Langerhans cells (LCs) are dendritic cells (DCs) localized in stratified epithelia, such as those overlaying skin, buccal mucosa, and vagina. The contribution of LCs to the promotion or control of immunity initiated at epithelial sites remains debated. We report in this paper that an immunogen comprising OVA linked to the B subunit of cholera toxin, used as delivery vector, was efficient to generate CTLs after vaginal immunization. Using Lang-EGFP mice, we evaluated the contribution of distinct DC subsets to the generation of CD4 and CD8 T cell responses. We demonstrate that the vaginal epithelium, unlike the skin epidermis, includes a minor population of LCs and a major subset of langerin DCs. Intravaginally administered Ag is taken up by LCs and langerin DCs and carried up to draining lymph nodes, where both subsets prime CD8 T cells, unlike blood-derived DCs, although with distinct capabilities. LCs prime CD8 T cells with a cytokine profile dominated by IL-17, whereas Lang DCs induce IFN-γ–producing T cells. Using Lang-DTR-EGFP mice to ensure a transient ablation of LCs, we found that these cells not only are dispensable for the generation of genital CTL responses but also downregulate these responses, by a mechanism that may involve IL-10 and IL-17 cytokines. This finding has implications for the development of mucosal vaccines and immunotherapeutic strategies designed for the targeting of DCs.

The genital mucosa constitutes the major site of sexually transmitted infections. This mucosa is a highly specialized immune system including effector mechanisms, such as secretory Ab production and T cell responses, playing a crucial role in protection against pathogen invasion (1, 2). Following genital exposure to HIV, the expression of secretory Abs and CTL responses in the genital compartment seems to be involved in the control of HIV infection by female sex-workers (35).

In addition, it has been demonstrated that mucosal routes of immunization (nasal, genital) are more efficient than parenteral vaccination for inducing secretory Abs in genital tissues (6, 7). Moreover, genital immunization with subunit vaccines induces the generation of CTLs in draining ileosacral lymph nodes (ILNs) (8, 9) as well as in genital mucosa, whereas systemic immunization is unable to evoke vaginal CTL responses (9). Nevertheless, a better understanding of the early events involved in the generation of CTLs initiated in the reproductive tract is required for the development of targeted vaccination strategies.

An effective cellular immune response requires that APCs process Ags and present epitopes to T cells. Dendritic cells (DCs) are known to be the most potent APCs able to prime both CD4+ and CD8+ T cells. In the immature stage, DCs act as sentinels, as they are located in most nonlymphoid tissues and in close contact with the epithelial surfaces, where they can sense and sample incoming pathogens. Among these, Langerhans cells (LCs) represent the immature DCs located in stratified epithelia, such as the skin and the buccal and vaginal mucosae. Because of their strategic location, LCs are thought to participate in maintaining peripheral tolerance and/or in sensing “danger” signals at epithelial surfaces (10). The implication of LCs in the priming of CD8 T cells after vaginal immunization still remains to be determined. To date, few studies have analyzed the DC subsets located in the mouse vaginal mucosa. Parr and Parr (11) have highlighted the heterogeneity of APCs, including DCs, in the vaginal mucosa. Different subsets of DCs within the vaginal epithelium have been described on the basis of their differential expression of CD205 and F4/80. Their frequency and exact location in the epithelium and vaginal stroma were dependent on the menstrual cycle (11). Iwasaki et al. (12, 13) have confirmed the heterogeneity of vaginal epithelial DCs (VEDCs) and have reported that VEDCs express very low levels or no langerin, a C-type lectin originally identified as a specific marker of LCs (14, 15). Recently, Langerin+ DCs have been also described in the mouse dermis (1618). Furthermore, knock-in mice that express the enhanced GFP (EGFP) and the receptor for diphtheria toxin under the control of the langerin gene have been engineered to study langerin+ DC functions in vivo (1921). These mice constitute a unique tool to examine the frequency and functions of murine LCs in most stratified tissues.

We have previously shown that the nontoxic B subunit of cholera toxin (CTB) constitutes an efficient delivery vector for inducing vaginal CTL responses against a colinked Ag after intravaginal immunization (9). In this paper, we have reported the contribution of vagina-derived DCs, including LCs and langerin DCs (Lang DCs), in the generation of local CTL responses following intravaginal immunization with an immunogen comprising CTB conjugated to OVA (CTB-OVA).

Female 6- to 8-wk-old C57BL/6 mice were from Charles River Laboratories (L'Arbresle, France). Lang-EGFP and Lang-DTR-EGFP transgenic mice were described elsewhere (19). LC ablation into Lang-EGFP/Lang-DTR-EGFP mice was accomplished by i.p. injections of diphtheria toxin (DT; 1μg) (List Biological Laboratories, Campbell, CA) 4 and 1 d prior to immunization. OT-I and OT-II transgenic mice were used as the source of OVA-specific CD4 and CD8 T cells. All mice were maintained in pathogen-free conditions, and studies were approved by our institutional review committee.

Cholera toxin (CT) was obtained from List Biological Laboratories. Recombinant CTB was from SBL Vaccines (Uppsala, Sweden). Albumin Chicken Egg Grade VII (OVA) was from Sigma-Aldrich (St. Louis, MO). OVA was conjugated to CTB, and conjugates were checked by ELISA, as described earlier (22, 23). OVA257–264 (SIINFEKL) and OVA323–339 (ISQAVHAAHAEINEAGR) peptides were from NeoMPS (Strasbourg, France). LPS contamination of preparations was <0.12 ng/ml (Limulus amebocyte test Pyrogen Plus; Cambrex BioWhittaker, East Rutherford, NJ).

Depoprovera-treated C57BL/6 or Lang-EGFP/Lang-DTR-EGFP mice (depleted or not in LCs by DT injections) were immunized vaginally at days 0, 14, and 21 with 20 μl CTB-OVA (40 μg), with CT (1 μg) or OVA (20 μg), and used to analyze ex vivo effector TCR-OVA CD8 T cells and to monitor in vivo cytolysis. In vivo cytolysis was conducted as previously described after injection of fluorescent targets (9, 24). For Ag presentation experiments, Lang-EGFP mice received a vaginal 20-μl dose of Ag or a 5-μl dose of Ag under the tongue (sublingual immunization). Organs (vagina, ILNs, or submandibular lymph nodes [SMLNs]) were collected at time points between 6 and 72 h postimmunization.

Vaginas were excised carefully to avoid any contamination with outer skin. Vaginal epithelial and lamina propria tissues were mechanically separated after overnight incubation at 4°C in dispase II (Roche Diagnostic Systems, Somerville, NJ) and cells dispersed after collagenase/dispase treatment (Roche Diagnostics) in the presence of DNase I (Roche Diagnostics).

Low-density DC-enriched suspensions from ILNs or SMLNs from C57BL/6 or Lang-EGF mice were prepared onto an optiprep gradient as previously described (25, 26). Phenotypic analysis or DC sortings were performed after multiparametric stainings with PE–anti-MHC class II (MHCII) (clone 2G9), PerCP–anti-CD8α (clone 53-6-7), Pe–Cy7–anti-CD11c (clone HL3), APC–Cy7–anti-CD11b (clone M1/70), APC–anti-CD45RB (B220) (clone RA3-6B2), or biotin-mPDCA (Miltenyi Biotec, Auburn, CA), revealed with APC-streptavidin and Alexa488-anti-langerin (clone 923B7; Dendritics, Lyon, France) (intracellular staining onto DC preparations from C57BL/6 mice only, according to the manufacturer’s instructions). Vaginal intraepithelial and lamina propria DCs were analyzed using Alexa488–anti-langerin, PE–anti-MHCII, PerCP–anti-CD11b, Pe–Cy7–anti-CD11c, APC–Cy7–anti-CD45 (clone 30-F11), APC–anti-Gr1 (clone RB6-8C5) or biotin–anti-CD80 (clone 16-10A1), biotin–anti-CD86 (clone GL1), anti-macrophage biotin-F4/80 (clone C1.A3-1; Caltag Laboratories, Burlingame, CA), biotin–anti-CD205 (clone NLDC-145; Cedarlane Laboratories, Hornby, Ontario, Canada), and biotin–anti-CD24 (clone M1/69), revealed with APC-streptavidin. For T cell analysis (either DC-primed T cells or cell suspensions from Lang-DTR-EGFP immunized mice), cell suspensions from cultures were stained with APC–Cy7–anti-CD45 (clone 30-F11), FITC–anti-CD3 (clone, PE–Cy7–anti-CD8α (clone 53-6-7), PE-conjugated–H-2Kb/SIINFEKL pentamer (ProImmune, Oxford, U.K.), and PerCP–anti-MHCII (clone 2G9). In some experiments, cells were then fixed in 2% (vol/vol) paraformaldehyde in PBS and washed before intracellular staining. Intracellular stainings for PE–anti–IL-17A (clone TC11-18H10) or for PE–anti-Granzyme B (clone 16G6, eBioscience, San Diego, CA) and APC–IFN-γ (cloneXMG1.2) were done following the manufacturers’ recommendations. When not indicated, Abs were from BD Biosciences (San Jose, CA). Fc receptors were blocked with anti-FcγRII/III mAb (2.4G2 clone). Isotype control stainings were performed. Analyses were performed at the Pasteur flow cytometry platform under the supervision of A. Loubat (Institut fédératif de recherche 50, Nice, France) onto a FACSCANTO cytometer and sortings onto a FACSARIA cytometer, equipped with FACSDIVA software (BD Biosciences). DC subsets were sorted with a purity >95%.

