Journal of Steroid Biochemistry and Molecular Biology 188 (2019) 103–110

Contents lists available at ScienceDirect

Journal of Steroid Biochemistry and Molecular Biology

journal homepage: www.elsevier.com/locate/jsbmb

Vitamin D-modulated dendritic cells delay lethal graft-versus-host disease T

through induction of regulatory T cells
An-Sofie Vanherwegen, Dana Paulina Cook, Gabriela Bomfim Ferreira, Conny Gysemans1,
⁎
Chantal Mathieu ,1
Clinical and Experimental Endocrinology, KU Leuven, Belgium

A R T I C LE I N FO A B S T R A C T

Keywords: Graft-versus-host disease (GVHD) is the most lethal complication after allogeneic bone marrow transplantation
Vitamin D (allo-BMT). Current approaches to prevent GVHD rely on donor lymphocyte/T cell depletion or general im-
GVHD munosuppression, leading to opportunistic infections and cancer relapse. Tolerogenic dendritic cells can induce
Tolerogenic dendritic cells regulatory T cells (Tregs) with the ability to suppress inflammation and prevent transplant rejection, making
Regulatory T cells
them an attractive cellular therapy to control GVHD. Active vitamin D (1α,25-dihydroxyvitamin D3;
1α,25(OH)2D3) promotes the generation of tolerogenic dendritic cells (1,25D3-DCs). This study aimed to de-
termine the ability of ex vivo generated 1,25D3-DCs to trigger the expansion of Tregs that are able to control
lethal xenogeneic GVHD in humanized NOD/LtSz-PrkdcscidIL2rγtm1Wjl (NSG) mice. We demonstrate that 1,25D3-
DCs express lower levels of HLA-DR and costimulatory molecules, such as CD80 and CD86, and produce higher
levels of IL-10 and TNF-α and lower amounts of IL-12, compared to vehicle-treated DCs. Moreover, these cells
express increased levels of various co-inhibitory molecules such as PD-L1 and ILT-3 and the glycoprotein CD52
that is known to suppress T cell activation. Consequently, 1,25D3-DCs are poor stimulators of alloantigen-primed
T cells, but foster the generation of antigen-specific suppressive Tregs. When adoptively transferred in huma-
nized NSG mice, these 1,25D3-DC-induced Tregs delayed GVHD caused by the co-transferred autologous human
peripheral blood mononuclear cells (PBMCs). These results indicate that 1,25D3-DC-induced Tregs can inhibit
xenogeneic GVHD and maintain their immunomodulatory function under conditions of inflammation.

1. Introduction
Allogeneic bone marrow transplantation (allo-BMT) has been
exploited as a curative treatment for hematological malignant and non-
malignant diseases since the 70 s [1]. The alloreactivity of these cells
against human leukocyte antigen (HLA) mismatches and minor histo-
compatibility antigens can trigger life-threatening graft-versus-host
disease (GVHD) [2]. Recent studies suggest that an imbalance in ef-
fector (Teff) and regulatory T cell (Treg) reconstitution after allo-BMT
can amplify alloreactivity and promote GVHD.
Vitamin D3 is a fat-soluble hormone that can be acquired from food
or be synthesized in the skin upon ultraviolet-B radiation. In addition to
its classic role in calcium/phosphate and bone homeostasis, 1α,25-di-
hydroxyvitamin D3 (1α,25(OH)2D3), which is the active metabolite of
vitamin D3, has been identified as a potent natural modulator of innate
and adaptive immunity [3]. In dendritic cells (DCs), generated from
bone marrow precursors or peripheral blood monocytes, 1α,25(OH)2D3
interferes with the differentiation and maturation processes [4], locking
DCs in a tolerogenic state [5]. DCs generated in the presence of
1α,25(OH)2D3 (1,25D3-DCs) maintain CD14 expression, express little
CD1a, show reduced expression of antigenpresenting (e.g. HLA-ABC and
-DR) and costimulatory molecules (e.g. CD80, CD86) and diminished
secretion of stimulatory cytokines (e.g. interleukin (IL)-12, IL-23) that
are known to regulate the polarization of Teff cells [6–8]. Moreover, the
expression of inhibitory molecules, including programmed death-ligand
(PD-L) 1, PD-L2, and immunoglobulin-like transcript (ILT)-3 and -4, is
upregulated and the production of regulatory cytokines like IL-10,
transforming growth factor (TGF)-β and tumor necrosis factor (TNF)-α
is enhanced [6,9,10]. As a result, 1,25D3-DCs influence T cell behavior
inducing hyporesponsiveness and shifting T cell polarization from T
helper (Th)1- and Th17-mediated inflammatory responses towards Treg
responses [8–10]. In addition to these effects, 1α,25(OH)2D3 has a
profound impact on the migratory capacity of DCs through upregula-
tion of the chemokine receptor CXCR3, guiding cells to inflamed

⁎
Corresponding author at: Clinical and Experimental Endocrinology, KU Leuven, Campus Gasthuisberg O&N1, Herestraat 49 box 902, 3000, Leuven, Belgium.
E-mail address: chantal.mathieu@uzleuven.be (C. Mathieu).
1
Shared senior author.

