| ObjectiveDendritic cells (DCs) are the most potent antigen-processing cells and havebeen widely applied in cancer vaccines. However, their abilities have often beensuppressed by tumors that secrete immunosuppressants such as transforminggrowth factor beta (TGF-β). We developed a new strategy using aTGF-β-insensitive DC vaccine and evaluated its effect in a murine renalcarcinoma (Renca) lung metastasis model. Tumor lysate-pulsed DCs (TP-DCs)were rendered TGF-βinsensitive by infecting them with a retrovirus containingthe dominant negative TGF-βtype II receptor (TβRIIDN). TβRIIDN blocked thetransduction of TGF-βsignals to the Smad family proteins, thus cutting offdownstream gene transcription in the nucleus. In this stusy we deliveredTβRIIDN DC vaccine to the mice bearing Renca pulmonary metastases to investigate if the vaccine can induce tumor-specific cytotoxic T lymphocyte(CTL) responses against Renca tumors, the situation of pulmonary metastasis,and survival times of the mice.Materials and methods1. Mice and cellsFemale Balb/c mice aged 6–8 weeks were obtained from the LaboratoryAnimal Research Center of the Fourth Military Medical University, Xi'an,China, and maintained in pathogen-free facilities at the State Key Laboratory ofCancer Biology following the established guidelines of the Animal EthicsCommittee of the Fourth Military Medical University. The mouse renalcarcinoma cell line Renca (ATCC, Rockville, MD) was maintained in acomplete medium (CM) containing RPMI-1640 medium (HyClone, Logan, UT)supplemented with 10% heat-inactivated fetal bovine serum (FBS; GIBCO,Gaithersburg, MD), 2 mM L-glutamine, 50μM 2-mercaptoethanol, 100 units/mlpenicillin, and 100μg/ml streptomycin (Sigma, St Louis, MO).2. Construction of TβRIIDN-GFP retroviral vectorTβRIIDN was excised from pcDNA3-TβRIIDN by BamHI/EcoRI digestionand inserted into the pMig-inteRIIDNl ribosomal entry sequence-greenfluorescence protein (herein designated MSCV-GFP) vector by first linearizingpMig with EcoRI and ligating an EcoRI/BamHI adapter(5'-AATTGGATCCGCGGCCGCG-3', 3'-CCTAGGCGCCGGCGCTTAA-5').These clones were designated as MSCV-TβRIIDN and were screened bysequencing for correct orientation and insert numbers.3. Production of infectious TβRIIDN-GFP retrovirusPantropic GP293 retroviral packaging cells (Clontech, San Diego, CA) wereseeded at a density of 2.5×106 per T-25 collagen I-coated flask (Biocoat; BD Biosciences, Mountain View, CA) and incubated for 24 h before plasmidtransfection in antibiotic-free 10% DMEM (Gibco, Grand Island, NY). Amixture of 2μg retroviral plasmid and 2μg vesicular stomatitis virus envelopeG protein (VSV-G) envelope plasmid was cotransfected in serum-free DMEMby using Lipofectamine-Plus (Invitrogen, Carlsbad, CA) according to themanufacturer's protocols. Briefly, the cells were transfected for 12 h followedby the addition of an equivalent volume of 10% DMEM and incubation for anadditional 12 h. The supeRIIDNtant was then aspirated, the cells were rinsedgently in PBS, and 3 ml fresh 10% DMEM was added to each flask. After 24 h,the virus-containing supeRIIDNtant was collected and used to infect target cells.4. Generation of bone marrow-derived DCsThe method to generate and expand bone marrow-derived DCs was mainlyaccording to the protocol of K. Inaba and M.B. Lutz with some modifications.Briefly, erythrocyte-depleted murine bone marrow cells were obtained underaseptic conditions from the femurs and tibiae of Balb/c mice and cultured at 2×105 cells/ml in a 100-mm diameter cell culture dish (Corning, NY, USA) with10 ml CM supplemented with 200 units/ml recombinant murinegranulocyte/macrophage colony stimulating factor (rmGM-CSF; PeproTech,London, UK) and 1000 units/ml recombinant murine interleukin-4 (rmIL-4;PeproTech, London, UK). The same amount of medium was added to theculture dish on day 3. On day 6, nonadherent DCs were harvested by gradientcentrifugation.5. Preparation