Induction of Immunological Tolerance to Adenoviral Vectors by Using a Novel Dendritic Cell-Based Strategy

Abstract

The success of helper-dependent adenoviral (HD-Ad) vector-mediated lung gene therapy is hampered by the host immune response, which limits pulmonary transgene expression following multiple rounds of vector readminstration. Here, we show that HD-Ad-mediated pulmonary gene expression is sustained even upon three rounds of readministration to immunodeficient mice, highlighting the need to suppress the adaptive immune response for sustained gene expression following vector readministration. Therefore, we devised a dendritic cell (DC)-based strategy for induction of immunological tolerance toward HD-Ad vectors. DCs derived in the presence of interleukin-10 (IL-10) are refractory to HD-Ad-induced maturation and instead facilitate generation of IL-10-producing Tr1 regulatory T cells which suppress HD-Ad-induced T cell proliferation. Delivery of HD-Ad-pulsed, IL-10-modified DCs to mice induces long-lasting immunological tolerance to HD-Ad vectors, whereby pulmonary DC maturation, the T cell response, and antibody response to HD-Ad vectors are suppressed even after three rounds of pulmonary HD-Ad readministration. Moreover, sustained transgene expression is also observed in the lungs of mice immunized with HD-Ad-pulsed, IL-10-modified DCs even after three rounds of pulmonary HD-Ad delivery. Taken together, these studies identify the use of DCs generated in the presence of IL-10 as a novel strategy to induce long-lasting immune tolerance to HD-Ad vectors.

INTRODUCTION

Adenoviral (Ad) vectors have been extensively studied for pulmonary gene therapy due to their ability to efficiently transduce a wide variety of proliferating and nonproliferating cells (2, 14, 31). Adenoviruses are a family of DNA viruses with a linear, double-stranded genome of about 36 kb. Initially, the use of first-generation adenoviral (FG-Ad) vectors demonstrated substantial host immune responses to viral antigens, leading to destruction of transduced cells and prevention of readministration (36). Significant improvement in the safety and efficacy of Ad-based vectors came with the development of helper-dependent adenoviral (HD-Ad) vectors, which do not encode any viral genes (24, 25, 29). In contrast to FG-Ad, HD-Ad vectors are able to mediate long-term, high-level transgene expression in the absence of the chronic toxicity observed with FG-Ad due to the absence of viral coding sequences. We and our collaborators have previously demonstrated unprecedented levels of transgene expression when HD-Ad vectors were delivered to the airway of rabbits (11) and baboons (1). Although with HD-Ad vectors the immune response is reduced, subsequent vector readministration can increase it to levels seen with FG-Ad vectors, thereby limiting transgene expression (12). Therefore, there is a need to induce tolerance within the host to the HD-Ad vector without compromising the immunity to other infections to mediate stable gene expression following multiple vector readministrations to the lung.

Dendritic cells (DCs) are professional antigen-presenting cells derived from the same bone marrow (BM) precursors as macrophages. DCs reside in the tissues as immature DCs which, in the presence of appropriate signals, turn into mature DCs. Mature DCs are potent stimulators of T cell proliferation and effector T cell development. In contrast to mature DCs, immature DCs have been implicated in the generation of peripheral tolerance through regulatory T cell (Treg) development (15). Tregs are critical in preventing autoimmunity by suppressing autoreactive T cells (34). Restimulation of cord blood-derived naïve CD4⁺ cells with immature DCs has been shown to induce development of Tregs, whereas restimulation with mature DCs leads to a Th1 effector phenotype (10). Therefore, the maturation status of DCs is critical in deciding between tolerance and immunity. Moreover, adoptive transfer of immature DCs into rats 7 days before cardiac transplant has been shown to significantly enhance graft survival through induction of Tregs (6). Although several different subsets of Tregs have been identified, the two most well characterized are the Foxp3⁺ Tregs and type 1 regulatory (Tr1) Tregs (15). Foxp3⁺ Tregs are induced by transforming growth factor β (TGF-β) and are characterized by the expression of the transcription factor Foxp3, whereas immature DCs can drive induction of Tr1 Tregs, which do not express Foxp3 and instead are characterized by production of interleukin-10 (IL-10). Therefore, we hypothesize that immature DCs presenting HD-Ad-derived epitopes may be used to induce tolerance to HD-Ad vectors.

In this study, we assessed for the feasibility of inducing immunological tolerance to HD-Ad vectors using immature DCs pulsed with HD-Ad vectors. DCs generated in the presence of IL-10 were refractory to HD-Ad-induced DC maturation and, instead of inducing T cell differentiation, primed differentiation of IL-10-producing regulatory T cells, which suppressed T cell proliferation. Delivery of HD-Ad-pulsed, IL-10-modified DCs to mice suppressed the adaptive immune response against HD-Ad vectors following intranasal delivery and also primed differentiation of IL-10-producing Tr1 Tregs in vivo, which could suppress HD-Ad-induced T cell proliferation. After three rounds of vector delivery, mice that received HD-Ad-pulsed, IL-10-modified DCs had significantly elevated levels of vector-encoded transgene expression compared to control groups. Altogether, these findings highlight a new DC-based strategy to induce tolerance to HD-Ad vectors which allows for sustained gene expression in the airways after multiple rounds of vector readministration.

