Vitamin D Regulates Cytokine Patterns Secreted by Dendritic Cells to Promote Differentiation of IL-22-Producing T Cells

One central mechanism, by which vitamin D regulates human immune responses, is the direct modulation of dendritic cells (DCs). However, the effect of vitamin D on several key DC functions, such as the secretion of central inflammatory cytokines, remains controversial. Moreover, whether vitamin D treatment of DCs regulates their ability to promote differentiation of IL-17-/IL-22-producing T cell subsets, such as Th17 and Th22 cell, is not known. Here, we report that vitamin D treatment during differentiation of monocytes into DCs markedly enhanced their ability to secrete TNF-α, IL-6, IL-1β and IL-23. Cytokines secreted by vitamin D-treated DC were significantly more potent in driving differentiation of IL-22-producing T cells, but not IL-17-producing T cells, as compared to secreted cytokines of not-vitamin D-treated DCs. Finally, we found that the differentiation of IL-22-producing T cells mediated by supernatants of vitamin D-treated DCs was dependent on TNF-α IL-6 and IL-23. In summary, our study suggests a novel role of vitamin D in regulating DC-mediated immune responses in humans.


Ethics statement
This study was conducted according to the principles expressed in the Declaration of Helsinki. The study was approved by the local Ethic Committee (Ethikkommission) of the University of Cologne, Germany. All donors provided written informed consent for the collection of peripheral blood and subsequent analysis.

Cell isolation and DC culture
Whole blood or 'buffy coats' from healthy donors were obtained with informed consent. PBMCs were isolated by Ficoll-Paque (GE Healthcare). Monocytes were isolated via MACS cell separation (Miltenyi) according to the manufactures instructions either using the monocyte untouched isolation kit II for DC surface molecule analyses or using the CD14 + microbeads for PCR experiments and to obtain supernatants for the T cell differentiation assays. DCs were differentiated by culturing 0.5x10 6 /ml monocytes in rGM-CSF (200 U/ml) and rIL-4 (100 U/ml) in RPMI with 10% FCS. The serum was either supplemented with 1,25D (10 −8 M) or 25D (10 −7 M) or used without additional 1,25D or 25D. At day six or seven DCs were either stained for the expression of antigen-presenting and co-stimulatory surface molecules or were stimulated with Pam 3 Cys-SKKK (1 μg/ml), LPS (10 ng/ml), or CD40L (5 μg/ml) in fresh media containing 10% FCS supplemented with 1,25D (10 −8 M) or 25D (10 −7 M) or used without addition of 1,25D or 25D. After 18-24 hours, supernatants were harvested from these cultures and the DCs were analyzed for surface molecule expression by FACS using a Calibur (BD Biosciences) and FlowJo software (Tree Star). We did not observe different DC numbers after differentiation in FCS supplemented with 1,25D (10 −8 M) or 25D (10 −7 M), as compared to differentiation in FCS alone. Supernatants were immediately frozen for future culture with T cells. TNF-α, IL-6, IL-1β, IL-12p40, IL-12p70, IL-23 and IL-10, as well as total and active TGFβ levels in the supernatants were determined by ELISA or CBA (BD Biosciences). To measure the total amount of TGF-β in DC supernatants samples were pre-treated with 1 M HCL for 5 minutes followed by neutralization with 1.2 M NaOH.

T cell differentiation assay
T cell differentiation assays were carried out as previously described [23] with slight modifications: Naïve CD4 + T cells were isolated from whole blood or 'buffy coats' from healthy donors by naïve CD4 + T cell isolation kit (Miltenyi) according to the manufacturer's instruction. 5x10 4 T cells were stimulated using 1:1 anti-CD3/anti-CD28 antibody-coated beads according to the manufacturer's instruction in the T cell expansion and activation kit (Miltenyi) in 96-well Ubottom plates and cultured in 10% human AB serum, 50% 1,25D diff -DC or serum-DC supernatant, and 40% fresh media. In blocking experiments anti-TNF-α, anti-IL-6 receptor-α, anti-TGF-β, anti-IL-1β and anti-IL-12/IL-23p40 monoclonal antibodies alone or in combination were added to the cultures at the indicated concentrations. On day five, rIL-2 was added at a final concentration of 50 U/ml and T cells were cultured for additional seven days. For intracellular cytokine staining T cells were re-stimulated with PMA (50 ng/ml) and Ionomycin (0.5 μg/ml) for five hours, for the final 2.5 hours in the presence of Brefeldin A, in fresh media.