https://doi.org/10.1016/j.jsbmb.2018.12.013
Received 1 October 2018; Received in revised form 17 December 2018; Accepted 31 December 2018
Available online 31 December 2018
0960-0760/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/BY-NC-ND/4.0/).
A.-S. Vanherwegen et al.
tissues.
In rodents, adoptively transferred tolerogenic DCs can efficiently
suppress lethal acute GVHD by the in vivo generation of inducible Tregs
(iTregs) [11–14]. However, high numbers of these cells are needed to
reduce the incidence and severity of GVHD and so far, complete pro-
tection from GVHD has never been achieved. In addition, a major fear is
the reversal of the ex vivo induced tolerogenic phenotype when exposed
to the in vivo pro-inflammatory environment of the host, resulting in
immunogenic DCs that could induce Teff cells, rather than Tregs. Thus,
intervention strategies that generate stable tolerogenic DCs that induce
Tregs with superior suppressive ability would be advantageous to
control GVHD as well as prevent transplant rejection or treat auto-
immunity. Although rodents continue to be important in vivo models for
studying the human immune system, we currently recognize the po-
tential limitations of generalizing data from mice to humans. The model
of NOD/LtSz-PrkdcscidIL2rγtm1Wjl (NSG) mice allows the study of an
extreme form of GVHD, namely a xenogeneic GVHD, when human
immune cells are infused.
In the current study, we explored whether 1,25D3-DCs can induce
suppressive Tregs that are able to prevent lethal xenogeneic GVHD in
NSG mice infused with human peripheral blood mononuclear cells
(PBMCs).
2. Materials and methods
2.1. Ex vivo generation of human monocyte-derived 1,25D3-DCs
Human monocytes were freshly isolated and differentiated into DCs
as previously described [6], with the approval of the ethics committee
(S-number 52679) of the University hospital of Leuven (UZ Leuven,
Leuven, Belgium). In brief, PBMCs were isolated by Lymphoprep gra-
dient centrifugation (Alere Technologies AS, Oslo, Norway) from blood
from healthy donors. Monocytes were purified from PBMCs by positive
selection using the CD14 Microbeads Kit (Miltenyi Biotec, Bergisch
Gladbach, Germany) according to manufacturer’s protocol. Isolated
monocytes were cultured in differentiation medium containing 500 IU/
ml human recombinant IL-4 (Gentaur, Kampenhout, Belgium) and 800
IU/ml human recombinant GM-CSF (Gentaur). Medium and cytokines
were refreshed on day 3. On day 6, maturation was induced in the
presence of 800 IU/ml GM-CSF (Gentaur), 100 ng/ml LPS (Sigma-Al-
drich, St. Louis, MO, USA) and 1,000 IU/ml human recombinant IFN-γ
(Roche, Basel, Switzerland). To obtain 1,25D3-DCs, 10−8 M
1α,25(OH)2D3 (Sigma-Aldrich) was added at day 0 and day 3. Similar
dilution of ethanol was used as vehicle control.
2.2. Flow cytometry analysis of surface markers on human DCs
Mature vehicle-treated and 1,25D3-DCs were analyzed for their
surface marker expression using flow cytometry. DCs were pretreated
with Fc block (anti-CD16/CD32) and subsequently labelled with the
following conjugated antibodies: CD1a, CCR7, CXCR3 (from BioLegend,
San Diego, CA, USA), CD14, HLA-DR, PD-L1, ILT-3 (from Thermo Fisher
Scientific/ eBioscience, San Diego, CA, USA), CD80, CD86 (from BD
Biosciences, San Jose, CA, USA), CD52 (Bio-Rad, Hercules, CA, USA).
Doublets were excluded based on signal height and widths. Dead cells
were stained by Zombie Yellow/Aqua Fixable Viability Kit (BioLegend)
according to manufacturer’s guidelines and excluded from analysis.
Samples were acquired on a Gallios™ Flow Cytometer (Beckman
Coulter) and analyzed with the FlowJo software (TreeStar, Ashland,
OR, USA).
2.3. Cytokine production in DC cultures and DC:T cell co-cultures
The V-plex Proinflammatory Panel 1 Human kit from MesoScale
Discovery (Rockville, MD, USA) was used according to the manu-
facturer’s instructions for the quantitative detection of IL-12p70, IL-10
104
Journal of Steroid Biochemistry and Molecular Biology 188 (2019) 103–110
and TNF-α secreted in the DC culture supernatant and similarly the
secretion of IFN-γ and IL-10 in the DC-T cell co-culture.
2.4. Measurement of indoleamine-2,3-dioxygenase (IDO)
Secretion levels of IDO in the DC culture supernatant was de-
termined by the RayBio® Human IDO ELISA Kit (RayBiotech, Norcross,
GA, USA).
2.5. In vitro allogenic T cell proliferation and flow cytometry analysis
CD3+ T cells were isolated from PBMCs of healthy donors using the
Pan T cell Kit II (Miltenyi Biotec) according to manufacturer’s guide-
lines. These purified CD3+ T cells were labelled with 1 μM carboxy-
fluorescein succinimidyl ester (CFSE; Thermo Fisher Scientific/
eBioscience) and subsequently co-cultured with DCs in a 1:10 DC:T cell
ratio for 4 days. T cell proliferation was determined in CD4+ and CD8+
T cell subsets by flow cytometry (antibodies from Thermo Fisher
Scientific/eBioscience). Doublets were excluded based on signal height
and widths. Dead cells were stained by Zombie Yellow Fixable Viability
Kit (BioLegend) and excluded from analysis. Samples were acquired on
a Gallios™ Flow Cytometer (Beckman Coulter) and analyzed with
FlowJo software (TreeStar).
2.6. Induction of antigen-specific Tregs and suppression assay
T cell lines were induced as described previously [10]. In brief,
CD4+ T cells were isolated from PBMCs using the Human CD4+ T cell
Isolation Kit (Miltenyi Biotec), according to manufacturer’s specifica-
tions. 1 × 106 purified CD4+ T cells were co-cultured with vehicle-
treated and 1,25D3-DCs from an haplo-identical donor at a 1:10 DC:T
cell ratio in IMDM (Thermo Fisher Scientific/Gibco) with 10% human
serum (Biowest, Nuaillé, France). After 5 days, T cells were recovered
and rested for 2 days in the presence of 10 ng/ml IL-7 and 5 ng/ml IL-15
(both from Genscript, Nanjing, China). On day 7 after primary stimu-
lation, the CD4+ T cells were restimulated under the same conditions. T
cells stimulated with 1,25D3-DCs are referred to as Treg lines, while
those stimulated with vehicle-treated DCs are referred to as Teff lines.
To assess the suppressive capacity of the generated T cell lines, CD4+
responder T cells (donor A) were labelled with 1 μM CFSE (Thermo
Fisher Scientific/eBioscience) and cultured in the presence of activated