MATERIALS AND METHODS

Generation of adenoviral vectors.

pC4HSU HD-Ad backbone vector (Ad 5 serotype [Ad5]) was obtained from Merck and was used for generation of empty adenoviral vectors. This vector contains “stuffer” fragments of human genomic DNA without repeat units and lacks retrovirus elements and known genes (30). Escherichia coli β-galactosidase (β-Gal) cDNA with a nuclear localization signal was separately cloned into an expression cassette containing control elements from the human cytokeratin 18 (K18) gene. This construct was then cloned into the AscI site of the pC4HSU HD-Ad vector. The infectivity of HD-Ad-K18LacZ, encoding LacZ under the control of the K18 promoter, was tested in COS7 cells. Determination of titers from blue-forming units allowed an estimate of the biological activity of the vector. For comparisons, we used an FG-Ad vector expressing β-Gal (FG-Ad-CMVlacZ) (where CMV is cytomegalovirus) and an HD-Ad vector expressing β-Gal (HD-Ad-CMVlacZ). HD-Ad vectors were obtained following a previously described protocol with minor modifications (26). Contamination by helper virus in the HD-Ad preparations was determined to be less than 0.1% by real-time PCR analysis.

Virus delivery.

Solutions of DEAE-dextran and viral particles were heated separately to 30°C, followed by addition of DEAE-dextran to the virus to a final concentration of 10 μg/ml. The mixture was then heated for 20 min at 30°C to allow complex formation, and delivery was performed intranasally (16). Six- to 10-week-old C57BL/6 mice or Rag-1-deficient mice (Charles River Laboratories, Wilmington, MA) were used for the experiment. All the animals were housed at SickKids animal facility or the Toronto Centre for Phenogenomics, and all the animal studies were conducted following institutional guidelines. Mice were briefly anesthetized by isoflurane (Aerrane) inhalation, and virus solution was placed in small drops on the nares, from where it was aspirated into the lungs.

β-Galactosidase activity assay.

Three days after vector delivery, mouse airways, including lungs and trachea, were isolated, homogenized, and assayed for β-galactosidase activity using a chemiluminescence kit (Galacto-Light; Applied Biosystems, Foster City, CA) and a microplate luminometer (LB96V; EG & G Berthold, Bad Wildbad, Germany), as described previously (12), and the values were normalized to the total protein concentration.

X-Gal staining.

Three days after vector delivery, lungs were isolated, and staining with 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) of the whole lung was performed as described previously (33). After being stained, tissues were washed and fixed in 4% paraformaldehyde (PFA) for 4 h, followed by transfer in 70% ethanol, and cleared using methyl salicylate prior to being photographed.

Generation of BM-derived dendritic cells (BMDCs).

C57BL/6 mice were used as a source of bone marrow cells. Briefly, humerus, tibia, and femur bones were isolated, and BM cells were flushed out using saline. BM cells were cultured in the presence of granulocyte-macrophage colony-stimulating factor ([GM-CSF] Peprotech, Rocky Hill, NJ) only to generate normal DCs or in the presence of 70 ng/ml IL-10 (Peprotech) to generate IL-10-modified tolerogenic DCs. Around day 3 of culture, nonadherent cells were removed, for they were comprised of mostly lymphoid lineages. DCs were ready for use on day 7, during which either they were used as is or incubated with HD-Ad particles for 24 h.

T cell proliferation assay.

DCs incubated overnight with adenoviral vectors were irradiated (25 Gy) and cocultured with naïve T cells for a period of 3 to 7 days. During the last 24 h of culture, bromodeoxyuridine (BrdU) was added to be incorporated by proliferating cells. T cell proliferation was measured using a cell proliferation chemiluminescence kit (Roche, Laval, QC, Canada).

Suppression assay.

HD-Ad-pulsed DCs were used as stimulators, and naive CD4⁺ T cells were used as responders. The stimulators (2.5 × 10⁴ cells/well) and responder cells (2.5 × 10⁵ cells/well) were cultured in 96-well round-bottom plates, and suppressors (CD4⁺ T cells) were isolated after 5 days of coculture of naive CD4⁺ T cells and HD-Ad-pulsed normal or IL-10-modified DCs. These suppressors were added at a ratio of 1:1, 1:5, or 1:10. Proliferation was assessed at day 4 of coculture using a cell proliferation chemiluminescence kit (Roche) following the manufacturer's instructions.

Assessment of regulatory T cell generation.

CD4⁺ naïve T cells were cultured with HD-Ad-incubated DCs. After 5 days, IL-10 production by CD4⁺ T cells was assessed by flow cytometry.

In vivo delivery of dendritic cells.

A total of 5 × 10⁶ DCs suspended in 150 μl of saline were delivered via thoracic injection as described previously (27).

BrdU labeling.

At different time points after vector delivery, mice were anesthetized using isoflurane inhalation, and then BrdU was administered intranasally in a volume of 50 μl at a concentration of 16 mg/ml. At 24 h after BrdU delivery, mice were killed, and lungs/draining mediastinal lymph nodes (MLN) were isolated.

In vivo labeling of DCs.

Mice were anesthetized using isoflurane inhalation, and 50 μl of 1 mg/ml fluorescein isothiocyanate (FITC)-dextran (Sigma-Aldrich) was delivered intranasally 2 h prior to viral delivery as described previously (13).

Proliferation and cytokine production by T cells from lymph nodes.

Six days after challenge with HD-Ad vector particles, mice were sacrificed, and MLN were isolated. T cells isolated from MLN by nylon wool enrichment were cocultured with HD-Ad-treated, irradiated BMDCs for a period of 72 h, and BrdU solution was added during the last 12 h to label the cells, after which proliferation was measured using a cell proliferation chemiluminescence kit (Roche) following the manufacturer's instructions. Medium was collected and used for measurement of cytokine production.

Preparation of single-cell suspensions from lungs.