Sorted DC subsets were cultured at concentrations between 2.5 × 103 and 10 × 103 DCs, with 105 CFSE-labeled naive OT-I and OT-II T cells for 4 d in flat-bottom 96-well plates without exogenous peptides. CFSE dilution was measured by flow cytometry. Controls in which T cells and DC subsets were preincubated with OVA peptides in vitro were performed to check the functionality of DC subsets after isolation. Naive CFSE-labeled OT-I and OT-II T cells were prepared from lymph nodes, as previously described (9).

After the priming step, 104 DC-primed T cells were stimulated with OVA-pulsed splenic APCs (5 × 103) for 3 d in round-bottom 96-well plates, and supernatants were collected and assayed for IFN-γ, IL-10, IL-17A, TNF-α, and IL-4, using mouse cytometric bead array kits (BD Biosciences) according to the manufacturer’s recommendations. Cytokine profiles (IL-17A, IFN-γ) were also evaluated by intracellular staining. Phorbol 12-myristate 13-acetate (200 ng/ml; Sigma-Aldrich), ionomycin (1 mg/ml; Sigma-Aldrich), and GolgiPlug (BD Biosciences) were added in the last 4 h of secondary cultures. Cells were then collected and extra- and intracellular stainings done as described above in the flow cytometry analysis and sortings discussion.

Organs (vagina, ILNs, spleen) from DT-treated and nontreated immunized Lang-DTR-EGFP mice were collected 7 d after the last immunization. Spleen and ILN cells were dispersed mechanically. Vaginas were excised carefully, and vaginal cells were mechanically dispersed after collagenase A digestion (Roche Diagnostics) of vaginal fine pieces for 45 min at 37°C in the presence of DNAse I (Roche Diagnostics). Cell suspensions were freed from erythrocytes by treatment with ammonium chloride. Cell suspensions were counted and directly stained, as described in the flow cytometry discussion, above, to determine the frequency and absolute numbers of TCR-OVA+ CD8+ T cells. In parallel, 5 × 105 total cells were stimulated with phorbol 12-myristate 13-acetate (200 ng/ml; Sigma-Aldrich) and ionomycin (1 mg/ml; Sigma-Aldrich) in the presence of GolgiPlug (BD Biosciences) for 4 h in round-bottom 96-well plates before analysis of IFN-γ and granzyme B expression by intracellular staining to determine the frequency of effector CD8 T cells.

Groups were compared using a one-way variance analysis and a Tukey test. A p value < 0.05 was considered significant. All statistical calculations were computed with Sigma Stat software (SPSS, Chicago, IL).

Using different OVA Ag formulations administered by the vaginal route in C57BL/6 mice, we demonstrated that CTB-OVA induced moderate OVA-specific CTLs (30% OVA-specific cytolysis) in ILNs, which drain the genital and rectal mucosae, but not the skin. The administration of CTB-OVA in the presence of CT, a strong mucosal adjuvant, potentiated the OVA-specific cytotoxic activity (90% OVA-specific cytolysis), whereas OVA used alone at the same concentration or coadministered with CT was inefficient (Fig. 1A) (9). Complementary to this, we previously demonstrated that the immunogen delivery properties of CTB were explained by its mucoadhesive properties toward cells expressing the ganglioside GM1 (all nucleated cells, including mucosal epithelial cells) and its ability to deliver immunogens in the MHC class I pathway (9).

FIGURE 1.

Vagina-derived DCs prime CD8 T cell responses after intravaginal immunization. A, OVA-specific cytolysis in ILNs from C57BL/6 mice (n = 5) immunized by the intravaginal route (d0, d14, d21) with different Ags was measured with an in vivo CTL assay 7 d later by flow cytometry. FACS profiles correspond to a representative experiment. Numbers represent the percentages of unpulsed CFSElow or OVA-pulsed CFSEhigh cells. DCs (B, C) and non-DC APCs (B) or DC subsets (D) were isolated from ILNs 36 h after intravaginal immunization. These cells (104) were cultured for 4 d with CFSE-labeled OT-I T cells (105) without any addition of antigenic peptide. B, Flow cytometry histograms correspond to CFSE dilution of gated CD8+ TCR-OVA+ T cells from a representative experiment, and percentages represent the frequency of dividing cells. C, Histograms represent the mean value (+SEM) of the percentages of dividing cells of three independent experiments for different DC/T ratios. D, DC subsets were sorted into Muc-DC, conventional cDC, and pDC. Flow cytometry histograms correspond to the proliferation of gated CD8+ TCR-OVA+ T cells primed by these different DC subsets. Data are representative of three independent experiments.

FIGURE 1.

Vagina-derived DCs prime CD8 T cell responses after intravaginal immunization. A, OVA-specific cytolysis in ILNs from C57BL/6 mice (n = 5) immunized by the intravaginal route (d0, d14, d21) with different Ags was measured with an in vivo CTL assay 7 d later by flow cytometry. FACS profiles correspond to a representative experiment. Numbers represent the percentages of unpulsed CFSElow or OVA-pulsed CFSEhigh cells. DCs (B, C) and non-DC APCs (B) or DC subsets (D) were isolated from ILNs 36 h after intravaginal immunization. These cells (104) were cultured for 4 d with CFSE-labeled OT-I T cells (105) without any addition of antigenic peptide. B, Flow cytometry histograms correspond to CFSE dilution of gated CD8+ TCR-OVA+ T cells from a representative experiment, and percentages represent the frequency of dividing cells. C, Histograms represent the mean value (+SEM) of the percentages of dividing cells of three independent experiments for different DC/T ratios. D, DC subsets were sorted into Muc-DC, conventional cDC, and pDC. Flow cytometry histograms correspond to the proliferation of gated CD8+ TCR-OVA+ T cells primed by these different DC subsets. Data are representative of three independent experiments.

Close modal

We further examined which type or types of APCs were involved in the priming of CD8 T cells after intravaginal immunization. CD11c+ DCs and CD11cMHCII+ APCs (non-DC APCs) were isolated from the ILNs of mice at different time points between 6 h and 72 h after intravaginal immunization with CTB-OVA+CT or OVA alone used as control. Enriched APCs were cultured with naive OVA-specific CD8 T cells. Preliminary studies indicated that good proliferative responses were induced 36 h after immunization (data not shown) (9). As illustrated in Fig. 1B, DCs isolated from mice 36 h after intravaginal immunization with CTB-OVA+CT were highly efficient in inducing the specific proliferation of CD8+ T cells, as measured by CFSE dilution (60% proliferating OT-I T cells), whereas non-DC APCs were not (4% proliferating OT-I T cells). The rate of proliferation induced by ILN DCs was dependent on the T/DC ratio (Fig. 1C).

We then evaluated whether the different DC subsets present in the ILNs were able to present the Ag acquired in vivo after intravaginal immunization. In the mouse, DC subtypes express the αN integrin CD11c and the MHCII molecule (27). On the basis of the phenotypical properties assigned to tissue-derived and blood-derived DCs of peripheral lymph nodes in different studies using the differential expression of MHCII and CD11c, we identified three major DC subsets in the ILNs: CD11cintMHCIIhigh DCs corresponding to mucosa-derived DCs (Muc-DCs), CD11chighMHCIIint corresponding to conventional blood-derived DCs (cDCs), and CD11cintMHCIIint DCs, which also express the mPDCA marker specific to plasmacytoid DCs (pDCs) (Fig. 1D, left panel; data not shown). Muc-DCs, obtained from mice immunized 36 h earlier with CTB-OVA+CT given by the vaginal route, but not the same cells from mice immunized with OVA alone, were able to induce the proliferation of CD8 OT-I T cells (66% versus 0.8% proliferating T cells, respectively) (Fig. 1D, right panel). Neither cDCs nor pDCs were able to present the CD8 OVA epitope to OT-I T cells, as shown by the absence of proliferation (Fig. 1D , right panel). Nevertheless, sorted cDCs and pDCs are functional after isolation, as they are able to prime OT-I T cells when pulsed in vitro with the OVA epitope (SIINFEKL) (data not shown). This strongly suggests that the Ag had been taken up locally by mucosal DCs and carried into draining ILNs. Furthermore, DC subsets isolated from ILNs at shorter time points (6 and 12 h) postimmunization were not able to present the Ag administered by the vaginal route, which sustains this conclusion (data not shown). Together, these results rule out the possibility that the topically-applied Ag had diffused alone through lymphatic vessels. Alternatively, we cannot exclude that the Ag may have diffused, but not in an immunologically relevant fashion.

