Abstract
The addition of monophosphoryl lipid A, a minimally toxic derivative of LPS, to nonmucosally administered vaccines induced both systemic and mucosal immune responses to coadministered Ags. This was dependent on an up-regulated expression of 1α-hydroxylase (CYP27B1, 1αOHase), the enzyme that converts 25-hydroxycholecalciferol, a circulating inactive metabolite of vitamin D3, into 1,25(OH)2D3 (calcitriol). In response to locally produced calcitriol, myeloid dendritic cells (DCs) migrated from cutaneous vaccination sites into multiple secondary lymphoid organs, including classical inductive sites of mucosal immunity, where they effectively stimulated B and T cell immune responses. The endogenous production of calcitriol by monophosphoryl lipid A-stimulated DCs appeared to be Toll-IL-1R domain-containing adapter-inducing IFN-β-dependent, mediated through a type 1 IFN-induced expression of 1αOHase. Responsiveness to calcitriol was essential to promote the trafficking of mobilized DCs to nondraining lymphoid organs. Collectively, these studies help to expand our understanding of the physiologically important roles played by locally metabolized vitamin D3 in the initiation and diversification of adaptive immune responses. The influences of locally produced calcitriol on the migration of activated DCs from sites of vaccination/infection into both draining and nondraining lymphoid organs create a condition whereby Ag-responsive B and T cells residing in multiple lymphoid organs are able to simultaneously engage in the induction of adaptive immune responses to peripherally administered Ags as if they were responding to an infection of peripheral or mucosal tissues they were designed to protect.
To be effective as a prophylactic or therapeutic treatment, vaccines must be capable of stimulating the generation of appropriate adaptive immune responses so that an effective state of protection against disease is provided to immunized hosts. Vaccine-elicited responses also need to be of sufficient magnitude and duration to provide a prolonged period of immunological protection against infection. This latter property is generally provided to vaccines by the inclusion of adjuvants, additives capable of augmenting the quantitative levels of Ag-specific Abs and/or effector T lymphocyte responses. The most effective adjuvants are comprised of chemical agents capable of activating the host’s innate immune system, stimulating responses by leukocytes that tend to mimic some of the consequences of a natural infection, influencing the development of Ag-specific effector responses to coadministered Ags.
In addition to augmenting the intensity and duration of adaptive immunity, many vaccine adjuvants also possess the capacity to affect the qualitative nature of elicited effector immune responses. It has long been appreciated that the inclusion of aluminum hydroxide (Alum) into vaccines is capable of enhancing elicited humoral immune responses in a Th2 cell-biased manner (1), whereas an inclusion of highly inflammatory CFA effectively stimulates humoral and cellular immune responses through mechanisms largely regulated by Th1 cells (2).
To achieve optimal states of host protection against a particular disease through vaccination requires the immune effector responses being appropriately controlled. Therefore, it is essential to also consider how a selected adjuvant will affect the qualitative nature of elicited immune responses. Such considerations need to be secured during an early phase in the vaccine development process, to maximize protective efficiency and minimize the chances of pathologic outcomes.
The TLR4 ligand, LPS, is easily isolated from the cell walls of Gram-negative bacteria. LPS is a potent adjuvant for systemically administered experimental vaccines, although the extreme inflammatory toxicity of LPS precludes it from being clinically useful as an adjuvant in humans (3). Efforts to reduce the toxicity of LPS, while retaining its potent adjuvant properties, have led to the development of monophosphoryl lipid A (MPLA),3 an LPS derivative capable of generating potent immune responses to coadministered Ags with minimal toxicity (4, 5). Like LPS, the adjuvant properties of MPLA appear to be TLR4 dependent (4, 5, 6). Evidence indicates that the major immunostimulating properties of MPLA are mediated by a preferential activation of the Toll-IL-1R domain-containing adapter-inducing IFN-β (TRIF) signaling pathway (5, 6). This is distinctive from LPS itself, which elicits strong inflammatory responses driven by activation of the inflammatory MyD88 signaling pathway (7). When MPLA is used as a vaccine adjuvant, the systemic immune responses elicited in immunized hosts are strongly Th1 biased (8, 9).
The primary objective of the present investigation is to establish whether MPLA, in a manner similar to what we have observed with LPS, is able to effectively stimulate the simultaneous induction of both common mucosal and systemic immune responses when added as an adjuvant to nonmucosally administered vaccines (10). We report that when used as an adjuvant in s.c. administered vaccines, MPLA augments serum Ab responses that are dominated by IgG2a, and simultaneously promotes the induction of common mucosal immune responses. This is evidenced by the presence of Ag-specific IgA Abs in all tested mucosal secretions. The mucosal adjuvant properties of MPLA appear to directly correlate with its capacity to induce expression of the vitamin D3-metabolizing enzyme 1α-hydroxylase (1αOHase) in activated myeloid dendritic cells (DCs). This enzyme effectively converts the circulating precursor hormone, 25-hydroxycholecalciferol (25(OH)D3), into its bioactive form 1,25(OH)2D3 or calcitriol (11). MPLA is able to alter the migratory properties of DCs leaving vaccination sites, allowing a subpopulation of these mobilized DCs to bypass sequestration in the draining lymph nodes (LNs) and localize within multiple secondary lymphoid organs throughout the host’s body, including Peyer’s patches (PPs). These studies also established that the ability of MPLA-activated DCs to migrate beyond draining LNs depends upon their capacity to respond to locally produced calcitriol. This altered migration of DCs by the activities of MPLA was additionally found to be TRIF dependent, and required both the induction and activities of type 1 IFNs. These novel findings should prove useful in aiding the rational design of new vaccine adjuvants, tailor made to promote the optimal induction and most beneficial types of effector responses needed to achieve maximal host protection against infectious diseases.
Materials and Methods
Animals and tissues
C3H/HeN and C57BL/6 mice (8–12 wk old) were obtained from the Charles River Breeding Laboratory. OT-II transgenic mice were purchased from The Jackson Laboratory. CD4+ T cells of OT-II mice are specific for OVA peptide 323–339 (OVA323–339) in context of MHC class II (I-Ab).
