Retinoic acid (RA), an active metabolite of vitamin A, and the hormonal form of vitamin D, 1,25-dihydroxyvitamin D (1,25D), are very active compounds, which regulate many important cellular processes, such as differentiation and proliferation [1
]. RA and 1,25D are the ligands for nuclear receptors, which act as transcription factors after binding the ligand. A dominating RA metabolite in humans is all-trans
-RA (ATRA), which binds with high affinity to all retinoic acid receptors (RARα, β and γ). A less abundant metabolite, which is nevertheless present in almost all tissues, is 13-cis
-RA, which most probably serves as a depot for isomerisation to ATRA or to 9-cis
-RA is hard to detect in human tissues, and it binds predominantly to retinoid X receptors (RXRα, β and γ) [3
]. 1,25D is a ligand to only one vitamin D receptor (VDR), which in its active state forms heterodimers with RXRs [5
]. After ligation, VDR and RARs undergo conformational changes that induce binding to specific sequences in the promoter regions of target genes. These sequences are called vitamin D response elements (VDRE) and retinoic acid response elements (RARE) [6
]. The role of RARα in blood development is very important, as the fusion of the RARα gene with the PML gene, caused by a translocation t(15;17), initiates acute promyelocytic leukemia where differentiation of the myeloid lineage is blocked at a promyelocyte stage and followed by an uncontrolled proliferation of leukemic blasts [7
]. It is also well known and widely accepted that VDR is present in multiple blood cells, and that the correct levels of 1,25D are essential for proper functioning of the immune system [8
]. Both compounds, ATRA and 1,25D, can be used in therapy to induce differentiation of acute myeloid leukemia (AML) blasts. ATRA induces differentiation of these blasts to granulocyte-like cells, and 1,25D to monocyte-like cells [9
]. Past research indicated a synergistic cancer differentiation effect by using a combination of 1,25D and ATRA [11
]. However, our recent experiments have revealed that in AML cell lines the effect of combination treatment varies, due to either down- or upregulation of VDR
expression in response to ATRA, depending on the AML cell line examined [12
Since beneficial effects of 1,25D and ATRA combination treatment in anticancer therapy have been reported and their wider use postulated [14
], the effects of such combination towards normal cells should be addressed. Hematopoiesis seems to be the most relevant process which might be influenced by ATRA and 1,25D. The roles of vitamin A and its most active metabolites during hematopoiesis have been extensively studied and are well appreciated [15
]. The actions of RA are multiple, and they start as early as in embryonal yolk sac and aorta-gonad-mesonephros, where RA causes the appearance of hematopoietic progenitors from the hemogenic endothelium [16
]. In adult hematopoiesis, RA is important for granulopoiesis, and it controls differentiation of B and T lymphocytes [15
]. However, it should be remembered that, due to difficulties in the use of human models of hematopoiesis, mice models have often been used in the experiments [15
]. The role of 1,25D in hematopoiesis is less well documented than that of ATRA; moreover, some of the data come from zebrafish models. It should be remembered that, in contrast to humans and mice, there are two forms of VDR in zebrafish [17
]. However, the available data show that the correct levels of 1,25D are necessary to maintain hematopoietic stem and progenitor cells (HSPCs) [18
]. It was also shown that in human hematopoietic stem cells (HSCs) exposed to physiological concentrations of 1,25D, markers of monocytic differentiation are induced [19
The gene encoding human VDR is located on chromosome 12. This gene is composed of 14 exons, and translation of VDR protein starts from the exon 2. Region 5′ of human VDR
gene is very complex, and is composed of the seven exons 1a–g. These exons, together with corresponding promoter regions, are alternatively used for VDR
transcription in different tissues. Transcripts starting from exon 1a and from exon 1d are regulated by the common promoter upstream to exon 1a, and the exons 1f and 1c have separate upstream promoters [20
]. Our recent experiments have revealed a new exon, 1g, regulated from the promoter of exon 1a. Exon 1g is used in VDR
transcripts present in AML cells [13
]. Multiple publications confirm that VDR
expression in humans is regulated in response to ATRA [12
], while there are conflicting reports concerning regulation of human VDR
by 1,25D [13
The murine VDR
gene is located on chromosome 15, and its composition is less complex than in humans. In the 5’ UTR region of VDR
gene, exons 1 and 2 were identified, which show strong homology to human 1a and 1c, respectively [27
]. Although exon 1d is well conserved (1d-like), transcripts containing this exon have not been reported in mice. The sequence similarity of the exons 1f and 1b is low between man and mice. Translation of mouse VDR protein starts from exon 3 [28
]. It has been shown that transcription of VDR
is upregulated in response to 1,25D in murine osteoblasts [29
This is why we decided to examine the effects of the 1,25D and ATRA combination on VDR gene expression in blood cells at various steps of their development. We were interested in discovering if, in normal human blood cells, transcriptional variants of the VDR gene are as multiple as in AML cells, and if they are regulated in response to ATRA and 1,25D. Since the availability of human hematopoietic cells for experiments is very limited, we decided to examine whether human cells could be replaced in this type of studies with murine blood model.
