Novel insight into the role of the vitamin D receptor in the development and function of the immune system

Abstract

Immune cells express the vitamin D receptor (VDR) and are therefore vitamin D targets. The Vdr protein can be readily measured in the kidney using antibodies to the Vdr and western blot. It is much more difficult to measure Vdr protein in the spleen because of the low level of VDR expression in resting immune cells. In order to more sensitively measure VDR expression, the Cre enzyme was inserted in the 3rd exon of the VDR gene and a reporter mouse that irreversibly expresses tdTomato was made. Mice that express one copy of the VDRCre gene were confirmed to be VDR +/− and mice that express two copies were confirmed to be VDR −/−. Initial characterization of the immune cells from the VDR +/− VDR^tdTomato+ compared to VDR+/+ wildtype (WT) littermates showed no effect of being hemizygous for the VDR on immune cell frequencies. High tdTomato expression was shown to be present in the bone marrow (BM) and thymus immune cell precursors. In the periphery, monocytes, neutrophils and macrophages had very high tdTomato+ (88–98%) expression while lymphocytes ranged from 60–70% tdTomato+. Tissue resident innate lymphoid cell (ILC) 1 and 3 cells were about 60–80% tdTomoto+, while ILC2 cells had very low tdTomato expression. Stimulation of VDR^tdTomato+ splenocytes showed that the tdTomato− CD4+ and CD8+ T cells proliferated more than their tdTomato+ counterparts. T cells were sorted for tdTomato+ and tdTomato− and then activated for 72h. Sorted tdTomato+ T cells expressed the VDR protein only after 72h post-activation. The sorted tdTomato− T cells proliferated more than the sorted tdTomato+ T cells. Interestingly, activation of the tdTomato− T cells failed to induce new tdTomato expression. The data suggest that an early immune precursor expresses the VDR. In the periphery, neutrophils and monocytes are almost all tdTomato+, while some immune cells (ILC2 and some T cells) may never express the VDR.

Introduction

The vitamin D receptor (VDR) is a ligand-activated nuclear receptor. The VDR binds the high affinity vitamin D ligand 1α,25-dihydroxyvitamin D (1,25(OH)₂D, 1,25D) to form a VDR/RXR heterodimeric complex that further recruits co-activators/co-repressors to regulate gene transcription [1]. Therefore, cells that express the VDR are direct vitamin D targets. Immune cells including dendritic cells, macrophage, B cells, and T cells all express the VDR [2–5]. Vitamin D and 1,25(OH)₂D inhibit T cell proliferation, IFN-γ and IL-2 production while enhancing regulatory T cell functions and IL-10 [6–8]. The immunoregulatory functions of vitamin D include suppression of immune mediated diseases like inflammatory bowel disease, diabetes and multiple sclerosis [9–12]. The immune system is a vitamin D target and cells of the immune system express the VDR.

The evidence for VDR expression in the immune system includes ligand binding assays, quantitative PCR, and western blots [13, 14]. The VDR is highly and constitutively expressed in the kidney, colon, and small intestine (SI) [13]. Other tissues including immune tissues like the spleen have very low levels of the VDR. In immune cells, activation upregulates the VDR and VDR protein is detectable by western blots at 2–3 days after activation in T cells [15]. Cytokines or toll like receptor agonists upregulate the VDR in macrophage [3]. There are several different antibodies available to detect the VDR but there are conflicting results as to the specificity and sensitivity of the VDR antibodies [13]. Wang and colleagues compared the available anti-VDR antibodies for VDR specificity and showed that the D-6 antibody had specific staining of the Vdr in the mouse kidney and SI [13]. The D-6 antibody failed to stain the brain by immunohistochemistry [16]. Recently Liu et.al published a paper that used a VDR reporter mouse to characterize VDR expression in the brain and showed reporter expression in the brain [17]. The available reagents for evaluating Vdr protein expression are not sensitive enough to identify individual cell types for expression of the Vdr.