TRIF−/− and vitamin D3 receptor (VDR−/−) knockout mice were bred from animals obtained from The Jackson Laboratory. All mice were kept under specific pathogen-free conditions at the Center of Comparative Medicine Animal Facility, University of Utah. The Center of Comparative Medicine at the University of Utah employs sentinel animals to monitor for the most prevalent murine pathogens and guarantees strict compliance with regulations established by the Animal Welfare Act. VDR−/− mice were maintained on a special diet containing 20% lactose, 2% calcium, and 1.25% phosphorus (TD96348; Harlan TekLad). Femurs from MyD88 knockout (MyD88−/−) and type 1 IFNR knockout (IFNR−/−) mice were provided by J. Weis (University of Utah, Salt Lake City, UT) from animals originally obtained from S. Akira (Osaka University, Osaka, Japan) and M.-K. Kaja (Washington University, Seattle, WA). Femurs from genetically mutant mice lacking the enzyme 1αOHase (1αOHase−/−) were provided by J. Whitcomb (University of Notre Dame, Notre Dame, IN).
Animal vaccination
C3H/HeN mice were s.c. immunized with vaccine formulations containing 1 μg of the diphtheria CRM 197 protein (DT; Wyeth-Ayerst Laboratories) in a colloidal suspension of aluminum hydroxide (Alum), as has been previously described (10). Experimental mice received the DT vaccine containing 20 μg of MPLA (Sigma-Aldrich). Mice immunized with DT-Alum plus 0.1 μg of calcitriol (a gift from M. Uskokovic, Hoffmann-LaRoche, Nutley, NJ) were used as a positive control for inducing common mucosal immune responses (12). Serum, vaginal washes, fecal samples, and nasal and lung lavage fluids were collected at various times post primary and booster immunizations. The levels and isotypes of anti-DT-specific Abs present in serum and mucosal secretions were quantitated by ELISA, as previously described (12, 13).
Tracking of mobilized DCs in vivo
DC tracking assays in vivo were performed, as described previously (10, 13). Fifty microliters of a 0.25% suspension of fluoresbrite carboxy YG (green) 0.2-μm microspheres (Polysciences) was injected into hind footpads of wild-type (WT) or genetically mutant mice in the presence or absence of 20 μg of MPLA or 0.1 μg of calcitriol. In experiments in which the phenotype of microsphere+ cells was being examined, cell samples were additionally stained with fluorescent mAbs directed against various cell surface markers (10).
Tracking of bone marrow-derived DCs (BMDCs) in vivo
CD11c+ DCs (BMDCs) were generated from bone marrow of adult WT, VDR−/−, MyD88−/−, IFNR−/−, TRIF−/−, or 1αOHase−/− mice on the C57BL/6 background using previously described protocols (10, 13). Enriched CD11c+ BMDCs (1 × 106/ml) were stimulated in vitro with 60 ng/ml MPLA in the presence or absence of the calcitriol precursor 25(OH)D3 (10−7 M; Sigma-Aldrich) or calcitriol (10−8 M). In some experiments, BMDCs (1 × 106/ml) were ex vivo activated with 20 μg/ml CpG ODNs (5′-TCC-ATG-ACG-TTC-CTG-ACG-TT-3′, synthesized by the University of Utah DNA core facility), 100 U/ml IFN-β (PBL InterferonSource), or both CpG ODNs and IFN-β in the presence of 25(OH)D3. Overnight-treated BMDCs were stained with 5 μM CFSE (Molecular Probes) and injected s.c. into the hind footpads of naive syngeneic C57BL/6 recipients (2 × 106/mouse). The ability of injected BMDCs to localize to various secondary lymphoid organs then was analyzed by FACS (10).
Western blot analysis
BMDCs stimulated with TLR ligands (30 or 90 ng/ml MPLA, 10 ng/ml Escherichia coli LPS, strain 0111:B4 (Sigma-Aldrich), or 20 μg/ml CpG ODNs, 100 U/ml IFN-β) or left untreated were used for analysis of 1αOHase expression, as previously described (10).
OT-II TCR transgenic CD4+ T cell activation by ex vivo pulsed DCs
OT-II TCR-transgenic CD4+ T cells, specific for the OVA323–339 peptide in association with I-Ab, were purified from the spleens (SPL) and LNs of naive OT-II mice (10–12 wk old) by positive selection using murine CD4+ T cell MicroBeads and AutoMacs (Miltenyi Biotec). Isolated T cells were stained with 5 μM CFSE and i.v. administered to naive WT or VDR−/− C57BL/6 recipients (4 × 106/mouse). After 24 h, adoptive recipients received a s.c. injection of BMDCs (2 × 106 cells/mouse) that were ex vivo activated with 60 ng/ml MPLA in the presence or absence of 10−7 M 25(OH)D3 or 10−8 M calcitriol. The DCs were simultaneously pulsed with 500 μg/ml OVA (Sigma-Aldrich). Four days later, individual secondary lymphoid organs were removed from sacrificed animals, and single-cell suspensions were prepared and stained with fluorescent anti-CD69 (H1.2F3; BD Pharmingen) Abs and analyzed by FACS.
OT-II TCR transgenic CD4+ T cell activation in vivo
Adoptive recipients of CFSE-labeled OT-II CD4+ T cells were s.c. immunized with vaccine formulations containing 50 μg of OVA in Alum with or without the addition of 20 μg of MPLA or 0.1 μg of calcitriol. Four days later, individual secondary lymphoid organs were removed, and single-cell suspensions were prepared and stained for CD69 expression before being analyzed by FACS.
Statistical analysis
Data are shown as the mean ± SD. Differences between control and experimental groups were evaluated by unpaired Student’s t test with two-tailed distributions. Values of p below 0.05 were considered significant.