of tumor lysate-pulsed DCs (TP-DCs)Tumor cells were suspended at a density of 1×107 cells/ml in CM. After 3cycles of rapid freezing (in liquid nitrogen) and thawing (37°C), the tumor cellsuspension was centrifuged at a low speed (400 rpm for 10 min). The supeRIIDNtant (tumor lysate) was collected and incubated overnight withpurified DCs at a ratio of 3:1 tumor cell equivalents to DCs. TP-DCs wereharvested and cultured in CM supplemented with 200 units/ml rmGM-CSF and1000 units/ml rmIL-4.6. Infection of DCs with retrovirus containing TβRIIDNTP-DCs were infected with either the MSCV retrovirus containing TβRIIDNand green fluorescent protein (GFP) or GFP control vector. Thus, 4 types ofDCs were established. The first type comprised tumor lysate-pulsedTGF-β-insensitive DCs (TGF-β-insensitive TP-DCs); the second type, TP-DCsinfected with virus containing the GFP control vector (GFP-transduced TP-DCs);the third type, non-transduced TP-DCs; and the fourth type, na?ve DCs that werecultured from the bone marrow cells of na?ve untreated donor animals.7. Flow cytometric analysisAll 4 types of DCs were cultured for another 2 days and then analyzed byFACS Calibur flow cytometry (Becton-Dickinson, San Jose, CA) with a panel ofantibodies specific to CD11c, MHC class II, CD40, CD80, and CD86(Immunotech, Miami, FL), as previously described.8.Western blot analysis for Smad 2 phosphorylationAll 4 types of DCs were treated for 24 h with or without 10 ng/ml TGF-β1(R&D Systems, Minneapolis, MN) in the culture medium. The cells were lysedusing 125μl CelLytic M (Sigma, St Louis, MO) per 106–107 cells. The proteinextract obtained was loaded onto a 10% acrylamide gel in Tris-HCl (Bio-Rad,Hercules, CA). Electrophoresis was carried out in Tris-glycine-SDS runningbuffer and transferred to a polyvinylidene difluoride membrane. The blots wereprobed with monoclonal antibodies against Smad 2 (BD Pharmingen, San Diego,CA), phospho-Smad 2 (Upstate Biotechnology, Lake Placid, NY), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH; Invitrogen, Carlsbad,CA).9. Thymidine incorporation assayTP-DCs (3×104 cells/24 wells) in the culture medium were treated with orwithout 10 ng/ml TGF-β1 for 24 h. Medium containing [3H]-thymidine (1μCi/ml; Amersham, Buckinghamshire, UK) was introduced, and the cells werecultured for another 16 h. The experiment was terminated by washing the cellswith warm serum-free medium, and the samples were harvested onto glass fiberfilter paper for analysis using the TriCarb 2500 TR scintillation counter (PackardBioScience, Meriden, CT).10. In vivo antitumor analysesMouse lung metastasis models were prepared by a single injection of 5×105Renca cells in the tail vein (day 0). On day 7, the tumor-bearing mice (n =10/group) were inoculated with all 4 types of DC vaccines (1×106 cells) via asubcutaneous injection on either flank. Vaccination was repeated on day 10. Allanimals were sacrificed on day 60. The lungs of all mice including those whodied prior to the end of the experiment were excised, weighed, and fixed with4% formaldehyde solution. All the fixed lung samples were embedded inparaffin and serially sectioned at 5μm until the block was exhausted. RoutineHE staining was performed for every 10th section.11. In vitro CTL assaysAnother group of mouse models (n = 5/group) was similarly established, andthe mice were vaccinated by the aforementioned protocol. On day 20, all 4groups of animals were sacrificed. Splenic CD8+ T cells were isolated using amurine T-cell CD8+ subset column kit (R&D Systems, Minneapolis, MN).Renca and another irrelevant cell line, namely, mouse breast carcinoma EMT-6, were used as targets. Briefly, target cells were labeled with 100μCi 51Cr/107cells. Different groups of CD8+ T cells were added to U-bottom platescontaining 5,000 cells/well with various effector to target ratios ranging from1:1 to 100:1. Equal volumes of RPMI-1640 