The Liu et. al VDR reporter mouse expressed Cre at the C terminus of the VDR gene and VDR mRNA expression was not affected in the transgenic mice [17]. We took a different approach that expressed the Cre recombinase in the first coding exon of the VDR gene. Expression of Cre resulted in the inactivation of the VDR gene. The VDRCre mouse was crossed with a reporter mouse that expressed a floxed stop codon in front of a tdTomato gene. The expression of the VDRCre transgene results in the irreversible expression of tdTomato. Using flow cytometry, we showed that about 80% of the bone marrow (BM) and thymus progenitors already expressed tdTomato. The immune system is an important vitamin D target and during development the hematopoietic stem cells in the adult BM has expressed the VDR. In the periphery most cell types develop from precursors that have previously expressed the VDR. ILC2 cells in the mucosa and some T cells in the spleen develop from VDR− precursors and some activated T cells did not express the VDR tdTomato reporter. The VDR^tdTomato+ mouse is a novel and useful tool for studying VDR expression in the immune system.

Results

Generation of VDRCre mouse model.

The exon 3 of the VDR gene was specifically targeted for Cre insertion using CRISPR technology (Albert Einstein College of Medicine transgenic mouse facility, Bronx, NY). An RNA sequence encoding the Cre recombinase gene was inserted into exon 3 by homology mediated direct repair, resulting in a knock-in of the Cre sequence into the VDR allele (Fig. 1A). The founder male (F0) was bred with WT female mice and the offspring were genotyped for expression of the WT and Cre transgene (Fig. 1B). Confirming the VDRCre+/− genotype, the F0 male expressed two bands one for the WT VDR gene of 229 bp and one for the Cre transgene of 177 bp (Fig. 1B). The VDRCre+/− offspring had two bands and the WT offspring (VDRCre−/−) had only one band of 229 bp.

Figure 1:

Generation of VDRCre mouse using CRISPR technology (Albert Einstein College of Medicine transgenic mouse facility, Bronx, NY). A) Guide RNA that targeted exon 3 of the VDR gene was used to direct Cas9 mediated cleavage. Homology directed repair was done to generate the VDRCre+/− founder (F0). B) The F0 male was bred with WT Cre −/− females and the offspring were genotyped. One reaction containing 3 primers were used to identify the WT allele at 229 kb and the Cre containing transgene at 177 kb.

Tissue Vdr expression.

The Vdr is highly and constitutively expressed in the kidneys and undetectable in the liver [13]. Kidney was used as the positive control for Vdr expression and liver as the negative control (Fig. 2A). As expected, Vdr mRNA and protein were high in the kidneys and low in the liver (Fig. 2A). Vdr mRNA was not detected in the liver, low in the spleen and lung and highest in the kidney (Fig. 2A). Western blots using the D-6 antibody showed some non-specific staining especially in tissues with low or no Vdr expression (Fig. 2A). The liver did not have any Vdr, the spleen had very low Vdr, the lung was intermediate and the kidney was high for the Vdr protein (Fig. 2A). The Vdr protein was quantified in the kidneys of the VDR +/+ (VDRCre−/−), VDR+/− (VDRCre+/−) and the VDR−/− (VDRCre+/+) offspring (Fig. 2B). The VDR+/− kidneys had a two-fold reduction in protein expression compared to the VDR +/+ kidneys and VDR−/− did not show any Vdr expression. In addition, the VDR Cre+/+ mice had alopecia (data not shown), which has been shown previously to occur in VDR KO mice [18, 19]. The data confirms the disruption of the VDR gene in the VDRCre+/− and VDR Cre+/+ mice.

Figure 2:

Vdr expression in VDR +/+, VDR +/− and VDR −/− offspring of the VDRCre+/− mice. A) WT mice were used as a source of tissue for Vdr mRNA and protein expression. Vdr expression was normalized to WT kidney and Gapdh expression (expressed as fold expression). Western blots were done for the Vdr and Gapdh followed by densitometric analysis. Fold change was calculated relative to WT kidney Vdr corrected for Gapdh. n=2–5 mice per group. B) Western blots of the kidneys from VDR +/+ (VDRCre−/−, n=8) VDR +/− (VDRCre+/−, n=9), and VDR −/− (VDRCre +/+, n=6) for Vdr and Gapdh. Fold change was calculated relative to VDR +/+ kidney and corrected for Gapdh. Values are mean + SEM of n=6–9 mice per group. One way ANOVA with Bonferroni multiple comparison tests were done, *P<0.5, **P<0.05.

Generation of the VDR^tdTomato reporter line.