Results
Nonmucosally delivered vaccines containing the adjuvant MPLA induce both common mucosal immunity and Th1-biased systemic immune responses
We recently determined that both common mucosal and systemic immune responses were simultaneously induced in response to Ags delivered nonmucosally, when the TLR4 agonist LPS was used as the vaccine adjuvant (10). In this study, an evaluation of MPLA, a minimally toxic derivative of LPS, was made to ascertain the types of adaptive immune responses induced when it was used as an adjuvant. Following the s.c. vaccination of mature adult mice with a vaccine formulation containing 1 μg of the DT in Alum (DT-Alum) and MPLA, both systemic and common mucosal immune responses were simultaneously elicited (Fig. 1⇓, A–C). Markedly elevated levels of anti-DT IgA Abs were detected in fecal, vaginal, nasal, and lung secretions following primary and/or booster recall responses (Fig. 1⇓, A and B, and data not shown). The mucosal Ab responses generated to DT-Alum administered with added MPLA were comparable to the responses generated following immunization of mice with a DT-Alum vaccine formulation containing calcitriol, the active form of vitamin D3 (Fig. 1⇓, A and B). Immunization of mice with DT-Alum alone was incapable of inducing Ag-specific mucosal immunity, while promoting good serum Ab responses (Fig. 1⇓, A–C). Both MPLA- and calcitriol-containing vaccines were capable of eliciting DT-specific serum IgG Abs that peaked at day 28 postprimary immunization and were higher in magnitude than the serum Ab responses induced by vaccines containing DT-Alum only (Fig. 1⇓C). Isotype analysis of the serum Abs revealed that MPLA addition to the vaccine formulation shifted the elicited systemic immune responses toward a Th1 dominance, evidenced by a low IgG1:IgG2a ratio, whereas the serum Ab responses generated to calcitriol-containing or DT-Alum-containing vaccines were biased toward a Th2 dominance (Table I⇓). Additional studies determined that a similar profile of mucosal and systemic humoral immune responses was generated in mice following vaccination with anthrax-protective Ag containing either MPLA or calcitriol as adjuvants (data not shown). Our data indicate that MPLA, as an adjuvant for nonmucosally delivered vaccines, simultaneously stimulates a robust common mucosal immune response and a Th1-biased serum Ab response to s.c. administered vaccines.
Vaccine formulations containing MPLA are able to promote the induction of both systemic and mucosal immune responses. Groups of mature adult C3H/HeN mice (five mice/group) were s.c. immunized with a vaccine containing 1 μg of DT-Alum in the presence or absence of MPLA (20 μg). Parallel groups of mice were immunized with DT-Alum alone, or with DT-Alum and 0.1 μg of calcitriol. Eighty days after the primary immunization, all groups of animals were s.c. reimmunized, as described above. Anti-DT IgA in fecal extracts (A), anti-DT IgA in vaginal lavage fluids (B), and anti-DT IgG in serum (C) were analyzed from samples collected at various time points postvaccination. Results are reported as the mean value of anti-DT Abs detected in the serum or mucosal secretions of five mice per group (±SD). #, Significantly higher Ab levels compared with the Ab levels in the group immunized with DT-Alum (p < 0.03); ∗, significantly higher Ab levels compared with the Ab levels in the group immunized with DT-Alum (p < 0.002). D, Fluorescent microspheres (0.2 μm) were injected into the hind footpads of mature adult C3H/HeN mice (three mice per group) in the presence or absence of MPLA (20 μg) or calcitriol (0.1 μg). After 48 h, individual lymphoid organs (PLNs, ALNs, SPL, and PPs) were analyzed for the presence of microsphere+ cells by FACS analysis. A total of 500,000 events was collected. Data presented as mean ± SD. ∗, Significantly higher numbers of microsphere+ cells present in the lymphoid organs compared with the numbers of microsphere+ cells migrated in the group injected with microsphere alone (p < 0.003). Results are representative of two independent experiments.
Employing MPLA as a vaccine adjuvant stimulates a Th1-biased systemic immune response
MPLA addition to vaccines promotes the migration of myeloid DCs from cutaneous sites of inoculation to multiple secondary lymphoid organs throughout the body
The incorporation of calcitriol or LPS into vaccine formulations as adjuvants causes an alteration to the migratory properties of maturing DCs emigrating from sites of immunization, allowing them to traffic and localize into both draining and nondraining secondary lymphoid organs throughout the body, including PPs of the small intestine (10, 13). To determine whether MPLA addition to vaccines affects DC migration in vivo, a solution containing 0.2-μm fluorescent latex microspheres was injected into the hind footpads of C3H/HeN mice in the presence or absence of added MPLA (10). A group of mice injected with microspheres plus calcitriol was used as a positive control, and the injection of microspheres alone served as a negative control. Forty-eight hours later, various secondary lymphoid organs were individually collected, and single-cell suspensions were prepared for examination by FACS analysis for the presence of microsphere+ cells. When microspheres were injected in the presence of either MPLA or calcitriol, microsphere+ cells could be detected in all tested draining and nondraining secondary lymphoid organs, including PPs (Fig. 1⇑D). The localization of microsphere+ cells was primarily restricted to the draining popliteal LNs (PLNs) following a s.c. injection of fluorescent microspheres alone (Fig. 1⇑D).
The majority of microsphere+ cells that localized into the draining LNs of C3H/HeN mice phenotyped as CD11c+CD11b+CD205− myeloid DCs (Table II⇓). The remaining microsphere+ cells phenotyped as CD11c+CD11b+CD205intCD8αlow (interstitial DCs) or CD11c+CD11b−CD205+CD8α+ (lymphoid DCs). Virtually all of the microsphere+ cells that localized to nondraining secondary lymphoid organs were DCs of myeloid origin, with minimal interstitial or lymphoid microsphere+ DCs detected (Table II⇓).
The majority of microsphere+ cells that localize within nondraining secondary lymphoid organs following a s.c. injection of fluorescent microspheres in the presence of MPLA are of myeloid origin
These data suggest that MPLA as an adjuvant, similar to calcitriol or LPS, affects the migratory properties of myeloid DCs leaving skin sites of vaccination, allowing their localization into both draining and nondraining secondary lymphoid organs.