In some experiments T cells were counted using Trypan blue exclusion and were stained for intracellular cytokines using the Cytofix/Cytoperm fixation/permeabilization solution kit (BD Biosciences) according to the manufacturer's instruction. For measuring secreted cytokines T cells were re-stimulated with PMA (50 ng/ml) and Ionomycin (0.5 μg/ml) for 18-24 hours (without addition of Brefeldin A) in fresh media. Cytokines in the T cell culture supernatants were measured by ELISA or CBA (BD Biosciences). To measure cytokine expression of naïve T cells after isolation (day 0) 5x10 4 naïve CD4 T cells were stimulated with PMA (50 ng/ml) and Ionomycin (0.5 μg/ml) or cultured in media alone immediately after isolation, for five hours, the final 2.5 hours in the presence of Brefeldin A, and subsequently stained for intracellular cytokines using the Cytofix/Cytoperm fixation/permeabilization solution kit (BD Biosciences).
PCR mRNA was isolated from the GM-CSF/IL-4 stimulated monocytes after 24h using the RNeasy mini kit (Qiagen) according to the manufacturer's recommended protocol. cDNA was prepared and mRNA levels assessed by qPCR as previously described [45]. Primer sequences for human cathelicidin, CYP24A1 and h36B4 were previously reported [44,45].

Statistics
P-values were calculated using two-tailed Student's t-tests. All treatment groups in each subfigure were conducted in parallel. n refers to the number of repeated experiments performed with cells from individual human donors.
Given that 1,25D levels are low and kept relatively constant in human serum, most immune cells in peripheral tissues rely on the local conversion of 25D into bioactive 1,25D [32]. Thus, we also investigated the effect of additional 25D (10 −7 M) (Fig 2) on the differentiation and/or stimulation of human DCs in the same experimental setup as for 1,25D. We detected more TNF-α 8 ng/ml vs. 3 ng/ml, p<0.05), IL-6 (45 ng/ml vs. 12 ng/ml, p<0.05) and IL-23 (364 pg/ ml vs. 31 pg/ml, p<0.01) (Fig 2) secretion, as well as a trend towards more IL-1β and IL-12p40 secretion upon stimulation via TLR2/1 by 25D diff -DCs as compared to serum-DCs (Fig 2). In contrast, we did not measure any significant difference in IL-10 secretion, or total or active TGF-β (Fig 2). With respect to 25D stim -DCs there was no significant difference in cytokine secretion as compared to serum-DCs. Moreover, the cytokine pattern secreted by 25D diff/stim -DCs was comparable to the pattern secreted by 25D diff -DCs.
Next, we asked if differences between the cytokine patterns observed in 1,25D diff DCs vs. serum-DCs are dependent on the activating stimulus. Thus, we compared cytokine secretion of 1,25D diff -DCs vs. serum-DCs upon activation with the TLR2/1L Pam 3 Cys, the TLR4L LPS, or CD40L. We found that all three ligands induced significantly more TNF-α and IL-6 secretion by 1,25D diff -DCs as compared to serum-DCs (Fig 3). 1,25D diff -DCs also secreted more IL-1β, IL-12p40 and IL-23 irrespective of the activating stimulus (Fig 3). IL-12p70 secretion was low and not different between 1,25D diff -DCs and serum-DCs (Fig 3). Moreover, we did not measure significantly different amounts of IL-10, or total or active TGF-β (Fig 3). Taken together, our data show that vitamin D treatment of differentiating DCs increases their secretion of TNF-α, IL-6, IL-1β, IL-12p40 and IL-23.