VDRCre+/− mice were crossed to the Ai4 reporter mice that have a floxed stop codon upstream of the tdTomato gene, driven by the Rosa26 promoter (Fig. 3A). Expression of the VDRCre transgene results in irreversible tdTomato expression following genomic excision of the stop codon. The VDR Cre+/− (VDR^tdTomato+) mice were visibly pink compared to the VDRCre−/− (WT) littermates (Fig. 3B). In addition, tissues known to express high levels of the VDR such as the kidney, colon and SI of the VDR^tdTomato+ were all pink compared to the same tissues from the WT littermates (Fig. 3B). Seventy percent of the lymphocytes in the spleen of VDR^tdTomato+ mice expressed tdTomato (Fig. 3C). Furthermore, 75–88% of the CD19+ B cells and 60% of the CD3+ T cells expressed tdTomato in the spleen (Fig. 3C and 4A). High VDR expressing tissues were pink in the VDR^tdTomato+ mice and a high fraction of the T and B cells in the spleen were tdTomato+.

Figure 3:

Characterization of the VDR^tdTomato reporter mouse. A) VDRCre+/− mice were crossed with Ai4 mice that have a floxed stop in front of the tdTomato flourescent protein in the Rosa26 gene. Cre expression results in the irreversible expression of tdTomato. The offspring of the cross were +/− for tdTomato and either VDRCre−/− (WT) or VDRCre+/− (VDR^tdTomato+). B) The faces, paws and tails of the VDR^tdTomato+ mice are pink compared to the WT littermates. In addition, the kidney, SI and colon of the VDR^tdTomato+ mice were visibly pink compared to the WT littermates. C) Lymphocytes in the spleen were gated based on side scatter and forward scatter. Red tdTomato flourescence is shown in WT and VDR^tdTomato+ lymphocytes. B cells were stained for CD19 and T cells were stained for CD3. The frequency of red tdTomato+ cells within the B cell and T cell population of VDR^tdTomato+ splenocytes.

Figure 4:

Comparison of immune cell frequencies in WT and VDR^tdTomato+ littermates. A) Cell frequency and tdTomato+ frequency of B cells, CD4+, CD8+ and NK cells were measured in the spleen. B) Hematopoietic progenitors in the BM of WT and VDR^tdTomato+ mice. LSK (Lin− Sca1⁺ckit⁺), hematopoetic stem and progenitor cells (HSPC), multipotent progenitor (MPP) and common helper innate lymphoid progenitor (CHILP) were identified by flow cytometry. B) Double negative (CD4⁻CD8⁻), double positive (CD4⁺CD8⁺) and single CD4⁺ or CD8⁺ positive T cells measured in the thymus of WT and VDR^tdTomato+ mice. C) NK cells, ILCs, T cells and macrophage were isolated from the lung tissue of WT and VDR^tdTomato+ mice. n=3–5 mice per group. Data are mean ± SEM. Two-way ANOVA and Student’s t-tests were used to assess statistical significance.

In VDR^tdTomato+ mice, one copy of the VDR gene is inactivated by knock-in of Cre recombinase. In order to rule out an effect of being hemizygous for the VDR on the numbers of immune cells, the frequencies of the major subtypes of immune cells were determined in VDR^tdTomato+ and WT littermates (Fig. 4). The immune cells in the peripheral tissues largely derive from the BM and T cells undergo further development in the thymus. The Lin⁻Sca1⁺cKit⁺ (LSK) give rise to hematopoietic stem and progenitor cell (HSPC, Lin⁻Sca1⁺cKit⁺CD135⁺), then multipotent progenitor (MPP, Lin⁻Sca1⁺cKit⁺CD135⁺CD150⁻CD48⁻) and then the common helper innate lymphoid progenitor (CHILP, CD127⁺a₄b₇+CD135⁻CD25⁻ID2⁺) cells [20]. The frequencies of LSK, HSPC, MPP and CHILP were similar in the BM of WT and VDR^tdTomato+ mice (Fig. 4B). T cells develop in the thymus from double negative (CD4⁻CD8⁻), to double positive (CD4⁺CD8⁺), to single positive CD4⁺ and CD8⁺ T cells respectively [21]. There was no difference in the frequency of thymic T cell progenitors and single positive T cells isolated from WT and VDR^tdTomato+ mice (Fig. 4B). Similarly, the frequency of T (CD3+) or B cells (CD19+), CD4+, CD8+ and macrophage (CD11b⁺MHCII⁺CD11c⁺) isolated from WT and VDR^tdTomato+ mice were not different in the spleen, lung, colon, or SI (Fig. 4 and data not shown). Loss of one copy of the VDR did not affect the frequency of ILCs, T cells and their progenitors.

tdTomato is highly expressed in most immune cells.

