EDITOR’S CHOICE Vitamin D differentially regulates colon stem cells in patient-derived normal and tumor organoids Asuncio n Fernandez-Barral1,2, Alba Costales-Carrera1,2, Sandra P. Buira3, Peter Jung4,5, Gemma Ferrer-Mayorga1,2, Marıa Jesu  s Larriba1,2, Pilar Bustamante-Madrid1,2, Orlando Domınguez , Francisco X. Real2,6, Laura Guerra-Pastria 6 n7, Miguel Lafarga8, Damian Garcıa-Olmo , Ramo 3,9  n Cantero , LuisDel Peso , Eduard Batlle4,11,12, Federico Rojo2,3, 9 1,10 Alberto Mun ~ oz 1,2 and Antonio Barba chano1,2 1 Departments of Cancer Biology and Biochemistry, Instituto de Investigaciones Biome dicas ‘Alberto Sols’, Spanish National Research Council (CSIC)-Autonomous University of Madrid (UAM) and IdiPAZ, Barcelona, Spain 2 Biomedical Research Networking Centres-Oncology (CIBERONC), Madrid, Spain 3 Departments of Pathology and Surgery, Instituto de Investigacion Sanitaria-Fundacio n Jimenez Dıaz, Madrid, Spain 4 Institute for Research in Biomedicine Barcelona (IRB), Barcelona, Spain 5 DKTK, German Cancer Consortium, Research Group, Institute of Pathology, Ludwig-Maximilians University, Munich, Germany German Cancer Research Centre (DKFZ), Heidelberg, Germany 6 Spanish National Cancer Research Centre (CNIO), Madrid, Spain 7 Department of Pathology, La Paz University Hospital-IdiPAZ, Madrid, Spain 8 University of Cantabria-IDIVAL, Santander, Spain 9 Colorectal Unit, Department of Surgery, La Paz University Hospital-IdiPAZ, Madrid, Spain 10 Biomedical Research Networking Centres-Respiratory Diseases (CIBERES), Madrid, Spain 11 Centro de Investigacion Biomedica en Red de Cancer (CIBERONC), Barcelona, Spain 12 ICREA, Passeig Lluıs Companys 23, 08010, Barcelona, Spain Keywords Intestine is a major target of vitamin D and several studies indicate an associ- colon cancer; colon stem cells; organoids; ation between vitamin D deficiency and inflammatory bowel diseases (IBD), stemness genes; vitamin D but also increased colorectal cancer (CRC) risk and mortality. However, the putative effects of 1a,25-dihydroxyvitamin D3 (calcitriol), the active vitamin Correspondence A. Mun ~oz and A. Barbachano, Departments D metabolite, on human colonic stem cells are unknown. Here we show by of Cancer Biology and Biochemistry, Instituto immunohistochemistry and RNAscope in situ hybridization that vitamin D de Investigaciones Biome dicas ‘Alberto receptor (VDR) is unexpectedly expressed in LGR5+ colon stem cells in Sols’, Spanish National Research Council human tissue and in normal and tumor organoid cultures generated from (CSIC)-Autonomous University of Madrid patient biopsies. Interestingly, normal and tumor organoids respond differen- (UAM) and IdiPAZ, Madrid, Spain tially to calcitriol with profound and contrasting changes in their transcrip- Tel: +34-91-5854451 tomic profiles. In normal organoids, calcitriol upregulates stemness-related E-mails: amunoz@iib.uam.es (AM); abarbachano@iib.uam.es (AB) genes, such as LGR5, SMOC2, LRIG1, MSI1, PTK7, and MEX3A, and inhi- bits cell proliferation. In contrast, in tumor organoids calcitriol has little Alberto Mun~oz and Antonio Barbachano effect on stemness-related genes while it induces a differentiated phenotype, share co-senior authorship and variably reduces cell proliferation. Concordantly, electron microscopy (Received 19 June 2019, accepted 12 July showed that calcitriol does not affect the blastic undifferentiated cell pheno- 2019) type in normal organoids but it induces a series of differentiated features in doi:10.1111/febs.14998 tumor organoids. Our results constitute the first demonstration of a regula- tory role of vitamin D on human colon stem cells, indicating a homeostatic effect on colon epithelium with relevant implications in IBD and CRC. Abbreviations ChIP, chromatin immunoprecipitation; CRC, colorectal cancer; EGF, epidermal growth factor; FBS, fetal bovine serum; GSEA, gene set enrichment analysis; IBD, inflammatory bowel diseases; IGV, integrative genome visualization; PBS, phosphate-buffered saline; PGE2, prostaglandin E2; RSPO1, R-spondin 1; VDR, vitamin D receptor. The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of 53 Federation of European Biochemical Societies. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made. Vitamin D regulates human colon stem cells A. Fern andez-Barral et al. Introduction phenotype and gene expression. Our results strongly support a major regulatory role of vitamin D on Calcitriol (1a,25-dihydroxyvitamin D3) is the active human colon normal and cancer stem cells that has vitamin D metabolite and a main regulator of the potential implications in important medical conditions physiology and homeostasis of the intestine. It induces such as IBD and CRC. Ca2+ and phosphate absorption, and contributes to epithelial integrity/barrier function, detoxification, and protection against infection [1]. These actions of cal- Results citriol are the consequence of it binding to and activat- ing vitamin D receptor (VDR), a member of the VDR is expressed in human colon crypt stem nuclear receptor superfamily that acts predominantly cells as a transcription factor modulating the expression of To analyze the expression of VDR protein in human a high number of genes in a tissue- and cell-dependent colon mucosa, we performed immunohistochemistry fashion [2,3]. Epidemiological and preclinical studies assays in healthy tissue from CRC patients. In the suggest an association between vitamin D deficiency same sections, to identify the colon stem cells, we car- and inflammatory bowel diseases (IBD), various ried out RNAscope in situ hybridization for detection extraskeletal disorders, and several neoplasias, particu- of the stem marker LGR5 [9] as all available anti- larly colorectal cancer (CRC) [1,3,4]. Supporting this, LGR5 antibodies showed a lack of specificity in test experimental results in immortal carcinoma cell lines experiments (data not shown). We observed a strong and animal models of CRC demonstrate a multilevel predominantly nuclear VDR signal in both top differ- protective action of calcitriol and other VDR agonists entiated enterocytes/cells and a population of cells [3,5]. located at the crypt bottom, while those in the mid- High level of Vdr RNA has been found in the small crypt displayed a lower signal (Figs 1A and S1). Quan- and large intestine/colon as compared to other mouse tification of data at cell level in 100 crypts from four tissues [6]. VDR was reported to be preferentially CRC patients (25 crypts/patient) revealed that VDR is expressed in top differentiated enterocytes, which sup- expressed in 64% of crypt bottom cells and that 87% ports its role in Ca2+ and phosphate absorption and of LGR5+ cells were positive for VDR expression barrier function, and to a lesser extent in cells at the (Fig. 1B). We additionally performed double crypt base [7,8]. However, a detailed analysis of VDR immunostaining of tissue microarrays of healthy colon protein expression and function in human colon tissue from 25 CRC patients with antibodies against VDR is lacking. Intestinal epithelium is the highest renewal and PTK7, a stemness-related cell marker [12]. In line body tissue, which has hampered its long-term cell cul- with LGR5 data, VDR (nuclear) and PTK7 (extranu- ture in vitro. However, the identification of LGR5 as a clear) proteins were largely co-expressed in crypt bot- marker of colon crypt bottom stem cells [9] made pos- tom cells (Fig. S1). These findings provide evidence, sible their isolation and unlimited growth as three-di- for the first time, that VDR is expressed in human mensional culture, forming structures called organoids colon stem cells in vivo. or mini-guts [10,11]. Organoid technology is a power- ful tool for multiple applications (diagnosis, personal- Human normal and tumor colon organoids ized medicine. . .) and has now allowed us to study VDR expression and calcitriol action in human colon express VDR and respond to calcitriol stem cells using CRC patient-derived organoids. To investigate the effect of vitamin D on human colon Here, we combined RNAscope in situ hybridization stem cells, we successfully generated a living biobank and immunohistochemistry to study the expression of of long-term normal and tumor organoid cultures LGR5 stem cell marker and VDR protein in human from 39 CRC patients who had undergone surgery for colon mucosa. Interestingly, our results show for the curative purposes (Table S1). Normal organoids had first time that a high proportion of crypt bottom stem relatively homogeneous cystic morphology formed by cells co-express LGR5 and VDR. This led us to investi- one or two cell layers, whereas tumor organoids were gate the role of calcitriol in colon stem cells. To this frequently multilayered and showed two alternative end, we generated a unique living biobank of CRC morphologies, cystic or compact (Fig. 1C) and resem- patient-derived organoids from matched healthy (nor- bled the primary tumor morphology (Fig. 1D). The