25(OH)D₃ (BML-DM-100-0001) and 1,25(OH)₂D₃ (BML-DM200-0050) were from Enzo Life Sciences, Inc., Ann Arbor, MI, USA. Stock solutions of 2.5 mM 25(OH)D₃ and 2.4 mM 1,25(OH)₂D₃ were prepared in anhydrous (≥99.5%) ethanol and stored at −20°C. To determine 1,25(OH)₂D₃ in the supernatants, we used the 1,25-dihydroxy vitamin D EIA kit (AC-62F1) from IDS, Tyne and Wear, UK according to the manufacturer’s instructions.

Mononuclear cells from blood were isolated by Lymphoprep (Axis-Shield, Oslo, Norway) density-gradient centrifugation using SepMate™ tubes (86460, Stemcell Technologies, Grenoble, France) from healthy donors after obtaining informed, written consent in accordance with the Declarations of Helsinki principles for research involving human objects. The study was approved by The Committees of Biomedical Research Ethics for the Capital Region in Denmark (H-16033682). Naïve CD4⁺ T cells were isolated and cultured as previously described (47). The purified naïve CD4⁺ T cells were cultured in serum-free X-VIVO 15 medium (BE02-060F, Lonza, Verviers, Belgium) at 37°C, 5% CO₂ at a cell concentration of 1 × 10⁶ cells/ml in flat-bottomed 24-well tissue culture plates (142475) from Nunc and stimulated with Dynabeads human T-activator CD3/CD28 beads (111.31D, Life Technologies, Grand Island, NY, USA) at a cell to bead ratio of 5:2 for up to 3 days. Cells present in the culture after 3 days were defined as activated T cells. In some experiments, 25(OH)D₃, 1,25(OH)₂D₃, or recombinant human IL-12 (219-IL, R&D Systems) was added at the indicated concentrations to the medium during the stimulation period. In Th1 polarization studies, purified naïve CD4⁺ T cells were cultured and stimulated as described above in the presence of recombinant human IL-12 (10 ng/ml) plus human IL-4 antibody (4 µg/ml, MAB204, R&D Systems). In some experiments, cells were counted after 72 h of activation using the automated cell counter NucleoCounter^® NC-100™ from Chemometec.

Monocytes used for generation of monocyte-derived DC were isolated from blood mononuclear cells using EasySep Human Monocyte Enrichment Kit (19059, Stemcell Technologies) according to the manufacturer’s protocol. Following isolation, monocytes were cultured in flat-bottomed six-well tissue culture plates (140675, Nunc) at a cell concentration of 5 × 10⁵ cells/ml for 6 days in medium (RPMI-1640, R5886, Sigma-Aldrich) with penicillin, streptomycin, l-glutamine, and 10% heat-inactivated FBS (10082-147, Gibco) supplemented with GM-CSF and IL-4 (both 50 ng/ml, AF-HDC, PeproTech) with a re-supplementation of medium and cytokines on day 3. On day 5, the differentiated DCs were supplemented with GM-CSF (50 ng/ml) and treated with either heat-killed M. tuberculosis (HKMT) (10 µg/ml, tlrl-hkmt, InvivoGen), Pam3CSK4 (300 ng/ml, tlrl-pms, Invivogen) as TLR2 ligand, lipopolysaccharides (LPS) from E. coli (50 ng/ml, L 5668, Sigma) as TLR4 ligand, or left untreated for 24 h, washed, and resuspended in X-VIVO 15 medium for use in mono- and co-cultures or in RPMI-1640 medium for use in 1,25(OH)₂D₃ titration experiments. For monocultures, 2.5–5 × 10⁵ cells/ml DC were plated in flat-bottomed 24-well tissue culture plates for 24 h in the presence or absence of 25(OH)D₃ and with or without recombinant human IFNγ (R&D Systems) or with increasing concentrations of 1,25(OH)₂D₃ for titration studies. For co-culture studies, naïve human CD4⁺ T cells were purified from a different donor as described above and co-cultured with 1 × 10⁵ cells/ml DC in a ratio of 1:10 DC to T cells in flat-bottomed 24-well tissue culture plates for another 6 days in X-VIVO 15 medium.