In the BM, the frequency of tdTomato+ cells ranged from 70–90% (Fig. 5A). Not surprisingly, the thymus precursors were also 70–85% tdTomato+ (Fig. 5A). The tdTomato reporter is irreversibly expressed in any cell that expressed the VDR. Therefore, all cells that develop from the tdTomato+ BM or thymic precursors would also be tdTomato+. In the spleen, CD19⁺ B cells were 75–88%, CD3⁺ T cells were 69–81%, and NK1.1⁺/CD3⁻ NK cells were 82–89% tdTomato+ (Fig. 4A). Splenic monocytes (CD11b⁺F4/80⁻ Ly6G⁻) were 94±1%, macrophage (CD11b⁻F4/80⁺) were 88±1% and neutrophils (CD11b⁺F4/80⁻ Ly6G⁺) were 98±0.2% tdTomato+. The frequency of tdTomato+ NK cells (Lin⁻CD127⁻ Tbet⁺Eomes⁺) in the lung and SI was high (80–85%) and similar to the frequency of tdTomato+ NK cells in the spleen (Fig. 5C–D). Innate lymphoid cells are found primarily in the mucosal tissues. ILC1 (Lin⁻CD127⁺Tbet⁺Eomes⁻) and ILC3 (Lin⁻CD127⁺RORgt⁺GATA3⁻) cells had high tdTomato expression (60–80%) in the lung, SI and colon (Fig. 5C–D). ILC2 (Lin⁻ CD127⁺GATA3⁺RORgt⁻) cells were largely tdTomato− in the lung and colon and only 37% of the ILC2 cells in the SI were tdTomato+. The data suggest that most of the peripheral immune cells derive from progenitors that have expressed the VDR. However, there is a small subset of immune cells including the tissue resident ILC2 cells that have developed from cells that have never expressed the VDR.

Figure 5.

tdTomato expression in BM, thymus and mucosal tissues. tdTomato expression was characterized in precursor cells in the A) BM, and B) thymus. Tissue resident ILCs in the C) lungs, D) SI and colon in adult VDR^tdTomato+ mice were evaluated by flow cytometry. Values are from n=2–6 mice. Data are mean ± SEM.

Even though the thymic progenitors are 70–80% tdTomato+, the spleen has about 25% of the T cells that are tdTomato− (Fig. 4). The CD4+ and CD8+ cells were subdivided into naive (CD62L^high/CD44^low) and effector (CD62L^low/CD44^high) cells and the frequency of tdTomato+ cells was evaluated (Table 1). We observed no significant difference in the frequency of tdTomato+ cells when comparing naive versus effector CD4+ or CD8+ T cells (Table 1). Vdr protein expression is very low in the spleen (Fig. 2A), and activation of T cells induces the expression of the VDR [15]. We hypothesized that activation would induce VDR expression in T cells and that the reporter gene would be expressed in all proliferating T cells. To determine whether activation would increase the VDR^tdTomato+ cell frequency, splenocytes were labelled with CFSE and stimulated with anti-CD3 antibodies. With proliferation the CFSE dye is diluted sequentially with each cell division. Cells that proliferate more have less CFSE fluorescence. The data showed that 84% CD4+ T and 75% CD8+ T cells were tdTomato+ before stimulation. With stimulation, the percentage of tdTomato+ CD4+ or CD8+ T cells decreased while the percentage of tdTomato− CD4+ or CD8+ T cells increased (Fig. 6A); which is consistent with greater proliferation of the tdTomato− T cells (Fig. 6A). Of the cells that had undergone proliferation (as measured by CFSE dilution), only 60% of the CD4+ cells and 30% of the CD8+ T cells expressed the tdTomato reporter gene after 72 hours (Fig. 6A). tdTomato expression in CD4+ and CD8+ T cells was highest before activation and declined over time post-stimulation (Fig. 6A). The reduction in tdTomato+ frequency was greater in CD8+ than CD4+ T cells (Fig. 6A). The data suggest that the tdTomato− T cells may be proliferating more quickly than the tdTomato+ T cells.