At different time points after HD-Ad vector delivery, mice were sacrificed by intraperitoneal (i.p.) injection of sodium pentobarbital (Euthanyl; Bimeda-MTC, QC, Canada). In order to collect bronchoalveolar lavage fluid (BALF), mouse lungs were lavaged as described previously (12). After lavage was performed, lungs were perfused with 10 ml of phosphate-buffered saline (PBS) containing 10 U/ml heparin via the right ventricle of the heart in order to remove blood cells from the lung vasculature. Perfusion was performed until the lungs turned completely white. Lungs were dissected, and after removal of draining MLN, lungs were minced and digested for 25 min at 37°C using 250 U/ml collagenase D (Roche) solution, with addition of EDTA (10 mM final) during the last 5 min of incubation. Fragments of digested lungs were passed through a 100-μm-pore-size cell strainer (BD Biosciences), and hypotonic lysis was used to remove erythrocytes. Similarly, draining MLN were digested and then suspended at appropriate concentrations.

Cytokine measurements.

Cytokine levels were measured using IL-1β, IL-10, gamma interferon (IFN-γ), and IL-12 enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems, Minneapolis, MN).

Flow cytometry.

All the antibodies were obtained from eBioscience (San Diego, CA) unless indicated otherwise. For in vitro DC experiments, DCs were labeled at 4°C with CD11c and maturation markers. In order to assess pulmonary T cell proliferation, cells from BALF/lung or MLN single-cell suspensions were stained for CD3 and CD4/CD8 with/without BrdU staining as described previously (3). Migration of FITC-dextran-labeled DCs was assessed by labeling of MLN single-cell suspensions with CD11c. DCs in the airways were identified by staining single-cell lung suspensions with CD11c and CD11b antibodies. Maturation of DCs in vivo was assessed by staining with CD86 antibody as a marker for DC maturation. Cells were gated according to their forward scatter (FSC) versus side scatter (SSC) characteristics to discriminate highly autofluorescent macrophages from DCs, as described previously (7). Intracellular cytokine staining of T cells was performed as described previously (17, 18). Intracellular cytokine staining for adenoviral uptake was performed using anti-hexon antibody (Chemicon, Temekula, CA) as described previously (35). Flow cytometry data were acquired for each of the experiments using a BD FACSCalibur instrument (where FACS is fluorescence-activated cell sorter) at the SickKids-University Health Network (UHN) Flow Cytometry Facility and were analyzed using FlowJo flow cytometry analysis software (Treestar, OR).

ELISA for anti-Ad antibodies.

A pan-specific (IgA, IgE, IgGs, and IgM) ELISA for mouse anti-human Ad5 antibodies was performed as previously described (5, 12). Ninety-six-well ELISA plates (Costar; Corning, Acton, MA) were coated overnight at 4°C with 5 × 10¹⁰ particles of HD-Ad particles per well in 100 mM bicarbonate buffer, pH 9.6. The plates were then washed four times with TBS-T (Tris-buffered saline [TBS] with 0.05% Tween 20) and blocked for 3 h at room temperature with 3% bovine serum albumin (BSA) in TBS. Mouse BALF diluted 1:100 or serum diluted 1:2,000 in TBS was added to the wells for overnight incubation at 4°C. The plates were then washed four times with TBS-T and incubated with anti-mouse Ig-biotin (BD Pharmingen, San Diego, CA), diluted 1:5,000 in TBS, for 3 h at room temperature. The plates were washed four times with TBS-T and then incubated with avidin-alkaline phosphatase (Sigma-Aldrich), diluted 1:50,000 in TBS, for 2 h at room temperature. The plates were washed four times with TBS-T, incubated with 1 mg/ml p-nitrophenyl phosphate (Sigma-Aldrich) in 100 mM diethanolamine buffer, pH 9.8, containing 0.5 mM MgCl₂ for 10 min at room temperature. The reaction was stopped by the addition of EDTA (40 mM final), and optical density was read at 405 nm.

RESULTS

Sustained gene expression is observed in Rag-1-deficient mice following readministration of HD-Ad vectors to the lung.

In order to assess the role of the adaptive immune response in mediating loss of transgene expression observed with HD-Ad readministration to the lung, we used Rag-deficient mice, which lack mature T and B cells and are deficient in adaptive immune responses (23). Vector delivery was carried out as illustrated in Fig. 1A. Single intranasal delivery of HD-Ad vectors encoding LacZ under the control of the K18 promoter (HD-Ad-K18LacZ) resulted in similar β-galactosidase activities in lungs of Rag-deficient and wild-type counterparts (Fig. 1B). However, upon double delivery, where mice first received empty HD-Ad particles (C4HSU) followed by HD-Ad-K18LacZ particles, β-galactosidase activity was significantly reduced in the lungs of wild-type mice compared to levels in Rag-deficient mice (Fig. 1C). Similarly, upon triple delivery where mice received two rounds of C4HSU delivery followed by HD-Ad-K18LacZ, β-galactosidase activity was further reduced in the lungs of wild-type mice compared to levels in Rag-deficient mice (Fig. 1D). Double delivery led to a 50% loss of β-galactosidase activity in the lungs of wild-type mice, which further increased to 82% loss of activity after triple delivery compared to mice that received single delivery (Fig. 1E). In contrast, only a 20 to 30% reduction in β-galactosidase activity was observed in the lungs of Rag-deficient mice upon both double and triple delivery. Taken together, these findings indicate that upon multiple readministrations of HD-Ad particles to the lung, there is a substantial loss in transgene expression, and the loss is largely mediated by an adaptive T and B cell response, which is targeted toward the HD-Ad vector-derived antigens.

Fig 1.

Gene expression is sustained in the lungs of Rag-deficient mice following multiple rounds of HD-Ad vector delivery. (A) Strategy for multiple HD-Ad readministrations to the lung. Levels of β-galactosidase activity were determined in the lung extracts from mice receiving single delivery (B), double delivery (C), and triple delivery (D) of HD-Ad vectors. (E) Histogram comparing loss in β-galactosidase activity in the lungs of mice following double or triple delivery relative to single delivery. Results are shown as mean β-galactosidase activity (relative light units [RLU]/μg of protein) ± standard deviation for 9 animals in each group (*, P < 0.05, for the Rag-deficient group versus the wild-type group). vp, virus particles.