To characterize the vitamin D-treated DCs in more detail, we analyzed the profile of surface molecule expression on vitamin D-treated DCs in comparison to serum-DCs before and after additional stimulation by TLR2/1L. On un-stimulated 1,25D diff -DCs we detected a significant lower expression of the antigen-presenting molecules HLA-DR and CD1a, as well as the costimulatory molecule CD80 as compared to serum-DCs (S2 & S3 Figs). The same pattern was observed for 25D diff -DCs compared to serum-DCs, yet only reaching statistical significance for  S3 Figs). Nevertheless, the expression pattern of HLA-DR, CD80, CD1a and CD14 as well as ILT3 in 1,25D diff -DCs and 25D diff -DCs as compared to serum-DCs within the TLR2/1-stimulated DC group was very similar to those in un-stimulated DCs. Also the mannose receptor CD206 and the chemokine receptors CCR5 and CCR7 remained at a low expression level within the stimulated DC group comparable to the un-stimulated DC group. Of note, PD-L1 was induced by TLR2/1L stimulation in all DC types (S2 & S4 Figs). In summary, vitamin D treatment of differentiating DCs resulted in a dual pro-/antiinflammatory phenotype characterized by a tolerogenic expression pattern of cell surface molecules, yet a markedly enhanced secretion of pro-inflammatory cytokines.

Supernatants of 1,25D diff -DCs promote differentiation of IL-22-producing T cells
One of the key functions of pro-inflammatory cytokines is the instruction of T cell polarization into effector T cells. Because it was shown that TNF-α and IL-6 promoted development of IL-22-producing T cells, and a combination of TNF-α, IL-6 and IL-1β promoted development of IL-17-producing T cells [23], we asked whether vitamin D treatment of differentiating DCs would enhance their ability to induce IL-17-and/or IL-22-producing T cells. Therefore, we activated 1,25D diff -DCs and serum-DCs with TLR2/1L and collected the supernatants of these cultures at 18-24 hours. Subsequently, we added the 1,25D diff -DC and serum-DC supernatants to naïve CD4 + T cells activated via their TCR. After culture for 12 days to allow differentiation into different T cell subsets, T cells were re-stimulated in fresh culture media for 18-24 hours and T cell cytokine secretion into the culture medium analyzed by ELISA or CBA. We found that T cells differentiated in the presence of TLR2/1-induced 1,25D diff -DC supernatants secreted significantly more IL-22 as compared to T cells cultured in the presence of TLR2/ 1-induced serum-DC supernatants (19.2 ng/ml vs. 11.4 ng/ml, p<0.01) (Fig 4). In contrast, we detected no significant differences in secretion of IFN-γ and IL-17a (Fig 4), as well as IL-4 ( Fig. A in S5 Fig), a signature cytokine for Th2 cell phenotypes, but a trend towards less IL-10, a signature cytokine for Tregs, respectively ( Fig. D in S5 Fig). However, in the presence of 1,25D diff -DC supernatants differentiated T cells secreted significantly more TNF-α as their serum-DC supernatant-treated counterparts (36.9 ng/ml vs. 19.9 ng/ml, p<0.05) (Fig. E in  S5 Fig).

