|Home | About | Journals | Submit | Contact Us | Français|
Type I interferons (IFNs) are considered to be important mediators of innate immunity due to their inherent antiviral activity, ability to drive the transcription of a number of genes involved in viral clearance, and their role in the initiation of innate and adaptive immune responses. Due to the central role of type I IFNs, we sought to determine their importance in the generation of immunity to a recombinant vaccine vector fowlpox virus (FPV). In analyzing the role of type I IFNs in immunity to FPV, we show that they are critical to the secretion of a number of innate and proinflammatory cytokines, including type I IFNs themselves as well as interleukin-12 (IL-12), tumor necrosis factor-alpha (TNF-α), IL-6, and IL-1β, and that deficiency leads to enhanced virus-mediated antigen expression. Interestingly, however, type I IFNs were not required for adaptive immune responses to recombinant FPV even though plasmacytoid dendritic cells (pDCs), the primary producers of type I IFNs, have been shown to be requisite for this to occur. Furthermore, we provide evidence that the importance of pDCs may lie in their ability to capture and present virally derived antigen to T cells rather than in their capacity as professional type I IFN-producing cells.
Type I interferons (IFNs) are considered to be one of the key regulators of innate and adaptive immunity following pathogenic challenge. While type I IFNs can act on infected cells directly to inhibit viral dissemination, they also have the ability to mediate the appropriate activation of other immune cell types to facilitate pathogen clearance. While plasmacytoid dendritic cells (pDCs) are generally considered to be the primary producers of high systemic levels of type I IFNs in both humans and mice, virtually every cell type is capable of low levels of secretion when required (15). As a result, type I IFNs have the potential to act on many different cell types and tissues, which can result in differential outcomes, depending on the local environment. The creation of mice deficient in the type I IFN receptor (IFNAR−/−) (36) has accelerated the understanding of the many important and often differential roles for type I IFNs in the antiviral response while concurrently illustrating their role as a contributing factor modulating many other pathologies (19, 30, 57). Mice deficient in IFNAR are defective in their ability to mount a normal innate and subsequent adaptive immune response to many viruses, including avian H5N1 influenza virus (49), vesicular stomatitis virus (VSV), herpes simplex virus (HSV), Newcastle disease virus (NDV) (20), Semliki Forest virus (SFV), lymphocytic choriomeningitis virus (LCMV), and vaccinia virus (VV) (36).
Innate immunity, the first line of defense against invading pathogens, is highly conserved among all multicellular organisms and is initiated in the host immune system by recognition of conserved structural motifs termed pathogen-associated molecular patterns (PAMPs) displayed by invading pathogens. These PAMPs are recognized by host cellular pattern recognition receptors (PRRs) with various chemical specificities, including the Nod-like receptors (NLRs), Rig-like receptors (RLRs), C-type lectin receptors (CLRs), and Toll-like receptors (TLRs) (28). In addition, a number of cytosolic receptors have recently been implicated in the detection of foreign double-stranded DNA (dsDNA) within infected cells (24, 25, 50). Toll-like receptors represent perhaps the best-studied family of PRRs and have been shown to be critical to both RNA and DNA viral recognition. Engagement of TLRs can instigate a range of immunological events through the activation of the nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) and interferon regulatory factor 3 and 7 (IRF-3/IRF-7) pathways, resulting in production of proinflammatory cytokines and type I IFNs, respectively (27). The secretion of type I IFNs also leads to the initiation of another signaling cascade mediated by signal transducer and activator 1 (STAT1) and STAT2 and Janus kinases (JAK-STAT pathway) (63). The JAK-STAT signaling pathway promotes the activation of a number of interferon-stimulated genes (ISGs), some of which lead to the alteration of natural cell cycle, induction of apoptosis in infected cells, triggered maturation of certain cell subsets, and regulation of both innate and adaptive effector cells, including natural killer cells, dendritic cells (DCs), and T and B cells (2). While most cell types are able to produce type I IFNs, pDCs are the most potent producers, usually secreting between 10 to 1,000 times as much IFN as other cell types (15).
Fowlpox virus (FPV), the prototypical member of the Avipox genus, infects chickens and turkeys, causing a drop in egg production, reduced growth rates, blindness, and, in some cases, death. Vaccination with live attenuated and nonattenuated viruses can be used to control this disease, and recombinant variants have been developed to provide protection against a variety of other economically significant diseases of poultry and other animal species (40, 52). More recently, however, the potential of recombinant FPV (rFPV) as a human vaccine vector has become the focus of attention. With viral replication intrinsically blocked in mammalian cells, rFPV has an impeccable safety profile, a large cloning capacity enabling the expression of multiple heterologous target genes, and the ability to elicit cell-mediated immunity (reviewed in reference 5). We have recently shown that immunization of mice with rFPV is accompanied by type I IFN secretion and a strong but transient cytotoxic T-cell response, which is critically dependent on CD4+ T-cell help. Plasmacytoid dendritic cells were found to be central to this response, and given their recognized ability to produce copious amounts of type I IFNs in response to virus challenge (15), we hypothesized that it was in this capacity that pDCs contributed to the formation of adaptive immunity to our rFPV-encoded antigen (11). This study expands these findings by formally investigating the role of type I IFNs and different DC populations in the induction of both innate and adaptive immune responses. Results demonstrated an absolute requirement for type I IFN secretion in innate immune responses to FPV; however, adaptive immune responses remained unaffected by the absence of IFN. Interestingly, we find the absence of the type I IFN receptor leads to increased recombinant antigen production in vitro and prolonged and enhanced antigen presentation in vivo. Contrary to our original hypothesis, though, it appears that the importance of pDCs lies in their ability to present FPV-encoded antigen as opposed to their ability to secrete high levels of type I IFNs. The findings presented herein represent a redundant role for type I IFNs in the formation of adaptive immunity to a nonreplicating viral vector.
Type I IFN receptor-deficient mice (IFNAR−/−) on a SV129 × B6 mixed background (36) were provided by the Australian Phenomics Facility (Australian National University, ACT, Australia). Control mice were produced in-house as F1 progeny from successful matings of normal C57BL/6J (B6) and SV129 mice purchased from Laboratory Animal Services ([LAS] University of Adelaide, S. A., Australia). OT-I T-cell receptor (TCR) transgenic (Tg) mice (8), which contain Tg CD8+ T cells that recognize the dominant chicken ovalbumin peptide consisting of residues 257 to 264 (OVA257-264) in the context of H2-Kb, and B6bm1 mice, which harbor a single natural mutation in the binding groove of H-2Kb and are therefore unable to present the H-2Kb-restricted OVA257-264 peptide to CD8+ T cells (39), were both kindly provided by W. R. Heath (Walter and Eliza Hall Institute of Medical Research, Melbourne, Victoria, Australia). Congenic CD45.1+ B6.SJL mice were purchased from the Animal Resource Centre (Canning Vale, W. A., Australia). All CD45.1+ OT-I (congenic OT-I) mice were obtained and used as F1 progeny from the successful pairing of CD45.1+ congenic B6 and homozygous OT-I TCR Tg mice. All mice were housed and bred under specific-pathogen-free (SPF) conditions at LAS, and all animal studies were approved and conducted following institutional ethical guidelines.
