Our previous studies demonstrated the potential of the sublingual (s.l.) route for delivering vaccines capable of inducing mucosal as well as systemic immune responses. Those findings prompted us to attempt to identify possible inductive mechanism of s.l. vaccination for immune responses. Within 2 h after s.l. administration with cholera toxin (CT), significantly higher numbers of MHC class II+ cells accumulated in the s.l. mucosa. Of note, there were brisk expression levels of both CCL19 and CCL21 in cervical lymph nodes (CLN) 24 h after s.l. vaccination with CT. In reconstitution experiments using OVA-specific CD4+ or CD8+ T cells, s.l. vaccination elicited strong Ag-specific T cell proliferation mainly in CLN. Interestingly, Ag-specific T cell proliferation completely disappeared in CD11c-depleted and CCR7−/− mice but not in Langerin-depleted, macrophage-depleted, and CCR6−/− mice. Similar to CD4+ T cell responses, induction of Ag-specific IgG (systemic) and IgA (mucosal) Ab responses were significantly reduced in CD11c-depleted and CCR7−/− mice after s.l. vaccination with OVA plus CT. Although CD8α− dendritic cells ferried Ag from the s.l. mucosa, both migratory CD8α− and resident CD8α+ dendritic cells were essential to prime CD4+ T cells in the CLN. On the basis of these findings, we believe that CCR7 expressed CD8α−CD11c+ cells ferry Ag in the s.l. mucosa, migrate into the CLN, and share the Ag with resident CD8α+CD11c+ cells for the initiation of Ag-specific T and B cell responses following s.l. challenge. We propose that the s.l. mucosa is one of the effective mucosal inductive sites regulated by the CCR7-CCL19/CCL21 pathway.
The mucosal immune system is a first line of defense against foreign Ags, including microbial and dietary Ags (1, 2). Many recent studies have focused on developing mucosal vaccines capable of effectively inducing both mucosal and systemic immune responses, thereby resulting in two layers of host protection (3, 4). Since the route of vaccine administration has a significant effect on the outcome of immune responses, numerous studies have attempted to develop novel mucosal vaccine delivery routes. The sublingual (s.l.)3 route has been used for many years to deliver low-molecular-weight drugs, including small immunogenic peptides (5, 6). The s.l. route is also effective and efficient for allergen-specific immunotherapy (7, 8). Recently, we demonstrated that s.l. administration of a prototype soluble protein together with cholera toxin (CT) as mucosal adjuvant induce a broad range of immune responses in mucosal and extramucosal tissues, including secretory and systemic Ab responses as well as mucosal and systemic cytotoxic T cell responses (9). In addition, s.l. vaccination with inactivated or live influenza virus resulted in effective protection against lethal virus infection (10).
Buccal mucosa, an oral mucosa, has received attention for induction of T cell responses, especially class I-restricted CD8+ effector cells in mice (11). CD8+ T cell cross-priming was induced by painting the buccal mucosa with the contact sensitizer 2,4-dinitro-1-fluorobenzene, leading to rapid recruitment of dendritic cells (DCs) into epithelium which is completely dependent on MIP-3α/CCL20 interaction (12). Moreover, the discovery of several related epithelial-expressed chemokines in the mucosal compartments (TECK/CCL25 in small intestine, CTACK/CCL27 skin, and MEC/CCL28 in diverse mucosal sites) highlights an important role for epithelial cells in controlling homeostatic lymphocyte trafficking, including localization of T cells and of IgA plasma cells (13).
In this study, we primarily focused on the mechanism and outcome of s.l. vaccination with a soluble protein Ag plus mucosal adjuvant (i.e., CT). Our findings demonstrate that the CCR7-CCL19/CCL21 pathway on CD11c+ DCs is responsible for efficient Ag-specific systemic and mucosal immune responses (including T and B cells) by the s.l. route.