Table 1.

Frequency of tdTomato+ cells in naive and effector T cells from the spleen.

	Unseparated1	Naive CD62Lhigh/CD44low	Effector CD62Llow/CD44high
CD4+	76 + 4 %	83 + 2 %	70 + 3%
CD8+	70 + 5%	78 + 5%	68 + 3%

¹Values are the mean + SD of n=3 mice.

Figure 6.

tdTomato− T cells divide more than tdTomato+ T cells. A) tdTomato expression in proliferating CD4+ and CD8+ T cells after 0–72h of stimulation. Data are from 1 representative of 2 independent experiments. B) tdTomato flouresence of sorted TCRβ+/tdTomato+, TCRβ+/tdTomato− T cells and 72h stimulated TCRβ+ tdTomato− T cells (lower panel). Comparison of CFSE dilution between sorted tdTomato+ and tdTomato− CD4+ or CD8+ T cells after 72h of stimulation. C) Vdr protein expression in sorted tdTomato+ T cells before (0h) and after (72h) stimulation. Ifn-γ production in cultures of sorted tdTomato− and tdTomato+ T cells stimulated for 48h or 72h. Values are mean ± SEM of n=3–7 cultures from 3 independent experiments, ****P<0.0001.

To investigate whether tdTomato− T cells may be proliferating more quickly than the tdTomato+ T cells, the tdTomato+/TCRβ+ cells and tdTomato−/TCRβ+ T cells were sorted to 98% purity, stained with CFSE and then stimulated with anti-CD3/CD28 antibodies (Fig. 6B). The tdTomato−CD4+ and CD8+ T cells proliferated more as indicated by lower CFSE staining than their tdTomato+ counterparts at 72 hours post-stimulation (Fig. 6B). In addition, IFN-γ secretion was higher in the supernatants of tdTomato− T cells than tdTomato+ T cells at 72 hours (Fig. 6C). Interestingly, activation of the sorted tdTomato− T cells for 72 hours did not induce expression of the tdTomato reporter (Fig. 6B). Western blot results showed that the tdTomato+ cells had undetectable levels of the VDR until after stimulation (Fig. 6C). The data suggest that VDR expression may separate CD4+ and CD8+ T cells into phenotypically distinct faster proliferating VDR negative and slower proliferating VDR positive subsets.

Discussion

The VDR^tdTomato+ mouse is a sensitive way to identify individual cell types that have previously expressed the VDR. Hematopoiesis first occurs in the mid-gestation embryo including the yolk sac, placenta and fetal liver. After birth, hematopoiesis largely occurs in the BM [22]. The LSK fraction, which contains hematopoietic stem and progenitor cells in the adult BM are 90% tdTomato+ and therefore the LSK has already expressed the VDR (Fig. 5). All cells derived from the tdTomato+ BM precursor should also be tdTomato+. Cell populations in the periphery that are tdTomato− are a result of the development of the minority of tdTomato− precursors. BM progenitors have previously been shown to be vitamin D targets. Hematopoietic cell lines and BM cells have been shown to express VDR mRNA [23–25]. BM from VDR KO mice had reduced numbers of common myeloid progenitors and more long-lived hematopoietic stem cells (Lin-Sca1+CD117+CD150+CD48-) [26]. In addition, embryonic hematopoietic stem and progenitor cells in a zebra fish model expanded more rapidly in response to 1,25(OH)₂D treatments [27]. Interestingly, there is a minor population of progenitors that have never expressed the VDR and give rise to immune cells in the peripheral tissues suggesting that Vdr expression during development marks two different compartments in these cell lineages. However, the majority of immune cells do develop from the VDR expressing progenitors and therefore have the potential to be regulated by vitamin D.