An important question in studies concerning nuclear receptors is whether or not they are transcriptionally active in the cells. It is therefore important to be able to study expression of the genes that are specific targets of regulation by either VDR or RARs. In the case of VDR, expression of the gene which encodes 24-hydroxylase of 1,25D (CYP24A1
) is the best measure of VDR’s activity. 24-Hydroxylase of 1,25D is the central enzyme in the catabolism of 1,25D to calcitroic acid. CYP24A1
is the most strongly regulated out of all 1,25D-target genes [31
]. This is why 1,25D-dependent upregulation of CYP24A1
confirms that VDR protein is expressed and active in cells. CYP24A1
is upregulated in response to 1,25D, but not in response to ATRA [13
], and its expression can also be detected in human HSCs [18
]. We decided to search for a similar sensor of RARs transcriptional activity. The effective concentration of the most active metabolite of vitamin A, ATRA is strictly controlled in the human body. In order to maintain safe concentrations of ATRA, its catabolism is also strictly regulated [32
]. The major catabolizing enzyme is retinoic acid 4-hydroxylase (CYP26A1), whose transcription is upregulated in response to ATRA [33
] by activating RARE, which is present in the CYP26A1
promoter region [34
]. Thus, our first aim was to verify that CYP26A1
is regulated by ATRA in RAR-positive cells, and that it is not regulated by 1,25D in VDR-positive cells.
The findings from studies of leukemia cell lines support the use of 1,25D as an anticancer agent, since 1,25D causes the growth arrest and differentiation of a wide variety of AML cell lines [39
]. Our results revealed that there are patients whose AML blasts respond to 1,25D analogs with differentiation, while the blasts of others are resistant [39
]. The possible reason of these differences may lie in the expression level of the VDR
gene and the VDR protein level in AML cells. We have recently found that low VDR
expression levels may be upregulated using ATRA and that unliganded RARα acts as transcriptional repressor to VDR
]. The possibility of using RA analogs to induce differentiation of blasts was investigated for many years and in the case of ATRA it was successfully introduced into clinics to treat one subtype of AML [41
]. Unfortunately, in other subtypes of leukemia, ATRA is not effective. This is why the combined use of 1,25D and ATRA, or a combination of their more active analogs, was postulated [14
]. In AML cells, the VDR
gene is transcribed in multiple variants, and some of them are transcriptionally regulated by ATRA. We thus wanted to ascertain whether similar transcriptional variants of VDR can be found in normal human blood cells. Our experiments revealed that transcripts VDR1a
can also be detected in normal blood cells, however they are upregulated in response to ATRA, in a manner similar to KG1 cell line, only in blood cells found in UCB. VDR
protein in UCB cells is transcriptionally active after exposure to 1,25D, and in PBM cells this transcriptional activity is much lower. Moreover, in contrast to AML cell lines, upregulation of VDR
expression is not followed by an increased transcriptional activity of VDR protein. This might suggest that possible side effects of combination treatment using 1,25D and ATRA in normal human blood cells do not synergize.
