Topical transcutaneous immunization (TCI) presents many clinical advantages, but its underlying mechanism remains unknown. TCI induced Ag-specific IgA Ab-secreting cells expressing CCR9 and CCR10 in the small intestine in a retinoic acid-dependent manner. These intestinal IgA Abs were maintained in Peyer’s patch-null mice but abolished in the Peyer’s patch- and lymph node-null mice. The mesenteric lymph node (MLN) was shown to be the site of IgA isotype class switching after TCI. Unexpectedly, langerin+CD8α− dendritic cells emerged in the MLN after TCI; they did not migrate from the skin but rather differentiated rapidly from bone marrow precursors. Depletion of langerin+ cells impaired intestinal IgA Ab responses after TCI. Taken together, these findings suggest that MLN is indispensable for the induction of intestinal IgA Abs following skin immunization and that cross-talk between the skin and gut immune systems might be mediated by langerin+ dendritic cells in the MLN.
Transcutaneous immunization (TCI)3 is a novel needle-free vaccination method that induces an immune response through the topical application of a vaccine Ag and adjuvant to the intact skin surface (1). When used with cholera toxin (CT) or heat-labile enterotoxin as adjuvant, TCI with heterologous protein induces robust serum IgG and secretory IgA (SIgA) Ab responses against both toxins and coadministered Ag in both the systemic and mucosal immune systems without systemic toxicity in human trials (2, 3). Such findings highlight the importance of this novel strategy for the induction of mucosal IgA Abs using intact skin as well as mucosal surfaces. However, the mechanism by which mucosal immune responses are induced via skin immunization has remained elusive. Our finding that TCI induced intestinal SIgA Abs suggests a possible linkage between the skin and gut immune responses, leading us to focus our study on the elucidation of that linkage.
Materials and Methods
C57BL/6 mice were purchased from the Charles River Laboratories. To generate mice lacking both Peyer’s patches (PP) and lymph nodes (LN), pregnant mice were injected i.v. with 200 μg of lymphotoxin-β receptor (LTβR)-Ig and 200 μg of TNFR55-Ig on gestational days 14 and 17 (4). To generate the PP-null mice, pregnant mice were injected i.v. with 600 μg of anti-IL-7Rα mAb on gestational day 14 (5). Langerin-diphtheria toxin receptor (DTR) and polymeric Ig receptor (pIgR)−/− mice were used (6, 7). Vitamin A-deficient C57BL/6 mice were prepared as previously reported (8).
Mice were immunized transcutaneously as described elsewhere (9). Briefly, an occlusive patch was applied to the shaved dorsum of each mouse for 24 h. The patch, which contained gauze soaked in 100 μg of tetanus toxoid (TT) plus 50 μg of CT (List Biological Laboratories), served as a delivery device for the Ag and adjuvant while also preventing the mice from disturbing the gauze. Once the gauze was removed, the mice were thoroughly washed and dried to prevent any oral contamination due to grooming. To analyze the induction of mucosal immune responses, mice were immunized by TCI three times at 2-wk intervals. TT was provided by Dr. Y. Higashi (Biken Foundation, Osaka University, Osaka, Japan).
Anti-CD205-biotin (clone NLDC-45; BMA Biomedicals) and anti-langerin-Alexa Fluor 488 (clone 929F3; AbCys) were used in accordance with the manufacturers’ instructions. Other Abs were purchased from BD Pharmingen. For staining of the B subunit of CT (CTB)86–103-I-Ab tetramers, CT-I-Ab tetramers were formed by incubation of CT-I-Ab monomers and streptavidin-PE (Molecular Probe) with a molecular ratio of 5:1 for 2 h followed by incubation with cells for 2.5 h at 37°C in the CO2 incubator.
Mononuclear cells (MNC) were dissociated from the lamina propria of the small intestine (SI) and the large intestine (LI) by digestion using a collagenase/DNase I enzyme solution after the removal of PP. Cells were then enriched by a discontinuous density gradient containing 40 and 75% Percoll (Amersham Bioscience).
The number of Ag-specific Ab-secreting cells (ASC) was determined by an ELISPOT assay according to an established protocol (10) with the aid of a stereomicroscope (SZ2-ILST; Olympus). The number of Ag-specific ASCs was standardized by the total MNCs because similar numbers of total IgA and IgG ASCs were observed in naive and TCI-immunized mice.
To evaluate the expression of chemokine receptors on Ag-specific ASCs, a chemotaxis assay and ELISPOT were combined as previously described (11).
Transplant of congenic bone marrow (BM) cells
Eight-week-old recipient CD45.2+ C57BL/6 mice were lethally irradiated with 950 rad by a Gammacell low dose-rate research irradiator (GC 3000 Elan; MDS Nordion) and were transferred i.v. with BM cells (1 × 106) obtained from congenic CD45.1+ C57BL/6 mice as described previously (12).
Data are expressed as the mean ± SD. Statistical comparison between experimental groups was performed using the Student t test.
