SEARCH

SEARCH BY CITATION

Keywords:

  • dendritic cells;
  • siRNA;
  • IDO;
  • breast cancer

Abstract

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Cancer immunotherapeutic agents (vaccines) in the form of antigen-loaded dendritic cells (DCs) reached an important milestone with the recent approval of Provenge, the first DC vaccine for treatment of prostate cancer. Although this heralds a new era of tumor immunotherapy, it also highlights the compelling need to optimize such DC-based therapies as they are increasingly tested and used to treat human patients. In this study we sought to augment and enhance the antitumor activity of a DC-based vaccine using siRNA to silence expression of immunosuppressive enzyme indoleamine 2,3-dioxygenase (IDO) in DCs. We report here that DCs loaded with tumor antigens, but with siRNA-silenced IDO expression, were introduced into 4T1 breast tumor-bearing mice, the treatment: (i) lengthened the time required for tumor onset, (ii) decreased tumor size compared to tumors grown for equal lengths of time in mice treated with antigen-loaded DCs without IDO silencing and (iii) reduced CD4+ and CD8+ T cell apoptosis. Furthermore, immunization with IDO-silenced DCs enhanced tumor antigen-specific T cell proliferation and CTL activity, and decreased numbers of CD4+CD25+Foxp3+ Treg. This study provides evidence to support silencing of immunosuppressive genes (IDO) as an effective strategy to enhance the efficacy of DC-based cancer immunotherapeutic.

Dendritic cells (DCs) act as central processing units of the immune system, using signals from the environment to either activate immune processes or promote tolerance of antigens. In the context of tumor immunotherapy and vaccine development, DCs are of particular interest because of their unique ability, when properly stimulated, to activate naïve T cells.1, 2 Ex vivo loading (pulsing) of DCs with tumor antigens followed by in vitro or in vivo activation has been demonstrated to successfully induce activity of cytotoxic T cells against foreign tumor cells, and tumor regression, in animal models and clinical trials.3–5 In human trials, heterogeneity of clinical responses has been associated with DC vaccination.6, 7 On the positive side, DC vaccine evoked specific cellular immune response and improved patient survival in certain patients.6, 8 On the negative side, some DC trials have actually demonstrated lack of positive response and even accelerated tumor growth.7

Activation of DCs is critical to stimulation of immunity. For example, we have previously reported that inhibition of IκB kinase (IKK) which is one of molecules in nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) signal transduction cascade in DCs leads to generation of tolerogenic DCs.9 Tolerogenic DCs express molecules, such as secreted transforming growth factor beta (TGF-β), interleukin 10 (IL-10), human leukocyte antigen G (HLA-G) and leukemia inhibitory factor (LIF) and nonsecreted programmed cell death 1 ligand 1(PD-1 L) and IDO, which promote tumor tolerance. In the context of cancer it is believed that tumors secrete signals that inhibit DC maturation and stimulate production of the same immunosuppressive molecules by DCs. Therefore, it is essential to describe new methods to engineer DCs that do not express immunosuppressive molecules, as a strategy to resist tumor-mediated reprogramming in response to treatment with anticancer vaccines. Given our experience with silencing immunostimulatory and immunosuppressive molecules by siRNA treatment of DCs,10, 11 we sought to use this technique to augment and enhance the antitumor activity of a DC-based vaccine.

Our target molecule for siRNA downregulation in these initial studies was the tryptophan-degrading enzyme indoleamine-2, 3-dioxygenase (IDO). IDO expression is associated with suppression of T cell responses because activated T cells are highly sensitive to local tryptophan depletion mediated by IDO, and catabolites produced as end products of tryptophan degradation by IDO induce T cell apoptosis.7, 12, 13 In addition, studies showing that IDO is essential for successful allogeneic pregnancy suggest that it is important in suppressing immune activity in the context of normal biological events.13, 14 In fact, inhibition of IDO has been show to induce immunologically mediated abortion. In the presence of tumors, IDO is found on tumor-associated macrophages and DCs15 and can actively promote tumor tolerance by stimulating generation of T regulatory (Treg) cells.16–19 DCs isolated from tumor-draining lymph nodes (TDLNs) of patients and mice have been shown to induce tumor tolerance.15, 20 Plasmacytoid DCs (pDCs present in both blood and peripheral lymphoid organs, including TDLNs) constitutively express immunosuppressive levels of IDO.21, 22 In vitro, pDCs isolated from TDLNs are potent suppressors of T cell responses to antigens. Although pDC is comprised of 0.5% of TDLNs, adoptive transfer of these pDCs from tumor-bearing mice into naive hosts created profound local T cell anergy.20 Depletion of IDO mRNA and subsequent protein in DCs by siRNA or by administration of the IDO inhibitor drug 1-methyl-D-tryptophan to recipient mice can reverse immunosuppression.20–22

In a previous study we demonstrated that systemically administrated IDO siRNA can reduce melanoma growth and enhance host antitumor immune responses in a murine model.11 In the present study we report on the effect of silencing IDO expression in DCs to generate a DC vaccine lacking immunosuppressive IDO, and specifically the antitumor effects of this novel DC vaccine on murine breast cancer growth.