mal) and tumor colon biopsies. Remarkably, matched analysis by RT-qPCR of VDR RNA expression normal and tumor organoids responded distinctly to revealed slightly higher levels in normal than in tumor calcitriol, which has contrasting effects on cell organoids (Fig. 2A). RNA expression of CYP24A1, 54 The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies. A. Fern andez-Barral et al. Vitamin D regulates human colon stem cells Fig. 1. Human colon stem cells express VDR. (A) Images of VDR immunofluorescence (red) and LGR5 RNAscope in situ hybridization (green) in human colon crypts. Scale bars: 50 lm left and 20 lm right. (B) Quantification of VDR and LGR5-positive cells in 25 crypts/patient (n = 4 patients). Bars represent mean  standard error of the mean (SEM). (C) Representative phase-contrast images of normal and tumor colon organoids. Scale bars: 500 lm. (D) Hematoxylin/eosin images showing the primary tumor architecture and tumor organoid phenotype. Scale bars: 100 lm. The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of 55 Federation of European Biochemical Societies. Vitamin D regulates human colon stem cells A. Fern andez-Barral et al. the most responsive calcitriol-target gene that encodes differentiation: heterochromatin patches, prominent the enzyme responsible for its degradation, was higher microvilli and adhesion structures (desmosomes) in tumor than in normal organoids, as occurs in tumor (Fig. 3A, lower panels), bundles of intermediate fila- vs. healthy colon tissue [13]. Calcitriol treatment ments, abundant, well-developed, rough endoplasmic increased CYP24A1 RNA to comparable levels in the reticulum cisternae, autophagic vacuoles, and Golgi two types of organoids (Fig. 2B). In addition, calcitriol complexes (Fig. 3B). increased both VDR and CYP24A1 proteins to a vari- We then explored whether the antiproliferative effect able extent in matched normal and tumor organoids of calcitriol reported in carcinoma cell lines [2,3] was from a subset of patients (Fig. 2C) and CA2, a protein reproduced in colon organoids. Calcitriol consistently typically expressed in normal colon organoids [14] diminished cell proliferation in all seven patient- (Fig. 2C). We also performed RNAscope in situ derived normal organoids under study and, to a vari- hybridization for LGR5 detection in organoids and, as able extent, in tumor organoids (Fig. 4A,B). No differ- in colon tissue, VDR protein and LGR5 RNA were ences were found in the number of organoids, which mostly co-expressed in normal and tumor organoids. suggests that calcitriol does not affect the clonogenicity VDR signal was increased by calcitriol in both types of either normal or cancer stem cells (Fig. 4C). Taken of organoids whereas that of LGR5 seemed to increase together, the VDR RNA level in normal and tumor apparently only in normal organoids (Fig. 2D). Alto- organoids show a statistically significant correlation gether, these results showed that both types of orga- with the antiproliferative response to calcitriol noids, derived respectively from human normal and (Fig. 4D,E), which may also depend on other cal- cancer stem cells, express VDR and respond to cal- citriol-related features and/or the tumor-specific muta- citriol. tional status. Together, these results indicate that calcitriol contributes to maintain the undifferentiated phenotype and restricts the proliferation of normal Calcitriol differentially modulates cell phenotype colon stem cells and their progeny, while it induces dif- and proliferation in normal and tumor colon ferentiation and variably inhibits the proliferation of organoids cancer stem cells. Based on the induction of differentiation by calcitriol on carcinoma cell lines [15], we next explored the effect Calcitriol differentially regulates the expression of this hormonal agent on the cell phenotype in both of stemness, differentiation, proliferation and types of organoids. Ultrastructural analysis of tumorigenesis genes in colon cancer patient- untreated normal organoids from three patients consis- derived normal and tumor organoids tently showed an undifferentiated cell phenotype with decondensed euchromatin and large nucleoli typical of Next we studied the long-term effect of calcitriol on high transcriptional activity, absence of intercellular the transcriptomic profiles of six matched normal and adhesion structures and microvilli, many free ribo- tumor organoid cultures by RNA-seq analysis upon somes, and few Golgi complexes and intermediate fila- 96 h of treatment. Given the differences in function ment bundles. Unexpectedly, calcitriol treatment did and mutational profiles between colon segments (right- not change the undifferentiated cell phenotype in nor- side, left-side) [16,17], to attenuate interindividual vari- mal organoids (Fig. 3A, upper panels). On the con- ability [18], we selected patients with left-side colon trary, cells of three tumor organoid cultures that (descending plus sigma) tumors (Table S1). Mutational showed a similar undifferentiated phenotype responded analysis of tumor organoids confirmed that they har- to calcitriol with changes compatible with bored patterns of genetic alterations that are typical of Fig. 2. Human colon patient-derived normal and tumor organoids express VDR and respond to calcitriol. (A) Box-plot of VDR RNA level in 18 normal and 9 tumor organoid cultures in relation to that of SW480-ADH colon cancer cells. Box plots represent median  max/min. Statistical analysis was performed by nonparametric Mann–Whitney test, *P < 0.05. (B) Box-plot of CYP24A1 RNA level in relation to that of SW480-ADH colon cancer cells (left) and its fold-change (right) in the organoid cultures used in (A) treated for 96 h with 100 nM calcitriol or vehicle. Box plots represent median  max/min. Statistical analysis was performed by nonparametric Kruskal–Wallis test using the Benjamini, Krieger and Yekutieli post test (left) and Mann–Whitney test (right), *P < 0.05, ***P < 0.001. (C) Western blot analysis of VDR, CYP24A1, and CA2 protein levels in whole-cell extracts from six matched normal and tumor organoid cultures treated for 96 h with 100 nM calcitriol or vehicle. b-actin was used as a loading control. The graphics below show the VDR/b-actin ratio. (D) Images of VDR immunofluorescence (red) and LGR5 RNAscope in situ hybridization (green) in human normal and tumor organoids treated for 96 h with 100 nM calcitriol or vehicle. Scale bars: 15 lm. 56 The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies. A. Fern andez-Barral et al. Vitamin D regulates human colon stem cells The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of 57 Federation of European Biochemical Societies. Vitamin D regulates human colon stem cells A. Fern andez-Barral et al. Fig. 3. Calcitriol induces cell differentiation in human tumor organoids. (A) Representative ultrastructural images of normal (patient #11) and tumor organoids (patient #38) treated with 100 nM calcitriol or vehicle for 96 h. Upper panel scale bars (from left to right): 2 lm, 2 lm, 2 lm, and 1 lm. Lower panel scale bars: 2 lm, 1 lm, 2 lm, and 1 lm. L, lumen; asterisks, heterochromatin aggregates; arrowheads, desmosomes; red-dotted line, intercellular region lacking mature adhesion structures. (B) Epithelial differentiation features induced by calcitriol in tumor organoids from patients #4 and #29. (1) Microvilli. (2) Heterochromatin clumps (arrows) and dilated intercellular spaces (asterisks). (3) Desmosomes (arrows). (4) Rough endoplasmic reticulum. (5) Golgi complexes. (6) Autophagic vacuoles (asterisk). Scale bars (from 1 to 6): 0.4 lm, 1 lm, 0.5 lm, 0.5 lm, 0.5 lm and 1 lm. 58 The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies. A. Fern andez-Barral et al. Vitamin D regulates human colon stem cells Fig. 4. Antiproliferative action of calcitriol in human normal and tumor organoids. (A) Proliferation assay of normal and tumor organoid cultures from seven patients treated with 100 nM calcitriol or vehicle for 10 days. Box-plot represents median  max/min. Statistical analysis was performed by nonparametric Wilcoxon signed-rank test, *P < 0.05, ***P < 0.001. (B) Phase-contrast images of human normal and tumor organoid cultures from (A). Broken line indicates the border of Matrigel. Scale bars: 1 mm. (C) Graph showing the number of normal and tumor organoids used in (A) after 10 days of calcitriol treatment. Box-plot represents median  max/min. Statistical analysis was performed by Wilcoxon signed-rank test. The exact P-values are indicated for nonsignificant results. (D) RT-qPCR analysis of VDR RNA level in normal and tumor organoids from (A) in relation to that of SW480-ADH colon carcinoma cells. Data are represented as mean  standard deviation (SD). (E) Scattegram showing the inverse correlation between VDR RNA levels and cell proliferation in organoids used in (A). Statistical significance was determined by Pearson test. sporadic CRC, with APC (5/6) and KRAS