Redirection of Human Autoreactive T-Cells Upon Interaction With Dendritic Cells Modulated by TX527, an Analog of 1,25 Dihydroxyvitamin D₃ Free

Human HLA-DR3 homozygous peripheral blood mononuclear cells were isolated by Ficoll gradient from HLA-typed buffy coats, obtained from healthy blood donors. Monocytes were obtained via positive selection on a MACS column after staining peripheral blood mononuclear cells with CD14-MicroBeads according to the supplier’s protocol (Milteny Biotec via CLB, Amsterdam, the Netherlands). Isolated monocytes (0.5 × 10⁶/ml, routinely >90% pure) were subsequently cultured for 6 days in RPMI 1640 medium (Gibco Life Technologies, Breda, The Netherlands) supplemented with 10% heat-inactivated FCS, 2 mmol/l glutamine, 100 IU/ml penicillin, 100 IU/ml streptomycin, 1,000 units/ml recombinant human IL-4 (Strathmann Biotec AG, Hannover, Germany), and 800 units/ml recombinant human granulocyte/macrophage colony-stimulating factor (GM-CSF) (Leucomax, Novartis Pharma, Arnhem, the Netherlands) in 24-well tissue culture plates (Costar, Cambridge, MA). Medium and cytokines were refreshed every 3 days. At day 6, when DCs had obtained an immature phenotype, DC1 maturation was induced by addition of 100 ng/ml Escherichia coli-derived lipopolysaccharide (LPS) (Sigma Aldrich Chemie, Zwijndrecht, The Netherlands), 1,000 units/ml recombinant human IFN-γ (Peprotech, Rocky Hill, NJ), and 800 units/ml GM-CSF. DCs were harvested after 48 h for further analysis.

In some experiments, macrophages were generated in parallel to DCs by culturing monocytes in RPMI 1640 medium supplemented with 10% FCS, 2 mmol/l glutamine, 100 IU/ml penicillin, 100 IU/ml streptomycin, and 50 ng/ml recombinant human M-CSF (Peprotech) in 24-well tissue culture plates. Medium and cytokines were refreshed at days 3 and 6.

1,25(OH)₂D₃ was obtained from J.P. Vandevelde (Duphar, Weesp, the Netherlands). The analog TX527 (19-nor-14,20-bisepi-23-yne-1,25(OH)₂D₃ (19,20) was synthesized by M. Vandewalle and P. De Clercq (University of Ghent, Belgium) and further obtained from Théramex S.A. (Monaco). Both 1,25(OH)₂D₃ and TX527 were used at a final concentration of 10⁻⁸ mol/l and refreshed at days 3 and 6, respectively. They were added to the culture medium either during the complete culture period from monocyte to immature DCs to mature DCs or only during the maturation phase.

For evaluating DCs’ surface expression, DCs were incubated for 30 min at 4°C with the following FITC- or PE-conjugated murine antibodies (all from Becton Dickinson, San Jose, CA): CD1a, CD14, CD83, CD86, CD80, CD54, HLA-DR, CD40, and control mIgG, all diluted in PBS containing 0.1% BSA. After extensive washing, DCs were subsequently analyzed on a FACS Calibur (Becton Dickinson).

Because mature DC1 cells are unresponsive to bacterial or viral compounds (21), 4 × 10⁵ DCs were incubated with 4 × 10⁵ CD154 expressing L cells (provided by Dr. C. van Kooten, Department of Nephrology, LUMC, the Netherlands, [22]) in the presence of 1,000 units/ml recombinant human IFN-γ to trigger cytokine production by untreated 1,25(OH)₂D₃ or TX527 modulated DCs. After 24 h, supernatants were collected and cytokine levels were determined by enzyme-linked immunosorbent assay (ELISA) by using commercial kits specific for IL-12p70 and IL-10 (R&D Systems, Minneapolis, MN).

