Stromal cells control the epithelial residence of DCs and memory T cells by regulated activation of TGF-β

Abstract

Cells of the immune system that reside in barrier epithelia provide a first line of defense against pathogens. Langerhans cells (LCs) and CD8(+) tissue-resident memory T cells (TRM cells) require active transforming growth factor-β1 (TGF-β) for epidermal residence. Here we found that integrins αvβ6 and αvβ8 were expressed in non-overlapping patterns by keratinocytes (KCs) and maintained the epidermal residence of LCs and TRM cells by activating latent TGF-β. Similarly, the residence of dendritic cells and TRM cells in the small intestine epithelium also required αvβ6. Treatment of the skin with ultraviolet irradiation decreased integrin expression on KCs and reduced the availability of active TGF-β, which resulted in LC migration. Our data demonstrated that regulated activation of TGF-β by stromal cells was able to directly control epithelial residence of cells of the immune system through a novel mechanism of intercellular communication.

Figure 1

Activation of latent TGF-β by α_vβ₆ inhibits homeostatic LC migration. (a) Quantification of LCs in archival skin specimens from untreated (control) and losartan-treated patients, detected by immunohistochemistry with antibody to CD1a (anti-CD1a) or anti-langerin. Each symbol represents an individual patient; horizontal lines indicate the average. (b) Microscopy of whole mounts of back epidermis from cohorts (n = 10 mice in each) of wild-type mice (WT), TGF-βRI–CA^LC mice (with inducible expression of constitutively active TGF-βRI in LCs) and TGF-β^LC mice (with inducible ablation of TGF-β in LCs) 9 d after the start of tamoxifen treatment, as well as epidermis from adult Itgb6^−/− mice, all stained for MHC class II (green). (c) Ratio of the number of LCs in the epidermis (Epi) (obtained by counting of the MHCII⁺ cells in b) or skin draining LNs (LN) (obtained by flow cytometry) of Itgb6^−/−, TGF-βRI–CA^LC and TGF-β^LC mice to that in their wild-type counterparts. Each symbol represents an individual mouse (with results from the same mouse joined by a solid line); dashed horizontal lines indicate a ratio of 1.0. (d) Microscopy of whole mounts of ear epidermis from tamoxifen-treated wild-type and TGF-βRI–CA^LC mice given intradermal injection of anti-α_vβ₆ or isotype-matched control antibody (Isotype) (n = 4 mice per genotype per group), stained for MHC class II (green). (e) Quantification of LCs per high-power field (HPF) in the mice in d. Each symbol represents LCs per HPF; small horizontal lines indicate the average. Scale bars (b,d), 100 μm. NS, not significant (P > 0.05); *P < 0.01 and **P < 0.0001 (two-tailed unpaired Student’s t test (a,c) or Tukey’s multiple comparisons test (e)). Data are representative of experiments with n = 42 total donors (a) or are representative of (b,d) or pooled from (c,e) three independent experiments.

Figure 2

Epidermal residence of LCs requires expression of α_vβ₆ by IFE epidermal KCs. (a) Microscopy of whole mounts of back epidermis from wild-type and Itgb6^−/− mice, stained for MHC class II (red) and α_vβ₆ (green); arrowheads indicate bulge region of telogen hair; asterisks indicate IFE region. (b) Flow cytometric analysis of the expression of α_vβ₆ by IM, IFE and bulge KCs (identified as in Supplementary Fig. 3b) from the back skin of wild-type and Itgb6^−/− mice (key). (c) Quantification of LCs in transverse sections of back skin from wild-type and Itgb6^−/− mice, identified by colocalization of langerin and MHC class II within the IFE region or IM. (d) Distance of IFE LCs from the hair follicle (HF) in sections as in c. (e) Quantitative RT-PCR analysis of Itgb6 mRNA in sorted epidermal populations of dendritic epidermal T cells (DETC), LCs, and IFE and IM KCs in wild-type mice; results are presented relative to those of the control gene Hprt. (f) Frequency of host- and donor-derived epidermal LCs in lethally irradiated wild-type→wild-type (WT→WT) or wild-type→Itgb6^−/− (WT→Itgb6^−/−) (CD45.2⁺) chimeras 6–8 weeks after transfer of BM from wild-type (CD45.1⁺) donors, analyzed by flow cytometry. (g) Microscopy of whole mounts of back epidermis from chimeras as in f, stained for MHC class II (green); autofluorescence shows hair shafts. Scale bars (a,g), 100 μm. ND, not detected. Each symbol (c,d,f) represents LCs per HPF (c), distance from HF (d) or an individual mouse (f); small horizontal lines indicate the mean. *P < 0.01 and **P < 0.0001 (two-tailed unpaired Student’s t test). Data are representative of three independent experiments with eight mice per genotype (a,b,g) or six independent experiments (e; mean + s.e.m.) or are pooled from two independent experiments with two mice per group (c,d) or three independent experiments (f).