and 1 M HCl were added to otherwells as the negative and positive controls, respectively. After incubation for 4 hunder standard conditions, supeRIIDNtants were collected, and theirradioactivity was measured using a gamma counter. The specific lysispercentage was calculated as follows: 100×[(Experimental 51Cr Release–Spontaneous 51Cr Release)/(Maximum 51Cr Release– Spontaneous 51CrRelease)].12. In vitro cytokine assaysThe sera of the 4 above mentioned mouse groups were harvested. The serumlevels of IFN-γand IL-2 were determined using an ELISA kit (R&D Systems,Minneapolis, MN) according to the protocol.13. Statistical analysisNumerical data were expressed as mean±standard deviation (SD). ANOVAand chi-square tests were performed to determine the differences in the meansamong the various treatment groups. P < 0.05 was considered statisticallysignificant. The SPSS 10.0.2 software package (SPSS Inc., Chicago, IL) wasused for analysis. The Kaplan-Meier survival curve was analyzed by thelog-rank test with the Graphpad Prism 4.02 software (Graphpad Software Inc.,San Diego, CA).Results:1. Exertion of the TβRIIDN gene did not change the phenotype of transducedDCsFACS analyses showed that the purity of freshly harvested na?ve DCs on day 6 was 84.88%, and the transfection efficacy of TβRIIDN-GFP and GFP onlywas 92.13% and 91.95%, respectively. There were no significant differences insurface marker CD11c among all the four types of DCs (p=0.976). But theexpression of surface costimulatory molecules indicating maturation, namely,MHC class II, CD40, CD80, and CD86, was considerably higher in the 3 typesof TP-DCs than in the na?ve DCs (p=0.000 in MHC classII, CD80 and CD86,p=0.013 in CD40). However for all the three types of TP-DCs, there were nosignificant differences among them (p=0.708 in MHC classII, p=0.903 in CD40,p=0.592 in CD80 and p=0.941 in CD86).2. Phosphorylated Smad 2 was undetectable in TGF-β-insensitive TP-DCsWestern blot analysis showed that Smad 2 existed in all 4 types of DCs.However, on treatment with TGF-β1, phosphorylated Smad 2 was detected inna?ve DC, GFP-transduced and nontransduced TP-DCs. In contrast,phosphorylated Smad 2 was not detected in TGF-β-insensitive TP-DCs aftertreatment with TGF-β1, confirming that TGF-βsignal transduction was blockedby the presence of the TβRIIDN gene.3. TGF-β-insensitive TP-DCs were resistant to the effects of TGF-β1Since TβRIIDN lacks the intracellular kinase domain, the transduction ofTGF-βsignals to downstream molecules in the cells is blocked. To confirmwhether the expression of TβRIIDN in DCs could overcome the effects ofTGF-β1, thymidine uptake was compared among TGF-β-insensitive,GFP-transduced, and nontransduced TP-DCs after coculturing them with 10ng/ml TGF-β1 for 24 h. Thymidine uptake decreased greatly in theGFP-transduced and nontransduced TP-DCs. However, the thymidine uptake ofthe TGF-β-insensitive TP-DCs decreased by only 14.3%±1.4%. Comparingwith these of GFP-transduced TP-DCs (62.3±3.5%) and nontransduced TP-DCs (61.7±4.2%), the differences were significant (p=0.000). But between these ofGFP-transduced TP-DCs and nontransduced TP-DCs, there was no significantdifferent (p=1.000).4. TGF-β-insensitive TP-DCs suppressed tumor growth and increased survivalrates in tumor-bearing miceTo evaluate the antitumor effect of the TGF-β-insensitive TP-DC vaccine, the3 types of TP-DC vaccines and na?ve DCs were subcutaneously delivered oneither flank of 10 mice in every group on days 7 and 10. The mice that receivedTGF-β-insensitive TP-DCs exhibited the least degree of tumor burden. Amongall the 4 groups, the TGF-β-insensitive TP-DC group had the lightest averagelung weights (0.907±0.426 g). Compared with those of na?ve DC group(1.345±0.055 g), nontransduced TP-DC group (1.262±0.224 g) andGFP-transduced TP-DC group (1.263±0.24 g), there were significant differences(p=0.001 vs. na?ve DC group, p=0.006 