Materials and Methods
Female BALB/c and C57BL/6 mice aged 5–6 wk were purchased from Charles River Laboratories Orient. Dr. K. Yamamoto (University of Tokyo, Tokyo, Japan) provided the OVA-TCR transgenic mice (DO11.10) on a BALB/c background, which express a TCR specific for OVA epitope (323–339), and Dr. M. Lipp (Max Delbruck Center for Molecular Medicine, Berlin, Germany) generously made available the CCR7−/− mice on a C57BL/6 background. Langerin-DTR mice were given by Dr. B. Malissen (Centre d’Immunologie de Marseille-Luminy, INSERM-CNRS-Université de la Mediterranée, Parc Scientifique et Technologique de Luminy, France). The OVA-specific TCR transgenic OT-II, OT-I, and CD11c-DTR mice of C57BL/6 background were purchased from The Jackson Laboratory. To deplete each target cell, 100 ng of diphtheria toxin (DT; Sigma-Aldrch) was injected by the i.p. route every 3 days into the CD11c-DTR or Langerin-DTR mice. To deplete macrophages, 100 μl of Cl2MBP-liposome was prepared and administered by both i.v. and i.p. routes as described previously (14, 15). Mice were maintained in the animal facility of the International Vaccine Institute under specific pathogen-free conditions and received sterilized food (certified diet MF; Charles River Laboratories Orient) and water ad libitum.
Mice were anesthetized by i.p. injection of ketamine (100 mg/kg body weight (BW); Yuhan) and xylazine hydrochloride (10 mg/kg BW; Bayer). For s.l. immunization, forceps were placed under the tongue of an anesthetized mouse, its mouth was stretched open, and Ag was administered by micropipette. The total volume of Ag plus adjuvant was kept to <10 μl to avoid swallowing effects. Mice were immunized via the s.l. route on days 0, 14, and 21 with a mixture of 100 μg of OVA (Sigma-Aldrich) and 2 μg of CT (List Biological Laboratories) and assessed at 5–7 days after the last immunization. In some experiments, mice received via s.l. challenge A/PR/8 influenza virus (10× LD50).
In vivo Ag uptake assay
To trace Ag uptake pattern by the s.l. route, mice were immunized via s.l. route with 40 μg of FITC-conjugated OVA (FITC-OVA; Molecular Probes) alone or FITC-OVA plus 2 μg of CT (List Biological Laboratories). The s.l. mucosa was harvested at 1, 2, and 4 h following s.l. immunization. To observe nuclei, the tissue was stained with 4′,6-diamidino-2-phenylindole (DAPI) (Molecular Probes) and viewed using confocal microscopy (Carl Zeiss). Each image was analyzed with Photoshop (Adobe Systems).
The s.l. tissues were fixed in 4% paraformaldehyde at 4°C for 1 h and dehydrated in sucrose solutions (10, 20, and 30%) and embedded in OCT compound (Sakura Finetec) as described previously (16). Cryostat sections (5 μm) were fixed in ice-cold acetone and blocked with FcRII/III mAb (clone 2.4G2; BD Pharmingen) in PBS. The tissues were stained with FITC-conjugated anti-CD11c mAb (clone HL3), FITC-conjugated anti-I-Ad
Cell preparation from s.l. mucosa sheet
To investigate cell populations of an epithelial sheet of s.l. mucosa, tissues including mouse tongue were properly trimmed by discarding unnecessary tissues, gingiva, tooth, and muscles. Then, the remaining tissues, including part of the tongue, were removed with sharp scissors. Tissues were incubated 40 min with 0.5 U/ml dispase II (50 U/ml; Sigma-Aldrich) in PBS at 37°C, gently washed with PBS, and peeled off the epithelial sheets. Chopped tissues with 0.5 mg/ml collagenase D (5 mg/ml; Roche) plus 1 μg/ml DNase I (0.1 mg/ml; Sigma-Aldrich) enzyme solution in RPMI 1640 medium containing 10% FBS were vigorously stirred at 37°C for 1 h. Cells were recovered by centrifugation and filtered using a 100-μm cell strainer (BD Falcon).
Cell preparation from draining LN
DCs were isolated from CLN according to previously established methods (17). Briefly, CLN were digested with 0.5 mg/ml collagenase D (5 mg/ml; Roche) plus 1 μg/ml DNase I (0.1 mg/ml; Sigma-Aldrich) enzyme solution in RPMI 1640 medium containing 10% FBS with continuous stirring at 37°C for 30 min. After EDTA was added at a final concentration of 10 mM, tissues were incubated for an additional 5 min at 37°C. Cells were filtered using a 100-μm cell strainer (BD Falcon) and recovered by centrifugation.