Material and Methods

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

Animals

Female BALB/c (H-2Kd) mice 6 weeks of age were purchased from The Jackson Laboratories, (Jackson Laboratories, Bar Harbor, ME), and kept in filter-top cages at the animal care and veterinary services facility at the University of Western Ontario (UWO) according to Canadian Council for Animal Care Guidelines. All animal procedures were ethically approved by The UWO Animal Use Subcommittee.

Mouse cancer cells

The 4T1 breast cancer cells and B16F10 melanoma cells were purchased from American type culture collection (ATCC, Manassas, VA) and cultured in RPMI-1640 supplemented with 2 mmol/L L-glutamine, 100 U/mL penicillin, 100 μg streptomycin, 50 μmol/L 2-mercaptoethanol, and 10% fetal calf serum (Invitrogen).

Generation of bone marrow-derived dendritic cells

DCs were generated from bone marrow progenitor cells as previously described.23, 24 Briefly, bone marrow cells were flushed from femurs and tibias of BALB/c mice. Red blood cells were lysed using ACK lysis buffer and remaining bone marrow cells were cultured in six-well plates in complete medium (RPMI-1640) containing 2 mmol/L L-glutamine, 100 U/mL penicillin, 100 μg streptomycin, 50 μmol/L 2-mercaptoethanol, and 10% fetal calf serum; all from Invitrogen Canada: Life Technologies (Burlington, Ontario, Canada), supplemented with recombinant granulocyte/macrophage colony stimulating factor (GM-CSF), 10 ng/mL (PeproTech, Rocky Hill, NJ) and recombinant mouse interleukin 4 (IL-4), 10 ng/mL (PeproTech). All cells were incubated at 37°C in 5% humidified CO2. Medium was changed in all cultures every 2 days.

siRNA synthesis and transfection

An siRNA (5′GUUCUAGAAGGAUCCUUGA3′) targeting IDO mRNA was generated in according to the target sequence selection method described by Elbashir et al.25 siRNA was synthesized by the manufacturer (Dharmacon, Lafayette, CO). siRNA specific to the Luciferase gene GL2 Duplex (Dharmacon), which was not present in treated cells, thus was used as a sham-silencing control. Targeting and control siRNAs were transfected into DCs as described previously.10, 23, 26 Briefly, 2 μg of double-stranded siRNA was mixed with 5 μL of GeneSilencer™ transfection reagent (Gene Therapy Systems, San Diego, CA) and added onto DCs. After 4 hr, an equal volume of RPMI-1640 supplemented with 20% FCS and 200 ng mL−1 GM-CSF and IL-4 was added onto transfected DCs.

Reverse transcriptase-polymerase chain reaction (RT-PCR)

Total RNA was extracted from cells using Trizol reagent (Invitrogen, Burlington, Ontario, Canada). Total RNA (3 μg) was used to synthesize cDNA using oligodT and reverse transcriptase (Invitrogen). DNA primers used to amplify murine IDO and GAPDH16 were: IDO, 5′- GGGCTTTGCTCTACCACATCCACT-3′ (forward) and 5′-ACATCGTCATCCCCTCGGTTCC-3′ (reverse); glyceraldehyde-3-phosphate dehydrogenase (GAPDH), 5′-TGATGACATCAAGAAGGTGGTGAA-3′ (forward) and 5′-TGGGATGGAAATTGTGAGGGAGAT-3′ (reverse).

Conventional RT-PCR was performed at 94°C for 3 min followed by 30 cycles at 94°C for 30 sec, 58°C for 30 sec, and 72°C for 30 sec with a PCR thermal cycler (Eppendorf, Hamburg, Germany). PCR products were run on 1.5% agarose gel and visualized by ethidium bromide under a UV light.

Quantitative PCR (qPCR) was performed using a Stratagene MX 4000 PCR instrument (Stratagene, Cedar Creek, TX) and SYBR Green (Stratagene) according to the manufacturer's protocol. The data were used to estimate the amount of IDO mRNA present in samples, relative to GAPDH mRNA levels. The expression of IDO in IDO-siRNA treated DC was normalized to that of GL2 siRNA treatment.

Western blot

Cells were lysed in RIPA buffer (Cell signaling, Pickering, Ontario, Canada) and 30 μg of total protein was separated on a 12% SDS polyacrylamide gel and electrotransferred onto a nitrocellulose membrane. The membrane was blocked with TBS buffer containing 10% milk for 1 hr at room temperature and then incubated with anti-IDO (1:200, Millipore, Billerica, MA) or anti-actin (1:5,000, Santa Cruz biotechnology, Santa Cruz, CA) antibodies overnight at 4°C. Horseradish peroxidase-conjugated immunoglobulins were used as secondary antibodies (Santa Cruz). Proteins were detected using an ECL system (GE HealthCare, Quebec, Canada).

