Topical 1,25-dihydroxyvitamin D₃ subverts the priming ability of draining lymph node dendritic cells

Abstract

The active form of vitamin D, 1,25-hydroxyvitamin D₃ [1,25(OH)₂D₃] is produced in skin following exposure to sunlight. It is also used topically to control inflammatory skin diseases by stimulating keratinocyte differentiation and suppressing immune responses. Administration of 1,25(OH)₂D₃ to the skin of mice increases the capacity of CD4⁺ CD25⁺ (Foxp3⁺) regulatory T cells residing in the skin-draining lymph nodes (SDLN) to suppress immune responses. We hypothesized that dendritic cells (DC) may migrate from the skin to the lymph nodes to regulate T-cell function. Increased proportions of skin-derived DC (CD11c⁺ ClassII⁺ DEC-205^hi CD8^lo) cells were detected in the SDLN 18 hr after topical 1,25(OH)₂D₃ treatment of mouse skin. The capacity of DC from the SDLN to take up, process and present antigen to co-cultured T cells was not modified following topical 1,25(OH)₂D₃. However, CD11c⁺ cells from the SDLN of 1,25(OH)₂D₃-treated mice induced a significantly smaller ear-swelling response in a T helper type 1/17-mediated model of contact hypersensitivity. CD4⁺ CD25⁺ cells isolated from the ear-draining lymph nodes (EDLN) of mice that received ear injections of CD11c⁺ cells from donor mice topically treated with 1,25(OH)₂D₃ more potently suppressed effector cell proliferation. In addition, EDLN cells from recipients of CD11c⁺ cells from 1,25(OH)₂D₃-treated mice produced increased interleukin-4 levels. The CD11c⁺ cells from the SDLN of mice treated with topical 1,25(OH)₂D₃ expressed increased levels of indoleamine 2,3-dioxygenase messenger RNA, a molecule by which topical 1,25(OH)₂D₃ may enhance the ability of DC to control the suppressive function of CD4⁺ CD25⁺ cells.

Introduction

Interest in the involvement of vitamin D in various body systems and disease settings has recently increased with observations that deficiency in this hormone is an emerging health problem in many countries. With ultraviolet B (UVB) irradiation (290–315 nm), 7-dehydrocholesterol in skin is converted into pre-vitamin D₃. This molecule then isomerizes with body heat into vitamin D₃. Vitamin D binding protein transports much of the vitamin D₃ to the liver and kidneys for further hydroxylation into 1,25-dihydroxyvitamin D₃ [1,25(OH)₂D₃], the active form of vitamin D. Various skin cells including melanocytes and keratinocytes also possess the enzymatic machinery to produce 1,25(OH)₂D₃ where local skin concentrations of 2–5 nm can be achieved following UVB irradiation.^1,2 Recent investigations have identified that 1,25(OH)₂D₃ is important within the skin immune system.

The function of various immune cells located in the skin and draining lymphatic tissue can be modified by 1,25(OH)₂D₃ including the Langerhans cells,³ mast cells^4,5 and regulatory T cells.^6,7 Skin homing receptors (such as CCR10) can be up-regulated on 1,25(OH)₂D₃-stimulated T cells.⁸ Topically applied 1,25(OH)₂D₃ is also an effective treatment for skin inflammatory diseases such as psoriasis.⁹ 1,25(OH)₂D₃ reduces morbidity in the skin through direct effects on keratinocyte proliferation and differentiation in psoriatic lesions (reviewed in ref. 9) and also through immunoregulation (reviewed in ref. 10). The ability of topical 1,25(OH)₂D₃ (or analogues) to modify immune responses has been demonstrated in both mice^7,11 and humans,^12,13 where this treatment suppresses contact hypersensitivity responses. Topical 1,25(OH)₂D₃ may also be effective in controlling allergic diseases including asthma, where subcutaneous immunotherapies incorporating 1,25(OH)₂D₃ reduce respiratory inflammation during allergic airway disease.¹⁴

We recently described how topical application of 1,25(OH)₂D₃ or UVB irradiation increases the suppressive ability of CD4⁺ CD25⁺ Foxp3⁺ cells isolated from the skin-draining lymph nodes (SDLN).⁶ The pathway by which topical 1,25(OH)₂D₃ controls regulatory T cells in the SDLN is unknown. A candidate skin cell, which could be responsible for directly interacting with regulatory T cells in the SDLN is the dendritic cell (DC). Indeed, topical 1,25(OH)₂D₃ (or an analogue) reduces the number of MHC class II⁺ cells in the skin of treated mice,^7,11 suggesting that these cells could leave the skin following 1,25(OH)₂D₃ treatment. In other studies, incorporation of 1,25(OH)₂D₃ (100 ng) into a subcutaneous vaccine consisting of diphtheria and alum promoted the migration of cutaneous DC to secondary lymphoid sites such as the Peyer's patches.¹⁵ This study details new investigations into the ability of topical 1,25(OH)₂D₃ to modulate DC activity in the skin and the SDLN. Topical 1,25(OH)₂D₃ altered the phenotypic profile of DC in the SDLN. The ability of CD11c⁺ cells from the SDLN to prime T helper type 1 (Th1)/Th17-driven contact-hypersensitivity responses was subverted by earlier skin treatment with 1,25(OH)₂D₃, with reduced responses linked with an increased suppressive activity of CD4⁺ CD25⁺ regulatory cells and enhanced Th2 responses.