mutations differentially expressed genes, which mostly coincide (4/6) (Fig. 5A). Accordingly, nuclear staining of with those previously reported (Fig. 5B; GSE64392) MLH1, PMS2, MSH2, and MSH6 proteins in all cases [14]. Paired analysis of data from normal and tumor (data not shown) and lack of POLE mutation ruled organoids of each patient followed by joint analysis of out a hypermutated MMR phenotype. all patients’ data revealed that calcitriol significantly Transcriptomic profiles of untreated normal and changed the expression of 2,107 genes (943 upregu- tumor organoids showed a high number of lated, 1,164 downregulated) in normal organoids and The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of 59 Federation of European Biochemical Societies. Vitamin D regulates human colon stem cells A. Fern andez-Barral et al. 60 The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies. A. Fern andez-Barral et al. Vitamin D regulates human colon stem cells Fig. 5. Transcriptomic profile changes by calcitriol in human normal and tumor organoids. (A) Overview of the mutations found in the tumor organoid cultures of six selected patients. (B) Volcano plot comparing human normal and tumor RNA-seq signatures from the six matched organoid cultures analyzed in (A). The x-axis shows the fold-change (Log2) and the y-axis shows the q-value (Log2). Each dot represents a gene. Blue/red dots represent genes that are down- or upregulated (respectively) in tumor vs. normal organoids. Dots above the line are significant. (C) Volcano plot comparing normal (upper blot) and tumor (lower blot) organoid signatures from (B) treated with 100 nM calcitriol or vehicle for 96 h. (D) Venn diagram showing the overlap between calcitriol-significant regulated genes in normal and tumor organoids. The number of genes included in each group is depicted and the complete list of genes is in Table S2. (E) Graphs representing the linear correlation between the effect of calcitriol on gene expression (left graph, induction; right graph, repression), computed as the ratio (Log10 fold-change) of the RNA-seq counts, in treated organoids over controls. Only genes significantly (FDR<0.05) regulated by the treatment in normal organoids are represented. Green line, theoretical perfect correlation (r2 = 1). Statistical analysis was performed by Multiple r- squared regression test. (F) Heatmap showing significant genes commonly regulated by calcitriol in matched normal and tumor organoids with an average Log2 fold-change > 1 and expression > 4 cpm. (G) Heatmaps showing the genes with the greatest variance between sample groups of Log2 fold-change upon calcitriol treatment in matched normal and tumor organoids (left) and tumor RNA expression of those genes compared to its normal counterpart (right). 1,182 genes (643 upregulated, 539 downregulated) in analyses confirmed the upregulation by calcitriol of tumor organoids, which were partially coincident (532, multiple stemness-related genes including LGR5, 19.3%) (Fig. 5C,D and Table S2; GSE100785). A MSI1, PTK7, SMOC2, LRIG1, and MEX3A, a mar- strong correlation was found between genes induced ker of slow-dividing subpopulation of LGR5+ stem by calcitriol in normal and tumor organoids, whereas cells, in normal organoids (Fig. 6A) [9,12,21,22]. In the correlation was much weaker for repressed genes tumor organoids, the effects of calcitriol on stemness- (Fig. 5E). Common calcitriol-target genes in both related genes were weaker as only MSI1 and LRIG1 types of organoids included known VDR-regulated were upregulated (Fig. 6A). genes such as CYP24A1 and TRPV6 (Fig. 5F and Consistent with the ultrastructural analysis, calcitriol Table S2). Notably, some clusters of genes showed dif- induced desmosomal (DSC2), microvilli (PLS1), and ferent quantitative or even inverse regulation by cal- adhesion (CDH1) genes in tumor organoids citriol in normal and tumor organoids (Fig. 5G, left). (Table S2). We also confirmed by RT-qPCR that sev- The expression level of these genes in untreated nor- eral genes expressed in differentiated enterocytes mal and tumor organoids is shown (Fig. 5G, right). (KRT20) or in goblet/secretory cells (KLF4, SPDEF, Calcitriol targets also included genes involved in a and MUC2) were downregulated by calcitriol in nor- variety of metabolic, detoxification, nuclear receptor- mal but not in tumor organoids (Fig. 6A). Addition- related, and other pathways (Table S3). Supporting ally, Gene Set Enrichment Analysis (GSEA) showed post-translational regulation by vitamin D, as that the differentiation signature EPHB2low vs. described in several systems [19,20], VDR was not EPHB2high of human colon cells [12] was significantly identified as a significant regulated gene in the RNA- enriched in calcitriol-treated tumor, but not normal, seq analysis. organoids (Fig. 6B). Among the top up/downregulated genes by calcitriol We also observed that a series of genes involved in in normal and tumor organoids, there were some cell proliferation (LRIG1, GRK5, RARRES1, TFF2, involved in cell stemness (MSI1, SMOC2), differentia- and TNS4) and tumorigenesis (ALDH3A1, BCAS1, tion (CEACAM7, HMGCS2, and MUC2), prolifera- S100P, PPP1R14D, and SERPINE1) were regulated tion (GRK5, RARRES1), and tumorigenesis by calcitriol, preferentially in normal organoids, in a (ALDH3A1, S100P) (Table S4). So, next we used an way compatible with the previously described antipro- independent cohort of 18 normal and 9 tumor orga- liferative action of calcitriol and a potential antitu- noids to a) confirm the effect of calcitriol on the moral effect (Fig. 6A). Concordant with its effects on expression of several genes that were found to be sig- cell proliferation, GSEA revealed an inverse associa- nificantly regulated in the RNA-seq assay and, b) tion of the calcitriol profile (RNA-seq) with E2F, study the effect of calcitriol on genes involved in those mTOR, and c-MYC proliferative signatures (Fig. 6C). processes or pathways relevant for colon biology. As GSEA also showed that untreated tumor organoids expected, differentiation genes were expressed at higher were enriched in the ‘Core transcriptional embryonic levels in untreated normal than in tumor organoids stem cell (ESC)-like’ pro-tumorigenic signature that, as (Fig. S2). Interestingly, the opposite was found for described by Wong and cols [23] is frequently activated stemness-related genes, which showed lower expression in aggressive human epithelial cancers. Remarkably, in untreated normal than in tumor organoids calcitriol was able to abolish this association in both (Fig. S2). In line with the RNA-seq data, RT-qPCR types of organoids (Fig. 6D). In contrast to data on The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of 61 Federation of European Biochemical Societies. Vitamin D regulates human colon stem cells A. Fern andez-Barral et al. 62 The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies. A. Fern andez-Barral et al. Vitamin D regulates human colon stem cells Fig. 6. Calcitriol distinctly changes gene expression profile of human normal and tumor organoids. (A) RT-qPCR analysis of the RNA level of selected genes in an independent series of 18 normal and nine tumor organoid cultures treated with 100 nM calcitriol or vehicle for 96 h. Box plots represent median  max/min. Statistical analysis was performed by One-sample t-test (normal organoids) and nonparametric Wilcoxon signed-rank test (tumor organoids), *P < 0.05, **P < 0.01, ***P < 0.001. The exact P-values are indicated for nonsignificant results. (B) GSEA comparing the reported human healthy colon EPHB2-based differentiation gene signature and the RNA-seq analysis of human tumor colon organoids treated with calcitriol. (C) GSEA comparing the proliferation-related molecular signatures databases and the RNA-seq analyses of human normal and tumor organoids treated with calcitriol. (D) GSEA comparing the pro-tumorigenic signature core ESC-like with the RNA-seq analysis of genes up/downregulated in tumor vs. normal organoids (upper graph), and with that of normal or tumor organoids treated with calcitriol (lower graphs). carcinoma cell lines cultured in 2D conditions [15], cal- treatment (Fig. 7A). As in other systems [27], a large citriol did not repress classical genes of the Wnt path- number of VDR-binding sites were located in intronic way except DKK4 (Fig. 6A), a gene that increases or intergenic regions (Fig. S5A). Identified peaks were invasiveness, angiogenesis, and chemoresistance [24,25] assigned to the nearest gene at a maximum distance of (Table S5). 