Flow cytometric analysis
Phenotypic analyses of APC were performed using the following Abs (all from BD Pharmingen unless noted): anti-CD45 (clone 30-F11), anti-CD11c (clone HL3), anti-CD8α (clone 53-6.7), anti-I-Ad
ELISA and ELISPOT assay
Ag-specific Ab titers were determined by ELISA as described elsewhere (18). Endpoint titers were expressed as the reciprocal log2 of the last dilution giving an OD at 450 nm of 0.1 greater than background. Mononuclear cells were obtained from the spleen, CLN, submandibular gland, and small intestine as described previously (19). Numbers of Ab-secreting cells were determined using ELISPOT assay in accord with an established protocol (20).
cDNA synthesis and real-time quantitative PCR
Analysis of in vivo Ag-specific T cell proliferation
Mononuclear cells were isolated from the spleen and LN of DO11.10 mice on a BALB/c background or OT-I mice on a C57BL/6 background. CD4+KJ1.26+ T cells or CD8+Vα2+ T cells were labeled with 10 μM CFSE (Molecular Probes). CD4+KJ1.26+ T cells (1 × 107) or CD8+Vα2+ T cells (1 × 107) were adoptively transferred into the sex-matched naive BALB/c mice or C57BL/6 mice via the i.v. route. One day after adoptive transfer, the recipient mice were vaccinated via the s.l. route with a mixture of 500 μg of OVA and 2 μg of CT. The magnitude of the CFSE dilution in the mononuclear cells of each tissue was analyzed according to the manufacturer’s instructions.
Analysis of in vitro Ag-specific T cell proliferation
Naive BALB/c mice were s.l. challenged with a lethal dose of A/PR/8 influenza virus (10× LD50) and CD11c+CD8α− or CD11c+CD8α+ DCs were isolated from CLN at indicated day after challenge. Hemagglutinin (HA)-specific CD4+ T cells were isolated from all LNs of HA-TCR transgenic mice on a BALB/c background. HA-specific CD4+ T cells (5 × 105) were cocultured for 3 days with each DC subsets at 37°C in a 5% CO2 incubator. The magnitude of the CFSE dilution in the CD4+ T cells was analyzed according to the manufacturer’s instructions. To determine OVA-specific T cell proliferation, we adopted the H1ova virus, which inserted the coding sequences for OVA peptide OVA323–339 into the HA molecules of H1N1 (provided by Dr. P. C. Doherty, St. Jude Children’s Research Hospital, Memphis, TN) and OT-II CD4 T cell as stimulator Ag or responder T cell.
FTY 720 treatment
To induce lymphocyte retention in secondary lymphoid organs, mice received FTY 720 (1 mg/kg BW; Cayman Chemicals) by i.p. injection 1 day after an adoptive transfer (21). The effect of FTY 720 treatment was monitored by regular analysis of peripheral blood lymphocytes and by counting tissue lymphocytes.
Neutralization of CCL19 and CCL21
In vivo staining of s.l. mucosa with CFSE
CFSE was dissolved at 25 mM in DMSO and subsequently was diluted to 8 mM in PBS. CFSE (10 μl) was administered to mice via the s.l. route after anesthesia.
Data and statistical analysis
Ab titers, defined as the reciprocal of the sample dilution giving an absorbance value 0.1 higher than the PBS-vaccinated control group, were expressed as geometric mean ± SD and compared using the t test (Sigmaplot program). Each experiment was repeated at least three times using four mice per group.