HPLC assay for tryptophan and kynurenine

Proteins in the culture media were removed as described previously.11 Briefly, 15% perchloric acid was added to cell culture media to precipitate proteins. About 10 μL of resultant supernatant was injected onto a Waters Acquity Ultra-performance Liquid Chromatograph (UPLC) CSH C18 column. Tryptophan was detected fluorometrically with the excitation and emission wavelengths 253 and 313 nm, respectively using an ACQUITY UPLC® fluorescence detector. Kynurenine was detected using an ACQUITY UPLC® photodiode array (PDA) detector set to 360 nm. Tryptophan and kynurenine were quantified using an external calibrator generated from standard curves.

Preparation of IDO-silenced DC vaccine

BALB/c bone marrow-derived DCs were cultured in the presence of 100 ng/mL recombinant murine GM-CSF and IL-4 by our standard method. Mouse 4T1 breast tumor cell lysate was used as tumor antigen and was prepared using a six-cycle freeze-thaw method.11 The 4T1 lysate was added to DC culture medium at a ratio of 1:3 (DCs/ 4T1 cells) and incubated for 24 hr to allow phagocytic loading of 4T1 lysate into DCs (antigen pulsing). Antigen-pulsed DCs were immediately transfected with IDO siRNA or control siRNA as described above. At 24 hr after siRNA transfection, DCs were stimulated to mature by adding LPS (10 ng mL−1) overnight. Control and IDO-silenced, antigen-pulsed mature DCs were collected for subsequent experiments.

Dextran phagocytosis assay

FITC-labeled-dextran (5 μg/mL) was added to DC culture for 1 hr at 37°C. Fluorescence of internalized dextran particles was detected by flow cytometry as a measure of DC capacity to take up antigen.

Mixed lymphocyte reaction

T cells were purified from the splenocytes of naïve C57BL/6 mouse using nylon wool columns, and were used as responders to DC activation (105 responder T cells/well). Allogeneic 8-day cultured and activated DCs (BALB/c origin) were used as the stimulators. A 72-hr mixed lymphocyte reaction (MLR) was performed and the cells were pulsed with 1 μCi of 3H-thymidine for the last 18 hr. The incorporation of 3H-thymidine was used to measure the proliferation of T cells. The cultures were harvested onto glass fiber filters (Wallac filtermat B, Wallac 1205-404). Radioactivity was counted using a Wallac Betaplate liquid scintillation counter.

CTL-mediated tumor cell lysis assay

The capacity of T cells from mice treated with DCs to lyse target 4T1 tumor cells was measured using a CytoTox 96 nonradioactive cytotoxicity assay kit (Promega, Madison, WI) according to the manufacturer's instructions. Briefly, naïve target 4T1 cells or melanoma cell line B16F10 cells were plated and incubated for 4 hr with CD8+ effector T cells isolated from 4T1 tumor-bearing mice using CD8+ magnetic microbeads (BD Bioscience, Mississauga, ON, Canada). The isolated CD8+ cells were confirmed, by flow cytometry, to be over 85% CD8+. A range of ratios of effectors to target cells was tested. Lactate dehydrogenase (LDH) release in response to effector T cells was measured in the buffer bathing unlysed target cells as a fraction of maximal LDH (LDH in equivalent numbers of unlysed target cells). Target cells incubated in the absence of effector cells were used as a comparator to control for spontaneous LDH release. Released LDH in culture supernatants was detected after a 30-min incubation using a coupled enzymatic assay. The intensity of the color formed was proportional to the number of lysed cells. Cytotoxic activity of CTL was calculated using the following formula: Cytotoxic activity % = [(absorbance) − (spontaneous effector cell LDH release) − (spontaneous target cell LDH release)]/[(maximal LDH release) − (spontaneous target cell LDH release)] × 100.

Administration of IDO-silenced DC vaccine to mice harboring 4T1 breast tumors

Cultured 4T1 cells were collected, suspended in PBS and injected (105 cells/100 μL/mouse) into the mammary fat pads of 6-week-old female BABL/c mice. On the day of 4T1 cell injection, or when tumors had grown to 5 mm in diameter, mice were also injected intravenously (i.v.) with 2 × 106 IDO-silenced, 4T1 Ag-pulsed DCs. DC injection was repeated three times (once per week for 3 weeks). Mice were checked for tumor formation on alternate days. When tumors appeared, tumor size was measured by a caliper and the volumes were estimated using the following formula: tumor volume = 0.5 × (tumor width)2 × (tumor length).

Tumor antigen-specific T cell response

T cell proliferation in response to tumor antigen in subsequent groups of mice was measured with a standard microtiter assay. T cells from lymph node or spleen were collected from mice treated with IDO-silenced DCs or control DCs, seeded in triplicate (5 × 105/well) in a 96-well-plate (Corning), in the presence of cultured BABL/c derived bone marrow DCs, and 4T1 breast cancer cell lysate (25 μg protein mL−1). After 72 hr, 3H-thymidine (1 μCi, Amersham) was added to each well. Cells were collected on glass fiber filters 18 hr later, using an automated cell harvester. 3H Radioactivity was counted using a Wallac Betaplate liquid scintillation counter.