1,25D diff -DC-supernatant mediated differentiation of IL-22-producing T cells is dependent on TNF-α/IL-6 and IL-23
TNF-α and IL-6 have been shown to be sufficient, as well required in the DC-mediated induction to drive development of IL-22-producing T cells [23]. Thus, we hypothesized that priming of IL-22-producing T cells by 1,25D diff -DCs is dependent on TNF-α/IL-6. To test our hypothesis we cultured activated, naïve CD4 + T cells with supernatants from TLR2/1-stimulated 1,25D diff -DCs in the presence of monoclonal anti-TNF-α/anti-IL-6R-α neutralizing antibodies. As a control we used a monoclonal anti-TGF-β neutralizing antibody. After culture for 12 days, we performed intracellular cytokine staining for IL-22. Blocking TNF-α/IL-6R-α, but not TGF-β, significantly inhibited the frequency of IL-22 + T cells (15.4% vs. 10.2% IL-22 + T cells, p<0.01) (Fig 6A). We also investigated cytokine secretion by the T cells and found that the addition of anti-TNF-α/anti-IL-6R-α antibodies resulted in a significant reduction of IL-22 secretion (22.9 ng/ml vs. 7.8 ng/ml, p<0.05) (Fig 6B). In contrast, blocking TGF-β did not have any significant effect (Fig 6B). Next, we explored the influence of IL-23, reported to play an important role in IL-22 + T cell commitment [24][25][26][27][28], as well as IL-1β. Blocking of IL-23, via the IL-12p40 subunit, significantly inhibited the frequency of IL-22 + T cells (11.9% vs. 5.4% IL-22 + T cells, p<0.05) (Fig 6C). However, blocking IL-1β did not have any significant effect (Fig 6C). Taken together, our data showed that 1,25D diff -DC-secreted TNF-α, IL-6 and IL-23 contributed to the development of IL-22 expressing T cells. To further characterize the Effect of 25D on differentiation and/or stimulation of monocyte-derived DCs. Monocytes were differentiated for six days using rGM-CSF and rIL-4 in media with 10% FCS in the presence (25D diff -DCs or 25D diff/stim -DCs) or absence (serum-DCs or 25D stim -DCs) of additional 25D (10 −7 M). Subsequently, 25D-DCs and serum-DCs were stimulated with TLR2/1L (1 μg/ml) or left untreated in the presence (25D diff/stim -DCs or 25D stim -DCs) or absence (serum-DCs or 25D diff -DCs) of additional 25D (10 −7 M), and cultured in fresh media with 10% FCS for 18-24 hours. TNF-α, IL-6, IL-1β, IL-12p40, IL-12p70, IL-23 and IL-10, as well as total and active TGF-β levels in culture supernatants were measured by ELISA or CBA (mean cytokine levels in ng/ml ± SEM, n = 4). Results shown in Fig 1 and Fig. 2 were conducted using cells from the same cell preparations of identical donors. *p<0.05, **p<0.01, # p 0.05 doi:10.1371/journal.pone.0130395.g002 Vitamin D-Treated DCs Promote IL-22 + T Cells individual and combined effects of TNF-α, IL-6 and IL-23 in this process, we performed additional blocking experiments. Blocking of IL-6 alone was more efficient than blocking TNF-α alone (12.6% vs 5.3% vs. 7.8%) (Fig 6D) and as potent as the combined blocking of TNF-α and IL-6 (12.6% vs 5.8%) (Fig 6D). In addition, blocking of IL-23 via the IL-12p40 subunit in combination with TNF-α/IL-6R-α was more potent than the individual effects and resulted in drastic inhibition of the differentiation of IL-22 + T cells (12.6% vs. 1.4% IL-22 + T cells) (Fig 6D). Of relevance, given the lack of a monoclonal blocking antibody specific for the IL-23p19 subunit, we cannot exclude that the inhibition of IL-12p70 signaling, due to blocking of the IL-12p40 subunit, has also an effect on inhibition of IL-22 + T cell differentiation in our setup. In summary, these data show that differentiation of IL-22-producing T cells by supernatants of 1,25D diff -DCs is dependent on TNF-α IL-6 and IL-23.