stimulator DCs (donor B) at a 1:10 DC:T cell ratio in 96-well round
bottom plates coated with 0.1 μg/ml anti-human CD3 monoclonal an-
tibody (clone OKT-3; Exbio Antibodies, Vestec, Czech Rebublic). Tregs
(donor A) were added at a 1:1 ratio to responder T cells (donor A). After
4 days, the proliferation of the responder T cells was analyzed on a
Gallios™ Flow Cytometer (Beckman Coulter). Prior to acquisition,
CountBright Absolute Counting Beads (Thermo Fisher Scientific/In-
vitrogen) were added and an equal amount of beads was obtained to
facilitate comparison between samples. To asses IFN-γ and IL-10 cyto-
kine production by the induced T cell lines, T cells were stimulated with
LPS-activated control DCs (donor A) for 24 h after which culture su-
pernatant was collected.
2.7. Xenogeneic graft-versus-host disease model
NOD/LtSz-PrkdcscidIL2rγtm1Wjl (NSG) mice were purchased from
The Jackson Laboratory (Bar Harbor, ME, USA) and inbred in the an-
imal facility of KU Leuven (Leuven, Belgium). Housing of all mice oc-
curred under semi-barrier conditions and animals were fed sterile food
and water ad libitum. Handling of mice and experimental procedures
were performed in accordance to guidelines of the institutional animal
care and use committee of the KU Leuven (project number ECD 221/
2013). Xenogeneic GVHD was induced as described previously [15]. In
brief, NSG mice received total body irradiation with 1.5 Gy one day
before intravenous injection of purified single-donor human PBMCs
A.-S. Vanherwegen et al.
(3 × 106). Simultaneously, PBS or autologous DCs (0.3 × 106) were
injected intraperitoneally. Mice were treated with antibiotic drinking
water (Baytril, Bayer) after xenogeneic PBMC transplantation to pre-
vent unspecific infections. Transplanted mice were assessed twice
weekly for weight loss and clinical scoring of GVHD symptoms. A cu-
mulative global disease score was established for five clinical para-
meters on a scale from 0 to 8: weight loss (2, if > 10%; 6, if > 20%),
anemia (1, white ears; 2, white ears and tail), posture (1, kyphosis), fur
(1, hair loss) and activity (1, stationary > 50% of the time). Mice were
euthanized when reaching a global score ≥6, with the date of death
recorded as the maximum score of 8 in accordance with institutional
animal ethics guidelines. Mice were bled every 2 weeks after transfer in
order to determine the presence of human cells by flow cytometry.
Blood cells were labelled with antibodies against CD3, CD4, FoxP3
(Thermo Fisher Scientific/eBioscience) and CD25 (Miltenyi Biotec) .
Doublets were excluded based on signal height and widths. Dead cells
were stained by Zombie Aqua Fixable Viability Kit (BioLegend) ac-
cording to manufacturer’s guidelines and excluded from analysis.
Samples were acquired on a Canto II Flow Cytometer (BD Biosciences)
and analyzed with the FlowJo software (TreeStar, Ashland, OR, USA).
2.8. Statistics
The normality of all datasets was tested. When data was normally
distributed, either unpaired Student’s t test or ANOVA with Sidak’s
post-test for multiple comparison was used. When data was not nor-
mally distributed, statistical significance was determined using the
Mann-Whitney U test or the Kruskal-Wallis test with subsequent Dunn’s
post-test. Differences in survival rates were assessed using the Mantel-
Cox Log-rank test. All statistical analyses were performed using
GraphPad Prism software (GraphPad Software, La Jolla, CA, USA).
Significance was defined as *p < 0.05; **p < 0.01; ***p < 0.001;
**** p < 0.0001.
3. Results
3.1. 1α,25(OH)2D3 treatment induces stable tolerogenic DCs capable of
homing to inflammatory sites
We first studied the cellular phenotype and function of 1,25D3-DCs
from healthy donors after an inflammatory challenge (LPS and IFN-γ).
We demonstrate that even after exposure to inflammation, 1,25D3-DCs
displayed a stable tolerogenic phenotype with reduced surface expres-
sion of the antigen-presenting molecules HLA-DR and CD1a and the
costimulatory molecules CD80 and CD86 (Fig. 1). The expression of the
monocytic marker CD14 was retained in 1,25D3-DCs exposed to in-
flammation. The inhibitory molecules PD-L1 and ILT-3 as well as the
newly identified tolerogenic marker CD52 [16] were increased 1.3-,
7.5- and 6.4-fold, respectively, at the cell surface after 1α,25(OH)2D3
treatment, a general feature of tolerogenic DCs [17]. Moreover, surface
expression of CXCR3, responsible for homing to inflammatory sites, was
upregulated in 1,25D3-DCs, while the chemokine receptor CCR7 was
similarly expressed as in control cells, still allowing these cells to mi-
grate to the lymphoid tissues (Fig. 1). Also, the cytokine pattern of DCs
was altered upon 1α,25(OH)2D3 exposure with decreased IL-12p70 and
elevated IL-10 and TNF-α secretion levels (Fig. 2). Additionally, 1,25D3-
DCs produced higher amounts of the indoleamine-2,3-dioxygenase
(IDO) enzyme, although this difference was not statistically significant
(Fig. 2). Taken together, the inhibition of full maturation, coupled with
decreased IL-12p70 and enhanced IL-10 and TNF-α production, hints at
a superior immunoregulatory capacity of 1,25D3-DCs.
3.2. 1α,25(OH)2D3-modulated dendritic cells induce allogeneic T cell
hyporesponsiveness
To determine whether these 1α,25(OH)2D3-induced changes in DC
105
Journal of Steroid Biochemistry and Molecular Biology 188 (2019) 103–110
phenotype and cytokine secretion resulted in an altered T cell stimu-
latory capacity, DCs were co-cultured with T cells. Unstimulated T cells
have a low basal proliferation rate, which can be boosted by anti-CD3/
CD28 stimulation (Fig. 3A). Addition of mature vehicle-treated DCs
stimulated the proliferative capacity of both CD4+ and CD8+ allo-
geneic T cells with a respective 4- and 5-fold increase, which was re-
duced to at least half in the condition with 1,25D3-DCs. Analysis of
cytokines secreted into the co-culture supernatant showed that T cells
that were co-cultured with 1,25D3-DCs produced lower amounts of IFN-
γ (Fig. 3B) a typical Th1 cytokine, and high amounts of the tolerogenic
cytokine IL-10 compared to mature vehicle-treated DCs (Fig. 3C). Of
note, with the used measuring technique, it is impossible to determine
whether the IL-10 cytokine levels in the co-culture supernatant are
produced by DCs or by T cells. These data show that 1,25D3-DCs are