To identify the nature of the Muc-DCs involved in the priming of CD8 and CD4 T cells, we analyzed the phenotype of DC subsets present in the vaginal mucosa and in the ILNs before and after Ag administration, using classical markers expressed by DC subsets of the skin (14, 28, 29). We used Lang-EGFP mice, in which langerin+ DCs, including LCs, express the EGFP (19). Cell suspensions from vaginal epithelium and lamina propria were analyzed by flow cytometry for their content in CD11c+MHCII+ DCs expressing langerin. To this end, we compared langerin expression in C57BL/6 mice, using a monoclonal anti-langerin Ab, and in Lang-EGFP mice, measuring EGFP expression. Only a small DC subset expresses langerin, which corresponds to ∼1–3% of CD11c+MHCII+ DCs, accounting for ∼1% of total epithelial cell suspensions from progesterone-treated mice (Fig. 2A, upper panel, Supplemental Fig. 1). This intraepithelial population represents a higher percentage in Lang-EGFP mice (Fig. 2A,, upper panel). We also analyzed the frequency of LCs in the vaginal epithelium in the absence of progesterone treatment. We observed that LC frequency in nontreated mice represented 12–20% of intraepithelial DCs, corresponding to only a 2- to 3-fold increase in LC numbers per vagina, compared with progesterone-treated mice (Fig. 2A,, lower panel). We then phenotypically characterized the DCs present in the vaginal mucosa of progesterone-treated C57BL/6 mice, using intracellular langerin staining (Fig. 2B, 2C). Vaginal LCs represent a rare but homogeneous population of mucoepithelial LCs expressing high levels of CD205 and CD24, and intermediate levels of costimulatory molecules CD80 and CD86 (Fig. 2B, right histograms). The majority of intraepithelial DCs do not express langerin (Lang DCs). They constitute a heterogeneous population, as shown by variable expression of F4/80 and CD24 according to previous observations (Fig. 2B,, left histograms; Refs. 1113). LCs or langerin+ DCs are undetectable in the vaginal lamina propria, which contains a subset of DCs quite similar to dermal DCs (Fig. 2C, Supplemental Fig. 2). Furthermore, vaginal DC subsets from the epithelium and the lamina propria are negative for CD8α, CD45RB/B220, and NK/DX5 markers (data not shown).

FIGURE 2.

Phenotype of intraepithelial and lamina propria DCs from vagina. AD, Vaginal epithelium and lamina propria cell suspensions were analyzed by flow cytometry after surface staining. A, Dot plots represent the expression of langerin versus MHCII gated on CD11c+MHCII+ DCs in vaginal epithelial cell suspensions from progesterone-treated (upper panels) and nontreated (lower panels) mice. Langerin expression is analyzed in C57BL/6 mice after intracellular staining with a monoclonal anti-langerin Ab (left panels) or in Lang-EGFP mice, measuring GFP fluorescence (right panels). LC percentages in CD11c+MHCII+ DCs and LC numbers per tissue are indicated in each dot. B and C, Dark-gray histograms represent expression of different markers for gated intraepithelial Lang DCs and LCs (B) and for lamina propria Lang DCs (C) in C57BL/6 mice. Light-gray histograms represent isotype controls. Numbers in histograms correspond to the mean fluorescence intensity of specific labelings. D, Vaginal epithelium cell suspensions from CTRL and CTB-OVA+CT–treated C57BL/6 mice were prepared at different intervals after intravaginal application. Absolute DC numbers per organ were determined after counting viable cells and analysis of DC subset percentages following multiparametric staining. Results are expressed as means (+SEM) of DC numbers per epithelial tissue from five independent experiments. *Significant difference (p < 0.05, Tukey test). CTRL, control.

FIGURE 2.

Phenotype of intraepithelial and lamina propria DCs from vagina. AD, Vaginal epithelium and lamina propria cell suspensions were analyzed by flow cytometry after surface staining. A, Dot plots represent the expression of langerin versus MHCII gated on CD11c+MHCII+ DCs in vaginal epithelial cell suspensions from progesterone-treated (upper panels) and nontreated (lower panels) mice. Langerin expression is analyzed in C57BL/6 mice after intracellular staining with a monoclonal anti-langerin Ab (left panels) or in Lang-EGFP mice, measuring GFP fluorescence (right panels). LC percentages in CD11c+MHCII+ DCs and LC numbers per tissue are indicated in each dot. B and C, Dark-gray histograms represent expression of different markers for gated intraepithelial Lang DCs and LCs (B) and for lamina propria Lang DCs (C) in C57BL/6 mice. Light-gray histograms represent isotype controls. Numbers in histograms correspond to the mean fluorescence intensity of specific labelings. D, Vaginal epithelium cell suspensions from CTRL and CTB-OVA+CT–treated C57BL/6 mice were prepared at different intervals after intravaginal application. Absolute DC numbers per organ were determined after counting viable cells and analysis of DC subset percentages following multiparametric staining. Results are expressed as means (+SEM) of DC numbers per epithelial tissue from five independent experiments. *Significant difference (p < 0.05, Tukey test). CTRL, control.

Close modal

We also observed a transient increase in Lang DC and LC numbers in the vaginal epithelium 6 h after intravaginal administration of CTB-OVA+CT, demonstrating that both populations were able to respond to this immunostimulatory Ag (Fig. 2D). We then analyzed the langerin expression by the DC subsets present in ILNs, and we found that ILNs contained only a very small subset of mucosa-derived LCs, accounting for 3% Muc-DCs (Fig. 3A). Furthermore, these LCs did not express CD8α, which is consistent with their mucosal origin. Their low percentage among Muc-DCs (3%) is in accordance with the low proportion of LCs in the genital mucosa (Figs. 2A, 3A). Interestingly, 36 h after intravaginal administration of CTB-OVA+CT, we observed in the ILNs 6-fold and 3-fold increases in the frequencies of mucosa-derived Lang DC and LC subsets, respectively, whereas cDC and pDC frequencies were not significantly modified (Fig. 3B; data not shown). The increase in DC frequency within ILNs was accompanied by their maturation, as indicated by a slight upregulation of costimulatory molecules CD80 and CD86, suggesting that these cells have recently emigrated from the mucosa after activation (Fig. 3C). Notably, mucosa-derived Lang DCs seemed to have a more mature phenotype than did LCs regarding CD80 expression after intravaginal administration of CTB-OVA+CT.

FIGURE 3.

Intravaginal immunization with CTB-OVA in the presence of CT induces the migration and maturation of Muc-DC subsets in ILNs. AC, DCs enriched from ILNs of mice 36 h after intravaginal immunization with CTB-OVA in the presence of CT or from nonimmunized control mice were analyzed by flow cytometry after multiparametric staining. A, Dot plot FACS profiles show cDCs (MHCIIintCDIIchi) and Muc-DCs (MHCIIhiCD11c+), including Lang DCs and LCs. B, Histogram bars correspond to the means (+SEM) of absolute numbers in the ILN cDC and in the mucosa-derived Lang DC and LC subsets per ILN in OVA-treated (white bars) and CTB-OVA+CT–treated (gray bars) mice from three independent experiments. *Significant difference (p < 0.05, Tukey test). C, Flow cytometry histograms correspond to anti-CD80 and anti-CD86 (black curves) or isotype Ab (gray curves) stainings of mucosa-derived Lang DC and LC subsets isolated from control or CTB-OVA+CT–treated mice. These data are representative of three independent experiments.

FIGURE 3.

Intravaginal immunization with CTB-OVA in the presence of CT induces the migration and maturation of Muc-DC subsets in ILNs. AC, DCs enriched from ILNs of mice 36 h after intravaginal immunization with CTB-OVA in the presence of CT or from nonimmunized control mice were analyzed by flow cytometry after multiparametric staining. A, Dot plot FACS profiles show cDCs (MHCIIintCDIIchi) and Muc-DCs (MHCIIhiCD11c+), including Lang DCs and LCs. B, Histogram bars correspond to the means (+SEM) of absolute numbers in the ILN cDC and in the mucosa-derived Lang DC and LC subsets per ILN in OVA-treated (white bars) and CTB-OVA+CT–treated (gray bars) mice from three independent experiments. *Significant difference (p < 0.05, Tukey test). C, Flow cytometry histograms correspond to anti-CD80 and anti-CD86 (black curves) or isotype Ab (gray curves) stainings of mucosa-derived Lang DC and LC subsets isolated from control or CTB-OVA+CT–treated mice. These data are representative of three independent experiments.

Close modal

These data suggest that vaginal LCs and Lang DCs are rapidly mobilized in the ILNs after intravaginal immunization with CTB-OVA+CT, and may be able to present CD4 and CD8 epitopes derived from OVA Ags encountered in the vaginal mucosa.