ILC1 and ILC3 cells in the tissues were tdTomato+, while ILC2 cells were tdTomato−. Like other immune cells, ILCs develop from lymphoid progenitors initially in fetal liver and later in BM. The common lymphoid progenitor gives rise to the CHILP that develop into tissue ILC. Over 80% of the CHILP in the BM express tdTomato. Tissue resident ILC can be found in the lung and SI by post-natal day 7 and increase to adult numbers by 4wks of age [28]. It was not unexpected that the ILC1 and ILC3 cells in the lung, SI and colon were largely tdTomato+. Interestingly, the tissue resident ILC2 cells in adult mice are unique from the ILC1/ILC3s in that they have been shown to develop from HSC in the fetal liver as well as CD4/CD8 DN fetal thymocytes [29–33]. The majority of the ILC2 cells in the murine lung expanded de novo from neonatal ILC2 instead of coming from the BM [32]. The data in Fig. 5 suggests that the ILC2 in the lung develop from a VDR negative precursor, likely from the unique ILC2 tissue resident pathway of development [29–33]. It would be interesting to know whether activation of the ILC2 could induce VDR/tdTomato expression. Conversely, the ILC1 and ILC3 cells in the mucosal tissues develop from a pool of cells that have expressed the VDR previously and as a result are vitamin D targets.

Vitamin D has a well described role in the regulation of T cell proliferation. Resting T cells express low levels of the VDR that are increased with activation [34–36]. 1,25(OH)₂D treatment inhibited T cell proliferation [4, 37] in part by blocking cell cycle progression [38]. T cells derived from VDR expressing progenitors proliferated less than T cells that had never expressed the VDR (Fig. 6). Interestingly, the effect was more pronounced in CD8+ versus CD4+ T cells (Fig. 6). CD8+ T cells from VDR KO mice proliferated faster both in vitro and in vivo than WT CD8⁺ T cells [39]. The existence of VDR negative and VDR positive subsets of CD4 and CD8 T cells is novel. 1,25(OH)₂D was not added to the in vitro cultures, but there was serum in the media that may have provided a source of vitamin D. It may be that the effects of 1,25(OH)₂D on proliferation is to stabilize VDR protein expression, in the VDR expressing subset, to partially constrain the T cell response following activation. The tdTomato+ T cells upregulated the VDR, proliferated more slowly and made less Ifnγ. The data are in line with the previously well described anti-proliferative and anti-inflammatory effects of vitamin D and the VDR in T cells [15, 39].

The VDR^tdtomato+ reporter mouse model provides new insights into functionally distinct subsets of immune cells, that derive from VDR dependent and independent developmental pathways. It would be important to know the precise timing of VDR expression during immune system development and which fetal progenitors express the VDR and at what age. Since there are tdTomato− T cells in the peripheral tissues, it would be interesting to compare the response of tdTomato+ and tdTomato− T cells and their gene expression profiles following in vivo immune activation or infection. It would be interesting to compare our results with the mouse created by Liu and colleagues [17]. However, Liu et. al used a similar reporter mouse and so the limitations of both models are that Vdr protein may not be present in tdTomato+ cells [17].

Conclusions

A mouse that expresses Cre under the control of the VDR gene was made and the result of the insertion was shown to inactivate the VDR gene. The cross with the tdTomato reporter mouse (VDR^tdTomato+) showed that tissues with high VDR expression were visually pink. The data showed that most immune cells in the periphery are derived from hematopoietic stem cells that have previously expressed the VDR. Almost all neutrophils, monocytes and macrophage were tdTomato+. ILC2 were mostly tdTomato− and had never expressed the VDR. A minority of the T cells in the periphery were tdTomato−. The tdTomato− CD8+ and CD4+ T cells proliferated more than their tdTomato+ counterparts. Interestingly the activated tdTomato− T cells did not express the VDR/tdTomato reporter following activation. The VDR^tdTomato+ mouse provides a novel tool to sensitively and selectively identify immune cells that have expressed the VDR.

Material and methods

Mice.