IFNγ and IL-13 concentrations in the supernatants were determined by ELISA according to the manufacturer’s protocol (Ready-Set-Go; eBioscience).

mRNA expression was measured by real-time RT-PCR (QPCR). For this, cells were lysed in TRI reagent (T9424, Sigma-Aldrich) followed by addition of 1-bromo-3-chloropropane (B9673, Sigma Aldrich) to separate the sample into an aqueous and an organic phase. The RNA was precipitated from the aqueous phase using isopropanol supplemented with glycogen (10814-010, Invitrogen), washed with ethanol, and dissolved in RNase free water. Equal amounts of total RNA were used for complementary DNA (cDNA) synthesis using the high-capacity RNA-to-cDNA™ Kit from Applied Biosystems (4387406) according to manufacturer’s protocol. For QPCR, 12.5 ng cDNA was mixed with TaqMan^® Universal Master Mix II with UNG (4440038, Applied Biosystems), the target primer, and the eukaryotic 18 S rRNA primer (1509311, Applied Biosystems). The samples were run on a Stratagene Mx3000P™ real-time PCR machine (Agilent Technologies). The thermal profile was set to 2 min at 50°C, a 10 min hot start at 95°C, followed by cycles of 15 s at 95°C and 1 min at 60°C. Signal intensity was measured at the end of the 60°C step, and the threshold cycle values were related to eukaryotic 18S rRNA. The following primers were used; IFNG (Hs00989291_m1), CYP27B1 (Hs01096154_m1), vitamin D receptor (VDR) (Hs01045840_m1), CYP24A1 (Hs00167999_m1), TBX21 (Hs00203436_m1), IL-12RB2 (Hs00155486_m1), CXCL9 (Hs00171065_m1), CXCL10 (Hs00171042_m1), and CAMP (cathelicidin antimicrobial peptide) (Hs00189038_m1) all from Applied Biosystems.

For Western blot analysis, whole cell lysates were obtained by treatment of the cells with lysis buffer (50 mM Tris pH 7.5, 150 mM NaCl, 1 mM MgCl₂) supplemented with 1% Triton X-100, 1× Protease inhibitor cocktail (5872S, Cell Signaling Technology, Beverly, MA, USA) and 5 mM EDTA. The samples were run under reducing conditions on polyacrylamide gels for 2 h at 120 volt in 1× NuPAGE MOPS SDS Running buffer (K855, Amresco, Solon, OH, USA) in an XCell SureLock^® Mini-Cell Module from Life Technologies. The proteins were transferred to nitrocellulose membrane sheets (LC2001, Life Technologies) in 1× NuPAGE Transfer buffer (NP0006-1, Life Technologies) supplemented with 10% ethanol for 60 min at 50 V in an XCell II™ Blot Module from Life Technologies. The membranes were subsequently blocked for 60 min in Tris-buffered saline supplemented with 5% milk powder (70166, Sigma-Aldrich) and 0.1% Tween 20 (P1379, Sigma-Aldrich) and incubated at 4°C for 24 h with primary antibodies diluted in Tris-buffered saline supplemented with 5% bovine serum albumin (A4503, Sigma-Aldrich) and 0.1% Tween 20. The membranes were washed and the proteins visualized following 60 min incubation at room temperature with secondary HRP-rabbit anti-mouse Ig and HRP-swine anti-rabbit Ig using ECL (RPN2232, Sigma Aldrich) technology. The anti-phospho-STAT1 (9171) and phospho-STAT4 (5267) antibodies were from Cell Signaling Technologies, the anti-STAT1 (SC-346), STAT4 (SC-486), and VDR (SC-13133) antibodies were from Santa Cruz Biotechnology, Santa Cruz, CA, USA and the anti-GAPDH (Ab9485) was from Abcam, Cambridge, UK. For band density quantification, ECL-exposed sheets were analyzed in a ChemiDoc MP Imaging System from Bio-Rad.