Materials and methods

Mice

Female 8-week-old BALB/c mice were purchased from the Animal Resources Centre (Murdoch, Western Australia, Australia). All experiments were performed according to the ethical guidelines of the National Health and Medical Research Council of Australia and with approval from the Telethon Institute for Child Health Research Animal Ethics Committee.

Topical vitamin D application

A 100-μl aliquot containing 125 ng 1,25(OH)₂D₃ (Sigma Chemical Company, St. Louis, MO, USA) diluted in ethanol, propylene glycol and water mixed at a 2 : 1 : 1 ratio was painted onto a clean-shaven 8-cm² area of dorsal skin surface of mice as described previously.⁶

Flow cytometric analysis of skin- or ear-draining lymph node cells

The SDLN (inguinal, axillary and brachial) or ear-draining lymph nodes (EDLN, auricular) were removed from mice, pooled within experimental groups and physically disaggregated. Staining of surface antigens was performed as previously described.⁶ An intracellular staining kit (eBiosciences, San Diego, CA, USA) was used to determine intracellular Foxp3 and interleukin-10 (IL-10) expression. At least 10 000 cells of interest were collected using either the fluorescence-activated cell sorter (FACS) LSRII or FACS Calibur (BD Biosciences, Heidelberg, Germany) flow cytometers. Data were analysed using FlowJo software (Treestar, Ashland, OR, USA).

Measuring in vitro antigen uptake and processing by skin-derived DC

Eighteen hours after topical 1,25(OH)₂D₃ treatment, the SDLN were removed and a single-cell suspension was generated by using a sterile scalpel blade to mince the lymph nodes, which were then digested with collagenase (type 4, 1 mg/ml, Worthington Biochemical, Lakewood, NJ, USA) and DNase (type 1, 0·1 mg/ml, Sigma) for 30 min at 37° with 5% CO₂. Digested lymph nodes were filtered and cells were treated for 5 min at 4° with anti-CD16/CD32 monoclonal antbodies (mAb; BD Biosciences), and then DQ-OVA (Molecular Probes, Invitrogen Australia, Mulgrave, Australia 10–10000 ng/ml) or Alexa488–ovalbumin (OVA) (Molecular Probes, 10–1000 ng/ml) for 2 hr at 37° or 4°. Alexa488-OVA is used as a measure of antigen uptake.¹⁶ Upon cleavage by proteolytic activity within cells, the fluorochrome DQ fluoresces and is used as a measure of antigen processing.¹⁶ These processes occur at 37°. Surface molecules on DC in the lymph nodes were detected by incubating cells with mAb as described above.

Purification of CD11c⁺ cells

Eighteen hours after topical 1,25(OH)₂D₃ treatment, the SDLN were removed and a single-cell suspension was generated as described above. CD11c microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany) were used with an AutoMACS cell separator (Miltenyi Biotec) to positively select for CD11c⁺ cells. Flow cytometry was used to determine that the purity of live isolated CD11c⁺ cells was routinely > 90%. The expression of activation markers such as CD86 on the CD11c⁺ cells was not modified by this positive selection process (data not shown).

Measuring in vitro antigen presentation by skin-derived DC

The fluorescent antigen, 0·5% fluorescein isothiocyanate (FITC; Sigma, diluted 1 : 1 in acetone : dibutylphthalate) was painted onto the skin of mice 30 min after topical treatment. The SDLN were again removed 18 hr after topical 1,25(OH)₂D₃ treatment with FITC⁺ CD11c⁺ cells sorted to 80–90% purity using a FACSAria cell sorter. These cells were co-cultured with CD4⁺ T cells purified from the SDLN of mice administered 0·5% FITC 7 days earlier in complete RPMI-1640 (RPMI medium; Gibco, Auckland, New Zealand) with 10% fetal calf serum, 2 mm l-glutamine, 50 μm 2-mercaptoethanol and 5 mg/ml gentamicin). For all co-cultures, methyl-[³H]thymidine (10 μl (0·25 mCi)/well; Amersham Pharmacia Biotech, Piscataway, NJ) was added after 72 hr at 37° with 5% CO₂ and cells were harvested at 96 hr. Culture supernatants were also obtained at 96 hr.

Cytokine detection in tissue culture supernatants

Interleukin-2, IL-4, IL-5, IL-10, IL-17 and interferon-γ (IFN-γ) were detected using rat anti-mouse IL-2, IL-4, IL-5, IL-10, IL-17 or IFN-γ enzyme-linked immunosorbent assay capture and detection mAb (BD Biosciences) in a dissociation-enhanced time-resolved fluorescence immunoassay with Europium as the label (sensitivity 25 pg/ml). Recombinant mouse IL-2, IL-4, IL-5, IL-10, IL-17 or IFN-γ (BD Biosciences) were used as the standards.