Fowlpox virus mild strain 3 (6) was used in its wild-type form (FPVWT) or as a recombinant version encoding the full-length ovalbumin (OVA) gene under the control of an early/late vaccinia promoter (FPVOVA). Both FPVWT and FPVOVA were routinely propagated and titrated in chicken embryonic fibroblast (CEF) cells within our laboratory, as previously described (11).
Recombinant Chinese hamster ovary (CHO)-derived human fms-like tyrosine kinase 3-ligand (FL) was kindly provided by Amgen Inc. Recombinant murine granulocyte macrophage colony stimulating factor (GM-CSF) was obtained from R&D Systems. Chicken ovalbumin peptide OVA257-264 (SIINFEKL) was synthesized on an ABI 431A peptide synthesizer using standard Fmoc (9-fluorenylmethoxy carbonyl) chemistry. The CpG oligodeoxynucleotide 2216 (G*GGGGACGATCGTCG*G*G*G*GG, where the asterisk indicates phosphorothioate-modified nucleotides) was custom made (GeneWorks Pty, Ltd). Whole chicken OVA grade V protein was purchased from Sigma Aldrich.
Various populations of DCs and macrophages were expanded in vitro from bone marrow (BM) cells cultured in either FL, GM-CSF, or macrophage colony stimulating factor (M-CSF) as previously described (7, 18). Briefly, BM was flushed from the femur and tibia of experimental mice with complete RPMI medium ([CM] 10% fetal calf serum [FCS], 2 mM l-glutamine, 10 mM HEPES, 100 U/ml penicillin, 100 μg/ml streptomycin, 10 μM 2-mercaptoethanol [2-ME]). Red blood cells (RBCs) were subsequently lysed, and cells were washed thoroughly prior to seeding at a concentration of 2 × 106 cells/ml in CM supplemented with FL (200 ng/ml) and at 1 × 106 cells/ml in CM supplemented with GM-CSF (100 ng/ml) or CM supplemented with M-CSF (20% L929 culture supernatant). All cells were cultivated in 24-well plates for 8 days prior to experimentation, with 50% of medium replenished as required. Following expansion, GM-CSF cultures exhibited a conventional DC (cDC) phenotype lacking a pDC population, and FL cultures yielded both cDC and pDC populations while M-CSF cultures expressed a macrophage phenotype. All phenotypes were confirmed by flow cytometric analysis (data not shown).
Pregnant female mice were sacrificed (13 to 14 days postcoitum), and uterine horns were dissected and rinsed briefly in ethanol. Embryos were separated from the uterus and placenta, and the brain and liver tissues were removed. Embryos were subsequently minced and resuspended in TrypLE select (Invitrogen) to facilitate release of cells from tissues. Following incubation (at 37°C for 20 min) cells were collected, resuspended in mouse embryonic fibroblast (MEF) medium (Dulbecco's modified Eagle's medium [DMEM] supplemented with 10% FCS, 2 mM l-glutamine, 50 U/ml penicillin, and 50 μg/ml streptomycin), plated at a concentration of 1 embryo per 10-cm tissue culture dish, and incubated (at 37°C in 5% CO2) until confluent. MEF cells were reseeded at a concentration of 1 × 106 cells/ml in 24-well plates 1 day prior to experimentation.
FL-expanded BM cells were prepared and cultured as described above in flasks. After differentiation (8 days), cells were counted and plated into six-well plates and either left uninfected or infected with FPVWT for 8 h prior to the addition of brefeldin A (10 μg/ml; Sigma) for a further 12 h. Cells were collected in Dulbecco's phosphate-buffered saline (DPBS), and Fc receptors were blocked (Fc block; 1 μg/106 cells; 15 min at 4°C). Cells were subsequently washed and stained with fluorescein isothiocyanate (FITC)-anti-B220 and phycoerythrin (PE)-anti-CD11b (for 45 min at 4°C; BD Pharmingen) to distinguish DC populations, i.e., pDCs (CD11blo B220+) and cDCs (CD11b+ B220−). Following staining, cells were washed three times and then fixed by resuspension at 5 × 106 cells/ml in a 50:50 combination of DPBS and 10% neutral buffered formalin for 8 min at room temperature. Cells were washed three times prior to permeabilization and staining. Antibodies used for staining intracellular cytokines were as follows: rabbit anti-IFN-α and rabbit anti-IFN-β (PBL Interferon Source), and biotin anti-interleukin-6 (IL-6), biotin anti-IL-12, and biotin anti-tumor necrosis factor alpha ([TNF-α] eBiosciences). Each antibody was diluted separately in saponin buffer (DPBS, 1% FCS, 0.1% saponin) and incubated with cells for 45 min at 4°C. Cells were then washed (three times) and purified, or biotinylated primary antibodies were detected either with biotin anti-rabbit IgG followed by streptavidin-PE-Cy5 (45 min at 4°C in saponin buffer) or with streptavidin-PE-Cy5. Cells were subsequently washed (three times), and all groups were analyzed immediately via flow cytometry (FACSCalibur).
Bone marrow cells from control mice were harvested and cultured in the presence of FL for 8 days as described above. Following expansion, cells were collected, counted, and stained with anti-B220 microbeads for pDC purification. Magnetically labeled pDCs were collected via application to a magnetized column and subsequent elution, as the per manufacturer's instructions (magnetic bead cell sorting [MACS]; Miltenyi Biotec). The flowthrough (B220− cell population) formed the cDC population. All cell populations were washed, counted, and plated (1 × 106/ml) in 24-well plates for further analysis.
For cytokine measurement, BM-derived cells (either unsorted or sorted into DC populations) or MEF cultures from IFNAR−/− and/or control mice were infected with FPVWT (at a multiplicity of infection [MOI] of 1) for 24 h, after which cell-free supernatants were harvested and stored at −20°C until used in enzyme-linked immunosorbent assays (ELISAs). Interferon-α (RMMA-1) and interferon-β (RMMB-1) antibodies and protein standards were purchased from PBL Interferon Source. Matched antibody pairs and corresponding standards for IL-12 (C18.2/C17.8), TNF-α (1F3F3D4/XT3), IL-6 (MP5-20F3/MP5-32C11), and IL-1β (B122/polyclonal) were all purchased from eBiosciences. All antibodies were used in accordance with manufacturer's instructions utilizing standard ELISA protocols.