S.l. vaccination with CT induces an accumulation of MHC class II+ cells in the s.l. mucosa
To see histological traits of s.l. mucosa, tissues were stained with H&E (Fig. 1⇓Aa). As common features of skin, s.l. surfaces are covered with stratified squamous epithelia. Beneath the epithelial layer, the lamina propria (LP) contains mononuclear cells devoid of organized lymphoid structures. In the immunohistochemical study, MHC class II+ and/or CD11b+ cells were detected in both epithelium and LP of mice at steady state (Fig. 1⇓Ab); the numbers were much increased within 2 h after CT administration via s.l. mucosa (Fig. 1⇓A, c and d). We also found CD11c+ and Langerin+ cells although the expression intensities were moderate (data not shown). In contrast, the NKM 16-02-4 mAb (22) positive cells (i.e., M cells) were not detected in s.l. mucosa at steady state or after s.l. CT administration (data not shown). To trace the Ag uptake pattern in s.l. mucosa, we next administered mice with FITC-OVA alone or FITC-OVA plus CT by the s.l. route in a time-dependent manner. The FITC-conjugated OVA adhered to the s.l. mucosa until 2 h after s.l. administration with/without CT; most FITC-OVA had disappeared by 4 h after administration (Fig. 1⇓B). Taken together, the histological traits of the s.l. mucosa resemble other mucosal tissues, such as the skin, buccal, and vaginal mucosa, which possess stratified epithelium and lack mucosa-associated lymphoid tissues, and show ability to penetrate Ag in the absence and presence of mucosal adjuvant.
Phenotype characteristics of s.l. mucosa
We next investigated phenotypic features of APC in the s.l. mucosa by FACS analysis. Under the steady-state condition, there are two major APC: F4/80+MHCII+ (12.53%) and CD11c+MHC+ (6.24%) cells, and a minor population of Langerin+MHC+ (1.90%) cells (Fig. 2⇓A). In our study, the majority of the DCs were CD11b+CD8α−mPDCA−CD4− myeloid DCs in s.l. mucosa (Fig. 2⇓A). To assess maturation patterns of F4/80+ and CD11c+ cells pre- and post-CT administration via the s.l. route, expression levels of costimulatory molecules (CD40, CD80, and CD86) and chemokine receptors (CCR6, CCR7, and CCR10) were assessed (Fig. 2⇓B). The levels of CD40, CD80, and CD86 on both F4/80+ and CD11c+ cells isolated from s.l. epithelium were highly expressed at steady state. Unexpectedly, we found lower expression levels of CD40, CD80, and CD86 on both F4/80+ and CD11c+ cells after s.l. administration than at steady state (Fig. 2⇓B). Both CCR6 and CCR7 were expressed on both F4/80+ and CD11c+ cells from s.l. epithelium in the steady state and were slightly decreased following s.l. administration with CT (Fig. 2⇓B).
Expression of various chemokines in the s.l. mucosa and CLN
We next assessed mRNA expression levels of chemokines and chemokine receptors in the s.l. mucosa and CLN in a time-dependent manner (Fig. 3⇓A). We found that CCL19 (ligand of CCR7), CCL27, and CCL28 (ligands of CCR10) in the s.l. mucosa were substantially expressed at steady state. Of note, mRNA expression of CCL19 was increased at 0.5 and 2 h but drastically reduced by 6 h after s.l. administration with CT (Fig. 3⇓A). Furthermore, mRNA expression levels of CCL27 and CCL28 were decreased after s.l. administration with CT. Overall, s.l. mucosa possessed mature professional APC and had unique chemokine expression patterns under both steady-state and CT challenge. Of note, mRNA expression levels of CCL21 in CLN were ∼30 times higher at 24 h following s.l. vaccination with CT than at steady state and other time points (Fig. 3⇓B). In addition, mRNA expression levels of CCL19, CCL20, and CCL28 were much higher 24 h after s.l. vaccination with CT than at 0, 2, 48, and 72 h. No significant changes were shown in CCR6, CCR7, and CCR10 (data not shown). Predominant enhanced levels of CCL21 in CLN following s.l. administration were also confirmed in protein levels by immunohistochemical analysis (Fig. 3⇓C). To identify which cell population could produce the chemokines, we analyzed the CCL19 and CCL21 expression by DC or macrophages isolated from CLN at 24 h after s.l. vaccination with CT. Neither DCs nor macrophages have the ability to express these chemokines (data not shown). Thus, it seems likely that epithelial or stromal cells but not DCs or macrophages might be involved in regulation of CCL19 and CCL21 after s.l. vaccination. We think that both CCL19 and CCL21 may play substantial roles in immune responses in the s.l. mucosa.