10 kb. As many as 559 genes were coincident between To confirm the specificity of the RNA-seq results, two ChIP-seq assays and 182 genes were coincident in we silenced VDR expression in normal organoids all three, including CYP24A1 and TRPV6 (Table S7; through lentiviral transduction of a shRNAmir against GEO107283). Moreover, GSEA revealed that the cal- VDR, which inhibited the regulation of several cal- citriol-direct target gene signature, defined as the 182 citriol-target genes (Fig. S3). Overall, these results genes commonly identified in all ChIP-seq assays, was show that calcitriol has different/opposite regulatory strongly associated with calcitriol-upregulated genes in effects on stemness and differentiation genes in human the RNA-seq analysis (Fig. 7B). This result agrees normal and tumor organoids. with the current model of VDR/nuclear receptor In addition, we generated colon organoids from action: direct binding to upregulated genes and pre- wild-type (wt) and, as control, Vdr-deficient (Vdr/) dominant indirect action by signaling interference, pro- mice and analyzed the transcriptomic profile induced tein–protein interaction or via microRNAs on by calcitriol treatment using RNA microarrays. As downregulated genes [28]. Remarkably, 68 genes of the expected, no genes were regulated by calcitriol in orga- 182-gene signature overlapped with those regulated (62 noids from Vdr/ mice, while 143 target genes were up- and 6 downregulated) by calcitriol in the RNA-seq identified in wt organoids. Remarkably, only 32 genes assay (Table S7). The stemness-related MSI1 and overlapped between those regulated by calcitriol in SMOC2 genes, several genes with antiproliferative human (RNA-seq analysis) and mouse (RNA microar- and/or tumor suppressive function (KCNIP3, rays, GSE105117) colon organoids (Fig. S4A), while PPP1R1C, PRKG2, PRR5L, GRK5, and RARRES1, 111 genes were exclusively regulated in mouse orga- the latter with a VDR-binding site 37 kb away from noids (Table S6). This result agrees with the reported 50 UTR), and CYP24A1 and TRPV6 (Fig. 7C) were lack of conservation of VDR response elements among the calcitriol-direct transcriptional targets between primates and rodents [26]. We also show that (Table S7). ChIP-seq assays also showed enrichment in the calcitriol expression profile in wt mouse organoids the consensus direct repeat (DR)3 VDR motif was related with the Mex3ahigh/Lgr5high mouse signa- (Fig. 7D). The comparison of our RNA-seq data anal- ture of slow-dividing intestinal stem cells [22] ysis with public transcription factor (TF) ChIP-seq (Fig. S4B). datasets confirmed VDR as the most significant TF involved in gene upregulation (Fig. S5B). Identification of direct calcitriol-target genes in human colon organoids Discussion To further elucidate the action of calcitriol in colon The association of vitamin D deficiency with the risk homeostasis, we searched for direct transcriptional tar- and severity of clinically relevant gut diseases such as get genes in normal organoids. To this end, three inde- IBD and CRC, together with the observation that cal- pendent chromatin immunoprecipitation-sequencing citriol induces a more differentiated phenotype of ente- (ChIP-seq) assays were performed using six additional rocytes in vivo and of colon carcinoma cells in 2D normal patient-derived organoids (2 per ChIP-seq) culture conditions [1,3,15], led us to analyze the effects treated with calcitriol for 2 h. A variable number of of calcitriol on human organoids. These 3D structures DNA binding sites for VDR were found after calcitriol constitute the state-of-the-art system to study normal The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of 63 Federation of European Biochemical Societies. Vitamin D regulates human colon stem cells A. Fern andez-Barral et al. 64 The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies. A. Fern andez-Barral et al. Vitamin D regulates human colon stem cells Fig. 7. Direct transcriptional calcitriol-target genes in human colon normal organoids. (A) Venn diagram showing the overlap between genes with VDR binding sites identified by ChIP-seq in three independent experiments using six human normal organoid cultures treated with 100 nM calcitriol for 2 h. (B) GSEA comparing the genes containing VDR-binding sites identified in the ChIP-seq assays and those regulated by calcitriol in normal organoids in the RNA-seq analysis. (C) Integrative Genome Visualization (IGV) representation of VDR-binding sites identified in the ChIP-seq assays in CYP24A1 and TRPV6 (controls), MSI1, GRK5, SMOC2, and RARRES1 genes. (D) Homer de novo motif analysis of VDR-binding sites in ChIP-seq studies. (E) Scheme showing the proposed action of vitamin D on normal and cancer colon stem cells. and tumoral epithelial stem cells in vitro. We report epidemiological studies indicate an association of vita- unanticipated differential effects of calcitriol on both min D deficiency with IBD, and experiments with ani- types of organoids that are relevant for understanding mal models suggest a beneficial effect of VDR agonists the relevance of vitamin D in normal and pathological in these conditions [31]. intestinal biology. We also determined that calcitriol declines, but does Vitamin D action in human colon has classically not abolish, cell proliferation in normal organoids. been investigated in immortalized carcinoma cell lines, Although Wnt and EGF pathways are essential for and recently in ex vivo short-lived colon biopsy sam- colon stem cell proliferation in vivo and, concordantly, ples and primary fibroblasts from patient-derived nor- for the generation of organoids in vitro [32,33], we mal and tumor tissue [29,30]. We report for the first show that calcitriol reduces cell proliferation in normal time that human colon stem cells express VDR in vivo. organoids, yet classical Wnt pathway genes were unaf- VDR is a major determinant of tissue responsiveness fected in our transcriptomic analyses. Instead, cal- to vitamin D. Others are the circulating level of the citriol upregulates LRIG1, marker of quiescent + 4 liver-synthesized calcitriol precursor 25-hydroxyvitamin stem cells [21,31] and also master regulator of epithe- D/calcidiol, the activity of CYP27B1 and CYP24A1 lial stem cells and tumor suppressor that encodes a enzymes that synthesize and degrade calcitriol, respec- multilevel tyrosine kinase receptor (EGFR and others) tively, and the nuclear expression of VDR co-repres- inhibitor [34,35]. Moreover, we observed in mouse, sors and co-activators [2,3]. This finding points to a due the lack of the equivalent human signature, that potential effect of calcitriol on the behavior of normal calcitriol induces a gene expression profile that is and cancer stem cells. related to the Mex3ahigh/Lgr5high signature of slow-di- To directly address these issues, we established a viding intestinal stem cells [22]. Additionally, calcitriol unique long-term culture collection of matched inhibits the activator of EGF pathway TFF2 gene [36] patient-derived normal and tumor organoids and and regulates the expression of the pro-tumorigenic tested the effects of calcitriol. First, we validated that gene TNS4 and of the negative regulators of colon car- VDR is co-expressed in vivo with LGR5 and PTK7, cinoma cell proliferation RARRES1 (also known as markers of colon crypt base columnar stem cells, and TIG1) and its target GRK5 [37]. Conceivably, these furthermore, it mediates in normal organoids the effects may contribute to the antiproliferative action of induction of LGR5 by calcitriol. The direct binding of calcitriol in crypt stem cells. VDR to, and transcriptional regulation of, other stem- Our results show that the gene regulation effect of ness-related genes such as MSI1 and SMOC2 in nor- calcitriol is much stronger in human than in mouse mal organoids further supports the stemness-inducing colon organoids, and that little coincidence exists action of calcitriol in colon tissue. Together, these between target genes in the two species. However, results strengthen the association of the vitamin D-in- Peregrina and collaborators reported the presence of duced gene expression program with the maintenance Vdr and a role of dietary vitamin D in the growth and of stemness of, perhaps, several colon stem cell popu- maturation of mouse intestinal Lgr5+ cells [38]. lations. Homeostasis of intestinal mucosa, one of the The finding that calcitriol has pro-differentiation and tissues with the highest cell renewal, depends on an variable antiproliferative effects lacking stemness induc- adequate balance between proliferation and differentia- tion in tumor organoids adds to the reported actions on tion of crypt bottom stem cells and their progeny in carcinoma cell lines and cancer-associated fibroblasts, response to a number of intrinsic and stromal signals which indicates that calcitriol is a multifaceted protec- such as Wnt, EGF, Notch, Hippo, BMP, and Hedge- tive agent against CRC. In line with the cancer stem cell hog. The new actions of vitamin D described in this model stating that tumors originate from the mutational study may contribute to the homeostasis of healthy alteration of resident tissue stem cells, human CRC is intestinal mucosa, and to the repair and regeneration mainly supported by mutated LGR5+ stem cells [39]. of damaged intestine upon