The VDRCre mouse model was generated by CRISPR technology at the Albert Einstein College of Medicine Transgenic Mouse Core (New York, NY). The guide RNA was targeted to exon 3 of the VDR gene. Exon 3 was selected because it is the first coding exon. The Cre-polyA cassette was inserted immediately downstream of codon 14 of the Vdr gene. According to the design, a fusion protein consisting of 14 amino acids at the N-terminus of the VDR gene and Cre is expressed under the control of the endogenous VDR promoter. Due to the presence of the polyA signal (located immediately downstream of Cre) and the damage to the open reading frame, the VDR gene will not be expressed. The resulting founder (F0) male mouse, with the confirmed homologous recombination mediated knock-in of the exogenous DNA, was mated with WT female C57BL/6 mice. The VDRCre+/− offspring were then maintained as a separate line and crossed with homozygous Ai4 (B6.Cg-Gt(ROSA)26Sor^{tm14(CAG-tdTomato)Hze/J}, Jackson Laboratory, Bar Harbor, ME) mice. The Ai4 mice have a floxed stop in front of the tdTomato flourescent protein in the Rosa26 gene. Offspring were genotyped for VDR and Cre expression (Fig. 1) as well as the tdTomato transgene. Supplementary (S)Table 1 contains the primer sequences for the genotyping. All mice were bred and maintained according to approved IACUC protocols at the Pennsylvania State University (University Park, PA).

RNA isolation and Quantitative PCR.

Spleen, liver, lung, kidney, thymus, SI and colon were collected and homogenized. For RNA analysis homogenates were in TRIzol reagent (Sigma-Aldrich, Missouri). Chloroform was added to the lysates and following centrifugation, RNA was isolated from the upper aqueous layer. The RNA was precipitated using isopropanol and 70% ethanol and dried. The purified RNA was dissolved in DEPC-treated water and quantified using a NanoDrop (ThermoFisher, Waltham, MA). One μg of RNA was reverse transcribed using AMV Reverse Transcriptase (Promega, Madison, WI) and the manufacturer’s instructions. Primers for Cre, Vdr and Gapdh were designed using NCBI primer BLAST and obtained from Integrated DNA Technologies (IDT, Coralville, IA). The primer sequences are listed in STable 1. mRNA was quantitated using SYBR green mix (Azura Genomics, Raynham, MA) and StepOne Plus system (Applied Biosystems, Carlsbad, CA). Gene expression was calculated using the deltadelta Ct method and corrected for Gapdh.

Western blotting.

Spleen, liver, lung, kidney, SI and colon were collected and homogenized in RIPA buffer (ThermoFisher, Waltham, MA) containing 1X protease inhibitor cocktail (Sigma Aldrich, St. Louis, MO) and 5mM sodium orthovanadate. Total protein was isolated and the concentration of total protein measured by Pierce BCA protein assay (ThermoFisher, Waltham, MA). 30μg protein was loaded onto 10% SDS page gel and transferred to 0.45mm PVDF membrane using Trans-gel blotting system (Biorad, Hercules, CA). The membrane was blocked with 10% non-fat milk and incubated overnight at 4°C with monoclonal mouse anti-VDR (1:100, Clone: D-6, Santa Cruz, Dallas, TX) and Rat anti-GAPDH (1:10,000, Clone: W17079a, Biolegend). HRP conjugated goat anti-mouse IgG (H+L) secondary antibody (1:5000, Novagen, Germany) and goat anti-rat IgG (1:20,000, Clone: Poly054, Biolegend) were used for VDR and GAPDH respectively. The D-6 antibody is to the C terminus of the human VDR and has been shown to specifically stain WT mouse duodenal lysates but not VDR KO (exon 3 targeted deletion) duodenal lysates by western blot [13, 19]. All antibody dilutions were prepared in tris buffered saline containing 0.05% tween and 5% non-fat dry milk. The blot was developed using Thermo West Femto SuperSignal kit (ThermoFisher, Waltham, MA). Densitometric analysis was done using Image J software. Fold change was calculated relative to WT kidney and normalized to Gapdh protein expression.