Fluorescein-labeled zymosan particles (Saccharomyces cerevisiae; Molecular Probes Europe, Leiden, the Netherlands) were used to evaluate the phagocytosis capacity of variously treated DCs and macrophages. Briefly, the various cell suspensions were incubated for 2 h with fluoresceinated particles (1–100 particles/cell), after which the cells were washed and dissolved in 100 μl of Trypan Blue. After extensive washing, the cells were resuspended in 2% paraformaldehyde and analyzed by FACS. Phagocytosis capacity is expressed as percentage of fluorescein-positive cells.

To test the T-cell stimulatory capacities of nontreated-1,25(OH)₂D₃ or TX527-treated DCs, we used a GAD65-specific T-cell clones derived from a prediabetic individual, which recognizes naturally processed GAD65 antigen (Diamyd, Stockholm, Sweden) and the minimal GAD65 epitope p (339–352) in the context of HLA-DR3, as described previously (18). For proliferation or cytokine induction, 1 × 10⁴ T-cells were incubated with an equal number of variously treated DCs, diluted in Iscove’s modified Dulbecco’s medium (IMDM), supplemented with 2 mmol/l glutamine and 10% pooled human AB serum. T-cells were cultured for, respectively, 3 days (proliferation) or 24 h (ELIspot) in 96-well flat-bottom plates (Costar, Cambridge, MA). After 3 days, cultures were pulsed with [³H]thymidine (0.5 μCi/well) for another 18 h, after which ³H incorporation was measured by liquid scintillation counting. Results are expressed as means ± SD of triplicate wells. For detection of the T-cell cytokines IL-2, IFN-γ, IL-10, and IL-13, T-cells were harvested by gently rinsing the wells and washing the collected T-cells in a large volume of IMDM (23). T-cells were subsequently plated on anti-cytokine antibody-precoated ELISA plates and cultured for 2 h (IL-2), 5 h (IFNγ, IL-13), or overnight (IL-10) in IMDM supplemented with 2% pooled human AB serum at 37°C 5% CO₂. After lysis of the cells with ice-cold deionized water, the plates were washed with PBS/0.05% Tween-20 and incubated with biotinylated detector antibody for 1 h at 37°C, followed by a second incubation with gold-labeled anti-biotin antibody (GABA) for 1 h at 37°C. All antibodies were diluted in PBS/1% BSA. After extensive washing with PBS/0.05% Tween-20, the plates were developed according to the manufacturer’s protocol (U-CyTech, Utrecht, the Netherlands). Spots were counted on an Olympus microscope and analyzed with Olympus Micro Image 4.0 software (Paes Nederland, Zoeterwoude, the Netherlands). Results are expressed as means ± SD of triplicate wells.

The Mann-Whitney U test was used to compare results obtained from untreated DCs with either TX527- or 1,25(OH)₂D₃-treated DCs in various experiments. Significance was defined at the 0.05 level.

To compare the effects of TX527 and its parent compound 1,25(OH)₂D₃ on in vitro DC generation, we added TX527 or 1,25(OH)₂D₃ to monocytes that were cultured with GM-CSF and IL-4 for 6 days to generate immature DCs. On the basis of earlier findings (10,24), the concentration chosen in our study was 10⁻⁸ mol/l for both drugs, although TX527 was found to be equally potent at 10- to 100-fold lower concentrations (data not shown).

Addition of 1,25(OH)₂D₃ or TX527 impaired downregulation of the monocyte marker CD14 and interfered with upregulation of the skin DC marker CD1a when compared with untreated immature DCs (Fig. 1, left). Levels of MHC class II (HLA-DR) as well as costimulatory molecules CD86, CD80, CD54, and CD40 were unchanged compared with nontreated immature DCs (data not shown for CD80, CD54, and CD40). The DC maturation marker CD83 was obviously not expressed on nontreated TX527 or 1,25(OH)₂D₃-treated immature DCs.

After exposure to LPS and IFN-γ to induce DC maturation into DC1 (25), TX527- or 1,25(OH)₂D₃-treated DCs continued to express CD14 but failed to upregulate CD1a or CD83 (Fig. 1, right). In addition, TX527 or 1,25(OH)₂D₃ treatment resulted in a 10-fold lower expression of HLA-DR and CD86, whereas expression of CD80, CD40, or CD54 was not different from control mature DCs (data not shown).