DCs generated from bone marrow cells cultured in the presence of GM-CSF and IL-10 are refractory to HD-Ad-induced maturation.

In the absence of HD-Ad stimulation, CD11c⁺ DCs generated in the presence of only GM-CSF (normal DCs) are in an immature state, indicated by very low levels of CD86 expression; however, upon incubation with HD-Ad particles there is massive induction of CD86 expression, with approximately 80% of DCs undergoing maturation (Fig. 2A). In contrast, DCs generated by culturing bone marrow cells in the presence of high concentrations of IL-10 (IL-10-modified DCs) were refractory to HD-Ad-induced CD86 expression, for only 12 to 13% of DCs became CD86⁺ following stimulation with HD-Ad particles (Fig. 2A). Furthermore, expression of other maturation markers such as CD80, major histocompatibility complex class II (MHC-II), and CCR7, which are induced upon exposure of DCs to HD-Ad vectors, was also reduced on IL-10-modified DCs compared to normal DCs following HD-Ad stimulation (Fig. 2B). Addition of IL-10 to normal DCs prior to HD-Ad stimulation did not prevent CD86 induction, highlighting the requirement of IL-10 throughout DC differentiation for generation of DCs refractory to HD-Ad-induced maturation (see Fig. S1 in the supplemental material).

Fig 2.

Dendritic cells generated in the presence of high concentrations of IL-10 are refractory to HD-Ad-induced maturation. (A) Representative FACS plots looking at CD86 expression on DCs cultured in the absence of IL-10 (normal DCs) or in the presence of IL-10 (IL-10 mod. DC) upon stimulation with HD-Ad vectors (+HD-Ad). (B) Representative FACS histogram of CCR7, CD80, and MHC II expression. (C) Representative FACS profile of hexon staining upon incubation of DCs with HD-Ad vectors. Concentrations of IL-1β (D) and IL-12 (E) released upon stimulation of DCs by HD-Ad vectors are shown. Values (mean ± standard error) are representative of six independent experiments (*, P < 0.05, for the normal DCs versus the IL-10-modified DCs).

Following incubation with HD-Ad particles, both normal DCs and IL-10-modified DCs showed similar levels of positivity for adenoviral hexon protein staining, indicating that IL-10-modified DCs were not impaired in uptake of adenoviral vector particles (Fig. 2C). Although the proportions of DCs positive for adenoviral hexon proteins were similar, mean fluorescence intensity (MFI) analysis indicated that the MFI of adenoviral hexon protein staining was higher in IL-10-modified DCs than in normal DCs, indicating that the overall uptake of HD-Ad particles was higher by IL-10-modified DCs. However, following HD-Ad stimulation, the levels of inflammatory cytokines IL-1β and IL-12 secreted by IL-10-modified DCs were significantly reduced compared to those for normal DCs were significantly reduced (Fig. 2D and E). Taken together these studies indicate that DCs derived from bone marrow cells in the presence of IL-10 are refractory to HD-Ad-induced maturation and inflammatory cytokine production.

HD-Ad-pulsed dendritic cells derived in the presence of IL-10 induce generation of IL-10-producing regulatory T cells.

Under basal conditions, without incubation with HD-Ad particles, DCs are poor stimulators of T cell proliferation (Fig. 3A). In contrast, following HD-Ad incubation, normal DCs induce robust T cell proliferation, which was also observed upon addition of exogenous IL-10 to the coculture. In contrast, following HD-Ad incubation, IL-10-modified DCs were significantly impaired in their ability to induce T cell proliferation (Fig. 3A). In order to further confirm the impairment of HD-Ad-pulsed, IL-10-modified DCs to prime HD-Ad-specific T cell proliferation, we also measured frequencies of responding T cells. The frequency of T cells responding to HD-Ad-pulsed, IL-10-modified DCs was significantly reduced compared to the frequency of T cells responding to HD-Ad-pulsed normal DCs (see Fig. S2 in the supplemental material). Taken together, these findings indicate that HD-Ad-pulsed, IL-10-modified DCs are impaired in driving HD-Ad-specific T cell proliferation.

Fig 3.

Dendritic cells generated in the presence of high concentrations of IL-10 are impaired in priming HD-Ad-specific T cell proliferation and instead induce generation of IL-10-producing Tr1 cells. (A) Measurement of naive T cell proliferation upon stimulation by HD-Ad-pulsed normal or IL-10-modified DCs in the presence/absence of exogenous IL-10 (RLU, relative light units, is a chemiluminescent measurement of BrdU incorporation by proliferating cells). (B) Representative FACS plots looking at IL-10 production by T cells cultured with HD-Ad-pulsed DCs. (C) Levels of IL-10 released by T cells stimulated by HD-Ad-pulsed DCs. (D) CD4⁺ T cells were isolated from the culture and then added to a coculture of naive CD4⁺ T cells (responders) and HD-Ad-pulsed normal DCs as suppressors at three different ratios of 1:1, 1:5, and 1:10. Four days later, cell proliferation was assessed by BrdU incorporation assay, and data are presented as percentages of suppression of T cell proliferation compared with that of naive CD4⁺ T cells cultured in the presence of HD-Ad-pulsed normal DCs without addition of any CD4⁺ T cells from other cocultures. Values (mean ± standard deviation) are representative of five independent experiments (*, P < 0.05, for IL-10-modified DCs versus normal DCs).