Ag-specific T cell proliferation occurs in the CLN after s.l. vaccination
To identify the place of primary Ag presentation after s.l. vaccination, we undertook an adoptive transfer assay with OVA-specific CD4+ T cells isolated from BALB/c background DO11.10 or CD8+ T cells on C57BL/6 background OT-I mice. Each T cell was labeled with the intracellular amine dye CFSE and transferred to recipient wild-type BALB/c or C57BL/6 mice. At 24 h after adoptive transfer, recipient mice were administered OVA plus CT via the s.l. mucosa, and T cell proliferation was shown in a time-dependent mode. OVA-specific CD4+ T cells proliferated mainly in the CLN, mediastinal LN, and spleen (Fig. 4⇓A). To further confirm direct draining LN for priming OVA-specific CD4+ T cells, we treated mice with the FTY 720 molecule, which prevents the egress of T cells from secondary lymphoid organs (23). At day 6 after s.l. vaccination in mice treated with FTY 720, proliferating OVA-specific CD4+ T cells were found only in the CLN but not in mediastinal or mesenteric LN or spleen (Fig. 4⇓A). These data demonstrate that the CLN is the primary draining LN for priming of CD4+ T cells after s.l. vaccination. Thus, Ag ferrying into draining LN for OVA-specific CD4+ T cell proliferation by s.l. vaccination could not be fully addressed by blood circulation. Similarly, CD8+ T cell priming was evaluated in C57BL/6 mice reconstituted with OVA-specific CD8+ OT-I T cells (Fig. 4⇓B). Highly proliferated CD8+ T cells were determined in the CLN but not in other LN at 3 and 6 days following s.l. administration of OVA with CT. These data clearly demonstrate that s.l. administration of protein Ag with mucosal adjuvant enhances Ag-specific CD4+ and CD8+ T cell activation that initially took place in the CLN.
CCR7 on CD11c+ cells are indispensable for priming OVA-specific CD4+ T and B cells following s.l. vaccination
To identify crucial APC involved in ferrying and presentation of Ag in draining LN after s.l. vaccination, we used several selective APC depletion systems (i.e., Langerin-DTR, CD11c-DTR, and depletion of macrophages). OVA-specific CD4+ T cells isolated from OT-II mice of C57BL/6 background were labeled with CFSE and transferred to recipient Langerin-depleted (Langerin-DTR) (24) or CD11c-depleted (CD11c-DTR) (25) or macrophage-depleted (treated by clodronate) (14) mice. Of note, CD11c+ DC-depleted mice completely lacked OVA-specific CD4+ T cell proliferation in both the CLN and mediastinal LN, whereas both Langerin-depleted and macrophage-depleted mice had similar levels of CD4+ T cell proliferation and differed from those of recipient wild-type C57BL/6 mice (Fig. 5⇓A). However, the percentage of divided cell populations at 6∼8 cycles was 15.7 or 21.6% in langerin-DTR or clodronate-treated mice, respectively, in contrast to 70.7% in wild-type mice (Fig. 5⇓, inset table). These findings further confirmed our findings of fewer CD11c DCs after treatment with DT in the Langerin-DTR or clodronate-treated mice. CD11c+ DCs were also decreased by treatment with DT (14.39-10.18%) or clodronate (19.36-16.60%) (supplemental Fig. 1).4 This analysis revealed that the majority of T cell priming in the CLN was mediated by the CD11c+ DCs, even though the T cell priming of Langerin-DTR and clodronate-treated mice was less than in wild-type mice.
We next evaluated the role of chemokine receptors in s.l vaccination using CCR6−/− or CCR7−/− mice and chemokine-blocking Abs. Of interest, OVA-specific CD4+ T cell proliferation was completely abrogated in CCR7−/− mice but not in CCR6−/− mice. In addition, recipient mice treated with both CCL19- and CCL21-neutralizing mAbs synergically diminished activation of Ag-specific CD4+ T cells in both CLN and mediastinal LN (Fig. 5⇑B). In contrast, identical levels of CD4+ T cell proliferation were detected following treatment with neutralizing CCL28 mAb (Fig. 5⇑B), albeit highly expressed in s.l. epithelium at steady state (Fig. 3⇑A). These data demonstrate that OVA-specific CD4+ T cell proliferation induced by s.l. vaccination must be tightly regulated by CD11c+ DCs in a CCR7-CCL19/21-dependent manner.