Flow cytometry

Characterization of DCs was performed by flow cytometry. DCs were collected and stained (30 min, 4°C) with antibodies specific for CD11C, CD40, CD80, and MHC II. All antibodies were purchased from eBioscience (BD Biosciences Pharmingen, San Diego, CA). Phenotypic analysis of Treg cells was performed by staining with PE-Cy5-, PE-, or FITC-conjugated anti-mouse CD4, CD25 and Foxp3 mAbs, respectively (eBioscience). Foxp3 expression was assessed by intracellular staining using a cell permeabilization kit (eBioscience). All flow cytometric analyses were performed using appropriate isotype controls (Cedarlane Laboratories, Hornby, Ontario, Canada). All cells were washed and analyzed by flow cytometry using a Becton Dickinson FACScan flow cytometer (BD Biosciences, San Jose, CA).

Annexin-V staining

T cells were isolated from draining lymph nodes (LNs) from treated and control mice, stained with PE-labeled anti-CD4 or PE-Cy5 labeled anti-CD8 mAbs at 4°C for 30 min, washed once with PBS plus 2% FBS, suspended in 100 μL PBS containing 2% FBS, and 1 μL FITC-labeled Annexin-V (CalBiochem, San Diego, CA) was added and incubated at room temperature for 15 min. Flow cytometry was used to detect fluorescence indicative of Annexin-V positivity.

Inhibitory function of Treg cells

CD4+CD25+cells were isolated from draining LNs using Treg isolation kits (Stemcell Technologies, Vancouver, Canada). The capacity of Treg cells to inhibit T cell proliferation was determined as previously by us.27 Briefly, CD4+ T cells were isolated using CD4+ cell enrichment kits (Stemcell Technologies) following the manufacturer's instruction. The CD4+CD25+ subset of CD4+ T cells was isolated using a CD25+ cell isolation kit (Stemcell Technologies). Purified CD4+CD25+ T cells were added to ongoing tumor antigen-specific T cell response reactions (4 × 104 cells per well) in combination with whole splenic cells isolated from tumor-bearing mice (2 × 105 cells per well) and 4T1 tumor cell lysate (50 μg/mL). At 48 hr after incubation, 1 μCi 3H-thymidine was added to each well of the culture. Cells were continued to culture for another 18 hr. Cells were harvested and radioactivity measured using a Wallac Beta plate liquid scintillation counter (as described for mixed lymphocyte reactions).

Immunohistochemistry staining

Tumor tissue from IDO-silenced, 4T1 Ag-pulsed DCs or control DCs were snap-frozen using liquid N2 and sectioned by microtome. Frozen sections were dehydrated and fixed in acetone and then preincubated in 3% H2O2 to block endogenous peroxidase activity, followed by incubation with anti-mouse Foxp3 mAb (Biolegend, San Diego, CA). After washings (three times) in PBS/BSA, sections were incubated with biotinylated anti-rabbit IgG and then incubated with Streptavidin coupled to peroxidase. Sections were reacted with a diaminobenzidine/peroxide mixture and mounted hematoxylin.

Statistical analysis

An unpaired Student's t test and one-way ANOVA followed by a Newman–Keuls were used to test significance of differences between groups and among three groups, respectively. A p value <0.05 was chose a priori as the benchmark for significance of differences.

Results

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

siRNA-mediated IDO knockdown does not alter DC maturation

Under physiological conditions in vivo, immature DCs sample their environment by phagocytosing potential antigens at a low basal rate. Encounter with one or more activating molecules including Toll-like receptor (TLR) agonists such as poly(I:C) and LPS leads to DC maturation and migration to secondary lymphoid organs where mature DCs present acquired antigens to T cells.28 When presented antigens interact with an appropriate T cell receptor, productive immunity ensues. The basis of tumor immunotherapy is to mimic this process by loading antigens ex vivo in the expectation that Ag-pulsed DCs will induce antitumor immune activation, so it is critical that DC activities essential for in vivo immune activation functions remain intact. Because IDO acts to suppress immune activation potential of DCs, we sought to knock down IDO expression in those cells in the context of cancer.

To demonstrate the efficacy of siRNA-mediated IDO mRNA reduction, we cultured bone marrow-derived DCs from BALB/c mice and transfected DCs with IDO-targeting siRNA in vitro. RNA was extracted from DCs 24 hr after transfection and IDO mRNA levels. IFN-γ upregulated the expression of IDO in DCs (Fig. 1a, left panel and Fig. 1b). siRNA treatment reduced IDO mRNA by >80% (Fig. 1a, right panel). The gene silencing of IDO siRNA was further confirmed at the protein level by Western blot (Fig. 1b) and enzyme activity using HPLC (Fig. 1c). Next, we examined the phenotype of DCs with silenced IDO expression by flow cytometric assessment of CD11C, CD40, MHCII and CD80. The levels of CD11C, CD40, MHC II and CD80 in IDO-silenced DCs were similar or identical to that seen in DCs treated with nontargeting control siRNA (control DCs), indicating that IDO siRNA did not affect these characteristic determinants of DC phenotype (Figs. 1d and 1e).