This TX527- or 1,25(OH)₂D₃-induced altered DC phenotype was accompanied by typical morphological features (Fig. 2). Upon culture for 6 days with GM-CSF and IL-4, DCs became nonadherent and expressed protruding veils (data not shown). After the addition of LPS and IFN-γ, maturing DCs formed large clusters of nonadherent cells with even more pronounced cytoplasmic projections, whereas continuous addition of TX527 and, to a lesser extent, 1,25(OH)₂D₃ yielded small clusters of mostly adherent growing spindle-shaped cells.

In addition to the induction of phenotypic differences, prolonged TX527 or 1,25(OH)₂D₃ treatment was found to affect DC cytokine production induced by CD40 triggering via CD40L-expressing cells (Fig. 3). Untreated DCs express high levels of IL-12p70, whereas TX527 or 1,25(OH)₂D₃ treatment almost completely abrogated IL-12p70 production. The production of IL-10 was not different from untreated DC.

Because TX527 treatment in particular yielded adherent growing cells, which lack the expression of typical DC markers, e.g., CD1a and CD83, and more closely resemble macrophages, we compared the phagocytosis capacity of variously treated DCs and macrophages generated in parallel (Table 1). The phagocytosis capacity of TX527- or 1,25(OH)₂D₃-treated DCs did not differ from untreated mature DCs. All three types of DCs showed a lower capacity to take up fluorescent particles when compared with macrophages or immature DCs.

Taken together, TX527 and its parent compound 1,25(OH)₂D₃ affect the differentiation of DCs from monocytes and their subsequent maturation, yielding morphologically altered cells that lack DC-specific features and express lower levels of cell surface molecules involved in antigen presentation. Moreover, TX527- or 1,25(OH)₂D₃-modulated immature DCs were unable to convert into classical mature DC1 cells upon withdrawal of TX527 or 1,25(OH)₂D₃ and addition of LPS + IFN-γ, as determined by FACS analysis and analysis of IL-12p70 levels induced by CD40 ligation (data not shown), demonstrating that the effects of both drugs on DC differentiation are persistent.

Because lower levels of MHC class II and CD86 expression as well as reduced cytokine production by TX527- or 1,25(OH)₂D₃-treated DCs could affect the immune-stimulatory potential of these DCs, we tested proliferative and cytokine responses of a GAD65-specific T-cell clone upon coculture with variously treated DCs. Typically, this T-cell clone displays a Th0 cytokine profile when stimulated with peptides that contain the 339–352 sequence, i.e., IFN-γ, IL-10, and IL-13 (18). To exclude any (selective) effects of TX527 or 1,25(OH)₂D₃ on T-cells (8,24), we washed DCs extensively before addition to T-cell cultures.

Because immature DCs are most efficient in processing complete protein into peptides (26), we cocultured variously treated immature DCs with T-cells in the presence of GAD65 protein or the minimal peptide epitope 339–352 (Fig. 4). TX527 or 1,25(OH)₂D₃ pretreatment of DCs had no significant effect on GAD65-induced proliferative responses of the T-cells, whereas cytokine responses were completely blocked (P < 0.001). When the minimal peptide epitope was directly loaded onto DCs, proliferation of the T-cell clone was again not influenced by TX527 pretreatment of DCs. However, peptide-induced cytokine responses were differently affected by TX527 pretreatment, varying from complete abrogation of IL-10 (P < 0.001), significantly reduced IFN-γ (P = 0.04), and unchanged IL-13 production, whereas 1,25(OH)₂D₃ pretreatment of DCs was found to block IL-10 levels (P = 0.002) without significantly affecting IFN-γ or IL-13.

Mature DCs, however, do not have the capacity to process proteins into peptides efficiently. Therefore, the T-cell clone was incubated only with its minimal peptide epitope in the presence of variously treated DCs (Fig. 5). APCs that were generated from monocytes in the continuous presence of TX527 or 1,25(OH)₂D₃ induced significantly lower proliferation (P = 0.002 for TX527 vs. P = 0.009 for 1,25(OH)₂D₃) and remarkably affected the cytokine profile of the T-cell clone. Both IFN-γ and IL-10 were almost completely inhibited (P = 0.002), whereas IL-13 was not affected. Of both drugs, TX527 pretreatment induced the most pronounced inhibitory effects on the response pattern of the committed T-cell.