Bone marrow cells were expanded in the presence of FL (200 ng/ml) in six-well plates (10 × 106 cells/5 ml; 8 days). Following expansion, cells were infected with FPVWT (MOI of 1) or mock infected (phosphate-buffered saline [PBS]) for the times indicated in Fig. Fig.3e.3e. Postinfection, cells were incubated on ice, dislodged using a cell scraper, and collected, and pellets were resuspended in modified radioimmunoprecipitation assay (RIPA) buffer (10 mM Tris [pH 7.4], 100 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1 mM NaF, 20 mM Na4P2O7, 2 mM Na3VO4, 0.1% SDS, 0.5% sodium deoxycholate, 1% Triton X-100, 10% glycerol, and complete EDTA-free protease inhibitor cocktail). Following cell lysis (for 10 min at 4°C), lysates were clarified via centrifugation and stored until analysis (−20°C). For Western blotting, samples were quantified for protein concentration by bicinchoninic acid (BCA) assay (Thermo Scientific), boiled (in 2× SDS loading buffer at 100°C for 5 min), fractionated by polyacrylamide gel electrophoresis (16 μg), and transferred onto polyvinylidene difluoride (PVDF) membranes (GE Healthcare). Membranes were subsequently blocked with 5% skim milk in TTBS (0.1% Tween 20-Tris-buffered saline [TBS]) and incubated with primary antibodies phospho-IκBα (5A5) (1:1,000; overnight at 4°C; Cell Signaling Technology) or β-actin (3:5,000; overnight at 4°C). Blots were washed and then incubated with alkaline phosphatase (AP)-conjugated anti-mouse IgG antibody (Rockland) prior to color development (BCIP/NBT [5-bromo-4-chloro-3-indolylphosphate/nitroblue tetrazolium] solution; Sigma). ImageJ software (http://rsb.info.nih.gov/ij/) was used for quantifying the intensity of Western blot bands compared to the relevant actin control.
FL-DCs or GM-CSF-DCs prepared as described above in 12-well plates (2 × 106) were infected with FPV (MOI of 1) or stimulated with PBS (uninfected) or CpG 2216 (1 μg/ml) for 24 h. Cells were subsequently harvested on ice using a cell scraper, washed, and stained with peridinin chlorophyll protein (PerCP)-Cy5.5-anti-B220 and PE-anti-CD11b and one of either FITC-anti-CD40, FITC-anti-CD80, or FITC-anti-CD86 (for 45 min at 4°C). Following extensive washing, cells were fixed and stored at 4°C prior to flow cytometric analysis (BD FACSCalibur). Plasmacytoid DCs were gated as B220+ CD11blo and cDCs were gated as B220− CD11bint/hi.
Mice were depleted of type I IFNs by administration of partially purified sheep anti-mouse interferon globulin or control sheep globulin (1:3 dilution in PBS; 200 μl administered intravenously [i.v.]) (20) 1 day prior to and on days 1 and 3 after immunization with FPVOVA. In some experiments, the adoptive transfer of donor cells into recipient B6bm1 mice required the in vivo depletion of endogenous T-cell subsets to prevent graft-versus-host disease. This was achieved by administering neutralizing anti-Thy1.2 monoclonal antibody (0.2 mg) via intraperitoneal (i.p.) injection 1 day prior to transfer of donor cells.
For cytotoxic T lymphocyte (CTL) assays, IFN-depleted, IFNAR−/−, and control mice were immunized with FPVOVA (1 × 107 PFU; i.p.), and splenocyte suspensions were prepared at the peak of the cell-mediated response, 5 days postimmunization. Following RBC lysis, cells were resuspended in incomplete RPMI medium ([ICM] CM lacking FCS) at 2 × 106 cells/ml and incubated in vitro in 10-cm tissue culture plates for 1 h to remove adherent cell populations. Nonadherent effector cells were subsequently harvested and assessed for OVA-specific and nonspecific cytolytic activity by incubation with 51Cr-labeled EL4 target cells (50 μCi/106 cells) either pulsed with the SIINFEKL peptide (2.5 μM) or not. Serial dilutions of effector cells, starting at a 200:1 effector/target (E/T) ratio, were cultured in U-bottom 96-well plates with an equal volume of target cells (104 cells) in CM for 18 h. Spontaneous 51Cr release was evaluated by incubating target cells alone while maximal release was determined by complete lysis of target cells in 1% NP-40 lysis buffer at the end of the assay. In all cases, the amount of 51Cr released from target cells was assessed by quantifying the amount of gamma radiation contained within 100 μl of cell-free supernatant (Packard Cobra gamma counter). Percent specific lysis was determined using the following equation: [(amount of experimental 51Cr release − amount of spontaneous 51Cr release)/(maximum amount of 51Cr release − amount of spontaneous 51Cr release)] × 100. Percent spontaneous lysis was <10% in all assays. Nonspecific lysis was routinely <5% and was subtracted from specific lysis results.
For T-helper cell proliferation assays, splenocyte samples from FPVOVA-immunized IFN-depleted, IFNAR, and control mice, obtained as described, above were resuspended in CM and seeded (2 × 105 cells/ml) with various concentrations of OVA protein (6.25 to 400 μg/ml) in flat-bottom 96-well plates. Three days later, cells were pulsed with [3H]thymidine (1 μCi/well), and 18 h later they were harvested onto glass fiber filters (Packard), and thymidine incorporation was measured on a β-scintillation counter (TopCount NXT; Perkin Elmer).
High-binding enzyme immunoassay/radioimmunoassay (EIA/RIA) plates (Costar) were coated with FPVWT (107 PFU/ml) diluted in 0.1 M Na2HPO4 (pH 9) and incubated overnight at 4°C. After being washed with PBS-0.05% Tween 20, wells were blocked with PBS-1% BSA for 1 h at 37°C. Serum samples diluted in PBS-0.05% Tween 20-1% BSA were then added and incubated for 2 h at 37°C, and bound anti-FPV antibodies were detected by incubation (1 h) with biotinylated anti-mouse IgG antibody (Rockland), followed by incubation (1 h) with streptavidin-horseradish peroxidase ([HRP] Rockland). Binding was quantified by the addition of o-phenylenediamine dihydrochloride (OPD) substrate (200 μl; Sigma), with color development stopped after 30 min by the addition of 3 M HCl (50 μl). Resultant substrate color change was measured by the optical density at 490 nm (OD490).