Data are shown as mean ± SEM. Statistical analyses were performed using Student’s t-test with a 5% significance level, paired observations, and equal variance. * indicates p ≤ 0.05.

Most previous studies on the effect of vitamin D on IFNγ production in human T cells have been performed using heterologous cell populations or purified CD4⁺ T cells in cell culture medium supplemented with serum and non-physiological, high concentrations (1–100 nM) of the active form of vitamin D (1,25(OH)₂D) (41–45). Various sources of serum contain different concentrations of the inactive form of vitamin D (25(OH)D), 1,25(OH)₂D, and the vitamin D-binding protein, all of great importance when investigating the effect of vitamin D on T cells (24, 46). To study the direct effect of 25(OH)D₃ and 1,25(OH)₂D₃ on IFNγ production in human Th cells under strictly defined conditions, we isolated naïve CD4⁺ T cells and stimulated them with human T-activator CD3/CD28 beads in serum-free medium in the absence or presence of 25(OH)D₃ or 1,25(OH)₂D₃. We found that 1,25(OH)₂D₃ strongly inhibited IFNγ production (~15-fold) at physiological concentrations (60–110 pM) (Figure 1A). Furthermore, we found that 25(OH)D₃ at physiological concentrations (50–125 nM) inhibited IFNγ production to the same degree as 1,25(OH)₂D₃ (Figure 1A), which confirms that T cells have the ability to convert the inactive 25(OH)D₃ to the active 1,25(OH)₂D₃ (46, 47). It is well known that anti-CD3/CD28 bead stimulation promotes expansion of CD4⁺ T cells (48). To investigate whether the lower levels of IFNγ in vitamin D-treated cells could be attributed to fewer cells, we stimulated CD4⁺ T cells for 72 h in the presence or absence of 100 nM 25(OH)D₃ and subsequently counted the cells using an automated cell counter. At this time point, we found that there were approximately 1.5-fold more CD4⁺ T cells in the cultures compared to the initial cell number independently of the absence or presence of vitamin D (Figure 1B). To exclude a general toxic effect of vitamin D on the T cells, we subsequently measured secretion of IL-13 in parallel with IFNγ. We found that vitamin D increased IL-13 secretion in parallel with its inhibition of IFNγ secretion (Figure 1C). Thus, the decreased levels of IFNγ were not caused by a reduced cell number or a toxic effect of vitamin D but most likely by a specific effect of vitamin D on IFNγ secretion. Previous studies have reported that 1,25(OH)₂D₃ inhibits production of IFNγ and augment the production of IL-4, and based on these observations it has been concluded that vitamin D restrains Th1 differentiation and promotes Th2 differentiation (41–45). However, to our knowledge, the effect of physiological concentrations of vitamin D on IFNγ production in T cells activated under Th1-inducing conditions has not been reported. To study this, we activated naïve CD4⁺ T cells in the absence or presence of IL-12 and 25(OH)D₃. In the absence of 25(OH)D₃, IL-12 increased IFNγ production (Figure 1D). Interestingly, although 25(OH)D₃ at physiological concentrations did reduce IFNγ production in IL-12 treated cells, these cells still produced equal amounts of IFNγ as T cells activated in the absence of 25(OH)D₃ and IL-12 (Figure 1D). From these experiments, we could conclude that 1,25(OH)₂D₃ inhibits the production of IFNγ in CD4⁺ effector T cells but that IL-12 partially rescues IFNγ production in the presence of vitamin D allowing production of similar amounts of IFNγ as in T cells activated in the absence of vitamin D and IL-12.