Production of Ag-specific Abs depends on CD11c- and CCR7-expressing cells
To further confirm the roles of CD11c+ DCs and CCR7 interaction on humoral immunity, we assessed OVA-specific Ab titers in both systemic (i.e., serum) and mucosal compartments (i.e., fecal extract, nasal wash, saliva, and vaginal wash) following s.l. vaccination with OVA plus CT as adjuvant. As expected, there were significantly lower levels of OVA-specific IgG in serum and IgA in serum and mucosal compartments in CD11c-depleted and CCR7−/− mice (Fig. 6⇓). Furthermore, there were significantly fewer OVA-specific Ab-secreting cells in both systemic (i.e., IgG in spleen and CLN) and mucosal compartments (i.e., IgA in submandibular gland and LP of small intestine) (Fig. 6⇓). In addition, s.l. vaccination with OVA plus CT elicited high levels of anti-CT subunit B (CTB) Ab in serum and mucosal compartments (data not shown). Taken together, these data demonstrate that humoral immune responses induced by s.l. vaccination are tightly regulated by CD11c+ DCs in a CCR7-dependent manner.
Identification of DC subsets for Ag trafficking and presenting after s.l. vaccination
We have shown that DCs might be essential for acquiring Ag in the s.l. mucosa and required in draining LN (i.e., CLN) to elicit adaptive immunity following s.l. administration. To determine the s.l. mucosa-specific DC subsets involved in Ag ferrying into CLN, mice were painted with CFSE solution s.l. 4 h before infection with A/PR/8 influenza virus (10× LD50). Then, CFSE-labeled DC subsets were assessed in CLN in a time-dependent manner. We found significantly higher numbers of CFSE-labeled CD11c+CD8α− cells starting at 12 h after infection. The levels were sustained for 2 days and disappeared gradually (Fig. 7⇓A). Of note, no detectable CFSE-labeled CD11c+CD8α+ cells were found in CLN at all time points (Fig. 7⇓A). We further assessed which DC subsets are essential for Ag presentation to CD4+ T cells after s.l. challenge with A/PR/8 influenza virus. Both CD8α− and CD8α+ DCs isolated from mouse CLN could stimulate HA-specific T cells at 2 and 4 days post-s.l. administration with A/PR/8 influenza virus (Fig. 7⇓B). In parallel, when we adopted the H1ova viruses by inserting the coding sequences for an OVA peptide (OVA323–339) into the HA molecules of H1N1 influenza A viruses, both CD8α− and CD8α+ DCs could induce OT-II CD4 T cell proliferation. Overall, several lines of evidence lead us to speculate that CD8α− DCs sample Ag from s.l. mucosa and ferry it into the draining CLN where both migrant CD8α− and resident CD8α+ DCs play an important role in priming CD4 T cell proliferation.
We have previously reported that the s.l. route is safer than the nasal mucosa for delivery of live or inactivated influenza viruses with mucosal adjuvant. We have also shown that s.l. administration of various immunogenic preparations can elicit broad-based systemic and mucosal immune responses and confer protection against a respiratory viral challenge (9, 10). Data presented here offer the first evidence that both of migratory CD8α− and resident CD8α+ DCs in the s.l. mucosa after challenge of mucosal adjuvant (i.e., CT) or influenza virus are responsible for the efficient priming on Ag-specific T and B cells in the CLN. Moreover, interaction of CCR7-CCL19/CCL21 plays a crucial role in this avenue.