Single-cell suspensions prepared from the major lymph nodes (LNs) of OT-I Tg mice were labeled with 5-(and -6)carboxyfluorescein diacetate succinimidyl ester (CFSE; Molecular Probes). Cells were incubated with CFSE (1 μM; 107 cells/ml) in ICM (for 10 min at 37°C) with regular mixing, as previously described (31). Following incubation, cells were washed three times, and labeling was confirmed by flow cytometric analysis prior to injection.
For analysis of in vivo proliferation, filtered (70-μm-pore-size filter) CFSE-labeled congenic OT-I LN cells were injected i.v. into recipient B6 or IFNAR−/− mice (107 cells) following immunization with FPVOVA (5 × 106 PFU in 50 μl; right footpad) and PBS (50 μl; left footpad). Three days later, cells were recovered from the draining popliteal LN of recipient mice and stained with PE-Cy5-conjugated anti-CD8 and PE-conjugated anti-CD45.1 (BD Pharmingen). Analysis via three-color flow cytometry identified the adoptively transferred population from the endogenous cell population via the congenic marker (CD45.1), with the extent of OT-I proliferation (as measured by CFSE dilution) expressed as the proliferative index (PI), as previously described (12).
Recipient B6bm1 mice were initially depleted of T and B cells via i.p. administration of α-murine Thy1.2 depleting antibody (TIB107 [30H12]; 200 μg). The following day mice were injected with FPVOVA-infected (MOI of 1 for 1 h) or SIINFEKL-pulsed (2.5 μM for 1 h) FL- or GM-CSF-expanded dendritic cells (5 × 106). Five hours later, mice received CFSE-labeled congenic OT-I T cells (107; i.v.). Three days later, cells were recovered from the draining popliteal LNs and analyzed for proliferation as described above.
Splenocyte suspensions prepared from an immunized B6 mouse were initially purified for pDCs by staining with anti-murine plasmacytoid dendritic cell antigen (mPDCA) microbeads (MACS; Miltenyi Biotec), followed by collection of the magnetically labeled PDCA-positive (PDCA+) cells via application to a magnetized column and subsequent elution, as per the manufacturer's instructions. Splenic cDCs (PDCA− CD11c+) were subsequently purified from the PDCA-negative cell population using anti-CD11c (N418) microbeads and subsequent exposure to the magnetized column followed by elution. In some cases sorted pDCs and cDCs were recombined following sorting (1:1 ratio) to enable functional analysis of the combined subsets.
Splenocytes from an immunized B6 mouse were harvested 24 h postinfection with FPV by cutting the dissected spleen into small pieces and treating the samples with collagenase (2 mg/ml in 2% FCS-CM; 20 μg/ml DNase I) for 30 min at 37°C with gentle shaking. EDTA (0.1 M; pH 7.22) was added for the final 5 min of incubation. Cells were subsequently passed through a cell strainer (70-μm pore size) to remove clumps and tissue and sorted as described above. Sorted cells were subsequently incubated individually or recombined with LN cells from Tg OT-I mice (4 × 104) at various ratios (5:1 to 0.625:1) in U-bottom 96-well plates for 48 h prior to pulsing with [3H]thymidine (1 μCi). Cells were incubated for a further 18 h prior to harvesting onto glass fiber filters and measurement of incorporated [3H]thymidine (TopCount NXT; Perkin Elmer).
Statistical comparisons were performed using GraphPad Prism, version 5, software. Analysis of variance (ANOVA) or Students t test was used to deduce significant differences among the results. The Bonferroni posttest comparison was used to report P values.
To define which cell type(s) produced type I IFNs in response to FPV, FL- or GM-CSF-expanded dendritic cells (DCs), BM-derived macrophages, or MEFs were infected in vitro with FPVWT and assayed for IFN-α and IFN-β secretion. FL-expanded DCs, comprised predominantly of pDCs and cDCs, were responsible for the majority of both IFN-α and IFN-β secreted (33.6 ng/ml and 13.9 ng/ml, respectively). In contrast, GM-CSF-expanded DCs produced no IFN-α above background levels and only a minimal amount of IFN-β (4.2 ng/ml) (Fig. (Fig.1a,1a, i and ii), and type I IFN secretion from macrophages and MEF cells was consistently below detection levels. To confirm that pDCs within the FL cultures were responsible for producing type I IFNs, we conducted intracellular cytokine staining for IFN-α and IFN-β on FL cultures infected with FPVWT and treated with brefeldin A. The results indicated that pDCs produced the majority of IFN-α, with 4.43% of pDCs producing IFN-α compared to 0.13% of cDCs Fig. 1b, i). In regard to IFN-β secretion, pDCs again produced the majority of IFN-β (1.74%) while cDCs failed to produce levels of IFN-β above background (0.30%) (Fig. (Fig.1b,1b, ii). Similarly, FL cultures sorted into pDC and cDC populations by magnetic sorting prior to infection with FPV corroborated that the majority of type I IFNs were produced by pDCs: 76 ng/ml IFN-α compared to 6.8 ng/ml in cDC cultures (Fig. 1c, i), and 5.63 ng/ml IFN-β compared to 0.32 ng/ml in cDCs (Fig. (Fig.1c,1c, ii). Together these results indicated that pDCs were the major producers of type I IFNs in response to FPV infection. To investigate the importance of the type I IFN receptor (IFNAR) in the secretion of type I IFNs, BM cells from receptor-deficient or control mice were expanded in the presence of FL and subsequently infected with FPVWT. There was a complete abrogation of IFN-α (Fig. 1d, i) and IFN-β (Fig. (Fig.1d,1d, ii) production in IFNAR-deficient FL-expanded DCs compared to infected control cultures (0.3 versus 60 ng/ml and 0.6 versus 9.6 ng/ml, respectively), clearly demonstrating that that FPV-mediated type I IFN secretion required positive feedback signaling through IFNAR.