We next evaluated the ability of mucosa-derived LCs and Lang DCs to prime OVA-specific CD8 and CD4 T cells. For this purpose, mucosa-derived Lang DC and LC subsets were isolated by cell sorting from DC-enriched low-density fractions from ILNs of Lang-EGFP mice 36 h after intravaginal administration of CTB-OVA+CT. Mucosa-derived Lang DC and LC subsets were sorted with a purity of 97% and 98%, respectively (Fig. 4A). Lang DCs and LCs were able to induce the proliferation of OVA-specific OT-I T cells (Fig. 4B,, left panels), which demonstrates that both subsets are capable of capturing the CTB-OVA Ag in the vaginal mucosa, processing it, and presenting pertinent epitopes to specific CD8 T cells. These results are consistent with the mature phenotype of these DC subsets, as observed by flow cytometry (Fig. 3C). Nevertheless, on a per cell number basis, LCs were less efficient than mucosa-derived Lang DCs for inducing CD8 T cell proliferation (69% proliferating OT-I T cells versus 93% proliferating OT-I T cells). Furthermore, LCs failed to induce proliferation of CD4 OT-II T cells, in contrast to mucosa-derived Lang DCs (Fig. 4B , right panels).

FIGURE 4.

Vagina-derived LCs and Lang DCs differentially present CD4 and CD8 OVA epitopes after intravaginal immunization with CTB-OVA and CT. A and B, At 36 h after intravaginal application of CTB-OVA+CT, ILNs were collected and mucosa-derived Langerin+ (LC) and Lang DC subsets were highly purified by sorting these two subsets from the gated Muc-DC subset (MHCIIhigh CD11c+). Percentages in dot plots indicate purity of sorted subsets (A). CFSE-labeled CD8 OVA-specific T cells (OT-I) and CFSE-labeled CD4 OVA-specific T cells (OT-II) were cocultured with sorted DC subsets (5000 per well) without any addition of exogenous Ags for 4 d. Flow cytometry histograms show specific proliferation of gated CD8 (OT-I) and CD4 (OT-II) T cells, as measured by CFSE dilution. Numbers in FACS profiles indicate percentages of dividing cells (B). Results are representative of two independent experiments.

FIGURE 4.

Vagina-derived LCs and Lang DCs differentially present CD4 and CD8 OVA epitopes after intravaginal immunization with CTB-OVA and CT. A and B, At 36 h after intravaginal application of CTB-OVA+CT, ILNs were collected and mucosa-derived Langerin+ (LC) and Lang DC subsets were highly purified by sorting these two subsets from the gated Muc-DC subset (MHCIIhigh CD11c+). Percentages in dot plots indicate purity of sorted subsets (A). CFSE-labeled CD8 OVA-specific T cells (OT-I) and CFSE-labeled CD4 OVA-specific T cells (OT-II) were cocultured with sorted DC subsets (5000 per well) without any addition of exogenous Ags for 4 d. Flow cytometry histograms show specific proliferation of gated CD8 (OT-I) and CD4 (OT-II) T cells, as measured by CFSE dilution. Numbers in FACS profiles indicate percentages of dividing cells (B). Results are representative of two independent experiments.

Close modal

We then asked whether Muc-DCs from another pluristratified mucosa, the sublingual mucosa, had similar ability to present Ag-derived epitopes to specific CD4 and CD8 T cells after local immunization. We chose the sublingual mucosa because it constitutes an interesting site of mucosal immunization, and because, similarly to the vaginal epithelium, the sublingual epithelium contains Lang DCs and LCs but in different percentages (68% Lang DCs versus 32% LCs) (Supplemental Fig. 3) (30). Mucosa-derived Lang DCs and LCs, cDCs, and pDCs were isolated from DC-enriched low-density fractions from draining SMLNs of Lang-EGFP mice at different time points between 12 h and 72 h after sublingual application of CTB-OVA+CT. Preliminary experiments showed that 24 h after sublingual application of CTB-OVA+CT, OVA-specific CD4 and CD8 epitopes were efficiently presented by sorted Muc-DCs but not by cDCs and pDCs (data not shown). At 24 h after immunization, both sublingually-derived Lang DC and LC subsets were able to induce the proliferation of OVA-specific CD8 T cells, demonstrating their ability to capture and process the OVA Ag in vivo (Fig. 5A,, left panel). As observed for vaginal LCs, sublingual LCs, compared with Lang DCs, were inefficient in inducing the proliferation of OVA-specific CD4 T cells (Fig. 5A,, right panel). Nevertheless, sorted LCs were functional, as indicated by their ability to induce strong CD8 and CD4 proliferation in the presence of exogenous OVA (Fig. 5B). The latter observations further exclude that the sorting strategy employed might have inadvertently affected functions of LCs. Furthermore, the culture of OVA-specific CD8 and CD4 T cells alone (Fig. 5D) or in the presence of Lang DC and LC subsets isolated from ILNs of control CT-treated mice (Fig. 5C) did not induce any T cell proliferation, which indicates that the proliferation observed postimmunization with CTB-OVA+CT was Ag specific.

FIGURE 5.

Sublingual LCs and Lang DCs have Ag-presenting properties similar to those of vaginal DCs after sublingual immunization with CTB-OVA and CT. AC, Draining SMLNs were collected from mice 24 h after sublingual topical application of CTB-OVA+CT (A, B) or from control mice (C), and mucosa-derived langerin+ (LC) and Lang DC subsets were highly purified by sorting these two subsets from the Muc-DC subset (MHCIIhigh CD11c+). CFSE-labeled CD8 OVA-specific T cells (OT-I) and CFSE-labeled CD4 OVA-specific T cells (OT-II) were cocultured with the sorted DC subsets, without any addition of exogenous Ags for 4 d (A, C), or in the presence of OVA in culture (B) or were cultured alone (D). Histograms represent specific proliferation of gated CD8 and CD4 T cells, as measured by CFSE dilution. Numbers in FACS profiles indicate percentages of dividing cells. Results are representative of two independent experiments.

FIGURE 5.

Sublingual LCs and Lang DCs have Ag-presenting properties similar to those of vaginal DCs after sublingual immunization with CTB-OVA and CT. AC, Draining SMLNs were collected from mice 24 h after sublingual topical application of CTB-OVA+CT (A, B) or from control mice (C), and mucosa-derived langerin+ (LC) and Lang DC subsets were highly purified by sorting these two subsets from the Muc-DC subset (MHCIIhigh CD11c+). CFSE-labeled CD8 OVA-specific T cells (OT-I) and CFSE-labeled CD4 OVA-specific T cells (OT-II) were cocultured with the sorted DC subsets, without any addition of exogenous Ags for 4 d (A, C), or in the presence of OVA in culture (B) or were cultured alone (D). Histograms represent specific proliferation of gated CD8 and CD4 T cells, as measured by CFSE dilution. Numbers in FACS profiles indicate percentages of dividing cells. Results are representative of two independent experiments.

Close modal

Altogether, our results demonstrate that mucosa-derived Lang DCs and LCs from two different pluristratified (vaginal and sublingual) mucosal tissues are involved in the priming of CD8 T cells induced by topical application of an Ag but not resident DCs (cDCs and pDCs). Moreover, these results demonstrate that Lang DCs and LCs have different stimulatory properties. This observation may be explained in part by the fact that LCs are poor activators of CD4 T cell proliferation, owing to their less mature phenotype and/or their inefficiency to present CD4 epitopes after in vivo Ag capture.

Because mucosa-derived Lang DCs and LCs had different immunostimulatory properties toward CD4 and CD8 T cells, we analyzed the cytokines produced by DC-primed OVA-specific T cells by cytokine detection in supernatants and by intracellular stainings, after a secondary culture in presence of OVA-pulsed splenic APCs. Mucosa-derived Lang DCs primed T cells with a strong proliferative activity (Fig. 4B,, upper panel), and this priming is associated with production of high levels of IFN-γ (4700 ± 290 pg/ml) but limited amounts of IL-10 (34 ± 20 pg/ml) in secondary cultures (Fig. 6A). In contrast, vagina-derived LCs primed OVA-specific CD8 T cells with moderate proliferative activity (Fig. 4B,, lower panel). Secondary culture supernatants of LC-primed T cells induced around 5-fold lower amounts of IFN-γ (1090 ± 580 pg/ml) and higher levels of IL-10 (220 ± 100 pg/ml) (Fig. 6A). The IFN-γ/IL-10 ratio of T cells primed by Lang DCs was 20-fold higher (IFN-γ/IL-10 ratio >100) than the one of T cells primed by mucosa-derived LCs (IFN-γ/IL-10 ratio <5). We also observed that secondary culture supernatants of LC-primed T cells produced substantial amounts of IL-17A cytokine (485 ± 140 pg/ml), contrary to Lang DCs (109 ± 60 pg/ml) or control groups (<20 pg/ml) (Fig. 6A). These results are confirmed by intracellular staining of DC-primed T cells (Fig. 6B). IFN-γ and IL-17A cytokines produced in these secondary cultures are mainly secreted by stimulated DC-primed T cells, not by APCs. Furthermore, LC-primed T cells are producers of IL-17 (1.9% IL-17–expressing T cells) and, to a lesser extent, of IFN-γ (0.9%), whereas Lang DC-primed T cells are major inducers of IFN-γ (5.9% IFN-γ–producing T cells). No specific IL-10 staining was detected by intracellular staining in these cultures. This can in part be explained by the fact that IL-10 is produced at lower levels by these cultures and by a too small percentage of stimulated T cells to be detected.