potent antigen-presenting cells that can suppress the activation of al-
logeneic Teff responses.
3.3. 1α,25(OH)2D3-modulated dendritic cells induce suppressive antigen-
specific Tregs and delay GVHD
Next, 1,25D3-DC-induced Treg lines were generated and assessed for
their functional capacity. Tregs, primed in the presence of 1,25D3-DCs,
potently suppressed the proliferation of responder T cells to allogeneic
mature DCs compared to Teff cells, primed with mature vehicle-treated
DCs, as evidenced by decreased proliferation and division indexes
(Fig. 3D).
Finally, we investigated whether 1,25D3-DCs were able to mediate
immune suppression when transferred in vivo. We used a xenogeneic
GVHD model induced by transferring human PBMCs into NSG mice that
were co-injected with PBS or autologous, mature vehicle-treated DCs or
1,25D3-DCs (Fig. 4A). All recipients of human PBMCs developed GVHD
(Fig. 4B) and died between day 16 and day 72 following transplanta-
tion. In mice receiving only human PBMCs, clinical signs of GVHD in-
cluding reduced weight, anemia and hair loss were evident from day 16
onwards (Fig. 4B). No significant delay in disease occurrence or mor-
tality was observed when control DCs were co-transferred with human
PBMCs (Fig. 4B, C). A single injection of co-transferred 1,25D3-DCs
markedly attenuated GVHD onset to 34 days on average (ranged from
day 28 until day 40), with reduced GVHD disease manifestations at a
time when most PBS or vehicle DC-treated mice had died (Fig. 4C).
Mice were bled every 2 weeks after transfer to determine the presence
of human cells and the in vivo induction of CD4+CD25+FoxP3+ Tregs
by flow cytometry. Interestingly, at the first time point of 2 weeks post-
engraftment, a significant higher percentage of human FoxP3-expres-
sing Tregs were present in the mice that received 1,25D3-DCs compared
to vehicle-treated DCs or only PBMCs, while the total CD4+CD25+
population remained equal over the different groups and over time
(Fig. 4D). However, 4 and 8 weeks post-engraftment this effect was no
longer detectable. This observation indicates that 1,25D3-DCs are able
to induce CD4+CD25+FoxP3+ Tregs in an in vivo setting, but that this
is a transient phenomenon.
4. Discussion
The active metabolite of vitamin D3, 1α,25(OH)2D3, potently
modulates the behavior of several immune cells. In DCs, 1α,25(OH)2D3
induces a tolerogenic profile. As DCs play a central role in regulating
the delicate balance between immunogenicity and tolerance, 1,25D3-
DCs possess a great potential for therapeutic applications in auto-
immune diseases and transplantation settings. In the current manu-
script, we confirmed previous studies showing the 1α,25(OH)2D3-
mediated downregulation of antigen-presentation, costimulation and
pro-inflammatory cytokine production. Moreover, an increased surface
expression of inhibitory molecules and tolerogenic cytokine secretion
was observed. 1,25D3-DCs maintained a specific migratory signature
upon 1α,25(OH)2D3 modulation, allowing them to migrate to the
A.-S. Vanherwegen et al.
Fig. 1. The tolerogenic phenotype of 1α,25(OH)2D3-modulated human dendritic cells. Surface marker analysis of mature vehicle-treated (CTR; black) and 1,25D3-
DCs (1,25D3; grey). The expression level of phenotypical markers HLA-DR, CD1a, CD80, CD86, CD14, PD-L1, ILT-3, CD52, CXCR3, and CCR7 was determined by flow
cytometry (n = 6–12). Bar graphs represent mean fluorescence intensity (MFI) ± SEM. Each dot represents a different DC donor. * p < 0.05, ** p < 0.01, ***
p < 0.001, **** p < 0.0001.
lymph nodes and the site of inflammation [6,10,18–22]. Interestingly,
these changes remained stable upon inflammatory challenge in vitro
[10,23], which is crucial for in vivo applications where 1,25D3-DCs will
have to exert their functions in a highly inflammatory environment. The
elevated expression of the new tolerogenic DC biomarker CD52 [16]
was validated as well.
Many microarray analyses have been performed on several cell
types, including immune cell subsets, to identify direct transcriptional
targets of 1α,25(OH)2D3. However, only early time points are useful to
distinguish primary and secondary target genes as the full 1,25(OH)2D3-
mediated transcriptional process is believed to take 2–3 h [24]. In ad-
dition, primary 1α,25(OH)2D3 targets are expected to have at least one
vitamin D responsive element (VDRE) in their promoter region.
Therefore, establishment of the chromatin immunoprecipitation se-
quencing (ChIP-seq) technique allowed the study of transcriptional
106
Journal of Steroid Biochemistry and Molecular Biology 188 (2019) 103–110
regulation in more detail on a genome-wide scale. For example, the
CD14 gene expression has been found to be early upregulated by
1α,25(OH)2D3 treatment [25,26]. Moreover, the promoter region of
CD14 comprises several VDR binding sites, confirming that CD14 is
indeed a direct target of active vitamin D3 [26]. Over the past few years,
1α,25(OH)2D3 has been shown to rapidly regulate several immune
genes in monocytes and DCs (e.g. CD274, CD40, CD80, HLA-DR)
[16,25–27]. With regard to the physiological actions of active vitamin
D, not only the primary 1α,25(OH)2D3 targets are important and sec-
ondary targets should be taken into account as well. In this regard,
1α,25(OH)2D3 induces other transcription factors that could regulate
the expression of secondary target genes [28]. Other mechanism of
indirect regulation by 1α,25(OH)2D3 include protein-protein interac-
tions with other transcription factors, such as NF-κB in the case of IL-12
[29], hereby preventing association with its genomic binding sites or
Fig. 2. 1α,25(OH)2D3-modulated human den-
dritic cells secrete low levels of stimulatory
cytokines and high levels of tolerogenic cyto-
kines and markers. Secretion levels of the cy-
tokines IL-12p70, IL-10 and TNF-α into the DC
culture supernatant were measured by MSD
multiplex (n = 6–11). Bar graphs represent
mean cytokine levels ± SEM. Production of
the enzyme IDO into DC culture supernatant
was determined by ELISA (n = 6). Bar graphs
represent mean IDO protein level ± SEM.
Each dot represents a different DC donor. *
p < 0.05, ** p < 0.01, *** p < 0.001, ****
p < 0.0001.
A.-S. Vanherwegen et al. Journal of Steroid Biochemistry and Molecular Biology 188 (2019) 103–110