To determine whether infection with FPV results in proinflammatory cytokine secretion, FL- or GM-CSF-expanded DCs or MEFs were infected with FPV and analyzed for the production of IL-6, IL-12, TNF-α, and IL-1β. It was clear that FL-expanded DCs produced the majority of IL-6 (~4 ng/ml) (Fig. (Fig.2a,2a, i) and IL-12 (~0.34 ng/ml) (Fig. 2b, i) at levels 3- to 4-fold greater than those observed in uninfected cultures, with almost no secretion detected in GM-CSF-expanded cultures or the MEF cultures. In contrast, IL-1β was produced only by GM-CSF-expanded DCs (~2.1 ng/ml) (Fig. 2d, i) while both FL- and GM-CSF-derived DCs were capable of producing TNF-α (~0.41 and 0.30 ng/ml, respectively) (Fig. 2c, i). This suggested that FL-expanded DCs produced the majority of proinflammatory cytokines; however, GM-CSF-expanded DCs also contributed importantly as secretors of IL-1β and TNF-α. These findings implied that either the pDCs were primarily responsible for the majority of the cytokine secretion or that another factor in the culture (produced by the pDCs) was required to stimulate secretion from the other cell types present.
To determine which DC subsets within the FL cultures were responsible for secreting each of the cytokines, intracellular cytokine staining was performed following infection with FPV. The results of these experiments showed that both the cDC and pDC subsets were capable of secreting IL-6 (8.62% and 1.63%, respectively) (Fig. (Fig.2a,2a, ii), IL-12 (33.65% and 13.32%) (Fig. (Fig.2b,2b, ii), and TNF-α (19.22% and 11.15%) (Fig. (Fig.2c,2c, ii); however, in all cases cDCs produced the majority of each of the cytokines. Unsurprisingly, neither pDCs or cDCs from the FL cultures produced IL-1β (Fig. (Fig.2d,2d, ii), given that none was detected in these cultures when they were analyzed by ELISA (Fig. 2d, i). Interestingly, though, when the DC subsets were purified prior to infection with FPV and cytokine production was measured, pDCs produced almost all of the IL-12 (0.22 ng/ml) (Fig. (Fig.2b,2b, iii) and TNF-α (0.05 ng/ml) (Fig. (Fig.2c,2c, iii) while cDCs did not produce any cytokine above background levels. Surprisingly cDCs and pDCs produced comparable levels of IL-6 in response to FPV (0.4 and 0.32 ng/ml, respectively) (Fig. (Fig.2a,2a, iii). Very low levels of IL-1β were secreted by cDCs while no IL-1β was generated by the pDCs above the amount observed in uninfected cultures (Fig. (Fig.2d,2d, iii). These results indicated that in isolation, pDCs produce the majority of cytokines, particularly IL-12 and TNF-α, while both subsets can produce IL-6 and only cDCs are capable of secreting IL-1β. In contrast, when pDCs and cDCs exist within the same culture, both subsets produce high levels of inflammatory cytokines, and cDCs actually produce these cytokines to higher levels than the pDCs. Together, these results indicate the likelihood that pDCs provide initial stimulation or help, via cross talk to cDCs, to enable their high-level secretion of proinflammatory cytokines.
We next examined proinflammatory cytokine secretion in FL- and GM-expanded and FPV-infected cultures derived from IFNAR−/− and control mice to ascertain whether positive-feedback signaling was necessary for this response. Here, it was found that FPV infection of IFNAR-deficient FL-expanded DCs resulted in a complete abrogation of IL-6, IL-12, and TNF-α production in comparison to infected control cells: for IL-6, 0.25 versus 3.17 ng/ml (Fig. (Fig.3a,3a, i); for IL-12, 0.05 versus 0.26 ng/ml (Fig. 3b, i); and for TNF-α, 0.02 versus 0.23 ng/ml (Fig. 3c, i). As expected, no IL-1β production was observed in any FL-expanded DC cultures (Fig. 3d, i). In regard to GM-CSF-expanded DCs, those derived from IFNAR-deficient mice had significantly reduced IL-1β secretion compared to control infected cultures (0.34 versus 1.6 ng/ml) (Fig. (Fig.3d,3d, ii). No IL-12 production was observed in any GM-CSF-derived DC cultures (Fig. (Fig.3b,3b, ii), whereas both IL-6 (Fig. (Fig.3a,3a, ii) and TNF-α (Fig. (Fig.3c,3c, ii) secretions were reduced in IFNAR-deficient cultures, although at overall lower levels than those observed from infected FL-expanded DC cultures. Taken together, these data indicated that, as with type I IFNs, the production of proinflammatory cytokines in response to infection with FPV was also dependent on feedback signaling through IFNAR.
To gain a fundamental understanding as to whether the defect in proinflammatory cytokine secretion in IFNAR-deficient mice was possibly due to an intrinsic defect in upstream signaling capabilities, NF-κB activation was also assessed. Bone marrow cells taken from control or IFNAR-deficient mice were expanded in vitro in the presence of FL and infected with FPVWT. At various time points, the levels of IκBα phosphorylation were measured by Western blot analysis of cell lysates as an indirect indicator of the state of NF-κB activation. These results indicated that mice deficient in IFNAR were still capable of activating NF-κB, as evidenced by the presence of levels of phospho-IκBα higher than background at 8 h postinfection; however, compared to the intensity of levels obtained from control infected cultures, it appeared that at the 8-h time point, the level of NF-κB activation was reduced (Fig. 3e, i). When the densities of the resulting phospho-IκBα bands were quantified as percentages of the corresponding actin controls, it was observed that the level of IκBα phosphorylation was approximately halved in the IFNAR−/− cells compared to the amount in control cells at 8 h postinfection (23.5% versus 51.9%), while it was actually slightly higher at 4 h postinfection (14.8% versus 9.94%) (Fig. (Fig.3e,3e, ii). Even though IFNAR-deficient cells retained the ability to activate NF-κB, this did not result in proinflammatory cytokine secretion, and it suggests that if type I IFNs themselves are also responsible for ultimately regulating this response, then this must be occurring downstream of the cytosolic aspects of the NF-κB signaling pathway.