injury. In line with this, The pro-differentiation effect of calcitriol on tumor The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of 65 Federation of European Biochemical Societies. Vitamin D regulates human colon stem cells A. Fern andez-Barral et al. organoid cells is in line, albeit less apparent, with that transferred to a 50 mL conical tube in fresh PBS. Colon observed in colon carcinoma cell lines grown on plastic crypts were separated from the mucosa after shaking and dishes [15]. Notably, VDR agonists have been proposed supernatant was centrifuged at 115 g for 5 min at RT. Crypts to promote terminal differentiation of normal upper were washed twice in washing buffer [Advanced DMEM/ crypt absorptive enterocytes [1], which however is not F12, 10 mM HEPES, and 10 mM Glutamax (Thermo Fisher observed in cells of normal colon organoids. This dis- Scientific)] and finally pelleted crypts were embedded in crepancy may rely on differences in the microenviron- Matrigel (Corning, MA, USA). Drops were seeded on pre- ment (location and other signals) and the distinct warmed, 48-well culture dishes. After Matrigel solidification at 37 °C complete ‘normal’ culture medium was added [50% phenotype of these two cell populations (top crypt ente- Advanced DMEM/F12, 50% Wnt3a-conditioned medium, rocytes/bottom crypt stem cells) in human tissue and 10 mM HEPES, 10 mM Glutamax, 10 mM Nicotinamide organoid culture, respectively. In this scenario, our (Sigma-Aldrich, St. Louis, MD, USA), 1x N2 (Thermo results favor a protective action of VDR agonists Fisher Scientific), 1x B27 (Thermo Fisher Scientific), 1 mM against CRC. Further supporting this we describe that N-acetyl-L-cysteine (Sigma-Aldrich), 1 : 500 Primocin, calcitriol downregulates a gene signature associated to 0.1 lgmL1 Noggin ( Peprotech, New Jersey, NJ, USA), aggressiveness and poor prognosis of breast and lung 1 lgmL1 Gastrin (Tocris, Bristol, UK), 1 lgmL1 RSPO1 carcinoma patients [23]. (Sinobiological, Beijing, China), 50 ngmL1 EGF (Pepro- In summary, the newly identified functions in tech), 0.02 lM PGE2 (Sigma-Aldrich), 1 lM LY-2157299 patient-derived organoids reveal an unanticipated (Axon-Medchem, Groningen, The Netherlands), and 10 lM homeostatic action of vitamin D in the human intesti- SB-202190 (Sigma-Aldrich)]. nal mucosa. This strongly supports a role of calcitriol To establish mouse normal colon organoids, colon crypts in the maintenance of colon crypt cell stemness and in were basically isolated as described for human crypts with the differentiation of colon cancer stem cells (Fig. 7E). minor modifications. Mouse colon was washed in PBS, incubated for 30 min with antibiotics (Primocin) and cut Materials and methods into small pieces. Next, tissue was incubated for 5 min with 8 mM EDTA at RT and 30 min at 4 °C in slow rotation. Human samples Then, it was washed in PBS to remove EDTA. Crypts were isolated after shaking and collected in a 50 mL conical tube Fresh human tissues were provided by IdiPAZ and Fun- after passing the solution through a 70-lm mesh filter. The daci on Jimenez Dıaz Biobanks, integrated into the Spanish crypts solution was centrifuged at 260 g for 5 min at 4 °C. Biobank Network ( www.redbiobancos.es), from individu- The pellet was washed in washing buffer and pelleted als diagnosed with colorectal cancer and subjected to sur- crypts were embedded in Matrigel on a prewarmed 48-well gery between 2013 and 2018. Normal tissue samples were culture dish. After Matrigel solidification, ‘mouse’ culture obtained from the area distal to the tumor and the histol- medium was added (‘normal’ culture medium minus Nicoti- ogy of the biopsies was evaluated by the pathology services namide, Gastrin, PGE2, and SB-202190). of La Paz University Hospital and Fundaci on Jimenez Human tumor organoid cultures were generated as fol- Dıaz. All human subjects gave informed consent. The study lows. Human tumor biopsies were washed in PBS several complied with ethical regulations and was approved by the times and incubated in a mixture of antibiotics (Primocin, Ethics Committee of La Paz University Hospital (HULP- Gentamycin, and Fungizone) for 30 min in rotation at RT. PI-1425) and the Fundacion Jimenez Dıaz (PIC-15/2014). To obtain single cells, tissue was cut into small pieces and digested enzymatically in a suspension of 1 mgmL1 colla- Establishment of 3D colon and tumor organoid genase type IV (Sigma-Aldrich) (in PBS) for 30 min at cultures 37 °C with continuous shaking in a waterbath. We forced cell disaggregation by passing the suspension through a To establish human normal colon organoids, colon crypts 18G syringe. Next, we added 5% FBS and single cells were were isolated from human biopsies as previously described collected in a conical 50 mL tube after passing the solution [10,11]. Briefly, colon mucosa biopsies were incubated with a through a 70-lm mesh filter and centrifugation at 250 g for mixture of antibiotics [Primocin (Invivogen, San Diego, CA, 5 min at 4 °C. To lyse erythrocytes, the pellet obtained USA), Gentamycin and Fungizone (Thermo Fisher Scien- after centrifugation was incubated in 157 mM NH4Cl for tific, MA, USA)] for 30 min in rotation at room temperature 5 min, washed in PBS and centrifuged again. Cells were (RT). Next, tissue was cut into small pieces and incubated then washed in washing buffer and finally pelleted cells twice with 10 mM dithiothreitol (DTT) for 5 min at RT. were embedded in Matrigel and seeded on prewarmed 48- Samples were then transferred to 8 mM EDTA solution for well culture dishes. After Matrigel solidification, ‘tumor’ 5 min at RT and 60 min in slow rotation at 4 °C. Samples culture medium was added (‘normal’ culture medium minus were washed in PBS until complete EDTA removal and Wnt3a-conditioned medium, Nicotinamide and RSPO1). 66 The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies. A. Fern andez-Barral et al. Vitamin D regulates human colon stem cells Growth and expansion of organoid cultures deoxycholate) containing protease and phosphatase inhibi- tors (10 lgmL1 leupeptin, 10 lgmL1 aprotinin, 1 mM Normal and tumor culture medium was changed every PMSF, 1 mM orthovanadate, and 1 mM NaF, all from other day. For passaging, we followed the protocol Sigma-Aldrich). Cell extracts were separated by SDS/ described [10,11], with modifications. Briefly, Matrigel-em- PAGE, transferred to PVDF membranes and incubated bedded organoids were collected using a scraper in a with the following primary antibodies: rabbit monoclonal- 15 mL conical tube and incubated with Cell Recovery VDR (D2K6W) (Cell Signaling, #12550), rabbit polyclonal- Solution (Corning) for 30 min on ice with continuous shak- CA2 (CUSABIO, #PA004370HA01HU, TX, USA), rabbit ing to remove Matrigel. After centrifugation at 260 g 5 min polyclonal-CYP24A1 (Santa Cruz Biotechnology, #sc- at 4 °C, organoids were washed in washing buffer and cen- 66851, Dallas, TX, USA) and goat polyclonal-b-actin trifuged again. Next, organoids were incubated with disag- (Santa Cruz Biotechnologies, #sc-1616). Different exposure gregation buffer [washing buffer solution containing times of the films were used to ensure that bands were not 1 mgmL1 dispase (Thermo Fisher Scientific)] for 10 min saturated. ImageJ was used for the semi-quantification of at RT in orbital rotation. Immediately afterwards, 2 mM VDR protein level. EDTA was added and the mixture was incubated for an additional 5 min. Following this, the solution was homoge- Electron microscopy nized by passing it through a 21G syringe, collected in a 15 mL conical tube, centrifuged at 250 g for 5 min at 4 °C For electron microscopy analysis, samples from three human and washed twice in washing buffer. Pelleted cells were normal (patients #11, #12, and #37; passages 2–11) and embedded in Matrigel and seeded on culture dishes. tumor organoid cultures (patients #4, #29, and #38; passages 10–21) treated with 100 nM calcitriol or vehicle for 96 h (n = 3 per group) were fixed with 3% glutaraldehyde (Mer- RNAscope in situ hybridization ck-Millipore, Burlington, MA, USA) in 0.12 M phosphate In situ hybridization for LGR5 on formalin-fixed paraffin- buffer, pH 7.4. Samples were then rinsed in 0.12 M phos- embedded samples (colon mucosa and normal/tumor orga- phate buffer, postfixed in 2% osmium tetroxide (Sigma- noids) was performed using the RNAscope 2.5 HD Assay- Aldrich), dehydrated in acetone and embedded in araldite RED (Advanced Cell Diagnostics; External Service, Dr. Ilse (Durcupan, Sigma-Aldrich, St. Louis, MO, USA). Ultrathin Rooman’s laboratory at Vrije Universiteit Brussel) according sections stained with uranyl acetate and lead citrate were to manufacturer’s instructions and with standard sample pre- examined with a JEOL 201 electron microscope. treatment (15 min Target Retrieval and 30 min Protease Plus). Subsequently, sections were stained for VDR using a Proliferation and clonogenicity assays rabbit monoclonal antibody (D2K6W #12550, Cell Signal- ing, Danvers, MA, USA) diluted 1:50 in Antibody Diluent Normal (passages 2–7) and tumor (passages 8–27) organoids (Agilent, Santa Clara, CA, USA) and applied for 20 min at were disaggregated with 1 mgmL1 dispase and around RT. For detection, a donkey anti-rabbit Alexa Fluor 488-la- 6000–7000 single cells were seeded on 20 lL Matrigel drops beled F(ab’)2 secondary antibody (Jackson ImmunoResearch in 48-well culture dishes. Organoids were incubated with nor- Europe, Cambridgeshire, UK) was used, diluted 1:500 in mal or tumor complete medium plus Y-27632 (Tocris) in the PBS and applied for 30 min at RT. Nuclei were counter- presence of 100 nM calcitriol or vehicle for 10 days. Medium stained with Hoechst 33342. Images were taken with Zeiss and calcitriol/vehicle were replaced every other day. At day LSM710 confocal microscope (Zeiss, Oberkochen, Ger- 10, normal and tumor organoids were used to study RNA many). VDR and LGR5 colors were changed using IMAGEJ expression and cell viability by estimating the amount of cel- software for a better visualization (VDR in red and LGR5 in lular ATP using the CellTiter-Glo Luminescent Cell Viability green). Images of at least 25 crypts/patient from four patients Assay (Promega, Madison, WI, USA) following the manufac- were taken in a confocal microscope. Quantification of VDR turer’s instructions. To analyze the clonogenicity, pictures of and LGR5 staining was performed manually by two people, drops were taken at days 0 and 10 of the cell viability assay considering the percentage of positive cells from a total of 46 and the number of organoids in each drop was counted using cells of the crypt bottom per crypt. MIPAR software [40]. Phase-contrast images of cultured orga- noids were captured with a DFC550 digital camera (Leica, Wetzlar, Germany) mounted on an inverted TS100 micro- Western blot scope (Nikon, Tokio, Japan). Normal (passages 1–5) and tumor (passages 3–8) organoids were pelleted after removing Matrigel with Cell Recovery Mutational status Solution and whole-cell extracts were prepared using RIPA buffer (0.05 M Tris/HCl pH 7.5, 0.1% SDS, 0.15 M NaCl, Mutations present in six tumor organoid cultures from 1% Triton X-100 (Sigma-Aldrich) and 1% sodium CRC patients (#1–6; passages 1–3) were analysed. Briefly, The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of 67 Federation of European Biochemical Societies. Vitamin D regulates human colon stem cells A. Fern andez-Barral et al. Matrigel was removed using ice-cold Cell Recovery follow- ATGCACAAAGCGGTTGTTGG-30 ) analysis, and their val- ing manufacturer’s instructions and DNA was extracted ues were normalized vs. the housekeeping gene Succinate Dehy- from pelleted organoids by incubation overnight (O/N) at drogenase Complex subunit A (SDHA) (forward 50 -TG 56 °C with lysis buffer [50 mM Tris/HCl pH 8.0, 100 mM GGAACAAGAGGGCATCTG-30 ; reverse 50 -CCACCACT EDTA pH 8, 100 mM NaCl, 1% SDS, and 20 mgmL1 GCATCAAATTCATG-30 ) using the comparative CT method. proteinase K (Merck-Millipore, MA, USA)]. Saturated All RT-qPCR were performed in a CFX384 Touch Real-Time NaCl buffer was added for 5 min and DNA was precipi- PCR Detection System (Bio-Rad). SW480-ADH cells were used tated with isopropanol and washed twice with 70% etha- as relative value to compare VDR and CYP24A1 expression in nol. DNA was genotyped by sequencing the amplified Figs 2A,B and 4D. To study the mRNA levels of other genes, product of a multiplexed-PCR reaction (Amplicon sequenc- organoids treated with vehicle were used as control. ing) using a proof-reading polymerase. Indexed libraries were pooled and loaded onto a MiSeq instrument (Illu- RNA-sequencing mina). Initial alignment was performed with BWA after primer sequence clipping and variant calling was done with Matched normal and tumor organoids from six patients (#1– the GATK Unified Genotyper and VarScan2 followed by 6; passages 1–3) were seeded in 48-well culture dishes and ANNOVAR annotation. Mutations were called at a mini- 48 h later were treated with 100 nM calcitriol (Sigma- mum 3% allele frequency. SNPs were filtered out with Aldrich) or vehicle for 96 h. Medium and calcitriol/vehicle dbSNP and 1000 genome datasets. All detected variants was replaced every other day. Total RNA was extracted and were manually revised and confirmed. RNA integrity was confirmed as optimal for all samples on an Agilent 2100 Bioanalyzer (RIN range 9.1–9.9). Six hun- dred nanograms of RNA samples was used for library con- Real-time quantitative PCR struction. PolyA+ fraction was purified and randomly Matrigel-embedded normal and tumor organoids (passages 1–6) fragmented, converted to double stranded cDNA and pro- were washed twice in PBS, lysed with TRIZOL (Thermo Fisher cessed through subsequent enzymatic treatments of end-re- Scientific) and total RNA was purified using the NucleoSpin pair, dA-tailing and ligation to adapters as in Illumina’s miRNA extraction kit (Machery-Nagel, D€ uren, Germany). For TruSeq Stranded mRNA Sample Preparation kit (Illumina, cDNA retrotranscription iScript cDNA Synthesis kit, (Bio-Rad, CA, USA). Adapter-ligated library was completed by PCR Hercules, CA, USA) was used. RT-qPCR analyses were per- with Illumina PE primers (eight cycles). The resulting puri- formed with the Taqman Universal PCR Master Mix (Applied fied cDNA library was applied to an Illumina flow cell for Biosystems, Waltham, CA, USA) using the following cluster generation and sequenced on an Ilumina HiSeq2000 FAM-labeled TaqMan probes (Applied Biosystems): by following manufacturer’s protocols. Eight samples were ALDH3A1 (Hs00964880_m1), AXIN2 (Hs00610344_m1), combined per sequencing lane and a minimum of 25 million BCAS1 (Hs00180227_m1), CA2 (Hs01070108_m1), CCND1 50 base single-reads were generated for each sample. (Hs00765553_m1), CEACAM7 (Hs00185152_m1), CYP Sequencing reads were aligned to the transcriptome with 24A1 (Hs00167999_m1), DKK1 (Hs00183740_m1), DKK4 TopHat2 [41]. Novel transcript discovery was not (Hs00205290_m1), GRK5 (Hs00992173), HMGCS2 attempted. TopHat was provided with known gene annota- (Hs00985427_m1), IFITM1 (Hs01652522_g1), IFITM2 (Hs0 tions and other transcript data obtained from Gencode 0829485_sH), JSRP1 (Hs00376079_m1), KLF4 (Hs0035883 (Version 26 – Ensembl 75) basic gene set for the GRCh37/ 6_m1), KRT20 (Hs00300643_m1), LGR5 (Hs00173664_m1), hg19 human genome assembly [42]. Gene expression level LRIG1 (Hs00394267_m1), MSI1 (Hs01045894_m1), MUC2 was calculated from TopHat alignments as the number of (Hs03005103_g1), MYC (Hs00153408_m1), PPP1R14D (Hs reads per gene computed using HTSeq [43] using default 00214613_m1), RARRES1 (Hs00161204_m1), S100P (Hs00 settings and gene features as defined in the GRCh37.75 195584_m1), SERPINE1 (Hs01126606_m1), SLC2A1 (Hs00 release of the human genome (gtf file). Differential gene 892681_m1), SMOC2 (Hs00405777_m1), SPDEF (Hs0102 expression analysis was performed with the Bioconductor 6050_m1), TFF2 (Hs00193719_m1), and TNS4 (Hs0026266 [44] DEseq2 package for the R statistical Software [45]. Sin- 2_m1). RNA expression values were normalized vs the VIC- gle enrichment analysis was performed with differentially labeled TaqMan probe for housekeeping gene large riboso- expressed genes between calcitriol vs. vehicle treatment. The mal protein (RPLPO) (H99999902_m1) using the compara- analysis was carried out with the functional annotation tool tive CT method. Additionally, we used Power SYBR Green included in the PathVisio bioinformatic resources [46,47]. PCR Master Mix (Applied Biosystems) and the following pri- mers to VDR (forward 50 AACGCTGTGTGGACAT Chromatin immunoprecipitation-sequencing CGGC-30 ; reverse 50 -GTCATGGCTTTCGTTGGACT-30 ), MEX3A (forward 50 -CAAGCTCTGCGCTCTCTACA-30 ; We performed three independent ChIP-seq experiments with reverse 50 -ATGAACACTGGTTCCTCGCC-30 ) and PTK7 human normal organoids (two culture organoids/experi- (forward 50 -TCTGGGAGACCTCAAGCAGT-30 ; reverse 50 - ment: ChIP #1, patients #34&35 at passages 12 and 6; ChIP 68 The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies. A. Fern andez-Barral et al. Vitamin D regulates human colon stem cells #2, patients #18&19 at passages 3 and 4; and ChIP #3, an Illumina flow cell for cluster generation and sequenced on patients #38&39 at passages 2 and 4) following the ChIP-seq an Illumina instrument by following manufacturer’s proto- Millipore protocol with modifications. Briefly, normal orga- cols. Sequencing reads were aligned to the GRCh37/hg19 gen- noids from two patients cultured on 6-well dishes were trea- ome assembly with Bowtie2 [48]. Enriched peak calling was ted with 100 nM calcitriol or vehicle for 2 h. DNA-protein performed with MACS2 [49] callpeak function, with –q-value crosslinking was done by incubation of Matrigel-embedded 0.15, –nomodel and –extsize 200 options using as control con- organoids with 1% formaldehyde (methanol free) for 8 min catenated same-patient alignment