FIGURE 6.

Vagina-derived LCs and Lang DCs differentially polarize T cells. A and B, CD8 OVA-specific T cells (OT-I) and CD4 OVA-specific T cells (OT-II) were cocultured with sorted DC subsets (5000 per well) from immunized or control mice without any addition of exogenous Ags for 4 d. A total of 10,000 DC-primed T cells or nonprimed T cells were then restimulated with OVA-pulsed splenic APCs at T/APC ratio = 2. A, Production of IFN-γ, IL-10, and IL-17A cytokines was measured from supernatants of these secondary cultures by cytometric bead array technique. Bars represent mean values + SEM from three independent experiments. B, IFN-γ and IL-17A expression by CD3+ T cells and MHCII+ cells of these secondary cultures was also determined after intracellular staining, as detailed in 1Materials and Methods. Results are presented as dot plots corresponding to IL-17A and IFN-γ expression by T cells (left panels) and MHCII+ cells (right panels) for different experimental conditions. Percentages of IL-17A+ and IFN-γ+ cells correspond to the values indicated in the upper left and lower right quadrants of each dot plot. Intracellular stainings with isotype control Abs are shown for T cells and MHCII+ cells. FACS profiles correspond to representative data. Two independent experiments gave similar data.

FIGURE 6.

Vagina-derived LCs and Lang DCs differentially polarize T cells. A and B, CD8 OVA-specific T cells (OT-I) and CD4 OVA-specific T cells (OT-II) were cocultured with sorted DC subsets (5000 per well) from immunized or control mice without any addition of exogenous Ags for 4 d. A total of 10,000 DC-primed T cells or nonprimed T cells were then restimulated with OVA-pulsed splenic APCs at T/APC ratio = 2. A, Production of IFN-γ, IL-10, and IL-17A cytokines was measured from supernatants of these secondary cultures by cytometric bead array technique. Bars represent mean values + SEM from three independent experiments. B, IFN-γ and IL-17A expression by CD3+ T cells and MHCII+ cells of these secondary cultures was also determined after intracellular staining, as detailed in 1Materials and Methods. Results are presented as dot plots corresponding to IL-17A and IFN-γ expression by T cells (left panels) and MHCII+ cells (right panels) for different experimental conditions. Percentages of IL-17A+ and IFN-γ+ cells correspond to the values indicated in the upper left and lower right quadrants of each dot plot. Intracellular stainings with isotype control Abs are shown for T cells and MHCII+ cells. FACS profiles correspond to representative data. Two independent experiments gave similar data.

Close modal

We then evaluated the in vivo contribution of vaginal LCs in the generation of TCR-OVA+ CD8+ T cells and CTL responses after intravaginal immunization with CTB-OVA+CT, as illustrated in Fig. 7A. For this purpose, we used Lang-EGFP/Lang-DTR-EGFP mice, in which DT injections induced a transient depletion of langerin+ cells (16). One group of mice (n = 6) received two consecutive DT injections at days 4 and 1 before intravaginal immunization, a protocol efficient for inducing a total depletion of vaginal LCs (Fig. 7B) and one that has no significant effect on vaginal Lang DC numbers (Supplemental Table I). At 1 wk after the third intravaginal immunization, the presence of functional OVA-specific CTLs was determined in ILNs, vaginal mucosa, and spleen in nontreated (−DT) and DT-treated (+DT) mice. The in vivo ablation of LCs did not suppress the generation of vaginal and systemic OVA-specific cytotoxic responses (Fig. 7C), which is not surprising owing to the high priming potential of Lang DCs after mucosal CTB-OVA in vivo uptake (Figs. 4B, 6). Interestingly, the in vivo ablation of vaginal LCs induced a statistically significant increase in cytotoxic activity (34%) in the vagina (Fig. 7C). Furthermore, we analyzed the frequency and absolute numbers of Ag-specific TCR-OVA+ CD8+ T cells after immunization in DT-treated (+DT) and nontreated (−DT) mice. As shown in Fig. 7D for a representative experiment, the frequency of TCR-OVA+ CD8+ T cells was increased in the vaginal mucosa of mice immunized in the absence of LCs (+DT), in comparison with nontreated (−DT) mice (29% versus 13% TCR-OVA+ CD8+ T cells in vagina). The increased frequency of Ag-specific T cells in the absence of LCs is also accompanied by a 3-fold increase in the absolute numbers of TCR-OVA+ CD8+ T cells per vagina in DT-treated mice compared with nontreated mice (Vagina, +DT: 14900 ± 1500 TCR-OVA+ CD8+ T cells versus vagina, −DT: 4600 ± 1500 TCR-OVA+ CD8+ T cells) (Fig. 7E). Interestingly, LC depletion is also associated with an increased frequency of granzyme B+ IFN-γ+ in CD8 T cells in the vaginal mucosa of DT-treated mice (Fig. 7F). The effect of LC depletion on Ag-specific CD8 T cells is restricted to vaginal effector T cells and has no significant effect on ILN and spleen effector T cells (Fig. 7C–E).

FIGURE 7.

In vivo ablation of LCs prior to intravaginal immunization favors the generation of mucosal CTLs. Lang-EGFP/Lang-DTR-EGFP mice received three intravaginal immunizations at days 0, 14, and 21 with CTB-OVA+CT. A group of mice received i.p. injections of DT on days 4 and 1 prior to immunizations, as illustrated in A. B, LC depletion was checked before immunization by flow cytometry analysis onto low-density enriched cell preparations from vaginal epithelium (B, left panel) and epidermis (B, right panel). FACS dot plots show the frequency of LC (CD11c+EGFP+cells) gated on CD45+ cells from nontreated mice (−DT, upper panel) and DT-treated mice (+DT, lower panel). At 7 d after the third immunization, OVA-specific CD8 T cells were analyzed. C, OVA-specific CD8 T cytolysis was measured in vivo after i.v. injection or direct injection in vaginal mucosa of SIINFEKL-pulsed CFSEhigh and unpulsed CFSElow target cells. Specific cytolysis was assessed by flow cytometry analysis of CFSElow or CFSEhigh cells in ILNs, vaginal mucosa, and spleen. Results are represented as means (+SEM) of percentages of specific cytolysis for nontreated (−DT, black bars) and DT-treated (+DT, gray bars) immunized mice from three independent experiments. *Significant difference (p < 0.05, Tukey test). D and E, TCROVA+ CD8+ T cell frequency and absolute numbers per organ were determined by flow cytometry after multiparametric stainings of cell suspensions from nonimmunized Lang-DTR-EGFP mice (CTRL), or nontreated (−DT) or treated (+DT) immunized Lang-DTR-EGFP mice. D, FACS dot plots show the percentages of CD8+ T cells gated on CD45+ cells and TCR-OVA+ (Kb/SIINFEKL pentamer staining) in CD8+ T cells from vaginal tissue for a representative experiment. E, TCR-OVA+ CD8+ T cell frequency (left graph) and absolute numbers (right graph) were obtained from flow cytometry analysis and cell counting per organ. Bars represent mean values (+SEM) from three independent experiments. *Significant difference (p < 0.05, Tukey test). F, Cell suspensions (5 × 105) from CTRL and immunized mice were stimulated with PMA/ionomycin for 4 h and stained with IFN-γ, granzyme B, and CD8 Abs, as detailed in 1Materials and Methods. Dot plots show the expression of IFN-γ and Grz B gated on CD8 T cells from CTRL, nontreated (−DT), and DT-treated (+DT) immunized groups for a representative experiment. Bars correspond to the mean percentages (+SEM) of Grz B+IFN-γ+ in CD8+ T cells from two independent experiments with five pooled organs per experiment. Grz B, granzyme B.

FIGURE 7.