One way for IL-12 to counteract vitamin D-mediated inhibition of IFNγ production in CD4⁺ T cells could be by limiting the ability of the cells to convert 25(OH)D₃ to 1,25(OH)₂D₃. To investigate this, we studied whether IL-12 affected the expression of CYP27B1, the enzyme responsible for converting 25(OH)D₃ to 1,25(OH)₂D₃. We activated naïve CD4⁺ T cells in the presence of 100 nM 25(OH)D₃ in the absence or presence of IL-12 and measured the expression levels of CYP27B1 mRNA in the cells. We found that IL-12 did not influence the expression of CYP27B1 (Figure 2A). Furthermore, we directly measured the production of 1,25(OH)₂D₃ in the supernatants and found that IL-12 did not affect the production of 1,25(OH)₂D₃ (Figure 2B). Another way IL-12 might counteract vitamin D-mediated inhibition of IFNγ production in CD4⁺ T cells could be by inhibiting the expression or function of the VDR. However, we found that IL-12 slightly increased VDR expression on the mRNA level (Figure 2C). To further investigate whether IL-12 affected VDR expression, we determined the effect of IL-12 on VDR protein levels. We did not observe any significant effect of IL-12 on VDR expression at the protein level (Figure 2D). CYP24A1 is the enzyme that initiates degradation of 1,25(OH)₂D₃, and CYP24A1 expression is dependent on functional 1,25(OH)₂D₃/VDR complexes. Consequently, we measured the CYP24A1 levels in cells treated with or without IL-12. We found that the cells produced very high amounts of CYP24A1 in the presence of 100 nM 25(OH)D₃ (approximately 100.000-fold upregulated compared to unstimulated T cells) although IL-12 slightly reduced CYP24A1 expression (Figure 2E). Taken together, these experiments indicated that IL-12 does not counteract vitamin D-mediated inhibition of IFNγ production in CD4⁺ T cells by reducing their ability to produce 1,25(OH)₂D₃ or by reducing their VDR expression or function.

Due to the inhibition of IFNγ production, it has been suggested that vitamin D impedes differentiation of Th1 cells. However, whether vitamin D actually inhibits Th1 differentiation is still not known. To directly study whether vitamin D affects Th1 differentiation, we stimulated naïve CD4⁺ T cells under classical Th1 conditions with IL-12 and anti-IL-4 in the absence or presence of 25(OH)D₃ and measured the expression levels of TBX21 and IL-12Rβ2. Compared to naïve T cells, cells activated for 24 h in the absence of 25(OH)D₃ upregulated TBX21 approximately 120-fold. Although not significantly, 25(OH)D₃ slightly reduced TBX21 upregulation at 24 h but had no inhibitory effect on TBX21 expression after 48 and 72 h (Figure 3A). Likewise, IL-12Rβ2 expression was slightly reduced after 24 h but unaffected by 25(OH)D₃ after 48 and 72 h of stimulation (Figure 3B). While vitamin D did not affect TBX21 and IL-12Rβ2 expression in cells stimulated for 48 and 72 h, it clearly reduced IFNγ expression in the same cells (Figure 3C). Thus, these experiments indicated that vitamin D does not block differentiation of Th1 cells but directly affects the transcription of the IFNG gene. Although reduced, IFNγ expression in Th1 cells treated with vitamin D was still highly upregulated (1000- to 2000-fold, Figure 3C) and they produced high levels of IFNγ (Figure 1D). In line with the strong production of IFNγ, we found that vitamin D did not affect STAT1 phosphorylation significantly during T cell activation (Figure 3D). Likewise, STAT4 phosphorylation was not significantly affected of vitamin D (Figure 3E).

IFNγ itself plays an important role during Th1 differentiation as STAT1 activated by IFNγ further activates transcription of TBX21 and IL12Rβ2 (8, 9). We have previously demonstrated that naïve human CD4⁺ T cells do not express the VDR but that they start to express it 24–48 h after TCR/CD28 stimulation (46, 49). This suggests that vitamin D cannot affect IFNγ production in the early stages of T cell activation due to the lack of VDR. To study the effect of 25(OH)D₃ and Th1-inducing conditions on the early kinetics of VDR expression, we stimulated naïve CD4⁺ T cells for 12, 24, and 48 h in the presence or absence of 25(OH)D₃ and Th1-inducing conditions. We could confirm that only very low levels of VDR are present after 12 and 24 h of stimulation, and that substantial VDR expression levels are found after 48 h of stimulation (Figure 4A). In agreement, we found that at 12 h of stimulation, vitamin D did not influence IFNγ production neither in cells activated in the absence or presence of Th1 conditions (Figure 4B). In contrast, Th1 conditions acted rapidly with increased production of IFNγ already 12 h after stimulation (Figure 4B). After 24 h of stimulation, a minor inhibitory effect of vitamin D on IFNγ production was seen (Figure 4C), and after 48 h vitamin D substantially inhibited IFNγ production (Figure 4D). These experiments indicated that IL-12-mediated signaling is initiated rapidly after initial T cell stimulation with increased IFNγ production as a consequence, and that this takes place before the VDR is expressed at sufficiently high levels to inhibit IFNγ production. These observations support that CD4⁺ T cells activated under Th1-inducing conditions start differentiation toward Th1 cells even in the presence of vitamin D.

