Commensal Microbes and Hair Follicle Morphogenesis Coordinately Drive Treg Migration into Neonatal Skin

Abstract

Regulatory T cells (Tregs) are required to establish immune tolerance to commensal microbes. Tregs accumulate abruptly in the skin during a defined window of postnatal tissue development. However, the mechanisms mediating Treg migration to neonatal skin are unknown. Here we show that hair follicle (HF) development facilitates the accumulation of Tregs in neonatal skin and that upon skin entry these cells localize to HFs, a primary reservoir for skin commensals. Further, germ-free neonates had reduced skin Tregs indicating that commensal microbes augment Treg accumulation. We identified Ccl20 as a HF-derived, microbiota-dependent chemokine and found its receptor, Ccr6, to be preferentially expressed by Tregs in neonatal skin. The Ccl20-Ccr6 pathway mediated Treg migration in vitro and in vivo. Thus, HF morphogenesis, commensal microbe colonization, and local chemokine production work in concert to recruit Tregs into neonatal skin, thereby establishing this tissue Treg niche early in life.

Keywords: commensal-specific immune tolerance; hair follicles; migration; regulatory T cells; skin commensal microbes; tissue development.

Figure 1. Treg Cells Localize to HFs in Neonatal Skin and Are Reduced in the Absence of HF Morphogenesis

(A) Immunohistochemical staining for GFP in skin from 13-day-old (D13) FoxP3^GFP mice. Dashed outline, hair follicle; In, infundibulum, Is, isthmus. Scale bar, 20 µm. Inset: magnified image demonstrating GFP⁺ cell morphology and size. (B) Pregnant K5⁺ females bred to K5/Dkk1 males were fed doxycycline starting at gestational date E13.5, preventing wnt-dependent hair follicle development in K5/Dkk1 but not single-transgenic (Ctl.) littermates. (C) Representative histology of skin from D13 K5/Dkk1 and Ctl. littermates, taken from site-matched areas on the scalp. (D) Flow cytometry plots of CD4⁺ T cells in D13 K5/Dkk1 versus Ctl. skin. (E and F) Percentage of Treg cells of CD4⁺ cells in D13 K5/Dkk1 versus Ctl. skin (E), as well as in gut lamina propria (LP) and skin-draining lymph nodes (SDLN) (F). Data were pooled from three replicate experiments on independent litters, each with ≥3 pups total.

Figure 2. Commensal Microbes Mediate Accumulation of Treg Cells in Neonatal Skin

(A) Representative flow cytometry plots of CD4⁺ T cells from skin of D13 germ-free (GF) versus specific-pathogen-free (SPF) mice. (B and C) Percentages (B) and absolute numbers (C) of Treg cells in skin of D13 GF versus SPF mice. (D–G) Absolute numbers of FoxP3neg CD4⁺ effector T cells (Teff) (D), CD8⁺T cells (E), dermal γδ T cells (F), and dendritic epidermal T cells (DETC) (G) in skin of D13 GF versus SPF mice. Data were pooled from two of three replicate experiments with ≥3 mice per group.

Figure 3. Discovery Approach to Identify Chemokine-Chemokine Receptor Pathways Responsible for Treg Cell Migration to Neonatal Skin

(A) Quantitative RT-PCR chemokine array was performed on RNA isolated from whole skin of three D6 and five D13 SPF mice. Volcano plot depicting log2FC versus log10 (p value) for all candidates tested. (B) Log2FC for those chemokines significantly increased in D13 versus D6 skin. (C) Identical qRT-PCR was performed on five D13 SPF versus three GF skin. Volcano plot depicting log2FC versus log10(p value) for all candidates tested. (D) Log2FC in D13 SPF versus GF skin for chemokine subset also increased in D13 versus D6 skin. (E) Treg cells were sorted from skin and LN from D13 SPF FoxP3^DTR mice and subjected to RNA sequencing. Average raw gene counts in D13 skin Treg cells are shown for receptors corresponding to chemokines increased in D13 skin. (F) Log10(p value) (corrected for multiple comparisons) of log2FC expression of receptors by skin versus SDLN Treg cells. Blue bar: p ≤ 0.05. See also Figure S1.