Fig. 3. 1α,25(OH)2D3-modulated human dendritic cells suppress allogeneic T cell activation and induce suppressive Tregs. (A) Mature CTR- and 1,25D3-DCs were
harvested and co-cultured with CD3+ T cells. After 4 days, proliferation of CD4+ (left) and CD8+ (right) T cells was assessed through CFSE dilution by flow
cytometry. Unstimulated (T alone) and anti-CD3/CD28-stimulated (aCD3/aCD28) T cells were included as negative and positive controls, respectively. Bar graphs
represent percentage of proliferation relative to CTR condition ± SEM (n = 3–5). Secretion of IFN-γ (B) and IL-10 (C) into the supernatant was measured after 4
days of co-culture by MSD multiplex (n = 4). (D) Mature CTR- and 1,25D3-DCs were co-cultured with CD4+ T cells to generate Teff or Treg lines, respectively.
Suppressive capacity of Teff and Treg lines determined in a suppression assay with autologous responder T cells in a 1:1 T cell line: T responder ratio. Plots show CFSE
dilutions in CD4+ cells of Teff lines primed with mature CTR- (left) or Treg lines primed with mature 1,25D3-DCs (right). Representative figure for two independent
experiments. Proliferation index and division index expressed as percent proliferation relative to the crowding control (n = 2). Bar graphs represent mean ± SEM.
Each dot represents a different DC donor. Data are pooled from at least two independent experiments. * p < 0.05, ** p < 0.01.