Activation of naïve TB-specific T cells depends on the migration of infected DC from the lungs to the mediastinal lymph nodes where the DC present TB-derived antigens for the T cells (5–7). It has been suggested that antigen-presenting cells mediate the vitamin D-induced inhibition of IFNγ production in T cells (36, 38, 40), and the absence of antigen-presenting cells in our experimental set up could maybe explain why we observed IFNγ production in purified T cells stimulated in the presence of vitamin D and IL-12. To investigate this possibility, we cultured naïve CD4⁺ T cells with allogeneic DC for 6 days with or without 25(OH)D₃ in the absence or presence of Th1-inducing conditions and subsequently measured IFNγ expression and production. Vitamin D strongly inhibited IFNγ mRNA expression (~15-fold) (Figure 5A) and production (~20-fold) (Figure 5B) in the absence of Th1-inducing conditions. Th1-inducing conditions increased IFNγ expression and production and counteracted the inhibitory effect of vitamin D. Thus, vitamin D only inhibited IFNγ mRNA expression ~2-fold and IFNγ production ~3-fold in the presence of Th1-inducing conditions (Figures 5A,B). Interestingly, IFNγ expression and production in cells cultured with vitamin D under Th1-inducing conditions were at least as high as in cells cultured without vitamin D and Th1-inducing conditions (Figures 5A,B). This indicated that DC do not shut down IFNγ production in T cells in the presence of vitamin D. Taken together, we could conclude that vitamin D strongly inhibits IFNγ expression and production in T cells in the absence of Th1-inducing conditions, and that Th1-inducing conditions partially rescue IFNγ production in the presence of vitamin D independently on the presence or absence of DC.

Cellular responses mediated by IFNγ are primarily caused by gene expression modulation. Primary IFNγ-responsive genes are induced early by binding of STAT1 dimers to gamma-activating sequences in the promoter of target genes such as IRF1, CXCL9, and CXCL10 (50). To study whether the levels of IFNγ produced in Th1 cells in the presence of vitamin D were sufficient to affect IFNγ-responsive genes, we measured the expression of CXCL9 and CXCL10 in the same co-cultures as described above. We found that vitamin D reduced CXCL9 and CXCL10 expression ~110 and ~25-fold in Th0 cells, respectively (Figures 5C,D). In contrast, vitamin D only reduced CXCL9 and CXCL10 expression ~2- and ~3-fold in Th1 cells compared to untreated Th0 cells. This demonstrated that the cellular responses to the levels of IFNγ produced in Th1 cells in the presence of vitamin D almost equaled the responses to the levels of IFNγ produced in Th0 cells in the absence of vitamin D.

As the importance of IFNγ in the immune response to mycobacteria is well established, it is puzzling why vitamin D, associated with a beneficial role in TB prevention and treatment, reduces IFNγ production. Cathelicidin plays an important role in the killing of especially intracellular pathogens such as M. tuberculosis in macrophages (51). Furthermore, it has been shown that cathelicidin expression is dependent on vitamin D, and that IFNγ increases the synthesis of cathelicidin in macrophages (21–23). To study how cathelicidin is regulated in DC, we cultured naïve CD4⁺ T cells with allogeneic DC for 6 days as described above with or without 25(OH)D₃ in the absence or presence of Th1-inducing conditions and subsequently measured cathelicidin expression. We found that cathelicidin upregulation was absolutely dependent on vitamin D but independent of Th1-inducing conditions (Figure 5E). This indicated that vitamin D without IFNγ is sufficient to induce cathelicidin upregulation, and that IFNγ without vitamin D does not affect cathelicidin expression in DC.