It is widely accepted that the eventual cell fate during hematopoiesis is governed by spatiotemporal fluctuations in transcription factor concentrations, which either cooperate or compete in driving target gene expression [42
]. Nuclear receptors for vitamin A and vitamin D do not belong to the set of the most important hematopoietic regulators, but their roles in blood cell differentiation are becoming apparent and appreciated. The availability of human HSC cells is limited, and their differentiation can’t be studied in vivo. This is why the majority of the available data about blood cell formation comes from murine models, however one should be aware that human hematopoiesis does not reflect murine hematopoiesis in all aspects [43
]. This might also concern the roles of VDR and RARs in blood cell formation. As mentioned in the Introduction, the organization and regulation of the VDR
locus in humans and mice are different, and the 5′ UTR region composition is less complex in mice. Despite the high resemblance of the symptoms of VDR
knock-out in mice and inactivating mutations of VDR
in humans, not all findings concerning vitamin D endocrine systems in mice are present in humans [44
]. This is why we addressed regulation of VDR
transcription in murine blood cells in parallel to human cells.
Our results indicate that there are differences in the regulation of VDR transcription between mice and man. In contrast to human cells, in murine cells ATRA does not influence VDR expression, even though RARs are present and transcriptionally active. On the other hand, the VDR gene in murine blood is positively auto-regulated by the VDR ligand, 1,25D. This positive feedback causes transcriptional activity of VDR to be particularly high in murine HSPC exposed to 1,25D. This might indicate that VDR is important for murine blood cells at early stages of their commitment. Such auto-regulation of VDR does not occur in human blood cells. Since there is no good alternative for testing the toxicity of potential drugs other than rodents, we suggest that observations concerning toxic effects of 1,25D in mice should be translated to humans with caution.
Another issue to consider is the fact that there are differences in the vitamin D system between mice and man. 1,25D is a steroid hormone, which is entirely produced by organisms from 7-dehydrocholesterol, and biologically activated by subsequent hydroxylations at carbons C25 and C1 [45
]. This requires exposure to the UVB light, and vitamin D must only be delivered with food for people who live in regions deficient in sunlight. Mice are much less likely to produce vitamin D in high amounts from exposure to sunlight due to their fur, as well as their nocturnal and underground activities. Therefore, the role and regulation of VDR
is likely to be different in these two distinct species, and should be taken into consideration when vitamin D compounds are being tested in mice. Very high positive auto-regulation of VDR
expression in murine blood cells at their early steps of development, which does not occur in humans, might cause unwanted side-effects of 1,25D, or of its highly active analogues, to be more pronounced in mice than in humans.
4. Materials and Methods
1,25D was purchased from Cayman Europe (Tallinn, Estonia) and ATRA was from Sigma-Aldrich (St. Louis, MO, USA). The compounds were dissolved in an ethanol to reach 1000× final concentrations, and subsequently diluted in the culture medium to the concentration required for experiments.
4.2. Cell Lines and Normal Cells
HL60 cells were acquired from the cell bank at the Institute of Immunology and Experimental Therapy in Wrocław, Poland and KG1 cells were purchased from the German Resource Center for Biological Material (DSMZ GmbH, Braunschweig, Germany). The cells were cultured in RPMI-1640 medium (Biowest, Nuaillé, France) with 10% fetal bovine serum (FBS), 2 mM l-glutamine, 100 units/mL penicillin and 100 µg/mL streptomycin (all from Sigma-Aldrich) and maintained at standard cell culture conditions.