In vivo ablation of LCs prior to intravaginal immunization favors the generation of mucosal CTLs. Lang-EGFP/Lang-DTR-EGFP mice received three intravaginal immunizations at days 0, 14, and 21 with CTB-OVA+CT. A group of mice received i.p. injections of DT on days 4 and 1 prior to immunizations, as illustrated in A. B, LC depletion was checked before immunization by flow cytometry analysis onto low-density enriched cell preparations from vaginal epithelium (B, left panel) and epidermis (B, right panel). FACS dot plots show the frequency of LC (CD11c+EGFP+cells) gated on CD45+ cells from nontreated mice (−DT, upper panel) and DT-treated mice (+DT, lower panel). At 7 d after the third immunization, OVA-specific CD8 T cells were analyzed. C, OVA-specific CD8 T cytolysis was measured in vivo after i.v. injection or direct injection in vaginal mucosa of SIINFEKL-pulsed CFSEhigh and unpulsed CFSElow target cells. Specific cytolysis was assessed by flow cytometry analysis of CFSElow or CFSEhigh cells in ILNs, vaginal mucosa, and spleen. Results are represented as means (+SEM) of percentages of specific cytolysis for nontreated (−DT, black bars) and DT-treated (+DT, gray bars) immunized mice from three independent experiments. *Significant difference (p < 0.05, Tukey test). D and E, TCROVA+ CD8+ T cell frequency and absolute numbers per organ were determined by flow cytometry after multiparametric stainings of cell suspensions from nonimmunized Lang-DTR-EGFP mice (CTRL), or nontreated (−DT) or treated (+DT) immunized Lang-DTR-EGFP mice. D, FACS dot plots show the percentages of CD8+ T cells gated on CD45+ cells and TCR-OVA+ (Kb/SIINFEKL pentamer staining) in CD8+ T cells from vaginal tissue for a representative experiment. E, TCR-OVA+ CD8+ T cell frequency (left graph) and absolute numbers (right graph) were obtained from flow cytometry analysis and cell counting per organ. Bars represent mean values (+SEM) from three independent experiments. *Significant difference (p < 0.05, Tukey test). F, Cell suspensions (5 × 105) from CTRL and immunized mice were stimulated with PMA/ionomycin for 4 h and stained with IFN-γ, granzyme B, and CD8 Abs, as detailed in 1Materials and Methods. Dot plots show the expression of IFN-γ and Grz B gated on CD8 T cells from CTRL, nontreated (−DT), and DT-treated (+DT) immunized groups for a representative experiment. Bars correspond to the mean percentages (+SEM) of Grz B+IFN-γ+ in CD8+ T cells from two independent experiments with five pooled organs per experiment. Grz B, granzyme B.

Close modal

Altogether, these data suggest that vaginal LCs are endowed with Ag-specific regulatory activities on vaginal effector CD8 T cell responses.

We have examined the contribution of vaginal DCs in the generation of CTL responses after intravaginal application of a model Ag. We have identified in the vaginal epithelium a minor DC subset corresponding to epidermal LCs and a major DC subset that does not express langerin (Lang DCs). Our study demonstrates that Muc-DCs, including LCs and Lang DCs, are involved in the priming of Ag-specific CD8 T cells. Vaginal LCs differentiate CD8 T cells producing IL-10 and IL-17A, whereas vaginal Lang DCs prime IFN-γ–producing T cells. Moreover, not only are vaginal LCs dispensable for the generation of intravaginally induced CTL responses but also their in vivo ablation prior to immunization enhances vaginal Ag-specific CD8 T cells and CTL responses, suggesting they have a regulatory activity on these responses. This may be partly explained by their ability to prime CD8 T cells exhibiting a cytokine profile dominated by IL-17A and, to a lesser extent, IL-10.

Regarding the phenotype of vaginal DC subsets, our work confirms the heterogeneity of vaginal DCs, a feature reported first by Parr and Parr (11) and, more recently, by Iwasaki et al. (12, 13). Our observations further document the existence of two DC subsets in the mouse vaginal epithelium, including LCs and Lang DCs, which contrasts with the situation seen in the skin, in which all epidermal DCs are LCs (Figs. 2, 7; Refs. 1921). Vaginal LCs represent a minor yet homogeneous DC subset expressing levels of langerin comparable to those in skin LCs (Figs. 2, 7). This observation contrasts with a recent report by Iijima et al. (13), claiming that VEDCs do not express langerin or express it at very low levels. This discrepancy may be explained by the use of different monoclonal anti-langerin Abs, and of different enzymatic tissue dispersion procedures to prepare vaginal cells, which can affect cell recovery and expression of certain markers. Indeed, different isoforms of langerin are described in the mouse, which can explain different anti-langerin Ab reactivities (31). Moreover, using Lang-EGFP mice, we confirmed the presence of a minor subset of LCs in the vaginal epithelium (Fig. 2A). Furthermore, the observation that we identified higher LC percentages in Lang-EGFP mice than in C57BL/6 mice suggests that the anti-langerin Ab we used does not recognize all langerin isoforms. Moreover, vaginal LCs express brightly positive CD24, recently described as a specific LC marker in the epidermis (32). We also characterized a subset of Lang DCs in the vaginal epithelium, which constitutes a heterogeneous subset with bimodal expression of the macrophage marker F4/80, the CD205 receptor, and the CD24 Ag. This finding suggests that some of these DCs might represent LCs that do not express langerin, as proposed by Iijima et al. (13). The relationship between this subset and LCs remains to be established.

In the absence of progesterone treatment, a condition in which the vaginal epithelium is thick, we observed a slight increase in LC numbers in the vaginal epithelium, which may be explained by changes in the epithelial architecture. This finding suggests that modifications of hormonal status may influence the distribution and functions of vaginal DC subsets, which requires further investigation. Moreover, it is worth mentioning that Lang DCs and LCs are present in the epithelium of another pluristratified mucosa, the sublingual mucosa (30) (Supplemental Fig. 3). The presence of Lang and LCs in mucosal epithelia seems to be a common feature of pluristratified mucosae, contrary to the skin. This observation, along with the absence of Langerin+ DCs or migratory Langerin+ LCs in the vaginal lamina propria (Fig. 2), two DC subsets previously described in the skin dermis (16, 17), emphasizes the particularity of the vaginal mucosa. Altogether, our data and previous reports (1113) highlight the DC heterogeneity in mouse vaginal mucosa and the differences between pluristratified mucosae and skin.

We observed that our Ag applied onto a pluristratified epithelium (vagina and sublingual mucosa) is cross-presented to CD8 T cells by Muc-DCs, including Lang DCs and LCs, but not by resident DCs (cDCs and pDCs) (Figs. 4, 5). This observation is in accordance with a report (33) demonstrating that after vaginal infection with HSV-1, Muc-DCs were those DCs primarily involved in the presentation of viral Ags to CD4 and CD8 T cells. Furthermore, our study demonstrates that vagina-derived LCs present CD8 epitopes to specific CD8 T cells. This observation contrasts with recent findings showing that vagina-derived LCs were not able to present viral Ags to specific CD4 and CD8 T cells after vaginal HSV-1 infection (33). This may be explained by the fact that vaginal LCs have different abilities to take up and process nonreplicative Ags (CTB-OVA) and replicative Ags (HSV-1). Moreover, Lee et al. (33) monitored the Ag-presenting properties of vagina-derived DCs and LCs by measuring their ability to prime IFN-γ–producing CD4 and CD8 T cells. We can also hypothesize that in that study, vagina-derived LCs prime Ag-specific T cells with another cytokine profile (Fig. 6).

Our observation that LCs are competent in vivo to cross-present CD8 epitopes after Ag uptake in mucosal tissues is consistent with data showing that OVA applied onto skin or expressed by keratinocytes was efficiently cross-presented to CD8 T cells by LCs (34, 35). Regarding vagina-derived Lang DCs, we and others have demonstrated that this DC subset efficiently presents both CD4 and CD8 epitopes after intravaginal application of different Ags (Fig. 4; Refs. 12, 33). This last DC subset is also characterized by the expression of the integrin CD11b and the absence of the CD8α molecule. Its ability to cross-present CD8 epitopes from tissue-derived Ags contrasts with previous data demonstrating that resident splenic CD8+ DCs were equipped to cross-present protein Ag, but not CD11b+ DCs (36). This finding confirms that tissue-derived CD11b+ DCs are functionally different from splenic CD11b+ DCs. Furthermore, the strong priming potential toward IFN-γ–producing CD4 and CD8 T cells of Lang DCs, their strategic location in the epithelium in direct contact with the vaginal lumen, suggests that they might play a critical role in induction of protective immunity against vaginal infections by intracellular pathogens. Altogether, our data and other reports reinforce the notion that cross-presentation by different DC subsets is dictated by the site and the type of Ag delivery (injection or topical application), as well as the nature of the Ag (invasive versus noninvasive Ags).

In skin, conflicting data regarding the role of epidermal LCs in the promotion or control of immunity, in particular in the induction of contact hypersensitivity (1921), have been initially reported. These discrepancies can be explained by different parameters, including DT-depletion protocols, application sites of Ags (ear skin versus abdominal skin), doses, and nature of the Ags used. Indeed, under defined stimulatory conditions, skin LCs seem to contribute to productive immune responses but are dispensable (37). In the vaginal environment, vaginal LCs preferentially prime CD8 T cells with potential regulatory/suppressive functions, which is consistent with the increase in T CD8 cytotoxic activity in the vagina of LC-depleted mice (Figs. 6, 7). Indeed, vagina-derived LCs primed CD8 T cells toward a mixed cytokine profile, dominated by IL-17A and, to a lesser extent, by IL-10 (Fig. 6), two cytokines believed to be produced by suppressive CD8 T cells during viral infections (38, 39). Alternatively, IL-17 produced by Th17 cells has also been described as inhibiting CD8 T cell cytotoxicity (40). Further studies are needed to definitely establish the role of IL-17– and/or IL-10–producing T cells in the regulation of vaginally induced CTLs.