The ability of MPLA to alter the migratory properties of DCs is dependent upon its capacity to stimulate the endogenous production of calcitriol from circulating precursors
Most of our previously reported studies were conducted with C3H/HeN strain mice (10, 12, 13). However, to identify molecular mechanisms responsible for altered DC migration and common mucosal immunity induction necessitated the investigation of animals on the C57BL/6 background. We, therefore, questioned whether a microsphere injection plus MPLA into footpads of C57BL/6 mice would yield similar results to those observed following microsphere injection into C3H/HeN mice. When microspheres plus MPLA were injected s.c. into C57BL/6 mice, the pattern of microsphere+ cell migration was similar to that observed with other strains of mice (data not shown). This indicates that C57BL/6 mice, or their BMDCs, could be used to characterize the mechanisms involved in promoting DC localization into nondraining secondary lymphoid organs following a s.c. delivery.
MPLA induces common mucosal immunity to Ags administered s.c. by facilitating the migration of maturing Ag-laden DCs from sites of vaccine administration to classical mucosal immunity induction site (e.g., PPs). This is similar to what occurs when calcitriol is used as the adjuvant. Both adjuvants might share the same underlying mechanism of the mucosal immunity induction. We therefore investigated whether BMDCs exposed to MPLA in vitro would alter their expression of 1αOHase, the enzyme that converts the inactive circulating form of vitamin D3 (25(OH)D3) into the bioactive calcitriol. The activation of DCs derived from bone marrow of WT mice with MPLA or LPS in vitro resulted in an up-regulation in protein expression of 1αOHase (Fig. 2⇓A). In contrast, their treatment with CpG ODNs failed to induce 1αOHase protein expression (Fig. 2⇓A). These data suggest that the ability of MPLA to alter the migratory properties of maturing DCs, eventually leading to the induction of common mucosal immune responses, may be due to its capacity to induce the local metabolism of 25(OH)D3 into calcitriol.
Localization of MPLA-activated DCs into nondraining secondary lymphoid organs depends upon the up-regulation of 1αOHase. A, BMDCs generated from bone marrow of C57BL/6 mice were stimulated with MPLA (30 and 90 ng/ml), LPS (10 ng/ml), or CpG ODNs (20 μg/ml) for 24 h at 37°C or left untreated. Cell lysates were then prepared and analyzed for the presence of 1αOHase by Western blot. Proteins were visualized by ECL Plus Western Blotting Detection System. The blot was later stripped and reprobed with Abs against β-actin to ensure an equal loading of protein. Results are representative of two experiments. B, BMDCs generated from bone marrow of WT or 1αOHase−/− C57BL/6 mice were treated with 60 ng/ml MPLA in the presence or absence of 10−7 M 25(OH)D3 or 10−8 M calcitriol. After 24-h incubation at 37°C, the BMDCs were washed, stained with CFSE, and s.c. injected into naive recipients (three mice/group). Forty-eight hours later, mice were sacrificed. Single-cell suspensions from individual lymphoid organs (PLNs, ALNs, SPL, and PPs) were prepared and analyzed by FACS analysis for the presence of CFSE+ cells. A total of 500,000 events was collected. Results are presented as mean ± SD. ∗, Significantly higher numbers of BMDCs migrated to the lymphoid organs than the numbers of BMDCs that migrated in the group in which they were treated with MPLA alone (p < 0.005). Results are representative of two independent experiments.
To more definitively demonstrate that the ability of MPLA to alter migratory properties of DCs depends upon 1αOHase activity, we generated DCs from the bone marrow of WT or 1αOHase−/− mice. The BMDCs were exposed for 24 h to MPLA in vitro, in the presence or absence of the inactive precursor 25(OH)D3, or bioactive calcitriol. After treatment, the BMDCs were stained with CFSE and injected into the footpads of WT naive syngeneic recipients to evaluate their migratory properties in vivo. WT BMDCs, activated with MPLA and 25(OH)D3 or calcitriol, were able to migrate to the draining and nondraining secondary lymphoid organs of recipient mice, whereas WT BMDCs activated in vitro with MPLA alone migrated only to the draining LNs (Fig. 2⇑B). BMDCs generated from 1αOHase−/− mice and activated with MPLA plus 25(OH)D3 migrated only to the draining LNs following their s.c. injection, whereas BMDCs from 1αOHase−/− mice activated with MPLA plus added calcitriol were capable of migrating into both draining and nondraining secondary lymphoid organs (Fig. 2⇑B).
These findings suggest that the ability of MPLA-matured DCs to migrate from skin sites of vaccine injection into multiple secondary lymphoid organs in the body requires an ability to endogenously produce bioactive calcitriol from its circulating precursors.
DC migration beyond the draining LNs in response to the adjuvant properties of MPLA is TRIF signaling pathway and type 1 IFN dependent
It has been well established that the effects of LPS and the less toxic MPLA on responsive cell types are dependent upon signaling through TLR4 (5, 6, 7). Stimulation of TLR4-bearing cells with these agonists results in the activation of both the MyD88- and TRIF-dependent pathways, although MPLA elicits a greater TRIF-biased signaling response than LPS (5, 6). To determine the signaling requirements necessary to alter the migratory properties of DCs mobilized by MPLA treatment, we generated DCs from bone marrow of WT, TRIF−/−, and MyD88−/− mice, and activated them in vitro with MPLA in the presence or absence of 25(OH)D3 or calcitriol. The matured BMDCs were then CFSE stained and injected s.c. into WT naive C57BL/6 recipients. BMDCs from TRIF−/− mice, activated with MPLA in the presence of 25(OH)D3, had a markedly reduced capacity to migrate and localize to nondraining secondary lymphoid organs when compared with WT BMDCs treated in a similar manner (Fig. 3⇓, A and B). The migration of TRIF−/− BMDCs treated with MPLA and calcitriol to nondraining lymphoid organs was comparable to the migration of WT BMDCs activated with MPLA plus calcitriol (Fig. 3⇓, A and B). BMDCs generated from MyD88−/− mice were fully capable of migrating into multiple secondary lymphoid organs following their activation with MPLA plus 25(OH)D3 or calcitriol (Fig. 3⇓C).