the induction of epigenetic changes [30]. Future ChIP-seq experiments
in human 1,25D3-DCs will provide more insights in genomic VDR
binding sites and allow better discrimination between primary and
secondary targets.
As a result of these aforementioned phenotypic changes, 1,25D3-
DCs are poor T cell stimulators and simultaneously drive the induction
of antigen-specific Tregs [10,31]. Our current data indeed showed
strongly reduced T cell proliferation and the generation of functional
Tregs in vitro and in vivo. Interestingly, repetitive priming of CD4+ T
cells with 1,25D3-DCs has been shown to result in the de novo induction
of IL-10-producing Tregs, and not only by the expansion of already
existing CD4+CD25+ Tregs [10,32]. The exact nature of these induced
Tregs (iTregs) remains controversial. 1,25D3-DCs can drive the differ-
entiation of IL-10-producing type 1 regulatory (TR1) cells and
CD4+CD25+FoxP3+ Tregs [10,33,34]. The characterization of these
TR1 cells was solely based on the observation that the iTregs produced
high amounts of IL-10 and was not confirmed by surface expression
analysis of TR1-specific parameters, such as CD49b and LAG3 [35]. It
remains to be elucidated which Treg subsets are exactly being induced
by 1,25D3-DCs in vitro and in vivo. Moreover, the exact mechanisms by
which 1,25D3-DCs induce antigen-specific Tregs to restore peripheral
tolerance or immune homeostasis are still not fully understood. A
crucial role for PD-L1, upregulated upon 1α,25(OH)2D3 modulation of
the DCs, has been reported as PD-L1 blockade resulted in Th1-like cells
without suppressive capacity [10]. It is thought that increased PD-L1
triggers IL-10 production by the DCs, thereby driving Treg differentia-
tion [36]. Moreover, 1,25D3-DCs produced elevated amounts of soluble
TNF after maturation, but mTNF was shown to be essential for the in-
duction of suppressive T cells through the mTNF-TNF receptor II sig-
naling pathway [20]. Tolerogenic DCs expressing IDO promote immune

