Dendritic cell (DC) manipulation by active vitamin D3: exploitation of molecular mechanisms for DC-based immunotherapies in type 1 diabetes

Vitamin D is a hormone that can be obtained from food (e.g. fatty fish and dairy products), but is mainly made in the skin where sunlight exposure converts it into its biologically active metabolite, 1,25-dihydroxyvitamin D3 (1,25(OH)2D3). The widespread expression of the vitamin D receptor (VDR) and vitamin D3-metabolizing enzymes in almost all cells of the immune system indicates a role for 1,25(OH)2D3 as a modulator of immune responses. In dendritic cells (DCs), 1,25(OH)2D3 interferes with the differentiation and maturation process resulting in the induction of a tolerogenic state. Tolerogenic DCs (tolDCs) are characterized by downregulated antigen-presentation, reduced costimulation and low pro-inflammatory cytokine production. As a result, tolDCs are poor T cell stimulators and induce T cell hyporesponsiveness. In addition, tolDC mediate a shift in T cell polarization from Th1/17 responses to more tolerogenic responses, with the induction of regulatory T cells (Tregs). On the other hand, 1,25(OH)2D3 are also able to directly modulate activated T cells. In these cells, 1,25(OH)2D3 decreases the production of pro-inflammatory cytokines and induces Tregs.

An important role for 1,25(OH)2D3 and its nuclear receptor in the development of type 1 diabetes (T1D) has been suggested by the discrepancy observed between absence of ligand and absence of receptor. Vitamin D3 deficiency in early life is associated with elevated risk for T1D, while the VDR null mouse model, lacking a functional VDR, do not present with aggravated disease. This difference indicates an important role for the unliganded VDR in the immune response. In the first part of this project, we investigated whether an unliganded VDR could affect the phenotype and function of murine bone marrow-derived DCs (BMDCs). Therefore, we studied myeloid BMDCs generated ex vivo from bone marrow precursors of VDR null, with a truncated, non-functional VDR, and VDR ∆AF-2 mice, with a mutated C‑terminal activation factor AF-2 domain thus rendering ligand-induced gene transcription impossible. To our surprise, the unliganded VDR did not affect BMDC phenotype or their T cell stimulatory capacity compared to VDR null BMDCs. These data indicate that an unliganded VDR does not elicit a more inflammatory phenotype in BMDCs.

Currently, no therapies for T1D are established which tackle the underlying immune attack. In general, immunotherapies focus on the inhibition of pathogenic immune cell activation and the (re-)establishment of self-tolerance against autoantigens. Given the interesting characteristics of Tregs and tolDCs to induce antigen-specific protection, cellular immunotherapies like adoptive T cell transfer and tolDC-based vaccines provide promising strategies for autoimmune diseases. In this regard, 1,25(OH)2D3 or its low‑calcemic analog TX527 are ideal for ex vivo modulation of autologous T cells and DCs.

Previously, our group showed successful ex vivo modulation of T cells isolated from healthy donors by 1,25(OH)2D3 as well as its analog TX527. However, for the clinical applicability of autologous adoptive T cell therapy, validation in samples from patients with T1D is required, especially as a defect in Treg function has been shown in T1D. Therefore, we further explored whether autologous T cells obtained from patients with T1D could be ex vivo modulated by 1,25(OH)2D3 and its analog TX527. Indeed, treatment with 1,25(OH)2D3 and TX527 increased the frequency of CD4+CD25highCD127low Tregs, with elevated CTLA-4 expression in this population. Moreover, exposure to 1,25(OH)2D3 and TX527 imprinted a unique homing signature and reduced the production of effector cytokines in the T cells isolated from T1D patients. In addition, both 1,25(OH)2D3 and TX527 promoted the formation of a stable Treg phenotype, which is essential for their possible clinical application as they will be reintroduced in an inflammatory microenvironment.

For the third part of this PhD project, we build further on a previous finding in our lab that demonstrated that the induction of the tolDC profile by 1,25(OH)2D3 is accompanied by an early and transcriptionally mediated metabolic reprogramming. We previously showed that glucose availability and glycolytic metabolism are crucial to induce and maintain tolDCs. This observation strokes with the emerging concept of immunometabolism which states that intracellular metabolism determines the phenotype and function of immune cells. In this regard, we aimed to find a primary target of vitamin D3 that could act as a metabolic switch in tolDCs. We identified the glycolytic enzyme PFKFB4 as strongly and directly upregulated by 1,25(OH)2D3 in differentiating monocytes. Pharmacological inhibition of PFKFB4 activity by the small molecule inhibitor 5-MPN could interfere with the increase in glucose metabolism, and more specifically glucose oxidation, which are hallmark features of tolDC. Moreover, PFKFB4 inhibition altered tolDC phenotype and their ability to induce functional Tregs. A better understanding of the metabolic pathways influenced by 1,25(OH)2D3 in human tolDCs will aid the development and improve the efficacy of tolDC-based immunotherapies.