In vivo priming assay

Eighteen hours after topical skin treatment, CD11c⁺ cells were purified from the SDLN of treated mice and 10⁵ cells in 20 μl 0·9% saline were adoptively transferred into the ear pinnae of naive BALB/c mice. CD11c⁺ cells were loaded with antigen before adoptive transfer by incubating cells with 1 mm dinitrobenzylsulphonic acid (DNBS, Sigma) for 30 min at 37°. Cells were washed three times before the transfer. The ears of some mice received a saline injection only to control for the non-specific inflammatory effects of transferring fluid into the ears. Seven days after the injection of cells, an ear-swelling response was elicited by painting 10 μl 0·2% 2,5-dinitrofluorobenzene (DNFB, Sigma, diluted in acetone) onto each ear pinna. An additional group of mice that received saline only were also challenged with DNFB. Forty-eight hours after the ear challenge with DNFB, the ear thickness was measured in a blinded manner, as previously described.¹⁷

Culturing ear-draining lymph node cells

Forty-eight hours after ear challenge with 0·2% DNFB, the EDLN were removed and cells were cultured at 10⁵ cells/200 μl per well (six replicates per treatment) in complete RPMI medium. DNBS was added to cultures at a concentration of 1 mm. Methyl-[³H]thymidine was added after 72 hr at 37° with 5% CO₂ and cells were harvested at 96 hr. Culture supernatants were also obtained at 96 hr.

Regulation of proliferation by CD4⁺ CD25⁺ cells from the EDLN

Forty-eight hours after ear challenge with 0·2% DNFB, CD4⁺ CD25⁺ cells (≥ 95%, as determined using flow cytometry) were purified from the EDLN of recipient mice by using the CD4⁺ CD25⁺ regulatory T-cell isolation kit (Miltenyi Biotec). Responder CD4⁺ CD25⁻ cells (≥ 95% pure) were purified from the EDLN of mice who received CD11c⁺ cells from vehicle-treated mice also by using the CD4⁺ CD25⁺ regulatory T-cell isolation kit (Miltenyi Biotec) and then labelled with 5 μm CFSE (Molecular Probes). CD4⁻ cells isolated from the same mice were used as an antigen-presenting cell population. CD4⁺ CD25⁻ cells (10⁵ cells/200 μl per well) were co-cultured with CD4^– cells at a constant ratio of 1 : 20 with 1 mm DNBS in complete RPMI. CD4⁺ CD25⁺ cells were then cultured at ratios of 1 : 1 or 1 : 10 with the CD4⁺ CD25⁻ responder cell population. After incubation at 37° for 92 hr with 5% CO₂, supernatants were harvested and cells were washed and stained with phycoerythrin-Cy5-conjugated anti-CD4 mAb (BD Biosciences). In some experiments anti-IL-10 mAb (BD Biosciences, 1 μg/ml) was added to the co-cultures. Data were analysed using the proliferation algorithm for FlowJo (version 8.7.3) on viable cells, which were gated according to forward- and side-scatter properties and CD4 expression with 50 000 cells collected on the flow cytometer. The percentage of total cells that had divided was determined by using measurements of the number of viable CD4⁺ CFSE^lo cells.

Measuring messenger RNA in CD11c⁺ cells

Eighteen hours after topical skin treatment, CD11c⁺ cells were purified from the SDLN and ≥ 5 × 10⁵ cells were snap-frozen in 350 μl buffer RLT plus (Qiagen, Doncaster, Vic, Australia). RNA was extracted using the RNAeasy plus minikit (Qiagen) according to the manufacturer's instructions. Complementary DNA was reversed transcribed from the RNA samples using the Quantitect kit (Qiagen). Using primers designed in-house, a real-time polymerase chain reaction (PCR) was then performed as previously described¹⁸ using 2× RT² qPCR Master Mix (SABiosciences, Frederick, MD, USA). The PCR was performed using the standard two-step cycling conditions on the ABI 7000 SDS (Applied Biosystems, Foster City, CA, USA). Melting curve analysis was used to assess the specificity of the assay. Expression levels were determined by a standard curve created from serial dilutions of the PCR product and normalized to the reference gene eukaryotic translation elongation factor 1α. Primer pairs were as follows: eukaryotic translation elongation factor 1α, 5′-CTGGAGCCAAGTGCTAATATGCC-3′ and 5′-GCCAGGCTTG-AGAACACC AGTC-3; CCL22 5′-GGTCCCTATGGTG-CCAATG-3′ and 5′-TTATCAAAACAACGCCAGGC-3′; indoleamine 2,3-dioxygenase (IDO) 5′-AGGCTGGCAAA-GAATCTCCT-3′ and 5′-AATGACAAACTCACGGACT-GG-3′; IFN-γ; 5′-TCAAGTGGCATAGATGTGGAAGAA-3′ and 5′-TGGCTCTGCAGGATTTTCATG-3′; IL-10, 5′-GG-TTGCCAAGCCTTATCGGA-3′ and 5′-ACCTGCTCCACT-GCCTTGCT-3′; IL-12p35 5′-TGGACCTGCCAGGTGTCTTAG-3′ and 5′-CAATGTGCTGGTTTGGTCCC-3′; IL-12p40 5′-AAGGAACAGTGGGTGTCCAG-3′ and 5′-GGCAAAC-CAGGAGATGGTTA-3′; relB; 5′-TCCGCGCCCGAGCTA-3′ and 5′-GCCAAAGCCGTTCTCCTTAA-3′ transforming growth factor-β (TGF-β), 5′-CACTGATACGCCTGAGTG-3′ and 5′-GTGAGCGCTGAATCGAAA-3′.