107
A.-S. Vanherwegen et al. Journal of Steroid Biochemistry and Molecular Biology 188 (2019) 103–110

Fig. 4. 1α,25(OH)2D3-modulated dendritic

cells suppress xenogeneic GVHD. (A)
Overview of the experimental set-up. NSG
mice were irradiated with 1,5 Gy (day -1).
On day 0, they received 3 × 106 human
PBMCs (intraveneously, i.v.) simulta-
neously with 0.3 × 106 mature DCs (CTR-
or 1,25D3-, as indicated) or PBS (in-
traperitoneally, i.p.). (B) Disease onset, thus
when mean disease score ≥ 1, and (C)
mean GVHD scores based on weight loss,
anemia, hair loss, posture and reduced mo-
bility were monitored twice per week. (D)
Mice were bled every 2 weeks (if still alive)
and the presence of human Tregs was de-
termined by flow cytometry. Percentage of
human CD4+CD25+ (top) and
CD4 CD25 FoxP3+ (bottom) T cells 2
+ +

weeks, 4 weeks and 8 weeks post-engraft-

ment. Graphs represent mean ± SEM
(n = 4–6). * p < 0.05.

tolerance since tryptophan depletion reduces T cell proliferation, while required to induce FoxP3+ Tregs, it is dispensable for Treg induction by
kynurenine pushes Treg differentiation via the aryl hydrocarbon re- 1,25D3-DCs [37]. Antigen-specific iTregs are believed to suppress dia-
ceptor (AhR). In addition, low tryptophan levels increase ILT-3 and ILT- betogenic Th1 cells via linked suppression and infectious tolerance
4 expression on human DCs [33]. Although ILT-3 expression by DCs is [38].