Both HD-Ad-pulsed normal DCs and HD-Ad-pulsed, IL-10-modified DCs were poor inducers of Foxp3⁺ Tregs (see Fig. S3 in the supplemental material). Therefore, we tested for their ability to drive generation of type 1 regulatory T cells (Tr1 Tregs), characterized by IL-10 production. HD-Ad-pulsed normal DCs induced IL-10 production in less than 1% of CD4⁺ T cells, which increased to 5 to 6% of CD4⁺ T cells upon culture with HD-Ad-pulsed, IL-10-modified DCs (Fig. 3B). Furthermore, naive CD4⁺ T cells cultured with HD-Ad-pulsed, IL-10-modified DCs secreted significantly elevated levels of IL-10 compared to T cells cultured with HD-Ad-pulsed normal DCs, further confirming enhanced propensity of HD-Ad-pulsed, IL-10-modified DCs to drive Tr1 differentiation (Fig. 3C). This IL-10 production was specifically induced upon interaction with HD-AD-pulsed, IL-10-modified DCs since in the absence of HD-Ad pulsing, no IL-10 production was observed (Fig. 3C). Next, we assessed the ability of Tr1 cells generated upon culture of naive CD4⁺ T cells with HD-Ad-pulsed, IL-10-modified DCs to suppress HD-Ad-specific T cell proliferation. Only the T cells from coculture with HD-Ad-pulsed, IL-10-modified DCs had the ability to suppress HD-Ad-specific T cell proliferation, indicating that culture of naive T cells with IL-10-modified DCs led to induction of Tr1 cells with an ability to suppress HD-Ad-specific T cell proliferation in vitro (Fig. 3D).

HD-Ad-pulsed, IL-10-modified DCs suppress pulmonary DC maturation and migration in vivo upon intranasal challenge with HD-Ad vectors.

In order to test the ability of IL-10-modified DCs to induce tolerance to HD-Ad vectors in vivo, we immunized mice with either HD-Ad-pulsed normal DCs or HD-Ad-pulsed, IL-10-modified DCs and subsequently assessed pulmonary DC maturation and migration following intranasal challenge with HD-Ad particles. Assessment of pulmonary DC maturation was carried out by monitoring expression of CD86 on pulmonary DCs (Fig. 4A and B). Delivery of saline followed by HD-Ad delivery led to induction of CD86 on pulmonary DCs, which was also observed upon delivery of normal DCs followed by delivery of HD-Ad vectors, indicating pulmonary dendritic cell maturation in response to HD-Ad vectors. In contrast, CD86 expression levels were reduced on pulmonary DCs from HD-Ad-challenged mice which received HD-Ad-pulsed, IL-10-modified DCs compared to HD-Ad-challenged mice which received HD-Ad-pulsed normal DCs, indicating that delivery of HD-Ad-pulsed, IL-10-modified DCs prevented pulmonary DC maturation following HD-Ad delivery (Fig. 4A). Furthermore, the profile of CD86 expression on pulmonary DCs from HD-Ad-challenged mice which received HD-Ad-pulsed, IL-10-modified DCs was in fact similar to the CD86 expression profile on pulmonary DCs from saline-challenged mice, indicating the potency of IL-10-modified DCs in suppressing pulmonary DC maturation in response to HD-Ad vectors. Delivery of HD-Ad vectors to saline-treated mice or to mice immunized with HD-Ad-pulsed normal DCs led to an approximately 40% increase in the proportion of pulmonary CD86⁺ DCs compared to mice exposed only to saline (Fig. 4B). In contrast, only a 20% increase in the proportion of pulmonary CD86⁺ DCs was observed in HD-Ad-challenged mice which were immunized with HD-Ad-pulsed, IL-10-modified DCs, further confirming the ability of HD-Ad-pulsed, IL-10-modified DCs to induce immune tolerance toward HD-Ad vectors.

Fig 4.

Upon delivery of HD-Ad-pulsed, IL-10-modified DCs, mice are impaired in HD-Ad-induced pulmonary DC maturation. (A) Representative FACS histogram showing CD86 expression on pulmonary DCs of mice exposed to saline, mice exposed to saline followed by HD-Ad vectors (saline then HD-Ad), mice immunized with HD-Ad-pulsed normal DCs followed by HD-Ad vectors (Normal DC then HD-Ad), or mice immunized with IL-10-modified DCs followed by HD-Ad vectors (IL-10 mod. DC then HD-Ad). (B) Increase in proportions of CD86⁺ pulmonary DCs relative to levels in saline-treated mice. (C) Proportions of FITC-dextran-positive DCs in the MLN of mice given the above treatments. Values (mean ± standard deviation) are representative of five independent experiments (*, P < 0.05, for IL-10-modified DCs treated with HD-Ad versus the two groups indicated).

In order to assess pulmonary DC migration, prior to HD-Ad challenge, mice were given FITC-dextran to label pulmonary DCs, and following HD-Ad challenge, proportions of FITC-dextran-positive DCs were quantified in the mediastinal lymph nodes (MLN) (Fig. 4C). Upon HD-Ad challenge, approximately 20% of the DCs in the MLN migrated from the lungs in saline-treated mice and mice immunized with HD-Ad-pulsed normal DCs. However, in the lymph nodes of mice immunized with HD-Ad-pulsed, IL-10-modified DCs, following HD-Ad challenge, only 10% of the DCs were migratory, indicating an impairment of pulmonary DCs to migrate to the draining lymph nodes following HD-Ad challenge. These findings further confirm that delivery of HD-Ad-pulsed, IL-10-modified DCs suppressed pulmonary DC migration to the draining lymph nodes following challenge with HD-Ad vectors.

HD-Ad-pulsed, IL-10-modified DCs suppress T cell infiltration following intranasal challenge with HD-Ad vectors.