Figure 4. Commensal Microbes Augment Ccl20 Expression in Developing HFs

(A and B) Fold change expression of Ccl20 versus β 2-microglobulin mRNA transcripts in skin from D6 SPF, D13 SPF, and D13 GF mice (A) and D13 K5/Dkk1 versus Ctl littermates (B), pooled from three litters. (C–E) Representative images of in situ hybridization (ISH) for Ccl20 mRNA transcript in the skin of D13 SPF (C), D6 SPF (D), and D13 GF (E) mice. Arrows, visualization of Ccl20 mRNA signal; *, hair follicle infundibulum. Scale bars, 60 µm. (F) Quantification of Ccl20 ISH staining density (number of positive cells per 20× field). (G) Fold change expression in Ccl20 (normalized to EIF3L) in human fetal skin after overnight exposure to skin commensal bacteria or bacterial surface molecules as compared with PBS-treated controls. LPS, lipopolysaccharide; LTA, lipoteichoic acid; PGN, peptidoglycan; S. epi, Staphylococcus epidermidis; P. acnes, Propionibacterium acnes; C. accolens, Corynebacterium accolens. All data are representative of two independent experiments. See also Figure S2.

Figure 5. Neonatal Skin Treg Cells and a Subset of Thymic Treg Cells Preferentially Express Ccr6

(A) Flow cytometry plot depicting Ccr6 expression on CD4⁺ T cells in neonatal skin. (B) Histogram depicting relative mean fluorescent intensity (MFI) of Ccr6 by flow cytometry on Treg cells, CD4⁺ Teff cells, and CD8⁺ T cells from neonatal skin, as well as WT and Ccr6^−/− Treg cells in SDLN. (C) Average Ccr6 MFI on these same populations. (D) Histogram depicting MFI of Ccr6 on neonatal thymic Treg cells (CD3⁺ CD4⁺ CD8^neg FoxP3⁺). (E) Flow cytometry plot depicting Ccr6 and CD25 co-staining by thymic Treg cells. (F) Average MFI of Ccr6 on thymic Treg cells from Ccr6^−/− mice versus wild-type mice versus CD25^hi subset. (G) Percent of all thymic Treg cells and CD25^hi thymic Treg cells that stain positively for Ccr6. All data are representative of three independent experiments with n ≥ 3 mice.

Figure 6. Ccl20 Preferentially Drives Neonatal Treg Cell Migration In Vitro

(A) CD4⁺ enriched D9 thymocytes were incubated in a transwell system with various chemokines, and cell migration to the lower transwell was assessed at 3 hr by flow cytometry. Treg cell migration index for various chemokines; index calculated as fold increase in Treg cells migrated in presence of chemokine (at 1,000 ng/ml) relative to no chemokine control. Each data point depicts a technical replicate. (B) Representative flow cytometry plot of CD4⁺ population in lower transwell in the presence of Ccl20 at 100 ng/ml versus no chemokine control. (C) Percentage of Treg cells migrated to lower transwell at varying doses of Ccl20. Calculated as number of Treg cells in lower transwell / number of Treg cells input to assay. (D) Percentage of FoxP3^neg CD4⁺ Teff cells migrated at varying doses of Ccl20. Data in (A) are representative of two independent experiments; (B)–(D) are representative of three independent experiments. See also Figure S3.

Figure 7. Ccr6 Expression on Treg Cells Facilitates Their Migration into Neonatal Skin

(A) CD4⁺ enriched thymocytes from CD45.1 WT and CD45.2 Ccr6^−/− mice were injected intraperitoneally into D9 rag2^−/− neonates. Flow cytometry plots depicting percentage WT versus Ccr6^−/− Treg cells in injected population on D0 and in spleen, skin-draining lymph nodes (SDLN), and skin of recipients 13 days post-transfer. (B–D) Skin:spleen ratio for WT and Ccr6^−/− Treg cells (i.e., percentage of WT Treg cells in skin divided by percentage of WT Treg cells in spleen) (B), skin:spleen ratio for WT and Ccr6^−/− CD4⁺ Teffs (C), and percentage of Ki-67⁺ WT and Ccr6^−/− Treg cells in skin (D). Data are representative of two independent experiments with n ≥ 5 mice. See also Figure S4.