Statistical analyses

Data were compared using Student's t test with the Prism statistical analysis programs for Macintosh (v5.0a). Differences were considered to be significant at P-values < 0·05.

Results

Topical 1,25(OH)₂D₃ alters the phenotypic profile of DC in the SDLN

The SDLN were removed from mice topically treated up to 48 hr earlier with vehicle or 1,25(OH)₂D₃. The total number and viability of SDLN cells was not modified by topical 1,25(OH)₂D₃ (data not shown). The profile of DC populations in the SDLN was determined using the cell surface markers CD11c and MHC class II to differentiate DC (CD11c^hi MHC class II^hi) from other lymph node cells (Fig. 1a). Populations of DC were further classified into populations expressing high and low levels of DEC205 and CD8 (Fig. 1a); markers used previously to distinguish skin-derived from lymph node-derived DC populations in the SDLN.¹⁹ At 18 hr after treatment with 1,25(OH)₂D₃, there were significant increases in the proportions of DC (CD11c^hi MHC class II^hi cells) in the SDLN that were DEC205⁺ CD8⁺ (by 50%) or DEC205⁺ CD8⁻ cells (by 28%; Fig. 1b). These differences persisted until 24 hr, but resolved by 48 hr post-skin-treatment (data not shown). DEC205⁺ CD8⁻ cells comprise a mixture of skin-derived DC, whereas DEC205⁺ CD8⁺ cells express Langerin, and may belong to a recently identified subpopulation of DC in the SDLN (reviewed in ref. 20).

Figure 1.

Topical vitamin D₃ [1,25(OH)₂D₃] modifies the phenotype of dendritic cell (DC) populations in the skin-draining lymph nodes (SDLN). In (a), the proportions of CD11c^hi MHC classII^hi cells differentiated according to the expression of CD8 and DEC-205 were determined in the SDLN of mice after topical vehicle or 1,25(OH)₂D₃ application. In (b) the proportion of CD11c^hi MHC classII^hi cells that were also DEC205⁻ CD8⁺, DEC205⁺ CD8⁺, DEC205⁺ CD8⁻ or DEC205⁻ CD8⁻ in the SDLN 18 hr after topical vehicle or 1,25(OH)₂D₃ treatment is shown for nine individual mice per treatment from three repeated experiments (mean ± SEM, *P < 0.05).

Topical 1,25(OH)₂D₃ down-regulates CD86 expression by some SDLN DC

The expression of CD40, CD80, CD86, B7-H3 and B7-H4 by the different SDLN DC subpopulations was examined at 18 hours post-skin-treatment. CD8 and DEC205 were again used to distinguish four DC subpopulations in the SDLN, where the DEC205⁺ CD8⁻ DC expressed the greatest quantity of each surface marker, except for B7-H3 (Fig. 2a, and data not shown). However, there was no change in the expression of CD40, CD80 or B7-H3 (data not shown) on any DC population from the SDLN with topical 1,25(OH)₂D₃ treatment. The expression of CD86 was down-regulated on DEC205⁺ CD8⁺ and DEC205⁺ CD8⁻ cells at 18 hr post-topical 1,25(OH)₂D₃ (Fig. 2a), while B7-H4 was up-regulated on DEC205⁺ CD8⁺ cells (data not shown).

Figure 2.

Topical vitamin D₃ [1,25(OH)₂D₃] reduces the expression of CD86 on dendritic cell (DC) populations in the skin-draining lymph nodes (SDLN) but not the ability of DC from the SDLN to uptake or process antigen in vitro. The geometric mean fluorescent intensity (GMFI) of (a) CD86 on DC (CD11c^hi MHC ClassII^hi) populations distinguished by the expression of DEC205 and/or CD8. Results are shown for nine individual mice per treatment from three repeated experiments (mean ± SEM, *P < 0.05). In (b–d), the ability of different DC populations from the SDLN to take-up or process antigen was assessed using ovalbumin (OVA) conjugated to the fluorescent molecules Alexa488 or DQ. Cells that take-up OVA-Alexa will also fluoresce when, during antigen processing, the usually quenched molecule DQ will begin to fluoresce. In (b), examples of OVA-Alexa⁺ and DQ-OVA⁺ cells are shown relative to a negative control (no OVA). The proportions of CD11c^hi MHC classII^hi DEC205⁺ CD8⁻ cells that expressed OVA-Alexa or DQ-OVA are shown when SDLN cells from mice treated 18 hr earlier with either vehicle or 1,25(OH)₂D₃ were incubated at 37° for 2 hr with various concentrations of (c) OVA-Alexa or (d) DQ-OVA, respectively.