Upon stimulation of mucosal surfaces by foreign Ags, DCs among the APC are immediately processed and transport Ag via afferent lymph to draining LN for presentation of Ag to T cells. Subsequently they induce secretory IgA Abs (26). This suggestive link between traffic pattern of DCs and their functions led to the investigation of the chemokine responsiveness of DCs during their development and maturation (27). Our study and others (12, 26, 28) have shown that type II mucosa tissues (e.g., skin, vaginal, and buccal), which are covered with stratified epithelia and devoid of mucosa-associated lymphoid tissues, mainly express CCL20, CCL19/CCL21, or CCL27/CCL28 individually. For instance, Gr1+ blood monocytes are rapidly recruited under the inflammatory condition into buccal mucosa and skin via a CCR6/CCL20-dependent mechanism, which is essential for CD8+ T cell cross-priming (12). Our study indicates that CCR6/CCL20 was not involved in priming of CD4+ T cells after s.l. immunization. However, another study demonstrated that CCR7-CCL19/CCL21 is a key regulator required for the entry of skin DCs into dermal lymphatics under both inflammatory and steady-state conditions (29). The spontaneous mutant mouse strain plt/plt lacks the CCL21-Ser gene and disrupts trafficking of DCs and T cells to lymphoid tissues (30). That study also found that stromal cells of the LN T cell zone play a central role in recruiting naive DCs and T cells for the initiation of immune responses (30). That treatment of neutralizing Ab with both CCL19 and CCL21, but not with CCL28 during s.l. vaccination, impaired CD4+ T cell priming in draining LN (Fig. 5⇑) suggests that DCs emigration to draining LN after s.l. immunization depends on the CCL19/CCL21 pathway. Furthermore, CCR7−/− mice but not CCR6−/− mice are unable to elicit CD4+ T cell activation and subsequent Ag-specific IgG and IgA Ab responses. Thus, we conclude that the CCR7-CCL19/CCL21 pathway plays a crucial role for DCs acquisition and presentation for induction of Ag-specific T cell proliferation and sIgA secretion following s.l. vaccination.
In general, mucosally administered Ags are taken up by microfold (M) cells, remarkable cells found in the follicle-associated epithelium (FAE) of the GALT and NALT (31, 32). We previously reported that not only FAE but also intestinal villi possess M cells capable of sampling Ags (i.e., protein and/or bacteria) and inducing sequential immune responses in an FAE-independent manner (33). In this study, we were unable to identify M cells in the epithelium of the s.l. mucosa in either the steady state or after CT or influenza virus challenge (data not shown). In addition, DC populations in the LP of the small intestine send dendrites through the epithelial layer without breaking its continuity and express molecules of tight junctions such as ZO-1, occludin, and claudin-1 (34). Another recent study found large numbers of CD11c+ DCs had infiltrated into the epithelium and LP of s.l. mucosa 1 day following s.l. vaccination with conjugate of OVA-CTB subunit (35). In this study, several lines of evidence suggest that DCs are crucial in the induction of immune response after s.l. immunization. The s.l. mucosa is enriched with CD11c+ cells in the epithelium, and the LP and topical application of a mucosal adjuvant induce the recruitment of CD11c+ cells in the mucosa. In addition, the depletion of CD11c+ cells abrogates priming of CD4+ T cells and dramatically reduces systemic and mucosal Ab responses (Figs. 5⇑A and 6⇑). Furthermore, numerous s.l. mucosa-derived CD8α−CD11c+ DCs were found in draining LN at 12 h post-s.l. challenge with influenza virus (Fig. 7⇑A). These results imply that CD8α−CD11c+ DCs newly recruited into the s.l. epithelium by stimulator (i.e., adjuvant or virus infection) are critically involved in sampling and ferrying Ag.
Of interest, whereas it appears that only CD8α− DCs migrate from the s.l. mucosa after viral infection to the draining LN, we found that both CD8α− and CD8α+ DCs were able to present Ag ex vivo to Ag-specific CD4+ T cells (Fig. 7⇑). A previous study showed that skin CD8α− DCs act as simple ferries for Ags and that resident CD8α+ DCs play a role in priming the CD8+ T cells in LN following cutaneous infection with HSV (36). Such a transfer was attributed to cross-presentation, an ability unique to CD8α+ DCs (36). In contrast, however, Allenspach et al. (37) demonstrated a cooperating effect of migratory and lymphoid-resident DCs. They reported that resident DCs trap the circulating Ag-specific T cells until Ag-bearing DCs arrive and migratory DCs reinforce initial T cell activation. More recently it has been shown that both CD8α+ and CD8α− DCs, but not plasmacytoid DCs, present viral Ags to CD4 and CD8 T cells in draining LN after mucosal and cutaneous HSV infection (38). Taken together, these findings suggest that CD8α− DCs derived from s.l. mucosa may have a role in acquiring specialized Ag and in sharing it with the resident CD8α+ DCs and finally that both CD8α− and CD8α+ DCs act as predominant APC to induce CD4 T cell proliferation in the CLN after s.l. challenge of influenza virus. Our group is now investigating the underlying mechanism that enables resident CD8α+ DCs in LN to obtain Ag by virus infection via mucosal tissues.