Results and Discussion
TCI induces intestinal IgA Abs in a retinoic acid-dependent manner
To investigate the interaction between skin and mucosal immunity, we administered TT as Ag and CT as mucosal adjuvant via TCI. TCI elicited TT-specific IgA and IgG Abs in the fecal extracts as well as sera (Fig. 1,A). In addition, other mucosal secretions, including saliva and vaginal and nasal washes, were also found to contain significant levels of TT-specific IgA Abs. As expected from these Ab results, a high number of TT-specific ASCs were detected in the lamina propria of the SI and LI after TCI with TT and CT (Fig. 1,B). In contrast, three parenteral immunizations (s.c. or i.p.) induced no Ag-specific ASCs in the gut (data not shown). To determine whether the TCI-induced IgA ASCs secreted SIgA Abs associated with the secretory component, IgA Ab responses after TCI were analyzed in pIgR−/− mice lacking this IgA secretion pathway (7). These mice showed complete loss of IgA in the fecal extracts and saliva, while levels of IgG Abs remained comparable to those seen in wild-type mice (Fig. 1 C). This result demonstrates that TCI can elicit the formation and secretion of SIgA into mucosal compartments.
To identify the expression of chemokine receptors on the TCI-induced intestinal Ag-specific IgA ASCs, we used a technique that combined chemotaxis assay and ELISPOT. CTB- and TT-specific IgA ASCs migrated to TECK/CCL25, CTACK/CCL27, or MEC/CCL28 (Fig. 1 D), showing that these cells expressed functional CCR9 and CCR10 and confirming previous observations of polyclonal IgA-ASCs (13, 14).
Intestinal dendritic cells (DCs) imprint gut-homing specificity on lymphocytes by retinoic acid (RA) (8); reciprocally, DCs from peripheral LNs imprint skin-homing specificity on T cells by vitamin D3 (15). In B cells, RA acts synergistically with IL-6 or IL-5 to induce IgA secretion (16). Therefore, most skin DCs are expected to induce activated Ag-specific T and B cells to home to the skin after TCI. However, TT-specific IgA ASCs in the gut were dramatically reduced in vitamin A-deficient mice after TCI (Fig. 1 E). This finding implies the existence of a unique subset of DCs after skin immunization that can induce homing of IgA ASCs into the gut in an RA-dependent manner.
Mesenteric lymph node (MLN) is indispensable for the induction of intestinal IgA ASCs after TCI
To evaluate the induction of Ag-specific CD4+ T cells by TCI, we used tetramer staining of the CT peptide-MHC class II complex. Increased levels of CT-specific CD4+ T cells were detected by day 12 after TCI in the spleen, the skin-draining cutaneous lymph node (CLN), and the MLN (Fig. 2,A). In contrast, no CT-specific CD4+ T cells were found in either PP or SI. When TCI was combined with FTY720 to prevent the recirculation of T lymphocytes, Ag-specific CD4+ T cells were also activated in the spleen, CLN, and MLN (Fig. 2,B), showing that CD4+ T cells are primed directly by Ag-bearing DCs in these tissues rather than by passive diffusion through blood circulation. We next sought to determine where IgA class switching occurs after TCI by using FTY720 treatment to confine the ASCs to the organ of isotype class switching (17). FTY720 treatment increased the number of IgA ASCs in the MLN and IgG ASCs in the CLN and reduced IgA ASCs in the SI (Fig. 2,C). To identify the privileged sites for induction of intestinal IgA Abs after TCI, we used mice that lacked PP as well as mice that lacked both LN and PP but retained isolated lymphoid follicles. After TCI, Ag-specific IgA ASCs were maintained in the gut of the PP-null mice (Fig. 2,D). However, the gut of the LN- and PP-null mice was bereft of IgA ASCs after TCI (Fig. 2 E). Taken together, these results strongly suggest that the MLN, rather than PP or CLN, is the privileged site for the induction of intestinal IgA responses after TCI. The MLN seems to hold a central position in immune anatomy, acting as a border between the gut and systemic immune systems.
Emergent langerin+ DCs in the MLN are key to the induction of intestinal IgA Abs after TCI
To investigate the mechanism underlying the induction of intestinal IgA by TCI, we next focused on the DCs. Langerin has been regarded as a specific marker of epidermal Langerhans cells (LCs) (18). Langerin+ DCs make up 30–50% of the total CD11c+ DCs in the CLN (Fig. 3,A). Strikingly, distinct CD205+langerinhigh DCs appeared in the MLN after TCI but not at steady-state conditions (Fig. 3,A). Langerin+ DCs in mice can be classified into tissue-derived LCs and blood-derived langerinlowCD8α+ DCs (6). Interestingly, the novel langerin+ DCs in the MLN did not express CD8α but expressed CD11b (Fig. 3,A), suggesting that langerin+ DCs in the MLN after TCI closely resemble migratory tissue-derived LCs but not blood-derived DCs. To further characterize these emergent langerin+ DCs in the MLN after TCI, we compared the expression patterns of costimulatory molecules and gut-homing integrins in the epidermis, CLN, and MLN (Fig. 3 B). Interestingly, the levels of CD80, CD86, and CD40 expressed on langerin+ DCs in the MLN were comparable to those in the CLN after TCI, but expression of gut homing-related integrins such as CCR9, α4β7, and CD103 (αE), were higher in the langerin+ DCs of the MLN than in those of the CLN.