Topical 1,25(OH)₂D₃ does not modify the in vitro antigen uptake, processing or presenting capacity of SDLN DC

The ability of these SDLN DC populations to take-up and process antigen was assessed using fluorescently tagged OVA proteins. Eighteen hours after skin treatment, SDLN cells were incubated with various concentrations of Alexa488-OVA or DQ-OVA, and the extents of antigen uptake and processing were determined, respectively. Examples of cells that had taken-up antigen (Alexa488-OVA⁺) or processed antigen (DQ-OVA⁺) are shown relative to a negative (no OVA) control in Fig. 2(b). The ability of DEC205⁺ CD8⁻ DC (or any of the other DC populations) in the SDLN to either take-up (Fig. 2c) or process (Fig. 2d) antigen was not modified by topical 1,25(OH)₂D₃. In a further assay, there was no difference in the ability of CD11c⁺ cells purified from the SDLN 18 hr after treatment with either vehicle or 1,25(OH)₂D₃, to present OVA peptide to co-cultured OVA-TCR⁺ CD4⁺ T cells (from DO11.10 mice) as no differences in the proliferation or cytokine production (IFN-γ and IL-10, data not shown) by co-cultured T cells was observed.

In an alternative assay, to test the ability of DC in the skin to take-up and process antigen in vivo, the fluorescent antigen FITC was applied to the skin of mice 30 min after topical 1,25(OH)₂D₃ treatment. Eighteen hours after topical 1,25(OH)₂D₃, the proportion of FITC⁺ CD11c⁺ cells was assessed in the SDLN of mice. In Fig. 3(a), an example of FITC expression by CD11c⁺ cells is depicted in mice treated (or not) with FITC 18 hr earlier. The proportion of CD11c⁺ FITC⁺ cells in the SDLN was not modified by topical 1,25(OH)₂D₃ (data not shown). The expression of CD86 (Fig. 3b) and CD80, but not of CD40 (data not shown), was down-regulated on CD11c⁺ FITC⁺ cells from mice treated topically with 1,25(OH)₂D₃. However, there was no difference in the capacity of CD11c⁺ FITC⁺ cells sorted from the SDLN 18 hr after topical vehicle or 1,25(OH)₂D₃ to induce the proliferation of co-cultured CD4⁺ T cells from FITC-sensitized mice (Fig. 3c). Together, these data indicate that the functional abilities of DC from the SDLN to take-up and process antigen both in vitro and in vivo were not modified by topical 1,25(OH)₂D₃.

Figure 3.

Topical vitamin D₃ [1,25(OH)₂D₃] does not modify the ability of skin dendritic cells (DC) to take-up antigen in vivo. To assess the ability of DC in the skin of topically treated mice to take-up antigen in vivo, the fluorescent antigen, fluorescein isothiocyanate (FITC), was applied (0.5%) to the skin of mice treated 30 min earlier with vehicle or 1,25(OH)₂D₃. In (a), FITC⁺ CD11c⁺ cells were identified in the skin-draining lymph nodes (SDLN) of mice 18 hr after skin treatment as compared with mice untreated with FITC (no FITC). In (b), the expression of CD86 on CD11c⁺ FITC^– cells and CD11c⁺ FITC⁺ cells for six individual mice per treatment from two repeated experiments (mean + SEM, *P < 0.05) is shown. In (c), the ability of CD11c⁺ FITC⁺ cells sorted from the SDLN 18 hr after topical treatment to induce the proliferation of co-cultured CD4⁺ T cells from FITC-sensitized mice is shown (mean ± SEM, for six replicate wells per treatment).

The in vivo priming capacity of CD11c⁺ cells from the SDLN of mice topically treated with 1,25(OH)₂D₃ is compromised

Our results to date suggested that DC populations in the SDLN of 1,25(OH)₂D₃-treated mice exhibit minor phenotypic differences with no change in their in vitro abilities to present antigen. An in vivo priming assay was used to further investigate the functional ability of these cells after topical application of 1,25(OH)₂D₃. Eighteen hours later, CD11c⁺ cells were purified from the SDLN and labelled with DNBS (Fig. 4a), a water-soluble analogue of the hapten, DNFB. Then, 10⁵ cells were adoptively transferred via subcutaneous injection into the ear pinnae of naive recipient mice (Fig. 4a), with another group of mice receiving a saline injection as a control. Seven days after ear injection, the ear pinnae were challenged with 0·2% DNFB to induce ear-swelling, which was measured 48 hr later (Fig. 4a). Significant ear-swelling was observed in mice that received CD11c⁺ cells from the SDLN of vehicle-treated donor animals (Fig. 4b). This response was significantly reduced (by 62%, relative to challenge-only controls) in recipients of CD11c⁺ cells from 1,25(OH)₂D₃-treated mice (Fig. 4b). The number of cells recovered from the EDLN cells was proportional to the extent of ear-swelling, such that the number of cells in recipients of CD11c⁺ cells from 1,25(OH)₂D₃-treated mice was significantly reduced relative to the vehicle treatment (Fig. 4c). Together, these observations indicate that topical 1,25(OH)₂D₃ treatment compromised the ability of CD11c⁺ cells in the SDLN to prime an immune response in vivo.

Figure 4.