Responsiveness to type 1 IFNs is required for MPLA-activated BMDC localization to nondraining secondary lymphoid organs. BMDCs generated from bone marrow of WT (A), TRIF−/− (B), MyD88−/− (C), or IFNR−/− (D) C57BL/6 mice were treated with 60 ng/ml MPLA in the presence or absence of 10−7 M 25(OH)D3 or 10−8 M calcitriol. After 24-h incubation at 37°C, the BMDCs were washed, stained with CFSE, and s.c. injected into naive recipients (three mice/group). Forty-eight hours later, mice were sacrificed and single-cell suspensions from individual lymphoid organs (PLNs, ALNs, SPL, and PPs) were prepared and analyzed by FACS analysis for the presence of CFSE+ cells. A total of 500,000 events was collected. Results are presented as mean ± SD. #, Significantly higher numbers of BMDCs migrated to the lymphoid organs compared with the numbers of BMDCs that migrated in the group in which they were treated with MPLA alone (p < 0.05); ∗, significantly higher numbers of BMDCs migrated to the lymphoid organs compared with the numbers of BMDCs in the group in which DCs were treated with MPLA alone (p < 0.003). Results are representative of three independent experiments.
One of the major downstream effector responses that occurs following activation of the TRIF signaling pathway in DCs is the induced synthesis of type 1 IFNs (e.g., IFN-αβ) (14, 15). We therefore questioned whether DCs derived from the bone marrow of mutant animals lacking responsiveness to type 1 IFNs could be mobilized to migrate into multiple secondary lymphoid organs following their maturation with MPLA and 25(OH)D3 or calcitriol. DCs generated from bone marrow of IFNR−/− mice were unable to migrate beyond the draining LNs when matured in vitro with MPLA in the presence of 25(OH)D3, whereas BMDC maturation with MPLA plus calcitriol promoted their migration into multiple secondary lymphoid organs after injection (Fig. 3⇑D). This indicates that cellular responsiveness to type 1 IFNs appears to be required for altering the migration of BMDCs activated with MPLA in vivo. This conclusion was further supported by finding that IFNR−/− BMDCs were unable to up-regulate the expression of 1αOHase protein in response to activation with MPLA in vitro (data not shown). Additionally, we found that WT BMDCs were able to up-regulate 1αOHase expression when directly exposed to IFN-β in vitro (Fig. 4⇓A). CpG ODNs, which by themselves were unable to up-regulate 1αOHase expression in DCs, significantly increased the levels of 1αOHase expression by BMDCs treated with IFN-β (Fig. 4⇓A). WT BMDCs treated in vitro with IFN-β and CpG ODNs in the presence of 25(OH)D3 migrated beyond the draining LNs and localized into multiple nondraining secondary lymphoid organs, including PPs, following their s.c. injection into naive syngeneic recipients (Fig. 5⇓B). These results indicate that exogenously added IFN-β can stimulate an up-regulation of 1αOHase in DCs, allowing calcitriol to be endogenously produced and promoting altered DC migration in vivo. It appears that a simultaneous stimulation of the MyD88-dependent and IFN-β signaling pathways helps to maximize up-regulation of 1αOHase expression and BMDC migration to nondraining secondary lymphoid organs.
BMDCs treated with IFN-β and simultaneously activated through MyD88-dependent pathway up-regulate 1αOHase expression and are capable of migrating beyond draining LNs. A, BMDCs generated from bone marrow of C57BL/6 mice were stimulated with MPLA (30 ng/ml), CpG ODNs (20 μg/ml), IFN-β (100U/ml), or CpG plus IFN-β for 24 h at 37°C or left untreated. Cell lysates were then prepared and analyzed for the presence of 1αOHase by Western blot. B, BMDCs generated from bone marrow of WT C57BL/6 mice were treated with 60 ng/ml MPLA, CpG ODNs (20 μg/ml), or CpG ODNs plus IFN-β (100U/ml) in the presence or absence of 10−7 M 25(OH)D3. After 24-h incubation at 37°C, the BMDCs were washed, stained with CFSE, and s.c. injected into naive recipients (three mice/group). Forty-eight hours later, mice were sacrificed and single-cell suspensions from individual lymphoid organs (PLNs, ALNs, SPL, and PPs) were prepared and analyzed by FACS analysis for the presence of CFSE+ cells. A total of 500,000 events was collected. Results are presented as mean ± SD. ∗, Significantly higher numbers of BMDCs migrated to the lymphoid organs than numbers of BMDCs that migrated in the group in which BMDCs were treated with CpG ODNs and 25(OH)D3 (p < 0.006). Results are representative of two independent experiments.
Calcitriol-responsive BMDCs migrate to nondraining secondary lymphoid organs and activate Ag-specific CD4+ T cells. A and B, BMDCs generated from bone marrow of WT or VDR−/− C57BL/6 mice were treated with 60 ng/ml MPLA in the presence or absence of 10−7 M 25(OH)D3 or 10−8 M calcitriol. After a 24-h incubation at 37°C, the BMDCs were washed, stained with CFSE, and s.c. injected into naive recipients (three mice/group). Forty-eight hours later, mice were sacrificed and single-cell suspensions from individual lymphoid organs (PLNs, ALNs, SPL, and PPs) were prepared and analyzed by FACS analysis for the presence of CFSE+ cells. A total of 500,000 events was collected. Results are presented as mean ± SD. ∗, Significantly higher numbers of BMDCs migrated to lymphoid organs compared with the numbers of BMDCs in the group in which BMDCs were treated with MPLA alone (p < 0.003). Results are representative of two independent experiments. C, C57BL/6 mice received an i.v. injection of CFSE-labeled CD4+ T cells isolated from OT-II donors (4 × 106/mouse). After a 24-h rest, the OT-II CD4+ T cell recipients received a s.c. injection of 2 × 106 cells/mouse of either WT or VDR−/− BMDCs treated as in A and B and additionally pulsed with 500 μg/ml OVA protein. Experimental animals (three mice/group) were sacrificed 4 days postimmunization, and their secondary lymphoid organs were analyzed for CSFE+CD69+ T cells by FACS analysis. Results are presented as mean ± SD. #, Significantly lower numbers of CD4+ T cells expressed CD69 marker compared with the numbers of CD4+ T cells that expressed CD69 in the group injected with WT BMDCs (p < 0.05); ∗, significantly lower numbers of CD4+ T cells expressed CD69 compared with the numbers of CD4+ T cells that expressed CD69 in the group injected with WT BMDCs (p < 0.002).
