-1 murine tumor cells subcutaneously and the tumor-bearing mice was immunized with IDO shRNA by gene gun at weekly intervals 7 days after implantation. Treatment of IDO shRNA delayed the growth of ML1-4a murine liver tumor cells in BALB/c mice (Fig. 1A). The results suggest that vaccination of IDO shRNA had a therapeutic potency on liver cancer cells in the subcutaneous tumor model (Fig. 1B). Therefore, we further evaluated the therapeutic efficacy in the metastatic and orthotopic tumor models.
In the metastatic tumor model, the mice were intravenously injected with ML1-4a cells and vaccinated with IDO shRNA, scramble IDO shRNA or ddH2O twice a week. Skin administration of IDO shRNA inhibited formation of tumor nodules in the lung (Fig. 2A) and the associated decrease of total lung weight (Fig. 2B). Metastatic liver cancer in the lung was also monitored using bioluminescent imaging in mice challenged with ML1-4a-luciferase cells. A significant reduction in luciferase activity was observed in mice vaccinated with IDO shRNA by in vivo imaging, indicating inhibition of metastasis (Fig. 2C,D).
In the orthotopic animal liver tumor model, ML1-4a or ML1-4a-luciferase cells were injected into the left liver lobe of mice to form liver tumors. All mice were bombarded with IDO shRNA once a week starting from day 7 post-injection. The mean tumor weight of mice treated with IDO shRNA is less than that of mice treated with scramble IDO shRNA or ddH2O after the third vaccination (Fig. 3A,B). The tumor size of ML1-4a-luciferase cells was also monitored by in vivo imaging analysis. Mice vaccinated with IDO shRNA had a lower luciferase activity than the control groups of mice (Fig. 3C,D). Altogether, IDO shRNA induced an antitumor effect in three tumor models.
Indoleamine 2,3-dioxygenase shRNA induced tumor-infiltrating T cells and cytotoxicity against tumor cells. To investigate whether delivery of IDO shRNA suppressed IDO expression of dendritic cells, we analyzed the mRNA level of IDO in orthotopic tumor-bearing mice. Indoleamine 2,3-dioxygenase shRNA did not downregulate IDO expression of whole splenocytes (Fig. 4A). However, IDO was significantly suppressed after IDO shRNA vaccination in plasmacytoid dendritic cells (Fig. 4B). We further determined whether lymphocyte cytotoxicity was activated by IDO shRNA in metastatic or orthotopic tumor-bearing mice. Compared with ddH2O and the scramble IDO shRNA group, the splenocytes from mice treated with IDO shRNA exerted higher cytotoxicity toward ML1-4a in both tumor models (Fig. 5A,B). In addition, ML1-4a cells were lysed by IDO shRNA-activated splenocytes but MBT-2 was not affected (Fig. 5C). Moreover, depletion of CD8+ T cells abolished the cytotoxic activity (Fig. 5D). In contrast, depletion of NK cells did not attenuate cytotoxicity (Fig. 5E). These results indicate that IDO shRNA vaccination induced specific cytotoxicity of splenocytes against ML1-4a tumor cells in a CD8+ T-cell-dependent manner.
Because CD8+ T cells were essential for eradication of tumor, we further analyzed the number of infiltrated immune cells in liver tumor nodules by immunochemical analysis. The number of infiltrated CD4, CD8 and NK cells in the IDO shRNA group is more than that of the scramble IDO shRNA and ddH2O groups (Table 1). These data indicate the IDO shRNA vaccination through the abdominal skin induced infiltration of immune cells in an orthotopic tumor model. Increased infiltration of immune cell infiltration might be associated with a therapeutic effect.
Table 1. Immune cell infiltration in orthotopic tumor
The total number of infiltrated T cells was counted from five random fields under a ×400 light microscope. The mean ± SD shown is the average of three independent experiments. *P <0.05, versus the ddH2O group. IDO, indoleamine 2,3-dioxygenase; NK, natural killer cells; shRNA, short hairpin RNA.
3 ± 3
2 ± 2
0 ± 0
Scramble IDO shRNA
2 ± 2
1 ± 1
1 ± 1
13 ± 5*
14 ± 2*
15 ± 6*
Indoleamine 2,3-dioxygenase shRNA induced Th1-biased immune responses in the lymph node and spleen. Dendritic cells have been shown to be activated by bombardment with gold-coated DNA and migrate to inguinal lymph nodes.(32) Our previous study also demonstrated that the enhanced green fluorescence protein (EGFP)-carrying dendritic cells mainly migrate to the inguinal lymph node after EGFP plasmid bombardment.(33) To determine whether IDO shRNA treatment influenced immune polarization in the lymphoid organ, we further determined the Th1/Th2 cytokines of immune cells in the spleen and lymph node. Higher levels of interleukin-12 and interferon-γ mRNA were detected in the spleen which was collected from mice vaccinated with IDO shRNA. In contrast, the mRNA level of interleukin-10 significantly decreased in the IDO shRNA vaccination group (Fig. 6A). The same pattern of Th1-biased cytokine expression was observed in lymphocytes (Fig. 6B). In addition, similar results were obtained in the orthotopic tumor model (Fig. 6C,D) and the subcutaneous tumor model (data not shown). The results suggest that vaccination of IDO shRNA induced Th1 polarization in the spleen and inguinal lymph node of tumor-bearing mice in all three tumor models.