Previous studies have demonstrated that IFNγ acts in synergy with TLR2 ligands to upregulate cathelicidin expression in human monocytes and macrophages (21, 22). To determine how vitamin D, IFNγ, and HKMT/TLR2 ligands affect cathelicidin expression in DC, we incubated DC in the absence or presence of 25(OH)D₃, IFNγ, and HKMT or Pam3CSK4, and measured the expression of cathelicidin. Surprisingly, we found that HKMT and Pam3CSK4 significantly downregulated cathelicidin expression both in the absence and presence of vitamin D (Figure 6A). In contrast, vitamin D strongly upregulated cathelicidin expression both in the absence and presence of HKMT or Pam3CSK4. Although HKMT and Pam3CSK4 downregulated cathelicidin expression ~13- and ~10-fold in the presence of vitamin D, respectively, the expression of cathelicidin was still ~215- and ~120-fold higher in DC stimulated with HKMT/Pam3CSK4 in the presence of vitamin D than in the absence of vitamin D (Figure 6A). IFNγ by itself did not affect cathelicidin expression.

To determine whether the HKMT/Pam3CSK4-induced inhibition of cathelicidin expression was caused by a reduced ability to produce or respond to 1,25(OH)₂D₃, we incubated DC in the absence or presence of 25(OH)D₃, IFNγ, and HKMT/Pam3CSK4 and measured the expression of CYP27B1, VDR, and CYP24A1, and the production of 1,25(OH)₂D₃. We found that HKMT/Pam3CSK4 increased CYP27B1 expression, especially in the presence of 25(OH)D₃ (Figure 6B). In accordance, although untreated DC produced high amounts of 1,25(OH)₂D₃, HKMT/Pam3CSK4 increased the 1,25(OH)₂D₃ production four- to fivefold (Figure 6C). In addition, HKMT slightly increased VDR expression in the presence of 25(OH)D₃ (Figure 6D). This suggested that HKMT/Pam3CSK4-treated DC should be at least as efficient as untreated DC to respond to 1,25(OH)₂D₃. This was supported by the expression of comparable levels of the 1,25(OH)₂D₃/VDR-dependent CYP24A1 in DC treated or untreated with HKMT/Pam3CSK4 in the presence of 25(OH)D₃ (Figure 6E). These observations indicated that the inhibitory effect of HKMT/Pam3CSK4 on cathelicidin expression was caused by a vitamin D/VDR-independent mechanism. To test this hypothesis, we incubated DC in the absence or presence of HKMT or Pam3CSK4 and increasing concentrations of 1,25(OH)₂D₃. In all concentrations of 1,25(OH)₂D₃ tested, we found that HKMT/Pam3CSK4 inhibited cathelicidin expression (Figure 6F). Interestingly, stimulation with the more complex HKMT resulted in a ~23-fold reduction in cathelicidin expression, whereas the pure TLR2 ligand Pam3CSK4 caused an ~8-fold reduction in cathelicidin expression at all concentrations of 1,25(OH)₂D₃ tested. This suggested that HKMT affected the DC through additional receptors than TLR2. In agreement, we found that the TLR4 ligand LPS also inhibited cathelicidin expression in DC (data not shown). In contrast to the effect on cathelicidin expression, HKMT/Pam3CSK4 did not affect CYP24A1 expression, supporting that HKMT/Pam3CSK4 did not inhibit the ability of DC to respond to 1,25(OH)₂D₃/VDR complexes (Figure 6G). Taken together, these data indicated that M. tuberculosis strongly inhibits cathelicidin expression in DC through TLR signaling but that vitamin D increases cathelicidin expression and thereby more than overcome the inhibitory effect of M. tuberculosis.