files from input treated shaking at 4 °C. Crosslinking was quenched with 1.25 mM (calcitriol) and untreated (vehicle). MACS2 bdgdiff function glycine in PBS for 5 min. We transferred the organoid solu- was used to obtain statistically enriched DNA regions using tion to a 50 mL conical tube for centrifugation at 700 g for the pileup data obtained from the peak calling. Peaks annota- 5 min at 4 °C. Supernatant was removed and the pellet was tion was performed using custom parameters in order to washed twice in cold PBS, transferred to a 1.5 mL tube and assign the peaks to the nearest gene in a range of 10 kb lysed in 1 mL ChIP-lysis buffer [1% SDS, 10 mM EDTA, upstream/downstream from the gene body. Manual inspec- 50 mM Tris/HCl pH 8.1 and 1x protease cocktail inhibitors tion of data and RNA-seq identified genes of interest was per- (Complete, EDTA-free, Sigma-Aldrich)] for 1 h. We used a formed using INTEGRATIVE GENOME VISUALIZATION (IGV) Tool 21G syringe to homogenize the solution and force the lysis. v.2.3.79 [50,51]. Samples were sonicated for 20 min at maximum intensity at 4 °C in a S220 Focused-ultrasonicator (Covaris). We took 40 lL of calcitriol and vehicle-samples as input control, and Functional enrichment analysis diluted 10-fold the remaining sonicated cell supernatant in We used Gene Set Enrichment Analysis (GSEA) [52] to ChIP-dilution buffer (0.01% SDS, 1.1% Triton X-100, assess the degree of association between our RNA-seq pro- 1.2 mM EDTA, 167 mM NaCl, Tris/HCl pH 8.1 and 1x pro- file (genes regulated by calcitriol in human colon normal/ tease cocktail inhibitor). After 45 min precleaning incubation tumor organoids), or microarray expression (genes regu- using magnetic beads (Magna ChIPTM Protein A+G Mag- lated by calcitriol in mouse colon normal organoids), with netic Beads, Merck-Millipore), the supernatant was incu- other signatures defined with genes enriched (differentiation bated with 10 lg rabbit monoclonal-VDR (D2K6W) gene set: EPHB2low vs. EPHB2high human normal colon antibody (Cell Signaling #12550, 2 mgmL1, made to order) epithelial cells [12]; proliferation gene sets: E2F, MYC and O/N in orbital rotation at 4 °C. Immunocomplexes were mTORC1 hallmark gene sets from MSigDB; pro-tumori- recovered using magnetic beads for 2 h in orbital rotation at genic gene set: Core ESC-like from MSigDB; slow-dividing 4 °C. Immunoprecipitated material was then washed in the mouse intestinal stem cells gen set: Mex3ahigh/Lgr5high following buffers in the indicated order: ChIP-dilution buf- mouse colon cells [22] and VDR ChIP-seq in human colon fer, ChIP-low salt buffer (0.1% SDS, 1% Triton X-100, normal organoids, Table S7). The over-representation of 2 mM EDTA, 150 mM NaCl, 20 mM Tris/HCl pH 8.0 and 1x transcription factor binding sites in genes regulated by cal- protease cocktail inhibitor), ChIP-high salt buffer (0.1% citriol was computed with the Bioconductor TFEA.ChIP SDS, 1% Triton X-100, 2 mM EDTA, 500 mM NaCl, 20 mM package. Tris/HCl pH 8 and 1x protease cocktail inhibitor), ChIP- LiCl buffer (0.25 M LiCl, 1% NP-40, 1% sodium deoxy- Statistical analysis cholate, 1 mM EDTA, 10 mM Tris/HCl pH 8.0 and 1x pro- tease cocktail inhibitor), ChIP-TE buffer (two washes) and Statistical analysis was performed using GraphPad Prism finally eluted in freshly prepared ChIP-elution buffer (1% (GraphPad, CA, USA). P-values < 0.05 were considered sig- SDS, 0.1 M NaHCO3). The eluted solution and input con- nificant and * indicates P < 0.05, **P < 0.01, and trols were incubated with 0.2 M NaCl at 65 °C ON in order ***P < 0.001. The exact P-values are indicated for nonsignifi- to de-crosslink the DNA complexes, and subsequently we cant results. Box plots shown in Figs 2A,B, 4A,C and 6A rep- added 9.5 mM EDTA, 36.5 mM Tris/HCl pH 6.5 and resent median  max/min. Graphs showing the 70 lgmL1 proteinase K for 1 h at 45 °C. DNA extraction quantification of VDR and LGR5 positive cells in human was performed using the QIAquick PCR purification kit crypts (Fig. 1B) represent mean  standard error of the (Qiagen, Venlo, The Netherlands). mean (SEM). Graphs showing the RNA levels in VDR inter- Fragmented DNA were processed through subsequent ference (Fig. S3), and relative VDR RNA levels of normal enzymatic treatments of end-repair, dA-tailing, and ligation and tumor organoids (Fig. 4D) represent mean  standard to adapters with NEBNext Ultra II DNA Library Prep Kit deviation (SD). for Illumina (New England BioLabs, Ipswich, MA, USA). Adapter-ligated libraries were completed by limited-cycle Acknowledgements PCR and extracted with a (single) double-sided SPRI size selection. Median fragment size is 340 bp from which 120 bp We thank our patients for their generosity. We are grate- correspond to adaptor sequences. Libraries were applied to ful to the Instituto de Salud Carlos III (ISCIII), RETICS The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of 69 Federation of European Biochemical Societies. Vitamin D regulates human colon stem cells A. Fern andez-Barral et al. Biobanks Network with the European Regional Devel- References opment Fund (FEDER); the Fundaci on Jimenez Dıaz 1 Barbachano A, Fernandez-Barral A, Ferrer-Mayorga Biobank PT17/0015/006 and La Paz University Hospi- G, Costales-Carrera A, Larriba MJ & Munoz A (2017) tal-IdiPAZ Biobank PT13/0010/0003. We also thank  The endocrine vitamin D system in the gut. Mol Cell Drs. E Gutierrez, E Alvarez, and L Asensio (La Paz Endocrinol 453, 79–87. University Hospital) for assisting with patient recruit- 2 Deeb KK, Trump DL & Johnson CS (2007) Vitamin D ment and consent, Prof. MB Demay for the donation of signalling pathways in cancer: potential for anticancer Vdr+/ mice, Dr. HG P almer for the mutational analysis, therapeutics. Nat Rev Cancer 7, 684–700. Dreamgenics for helping with the bioinformatics analy- 3 Feldman D, Krishnan AV, Swami S, Giovannucci E & ses, MT Berciano for her help with the electron micro- Feldman BJ (2014) The role of vitamin D in reducing scopy studies, MG Gonz alez-Bueno for her assistance cancer risk and progression. Nat Rev Cancer 14, 342–357. with animal maintenance and genotyping, Dr. Y Here- 4 Mondul AM, Weinstein SJ, Layne TM & Albanes D mans and Dr. I Rooman for RNAscope in situ (2017) Vitamin D and cancer risk and mortality: state of hybridization and L Banham for manuscript editing. the science, gaps, and challenges. Epidemiol Rev 39, 28–48. The studies were funded by Networks of Excellence from 5 Ferrer-Mayorga G, Larriba MJ, Crespo P & Munoz A Spanish Ministry of Science, Innovation and Universities (2019) Mechanisms of action of vitamin D in colon (MICINN) SAF2016-76377-R, ‘Nuclear Receptors in cancer. J Steroid Biochem Mol Biol 185, 1–6. Cancer, Metabolism and Inflammation’ (NuRCaMeIn) 6 Yamamoto H, Miyamoto K, Li B, Taketani Y, Kitano SAF2017-90604-REDT to A.M., ISCIII-Biomedical M, Inoue Y, Morita K, Pike JW & Takeda E (1999) Research Networking Centres-Oncology (CIBERONC) The caudal-related homeodomain protein Cdx-2 CB16/12/00273 to A.M. and A.B.; CB16/12/00453 to regulates vitamin D receptor gene expression in the F.X.R.; CB16/12/00342 to E.B. and CB16/12/00241 to small intestine. J Bone Miner Res 14, 240–247. F.R. ISCIII-Biomedical Research Networking Centres- 7 Modica S, Gofflot F, Murzilli S, D’Orazio A, Salvatore Respiratory Diseases (CIBERES) CB15/00037 to L.dP, L, Pellegrini F, Nicolucci A, Tognoni G, Copetti M, Valanzano R et al. (2010) The intestinal nuclear and ISCIII-FEDER PI15/00934 to F.R. receptor signature with epithelial localization patterns and expression modulation in tumors. Gastroenterology Conflict of interest 138, 636–48, 648 e1–12. 8 Ahearn TU, McCullough ML, Flanders WD, Long Q, The authors declare no conflict of interest. Sidelnikov E, Fedirko V, Daniel CR, Rutherford RE, Shaukat A & Bostick RM (2011) A randomized clinical trial of the effects of supplemental calcium and vitamin Data availability D3 on markers of their metabolism in normal mucosa Data from the RNA-seq analysis of patient-derived of colorectal adenoma patients. Cancer Res 71, 413–423. colon normal and tumor organoids, the VDR ChIP- 9 Barker N, van Es JH, Kuipers J, Kujala P, van den seq analysis of human colon normal organoids, and Born M, Cozijnsen M, Haegebarth A, Korving J, microarrays of mice colon normal organoids that sup- Begthel H, Peters PJ et al. (2007) Identification of stem port the findings of this study have been deposited in cells in small intestine and colon by marker gene Lgr5. the Gene Expression Omnibus (GEO) under the fol- Nature 449, 1003–1007. lowing accession codes: GSE100785 (RNA-seq), 10 Sato T, Stange DE, Ferrante M, Vries RG, Van Es JH, Van den Brink S, Van Houdt WJ, Pronk A, Van Gorp GSE107283 (ChIP-seq), and GSE105117 (microarray). J, Siersema PD et al. (2011) Long-term expansion of epithelial organoids from human colon, adenoma, adenocarcinoma, and Barrett’s epithelium. Author contributions Gastroenterology 141, 1762–1772. AM, EB, FXR, AF-B, and AB designed the research 11 Jung P, Sato T, Merlos-Suarez A, Barriga FM, Iglesias studies; AF-B, AB, and AC-C conducted most experi- M, Rossell D, Auer H, Gallardo M, Blasco MA, ments with the help of PB-M, PJ, GF-M and MJL; Sancho E et al. (2011) Isolation and in vitro expansion RC, LG-P, and DGO obtaining the surgical samples; of human colonic stem cells. Nat Med 17, 1225–1227. LdP, OD, AC-C, and AF-B performed the bioinfor- 12 Jung P, Sommer C, Barriga FM, Buczacki SJ, matics analyses of the data; ML conducted the elec- Hernando-Momblona X, Sevillano M, Duran-Frigola tron microscopy analysis, and FR and SPB conducted M, Aloy P, Selbach M, Winton DJ et al. (2015) the immunohistochemical analyses; AM, FXR, and Isolation of human colon stem cells using surface AB wrote the manuscript. expression of PTK7. Stem Cell Reports 5, 979–987. 