Epicutaneous allergen administration without antigen delivery device induces local T cell response and alleviates food allergic enteropathy
Epicutaneous immunotherapy (EPIT) in which immune responses are modulated by epicutaneous (EC) antigen administration has been reported 1e5 and anticipated as a novel allergen immunotherapy (AIT). 6,7 EC immunization has also been shown to inhibit immune response in animal models of contact sensitivity, 8 multiple sclerosis, 9 rheumatoid arthritis, 10 hapten-induced colitis 11 and in multiple sclerosis patients. 12 However, most reports of EPIT employ a particular device to deliver antigen on
... eliver antigen on skin without scratching or tape stripping, or apply allergen with TLR ligands. 13 In order to investigate its mechanism and efficacy more simply, we aimed to establish a new EPIT model without any antigen delivery device in a food allergy model. To examine whether EC OVA administration to intact skin without any device can deliver OVA to skin draining lymph nodes (DLNs) of sensitized mice, Alexa-fluor 647 conjugated OVA in PBS was applied to dorsal skin of both ears following pre-wetting with PBS. We examined the frequency of Alexa 647 þ cells in CD11c cells of three groups of mice; the OVA-Alexa group, a group treated with unconjugated Alexa N-hydroxysuccinimide ester and one treated with PBS (Supplementary Fig. 1 ; for more information concerning methods of this study, see the Supplementary Methods section). There was a significant increase in Alexa þ cells when OVA-Alexa was applied, while no significant increase was observed in the case of unconjugated Alexa. These results suggested that applied OVA are taken up in a healthy intact skin, delivered to secondary lymphoid tissues by CD11c þ cells. Similar delivery by dendritic cells have been observed for myelin peptides applied transdermally to multiple sclerosis patients in human studies. 14 Next, we examined whether EPIT can suppress allergic symptoms elicited by allergen food challenge (Fig. 1A) . Sensitized mice were epicutaneously administrated OVA in PBS (EPIT group) or PBS only (control group) repeatedly for 3 weeks (EPIT course). Following to EPIT course, both group of mice were fed with egg white (EW)-diet to induce allergic symptoms. In this food allergy model, allergic symptoms such as weight loss, T cell-mediated enteropathy, IgE production, and Th2 response are elicited by EW-diet feeding. 15 To evaluate efficacy of EPIT for suppressing food allergy symptoms, the rate of body weight change relative to the start of feeding (day 22) was calculated. Body weight of the control group mice decreased, while that of the EPIT group was significantly alleviated (Fig. 1B) . There was no significant difference between the two groups during the EPIT course ( Supplementary Fig. 2) . Rectal temperature was measured before EW-diet feeding (day 22) and after antigen challenge (day 27), respectively. There was no significant difference in the two groups before EW-diet feeding, while after allergen challenge with EW-diet feeding, rectal temperature drop was significantly rescued in the EPIT group compared to the control group with dropped rectal temperature ( Fig 1C) . Intestinal inflammation observed in the jejunal tissue was shown in Figure 1D . Typical morphological signs of inflammation, such as villous branching, goblet cell hyperplasia and infiltration into the mucosal epithelium in the control group were severer compared to the EPIT group. The ratio of villous height to crypt depth as parameters for monitoring morphology 16,17 decreased in the control group by EW-feeding than that of the EPIT group, showing the jejunal enteropathy significantly relieved by EPIT (Fig. 1E) . These results clearly show that EPIT without any device is effective to suppress systemic allergic symptoms, including T cell-mediated food allergic enteropathy. Next, we analyzed the parameters related to suppression of allergic symptoms in this EPIT model. We measured OVA-specific IgE antibody levels in sera by ELISA. There was no significant difference between the two groups after immunization (day 0) and EPIT course (day 22). On day 27, robust OVA-specific IgE production elicited by EW-feeding was observed in the control group, which was significantly suppressed in the EPIT group (Fig 1F) . Foxp3 þ Tregs are known for their suppressive function in antigen specific manner. After EW-diet feeding, we obtained spleen (SPL), mesenteric lymph nodes (MLNs), DLNs and analyzed Foxp3 expression for individual mice. In SPL and MLN, the frequency of Foxp3 expression was significantly higher in the EPIT group than that of control group (Fig. 1G) . On the other hand, in the skin DLN, such significant difference was not observed. To further analyze the mechanism of EPIT, we examined the effect of EPIT before EW-feeding ( Fig. 2A) . We considered that immune responses may be induced in the skin DLN during the EPIT course. CD4 þ T cells from DLN after EPIT course were cultured under OVA stimulation. In CD4 þ T cells from DLN of EPIT group, Th2 response to OVA stimulation, such as IL-4 production (Fig. 2B) , and cell proliferation ( Fig 2C) were induced OVA-dosedependently. Effects on other cytokines such as IFN-g and IL-10, and on SPL and MLN-derived CD4 þ T cells by EPIT were not stable (data not shown). To identify OVA-specific Tregs, we tracked the cells proliferated in response to OVA stimulation using 5,6calboxyfluorscein diacetate (CFSE) and analyzed Foxp3 expression by CD4 þ T cells (Supplementary Fig. 3 ). In DLN of the EPIT group, the proportion of OVA-specific Foxp3 þ cells in CD4 þ T cells was Peer review under responsibility of Japanese Society of Allergology.