To evaluate the effects of TX527 or 1,25(OH)₂D₃ on the process of LPS- and IFN-γ-induced DC maturation (phase II), we added TX527 or 1,25(OH)₂D₃ during the last 2 days of culture of immature DCs generated in the presence of GM-CSF and IL-4 only. Introduction of TX527 or 1,25(OH)₂D₃ resulted in upregulation of CD14 on a low percentage of mature DCs (mean fluorescence intensity [MFI] 225.73 in TX527- and MFI 129.76 in 1,25(OH)₂D₃-treated DCs vs. MFI 49.38 in control untreated DCs). Addition of TX527 or 1,25(OH)₂D₃ did not affect expression of CD1a, CD83, CD80, CD86, CD40, CD54, or HLA-DR, because levels of these markers expressed on TX527- or 1,25(OH)₂D₃-modulated DCs were identical to levels expressed by nontreated mature DCs (data not shown). Nevertheless, DC cytokine profiles released upon CD40 triggering were significantly altered (Fig. 6). In contrast to untreated mature DCs, which express high levels of IL-12p70 and low levels of IL-10, TX527 or 1,25(OH)₂D₃ treatment during maturation resulted in DCs that were unable to secrete IL-12p70 but secreted four- to eightfold increased levels of IL-10 upon CD40 ligation. Thus, TX527, as well as the parent compound, redirects mature DCs, resulting in functionally different DCs compared with nontreated mature DCs.

Despite that TX527 or 1,25(OH)₂D₃ treatment did not affect peptide-induced proliferative responses of the T-cell clone, cytokine responses were affected to various extents (Fig. 7). Again, TX527-modulated DCs were more effective than 1,25(OH)₂D₃-treated DCs in redirecting cytokine responses, as demonstrated by a complete reduction in IL-10 (P < 0.001) and significant reduction in IFN-γ (P = 0.002) release. 1,25(OH)₂D₃-pretreated DCs significantly affected IL-10 and IL-13 (P = 0.002) but not IFN-γ production.

In this study, we demonstrated that TX527 is a potent 1,25(OH)₂D₃ analog that is capable of targeting DCs at various stages throughout their in vitro generation from monocytes. TX527 was found to be biologically active in a broader range (10⁻⁷ to 10⁻¹⁰ mol/l) when compared with its parent compound 1,25(OH)₂D₃ (10⁻⁷ to 10⁻⁸ mol/l), yielding DCs with the highest potential to alter the cytokine profile of a committed autoreactive T-cell clone. This modulating effect was mediated via the antigen-presenting DCs, because TX527 or 1,25(OH)₂D₃ was not added during T-cell cultures.

Most classical immunomodulators and immune suppressants modulate T-cells through direct interaction with T-cell proliferation or cytokine production. Very few strategies also affect antigen-presenting DCs, the central cell of the immune system able to polarize the immune system into different directions, e.g., Th1 versus Th2 (27). The active form of vitamin D₃, 1,25(OH)₂D₃, is one of the few products available (10,12), together with glucocorticoids (28–31), that is able to change DC differentiation and function in vitro.

In the present study, we demonstrated that a novel nonhypercalcemic 1,25(OH)₂D₃ analog, TX527, is able to modify significantly the phenotype of DCs as characterized by major changes in cell surface marker expression, morphology, and cytokine profile. In contrast to another study (12), we demonstrated that TX527 or 1,25(OH)₂D₃ treatment does not simply abrogate differentiation and maturation but leads to the emergence of a different cell that lacks typical DC features. Moreover, these modulated DCs were not able to equalize the high phagocytosis capacity that is observed for untreated macrophages or immature DCs. This is a strong argument against the postulated 1,25(OH)₂D₃-mediated arrest in the immature stage of DC development (11). It also confirms that 1,25(OH)₂D₃ or TX527 does not simply mediate differentiation toward macrophages (12). In the past, however, a clear, positive effect of 1,25(OH)₂D₃ on the differentiation of macrophages was shown under different culture conditions (32). Monocytes may be considered as relative immature precursors with multiple differentiation potentials, depending on the microenvironment during differentiation (33). In vitro, cytokines and inflammatory substances play an important role in the final determination of whether monocytes will acquire DC, macrophage, osteoclastic, or other characteristics and functions (34,35). Additional analysis of phenotype and functionality will be necessary to characterize fully the cell type generated by continuous presence of TX527 or 1,25(OH)₂D₃.