thumbnail image

Figure 1. Silencing IDO with siRNA does not change DC maturation. (a) Silencing IDO in DCs: BALB/c-derived DCs were cultured for 6 days and transfected with siRNA targeting IDO or GL2 (control siRNA). About 6 hr after transfection, IFN-γ (200 U mL−1) was added to the culture. IDO mRNA levels were measured 24-hr post-transfection by conventional RT-PCR (left panel) and qPCR (right panel) as described in Material and Methods. (b) IDO expression at the level of protein. DCs were transfected with IDO siRNA and treated with IFN-γ as described above. At 48 hr after transfection with IDO siRNA, total protein were extracted from cells and separated by PAGE. IDO expression in DCs at the protein level was detected by Western blot. (c) Tryptophan and Kyurenine: Culture medium was collected from above transfected DCs culture (b) 48 h after transfection. The level of tryptophan and kynurenine in the medium were detected by HPLC as described in Material and Methods. (d) DC phenotype: DCs were cultured and transfected with IDO siRNA or control siRNA as in (a) and, 48 hr later, collected and stained with Abs against with mouse CD11C, MHC II, CD40 and CD80. Staining was assessed by flow cytometry. (e) Bar graph depicting the level of CD11C, MHC II, CD80 and CD40 expression in IDO or GL2 siRNA transfected DCs. *Different from cells treated with control GL2 siRNA (p < 0.05, student t test).

Download figure to PowerPoint

Knockdown of IDO in DCs does not affect the uptake of antigen but enhances allogeneic T cell response

Uptake and processing of antigen(s) by DCs, as cancer vaccine, is an essential first step in T cell activation. We therefore examined whether gene silencing influences DC capacity to capture antigenic molecules. Dextran phagocytosis measurement is a well-established surrogate measurement for capacity to phagocytose antigenic materials.29 FITC-labeled dextran (40 kD, green fluorescence) was added to DCs 24 hr after transfection of siRNA. DCs were then incubated at 37°C for 1 hr, and fluorescence of internalized dextran particles was detected by flow cytometry. IDO-silenced DCs had a similar capacity to take up FITC-dextran to control DCs (Fig. 2a, upper panel), indicating that IDO silencing with siRNA did not alter the capacity of DCs to take up antigenic molecules. After stimulation with LPS, the capacity of mature DCs to take up dextran was decreased in both IDO silenced DCs and control DCs. The degree of decrease was virtually the same in the two groups (Fig. 2a, lower panel).

thumbnail image

Figure 2. Knockdown of IDO in DCs does not affect the uptake of antigen but enhances allogeneic T cell response. (a) IDO-silenced DCs retain capacity to phagocytose: DCs were cultured and transfected with IDO siRNA (IDO-silenced DCs) or control siRNA (control DCs) as described in the legend to Figure 1 in the absence (upper panel) or presence of LPS (lower panel). FITC-labeled dextran was added 24 hr later and the amount phagocytosed by cells assessed by flow cytometry as described in Material and Methods. (b) IDO-silenced DCs enhanced allogeneic T cell response. BALB/c mice-derived DCs were transfected with siRNA as described in the legend to Figure 1 and allogeneic (C57B/6) T cells were incubated with siRNA-treated DCs at the indicated concentrations with or without incubation with kynurenine (50 μmol L−1) for 72 hr (+ Ky). Proliferation was determined by 3H-thymidine incorporation. Data are representative of three independent experiments.* (p < 0.05, one-way ANOVA).

Download figure to PowerPoint

IDO over-expressing DCs depleted tryptophan essential for T cell survival leading to inhibiting allogeneic T cell proliferation in vitro.30, 31 IDO-mediated inhibition of T cell proliferation is reversed by addition of tryptophan or a small molecule IDO inhibitor (1-MT).32 We next assessed the potential for knockdown of IDO to alter the antigen-presenting function of DCs. In a mixed lymphocyte reaction, IDO-silenced DCs increased their capacity to stimulate allogeneic T cell proliferation (by 22–26%, at three different ratios of DC: stimulatory T cells) compared with control DCs, or wild type DCs (Fig. 2b). Addition of kynurenine to the T cell culture reversed the effect of downregulating IDO gene expression (Fig. 2b).