Figure 3

LC residence is controlled by spatially distinct expression of α_vβ₆ and α_vβ₆ on KC subsets through activation of latent-TGF-β. (a) Quantitative RT-PCR analysis of Itgb8 mRNA in sorted epidermal populations from wild-type mice (as in Fig. 2e). (b) Quantification of LCs in skin-draining LNs of wild-type and Itgb8^ΔKC mice. (c) Quantification of LCs (per HPF) in IM of wild-type and Itgb8^ΔKC mice (as in Fig. 2c). (d) Microscopy of whole mounts of tail epidermis (top) and transverse sections of back skin (bottom) from wild-type and Itgb8^ΔKC mice, stained for langerin (red) and with the DNA-binding dye DAPI (blue); dashed lines, dermal-epidermal junction; SG, sebaceous gland; asterisks (bottom), IM. (e) Quantification of LCs in the IFE region and IM of the stained transverse skin sections in d. (f) Microscopy of whole mounts of back epidermis from wild-type, Itgb6^−/− and Itgb6^−/−Itgb8^ΔKC mice, stained for MHC class II (green); autofluorescence shows hair shafts. (g) Quantification of LCs in f. (h) Luciferase activity in mink lung TGF-β reporter cells cultured together with wild-type or Itgb8^ΔKC primary KCs pre-incubated with anti-α_vβ₆ or isotype-matched control antibody; results are presented as relative light units (RLU), normalized to those of wild-type control cells. Each symbol (b,c,e,g,h) represents an individual mouse (b,c,g), LCs per HPF (e) or RLU values (f); small horizontal lines indicate the mean (±s.e.m. in h). Scale bars, 50 mu;m (d) or 100 mu;m (f). *P < 0.05 and **P < 0.0001 (two-tailed unpaired Student’s t test (a–c,e) or Tukey’s multiple-comparisons test (g,h)). Data are representative of six independent experiments (a; mean + s.e.m) or three independent experiments (d,f), or are pooled from three independent experiments (b,c,e,g) with n = 5 mice per group (d,e) or three to six independent experiments (h).

Figure 4

UV irradiation promotes LC migration through diminished integrin expression and TGF-β activation. (a) Microscopy of whole mounts of back epidermis from tamoxifen-treated wild-type and TGF-βRI–CA^LC mice (n = 5 per group) 4 d after sham treatment (-UV) or UVB irradiation (+UV), stained for MHC class II (green) and langerin (red). Scale bars, 100 mu;m. (b) Quantification of LCs (per HPF) in a. (c) Flow cytometry analyzing the expression of CD86 and CCR7 by epidermal LCs from mice as in a, gated as CD45⁺MHCII⁺CD11b^int langerin-positive cells. Numbers in quadrants indicate percent cells in each. (d) Quantitative RT-PCR analysis of Itgb6 mRNA (left vertical axis) and Itgb8 mRNA (right vertical axis) in flow cytometry–sorted populations of IFE and IM KCs from wild-type mice 18 h after sham treatment or UVB irradiation (presented as in Fig. 2e). (e,f) Quantitative RT-PCR analysis of Itgb6 mRNA (e) and Itgb8 mRNA (f) in mouse primary KC cultures 24 h after sham treatment or UVB irradiation (presented as in Fig. 2e). (g) Luciferase activity of mink lung TGF-β reporter cells cultured together with primary keratinocytes given sham treatment or UVB irradiation; results are normalized to those of sham-treated cells. Each symbol (b,d–g) represents LCs per HPF (b), an independent experiment (d–f) or RLU values (g); small horizontal lines indicate the mean (±s.e.m. in d–g). *P < 0.01, **P < 0.001 and ***P < 0.0001 (two-tailed Mann-Whitney test (b,f) or two-tailed unpaired t test (d,e,g)). Data are representative of three independent experiments (a,c) or are pooled from three independent experiments with five mice per group (b) or six to seven independent experiments (d–g).