In order to further confirm that delivery of HD-Ad-pulsed, IL-10-modified DCs induced long-lasting tolerance toward HD-Ad vectors in vivo, mice were challenged with one or three rounds of HD-Ad delivery, and subsequently we assessed T cell infiltration in the bronchoalveolar lavage fluid (BALF) (Fig. 5A and B). In response to a single delivery of HD-Ad vectors, approximately 25 to 30% of the cells in BALF were T cells, which were also observed in mice immunized with HD-Ad-pulsed normal DCs (Fig. 5A; see also Fig. S4 in the supplemental material). Furthermore, upon three rounds of HD-Ad delivery, the proportion of BALF T cells increased to 50% (Fig. 5B; see also Fig. S4). In contrast, T cell proportions in the BALF of HD-Ad-challenged mice immunized with HD-Ad-pulsed, IL-10-modified DCs were significantly reduced to only 5 to 7% upon single HD-Ad delivery and only 10 to 15% following three rounds of HD-Ad delivery, highlighting the ability of HD-Ad-pulsed, IL-10-modified DCs to mediate long-term tolerance to HD-Ad vectors (Fig. 5A and B).

Fig 5.

Reduced T cell infiltration is observed in lungs of HD-Ad-challenged mice immunized with HD-Ad-pulsed, IL-10-modified DCs. Mice were immunized with HD-Ad-pulsed normal or IL-10-modified DCs and then were challenged intranasally with a single dose of HD-Ad (5 × 10¹⁰ vector particles) vectors or three doses of HD-Ad vectors separated by 3-week intervals. Histograms compare the proportions of CD3⁺ T cells in the BALF of mice on day 7 following the first delivery (A) or third delivery (B) of HD-Ad vectors, assessed by flow cytometry. (C) Ratio of CD4/CD8 T cells in the lungs of mice. (D) Absolute count of CD8⁺ T cells in the lungs of mice. (E) Levels of IFN-γ secreted by T cells from MLN cells following stimulation with HD-Ad-pulsed normal DCs after first or third delivery of HD-Ad vectors (n = 12 per group). Values are mean ± standard deviation (*, P < 0.05, suggesting statistical significance of the group of IL-10-modified DCs challenged with HD-Ad versus other groups).

We also measured the ratios of CD4 to CD8 T cells in the lungs of mice following challenge with HD-Ad vectors, with a lower ratio indicative of an increased CD8⁺ T cell infiltration (Fig. 5C). Following a single delivery of HD-Ad vectors, a CD4/CD8 ratio of 1.5 to 2 was observed in the lungs of saline-treated mice or mice immunized with HD-Ad-pulsed normal DCs. However, this ratio was further reduced following three rounds of HD-Ad delivery, indicating an increase in pulmonary CD8⁺ T cell infiltration. However, a CD4/CD8 ratio of 4 was observed in the lungs of mice immunized with HD-Ad-pulsed, IL-10-modified DCs following challenge with a single dose or three doses of HD-Ad vectors, indicating a reduction in infiltration of CD8⁺ T cells into the lung and thereby mediating long-term tolerance induction (Fig. 5C). Moreover, absolute cell counts of CD8⁺ T cells in the lungs of challenged mice confirmed that delivery of IL-10-modified DCs led to reduction in infiltration of CD8⁺ T cells in the lung after both the first and third rounds of HD-Ad vector challenge (Fig. 5D). Furthermore, we also measured IFN-γ levels upon stimulation of T cells from MLN cells with HD-Ad-pulsed normal DCs (Fig. 5E). T cells from HD-Ad-challenged, saline-treated mice or mice immunized with HD-Ad-pulsed normal DCs secreted elevated IFN-γ levels following HD-Ad stimulation, which were further elevated upon challenge of mice with three doses of HD-Ad vectors (Fig. 5E). In contrast, T cells from mice immunized with HD-Ad-pulsed, IL-10-modified DCs and challenged with a single dose or three doses of HD-Ad vectors secreted significantly reduced levels of IFN-γ following HD-Ad stimulation (Fig. 5E). These results indicate that HD-Ad-pulsed, IL-10-modified DCs mediate long-term tolerance induction to HD-Ad vectors in vivo with a reduced CD8⁺ T cell response against HD-Ad vectors.

HD-Ad-pulsed, IL-10-modified DCs suppress T cell proliferation following HD-Ad vector challenge.

Following a single delivery of HD-Ad particles, approximately 18 to 20% of the T cells were proliferating in the MLN of saline-treated mice or mice immunized with HD-Ad-pulsed normal DCs (Fig. 6A). In contrast, the proportion of proliferating T cells was reduced to only 6% in the MLN of mice immunized with HD-Ad-pulsed, IL-10-modified DCs. Following three rounds of HD-Ad delivery, the proportions of proliferating T cells increased to 35 to 40% in the MLN of saline-treated mice or mice immunized with HD-Ad-pulsed normal DCs, which was reduced to only 12% in the MLN of mice immunized with HD-Ad-pulsed, IL-10-modified DCs (Fig. 6A). Similarly, proliferation of T cells from MLN in response to HD-Ad vector stimulation was significantly reduced from mice immunized with HD-Ad-pulsed, IL-10-modified DCs compared to saline-treated mice or mice immunized with HD-Ad-pulsed normal DCs following either a single round or three rounds of HD-Ad delivery (Fig. 6B). Taken together, these findings confirmed the ability of HD-Ad-pulsed, IL-10-modified DCs to induce long-lasting immunological tolerance toward HD-Ad vectors.

Fig 6.

Reduced T cell proliferation in response to pulmonary HD-Ad challenge is observed in mice immunized with HD-Ad-pulsed, IL-10-modified DCs. Mice were immunized with HD-Ad-pulsed normal or IL-10-modified DCs and then were challenged intranasally with a single dose of HD-Ad vectors (5 × 10¹⁰ vector particles) or three doses of HD-Ad vectors separated by 3-week intervals. (A) Representative FACS histogram showing BrdU⁺ cells among CD3⁺ T cells in the MLN on day 7 following HD-Ad delivery. (B) Proliferation of T cells from MLN upon stimulation with HD-Ad-pulsed normal DCs in vitro (RLU, relative live units, is a chemiluminescent measurement of BrdU incorporation by proliferating cells). (C) Levels of IL-10 released upon stimulation of T cells from MLN upon stimulation with HD-Ad-pulsed normal DCs in vitro (n = 12 per group). Values are mean ± standard deviation (*, P < 0.05, for IL-10-modified DC group versus the saline-treated and normal DCs).