In an effort to determine whether IFNAR−/− mice might have a defect which may inhibit the maturation of their dendritic cells and, hence, secretion of proinflammatory cytokines, bone marrow cells cultured from IFNAR−/− and control mice in the presence of FL were stimulated for 24 h with either PBS, FPVWT, or CpG and then analyzed for upregulation of the DC activation markers CD40, CD80, and CD86 in both pDC and cDC populations. Results showed broadly similar upregulation of all activation markers in both SV129 × B6 (Fig. (Fig.4a)4a) and IFNAR−/− (Fig. (Fig.4b)4b) pDC and cDC populations following infection with FPV. FPVWT-infected pDCs from SV129 × B6 mice had a clear but small upregulation of CD40 (Fig. 4a, i), slight upregulation of CD80 (Fig. (Fig.4a,4a, ii), and strong upregulation of CD86 (Fig. (Fig.4a,4a, iii). In comparison, cDCs from the same culture showed the same pattern of enhanced upregulation of all three maturation markers, with strong upregulation of CD40 (Fig. (Fig.4a,4a, iv), CD80 (Fig. 4a, v), and CD86 (Fig. (Fig.4a,4a, vi), and in all cases was very similar to results observed following CpG stimulation. In the IFNAR−/− cultures, results showed a very similar outcome, with low levels of upregulation of CD40 and CD80 observed in pDCs (Fig. 4b, i) and (Fig. (Fig.4b,4b, ii) and higher levels of CD86 (Fig. (Fig.4b,4b, iii); however, the extent of this shift was lower than that seen in the wild-type cultures. Analysis of the results from the IFNAR−/− cDCs showed that CD40 (Fig. (Fig.4b,4b, iv), CD80 (Fig. 4b, v), and CD86 (Fig. (Fig.4b,4b, vi) were all upregulated following infection with FPVWT, in a similar manner to upregulation induced following CpG stimulation. Together, these results indicate no obvious defect in DC maturation between control (SV129 × B6) and IFNAR−/− cultures and indicate that DC maturation occurs to a similar extent in both the pDC and cDC populations.
Given our findings that IFNAR expression was critical for type I IFN and proinflammatory cytokine secretion in response to FPV infection, we next sought to ascertain whether these defects impaired the subsequent induction of an adaptive immune response to FPVOVA. To this end, control or IFNAR-deficient mice were immunized with FPVOVA, and resultant adaptive immune responses were analyzed 5 days later. Interestingly, we found that both control and IFNAR-deficient mice exhibited a strong capacity to lyse OVA-specific target cells in a direct CTL assay (at a 200:1 effector/target ratio, 64% and 48% lysis, respectively) compared with unimmunized control animals (4% lysis at an E/T ratio of 200:1) (Fig. 5a, i). Furthermore, there was no statistically significant difference in the FPVOVA-mediated increase in anti-OVA helper T-cell recall proliferation responses between control and receptor-deficient mice (11,647 cpm versus 14,486 cpm) (Fig. 5b, i). These similarities in cell-mediated immune responses were also evident in anti-FPV humoral immunity, where the amounts of anti-FPV IgG antibodies found in serum samples taken from control and IFNAR-deficient FPVOVA-immunized mice (OD of 0.47 versus OD of 0.38) (Fig. 5c, i) were not altered.
To confirm that signaling through IFNAR and the production of type I IFNs themselves in response to FPV infection were, in fact, inconsequential to the formation of adaptive immune responses to our FPV-encoded antigen, serum type I IFNs were neutralized by the administration of sheep antiserum to IFN-α/β prior to and following the immunization of mice with FPVOVA. Here, in support of our previous findings, we found no significant difference in anti-OVA cytolytic activity in mice depleted of type I IFNs compared to control immunized mice (35% versus 34% lysis; E/T ratio of 200:1) (Fig. (Fig.5a,5a, ii). Similarly, T-helper cell proliferation was not affected (1,585 cpm versus 1,720 cpm) (Fig. (Fig.5b,5b, ii), nor were anti-FPV IgG serum antibody responses (OD of 1.1 versus OD of 0.9) (Fig. (Fig.5c,5c, ii). When taken together, these results support the notion that production of type I IFNs and the capacity to signal through IFNAR did not contribute to the development of adaptive immune responses against FPVOVA.
As adaptive immune responses to FPVOVA remained unaffected by the absence of both IFNAR and type I IFNs, we sought to determine whether this might be due to the intrinsic antiviral activities of the type I IFNs themselves in reducing viral protein synthesis and in enhancing the clearance of virus-infected cells. Accordingly, bone marrow cells from IFNAR−/− and control mice were expanded in FL prior to infection with FPVOVA and subsequent analysis of secreted OVA protein. Results from these experiments showed that IFNAR−/− FL-DC cultures routinely contained 2- to 3-fold more OVA than FPVOVA-infected control cultures, while uninfected cells and those infected with FPVWT failed to produce any OVA (Fig. (Fig.6a).6a). To examine whether IFNAR deficiency altered the levels of major histocompatibility complex class I (MHC-I)-restricted antigen presentation in vivo, control and IFNAR-deficient mice were administered FPVOVA via footpad injection, and on days 0, 3, 6, and 9 postinfection, mice received CFSE-labeled OT-I Tg T cells. Three days later, transferred Tg T cells were recovered from the footpad-draining (popliteal) LN and assessed for CFSE dye dilution as a direct measure of OVA-specific Tg CD8+ T-cell proliferation resulting from MHC-I-restricted OVA antigen expression (31, 45). The degree of OT-I proliferation has previously been shown to correlate with the level of OVA antigen expression and, hence, provides an opportunity to track the location and extent of antigen expression in vivo (29, 34). As shown in representative proliferation plots from individual mice (Fig. 6b, i), OT-I Tg T cells recovered from IFNAR-deficient mice exhibited a significantly enhanced proliferative response compared to those observed in control mice on days 0 and 3 post-FPVOVA infection (day 0, PI of 19 versus 8; day 3, PI of 12 versus 4). This is indicative of an increase in the magnitude of FPV-derived MHC-I-restricted OVA antigen expression at these time points. While the extent of OT-I T-cell proliferation in IFNAR−/− mice decreased at a rate similar to that observed in control mice, it was clear that even at days 6 and 9 postinfection, OT-I Tg T cells were still entering into cellular proliferation (Fig. (Fig.6b,6b, ii), which was suggestive of a delay in clearance of virus-infected cells. This was in contrast to control infected mice, where almost complete clearance of FPVOVA was observed 3 days after infection. These data support the notion that type I IFN signaling through IFNAR is necessary for efficient clearance of virus-infected cells and may also indicate a possible compensatory mechanism, whereby a failure to directly influence anti-OVA adaptive immunity through the conventional stimulatory role of type I IFNs might be overcome by enhanced and prolonged viral antigen expression.