In vivo, 1,25(OH)₂D₃ analogs induce an immune shift that is characterized by downregulation of Th1 cytokines and upregulation of Th2 cytokines in response to pancreatic autoantigens such as GAD65 (16). Our in vitro data on committed human autoreactive T-cells indicate that TX527 treatment of DCs indeed inhibited IFN-γ release but did not increase Th2-linked IL-13 production. It is conceivable that this partial discordance with the in vivo data results from a direct action of 1,25(OH)₂D₃ on T-cells during in vivo treatment, because 1,25(OH)₂D₃ was shown to promote directly Th2-associated cytokine production in the absence of APCs (8).

Another important finding in our study is the redirection by TX527-modulated DCs of committed T-cells toward a different cytokine profile without completely blocking their proliferative capacity. The persistence of the proliferative capacity suggests that the autocrine stimulation of IL-2 was unaffected. Indeed, IL-2 production was not changed from that of autoreactive T-cells stimulated with untreated DCs (data not shown). This finding indicates that for complete blockade of committed autoreactive T-cells, a combined immunosuppressive therapy that targets both IL-2-dependent (e.g., cyclosporine, tacrolimus, or dacliximab [36]) and -independent pathways (TX527 or 1,25(OH)₂D₃) of T-cell activation would be preferred. This is in accordance with data on prevention of recurrent autoimmune disease after syngeneic islet transplantation in diabetic NOD mice treated with a combination of a vitamin D₃ analog and cyclosporine, whereas the individual drugs were less effective (15).

In addition, treatment with 1,25(OH)₂D₃ in combination with the IL-2-independent drug mycophenolate mofetil was recently shown to induce transplantation tolerance associated with an increased frequency of regulatory T-cells that bear the IL-2 receptor (CD152⁺CD4⁺CD25⁺ [17]). Mycophenolate mofetil is a selective lymphocyte inhibitor that blocks the progression of T-cells into the S phase without affecting the IL-2 pathway (24). The IL-2 pathway is known to be critical in the development of CD25-expressing regulatory T-cells (37) that mediate transplantation tolerance (17) and control autoimmunity in NOD mice (38).

1,25(OH)₂D₃- or analog-induced changes in cytokine profile of committed T-cells may in vivo result in impaired autoimmune effector cells. In addition, naive T-cells were shown to become hyporesponsive through priming by 1,25(OH)₂D₃-modulated DCs (10,12). When extrapolating these in vitro data to the in vivo situation, where the generation of suppressor cells by 1,25(OH)₂D₃ and its analogs has been confirmed in different settings (4,17), these studies collectively define 1,25(OH)₂D₃, especially its nonhypercalcemic analogs, as ideal immunosuppressants in autoimmunity and as candidate drugs to be combined with typical anti-T-cell immunosuppressants to intervene in hyperactivated immune systems, e.g., in islet transplantation. We conclude that a new structurally modified analog of vitamin D₃, TX527, modulates morphological features and cytokine production of human DCs, yielding altered DCs that redirect an already committed T-cell clone toward a cell with a reduced pathogenic potential.

This work was supported by Grant 2000.00.056 and 1998.01.001 from the Diabetes Fonds Nederland, Grant 99/10 from the Geconcerteerde OnderzoeksActie, and Grant 3.0332.98 from the Flemish Research Foundation (Fonds voor Wetenschappelijk Onderzoek).

Erica F.M. Meulenberg is greatly acknowledged for excellent technical assistance.