Langerhans cells (LC) can capture intraepithelial pathogens and migrate to regional LN to present them to naive T cells (39). S.l. mucosa is an enriched DCs site that shares several phenotypic characteristics with epidermal LC and may be a potent tissue for sampling Ags and initiating an immune response (9). In this regard, it has been proposed that allergen is captured within the oral mucosa by LC and transported to local LN, which may favor the induction of mucosal tolerance in humans (40). However, in our present study, Langerin-depleted mice had normal levels of CD4+ T cell activation (Fig. 5⇑) and both systemic IgG and mucosal IgA induction (data not shown). Thus, we suspect that Ag uptake in the epithelium of the s.l. mucosa may occur completely independently of Langerin+ DCs.
Another recent study found that the CCL27 is an indispensable component of T cell-mediated cutaneous immunity. After release from the epidermis it rapidly accumulates in the skin-draining LN and subsequently is involved in recruiting CCR10-expressing T cells after a single treatment of mouse skin with the contact sensitizer 2,4-dinitro-1-fluorobenzene (28). Of note, much higher levels of CCL27 protein expression were detected in lymphatic fluid of healthy volunteers than that of patients with lymphedema (28). In the present study, we found highly enhanced CCL27 and CCL28 mRNA expression at steady state but significantly less expression beginning 30 min after s.l. vaccination with CT (Fig. 3⇑A). Interestingly, at 24 h post-s.l. vaccination with CT, we detected high CCL28 mRNA levels in CLN (Fig. 3⇑B). It seems likely that accumulated CCL28, which derives from s.l. mucosa by CT challenge, can recruit CCR10+ T cells into CLN that might be essential for mucosal IgA secretion. This issue needs to be elucidated in a future study.
In murine studies, s.l. administration of OVA conjugated with CTB efficiently induced T cell tolerance through generation of Foxp3+CD25+CD4+ regulatory T cells that developed apoptosis and depletion of effector T cells (35, 41). It seems likely that different immune modulators (e.g., CTB vs CT) modify APC in the s.l. mucosa and set the T cells in the draining LN for tolerance rather than immunity.
The present study identified an indispensable role of DC regulated by CCR7 and its ligands, CCL19 and CCL21, for induction of efficient Ag-specific systemic IgG and mucosal IgA responses by s.l. challenge when proper mucosal adjuvants are used. We propose that selection of vaccine Ag dose, Ag type, and adjuvants with the capacity to give rise to CCR7 and CCL19/CCL21 production in the s.l. mucosa and the draining LN should be given priority in the design of s.l. vaccines.
The authors have no financial conflict of interest.
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 was supported by the governments of the Republic of Korea, Sweden, and Kuwait and a Korea Science and the Korean Ministry of Science and Technology, and Regional Technology Innovation Program of the Ministry of Commerce, Industry, and Energy (RT104-01-01).
↵2 Address correspondence and reprint requests to Dr. Mi-Na Kweon, Mucosal Immunology Section, International Vaccine Institute, Seoul National University Research Park, Kwanak-Gu, Seoul, 151-818 Korea. E-mail address:
↵3 Abbreviations used in this paper: s.l., sublingual; BW, body weight; CLN, cervical lymph node; CT, cholera toxin; CTB, CT subunit B; DAPI, 4′,6-diamidino-2-phenylindole; DC, dendritic cell; DT, diphtheria toxin; FAE, follicle-associated epithelium; HA, hemagglutinin; LC, Langerhans cell; LN, lymph node; LP, lamina propria.
↵4 The online version of this article contains supplemental material.
- Received October 24, 2008.
- Accepted March 23, 2009.
- Copyright © 2009 by The American Association of Immunologists, Inc.