Differential roles of migratory and resident DCs in T cell priming after mucosal or skin HSV-1 infection

Abstract

Although mucosal surfaces represent the main portal of entry for pathogens, the mechanism of antigen presentation by dendritic cells (DCs) that patrol various mucosal tissues remains unclear. Instead, much effort has focused on the understanding of initiation of immune responses generated against antigens delivered by injection. We examined the contributions of migratory versus lymph node-resident DC populations in antigen presentation to CD4 and CD8 T cells after needle injection, epicutaneous infection, or vaginal mucosal herpes simplex virus (HSV) 1 infection. We show that upon needle injection, HSV-1 became lymph-borne and was rapidly presented by lymph node-resident DCs to CD4 and CD8 T cells. In contrast, after vaginal HSV-1 infection, antigens were largely presented by tissue-derived migrant DCs with delayed kinetics. In addition, migrant DCs made more frequent contact with HSV-specific T cells after vaginal infection compared with epicutaneous infection. Thus, both migrant and resident DCs play an important role in priming CD8 and CD4 T cell responses, and their relative importance depends on the mode of infection in vivo.

Figure 1

Both CD8α⁺ and CD8α⁻ DCs, but not pDCs, present viral antigens to CD4 and CD8 T cells after mucosal and cutaneous HSV-1 infection. Groups of mice were infected with 10⁶ PFU HSV-1 via f.p. (A–C and G–I) or ivag (D–F and J–L) and at 20 h (A–F) or 72 h (G–L) after infection, the draining lymph nodes were collected, DCs were enriched, and CD8α⁺ DCs, pDCs, and DN DCs were sorted as described in Fig. S1 (A and B; available at http://www.jem.org/cgi/content/full/jem.20080601/DC1). 25,000 sorted DCs were coincubated with 10⁵ HSV-1–specific naive gBT-I cells (A, D, G, and J), bulk effector CD8 T cells (B, E, H, and K), or bulk effector CD4 T cells (C, F, I, and L) for 72 h, and IFN-γ secretion was measured. These results are representative of three similar experiments. Data are means ± SD.

Figure 2

NK cell contamination impairs DN DC presentation of viral antigens to CD4 and CD8 T cells. (A) FACS sorting strategy of DC subsets according to previously described methods. CD11c⁺ DCs were enriched by negative selection of T cells, B cells, granulocytes, and erythrocytes, and were labeled with anti-CD11c, anti-CD8, and anti-CD45RA to distinguish pDCs, CD8 DCs, and DN DCs (references 8, 10, 11). The sorted populations were analyzed for the purity of the cell types using anti-MHC II and anti–pan-NK antibodies (percentages are shown). (B and C) Using the strategy described in A, DCs from the draining lymph nodes of f.p. HSV-1–infected mice at 72 h after infection were sorted and co-cultured with HSV-specific CD4 (B) and CD8 (C) T cells for 72 h. IFN-γ was measured by ELISA. Antigen presentation by DN DC populations containing NK cell contamination (as in A) or purified DN DCs without NK cells was assessed. Similar results were obtained from three independent experiments. Data are means ± SD. SSC, side scatter.

Figure 3

Relative numbers of DC subsets in the skin and vagina draining lymph nodes. Total mean numbers of DCs belonging to the three subsets in the draining lymph nodes of mice (n = 5 per group) infected with HSV-1 via the f.p. (popliteal; A) or the ivag (iliac and inguinal; B) route at 0, 20, and 72 h after infection were assessed. These results are representative of three similar experiments.

Figure 4

Mucosally applied antigens do not directly access the lymph node. Mice were injected with FITC-dextran (100 µg per mouse) in the f.p. by needle (A and C) or into the vaginal cavity with a pipette (B and D). The applied antigen was tracked at the site of inoculation, in the lymph draining site of inoculation, and within the draining lymph nodes 30 min (A and B) or 24 h (C and D) later. These figures are representative of three similar experiments. Bars, 100 µm.

Figure 5

Characterization of the migrant versus lymph node–resident DC populations in the draining lymph nodes. Mice were painted on the flank skin with 1% FITC solution in acetone (A) or inoculated vaginally with 1% FITC in DMSO (B). At 72 h, draining lymph nodes (popliteal, A; inguinal and iliac, B) were collected, and NK cell–depleted CD11c⁺ DCs were analyzed by FACS. FITC labeling profiles for EpCAM⁺ Langerhans cells or EpCAM⁻ non–Langerhans cell DCs are depicted (percentages are shown). (C) CD11c⁺ CD8α⁺ EpCAM⁻ DCs were analyzed by FACS 72 h after FITC painting of the skin and FITC inoculation ivag (percentages are shown). (D) Mice were painted on the flank skin with 1% FITC solution in acetone. One group of mice received PTX injection at the site of FITC painting 18 h earlier. At 72 h, draining lymph nodes were collected, and DCs were analyzed by FACS for FITC incorporation. These figures are representative of three similar experiments.

Figure 6

Migrant submucosal DCs present viral antigen to both CD4 and CD8 T cells after vaginal HSV-1 infection. (A) 72 h after ivag HSV-1 infection or mock infection, draining lymph node NK cell–depleted CD8α⁻ B220⁻ CD11c⁺ DCs were sorted into Langerhans cell (EpCAM⁺), submucosal DC (EpCAM⁻ CD205^int), and lymph node–resident DN DC (CD205^lo) populations. The purity of the sorted populations is indicated in the rightmost columns (percentages are shown). (B) HSV-1–specific CD4 and CD8 T cells were co-cultured with the sorted DC populations from A in the absence of exogenously added antigens. IFN-γ secretion from T cells was measured after 72 h of culture by ELISA. These data are representative of three similar experiments. Data are means ± SD. (C) EpCAM⁺-sorted DCs (as in A, top row) were stained intracellularly with antilangerin or with an isotype control.