siRNA-mediated IDO knockdown in DCs enhances antitumor effect of DC vaccine

To test whether IDO-silenced DC vaccine enhance the antitumor effect of DC vaccine, lysates of 4T1 cells were used as a source of mixed tumor antigens to load DCs subsequently used to treat mice injected with mouse breast tumor cells. Female BALB/c mice (n = 9 per group) were injected with IDO-silenced DCs (2 × 106 cells, i.v.), or control siRNA transfected DCs (control DCs), or DCs which were not pulsed with tumor antigen (No-Ag DCs) at the same time as injection of 4T1 cells into the mammary fat pad. Mice were then treated with IDO-silenced DCs, or control DCs, or No-Ag-DCs weekly for 3 weeks and tumor size was measured by caliper once tumors became visible. Formation of visible tumor was delayed in mice treated with IDO-silenced DCs or control DC, compared to mice untreated with DCs of any kind (Fig. 3a). Treatment with No-Ag DCs did not delay tumor formation (Fig. 3a). Tumor onset occurred on Day 9 in mice treated with control DCs compared to tumor onset by Day 5 in mice that did not receive DC vaccine, indicating that IDO expression in DCs did not completely abrogate antitumor activity. However, tumor onset in mice treated with IDO-silenced DCs did not occur until Day 16, an ∼50% greater delay than that mediated by control DCs (Fig. 3b). Tumors in mice treated with IDO-silenced DCs grew more slowly than those in mice untreated with DCs or treated with control DCs: on Day 16 of growth, tumors in mice treated with IDO-silenced DCs were 83% smaller than in mice treated with control DCs and nearly 95% smaller than in mice untreated with DCs. On Day 30, they were 35% smaller than the control DCs group and over 50% smaller than the untreated group. Therefore, treatment with IDO-silenced DCs significantly reduced tumor growth (Fig. 3b). The mean weight of tumors excised at the time of test animal sacrifice was lower in mice treated with IDO-silenced DCs than in mice treated with control DCs, No-Ag DCs, and was also lower than in mice untreated with DCs (Fig. 3c).

thumbnail image

Figure 3. Silencing IDO in DCs slows tumor progression. (a) Silencing IDO in DCs delays tumor onset. The 4T1 cells were injected into the mammary fat pad of female BALB/c mice (n = 9 per group) followed by injection of DCs transfected with IDO siRNA (IDO-silenced DCs) or control, nontargeting siRNA (control DCs) or DCs that were not loaded with tumor antigen (No-Ag-DCs) as described in Material and Methods. The day of that palpable tumors were detected was defined as the day of tumor onset. (b) Tumor growth: Mice (n = 9 per group) were treated as for (a) and tumor volume was estimated as described in Material and Methods. *Different from control DCs or wild type (WT) DCs (p < 0.05, student t test). (c) Tumor weight: Mice from the experiment described in (a) were challenged with 4T1 cells and treated with DC vaccine as in (a). Tumors were excised from mice and weighed 30 days after tumor cell injection. (d) Treatment of existing tumors with IDO-silenced DC vaccine: The 4T1 cells were injected into the mammary fat pad of female BALB/c mice (n = 9 per group). When tumors reached 5 mm in diameter, mice were treated with IDO-silenced DCs or control DCs transfected with nontargeting siRNA (control DCs). Tumor volumes were estimated at various times as described in Material and Methods. * p < 0.05, ** p < 0.001, one-way ANOVA.

Download figure to PowerPoint

Furthermore, we tested whether treatment with IDO-silenced DCs was effective when applied to established tumors. We allowed 4T1-derived tumors to grow to palpable size (tumor diameter = 5 mm) prior to treatment with the same number of IDO-silenced DCs or control DCs as described for treatment of nascent tumors (i.e., concurrent injection of DCs and tumor cells). Compared to control mice that did not receive DC treatment, growth of palpable tumors was partially suppressed in mice treated with either IDO-silenced DCs or control DCs (Fig. 3d). However, there was no significant difference between treatment with IDO-silenced DCs and with control DCs.

Silencing IDO in DCs reduces T cell apoptosis and increases cytotoxic T lymphocyte numbers and activity in vivo

To further characterize the mechanism underlying enhanced antitumor activity mediated by IDO-silenced DCs, we detected the apoptosis in the T cells from the TDLN of the mice treated with IDO-silenced DCs or control DCs. T cell apoptosis in these pooled cells (i.e., the fraction of CD8+ or CD4+ T cells actively undergoing apoptosis) was measured in vitro as described in Material and Methods. siRNA-mediated IDO silencing in DCs reduced the number of apoptosing CD8+ T cells and apoptosing CD4+ T cells by over 40% (Fig. 4a). In a tumor antigen-specific T cell response assay, T cells from mice treated with IDO-silenced DCs underwent ∼34% more proliferation in response to challenge with tumor cell lysate than mice treated with control DCs, and 136% more than those from mice untreated with DCs (Fig. 4b). Although treatment with control DCs also increased proliferative response to lysate (by ∼75%), IDO silenced DCs significantly enhanced their capacity to induce T cell responsiveness to antigen challenge. We further examined the cytotoxic activity of CTL by a CTL-mediated tumor cell lysis assay. CD8+ T cells were isolated from TDLNs from mice treated with IDO-silenced DCs or control DCs using CD8+ microbeads. CD8+ cells from mice treated with IDO-silenced DCs showed stronger activities of CTL against 4T1 cells than those isolated from mice treated with control DCs (30.8% vs. 17.0% at an E/T ratio of 50, p < 0.01) (Fig. 4c left panel), while they did not efficiently lyse unrelated B16F10 cells (Fig. 4c right panel).