Our in vitro studies identified the ability of HD-Ad-pulsed, IL-10-modified DCs to drive induction of IL-10-producing Tr1 cells. In order to confirm the induction of Tr1 cells in vivo, we measured IL-10 levels following stimulation of T cells from MLN of mice immunized with HD-Ad vectors (Fig. 6C). T cells from MLN of saline-treated mice or mice immunized with HD-Ad-pulsed normal DCs secreted very low levels of IL-10 following HD-Ad stimulation. In contrast, T cells from MLN of mice immunized with HD-Ad-pulsed, IL-10-modified DCs secreted significantly elevated levels of IL-10 following HD-Ad stimulation, indicating that delivery of HD-Ad-pulsed, IL-10-modified DCs was able to drive Tr1 generation in vivo.

HD-Ad-pulsed, IL-10-modified DCs suppress antibody responses following HD-Ad vector challenge.

In order to measure the antibody response to HD-Ad vectors, we measured adenovirus-specific antibody titers in the serum and BALF (Fig. 7). Following single delivery of HD-Ad vectors, low levels of anti-Ad antibody production were observed in the serum and BALF of saline-treated mice or mice immunized with HD-Ad-pulsed normal DCs, and levels were significantly reduced both in the serum and BALF of mice immunized with HD-Ad-pulsed, IL-10-modified DCs following single HD-Ad challenge. Upon three rounds of HD-Ad vector challenge, the titers of anti-Ad antibodies were severalfold higher in the serum and BALF of saline-treated mice or mice immunized with HD-Ad-pulsed normal DCs than in mice that received a single round of HD-Ad vector delivery. However, the levels of anti-Ad antibody titer were significantly reduced both in the serum and BALF of mice immunized with HD-Ad-pulsed, IL-10-modified DCs following three rounds of HD-Ad challenge. Taken together, these findings indicate that delivery of HD-Ad-pulsed, IL-10-modified DCs induce long-lasting tolerance characterized by suppression of the antibody response against HD-Ad vectors.

Fig 7.

Reduced anti-Ad antibodies in response to pulmonary HD-Ad challenge is observed in mice immunized with HD-Ad-pulsed, IL-10-modified DCs. Mice were immunized with HD-Ad-pulsed normal DCs or HD-Ad-pulsed, IL-10-modified DCs and then were challenged intranasally with a single dose of HD-Ad vectors (5 × 10¹⁰ vector particles) or three doses of HD-Ad vectors separated by 3-week intervals. Histograms comparing anti-Ad antibody levels in serum (A) and BALF (B) following HD-Ad challenge are shown (n = 8 per group). Values are mean ± standard deviation (*, P < 0.05, for the IL-10-modified DCs versus the saline-treated and normal DCs). OD, optical density.

Sustained gene expression is observed in the lungs of mice which receive HD-Ad-pulsed, IL-10-modified DCs following multiple rounds of HD-Ad-mediated gene delivery.

The aim of tolerance induction to HD-Ad vectors is to prevent loss of gene expression associated with multiple rounds of HD-Ad vector readministration. Therefore, to assess sustained gene expression following multiple rounds of HD-Ad gene delivery in mice with tolerance to HD-Ad vectors, we devised a gene delivery protocol as shown in Fig. 8A. Single delivery of HD-Ad-K18LacZ was associated with robust β-galactosidase activity in the lung along with X-Gal staining all throughout the airways (Fig. 8B and C). However, in nontolerant mice which received multiple rounds of readministration, β-galactosidase activity in the lung was significantly reduced, with very faint LacZ staining in the airways (Fig. 8B and C). Similar results were observed in tolerant mice that received ovalbumin (OVA)-pulsed, IL-10-modified DCs, followed by multiple rounds of readministration, indicating that delivery of OVA-pulsed, IL-10-modified DCs failed to induce tolerance to HD-Ad vectors. In contrast, β-galactosidase activity was only slightly reduced in the lungs of mice which received HD-Ad-pulsed, IL-10-modified DCs following multiple rounds of vector delivery and was significantly higher than that in the nontolerant group or the group with tolerance to OVA. Taken together, these findings indicate that delivery of IL-10-modified, HD-Ad-pulsed DCs induces tolerance to HD-Ad vectors, which prevents loss in gene expression associated with readministration of HD-Ad vectors to the lung.

Fig 8.

Gene expression following multiple rounds of pulmonary HD-Ad vectors is sustained in mice immunized with HD-Ad-pulsed, IL-10-modified DCs. (A) Outline of the readministration strategy. (B) β-Galactosidase activity in the lungs of HD-Ad-challenged mice. Results are shown as mean β-galactosidase activity (RLU/μg of protein) ± standard deviation. (C) X-Gal staining of the lungs following HD-Ad delivery (n = 15 per group). *, P < 0.05, for the tolerant (HD-Ad) and single-delivery groups compared to the other groups.