Topical vitamin D₃ [1,25(OH)₂D₃] reduces the in vivo priming ability of CD11c⁺ cells from the SDLN. In (a), the ability of CD11c⁺ cells from the SDLN of mice treated topically with 1,25(OH)₂D₃ to prime immune responses in vivo was assessed. 18 hr after topical treatment, the SDLN were removed and CD11c⁺ cells were purified and labelled with 1 mm DNBS. Cells (10⁵) were then adoptively transferred (by subcutaneous injection) into the ear pinnae of naive mice, with a group of mice receiving saline only. After 7 days, ear pinnae of mice were challenged with 10 μl 0.2% DNFB (in acetone), with ear-swelling measured 48 hr later. In (b), the ear-swelling response 48 hr after ear challenge with DNFB is shown for recipients of CD11c⁺ cells from the skin-draining lymph nodes (SDLN) of vehicle-treated (Vehicle CD11c⁺) or 1,25(OH)₂D₃-treated (VitD CD11c⁺) mice. Responses are also shown for recipients of saline only (Saline) and mice that were only challenged with DNFB (Challenge only) (mean + SEM, n = 5 mice/treatment, *P < 0.05). In (c), the number of ear-draining lymph node (EDLN) cells determined 48 hr after ear challenge of recipient mice are shown, with results for the Vehicle CD11c⁺, VitD CD11c⁺ and Challenge only treatments described above (mean ± SEM, pooled EDLN cells from n = 5 mice/treatment for four or five repeated experiments, *P < 0.05).

CD11c⁺ cells from the SDLN of 1,25(OH)₂D₃-treated mice skew immune responses towards a Th2 phenotype

The EDLN cells were removed from recipient mice 48 hr after ear challenge (Fig. 4a) and the capacity of these cells to secrete various cytokines was assessed by culture for 96 hr, with or without DNBS. Although the secretion of IL-2, IL5, IL-10 and IL-17 by EDLN cells was not modified, IL-4 levels were significantly enhanced in the supernatants of EDLN cells from recipients of CD11c⁺ cells from 1,25(OH)₂D₃-treated mice (Fig. 5, with data not shown for IL-2 or IL-5). When DNBS was included in the culture, IFN-γ levels were reduced in supernatants of EDLN from recipients of CD11c⁺ cells from 1,25(OH)₂D₃-treated mice. No difference in the proliferation of EDLN cells over this 96-hr time–course was detected for EDLN from recipients of CD11c⁺ cells from vehicle- or 1,25(OH)₂D₃-treated mice (data not shown). Hence, the reduced ability of CD11c⁺ cells from 1,25(OH)₂D₃-treated mice to prime immune responses in vivo may be partly the result of a switch towards a Th2 phenotype.

Figure 5.

Topical vitamin D₃ [1,25(OH)₂D₃] enhances interleukin-4 (IL-4) secretion by ear-draining lymph node (EDLN) cells from recipients of CD11c⁺ cells from the SDLN. The EDLN cells were removed from the recipients of CD11c⁺ cells and from vehicle-treated or 1,25(OH)₂D₃-treated mice (as described in Fig. 4a) 48 hr after ear challenge and cultured for 96 hr with and without 1 mm DNBS. Concentrations of IL-4, IL-10, IL-17 and interferon-γ (IFN-γ) in supernatants are shown for results pooled from two individual experiments (mean + SEM, pooled EDLN cells from n = 5 mice/treatment for three replicate wells/treatment, *P < 0.05).

CD11c⁺ cells from the SDLN of 1,25(OH)₂D₃-treated mice increase the regulatory activity of CD4⁺ CD25⁺ cells in vivo

To determine if the reduced priming ability of CD11c⁺ from 1,25(OH)₂D₃-treated mice was the result of an effect on regulatory T cells, EDLN were removed 48 hr after ear challenge (Fig. 4a). There was no difference in the number of CD4⁺ CD25⁺ Foxp3⁺ cells identified in the EDLN, with 3·9% ± 0·1% and 3·9% ± 0·3% (mean ± SEM, n = 3 repeat experiments) in recipients of CD11c⁺ cells from vehicle or 1,25(OH)₂D₃-treated mice, respectively. There was also no difference in mean fluorescence intensity of Foxp3 expressed by CD4⁺ CD25⁺ cells in the EDLN of the vehicle or 1,25(OH)₂D₃ treatments, with most of the cells (≥ 90%) expressing Foxp3 (data not shown). To examine the suppressive activity of CD4⁺ cell populations, CD4⁺ CD25⁺ and CD25⁺ CD25^– cells were isolated from the EDLN 48 hr after ear challenge. The ability of these cells to modulate the proliferation of a responding cell population was determined by incubating them at various ratios with CFSE-labelled responder cells together with CD4^– cells (for antigen presentation) and DNBS. The dilution of CFSE in the responder cells in the cultures was determined after 96 hr of culture. CD4⁺ CD25⁺ cells from the EDLN of recipients of CD11c⁺ cells had a greater suppressive capacity than CD4⁺ CD25⁺ cells from mice challenged with DNFB (no cell transfer, Fig. 6). Previous topical 1,25(OH)₂D₃ treatment further increased the capacity of CD4⁺ CD25⁺ cells isolated from the EDLN to suppress the proliferation of co-cultured responder cells (Fig. 6). The proliferation of responder cells was not affected by CD4⁺ CD25⁻ cells from either experimental recipient mice (Fig. 6). The increased ability of CD4⁺ CD25⁺ cells from recipients of CD11c⁺ cells from 1,25(OH)₂D₃-treated mice was not dependent on IL-10 because the inclusion of anti-IL-10 mAb in these co-cultures did not prevent their enhanced ability to suppress responder cell proliferation (data not shown). Therefore, CD11c⁺ cells from the SDLN of mice treated topically with 1,25(OH)₂D₃ increase the suppressive activity of CD4⁺ CD25⁺ cells in recipient mice.