Only calcitriol-responsive DCs that had migrated to nondraining secondary lymphoid organs could effectively activate Ag-specific CD4+ T cells
We next questioned whether DC responsiveness to calcitriol is required for DC localization to nondraining secondary lymphoid organs. BMDCs were generated from WT and VDR−/− mice and activated with MPLA in the presence of either 25(OH)D3 or calcitriol. Treated BMDCs were then stained with CFSE and injected s.c. into WT naive C57BL/6 recipients. Forty-eight hours later, animals were sacrificed and various secondary lymphoid organs were evaluated for the presence of CFSE+ BMDCs. In contrast to WT BMDCs activated with MPLA and either 25(OH)D3 or calcitriol, which could migrate to nondraining secondary lymphoid organs, BMDCs generated from the bone marrow of VDR−/− mice and activated in a similar manner migrated only to the draining LNs and were not found in nondraining lymphoid organs (Fig. 5⇑, A and B). These data indicate that in addition to an ability to endogenously produce calcitriol from circulating precursors, DC responsiveness to calcitriol stimulatory effects is also necessary to promote DC migration beyond the draining LNs following their exposure to MPLA.
We also investigated whether the migration of calcitriol-responsive DCs is required to efficiently activate Ag-specific T cells residing in draining and nondraining secondary lymphoid organs. To test this experimentally, adoptive recipients of CFSE-stained OVA peptide-specific OT-II CD4+ T cells were injected s.c. with OVA-pulsed WT or VDR−/− BMDCs activated ex vivo with MPLA and either 25(OH)D3 or calcitriol. Adoptive recipients that received an injection of naive or OVA-only-pulsed DCs served as controls. Four days later, the experimental mice were sacrificed, and their secondary lymphoid organs were individually evaluated by FACS for the presence of CD69+ T cells. CD69 represents an early activation marker for CD4+ T cells (16). Results presented in Fig. 5⇑C clearly demonstrate that WT OVA-pulsed BMDCs that were activated with MPLA in the presence of 25(OH)D3 or calcitriol and injected into OT-II CD4+ T cell recipients stimulated an increased CD69 expression on OT-II CD4+ T cells residing in draining (PLNs) and nondraining (axillary LNs (ALNs and PPs) lymphoid organs. Mice injected with OVA-pulsed VDR−/− BMDCs that were activated with MPLA plus 25(OH)D3 or calcitriol ex vivo expressed elevated numbers of CD69-positive OT-II CD4+ T cells only in their draining LNs (Fig. 5⇑C).
To strengthen our observation that responsiveness to calcitriol is needed for promoting the migration of maturing DCs to nondraining lymphoid organs, WT mice and VDR−/− mice were s.c. injected with fluorescent microspheres. The microspheres were injected in the presence or absence of MPLA or calcitriol. As presented in Fig. 6⇓, A and B, microsphere+ DCs could only be detected in the draining LNs of VDR−/− mice injected with microspheres alone, or with added MPLA or calcitriol, whereas WT mice injected with microspheres plus MPLA or calcitriol demonstrated increased numbers of microsphere+ cells in all evaluated draining and nondraining secondary lymphoid organs.
Only calcitriol-responsive DCs are capable of migrating to nondraining secondary lymphoid organs and activating CD4+ T cells in vivo. A and B, Fluorescent microspheres (0.2 μm) were injected into the hind footpads of mature adult WT or VDR−/− mice (three mice/group) in the presence or absence of MPLA (20 μg) or calcitriol (0.1 μg). After 48 h, individual lymphoid organs (PLNs, ALNs, SPLs, and PPs) were analyzed for the presence of microsphere+ cells by FACS. A total of 500,000 events was collected. Data are presented as mean ± SD. #, Significantly higher numbers of microsphere+ cells migrated compared with the numbers of microsphere+ cells migrated in the group injected with microsphere alone (p < 0.02); ∗, significantly higher numbers of microsphere+ cells migrated compared with the numbers of microsphere+ cells migrated in the group injected with microsphere alone (p < 0.001). Results are representative of three independent experiments. C, WT or VDR−/− C57BL/6 mice received an i.v. injection of CFSE-labeled CD4+ T cells isolated from naive OT-II donors (4 × 106/mouse). After a 24-h rest, the OT-II CD4+ T cell recipients were immunized into their hind footpads with 50 μg of OVA in Alum in the presence of 20 μg of MPLA or 0.1 μg of calcitriol. OT-II CD4+ T cell recipients that received footpad injection of 50 μg of OVA in Alum or Alum alone were used as controls. Mice were sacrificed 4 days postimmunization, and their secondary lymphoid organs were individually analyzed for CSFE+CD69+ T cells by FACS. ∗, Significantly lower numbers of CD4+ T cells expressed CD69 compared with the numbers of CD4+ T cells that expressed CD69 in the WT mice that received OT-II CD4+ T cells (p < 0.001).