Human UCB was obtained post-delivery at the First Department of Obstetrics and Gynecology, Wrocław Medical University (Wrocław, Poland) from mothers who gave informed consent for this study. The study was accepted by the local Ethical Committee. Two to eight mL of cord blood were diluted with PBS in 1:1 ratio. Diluted blood was carefully layered onto the equal volume of Histopaque 1077 (Sigma-Aldrich), and centrifuged at 400× g for 30 min. The opaque interface containing mononuclear cells was moved to fresh sterile tube, and washed three times with PBS. The cells were transferred to Biotarget-1 (Biological Industries, Kibbutz Beit-Haemek, Israel) medium containing 4 mM l-glutamine, 100 units/mL penicillin and 100 µg/mL streptomycin and maintained at standard cell culture conditions.
The experiments using animals were performed according to the procedures approved by the First Local Ethical Commission for Animal Experimentation in Wrocław at the Institute of Immunology and Experimental Therapy (permit numbers 21/2016/W, 21/2016/U, 20/2016/U issued on 5 January 2016). Cell suspensions from 8 week old C57BL/6 mice were prepared as follows: bone marrow cells were isolated by washing the femur and tibia with ice-cold PBS stream. Spleen and thymus were washed with ice-cold PBS and strained through 30-μm mesh. Kidneys were washed to remove blood, cut into small 1–2 mm2 pieces and incubated in 0.5 mL with colagenase type II (1 mg/mL) and DNAse I (10 units/mL) at 37 °C for 40 min. Cells isolated from bone marrow, spleen and thymus were treated with red blood cell lysis buffer (155 mM NH4Cl, 10 mM KHCO3, 0.1 mM ethylenediaminetetraacetic acid) to remove erythrocytes. All tissues and cells were mechanically dissociated by the syringe trituration, washed twice with PBS by centrifugation (400 rcf, 5 min, 4 °C) and resuspended in PBS supplemented with 5% FBS. Single cell suspension was filtered through 70-μm mesh.
4.3. Sorting of Blood Cells and Flow Cytometry
Human HSCs were sorted from cord blood mononuclear cells using the Miltenyi MACS CD34 Isolation Kit (Miltenyi Biotec, Bergisch Gladbach, Germany) in accordance with the manufacturer’s instructions. Briefly, cord blood mononuclears were resuspended in Separation Buffer (PBS with 10% bovine serum albumin (BSA)) and incubated with FcR Blocking Reagent and magnetic microbeads conjugated to monoclonal mouse anti-human CD34 antibody for 30 min at 4 °C. Labeled cell suspension was sorted using magnetic separator. After three washes with Rinsing Solution (PBS supplemented with 2 mM EDTA and 0.5% BSA), the column was removed from the separator and labeled CD34+ cells were eluted with 1 mL of Rinsing Solution. To determine the purity of CD34+ cell fraction, 1 × 105
cells were stained with phycoerythrin (PE)-conjugated anti-CD34 (Becton Dickinson, San Jose, CA, USA) monoclonal antibody for 60 min on ice. Isotype-identical monoclonal antibodies served as controls. Next, the stained cells were analyzed using flow cytometry (BD Accuri™ C6, Becton Dickinson, San Jose, CA, USA). The purity ranged from 92% to 95%, and sample staining is presented in Figure A1
. CD34+ cells were grown in Stemline Hematopoietic Stem Cell Expansion Medium with 4 mM l
-glutamine, 100 units/mL penicillin and 100 µg/mL streptomycin, recombinant human cytokines (all from ImmunoTools, Friesoythe, Germany): stem cell factor (100 ng/mL), thrombopoietin (100 ng/mL) and granulocyte colony-stimulating factor (100 ng/mL).
