Requirement of CCL17 for CCR7- and CXCR4-dependent migration of cutaneous dendritic cells

Abstract

Chemokines are known to regulate the steady-state and inflammatory migration of cutaneous dendritic cells (DCs). The beta-chemokine CCL17, a ligand of CCR4, is inducibly expressed in a subset of DCs and is strongly up-regulated in atopic diseases. Using an atopic dermatitis model, we show that CCL17-deficient mice develop acanthosis as WT mice, whereas dermal inflammation, T helper 2-type cytokine production, and the allergen-specific humoral immune response are significantly decreased. Notably, CCL17-deficient mice retained Langerhans cells (LCs) in the lesional skin after chronic allergen exposure, whereas most LCs emigrated from the epidermis of allergen-treated WT controls into draining lymph nodes (LNs). Moreover, CCL17-deficient LCs showed impaired emigration from the skin after exposure to a contact sensitizer. In contrast, the absence of CCR4 had no effect on cutaneous DC migration and development of atopic dermatitis symptoms. As an explanation for the major migratory defect of CCL17-deficient DCs in vivo, we demonstrate impaired mobility of CCL17-deficient DCs to CCL19/21 in 3D in vitro migration assays and a blockade of intracellular calcium release in response to CCR7 ligands. In addition, responsiveness of CCL17-deficient DCs to CXCL12 was impaired as well. We demonstrate that the inducible chemokine CCL17 sensitizes DCs for CCR7- and CXCR4-dependent migration to LN-associated homeostatic chemokines under inflammatory conditions and thus plays an important role in cutaneous DC migration.

Fig. 1.

Impaired migration of DCs from lesional skin in the absence of CCL17. CCL17E/+ (E/+) and CCL17E/E (E/E) mice (A and B), or WT and CCR4KO mice (C) were treated with OVA or NaCl only. At 24 h after the last treatment, the lesional skin was isolated and analyzed by immunofluorescence staining. (A) MHCII⁺ cells (red) are localized in the epidermis (marked with a white “E”) and dermis (marked with a white “D”). The demarcation of the epidermis and dermis is indicated as a white line. (Scale bar: 100 μm.) (B and C) MHCII⁺ epidermal cells were counted in between six and eight independent fields per section, and mean number ± SEM per mm of skin was calculated. CCL17E/+ mice and WT are denoted by black bars; CCL17E/E and CCR4KO mice, by gray bars. The data are representative of two or three independent experiments with between five and seven mice per group.

Fig. 2.

Emigration of DCs after contact sensitization requires CCL17. (A) CCL17E/+ or CCL17E/E mice were treated with 0.5% DNFB on the ears for 4 h. After culture for 2 days, the skin layers were separated, and the epidermis was stained for Langerin. (Scale bar: 100 μm.) (B and C) Quantification of the number of LCs per mm² of epidermis of CCL17E/+ and CCL17E/E mice (B), and WT and CCR4KO mice (C) for the different treatment groups as indicated. (D and E) 0.5% DNFB was applied on the shaved back of CCL17E/+ and CCL17E/E mice. After 2 and 4 days, the numbers of immigrated dDCs (E) and LCs (D) in skin-draining LNs were compared with steady-state levels in untreated mice (0). CCL17E/+ and WT mice are denoted by black bars; CCL17E/E and CCR4KO mice, by gray bars. Mean values ± SEM of two or three independent experiments are shown.

Fig. 3.

Migration toward CCR7 and CXCR4 ligands is impaired in the absence of CCL17. (A–C) Chemotactic migration in 3D collagen matrices of BM-DCs from CCL17E/+ and CCL17E/E mice in response to CCL19 or medium only. Plots of 60 cells/sample tracked by live cell imaging from one representative experiment out of three are depicted. Quantifications of directionality (A), velocity (B), and y-forward migration index (C) represent mean values ± SEM from all three independent experiments. (D) BM-DCs were stimulated on day 6 with LPS overnight and transferred into transwell inserts. Cells were incubated with CCL19, CXCL12, or CCL21 as indicated, and migrated DCs were counted. Results of three independent experiments were pooled, and mean values ± SEM were calculated. (E) BM-DCs were stimulated on day 6 with 100 ng/mL of LPS overnight and with 1,000 ng/mL of CCL17 at different time points. Cells were incubated with CCL19, and migrated DCs were counted. (F) BM-DCs were stimulated on day 6 with 100 ng/mL of LPS and graded doses of recCCL17 overnight as indicated. Cells were incubated with CCL19. (G) BM-DCs were stimulated on day 6 with 100 ng/mL of LPS and 1,000 ng/mL of recCCL17 overnight. Cells were stimulated with CXCL12. CCL17E/+ mice are denoted by black bars; CCL17E/E mice, by gray bars.

Fig. 4.

Normal expression of CCR7 and CXCR4, but impaired Ca²⁺ flux, in CCL17E/E DCs. Unstimulated and LPS-stimulated BM-DCs of CCL17E/E (green line) and CCL17E/+ (red line) mice were stained with anti-CCR7 (A, Left), anti-CXCR4 (A, Right), or anti-CCR4 (B). As a negative control, splenic cells of CCR4KO mice were stained with the CCR4 antibody. The negative control for CCR7 and CXCR4 was unstained cells. The data are representative of three independent experiments. (C and D) LCs were isolated from the dorsal skin of C57BL/6 mice (black line), CCR4KO mice (blue dotted line), CCL17E/+ mice (green line), and CCL17E/E mice (red line) and stained for MHCII and CCR7 either immediately (Left) or after stimulation with LPS overnight (Right). The histograms in D depict CCR7 expression of gated MHCII⁺ cells as indicated in C. (E–H) LPS-stimulated BM-DCs from CCL17E/+ and CCL17E/E mice were loaded with Indo-1 AM, and CCL19-induced Ca²⁺-flux was analyzed by calculating the ratio of calcium-bound to free Indo-1 over time. Shown is the baseline ratio (0–50 s), followed by stimulation with CCL19 or ionomycin (50–200 s). (I–N) Before staining with Indo-1 AM, CD11c⁺ BM-DCs from CCL17E/E, CCL17E/+, and WT mice were enriched by MACS on day 6 and stimulated with LPS overnight. CXCL12-induced Ca²⁺ flux was calculated as described above (I, K, and M). As a control, DCs were treated with ionomycin (J, L, and N).