thumbnail image

Figure 4. Silencing IDO reduced apoptosis of T cells and enhanced T cell proliferation and cytotoxic activity of CTLs. (a) Silencing IDO reduced apoptosis of T cells: BALB/c mice were challenged with 4T1 cells and treated with DCs as described in the legend to Figure 3. On Day 30 after tumor cell injection, T cells from tumor-draining lymph nodes were collected, triple-stained with Abs against CD4, CD8 and Annexin-V, and analyzed by flow cytometry. (b) Tumor antigen-specific T cell responses: Viable T cells were isolated from breast cancer mice treated with IDO silenced DCs or control DCs. Tumor antigen-specific T cell responses were assessed as described in Material and Methods. T cells were stimulated in vitro with 4T1 cell lysate in the presence of in vitro cultured Day 6 bone marrow DCs from naïve BABL/c mice and induced proliferation was measured by 3H-thymidine incorporation. (c) Cytotoxic activity of CTLs: CD8+ CTLs were isolated from tumor-draining lymph nodes of tumor-bearing mice treated with IDO-silenced DCs or control DCs as described in Material and Methods. CTLs were incubated with 4T1 cells (left panel), B16F10 (right panel) at different effectors (CTL): target cells (4T1) ratios for 4 hr, respectively. Target cell lysis was determined by LDH release. Results represent one of three repeat experiments (n = 3 per experimental group in each experiment). *p < 0.05, one-way ANOVA.

Download figure to PowerPoint

Silencing IDO reduces Foxp3+ Treg generation

Naturally occurring Foxp3+ Treg cells exist at a very low frequency (0.5–5% of the total T cell population in peripheral compartments). In the tumor milieu, however, expansion of the Foxp3+ Treg compartment has been reported and hypothesized to lead to promotion of tumor tolerance.33 IDO expressing DCs can induce Treg generation and, in turn, Treg cells can upregulate IDO expression in DCs to strengthen the immunosuppressive microenvironment in cancer patients.7 We examined Tregs in breast cancer-bearing mice by flow cytometric analysis of intracellular Foxp3 in CD4+/CD25+ Treg cells. Notably, the fraction of CD25+Foxp3+ Treg cells in CD4+ cells derived from TDLNs in mice treated with IDO-silenced DCs was reduced to 11.8% ± 0.8% compared to 20.0% ± 1.6% in mice treated with control DCs and 18.4% ± 1.2% in untreated mice (Fig. 5a). Therefore, treatment with IDO-silenced DCs resulted in reduction in numbers of Treg cells capable of mediating tumor tolerance.

thumbnail image

Figure 5. Silencing IDO limits Treg cell generation in vivo. (a) Percentage of Treg population in CD4+ T cells from tumor draining lymph nodes. Tumor draining lymph node cells were isolated from tumor-bearing mice treated with IDO-silenced DCs or control DCs, as described in the legend to Figure 4, triple-stained with Abs against CD4, CD25 and Foxp3, and Ab binding assessed by flow cytometry. The percentage of CD25+Foxp3+ cells among cells first gated for CD4+ status was calculated. Data are pooled from three independent experiments (p < 0.05, one way ANOVA analysis). (b) Inhibitory function of Tregs: CD4+CD25+ cells were isolated from the mice as described in (a) and Material and Methods. Isolated CD4+CD25+ cells were added to an ongoing tumor-Ag specific T cell response assay, in which T cells isolated from breast cancer-bearing mice were used as responder cells in the presence of 4T1 tumor cell lysate and naïve DCs as antigen presenting cells. Representative data from one experiment of three are shown (n = 3 per experimental group in each experiment; *p < 0.05, One way ANOVA analysis). (c) Foxp3+ cell infiltrating into tumor tissue: Tumors were excised from non treatment group (A), treatment with control siRNA transfected DCs (B), and with IDO-silenced DCs group (C). Frozen tumor tissue was sectioned and stained with anti-mouse Foxp3 Ab. Tissues were visualized microscopically (×200 magnification). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Download figure to PowerPoint

In addition, we tested the inhibitory function of Treg cells in a tumor-Ag specific T cell response in which CD4+CD25 T cells were isolated from untreated tumor-bearing mice. We isolated CD4+CD25+Tregs from mice treated with IDO-silenced DCs or control DCs on Day 30 after tumor cell challenge. Tregs from mice treated with either IDO-silenced DCs or control DCs strongly inhibited 4T1-tumor-Ag specific T cell responses (Fig. 5b).

Recently studies have shown that tumor-infiltrating Tregs play a role in maintaining local immune suppression.34 We therefore examined whether treatment with IDO-silenced DCs affected the number of tumor-infiltrating Tregs detected by immunohistochemistry. FoxP3+ cells were obvious in tumor tissue excised from mice treated with control DCs, but not in tumors from mice treated with IDO-silenced DCs (Fig. 5c).