vs. nontransduced TP-DC andGFP-transduced TP-DC group). Interestingly, all the animals in the na?ve DCgroup died before day 33 of the experiment due to poor health conditions, whileonly 4 of 10 mice in the TGF-β-insensitive TP-DC group died at the end of theexperiment. According to the log-rank test, there were significant differencesamong three TP-DC groups and na?ve DC group (p=0.000). Comparing withthose of the other two TP-DC groups, the survival rate of TGF-β-insensitiveTP-DC group was significant higher (p=0.032 vs. GFP-transduced TP-DC groupand p=0.029 vs. nontransduced TP-DC group).5. TGF-β-insensitive TP-DCs induced potent tumor-specific CTL responseTo determine whether the antitumor response induced by TGF-β-insensitiveTP-DCs was tumor specific, splenic CD8+ T cells of the other 4 mouse groupswere isolated. The ability of CTLs to lyse Renca cells was assayed using a standard 51Cr-release assay. It revealed that compared with the na?ve DC group,higher cytotoxicities against the Renca cells were observed in mice treated withTP-DC (11-fold TGF-β-insensitive TP-DC group vs. na?ve DC group, 5-foldGFP-transduced TP-DC group vs. na?ve DC group and 5-fold nontransducedTP-DC group vs. na?ve DC group at an E:T cell ratio of 100:1). But the mostpotent Renca-specific CTL response was induced by the TGF-β-insensitiveTP-DCs in the tumor bearer (76.4% killing activity at an E:T cell ratio of 100:1).The tumor-killing ability of CD8+ T cells from TGF-β-insensitive TP-DCsgroup was 2-fold that of GFP-transduced TP-DCs and nontransduced TP-DCs atthe same ratio. No apparent lyses were observed against irrelevant EMT-6 cellsin all the four groups.6. TGF-β-insensitive TP-DCs induced high serum levels of INF-γand IL-2On the tenth day after vaccination, the serum level of IL-2 of na?ve DC groupwas 29.13±6.37 pg/ml. Comparing with those of nontransduced TP-DC group(830.17±30.15 pg/ml), GFP-transduced TP-DC group (850.34±37.16 pg/ml) andTGF-β-insensitive TP-DC group (1540.25±37.16 pg/ml), there were significantdifferences (p=0.000). And between those of TGF-β-insensitive TP-DC groupand the other two TP-DC groups, the differences were also significant (p=0.000).For the serum level of INF-γ, the results were almost the same. The differencesof serum levels between na?ve DC group (28.24±7.07 pg/ml) and nontransducedTP-DC group (153.14±7.07 pg/ml), GFP-transduced TP-DC group (163.25±9.09pg/ml) and TGF-β-insensitive TP-DC group (285.42.25±18.19 pg/ml) weresignificant (p=0.000). Also between those of TGF-β-insensitive TP-DC groupand the other two TP-DC groups the serum levels of INF-γwere significantlydifferent (p=0.000). A greater increase in levels of serum IL-2 and IFN-γwasobserved in the TGF-β-insensitive TP-DCs group, indicating that immune cells were most strongly activated in these hosts.Conclusion1. We successfully constructed a retrovirus containing dominant-negativeTGF-βtype II receptor (TβRIIDN).2. Balb/c murine bone marrow DCs were pulsed with freeze–thawed Rencatumor lysate and induced Renca specific DCs.3. The tumor lysate-pulsed DCs were rendered TGF-βinsensitive by infectingwith a retrovirus containing dominant-negative TGF-βtype II receptor(TβRIIDN), leading to the blockade of TGF-βsignals to members of theSmad protein family.4. Expression of TβRIIDNdid not affect the phenotype of transduced DCs.Expression of MHC II, CD40, CD80, and CD86 was the same on TGF-βinsensitive TP-DC comparing with these on GFP TP-DC and TP-DC.5. TGF-βinsensitive TP-DC suppressed tumor growth and increased survivalrate of Renca pulmonary metastases mice.6. The most potent Renca-specific splenic CTL response was induced by theTGF-?-insensitive DCs in the tumor bearer (76.4% killing activity at aneffector:target cell ratio of 100:1). No apparent lysis was observed againstirrelevant EMT-6 cells.7. TGF-βinsensitive TP-DC induced higher IFN-γand IL-2 level in vivo.
|
PDF Full Text Request