To investigate the origin of langerin+ DCs in the MLN after TCI, we used LC chimeras, showing that recipient skin LCs become resistant to irradiation (12). At 8 wk after a congenic BM transplant, the presence of langerin+ DCs in the MLN was evaluated after TCI. Langerin+ DCs, most of which were CD45.1+ donor-derived cells, emerged in the MLN but not in the spleen and PP (Fig. 3 C). Regardless of TCI, ∼90% of langerin+ DCs in the CLN were also CD45.1+ donor-derived cells. These results suggest that epidermal LCs can be resistant to irradiation, but most langerin+ DCs in the LN are replenished from BM-derived precursors. Recent articles have identified a novel population of circulating langerin+ DCs in the dermis, skin-draining LN, and lung; this population is donor- derived and does not originate in the epidermal LC (19, 20, 21). Their recruitment into the LN requires CCR7. After TCI, langerin+ DCs of the MLN closely resemble dermal circulating langerin+ DCs. In this regard, TCI-induced intestinal IgA responses were significantly decreased, but serum IgG and IgA Ab levels were maintained in the CCR7−/− mice (data not shown), suggesting that Ag trafficking to the MLN is tightly regulated by CCR7 signaling. Circulating langerin+ DCs and the TCI-induced langerin+ DC population in MLN express CD103 whereas epidermal LCs do not. In addition, we challenged mice by painting tetramethylrhodamine isothiocyanate (TRITC) and CT on the skin but could not find any TRITC+langerin+ DCs in the MLN (data not shown). Therefore, we speculate that circulating langerin+ DCs could rapidly differentiate in or migrate to the MLN in a CCR7-dependent manner after TCI. Inflammation signals, such as CT, could favor differentiation of BM precursors into langerin+ DCs. The precise mechanism underlying this alternative possibility is currently under investigation in our laboratory.
We used langerin-DTR knock-in mice to determine whether intestinal IgA responses could be induced by TCI in the absence of langerin+ DCs of MLN (6). In the langerin-DTR mice, diphtheria toxin (DT) depleted 98–99% of langerin+ DCs in the skin and CLN at day 2 after injection (data not shown). When langerin-DTR mice were immunized by TCI together with DT treatment, the level of TT-specific SIgA Abs in the fecal extracts was significantly decreased in the absence of langerin+ DCs, while the levels of IgG and IgA Abs in the sera remained unchanged (Fig. 3,D). Consistent with this, the number of TT-specific IgA ASCs in the SI was also significantly reduced after in vivo ablation of langerin+ DCs (Fig. 3 E). These results suggest that intestinal IgA Ab responses following TCI cannot occur in the absence of langerin+ DCs whereas IgG Ab responses can, due to the compensation provided by other DCs. Collectively, these findings suggest that langerin+ DCs are indispensable for the induction of intestinal IgA Abs following TCI.
In summary, we have identified a novel mechanism underlying the induction of intestinal IgA Ab responses after TCI; langerin+ DCs in the MLN are indispensable for the RA-dependent induction of intestinal IgA Abs following TCI. Such a connection is plausible, because skin and mucosal surfaces share some immunological similarities, not the least of which is that both are constantly exposed to the outside environment. Taken together, our results suggest a new paradigm for cross-talk between the skin and gut immune systems, one in which langerin+ DCs in the MLN play a key role.
We thank Drs. Makoto Iwata (Tokushima Bunri University, Tokushima, Japan) and Masafumi Yamamoto (Nihon University, Matsudo, Japan) for generous gifts of reagents and helpful discussions.
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.
This work was supported by the governments of the Republic of Korea, Sweden, Japan, and Kuwait, by a Korean Research Foundation Grant funded by the Korean government (KRF-2005-015-E00117), and by grants from the Ministry of Education, Science, Sports, and Culture and Ministry of Health and Labor in Japan.
Abbreviations used in this paper: TCI, transcutaneous immunization; ASC, Ab-secreting cell; BM, bone marrow; CLN, cutaneous lymph node; CT, cholera toxin; CTB, B subunit of CT; DC, dendritic cell; DT, diphtheria toxin; DTR, diphtheria toxin receptor; LC, Langerhans cell; LI, large intestine; LN, lymph node; LTβR, lymphotoxin-β receptor; MLN, mesenteric lymph node; MNC, mononuclear cell; pIgR, polymeric Ig receptor; PP, Peyer’s patch; RA, retinoic acid; SI, small intestine; SIgA, secretory IgA; SP, spleen; TT, tetanus toxoid.