Discussion

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

DC-based immune cancer therapies have advanced into human clinical trials. At least one product has been FDA-approved for treatment of hormone-refractory prostate cancer as autologous cellular immunotherapy.35, 36 Indeed, DC-based vaccine immune therapy for cancer has been demonstrated to effectively generate tumor-specific cytotoxic T lymphocytes.37 Although this immune-based approach to cancer treatment has been shown to confer a therapeutic advantage both in principle and in practice, clinical outcomes do not predictably correspond with vaccine-induced induction of antitumor immune response.38 One potential strategy to augment DC-base therapies is to neutralize the recently-recognized immunosuppressive molecule IDO.32 This tryptophan-catabolizing enzyme is known to promote the capacity of tumor cells to escape attack by CTLs,39 and is associated with both local and systemic host tolerance of tumor growth and progression to increasingly malignant states. In the present study we demonstrate that enhanced antitumor response and cytotoxic activity of CTLs were observed when IDO gene expression in DCs was silenced using siRNA. Moreover, siRNA-mediated IDO silencing enhances the antitumor effect of DC-based vaccines.

DCs play a pivotal role in cancer immune responses. They induce adaptive immune responses by acquiring, processing, and presenting antigens to T cells. DC-based vaccines in which DCs are generated ex vivo, pulsed with tumor-associated antigens, and then exposed to maturation stimuli, are promising anticancer agents that can harness the specificity and potency of the immune system for therapeutic use in combating various types of cancers.40, 41 The most advanced cancer vaccine studies with DCs have been undertaken in melanoma, principally because of the relatively well-defined nature of tumor-associated antigens (TAAs) in those tumors.38, 42 The first reported anticancer DC vaccine study used DCs loaded with tumor peptides or autologous tumor cell lysate in the presence of the helper antigen keyhole limpet hemocyanin (KLH), an approach that was effective in inducing objective responses in advanced melanoma patients including regression of metastases in various organs.5 Since then, many preclinical and clinical studies of DC vaccines have been conducted with the aim of enhancing the clear capacity of antigen-pulsed DCs to elicit immune responses against tumors. In this study we used whole tumor lysate that evoked a strong antitumor immune response. Additionally, we found that siRNA-mediated silencing of IDO in DCs did not induce apparent altered expression of biological cell surface DC marker molecules, supporting the hypothesis that IDO reduction had little or no effect on the biological character of DCs. Our work suggests that manipulating DCs with siRNA does not degrade their function in tumor vaccination, in agreement with earlier findings.22

IDO is expressed in human carcinomas arising in many different tissues, and has consequently been implicated as an important factor in tumor evasion of host immunity across a broad range of tumor types.32 In addition to IDO expression in tumor cells themselves, many subtypes of DCs including monocyte-derived DCs,43 plasmacytoid DCs,20 CD8α+ DCs,44 “IDO-competent” DCs45 and IFN-γ-activated DCs used in DC vaccination express IDO in varying degrees. These DCs suppress immune responses through their ability to induce apoptosis of activated T cells,46 to mediate antigen-specific T cell anergy in vivo20 and to enhance production of Treg cells reported at sites of vaccination with IDO-positive DCs in human patients.7, 17, 18, 20, 47 Treg cells generated after DC vaccination can suppress host immune responses against tumor cells and, in fact, depletion of regulatory Treg cells enhances DC vaccine-mediated antitumor immunity in cancer patients.48, 49 The immunosuppressive effect of IDO can be completely reversed by the small molecule IDO inhibitor 1-methyl tryptophan (1-MT)32 or by IDO siRNA.11 It has been reported that treatment with DCs pulsed with tumor lysate, in combination with 1-MT, reduced the number of Treg cells and improved the anti-tumor efficacy of a DC vaccine; on the other hand, DC vaccine with normal IDO expression (or treated with 1-MT alone) could not induce a strong antitumor effect in pancreatic cancer.21 In the current study we demonstrate that knocking down IDO expression in DCs resulted in stronger allogeneic T cell proliferation in vitro. In addition, administration of IDO-silenced DC vaccine to mice harboring mouse 4T1 breast tumors decreased apoptosis of CD4+ and CD8+ T cells in vivo. Furthermore, knockdown of IDO levels in DCs limited Treg generation, resulting in enhanced antitumor effects.

In conclusion, silencing the immunosuppressive gene IDO in DC using siRNA improved, in a mouse tumor model, the antitumor effects of a conventional DC vaccine. Targeted reduction of IDO in dendritic cells may be a useful strategy to enhance the effectiveness of DC vaccines in the clinic, and to improve the potential of immunotherapy to delay tumor growth in cancer patients.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References

This study is partially supported by a grant to W.-P.M. and J.K. from the Canadian Breast Cancer Foundation (CBCF), a CBCF Fellowship to X.Z., and grant from the Schulich School of Dentistry and Medicine and the Lawson Health Research Institute to W.-P.M. and X.Z. This study is also supported by a grant from London Regional Cancer Program (W.-P.M and J.K.). X.Z. is supported by a postdoctoral Fellowship for the CIHR Strategic Training Program in Cancer Research and Technology Transfer (CaRTT).

References

  1. Top of page
  2. Abstract
  3. Material and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References