Discussion
In this study, we report that vitamin D treatment during differentiation of human DCs increased their ability to promote aspects of host protective immunity. A key finding was the strong enhancing effect of vitamin D on TNF-α, IL-6, IL-1β, as well as IL-23 secretion by DCs. Subsequently, 1,25D diff -DCs and serum-DCs were stimulated with TLR2/1L (1μg/ml), TLR4L (10 ng/ml) or CD40L (5 μg/ml) or left untreated, and cultured in fresh media with 10% FCS for 18-24 hours. TNF-α, IL-6, IL-1β, IL-12p40, IL-12p70, IL-23 and IL-10, as well as total and active TGF-β levels in culture supernatants were measured by ELISA or CBA (mean cytokine levels in ng/ml ± SEM, n of each condition indicated by numbers under the bars). *p<0.05, **p<0.01, ***p<0.001 doi:10.1371/journal.pone.0130395.g003 In contrast, we did not detect major differences in IL-12p70, IL-10 or TGF-β secretion. Together, these data show that vitamin D treatment of differentiating human DCs favors a pro-inflammatory cytokine profile, a fact that, even though reported for TNF-α, IL-6, IL-23 in the past [47,48,53,54], has not been interpreted as such [60]. Strikingly, the cytokines secreted by TLR2/1L-induced 1,25D diff -DCs were potent in driving differentiation of IL-22 + CD4 + T cells, demonstrating a novel role of vitamin D in regulating DC-mediated instruction of T cell responses. In addition, we demonstrate that the 1,25D diff -DC-mediated instruction of IL-22 + CD4 + T cells was dependent on TNF-α/IL-6 and IL-23. Our findings stand in accordance with several current reports using different models, including mouse and human. TNF-α and IL-6 alone or in combination induced IL-22 production from naïve T cells [23,24]. Moreover, collective evidence indicates that IL-23 contributes essentially to the induction of IL-22 production from different immune cells, including CD4 + T cells [24][25][26][27][61][62][63][64]. However, the exact role of different cytokine combinations in driving differentiation of IL-22 + T cell remains controversial [59], but it seems likely that this depends on the experimental setup and/or biological context. In our experiments, TGF-β, although secreted, seemed not to significantly regulate differentiation of IL-22-producing T cells [23,29,59], because neutralizing TGF-β in 1,25D diff -DC supernatants had no effect. Of note, we targeted the IL-12-/IL-23-shared p40 subunit as an IL-23p19 specific monoclonal blocking antibody was not available. Therefore, we are not able to rule out that IL-12p70 could also contribute to IL-22 + T cell differentiation. Of relevance, even though vitamin D treatment of differentiating DCs enhanced their secretion of IL-1β [65] and IL-23, both of which have been linked to the development and maintenance of Th17 cells [16][17][18], in our in vitro T cell differentiation model, we did not measure an increase in Th17 cells. However, it is noteworthy that the relative and absolute amounts of DC-secreted IL-1β and IL-23 detected in our cultures were 25 to 100 times lower than the amounts of recombinant cytokines used in previous protocols [16][17][18], which may explain the apparent differences.
Given the lack of an established pathogen-DC-T cell co-culture system to investigate differentiation of IL-22-producing cells from naïve T cells, we adopted a model, in which T cell are activated by CD3/CD28-coated beads and cultured with DC supernatants according to Duhen et al. [23], a key paper showing that TNF-α and IL-6 drive differentiation of IL-22-producing T cells. Therefore, our study does not experimentally take contact-dependent mechanisms into account, yet addressed differences in T cell instruction by vitamin D-treated vs. serum-DCs in a bystander fashion. Interestingly, even though DCs drove IL-22 + T cell differentiation in a classical mixed-lymphocyte reaction and in an APC-CD3-autologous T cell co-culture, direct contact of DCs and T cells seemed not to be required for this process [23,59]. In fact, cytokines were sufficient to drive IL-22 + T cell commitment in CD3/CD28-activated T cells [7,23,29,59], consistent with our presented data. Nevertheless, it will be interesting in future studies to investigate the effect of vitamin D on critical steps in DC antigen-presentation to naïve T