Figure 6.

Topical vitamin D₃ [1,25(OH)₂D₃] enhances the suppressive ability of CD4⁺ CD25⁺ cells from the ear-draining lymph nodes (EDLN) of recipients of CD11c⁺ cells from the skin-draining lymph nodes (SDLN). The EDLN cells from the recipients of CD11c⁺ cells from vehicle-treated or 1,25(OH)₂D₃-treated or challenge-only mice (as described in Fig. 4a), were removed 48 hr after ear challenge. CD4⁺ CD25⁺ or CD4⁺ CD25⁻ cells were purified from the EDLN and cultured with a CFSE-labelled CD4⁺ CD25⁻ responding T-cell population, CD4⁻ antigen-presenting cells and 1 mm DNBS for 96 hr. CD4⁺ CD25⁺ or CD4⁺ CD25⁻ cells were cultured with the responding cells at a ratio of 1 : 1 or 1 : 10. The proliferation of the responding cells as determined by dilution of CFSE is shown as the proportion of CD4⁺ CFSE^lo cells. Results were pooled from three repeated experiments (mean + SEM, pooled EDLN cells from n = 5 mice/treatment for three replicate wells/treatment, *P < 0.05).

Topical 1,25(OH)₂D₃ enhances indoleamine 2,3-dioxygenase messenger RNA expression by CD11c⁺ cells in the SDLN

To investigate a potential mechanism by which topical 1,25(OH)₂D₃ compromises the ability of CD11c⁺ cells to prime immune responses in vivo, the expression of eight messenger RNAs (mRNAs) was assessed, including CCL22, IDO, IFN-γ, IL-10, IL-12p35, IL-12p40, relB and TGF-β. Of these eight mRNAs investigated for cells from three independent experiments, only that for IDO was significantly up-regulated in CD11c⁺ cells isolated from the SDLN of mice treated topically with 1,25(OH)₂D₃ (Fig. 7).

Figure 7.

Topical vitamin D₃ [1,25(OH)₂D₃] increases indoleamine 2,3-dioxygenase messenger RNA (IDO mRNA) expression in CD11c⁺ cells from the skin-draining lymph nodes (SDLN). CD11c⁺ cells were purified from the SDLN of mice 18 hr after topical treatment with vehicle or 1,25(OH)₂D₃ and the expression of mRNAs for eight genes was determined as shown relative to the house-keeping gene Eef1α. Results are from three repeated experiments (mean + SEM, pooled CD11c⁺ cells from n = 10 mice/treatment, *P < 0.05).

Discussion

In this study, a single topical application of 1,25(OH)₂D₃ modified the immunostimulatory capacity of DC residing in the SDLN. Eighteen hours after topical 1,25(OH)₂D₃, CD11c⁺ DC from the SDLN primed a significantly smaller ear-swelling response in recipient mice, and enhanced the suppressive capacity of CD4⁺ CD25⁺ regulatory T cells. These results provide a mechanism for the suppressive effects of topical 1,25(OH)₂D₃ on immune responses, where DCs prime a reduced response and alter the function of regulatory T cells residing in the SDLN. We also identify a candidate immunomodulatory molecule, IDO, which may be expressed by 1,25(OH)₂D₃-modified DC to affect T-cell function. These are novel findings, which for the first time link experimentally the effects of 1,25(OH)₂D₃ with DC function and regulatory T-cell activity in vivo.

In vitro stimulation with 1,25(OH)₂D₃ can induce human or murine DC to acquire ‘tolerogenic’ qualities with reduced abilities to secrete IL-12 and to present antigen to co-cultured T cells, and enhanced capacities to induce regulatory T cells (reviewed in ref. 21). These observations are complemented by recent findings, where many gene targets identified in 1,25(OH)₂D₃-treated DC are associated with a tolerogenic phenotype.²² Although the effects of 1,25(OH)₂D₃ on DC in vivo are less well described, similar tolerogenic effects have been observed with 1,25(OH)₂D₃-modified DC indirectly promoting allograft survival, suppressing contact hypersensitivity responses, reducing allergic airway disease and enhancing T regulatory cell activity.^7,11,14,23 By transferring the CD11c⁺ cells from 1,25(OH)₂D₃-treated mice into naive recipient mice, we have confirmed these earlier observations, and formalized a direct link between the action of 1,25(OH)₂D₃ on DC function and subsequent regulation of CD4⁺ CD25⁺ cells.