Hematopoietic stem and progenitor cells were isolated from murine bone marrow using Mouse Hematopoeitic Progenitor Cell Isolation Kit (Stemcell, Cologne, Germany) according to the manufacturer’s recommendations. Briefly, the cells were resuspended in PBS (with 2% FBS and 1 mM EDTA, rat serum 50 μL/mL at the density of 1 × 108 cells/mL) and incubated with EasySep Mouse Hematopoietic Progenitor Cell Isolation Cocktail (50 μL/mL) for 15 min at 4 °C. Next, EasySepTM Streptavidin RapidSpheres (75 μL/mL) were added and after 10 min of incubation the cell suspension was sorted with magnets. The purity of the obtained population was monitored by flow cytometry (FACS-Calibur, Becton Dickinson, San Jose, CA, USA) using anti-c-kit-APC (eBioscience, Vienna, Austria) and anti-Sca-1-FITC (eBioscience) staining. According to the manufacturer, the lineage antigen-negative cell content of the isolated fraction typically ranges from 60% to 84%. In our experiments c-kit+ cells constituted 65% of sorted population. Stemline Hematopoietic Stem Cell Expansion Medium (Sigma-Aldrich) and recombinant murine cytokines (all from ImmunoTools): stem cell factor (50 ng/mL), Flt3-ligand (50 ng/mL), thrombopoietin (50 ng/mL) and interleukin-6 (10 ng/mL) were used for further ex vivo culture of the isolated cells.
Murine spleen cells were stained with anti-CD3-APC and anti-CD19-PE antibodies (Becton Dickinson) to isolate mature T- and B-cells, respectively (Figure A3
a). Bone marrow cells were stained with anti-CD45-FITC antibody (Becton Dickinson) to isolate granulocytes, using CD45/SSC-based sorting criteria (Figure A3
b). Cells were stained in 0.5 mL PBS supplemented with 2% FBS using 1 μg of each antibody for 30 min on ice. Cells were sorted using FACS-Aria (Becton Dickinson).
4.4. 5′-RACE Assay
In order to identify the transcriptional start sites for murine VDR
transcript(s), 5’-RACE was used [46
]. Ten micrograms of total RNA were isolated from intestine, kidney or bone marrow of C57BL/6 mice and then processed as before: digested with calf alkaline phosphatase (CIP, New England Biolabs, Ipswich, MA, USA) in the presence of RiboLock RNAse inhibitor for 1 h at 37 °C and purified by extraction with TRI Reagent (Sigma-Aldrich). Half of the CIP-digested RNA was treated with tobacco acid pyrophosphatase (TAP, Epicentre, Madison, WI, USA) for 1 h at 37 °C in 10 μL reaction mixture containing TAP buffer, 0.5 units of TAP and 20 units of RiboLock RNAse inhibitor. 2 μL of TAP-treated RNA was ligated with a RNA oligonucleotide (5’-GCUGAUGGCGAUGAAUGAACACUGCGUUUGCUGGCUUUGAUGAAA-3′) for 1 h at 37 °C in a 10 μL reaction mixture containing 0.3 μg of the oligonucleotide, 5 U of RNA ligase (New England Biolabs), RNA ligase buffer and 20 U of RiboLock RNase inhibitor. 2 μL of RNA was then reverse transcribed using SuperScript III reverse transcriptase (Invitrogen, Carlsbad, CA, USA) and random hexamers. The cDNA was amplified in nested PCR reactions (2 × 20 cycles, annealing temp. 52 °C, in the presence of 1.2 M betaine) using primers complementary to 5′-adapter (5′-GCTGATGGCGATGAATGAACACTG-3′, 5′-CGCGGATCCGAACACTGCGTTTGCTGGCTTTGATG-3′) and exon 5 and 4 of VDR
gene (5’-TCTGTGAGGATGAACTCCTTCATC-3′, 5′-TCCTTGGTGATGCGGCAATCTC-3′). The amplification products were directly cloned into pGEMT-easy vector. The individual plasmid clones were sequenced using SP6 primer (5’-ATTTAGGTGACACTATAG-3′) and BigDye 3.1 Terminator Cycle Sequencing Kit (Life Technologies, Carlsbad, CA, USA). The sequencing reaction was analyzed using ABI Prism 310 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA). The obtained sequences of VDR
transcripts were aligned with the genomic sequence of VDR
gene using Spidey software (https://www.ncbi.nlm.nih.gov/spidey/
National Center for Biotechnology Information, Bethesda, MD, USA) to identify exons and transcriptional start sites.