cells using extracellular antigens, including specific antigen-binding and-uptake, as well as intracellular processing (involving degradation by proteasomes and lysosomal-associated proteases, loading on MHC molecules, etc.) and quality of MHC:peptide complexes, co-stimulation etc. [66,67]. Supernatants of TLR2/1-induced 1,25D diff -DCs and serum-DC were added to naïve CD4 + T cells activated with CD3/CD28-coated beads (as described in Fig 4). After five days, rIL2 was added to all cultures. On day 12, T cells were re-stimulated with PMA/Ionomycin for five hours, the last 2.5 hours of culture in the presence of Brefeldin A, in fresh media and intracellular cytokine expression of IL-22, IFN-γ or IL-17a was measured. Vitamin D-Treated DCs Promote IL-22 + T Cells  Fig 6. 1,25D diff -DC-supernatant mediated priming of IL-22-producing T cells is dependent on TNF-α IL-6 and IL-23. Supernatants of TLR2/ 1-stimulated 1,25D diff -DCs were added to naïve CD4 + T cells activated via CD3/CD28-coated beads (as described in Fig 4) in the presence or absence of During immune responses in human tissues, such as the skin, monocytes are recruited to the site of inflammation, where they locally differentiate into effector macrophages and DCs [68,69]. Thus, one could speculate that the vitamin D effect on differentiating DCs provides a mechanism, by which vitamin D amplifies the initial inflammatory immune responses in the context of infection. In contrast, when we stimulated already differentiated DCs in the presence of vitamin D, we observed a tendency towards decreased secretion of inflammatory cytokines consistent with previous reports [49,52]. Thus, one could speculate that under steady-state conditions, inflammatory responses by tissue DCs are suppressed by vitamin D, preventing exaggerated inflammation and tissue destruction. Moreover, we observed a dual pro-inflammatory/anti-inflammatory phenotype of vitamin D-treated DCs characterized by a tolerogenic expression pattern of cell surface molecules consistent with previous studies [47, 49, 51, 53-55, 70, 71], but also by a markedly enhanced secretion of pro-inflammatory cytokines. This could suggest that monocytes, within a continuous spectrum of DC-macrophage polarization, differentiate more towards a macrophage phenotype in the presence of vitamin D [72]. Besides, the dual inflammatory/anti-inflammatory phenotype implies that vitamin D promotes the initiation of an innate inflammatory response and at the same time could balance the acquired immune response. In this regard, an increased production of T cell IL-22 may result in restoration of tissue homeostasis and in combination with TNF-α in induction of antimicrobial peptides [3,5,73,74]. Generally, IL-22 promotes epithelial innate immune mechanisms, which can either be harmful or protective: IL-22 contributes to host defense against extracellular bacterial infections, tissue homeostasis and inflammation, in particular at epithelial barriers like bowel, lung and skin [75]. In detail, IL-22 has been described to have pro-and anti-inflammatory activities depending on the context, e.g. the specific tissue microenvironment, the infectious agent and the cytokine milieu, in which IL-22 is expressed [76]. For example, while IL-22 can act synergistically with IL-17a in promoting pathological airway inflammation [76], both contribute to protective antimicrobial peptide expression in the skin [63,77]. Also, the cosecretion of TNF-α and IL-22 was essential to trigger antimicrobial peptide expression in keratinocytes, and was shown to be the optimal combination for the skin immune response against Candida albicans in a 3D-skin model [3,7]. Therefore, especially the induction of 'polyfunctional' Th22 cells, T cells co-expressing IL-22 in conjunction with other cytokines, seems to be critical. We found that 1,25D diff -DCs promoted differentiation of total IL-22 + T cells, IL-22/ IFN-γ co-expressing T cells and an enhanced TNF-α secretion by T cells. Furthermore, we did not observe an increase in IL-4 expression and secretion, or IL-10 secretion. Of interest, compatible with two previous human studies, IL-4 was only co-expressed by a small fraction of IL-22 + T cells [58,59]. However, our findings on IL-4 and IL-10 stand in