DISCUSSION

Adaptive immune response against HD-Ad vectors is a major barrier in limiting transgene expression following vector readministration to the lung. This was confirmed in Rag-deficient mice, which have a defective adaptive immune response, for upon HD-Ad vector readministration, transgene expression was sustained in the lungs of these mice compared to that in their wild-type counterparts. Moreover, upon multiple rounds of HD-Ad vector readministration to wild-type mice, there was an increase in CD8⁺ T cell infiltration to the lung along with a severalfold increase in T cell proliferation and anti-Ad antibody titers, confirming an association of HD-Ad vector readministration with magnification of an adaptive immune response toward these vectors. Therefore, it is important to devise strategies to induce tolerance to HD-Ad vectors so that vector readministration can be carried out and gene expression can be sustained. Our results show that generation of DCs in the presence of high concentrations of IL-10 (IL-10-modified DCs) leads to generation of tolerogenic DCs that are impaired in undergoing maturation in response to HD-Ad vectors and instead induce generation of IL-10-secreting Tr1 Tregs with an ability to suppress HD-Ad-specific T cell proliferation. Moreover, delivery of HD-Ad-pulsed, IL-10-modified DCs to mice leads to induction of tolerance to HD-Ad vectors. In tolerant mice, maturation of pulmonary DCs in response to intranasal HD-Ad delivery is impaired. Moreover, upon several rounds of HD-Ad challenge, in tolerant mice the T cell response toward HD-Ad vectors is significantly diminished along with a significant reduction in anti-Ad antibody titers, confirming the ability of HD-Ad-pulsed, IL-10-modified DCs in inducing long-term tolerance toward HD-Ad vectors.

The effects of IL-10 on DCs have been investigated in maintaining the immature status of DCs. Initial studies indicated that IL-10 results in inhibition of DC-driven IFN-γ production by purified CD4⁺ and CD8⁺ T cells (21). IL-10 has been further shown to inhibit upregulation of CD86 on DCs. Addition of IL-10 to in vitro cultures of skin-derived dermal DCs resulted in significant reduction of CD86 and CD80 expression compared to levels in mature dermal DCs. At the same time, IL-10-treated dermal DCs showed an approximately 50% reduction in their ability to activate T cells (22). IL-10 treatment also results in enhanced phagocytosis by DCs, which, when coupled with downregulation of costimulatory molecules, results in an immature DC phenotype. The effects of IL-10 on DCs correlate well with inhibition of primary T cell responses, for addition of anti-IL-10 antibodies promotes DC maturation (4). Retroviral transduction of Epstein-Barr virus (EBV) IL-10 into DCs has also been shown to result in an immature phenotype (32). Moreover, recent studies have shown that delivery of IL-10-differentiated DCs can mediate tolerance induction in asthmatic mice through generation of Tregs (9, 20).

In our study, we were able to induce HD-Ad-specific tolerance using IL-10-modified DCs pulsed with HD-Ad vectors. The Tregs generated in vitro were IL-10 secreting and Foxp3⁻ with the ability to suppress HD-Ad-specific T cell proliferation. These Tregs were likely Tr1 Tregs, for human tolerogenic DCs generated in vitro using IL-10 have been shown to prime differentiation of antigen-specific Tr1 Tregs (8). Delivery of HD-Ad-pulsed, IL-10-modified DCs to mice suppressed pulmonary DC maturation and migration in response to HD-Ad vectors. It is likely that IL-10-modified DCs in response to HD-Ad vectors secrete IL-10 in vivo, which has an immunosuppressive effect on pulmonary DCs, thereby suppressing their maturation and migration in response to HD-Ad vectors. Moreover, HD-Ad-pulsed, IL-10-modified DCs likely prime HD-Ad-specific Tr1 cells in vivo since T cells from the draining lymph nodes of HD-Ad-challenged mice immunized with HD-Ad-pulsed, IL-10-modified DCs secreted elevated levels of IL-10, indicating that HD-Ad-pulsed DCs primed induction of Tr1 cells in vivo. Transfer of Tr1 regulatory T cells has been shown to abrogate DC-mediated immune response in an IL-10-dependent manner (37). Therefore, following pulmonary HD-Ad challenge, it is likely that Tr1 cells specific to HD-Ad vectors secrete IL-10, which can have an anti-inflammatory effect on pulmonary DCs, thereby suppressing their maturation. Moreover, Tregs have also been shown to directly suppress DC maturation via binding of lymphocyte activation gene 3 (LAG3) on Treg surfaces to MHC-II on DC surfaces (19).

IL-10-modified DCs were also able to significantly inhibit HD-Ad-induced T cell proliferation in mice. This was likely due to impairment of pulmonary DC migration upon HD-Ad delivery, which suppressed induction of the T cell response against HD-Ad vectors and also due to induction of HD-Ad Tregs, which further suppressed the T cell response. The T cell response is critical for induction of a B-cell-mediated antibody response (28). Therefore, suppression of a T cell response to HD-Ad vectors upon immunization with HD-Ad-pulsed, IL-10-modified DCs also suppressed generation of anti-Ad antibody titers. In accord with induction of long-lasting tolerance, transgene expression was significantly enhanced in the lungs of mice immunized with HD-Ad-pulsed, IL-10-modified DCs.

Our findings also show that generation of tolerance was specific toward HD-Ad vectors since delivery of IL-10-modified DCs pulsed with OVA was not associated with sustained gene expression following multiple rounds of HD-Ad readministration. This indicates that induction of tolerance toward HD-Ad vectors will not compromise immunity against other pathogens.

It is important to note that our findings indicate that for tolerance induction to HD-Ad vectors, DCs need to be exposed to IL-10 while they are undergoing differentiation. Therefore, strategies aimed at delivering IL-10-encoding vectors to induce tolerance may not have the potential to induce tolerance to adenoviral vectors. In a clinical setting, perhaps blood monocytes from patients can be differentiated into DCs in vitro in the presence of high concentrations of IL-10, which can then be subsequently pulsed with HD-Ad vectors and delivered to the patients prior to gene therapy for induction of tolerance to HD-Ad vectors.

This strategy to induce antigen-specific tolerance has important implications in regenerative medicine because this strategy can be applied to other gene therapy vectors and to protein and cell therapy along with transplantation.