Finally, we investigated whether the activation of Ag-specific T cells residing in draining and nondraining secondary lymphoid organs by peptide-presenting DCs that had migrated from a nonmucosal vaccination site in response to MPLA was dependent upon responsiveness to calcitriol. CFSE-stained OVA peptide-specific OT-II CD4+ T cells were i.v. injected into groups of WT and VDR−/− mice. One day later, mice were vaccinated with OVA-Alum and MPLA or OVA-Alum and calcitriol. A third group of adoptive recipients of OT-II CD4+ T cells received an injection of OVA-Alum to serve as a negative control. Four days later, all experimental mice were sacrificed, and their secondary lymphoid organs were individually evaluated by FACS for the presence of activated OT-II CD4+ T cells by assessing the expression of CD69 on CFSE+ cells. WT mice immunized with OVA in the presence of MPLA or calcitriol as the adjuvant showed increased levels of CD69 expression on OT-II CD4+ T cells residing in nondraining (ALNs and PPs) secondary lymphoid organs. VDR−/− mice immunized with OVA, regardless of the adjuvant used, expressed only small increases in CD69+ OT-II CD4+ T cells in their draining LNs and no CD69+ OT-II CD4+ T cells in their nondraining peripheral or mucosal lymphoid organs. These findings indicate that DC responsiveness to calcitriol appears to be essential for their migration to nondraining secondary lymphoid organs in vivo subsequent to activation. In addition, mice lacking responsiveness to calcitriol appeared to be generally compromised in the ability of their DCs to activate normal Ag-specific CD4+ T cells in response to vaccination.
Discussion
Recently, we reported that the TLR4 agonist LPS, when used as an adjuvant for nonmucosally delivered vaccines, simultaneously promotes the induction of both systemic and common mucosal immune responses in immunized mice (10). Unfortunately, the proinflammatory properties of LPS preclude its clinical use as an adjuvant in humans. Therefore, MPLA, a minimally toxic derivative of LPS, was evaluated for its potential use as a mucosal adjuvant (5, 6). When added to vaccine formulations, MPLA stimulated the generation of both systemic and mucosal immune responses to s.c. administered Ags. MPLA addition promoted the migration of DCs leaving vaccination sites into multiple secondary lymphoid organs. The migratory DCs that localized in the draining LNs were a heterogeneous mixture of myeloid (CD11c+CD11b+CD205−), dermal interstitial (CD11c+CD11b+CD205intCD8αlow), and LN resident lymphoid (CD11c+CD11b−CD205+CD8α+) DCs. In contrast, the mobilized DCs that migrated to and localized within nondraining secondary lymphoid organs phenotyped as myeloid DCs, indicating that this DC type represents the major migrating cell population that can effectively transport both processed and native Ag from cutaneous sites of vaccination into multiple lymphoid organs throughout the body. These migratory DCs are then able to present Ag peptides and native Ag to Ag-responsive T and B cells residing within these diverse lymphoid compartments (17, 18, 19).
Mobilized DCs that ultimately localize into nondraining lymphoid organs are probably derived from circulating monocytes that had infiltrated into inflamed tissues at sites of vaccination. Within inflammatory tissue sites, some infiltrating monocytes differentiate into macrophages and remain within the inflamed tissue, whereas others acquire Ag, become activated, differentiate into CD11c+CD11b+ DCs, and enter the afferent lymphatic drainage for transport to draining LNs (20, 21). Once localized within the draining LNs, mobilized DCs migrate to T cell-rich areas via a CCR7-dependent process, where they present Ag peptides to responsive CD4+ T cells or transfer their cargo to LN-resident CD8α+ DCs for subsequent presentation to CD8+ T cells (21, 22, 23). Following selected types of stimulant-induced activation, however, many immature myeloid DCs that become mobilized from vaccination sites bypass sequestration in the draining LN, and gain the capacity to localize into any secondary lymphoid organ or the bone marrow (10, 13, 24, 25).
To gain insight into the cellular processes involved, we used a model in which MPLA-activated myeloid DCs from normal or genetically mutant mice were s.c. injected into WT recipients, followed by an assessment of their migratory properties in vivo. BMDCs from 1αOHase-deficient mice, following activation ex vivo with MPLA in the presence of 25(OH)D3, were incapable of migrating beyond the draining LNs following their s.c. injection, suggesting that a local production of calcitriol is required for the migration of DCs beyond the draining LNs.
A common characteristic of the TLR agonists that can stimulate the expression of 1αOHase in DCs and promote their migration beyond draining LNs subsequent to injection (LPS, poly(I:C), and MPLA) is their ability to activate the TRIF- dependent signaling pathway (26). An analysis of the migratory properties of TRIF−/− BMDCs activated with MPLA in the presence of 25(OH)D3 and injected into naive recipients demonstrated a diminished capacity to migrate into nondraining lymphoid organs when compared with WT BMDCs activated in a similar manner. This decrease in DC migration was paralleled by a partial reduction in MPLA-induced 1αOHase (data not shown), indicating that the expression of this enzyme is not absolutely dependent upon TRIF activation.
MPLA activates DCs predominantly through the TRIF-dependent signaling pathway with some involvement of the MyD88-dependent pathway (5, 6). This probably explains the diminished inflammation induced by MPLA when compared with LPS (6). An analysis of the involvement of MyD88-dependent pathway in regulation of the DC migration revealed that BMDCs generated from MyD88−/− mice, matured in vitro with MPLA and 25(OH)D3, behaved like WT DCs treated in a similar manner, suggesting that the MyD88 signaling pathway is not critical for allowing MPLA-activated DC migration beyond draining LNs.
A major downstream effect of TRIF pathway activation is the stimulation of type 1 IFN production (26). The inability of MPLA-activated IFNR−/− BMDCs to migrate beyond the draining LNs suggests that production and responsiveness to type 1 IFNs may be essential for allowing DC migration into nondraining lymphoid organs. Following their activation by MPLA ex vivo, IFNR−/− BMDCs did not up-regulate 1αOHase expression (data not shown).