Figure 5

α_vβ₆ and α_vβ₈ are required for the residence of CD8⁺ T_RM cells in epidermis. (a) Microscopy of whole mounts of back epidermis from wild-type, Itgb6^−/−, Itgb6^−/−Itgb8^ΔKC and LC-deficient (ΔLC) mice given injection of Thy-1.1⁺ P14 cells on day -1 and infected with LCMV on day 0, followed by topical treatment with DNFB on day 3 (for epidermal seeding of P14 T cells), stained for Thy-1.1 (red) and MHC class II (green) on day 42 after infection. (b) Quantification of Thy-1.1⁺ cells in a. (c) Microscopy of whole mounts of back epidermis from wild-type, Itgb6^−/− and Itgb6^−/−Itgb8^ΔKC treated as in a, stained for Thy-1.1 (red) and MHC class II (green) on day 7 after infection. (d) Quantification of Thy-1.1⁺ cells in c. (e) Frequency of CD103⁺Thy-1.1⁺ cells in epidermis of the mice in c, assessed by flow cytometry on day 7 after infection. (f) Microscopy of whole mounts of back epidermis from wild-type mice treated intraperitoneally with anti-α_vβ₆ or isotype-matched control antibody once a week for 4 weeks starting at 56 d after infection with LCMV and epidermal seeding of T_RM cells as in a, stained for Thy-1.1 (red) and MHC class II (green). (g) Quantification of Thy-1.1⁺ cells in f. Autofluorescent hair follicles are visible in a,c,f. Scale bars (a,c,f), 100 μm. *P < 0.05, **P < 0.01 and ***P < 0.0001 (Tukey’s multiple-comparisons test (b,e) or two-tailed Mann-Whitney test (g)). Data are representative of two to three independent experiments with n = 4 mice per group (a,b; mean + s.e.m. in b) or two independent experiments with n = 4–5 mice per group (c–g; mean + s.e.m. in d,e,g).

Figure 6

α_vβ₆ is required for residence of T_RM cells in intestinal epithelium. (a) Microscopy of small intestine from wild-type and Itgb6^−/− mice 42 d after infection with LCMV, stained for Thy-1.1 (red) and collagen IV (ColIV) (green) and with DAPI (blue); arrowheads indicate T_RM cells in the epithelial layer. (b,c) Quantification of Thy-1.1⁺ cells (per 10⁶ nucleated cells) in the epithelial layer (above collagen IV (arrowheads) in a) (b) and lamina propria (LP) (below collagen IV in a) (c) of the small intestine (SI) of mice as in a. (d) Microscopy of small intestine from wild-type and Itgb6^−/− mice 7 d after infection with LCMV, stained as in a. (e,f) Quantification of Thy-1.1⁺ cells in the epithelial layer (as in b) (e) and lamina propria (as in c) (f) of the small intestine of mice as in d. (g) Microscopy of small intestine from wild-type mice infected with LCMV and then treated intraperitoneally with anti-α_vβ₆ or isotype-matched control antibody starting at 56 d after infection, stained as in a. (h,i) Quantification of Thy-1.1⁺ cells in the epithelial layer (as in b) (h) and lamina propria (as in c) (i) of the small intestine of mice as in g. Scale bars (a,d,g), 50 μm. *P < 0.05 (two-tailed Mann-Whitney test). Data representative of three independent experiments with n = 4 mice per group (a–c; mean + s.e.m. in b,c) or two independent experiments with n = 4–5 mice per group (d–i; mean + s.e.m. in e,f,h,i).