108
A.-S. Vanherwegen et al.
In the present report, we showed that 1,25D3-DCs secrete IL-10 and
were fully active in vitro and in vivo. The xenogeneic GVHD model in
NSG mice provides the unique opportunity to test the functionality of
human tolerogenic DCs in an in vivo situation. NSG mice are a model for
immunodeficiency as these animals lack functional T cells, B cells and
NK cells. Moreover, NSG mice do not express the IL-2Rγ resulting in
defective cytokine signaling pathways [39]. In this mouse model,
human PBMC transfer will cause GVHD due to the immune reaction of
human T cells against murine tissues. The severity of the induced GVHD
depends on the dose of PBMCs infused. Interestingly, infusion of en-
riched Tregs have been shown to delay the development of xenogeneic
GVHD in NSG mice [40]. In the current study, the co-transfer of 1,25D3-
DCs alongside human PBMCs could significantly delay xenogeneic
GVHD, suggesting the induction of Tregs in this model and demon-
strating the stability of the tolerogenic profile in the highly pro-in-
flammatory milieu in GVHD. Although we did not completely prevent
GVHD, it would be interesting to study whether higher DC doses or
repetitive dosing need to be explored. Our results indeed showed that
the induction of Tregs in vivo is a transient phenomenon that probably
requires multiple doses of 1,25D3-DCs to obtain a stable Treg popula-
tion. Moreover, we believe that 1,25D3-DCs are a safer and more spe-
cific way to interfere with GVHD in the allo-BMT setting compared to
general immunosuppression with pharmacologic inhibitors. This ob-
servation is also of great importance for the current application of
1,25D3-DCs in several autoimmune disorders, like type 1 diabetes
(Trialregister.nl; NTR5542) and multiple sclerosis (MS) (Clinical-
trial.gov; NCT02903537 and NCT02618902). In these circumstances,
monocytes from autoimmune patients are isolated, treated ex vivo with
high doses of 1α,25(OH)2D3, and loaded with disease-relevant antigens
(e.g. proinsulin in type 1 diabetes or myelin peptides in MS). Once the in
vitro induction of the tolerogenic profile of these antigen-specific
1,25D3-DCs is confirmed, the cells are transferred back into the patients
in order to induce Tregs specific for the antigens used. Our data indicate
that it is highly unlikely that these cells would revert to an im-
munogenic phenotype and aggravate disease upon transfer into a pro-
inflammatory environment in vivo
In conclusion, we verified the potential of 1α,25(OH)2D3 to induce
tolerogenic DCs, that remain stable when challenged with pro-in-
flammatory signals both in vitro and in vivo, strengthening the potential
use of these 1,25D3-DCs for tolerance induction in autoimmune diseases
and open therapeutic routes for the treatment of GVHD.
Competing interests
There are no conflicts of interest to be disclosed.
Acknowledgements
We would like to thank Frea Coun and Dries Swinnen for their
technical assistance. We thank the staff at the VIB-KU Leuven FACS
Core Facility for assistance with flow cytometry and the staff of the
laboratory of Experimental Radiotherapy for assistance with the irra-
diations. This work was supported by grants from the Flemish Research
Foundation (FWO Vlaanderen) through an aspirant mandate to AS.V.
(11Y5915N) and D.P.C. (11Y6716N), FWO G.0672.14N, FWO Krediet
aan Navorser 1.5140.15N, the Boehringer Ingelheim Fund and the KU
Leuven (GOA2014/10).
References
[1] G.L. Phillips, Allogeneic hematopoietic stem cell transplantation (HSCT) for high-
risk acute myeloid leukemia (AML)/myelodysplastic syndrome (MDS): How can we
improve outcomes in the near future? Leuk. Res. 36 (2012) 1490–1495.
[2] J.L.M. Ferrara, P. Reddy, Pathophysiology of graft-versus-host disease, Semin.
Hematol. 43 (2006) 3–10.
[3] A.S. Vanherwegen, C. Gysemans, C. Mathieu, Regulation of immune function by
vitamin d and its use in diseases of immunity, Endocrinol. Metab. Clin. North Am.
109
Journal of Steroid Biochemistry and Molecular Biology 188 (2019) 103–110
46 (2017) 1061–1094.
[4] R.M. Steinman, Some interfaces of dendritic cell biology, Apmis 111 (2003)
675–697.
[5] L. Adorini, G. Penna, Dendritic cell tolerogenicity: a key mechanism in im-
munomodulation by vitamin D receptor agonists, Hum. Immunol. 70 (2009)
345–352.
[6] G.B. Ferreira, et al., Differential protein pathways in 1,25-dihydroxyvitamin D 3
and dexamethasone modulated tolerogenic human dendritic cells, J. Proteome Res.
11 (2012) 941–971.
[7] G.B. Ferreira, et al., 1,25-Dihydroxyvitamin D3 alters murine dendritic cell beha-
viour in vitro and in vivo, Diabetes Metab. Res. Rev. 27 (2011) 933–941.
[8] A.G.S. Van Halteren, et al., Redirection of human autoreactive T-cells upon inter-
action with dendritic cells modulated by TX527, an analog of 1,25 dihydrox-
yvitamin D3, Diabetes 51 (2002) 2119–2125.
[9] G.B. Ferreira, et al., 1,25-Dihydroxyvitamin D3 promotes tolerogenic dendritic cells
with functional migratory properties in NOD Mice, J. Immunol. 192 (2014).
[10] W.W.J. Unger, S. Laban, F.S. Kleijwegt, A.R. Van Der Slik, B.O. Roep, Induction of
Treg by monocyte-derived DC modulated by vitamin D3 or dexamethasone:
Differential role for PD-L1, Eur. J. Immunol. 39 (2009) 3147–3159.
[11] P. Reddy, et al., Histone deacetylase inhibition modulates indoleamine 2,3-dioxy-
genase–dependent DC functions and regulates experimental graft-versus-host dis-
ease in mice, J. Clin. Invest. 118 (2008) 2562–2573.
[12] W. Reichardt, et al., Impact of mammalian target of rapamycin inhibition on lym-
phoid homing and tolerogenic function of nanoparticle-labeled dendritic cells fol-
lowing allogeneic hematopoietic cell transplantation, J. Immunol. 181 (2008)
4770–4779.
[13] K. Sato, N. Yamashita, N. Yamashita, M. Baba, T. Matsuyama, Regulatory dendritic
cells protect mice from murine acute graft-versus-host disease and leukemia re-
lapse, Immunity 18 (2003) 367–379.
[14] A. Chorny, E. Gonzalez-Rey, A. Fernandez-Martin, D. Ganea, M. Delgado,
Vasoactive intestinal peptide induces regulatory dendritic cells that prevent acute
graft-versus-host disease while maintaining the graft-versus-tumor response, Blood
107 (2006) 3787–3794.
[15] J. Cuende, et al., Monoclonal antibodies against GARP/TGF-β1 complexes inhibit
the immunosuppressive activity of human regulatory T cells in vivo, Sci. Transl.
Med. 7 (2015) 284ra56.
[16] T. Nikolic, et al., Differential transcriptome of tolerogenic versus inflammatory
dendritic cells points to modulated T1D genetic risk and enriched immune reg-
ulation, Genes Immun. 18 (2017) 176–183.
[17] J.S. Manavalan, et al., High expression of ILT3 and ILT4 is a general feature of
tolerogenic dendritic cells, Transpl. Immunol. 11 (2003) 245–258.
[18] G. Penna, L. Adorini, 1, 25-Dihydroxyvitamin D3 Inhibits Differentiation,
Maturation, Activation, and Survival of Dendritic Cells Leading to Impaired
Alloreactive T Cell Activation, J. Immunol. 164 (2000) 2405–2411.
[19] L. Piemonti, et al., Vitamin D3 affects differentiation, maturation, and function of
human monocyte-derived dendritic cells, J. Immunol. 164 (2000) 4443–4451.
[20] F.S. Kleijwegt, et al., Critical role for TNF in the induction of human antigen-specific
regulatory T cells by tolerogenic dendritic cells, J. Immunol. 185 (2010)
1412–1418.
[21] G. Ferreira, et al., Vitamin D3 induces tolerance in human dendritic cells by acti-
vation of intracellular metabolic pathways, Cell Rep. 10 (2015) 711–725.
[22] G.B. Ferreira, et al., Proteome analysis demonstrates profound alterations in human
dendritic cell nature by TX527, an analogue of vitamin D, Proteomics 9 (2009)
3752–3764.
[23] M. Naranjo-Gomez, et al., Comparative study of clinical grade human tolerogenic
dendritic cells, J. Transl. Med. 9 (2011) 89.
[24] C. Carlberg, M.J. Campbell, Vitamin D receptor signaling mechanisms: integrated
actions of a well-defined transcription factor, Steroids 78 (2013) 127–136.
[25] S. Heikkinen, et al., Nuclear hormone 1α,25-dihydroxyvitamin D3 elicits a genome-
wide shift in the locations of VDR chromatin occupancy, Nucleic Acids Res. 39
(2011) 9181–9193.
[26] L. Széles, et al., 1,25-Dihydroxyvitamin D3 is an autonomous regulator of the
transcriptional changes leading to a tolerogenic dendritic cell phenotype, J.
Immunol. 182 (2009) 2074–2083.
[27] A. Neme, V. Nurminen, S. Seuter, C. Carlberg, The vitamin D-dependent tran-
scriptome of human monocytes, J. Steroid Biochem. Mol. Biol. 164 (2016) 180–187.
[28] V. Nurminen, et al., The transcriptional regulator BCL6 participates in the sec-
ondary gene regulatory response to vitamin D, Biochim. Biophys. Acta - Gene Regul.
Mech. 1849 (2015) 300–308.
[29] D. D’Ambrosio, et al., Inhibition of IL-12 production by 1,25-dihydroxyvitamin D3.
Involvement of NF-kappaB downregulation in transcriptional repression of the p40
gene, J. Clin. Invest. 101 (1998) 252–262.
[30] V. Nurminen, A. Neme, S. Seuter, C. Carlberg, The impact of the vitamin D-
modulated epigenome on VDR target gene regulation, Biochim. Biophys. Acta -
Gene Regul. Mech. 1861 (2018) 697–705.
[31] S.P. Cobbold, E. Adams, K.F. Nolan, F.S. Regateiro, H. Waldmann, Connecting the
mechanisms of T-cell regulation: dendritic cells as the missing link, Immunol. Rev.
236 (2010) 203–218.
[32] G. Penna, et al., Expression of the inhibitory receptor ILT3 on dendritic cells is
dispensable for induction of CD4+Foxp3+ regulatory T cells by 1,25-dihydrox-
yvitamin D3, Blood 106 (2005) 3490–3497.
[33] R. Kushwah, J. Hu, Role of dendritic cells in the induction of regulatory T cells, Cell
Biosci. 1 (2011) 20.
[34] D.X. Beringer, et al., T cell receptor reversed polarity recognition of a self-antigen
major histocompatibility complex, Nat. Immunol. 16 (2015) 1153–1161.
[35] N. Gagliani, et al., Coexpression of CD49b and LAG-3 identifies human and mouse T
A.-S. Vanherwegen et al.
regulatory type 1 cells, Nat. Med. 19 (2013) 739–746.
[36] H. Kuipers, et al., Contribution of the PD-1 ligands/PD-1 signaling pathway to
dendritic cell-mediated CD4+ T cell activation, Eur. J. Immunol. 36 (2006)
2472–2482.
[37] G. Penna, et al., 1,25-Dihydroxyvitamin D3 selectively modulates tolerogenic
properties in myeloid but not plasmacytoid dendritic cells, J. Immunol. 178 (2007)
145–153.
[38] F.S. Kleijwegt, et al., Transfer of Regulatory Properties from Tolerogenic to
110
Journal of Steroid Biochemistry and Molecular Biology 188 (2019) 103–110
Proinflammatory Dendritic Cells via Induced Autoreactive Regulatory T Cells, J.
Immunol. 187 (2011) 6357–6364.
[39] L.D. Shultz, M.A. Brehm, J.V. Garcia-Martinez, D.L. Greiner, Humanized mice for
immune system investigation: progress, promise and challenges, Nat. Rev.
Immunol. 12 (2012) 786–798.
[40] M. Hannon, et al., Infusion of clinical-grade enriched regulatory T cells delays ex-
perimental xenogeneic graft-versus-host disease, Transfusion 54 (2013).