Because responsiveness to type 1 IFNs appears to represent an important factor for inducing 1αOHase, we additionally tested whether type 1 IFNs themselves could stimulate DCs to express 1αOHase. Treatment of WT BMDCs with IFN-β ex vivo was found to increase expression of 1αOHase. Coactivation with CpG ODNs markedly enhanced enzyme expression levels, whereas activation with CpG ODNs alone was incapable of inducing 1αOHase. Our data suggest, therefore, that stimulation of the MyD88-dependent signaling pathway will optimize 1αOHase expression by DCs exposed to type 1 IFNs and enhance their ability to migrate into nondraining lymphoid organs.
Although our studies have focused on the capacity of MPLA to stimulate 1αOHase expression, and by inference endogenous calcitriol production, DCs themselves do not have to be the actual local producers of calcitriol. Skin keratinocytes, monocytes, and macrophages are known to produce calcitriol from 25(OH)D3 following selected types of stimulation and could potentially provide their hormonal influence to neighboring DCs (27, 28, 29, 30, 31, 32, 33, 34). A more critical question is whether responsiveness to calcitriol is essential for mobilized DCs to traffic to nondraining secondary lymphoid organs.
Following the activation of VDR−/− BMDCs with MPLA in the presence of 25(OH)D3, or calcitriol, the VDR−/− DCs were unable to migrate beyond the draining LNs when injected s.c. into naive hosts. This result directly correlated with the finding that Ag-specific CD4+ T cells residing in nondraining lymphoid organs could not be stimulated when OVA-pulsed VDR−/− DCs were activated with MPLA plus 25(OH)D3 and s.c. injected into WT adoptive recipients of CD4+ T cells from OT-II TCR transgenic mice. These findings indicate that responsiveness to calcitriol is necessary to allow DC localization to nondraining secondary lymphoid organs from cutaneous sites of vaccination.
Collectively, our results demonstrate that locally produced calcitriol, and vitamin D3 responsiveness by DCs, represent essential components for allowing Ag-laden DCs that have become mobilized from sites of vaccination to traffic to nondraining secondary lymphoid organs. We have previously demonstrated that calcitriol-treated DCs up-regulate their surface expression of CCR7 more slowly than untreated DCs in response to their activation through TLR4 (10). This temporal suppression in CCR7 expression was paralleled by an inhibition in chemotaxis to CCR7 ligands without any inhibition in chemotactic responsiveness to sphingosine-1-phosphate. Entry of DCs that become mobilized from sites of vaccine administration into nondraining lymphoid organs is followed by effective Ag presentation, evidenced by both the assessment of Ag-specific T cell activation, and from the successful induction of common mucosal immunity to test Ags administered nonmucosally. Such types of immune responses can only occur under conditions in which Ag-laden DCs are induced to traffic into nondraining lymphoid organs.
Recent data indicate that locally produced calcitriol in skin is necessary for controlling the homing properties of Ag-specific T cells, directing their localization to inflamed cutaneous tissue sites via mechanisms that involve the up-regulated expression of CCR10 (29). Our preliminary data suggest that MPLA-activated DCs that have localized to draining and nondraining peripheral lymphoid organs stimulate responsive T cells to express CCR10, whereas DCs that localize to the PPs are somehow able to stimulate responsive T cells to express both CCR10 and CCR9 (data not shown). It may turn out that tissue-localized end organ metabolism of vitamin D3 plays an underappreciated role in controlling both the migratory properties of Ag-presenting DCs, as well as the migratory properties of effector T and B cells.
The use of particular TLR agonists as adjuvants for nonmucosally delivered vaccines has multiple advantages. Small controlled doses of Ag can be precisely delivered by needle injection into the skin or muscle. Mucosally delivered vaccines, despite their being easy to administer, however, require much higher doses of Ag to effectively stimulate adaptive immunity and cannot be accurately controlled (12, 35). Unfortunately, orally delivered vaccines do not induce good adaptive immune responses unless toxic adjuvants such as cholera toxin are coadministered (36, 37). Although intranasal delivery of vaccines is able to induce both systemic and mucosal immunity, the mucosal immunity induced is limited to the respiratory and urogenital tracts (38).
The use of TLR agonists, like MPLA, as adjuvants for nonmucosally delivered vaccines is superior to the direct use of calcitriol itself. We have previously observed that calcitriol addition to s.c. delivered vaccine formulations stimulates a Th2-biased systemic immune response, whereas MPLA as a vaccine adjuvant promotes common mucosal immunity with Th1-biased serum immune responses. Th1 immunity is considered to be important for protection against many infectious diseases (39, 40, 41).
The ability of adjuvants such as MPLA to induce both systemic and mucosal immunity to administered vaccines may have important implications in the development of protective host immunity. The vitamin D3-dependent ability of myeloid DCs to transport Ags into multiple lymphoid tissues and present these Ags to responsive T and B cells allows the simultaneous induction of immune effector responses normally controlled by unique microenvironment differences that exist within distinct lymphoid organ inductive sites. This would allow the peripheral, central, and mucosal tissues of vaccinated hosts to all benefit immunologically, by expanding the types of immune effector responses being induced.
Acknowledgments
We thank James Whitcomb, University of Notre Dame, for providing bone marrow cells from 1αOHase−/− mice.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 This work was supported by funds from the Department of Pathology and National Institutes of Health Grant AI 059242.
↵2 Address correspondence and reprint requests to Dr. Raymond A. Daynes, Department of Pathology, SOM, University of Utah, 30 North 1900 East, Salt Lake City, UT 84132. E-mail address: ray.daynes{at}path.utah.edu
↵3 Abbreviations used in this paper: MPLA, monophosphoryl lipid A; 1αOHase, 1α-hydroxylase; 25(OH)D3, 25-hydroxycholecalciferol; ALN, axillary lymph node; BMDC, bone marrow-derived dendritic cell; DC, dendritic cell; DT, diphtheria CRM 197 protein; LN, lymph node; PLN, popliteal LN; PP, Peyer’s patch; SPL, spleen; TRIF, Toll-IL-1R domain-containing adapter-inducing IFN-β; VDR−/−, vitamin D3 receptor; WT, wild type.
- Received December 24, 2008.
- Accepted January 23, 2009.
- Copyright © 2009 by The American Association of Immunologists, Inc.