We have previously shown that pDCs are essential for the induction of a cytolytic response directed against an FPV-expressed antigen (11); however, the results from our current study suggest that this effect did not lie in the capacity of pDCs to secrete large amounts of type I IFNs, as originally postulated, as here we clearly show that they appear to make no contribution to the induction of adaptive immune responses to recombinant FPV. Recent reports have indicated that pDCs can also present antigens and therefore contribute directly to the development of adaptive immune responses (35, 55). With this in mind, we sought to determine whether the crucial role of pDCs in the formation of anti-FPV adaptive immunity may lie in their antigen presentation functions. In the first instance, FL- and GM-CSF-expanded DC cultures were left uninfected, infected with FPVOVA, or pulsed with the peptide SIINFEKL in vitro, washed, and injected into T-cell-depleted B6bm1 mice (mice unable to present the dominant OVA peptide SIINFEKL in the context of H-2Kb to cognate T cells). Recipient mice subsequently received CFSE-labeled OT-I T cells, and proliferative responses were analyzed 3 days later. As expected, no proliferation of OT-I Tg T cells was observed in the mice which received uninfected FL- or GM-CSF-expanded DC cultures (PIs of 1.08 and 1.06) (Fig. (Fig.7a),7a), whereas the same cultures pulsed with SIINFEKL proliferated strongly (PI of 4.97 versus PI of 3.39). Interestingly, only FL-expanded DCs infected with FPVOVA resulted in a high proliferative response in vivo (PI of 17.6), while GM-CSF-expanded DCs infected with FPVOVA could not drive OT-I Tg T-cell proliferation (PI of 1.07). To confirm that pDCs were important for antigen presentation in vivo, endogenous pDC (PDCA+) and cDC (PDCA− CD11c+) populations were sorted from splenocyte suspensions prepared from B6 mice injected with FPVOVA 24 h prior to isolation. Sorted populations were cocultured with OT-I Tg T cells, and T-cell proliferation was subsequently measured by radioactive thymidine incorporation. Interestingly, pDC and recombined pDC/cDC cultures had the highest levels of antigen presentation as assessed by OT-I Tg T-cell proliferation (352.8 and 395.8 cpm, respectively) while cDC cultures stimulated lower levels of T-cell proliferation (198.4 cpm) (Fig. (Fig.7b).7b). This finding suggested that in this particular scenario, pDCs were more efficient at presenting OVA than cDCs.
We have previously shown that the adaptive immune response to OVA antigen expressed by recombinant FPV is predominately cell mediated and lacking in any significant memory component, considered in part to be due to a failure of FPV to illicit strong helper T-cell responses and a defective ability to provide DC licensing signals (11). In addition, we have shown that adaptive immune responses to FPV-encoded antigen are highly dependent on pDCs, which was originally considered to be attributable to a reliance on these cells to provide type I IFNs. Here, we expand on these findings and confirm that pDCs are responsible for the secretion of the majority of type I IFNs in response to FPV and that this is dependent on positive feedback signaling through IFNAR.
These findings are consistent with the results of other studies utilizing similar viral model systems which show that pDCs are the major producers of type I IFNs (56) and that high levels of secretion are dependent on positive feedback signaling through the IFNAR (23). In most virus-infected cells, the transcription factor IRF-7 is considered the master regulator of type I IFN production (23), and expression of IRF-7 itself is dependent in part on IFN-β signaling through IFNAR. Furthermore, IFN-β is produced following the upregulated expression and kinase-dependent activation of IRF-3, a process mediated by PRR recognition of virus-derived nucleic acid (reviewed in references 41 and 53). This, in turn, signals IFNAR-dependent IRF-7 transcription, and together with IRF-3, both become activated via PRR-dependent kinase activity, ultimately driving high levels of type I IFN production (51). Plasmacytoid dendritic cells possess the ability to constitutively express IRF-7 (16) and, hence, have the capacity to secrete type I IFN in the absence of the IFNAR-restricted positive-feedback loop (3). Therefore, we hypothesize that the initial spike of feedback-independent IFN secretion, which is normally sufficient to initiate the feedback amplification loop, is either not operating here or is simply below the level of detection in the IFNAR−/− bone marrow cultures. We therefore predict that the secretion levels we observe in IFNAR-intact bone marrow cultures and animals results from repeated cycles of amplification via a mechanism of positive-feedback signaling through IFNAR (43).
In addition to type I IFNs, proinflammatory cytokines are likewise critically important contributors to effective innate immunity, as well as playing a role in determining how the adaptive response will be biased (38). Interleukin-6 secretion has a profound impact on inflammation through its ability to direct leukocyte recruitment, activation, and eventually apoptosis and also through its contribution to the differentiation of B cells into antibody-producing plasma cells, in conjunction with type I IFNs (26). In response to infection with FPV, FL-expanded DCs secreted the majority of IL-6, and within this culture both the pDC and cDC subsets were capable of producing the cytokine, with neither subset requiring the other for IL-6 generation. Interestingly, IL-6 production was almost completely abrogated in IFNAR−/− cultures, indicating that in a similar manner to the secretion of type I IFNs, IFNAR expression and, thus, high levels of type I IFNs are required to make IL-6 (21, 62). IL-12 production results in the polarization of the ensuing adaptive immune responses, promoting a cell-mediated phenotype through its ability to influence the development of type I helper T cells from the common precursor cell population (46). The majority of IL-12 produced in response to FPV also originated in the FL-expanded cultures, suggesting that its secretion is dependent on pDCs, either directly through their ability to actively secrete IL-12 (9) and/or indirectly through the provision of factors, such as type I IFNs which are required for IL-12 secretion from other cell types (9, 17, 42). When the identity of the subset of DCs secreting IL-12 within the FL-expanded culture was defined, it was clear that the cDCs within this culture produced the majority of the IL-12 although the pDC subset was also clearly capable of this. However, when the cultures were sorted prior to infection with FPV, it appeared that when in isolation, it was the pDCs and not cDCs that could produce IL-12. Therefore, it appears that the presence of pDCs is required for high-level expression of IL-12 from cDCs and indicates that pDCs provided an initial stimulatory signal for this to occur. The same result was observed with regard to TNF-α secretion, with cDCs producing the majority of the TNF-α within the FL-expanded culture; when the cultures were isolated prior to infection, however, only the pDCs from the same culture could do this. Tumor necrosis factor alpha is an important antiviral cytokine as it is capable of driving apoptosis to eliminate virus-infected cells or promote inflammation and cell survival through activation of various downstream transcription factors following binding to its cognate receptor (4). When examining the role for IFNAR in IL-12 and TNF-α secretion, we observed that in both cases IFNAR and, by inference, type I IFNs were required for high-level secretion of these cytokines. This observation is supported by other studies which suggest that simultaneous activation of both STAT1 and NF-κB is important in the priming of IL-12 production in DCs (54) and that TNF-α secretion can be influenced by type I IFN secretion (32). Finally IL-1β, an important mediator in the inflammatory response and a downstream indicator of inflammasome activation (37), was produced exclusively in GM-CSF-expanded DCs although when cDCs were sorted and concentrated from FL-expanded cultures, they, too, were also able to produce IL-1β, albeit at low levels. Again, as with the other proinflammatory cytokines, IL-1β secretion was almost completely abrogated in DCs from IFNAR−/− mice, demonstrating a critical role for type I IFNs not only in the secretion of proinflammatory cytokines but also possibly in activation of the inflammasome (21, 62).