4.5. cDNA Synthesis and Real-Time PCR
For PCR analyses, the cells were stimulated with 10 nM 1,25D and/or 1 μM ATRA for 96 h. RNA from unstimulated and stimulated cells was isolated using either TRI Reagent (Sigma-Aldrich), Extractme Total RNA Kit (DNA-Gdańsk, Gdańsk, Poland) (for >106 cells) or PicoPure RNA Isolation Kit (ThermoFisher Scientific, Waltham, MA, USA) (for <106 cells) according to manufacturer’s recommendations. Reverse transcription of 100 ng of total RNA (34 ng in case of human CD34+ cells) was done with High-Capacity cDNA Reverse Transcription Kit (ThermoFisher Scientific) using random hexamers. The Real-time PCR analysis was performed using Real-time PCR–PowerUp™ SYBR Green Master Mix (Applied Biosystems) or SensiFAST SYBR® No-ROX Kit (Bioline, London, UK). For murine samples the reaction consisted of 40 cycles (95 °C for 15 s and 60 °C for 60 s), preceded by uracil-DNA glycosylase and AmpliTaq DNA polymerase activation at 50 °C for 120 s and 95 °C for 120 s, respectively, and was performed on BioRad CFX Connect apparatus (Bio-Rad Laboratories Inc., Hercules, CA, USA). The thermal profile for human samples consisted of 45 cycles (95 °C for 5 s, 54/58 °C for 10 s, 72 °C for 5 s) followed by one step at 95 °C for 2 min, the reaction was performed using CFX Real-time PCR System (Bio-Rad).
The following primer pairs were used:
hGAPDH: forward 5’-CATGAGAAGTATGACAACAGCCT-3′, reverse 5’-AGTCCTTCCACGATACCAAAGT-3′;
hVDR: forward 5’-CCTTCACCATGGACGACATG-3′, reverse 5’-CGGCTTTGGTCACGTCACT-3′;
hVDR1a: forward 5’-GCGGAACAGCTTGTCCACCC-3′, reverse 5’-GAAGTGCTGGCCGCCATTG-3′;
hVDR1d: forward 5’-GCTCAGAACTGCTGGAGTGG-3′, reverse 5’-GAAGTGCTGGCCGCCATTG-3′;
hVDR1g: forward 5’-TTGCTCATCCAGCTTCCCAGAC-3′, reverse 5’-GAAGTGCTGGCCGCCATTG-3′;
hCYP24A1: forward 5’-CTCATGCTAAATACCCAGGTG-3′, reverse 5’-TCGCTGGCAAAACGCGATGGG-3′;
hCYP26A1: forward 5’-CGCATCGAGCAGAACATTCG-3′, reverse 5’-GCTTTAGTGCCTGCATGT-3′;
mGAPDH: forward 5’-AACTTTGGCATTGTGGAAGG-3′, reverse 5’-ACACATTGGGGGTAGGAACA-3′;
mVDR: forward 5’-CACCTGGCTGATCTTGTCAGT-3′, reverse 5’-CTGGTCATCAGAGGTGAGGTC-3′;
mCYP24A1: forward 5’-CACGGTAGGCTGCTGAGATT-3′, reverse 5’-CCAGTCTTCGCAGTTGTCC-3′;
mCYP26A1: forward 5’-GCAGGCACTAAAACAATCGTC-3′, reverse 5’-GCTGTTCCAAAGTTTCCATGTC-3′. Relative quantification of gene expression was analyzed with the ΔΔCt method using GAPDH as the endogenous control.
4.6. Statistical Analysis
The sample distribution was assessed using the Shapiro-Wilk test. For samples with normal distribution, t-test was used to assess significance of differences. For the remaining samples, a non-parametric one-way ANOVA test followed by a Mann-Whitney U test was used for assessing the significance of the differences in gene expression levels.