Traditional treatment has shown limited responses on liver cancer; therefore, it is important to develop novel therapeutic approaches. Several targeted therapeutic strategies have been evaluated in mouse tumor models and clinical trials. Blockage of the BRAF/VEGFR/PDGFR or EGFR signaling pathway by inhibitors gave positive results in clinical trials.(8,10) Suicide gene therapy was effective in murine liver cancer models.(34,35) In addition, interleukin 12 treatment is able to suppress hepatocellular carcinoma in several mouse tumor models.(36,37) In the present study, we demonstrated that suppression of IDO in dendritic cells delayed tumor progression in subcutaneous, orthotopic and metastatic tumor models. Increased infiltration of T cells in orthotopic tumor was observed in IDO shRNA-treated mice. Vaccination of IDO shRNA induced a Th1-biased cytokine profile and lymphocyte cytotoxicity against the ML1-4a tumor cells. Our previous study and this study demonstrated that gene gun delivery of naked DNA vaccine is able to induce Th1-biased immune responses, which is usually required for antitumor cytotoxic responses.(33) These results suggest that skin administration of IDO shRNA is a potential treatment for liver cancer.
Indoleamine 2,3-dioxygenase-expressing plasmacytoid dendritic cells activate regulatory T cells in tumor draining lymph nodes.(19) Blockage of IDO activity by systemic IDO inhibitor induces antitumor efficacy.(22) We have further demonstrated that administration of IDO shRNA by gene gun activated plasmacytoid dendritic cells in the spleen and induced antitumor immune responses. These dendritic cells might then be primed with circulating tumor cells or tumor antigens in skin tissue and induce a systematic antitumor immunity. Antitumor immunity is not only effective in suppressing the orthotopic liver tumor, but also effective in inhibiting the proliferation of circulating tumor cells in an experimental metastatic animal model. Indoleamine 2,3-dioxygenase shRNA induced CD8+ T-cell-mediated cytotoxicity (Fig. 4). However, the number of CD4+ FoxP3+ regulatory T cells did not significantly decrease in IDO shRNA vaccinated inguinal lymph nodes and spleen (data not shown). A previous study indicated that inactivation of IDO in cancer cells with IDO inhibitor increased the proliferation and activity of immune cells.(22) In the present study, IDO mRNA expression in orthotopic tumor cells and subcutaneous tumor cells was not altered after IDO shRNA vaccination (data not shown). The results suggest that dendritic cells were the major target of IDO shRNA in our experiments.
Indoleamine 2,3-dioxygenase and IL-6 are two molecules that play a role in the myeloid-derived suppressor cell (MDSC)-mediated immunosuppression pathway in the tumor microenvironment.(38) Indoleamine 2,3-dioxygenase and IL-6 might generate a feedback loop to expand the population of MDSC.(39) The feedback loop might not be involved in mediating antitumor immunity in our experimental model because IL-6 expression was not suppressed in the inguinal lymph node, spleen and tumor after IDO shRNA vaccination (Figs 4,S1). In contrast, Th1 cytokines were observed in the spleen and lymph nodes in orthotopic and metastatic animal tumor models, implying activated skin dendritic cells might cause activation of immune responses in the spleen in addition to the nearby lymphoid organ. In our experiments, shRNA is delivered without coating on gold particles by gene gun. Naked DNA might induce a Th1-biased immunological response, which might overcome MDSC-recruiting cytokines and the suppressive microenvironment in the spleen.
The combination of IDO shRNA and Her2/neu DNA vaccine can further boost the cancer therapeutic effect in a subcutaneous animal tumor model.(25) Because overexpression of neu is not frequently observed in human liver cancer, other tumor antigens might be used for combined studies in the future. Several antigens such as α-fetoprotein, SSX-2, MAGE-A and telomerase reverse transcriptase activate tumor-specific CD8+ T cells in liver cancer.(12–15) Co-administration of IDO shRNA and DNA vaccine targeting these tumor antigens might provide an even stronger therapeutic effect. In contrast, the combination of IDO shRNA and novel target therapy drugs might also be a feasible approach for treating liver cancer. Sunitinib is a small molecule targeting VEGFR-1 and VEGFR-2 to inhibit angiogenesis, and does not affect the induction of antigen-specific T cells.(40) The combination of Sunitinib and DNA vaccine might provide another novel treatment for liver cancer through inhibition of angiogenesis and elevation of immune responses.
In summary, IDO shRNA induced antitumor immunity through skin delivery in orthotopic and experimental metastatic animal models. Tumor-specific cytotoxicity is significantly evaluated and the Th1 cytokine profile is observed. Antitumor immune responses can be activated through non-invasive shRNA delivery on local skin without ex vivo manipulation, which might provide a new potential treatment against liver cancer.
This study was supported by grants NSC99-2323-B006-004 and NSC97-2320-B006-003-MY3 from the National Science Council, Taiwan.