Following vaginal immunization with a CT-based immunogen, vaginal LCs prime IL-17–producing T cells, contrary to vaginal Lang DCs, which mainly differentiate IFN-γ–producing CD4 and CD8 T cells, two T cell profiles known to be induced by CT (Fig. 6). This finding indicates that these tissue-derived DCs differentially respond to the same microbial stimulus and may have different intrinsic properties. This is consistent with data demonstrating that human skin migratory LCs are more efficient than dermal DCs in inducing IL-17–producing CD4 T cells in an in vitro allogeneic system (41). Nevertheless, other mucosal DC subsets, such as intestinal CD11b+ DCs, are also noted to elicit IL-17 production (42). Further studies using different microbial agents are needed to establish whether this property is restricted to mucosal LCs present in pluristratified epithelia.

Our findings support the contention that vaginal LCs prime Ag-specific CD8 T cells, which dampen the generation of mucosal CTLs. This study demonstrates that vaginal LCs not only are dispensable for the generation of mucosal CTL responses but also exert a downregulating activity on these responses, by a mechanism that may involve IL-17 and, to a lesser extent, IL-10. This finding has implications for the rational design of vaccines and immunotherapeutic strategies based on tissue-specific targeting of DCs.

We thank Thierry Juhel and Selma Bekri for expert technical assistance.

Disclosures The authors have no financial conflicts of interest.

This work was supported by Institut National de la Santé et de la Recherche Médicale, Agence Nationale de Recherche sur le syndrome d’immunodéficience acquise (SIDA), Association Ensemble contre le SIDA (SIDACTION), and Association de Recherche sur le Cancer.

The online version of this article contains supplemental material.

Abbreviations used in this paper:

cDC

conventional blood-derived dendritic cell

CT

cholera toxin

CTB

B subunit of cholera toxin

CTB-OVA

CTB conjugated to OVA

CTRL

control

DC

dendritic cell

DT

diphtheria toxin

EGFP

enhanced GFP

Grz B

granzyme B

ILN

ileosacral lymph node

Lang DC

langerin DC

LC

Langerhans cell

MHCII

MHC class II

Muc-DC

mucosa-derived DC

non-DC APC

CD11cMHCII+ APC

pDC

plasmacytoid DC

SMLN

submandibular lymph node

VEDC

vaginal epithelial DC.