contrast to a mouse study, showing an increased IL-4 and IL-10 expression by murine CD4 + CD45RB high naïve T cells, isolated from spleens of OVA 323-339 -specific TCR-transgenic DO11.10 mice and stimulated with OVA-peptide loaded mitomycin-treated wild-type splenocytes, when 1,25D was added to the co-culture [58,59,78]. This probably reflects species differences, or differences in different monoclonal blocking antibodies as indicated. After five days, rIL2 was added to all cultures. On day 12, T cells were restimulated with PMA/ Ionomycin for five hours, the last 2.5 hours of culture in the presence of Brefeldin A, in fresh media and intracellular cytokine expression of IL-22, IFN-γ or IL-17a was measured. Cytokine secretion was evaluated after 18-24 hours without further addition of Brefeldin A. (A) Anti-TNF-α, anti-IL-6R-α (5 μg/ml each) or anti-TGF-β (10 μg/ml). T cell-derived IL-22 assessed by ELISA (mean of cytokine levels in ng/ml ± SEM, n = 5). (B) Anti-TNF-α, anti-IL-6R-α (5 μg/ml each) or anti-TGF-β (10 μg/ml). Vitamin D-Treated DCs Promote IL-22 + T Cells the experimental setup. Nevertheless, our data suggest that human vitamin D-treated DCs, by enhancing differentiation of IL-22 + and IL-22 + /IFN-γ + T cells, as well as TNF-α secretion by T cells, can contribute to host defense responses at epithelial surfaces [3].
The effect of vitamin D on DCs could cooperate with other vitamin D-mediated mechanisms that promote protective host defense responses. We and others have previously shown that vitamin D was required for the human host defense response against intracellular pathogens [44,45,79,80]. Moreover, vitamin D induced expression of the skin-homing receptor CCR10 on human T cells [23,40].
Without any doubt further investigations including animal models are needed to decipher the role of vitamin D in regulating IL-22 T cell responses in vivo. Nevertheless, our findings are relevant for the clinical use of vitamin D-treated GMP-produced DCs [81,82]. On one hand, if used as therapeutics in autoimmune diseases the effect of vitamin D on the pro-inflammatory cytokine production by DCs could aggravate aspects of inflammation. On the other hand, increased production of IL-22 and TNF-α could not only promote tissue homeostasis in inflammatory diseases [30,83,84], but also regulate host defense by inducing antimicrobial peptide production at epithelial barriers in infections [3,5,6].
Clinically, vitamin D deficiency has paradoxically not only been linked to poorer outcomes in autoimmunity, but also in infectious diseases [56]. Nevertheless, the fact that many immune mechanisms, which contribute to detrimental inflammation in autoimmunity are identical to those that mediate host protection, has raised questions, if and how, vitamin D can promote protective acquired immune responses in the context of infections. One explanation could be derived from the concept that the reported promoting effect of vitamin D on the innate macrophage response in the context of infection [44,45,79,80] simply outweighs potential inhibitory effects on the acquired response, and/or that vitamin D provides a negative feed-back loop on acquired immunity to limit excessive inflammation. However, in the present study, we provide evidence that human DCs differentiated in the presence of vitamin D do not solely exhibit an anti-inflammatory phenotype. In fact, they are superior to serum-DCs in secreting key host defense cytokines and promoting differentiation of IL-22-producing T cells in a bystander manner, thereby indicating that vitamin D promotes aspects of both pro-inflammatory and anti-inflammatory immune responses in humans. Freshly isolated naïve CD4 + T cells were activated with CD3/CD28-coated beads in the presence of TLR2/1-induced serum-DC (white bar) or 1,25D diff -DCs supernatant (black bar) or absence of supernatant (beads only control, grey bar). After five days, rIL2 was added to all cultures. On day 12, T cells were re-stimulated with PMA/Ionomycin in fresh media and counted with Trypan blue exclusion prior to intracellular cytokine staining. The asterisks directly above the bars indicate the p-values calculated in comparison to day 0 (mean of absolute cell counts per well ± SEM, n = 7). Ã p<0.05 and ÃÃ p<0.01 (EPS)