While topical 1,25(OH)₂D₃ modulated the function of DC upon adoptive transfer in vivo, we did not observe any significant changes in the ability of CD11c⁺ cells from the SDLN to take-up, process or present antigen in vitro, even when the antigen (FITC) was painted onto skin immediately following 1,25(OH)₂D₃ application. Furthermore, co-stimulatory molecule expression was only modestly down-regulated on skin-derived DC populations in the SDLN by topical 1,25(OH)₂D₃. These observations are reminiscent of those published recently where there was no difference in antigen (FITC) carriage or co-stimulatory molecule expression by DC from the skin of vitamin D₃-deficient or -replete mice.²⁴ In our studies, we did not define the skin-derived DC population affected by topical 1,25(OH)₂D₃. SDLN-resident and not skin-derived DC could be the target cells. However, topical 1,25(OH)₂D₃ affected the numbers and phenotype (e.g. CD86 expression) of multiple DC subtypes (i.e. DEC205⁺ CD8⁺ and DEC205⁺ CD8⁻) in the SDLN suggesting that more than one type of DC may be affected by topical 1,25(OH)₂D₃. The expression of CD8 and DEC-205 on DC in the SDLN was similar to that described previously for mice in steady-state conditions.¹⁸ However, the co-expression of DEC-205 and CD8 on DC may be different in other lymphoid organ or tissue sites such as the spleen²⁵ or following antigenic stimulation.

This study identified IDO as a candidate molecule by which CD11c⁺ cells from 1,25(OH)₂D₃-treated mice may increase regulatory T-cell activity in 1,25(OH)₂D₃-treated mice. The expression of a panel of genes was examined, as they have previously been linked with a tolerogenic DC phenotype (CCL22,²⁶ IDO,²⁷ relB,²⁸ IL-10, TGF-β, reviewed in ref. 21) or an immunostimulatory phenotype (IL-12p35, IL-12p40, reviewed in ref. 21) with the expression of IFN-γ mRNA also determined as a negative control. The mRNA of IDO was the only mRNA of this eight gene panel to be significantly regulated in CD11c⁺ cells isolated from the SDLN of mice topically treated with 1,25(OH)₂D₃. However, previous studies of IDO expression by human cells cultured with 1,25(OH)₂D₃ have had conflicting results. Increased IDO enzymatic activity and enhanced IDO mRNA expression was detected in human CD4⁺ T cells cultured with 1,25(OH)₂D₃.²⁷ In contrast, human DC differentiated from peripheral blood monocytes in the presence of 1,25(OH)₂D₃ expressed an immature phenotype with an enhanced capacity to generate CD4⁺ T cells with regulatory activity, but they had reduced IDO enzymatic activity.²⁹ Perhaps the difference between the findings of Pedersen et al.²⁹ and our own is the result of the presence of 1,25(OH)₂D₃ during DC differentiation in vitro, as opposed to an acute dose of 1,25(OH)₂D₃ administered to the skin in vivo, where the presence of other cells and mediators could affect skin DC phenotype and function. In addition, our results appear to be independent of changes in relB mRNA expression (reviewed in ref. 28).

Others have postulated that the increased expression of receptor activator of nuclear factor-κB ligand (RANKL) by 1,25(OH)₂D₃-exposed keratinocytes may be responsible for modifying the downstream activity of epidermal DC and regulatory T cells.^7,30 Following topical administration of a 1,25(OH)₂D₃ analogue or UV irradiation of skin, keratinocytes up-regulate RANKL expression.^7,30 The DC from mice with keratinocytes over-expressing RANKL have an increased propensity to stimulate the proliferation of co-cultured regulatory T cells, although the suppressive activity of these cells was not modified on a per cell basis.³⁰ In contrast, following topical 1,25(OH)₂D₃, we observed that DC from the SDLN have an increased ability to promote the suppressive activity, but not the numbers of regulatory T cells.⁶ It is therefore unclear what role exists in our studies for increased RANKL expression by 1,25(OH)₂D₃-exposed keratinocytes to promote the tolerogenic potential of DC.

The 1,25(OH)₂D₃ can regulate Th2 immune responses in vitro. However, controversy exists as to how 1,25(OH)₂D₃ regulates IL-4 secretion and GATA-3 expression during Th2 differentiation of CD4⁺ T cells.^31–33 From our studies, DC affected by 1,25(OH)₂D₃in vivo possess the capacity to switch immune responses in the EDLN of recipient mice towards a Th2 type environment, as demonstrated by the increased IL-4 levels. Our observations are similar to those detected in a model of hapten-induced colitis, where 1,25(OH)₂D₃ (0·2 μg/kg intraperitoneally) reduced the severity of colitis and up-regulated Th2 markers like IL-4 and GATA-3 in the colons of treated mice.³⁴

In conclusion, topical 1,25(OH)₂D₃ modulates the activity of SDLN-resident CD4⁺ CD25⁺ cells through its tolerogenic effects on DC. These cells also promote a Th2 immune environment and perhaps through an IDO-dependent mechanism suppress Th1/17-driven immune responses. Our results support the continued use of 1,25(OH)₂D₃ (or its analogues) as a topical treatment of inflammatory skin diseases, for the induction of a tolerogenic environment to control or limit the severity of immune disease.

Glossary

Abbreviations:

1,25(OH)₂D₃

1,25-dihydroxyvitamin D₃

DC

dendritic cell

DNBS

dinitrobenzylsulphonic acid

DNFB

2,4-dinitrofluorobenzene

EDLN

ear-draining lymph nodes

FACS

fluorescence-activated cell sorting

FITC

fluorescein isothiocyanate

IDO

indoleamine 2,3-dioxygenase

IFN

interferon

IL

interleukin

mAb

monoclonal antibody

OVA

ovalbumin

RANKL

receptor activator of nuclear factor-κB ligand

SDLN

skin-draining lymph nodes

TGF

transforming growth factor

Th1

T helper type 1