1
Kuttey
W. H.
,
Mestecky
J.
,
Wira
C. R.
.
2004
.
Mucosal immunity in the human female reproductive tract
. In
Mucosal Immunology
, 3rd Ed.
Mestecky
J.
,
Lamm
M. E.
,
Strober
W.
,
Bienenstock
J.
,
McGhee
J. R.
,
Mayer
L.
, eds.
Academic Press
,
San Diego, CA
,
p. 1631
1646
.
2
Parr
E. L.
,
Parr
M. B.
.
2004
.
Female genital tract infections and immunity in animal models.
In
Mucosal Immunology
, 3rd Ed.
Mestecky
J.
,
Lamm
M. E.
,
Strober
W.
,
Bienenstock
J.
,
McGhee
J. R.
,
Mayer
L.
, eds.
Academic Press
,
San Diego, CA
,
p. 1614
1630
.
3
Kaul
R.
,
Makadzange
T.
,
Rowland-Jones
S.
.
2000
.
AIDS in Africa: a disaster no longer waiting to happen.
Nat. Immunol.
1
:
267
270
.
4
Kaul
R.
,
Plummer
F. A.
,
Kimani
J.
,
Dong
T.
,
Kiama
P.
,
Rostron
T.
,
Njagi
E.
,
MacDonald
K. S.
,
Bwayo
J. J.
,
McMichael
A. J.
,
Rowland-Jones
S. L.
.
2000
.
HIV-1-specific mucosal CD8+ lymphocyte responses in the cervix of HIV-1-resistant prostitutes in Nairobi.
J. Immunol.
164
:
1602
1611
.
5
Rowland-Jones
S. L.
,
Pinheiro
S.
,
Kaul
R.
,
Hansasuta
P.
,
Gillespie
G.
,
Dong
T.
,
Plummer
F. A.
,
Bwayo
J. B.
,
Fidler
S.
,
Weber
J.
, et al
.
2001
.
How important is the ‘quality’ of the cytotoxic T lymphocyte (CTL) response in protection against HIV infection?
Immunol. Lett.
79
:
15
20
.
6
Johansson
E. L.
,
Rask
C.
,
Fredriksson
M.
,
Eriksson
K.
,
Czerkinsky
C.
,
Holmgren
J.
.
1998
.
Antibodies and antibody-secreting cells in the female genital tract after vaginal or intranasal immunization with cholera toxin B subunit or conjugates.
Infect. Immun.
66
:
514
520
.
7
Eriksson
K.
,
Quiding-Järbrink
M.
,
Osek
J.
,
Möller
A.
,
Björk
S.
,
Holmgren
J.
,
Czerkinsky
C.
.
1998
.
Specific-antibody-secreting cells in the rectums and genital tracts of nonhuman primates following vaccination.
Infect. Immun.
66
:
5889
5896
.
8
Kato
H.
,
Bukawa
H.
,
Hagiwara
E.
,
Xin
K. Q.
,
Hamajima
K.
,
Kawamoto
S.
,
Sugiyama
M.
,
Sugiyama
M.
,
Noda
E.
,
Nishizaki
M.
,
Okuda
K.
.
2000
.
Rectal and vaginal immunization with a macromolecular multicomponent peptide vaccine candidate for HIV-1 infection induces HIV-specific protective immune responses.
Vaccine
18
:
1151
1160
.
9
Luci
C.
,
Hervouet
C.
,
Rousseau
D.
,
Holmgren
J.
,
Czerkinsky
C.
,
Anjuère
F.
.
2006
.
Dendritic cell-mediated induction of mucosal cytotoxic responses following intravaginal immunization with the nontoxic B subunit of cholera toxin.
J. Immunol.
176
:
2749
2757
.
10
Romani
N.
,
Holzmann
S.
,
Tripp
C. H.
,
Koch
F.
,
Stoitzner
P.
.
2003
.
Langerhans cells—dendritic cells of the epidermis.
APMIS
111
:
725
740
.
11
Parr
M. B.
,
Parr
E. L.
.
1991
.
Langerhans cells and T lymphocyte subsets in the murine vagina and cervix.
Biol. Reprod.
44
:
491
498
.
12
Zhao
X.
,
Deak
E.
,
Soderberg
K.
,
Linehan
M.
,
Spezzano
D.
,
Zhu
J.
,
Knipe
D. M.
,
Iwasaki
A.
.
2003
.
Vaginal submucosal dendritic cells, but not Langerhans cells, induce protective Th1 responses to herpes simplex virus-2.
J. Exp. Med.
197
:
153
162
.
13
Iijima
N.
,
Linehan
M. M.
,
Saeland
S.
,
Iwasaki
A.
.
2007
.
Vaginal epithelial dendritic cells renew from bone marrow precursors.
Proc. Natl. Acad. Sci. USA
104
:
19061
19066
.
14
Valladeau
J.
,
Ravel
O.
,
Dezutter-Dambuyant
C.
,
Moore
K.
,
Kleijmeer
M.
,
Liu
Y.
,
Duvert-Frances
V.
,
Vincent
C.
,
Schmitt
D.
,
Davoust
J.
, et al
.
2000
.
Langerin, a novel C-type lectin specific to Langerhans cells, is an endocytic receptor that induces the formation of Birbeck granules.
Immunity
12
:
71
81
.
15
Valladeau
J.
,
Clair-Moninot
V.
,
Dezutter-Dambuyant
C.
,
Pin
J. J.
,
Kissenpfennig
A.
,
Mattéi
M. G.
,
Ait-Yahia
S.
,
Bates
E. E.
,
Malissen
B.
,
Koch
F.
, et al
.
2002
.
Identification of mouse langerin/CD207 in Langerhans cells and some dendritic cells of lymphoid tissues.
J. Immunol.
168
:
782
792
.
16
Poulin
L. F.
,
Henri
S.
,
de Bovis
B.
,
Devilard
E.
,
Kissenpfennig
A.
,
Malissen
B.
.
2007
.
The dermis contains langerin+ dendritic cells that develop and function independently of epidermal Langerhans cells.
J. Exp. Med.
204
:
3119
3131
.
17
Ginhoux
F.
,
Collin
M. P.
,
Bogunovic
M.
,
Abel
M.
,
Leboeuf
M.
,
Helft
J.
,
Ochando
J.
,
Kissenpfennig
A.
,
Malissen
B.
,
Grisotto
M.
, et al
.
2007
.
Blood-derived dermal langerin+ dendritic cells survey the skin in the steady state.
J. Exp. Med.
204
:
3133
3146
.
18
Bursch
L. S.
,
Wang
L.
,
Igyarto
B.
,
Kissenpfennig
A.
,
Malissen
B.
,
Kaplan
D. H.
,
Hogquist
K. A.
.
2007
.
Identification of a novel population of Langerin+ dendritic cells.
J. Exp. Med.
204
:
3147
3156
.
19
Kissenpfennig
A.
,
Henri
S.
,
Dubois
B.
,
Laplace-Builhé
C.
,
Perrin
P.
,
Romani
N.
,
Tripp
C. H.
,
Douillard
P.
,
Leserman
L.
,
Kaiserlian
D.
, et al
.
2005
.
Dynamics and function of Langerhans cells in vivo: dermal dendritic cells colonize lymph node areas distinct from slower migrating Langerhans cells.
Immunity
22
:
643
654
.
20
Bennett
C. L.
,
van Rijn
E.
,
Jung
S.
,
Inaba
K.
,
Steinman
R. M.
,
Kapsenberg
M. L.
,
Clausen
B. E.
.
2005
.
Inducible ablation of mouse Langerhans cells diminishes but fails to abrogate contact hypersensitivity.
J. Cell Biol.
169
:
569
576
.
21
Kaplan
D. H.
,
Jenison
M. C.
,
Saeland
S.
,
Shlomchik
W. D.
,
Shlomchik
M. J.
.
2005
.
Epidermal langerhans cell-deficient mice develop enhanced contact hypersensitivity.
Immunity
23
:
611
620
.
22
Rask
C.
,
Holmgren
J.
,
Fredriksson
M.
,
Lindblad
M.
,
Nordström
I.
,
Sun
J. B.
,
Czerkinsky
C.
.
2000
.
Prolonged oral treatment with low doses of allergen conjugated to cholera toxin B subunit suppresses immunoglobulin E antibody responses in sensitized mice.
Clin. Exp. Allergy
30
:
1024
1032
.
23
Svennerholm
A.
,
Lange
S.
,
Holmgren
J.
.
1978
.
Correlation between intestinal synthesis of specific immunoglobulin A and protection against experimental cholera in mice.
Infect. Immun.
21
:
1
6
.
24
Coles
R. M.
,
Mueller
S. N.
,
Heath
W. R.
,
Carbone
F. R.
,
Brooks
A. G.
.
2002
.
Progression of armed CTL from draining lymph node to spleen shortly after localized infection with herpes simplex virus 1.
J. Immunol.
168
:
834
838
.
25
Anjuere
F.
,
Ardavin
C.
.
2001
.
Isolation of mouse thymic dendritic cells
. In
Methods in Molecular Medicine: Dendritic Cell Protocols
, Chapter 3,
Vol. 64
.
Robinson
S. P.
,
Stagg
A. J.
, eds.
Humana Press
,
Totowa, NJ
,
p. 23
28
.
26
Anjuère
F.
,
Luci
C.
,
Lebens
M.
,
Rousseau
D.
,
Hervouet
C.
,
Milon
G.
,
Holmgren
J.
,
Ardavin
C.
,
Czerkinsky
C.
.
2004
.
In vivo adjuvant-induced mobilization and maturation of gut dendritic cells after oral administration of cholera toxin.
J. Immunol.
173
:
5103
5111
.
27
Metlay
J. P.
,
Witmer-Pack
M. D.
,
Agger
R.
,
Crowley
M. T.
,
Lawless
D.
,
Steinman
R. M.
.
1990
.
The distinct leukocyte integrins of mouse spleen dendritic cells as identified with new hamster monoclonal antibodies.
J. Exp. Med.
171
:
1753
1771
.
28
Anjuère
F.
,
Martín
P.
,
Ferrero
I.
,
Fraga
M. L.
,
del Hoyo
G. M.
,
Wright
N.
,
Ardavín
C.
.
1999
.
Definition of dendritic cell subpopulations present in the spleen, Peyer’s patches, lymph nodes, and skin of the mouse.
Blood
93
:
590
598
.
29
Henri
S.
,
Vremec
D.
,
Kamath
A.
,
Waithman
J.
,
Williams
S.
,
Benoist
C.
,
Burnham
K.
,
Saeland
S.
,
Handman
E.
,
Shortman
K.
.
2001
.
The dendritic cell populations of mouse lymph nodes.
J. Immunol.
167
:
741
748
.
30
Cuburu
N.
,
Kweon
M. N.
,
Song
J. H.
,
Hervouet
C.
,
Luci
C.
,
Sun
J. B.
,
Hofman
P.
,
Holmgren
J.
,
Anjuère
F.
,
Czerkinsky
C.
.
2007
.
Sublingual immunization induces broad-based systemic and mucosal immune responses in mice.
Vaccine
25
:
8598
8610
.
31
Riedl
E.
,
Tada
Y.
,
Udey
M. C.
.
2004
.
Identification and characterization of an alternatively spliced isoform of mouse Langerin/CD207.
J. Invest. Dermatol.
123
:
78
86
.
32
Stutte
S.
,
Jux
B.
,
Esser
C.
,
Förster
I.
.
2008
.
CD24a expression levels discriminate Langerhans cells from dermal dendritic cells in murine skin and lymph nodes.
J. Invest. Dermatol.
128
:
1470
1475
.
33
Lee
H. K.
,
Zamora
M.
,
Linehan
M. M.
,
Iijima
N.
,
Gonzalez
D.
,
Haberman
A.
,
Iwasaki
A.
.
2009
.
Differential roles of migratory and resident DCs in T cell priming after mucosal or skin HSV-1 infection.
J. Exp. Med.
206
:
359
370
.
34
Stoitzner
P.
,
Tripp
C. H.
,
Eberhart
A.
,
Price
K. M.
,
Jung
J. Y.
,
Bursch
L.
,
Ronchese
F.
,
Romani
N.
.
2006
.
Langerhans cells cross-present antigen derived from skin.
Proc. Natl. Acad. Sci. USA
103
:
7783
7788
.
35
Mayerova
D.
,
Parke
E. A.
,
Bursch
L. S.
,
Odumade
O. A.
,
Hogquist
K. A.
.
2004
.
Langerhans cells activate naive self-antigen-specific CD8 T cells in the steady state.
Immunity
21
:
391
400
.
36
Schnorrer
P.
,
Behrens
G. M.
,
Wilson
N. S.
,
Pooley
J. L.
,
Smith
C. M.
,
El-Sukkari
D.
,
Davey
G.
,
Kupresanin
F.
,
Li
M.
,
Maraskovsky
E.
, et al
.
2006
.
The dominant role of CD8+ dendritic cells in cross-presentation is not dictated by antigen capture.
Proc. Natl. Acad. Sci. USA
103
:
10729
10734
.
37
Wang
L.
,
Bursch
L. S.
,
Kissenpfennig
A.
,
Malissen
B.
,
Jameson
S. C.
,
Hogquist
K. A.
.
2008
.
Langerin expressing cells promote skin immune responses under defined conditions.
J. Immunol.
180
:
4722
4727
.
38
Elrefaei
M.
,
Ventura
F. L.
,
Baker
C. A.
,
Clark
R.
,
Bangsberg
D. R.
,
Cao
H.
.
2007
.
HIV-specific IL-10-positive CD8+ T cells suppress cytolysis and IL-2 production by CD8+ T cells.
J. Immunol.
178
:
3265
3271
.
39
Intlekofer
A. M.
,
Banerjee
A.
,
Takemoto
N.
,
Gordon
S. M.
,
Dejong
C. S.
,
Shin
H.
,
Hunter
C. A.
,
Wherry
E. J.
,
Lindsten
T.
,
Reiner
S. L.
.
2008
.
Anomalous type 17 response to viral infection by CD8+ T cells lacking T-bet and eomesodermin.
Science
321
:
408
411
.
40
Hou
W.
,
Kang
H. S.
,
Kim
B. S.
.
2009
.
Th17 cells enhance viral persistence and inhibit T cell cytotoxicity in a model of chronic virus infection.
J. Exp. Med.
206
:
313
328
.
41
Mathers
A. R.
,
Janelsins
B. M.
,
Rubin
J. P.
,
Tkacheva
O. A.
,
Shufesky
W. J.
,
Watkins
S. C.
,
Morelli
A. E.
,
Larregina
A. T.
.
2009
.
Differential capability of human cutaneous dendritic cell subsets to initiate Th17 responses.
J. Immunol.
182
:
921
933
.
42
Denning
T. L.
,
Wang
Y. C.
,
Patel
S. R.
,
Williams
I. R.
,
Pulendran
B.
.
2007
.
Lamina propria macrophages and dendritic cells differentially induce regulatory and interleukin 17-producing T cell responses.
Nat. Immunol.
8
:
1086
1094
.