Isolation of cells

BM cells were isolated from both femurs using a 25-gauge needle (BD Biosciences, San Jose, CA) and phosphate buffered saline (PBS, pH=7.2) to expel the BM cells out. Single cell suspensions were obtained by sequentially passing the cells through 18G and 16G needles, followed by treatment with red blood cell lysis buffer. Single cell suspensions of the spleen and thymus were obtained by homogenizing the whole tissue in PBS (pH=7.2), followed by treatment with red blood cell lysis buffer. After removing the Peyer’s patches from the SI, the SI and colons were cut open longitudinally and sliced into smaller fragments. The tissues were incubated twice with Ca²⁺/Mg²⁺ free-hanks balanced salt solution (HBSS) containing 1mM 1,4 dithiothreitol (DTT, Sigma Aldrich) and 10mM EDTA at 37°C, 20 minutes each at 250 rpm in a shaking incubator. The remaining tissue fragments were further minced and incubated for 1 hour in collagenase type 1 (300 U/mL, Worthington, Lakewood, NJ) containing RPMI-C media at 37°C, 250 rpm. The lamina propria containing cells were layered onto 40/80% Percoll (Sigma Aldrich) gradients to remove debris. The interface contained the purified lamina propria cells. Lung tissue was minced and incubated with Ca²⁺/Mg²⁺ free HBSS containing 1mM 1,4 dithiothreitol (Sigma Aldrich) and 10mM EDTA at 37°C for 30 minutes. The tissue pieces were further digested in RPMI-C containing collagenase type I (300U/ml, Worthington) and DNase I (Sigma-Aldrich) for 1 hour at 37°C. Single cell suspension was then obtained by filtering through a 70mm filter and treatment with red blood cell lysis buffer.

Flow cytometry

A list of the antibodies used for flow cytometry can be found in STable 2. Cells were stained in PBS (pH = 7.2) containing 5% FBS and 0.01% sodium azide for 20 minutes at 4°C, followed by fixation with 2% paraformaldehyde. For intracellular staining, cells were fixed and permeabilized for 15 minutes at room temperature and stained with intracellular antibodies in permeabilization buffer, overnight at 4°C (eBioscience). All samples were run using BD LSR II Fortessa at the Flow Core Facility (Pennsylvania State University) and analyzed using Flo Jo ver 10.6 (BD Biosciences, San Jose, CA). Single positive and fluorescent minus one controls were used for setting the gates. Flow cytometric markers used for identification of each cell type are listed in STable 3.

T cell sorting and cultures

Splenocytes from VDR^tdTomato mice were labeled with 1ng/ml anti− T cell receptor (TCR)β PEcy-7 (H57–59.7) flow antibody (BD Biosciences). After staining, cells were centrifuged and resuspended in sterile sorting buffer (1x PBS with 1mM EDTA, 25mM HEPES pH 7.0 and 1% FBS). The cells were sorted into TCRβ⁺ and tdTomato+ or tdTomato− using a Coulter MoFlo Astrios (Beckman Coulter, Muskogee, OK) into tubes containing RPMI-C with 50% FBS. Sorted tdTomato⁺ TCRβ⁺ and tdTomato⁻TCRβ⁺ were used for isolation of protein and western blots, culturing and cytokine analysis.

Cells were labeled with carboxyfluorescein succinimidyl ester (CFSE) using CellTrace™ CFSE Cell Proliferation Kit (Thermo Fisher Scientific, Waltham, MA). Plate bound anti-CD3ε (10 μg/ml, TONBO Biosciences, San Diego, CA) was used to stimulate CFSE stained splenocytes. One million sorted T cells were stained with CFSE and stimulated with plate bound anti-CD3ε antibody and anti-CD28 (2 μg/ml, BD Biosciences) for 0–72h. Supernatants were collected from the cultures after 48h or 72h of stimulation. Ifn-γ in the supernatant was evaluated using an ELISA kit and following the manufacturer’s instructions (BD Biosciences).

Statistical analysis

Results are represented as mean ± SEM values from pooled experiments. Statistical analysis was performed using Prism ver. 9 (GraphPad, San Diego, CA) using One-way ANOVA, Two-way ANOVA, Mixed-effects analysis and Mann-Whitney t-test, or Bonferroni post-hoc tests as applicable. P-value < 0.05 was considered significant.

Highlights.

• A reporter mouse that expresses a fluorescent protein under the control of the vitamin D receptor (VDR) gene was used to identify individual cells for VDR expression.

• Early hematopoietic precursors express the vitamin D receptor. • Most macrophage and neutrophils in the periphery have previously expressed the vitamin D receptor.

• Some T cells and innate lymphoid type 2 cells develop from precursors that have never expressed the vitamin D receptor.

• In T cells expression of the vitamin D receptor corresponded to slower proliferation and less Ifn-γ production.