As it appeared that type I IFNs were required for productive secretion of proinflammatory cytokines, it was important to establish whether IFNAR−/− mice carried any inherent defects in upstream signaling which might lend themselves to such outcomes. While the type I IFN and NF-κB signaling pathways are generally considered separate entities, leading to activation of interferon-stimulatory genes mediated by the JAK-STAT pathway (63) or the regulation of over 400 diverse gene targets in the case of NF-κB (1), there is some evidence to suggest that type I IFN secretion itself can drive NF-κB. This activation can result from the serine phosphorylation and degradation of IκBα, mediated by phosphatidylinositol 3-kinase and Akt (14, 58, 59). Type I IFN-mediated activation of NF-κB was analyzed here through measurement of IκBα phosphorylation in IFNAR−/− cultures. Results of these experiments showed reduced levels of IκBα phosphorylation in infected IFNAR−/− bone marrow cells, indicating that defects in proinflammatory cytokine secretion may be partially attributable to aberrant NF-κB activation. This result supports prior observations that type I IFNs can drive NF-κB activation, although it appears unlikely that the complete abrogation of cytokine secretion is entirely due to this phenomenon as there is clearly a degree of residual NF-κB activation in these cells following infection with FPV.
Another possible mechanism which could result in an abrogation of the proinflammatory cytokine response to FPV was a defect in the maturation in IFNAR−/− DCs. Previous studies with viruses including human adenovirus (AdV) have shown that DCs from IFNAR−/− mice are incapable of undergoing full maturation in response to viral infection (22, 47) although other studies indicate little or no defect in DC maturation (33). The results of our experiments showed no obvious defect in the maturation of DCs from IFNAR−/− mice in response to FPV as upregulation of CD40, CD80, and CD86 was similar to that observed in the control groups, and the profiles were also broadly the same as those seen following stimulation with the synthetic TLR9 agonist, CpG. Together, these results showed that FPV is capable of stimulating DC maturation in a similar fashion to CpG, and they indicate that type I IFNs are not required for this to occur.
Upon examination of adaptive immune responses to FPVOVA in IFNAR−/− mice, we predicted that in light of our findings surrounding the importance of type I IFNs in type I IFN and proinflammatory cytokine secretion, an obvious defect in the formation of an adaptive immune response to FPVOVA would result. However, no significant differences were observed between IFNAR−/− and control animals with regard to cell-mediated or humoral immune responses following immunization with FPVOVA. This was profound, considering the host of functions that type I IFNs have in forming these responses (2, 16). To shed some light on the ability of IFNAR−/− mice to generate normal adaptive immunity, the extent of antigen production within IFNAR−/− cells and mice was assessed to determine whether more antigen was produced due to the absence of the inhibitory effects imposed on viral transcription and translation by type I IFNs. To address this question, FL-expanded bone marrow cells from IFNAR−/− and control mice were infected with FPVOVA, and levels of OVA expression were measured. The results indicated that IFNAR−/− cells consistently produced greater quantities of OVA than control cells, suggesting that type I IFNs dampen viral antigen expression, a phenomena also seen following MCMV and VV infection (10, 60). Interestingly, IFNAR−/− mice also displayed increased levels of MHC-I-restricted viral antigen expression and systemic distribution, as evidenced by OVA-specific T-cell proliferation in all lymph nodes, not just those draining the footpad. Due to the importance of type I IFNs in facilitating viral clearance (48), it is perhaps not surprising, then, that IFNAR−/− mice have impaired clearance of virus-infected cells even though FPV does not replicate in mammals. While it appears by the simplest explanation that type I IFNs have no role in adaptive immune responses to FPV infection, this result, detailing enhanced viral antigen expression and persistence in IFNAR−/− mice, raises the possibility that IFNAR−/− mice fail to display any defect in adaptive immune responses simply as a consequence of enhanced and prolonged antigen expression. However, it is unlikely that this is the sole reason as there was no evidence of enhanced anti-OVA adaptive immunity in mice immunized with FPVOVA, which had also been treated with type I IFN-depleting antibodies.
While we have previously shown an absolute requirement for pDCs for the induction of adaptive immunity to FPV, the results described here indicate no obvious requirement for type I IFNs. This is profound, considering the important role type I IFNs play in regulating adaptive immunity to a number of other viral infections and, of particular note here, the dsDNA viruses VV and adenovirus (AdV5), which lack the ability to elicit an adaptive immune response in the absence of type I IFNs (61, 62). We therefore propose that the main role of pDCs here is not to provide type I IFNs but to act as antigen-presenting cells (APCs) (13), a property recently shown to involve ligation of TLR7 and TLR9 (35), which are, coincidently, the two same TLRs that we have found to be responsible for innate recognition of FPV (E. L. Lousberg, submitted for publication). In antigen presentation experiments which involved the proliferation of OT-I T cells in vitro and in vivo, we demonstrated that pDCs had an important role in antigen presentation and were in fact responsible for presenting antigen to a significantly greater extent than cDCs when DCs were harvested from a previously immunized mouse and coincubated with OT-I T cells. Whether this is due to increased antigen uptake or cross-presentation by pDCs is currently not known but remains the subject of ongoing investigations.
Despite these findings, we do not believe that the induction of adaptive immunity to FPV-encoded antigen rests solely with antigen presentation functions of pDC as cDCs are also clearly capable of this. Rather, there exists a defect or defects in the DC subset cross talk and networking capacity in rFPV-immunized mice compared to the cross talk and response elicited by tropic viruses. This is because the total DC population is comprised of a network of different subsets with distinct functions that combine to sense pathogenic challenge and to orchestrate innate and adaptive immune responses (44). Indeed, we saw evidence of DC cross talk in secretion of proinflammatory cytokines, where pDCs were necessary for high-level secretion of IL-12 and TNF-α in particular. Together, then, these findings provide fundamental insights into how FPV is recognized by the mammalian immune system and give significantly valuable new directions for possible future rational modifications to FPV to ultimately improve and enhance immunization responses to new vaccines based on this durable and compliant vector technology.
The work was supported in part by funding from an Australian Research Council Linkage Project Grant LP05618109 (to J.D.H.), and E.L.L. was supported by the Cancer Council of South Australia through provision of a Ph.D. stipend. Material support for this study was provided by Virax Holdings Ltd.
We acknowledge Fares Al-Ejeh for critical reading of the manuscript.
Published ahead of print on 21 April 2010.