70 The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies. A. Fern andez-Barral et al. Vitamin D regulates human colon stem cells 13 Sun H, Wang C, Hao M, Sun R, Wang Y, Liu T, 24 Pendas-Franco N, Garcia JM, Pena C, Valle N, Palmer Cong X & Liu Y (2016) CYP24A1 is a potential HG, Heinaniemi M, Carlberg C, Jimenez B, Bonilla F, biomarker for the progression and prognosis of human Munoz A et al. (2008) DICKKOPF-4 is induced by colorectal cancer. Hum Pathol 50, 101–108. TCF/beta-catenin and upregulated in human colon 14 van de Wetering M, Francies HE, Francis JM, cancer, promotes tumour cell invasion and angiogenesis Bounova G, Iorio F, Pronk A, van Houdt W, van and is repressed by 1alpha,25-dihydroxyvitamin D3. Gorp J, Taylor-Weiner A, Kester L et al. (2015) Oncogene 27, 4467–4477. Prospective derivation of a living organoid biobank of 25 Ebert MP, Tanzer M, Balluff B, Burgermeister E, colorectal cancer patients. Cell 161, 933–945. Kretzschmar AK, Hughes DJ, Tetzner R, Lofton-Day 15 Palmer HG, Gonzalez-Sancho JM, Espada J, Berciano C, Rosenberg R, Reinacher-Schick AC et al. (2012) MT, Puig I, Baulida J, Quintanilla M, Cano A, de TFAP2E-DKK4 and chemoresistance in colorectal Herreros AG, Lafarga M et al. (2001) Vitamin D(3) cancer. N Engl J Med 366, 44–53. promotes the differentiation of colon carcinoma cells by 26 Dimitrov V & White JH (2016) Species-specific the induction of E-cadherin and the inhibition of beta- regulation of innate immunity by vitamin D signaling. J catenin signaling. J Cell Biol 154, 369–387. Steroid Biochem Mol Biol 164, 246–253. 16 Bauer KM, Hummon AB & Buechler S (2012) Right- 27 Pike JW & Christakos S (2017) Biology and side and left-side colon cancer follow different pathways mechanisms of action of the vitamin D hormone. to relapse. Mol Carcinog 51, 411–421. Endocrinol Metab Clin North Am 46, 815–843. 17 Yamauchi M, Morikawa T, Kuchiba A, Imamura Y, 28 Weikum ER, Knuesel MT, Ortlund EA & Yamamoto Qian ZR, Nishihara R, Liao X, Waldron L, Hoshida KR (2017) Glucocorticoid receptor control of Y, Huttenhower C et al. (2012) Assessment of transcription: precision and plasticity via allostery. Nat colorectal cancer molecular features along bowel Rev Mol Cell Biol 18, 159–174. subsites challenges the conception of distinct 29 Mapes B, Chase M, Hong E, Ludvik A, Ceryes K, dichotomy of proximal versus distal colorectum. Gut Huang Y & Kupfer SS (2014) Ex vivo culture of 61, 847–854. primary human colonic tissue for studying 18 Cristobal A, van den Toorn HWP, van de Wetering M, transcriptional responses to 1alpha,25(OH)2 and 25 Clevers H, Heck AJR & Mohammed S (2017) Personalized (OH) vitamin D. Physiol Genomics 46, 302–308. proteome profiles of healthy and tumor human colon 30 Ferrer-Mayorga G, Gomez-Lopez G, Barbachano A, organoids reveal both individual diversity and basic features Fernandez-Barral A, Pena C, Pisano DG, Cantero R, of colorectal cancer. Cell Rep 18, 263–274. Rojo F, Munoz A & Larriba MJ (2017) Vitamin D 19 Peleg S & Nguyen CV (2010) The importance of receptor expression and associated gene signature in nuclear import in protection of the vitamin D receptor tumour stromal fibroblasts predict clinical outcome in from polyubiquitination and proteasome-mediated colorectal cancer. Gut 66, 1449–1462. degradation. J Cell Biochem 110, 926–934. 31 Limketkai BN, Mullin GE, Limsui D & Parian AM 20 Kongsbak M, von Essen MR, Boding L, Levring TB, (2017) Role of vitamin D in inflammatory bowel Schjerling P, Lauritsen JP, Woetmann A, Odum N, disease. Nutr Clin Pract 32, 337–345. Bonefeld CM & Geisler C (2014) Vitamin D up- 32 Barker N, Ridgway RA, van Es JH, van de Wetering M, regulates the vitamin D receptor by protecting it from Begthel H, van den Born M, Danenberg E, Clarke AR, proteasomal degradation in human CD4 + T cells. Sansom OJ & Clevers H (2009) Crypt stem cells as the PLoS One 9, e96695. cells-of-origin of intestinal cancer. Nature 457, 608–611. 21 Munoz J, Stange DE, Schepers AG, van de Wetering 33 Fumagalli A, Drost J, Suijkerbuijk SJ, van Boxtel R, de M, Koo BK, Itzkovitz S, Volckmann R, Kung KS, Ligt J, Offerhaus GJ, Begthel H, Beerling E, Tan EH, Koster J, Radulescu S et al. (2012) The Lgr5 intestinal Sansom OJ et al. (2017) Genetic dissection of colorectal stem cell signature: robust expression of proposed cancer progression by orthotopic transplantation of quiescent ‘+4’ cell markers. EMBO J 31, 3079–3091. engineered cancer organoids. Proc Natl Acad Sci USA 22 Barriga FM, Montagni E, Mana M, Mendez-Lago M, 114, E2357–E2364. Hernando-Momblona X, Sevillano M, Guillaumet- 34 Ordonez-Moran P & Huelsken J (2012) Lrig1: a new Adkins A, Rodriguez-Esteban G, Buczacki SJA, Gut M master regulator of epithelial stem cells. EMBO J 31, et al. (2017) Mex3a marks a slowly dividing 2064–2066. subpopulation of Lgr5+ intestinal stem cells. Cell Stem 35 Neirinckx V, Hedman H & Niclou SP (2017) Cell 20 (801–816), e7. Harnessing LRIG1-mediated inhibition of receptor 23 Wong DJ, Liu H, Ridky TW, Cassarino D, Segal E & tyrosine kinases for cancer therapy. Biochim Biophys Chang HY (2008) Module map of stem cell genes Acta 1868, 109–116. guides creation of epithelial cancer stem cells. Cell Stem 36 Engevik KA, Hanyu H, Matthis AL, Zhang T, Frey Cell 2, 333–344. MR, Oshima Y, Aihara E & Montrose MH (2019) The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of 71 Federation of European Biochemical Societies. Vitamin D regulates human colon stem cells A. Fern andez-Barral et al. Trefoil Factor 2 activation of CXCR36 requires calcium 50 Thorvaldsdottir H, Robinson JT & Mesirov JP (2013) mobilization to drive epithelial repair in gastric Integrative Genomics Viewer (IGV): high-performance organoids. J Physiol 597, 2673–2690. genomics data visualization and exploration. Brief 37 Wu CC, Tsai FM, Shyu RY, Tsai YM, Wang CH & Jiang Bioinform 14, 178–192. SY (2011) G protein-coupled receptor kinase 5 mediates 51 Robinson JT, Thorvaldsdottir H, Winckler W, Guttman Tazarotene-induced gene 1-induced growth suppression M, Lander ES, Getz G & Mesirov JP (2011) Integrative of human colon cancer cells. BMC Cancer 11, 175. genomics viewer. Nat Biotechnol 29, 24–26. 38 Peregrina K, Houston M, Daroqui C, Dhima E, Sellers 52 Subramanian A, Tamayo P, Mootha VK, Mukherjee S, RS & Augenlicht LH (2015) Vitamin D is a Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub determinant of mouse intestinal Lgr5 stem cell TR, Lander ES et al. (2005) Gene set enrichment analysis: functions. Carcinogenesis 36, 25–31. a knowledge-based approach for interpreting genome- 39 Fujii M & Sato T (2017) Defining the role of Lgr5(+) wide expression profiles. Proc Natl Acad Sci USA 102, stem cells in colorectal cancer: from basic research to 15545–15550. clinical applications. Genome Med 9, 66. 40 Sosa JM, Huber DE, Welk B & Fraser HL (2014) Supporting information Development and application of MIPARTM: a novel Additional supporting information may be found software package for two- and three-dimensional online in the Supporting Information section at the end microstructural characterization. Int Mat Manuf Innovation 3, 123–140. of the article. 41 Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R & Fig. S1. Human colon crypt bottom cells express Salzberg SL (2013) TopHat2: accurate alignment of PTK7 and VDR. transcriptomes in the presence of insertions, deletions Fig. S2. Gene expression profile of untreated human and gene fusions. Genome Biol 14, R36. normal and tumor orgnaoids. 42 Harrow J, Frankish A, Gonzalez JM, Tapanari E, Fig. S3. VDR silencing in human colon normal orga- Diekhans M, Kokocinski F, Aken BL, Barrell D, noids. Zadissa A, Searle S et al. (2012) GENCODE: the Fig. S4. Gene expression profile of mouse colon orga- reference human genome annotation for The ENCODE noids by calcitriol treatment. Project. Genome Res 22, 1760–1774. Fig. S5. Genomic distribution of VDR-binding sites 43 Anders S, Pyl PT & Huber W (2015) HTSeq–a Python and their association to calcitriol regulated genes. framework to work with high-throughput sequencing Table S1. Clinicopathological characteristics of CRC data. Bioinformatics 31, 166–169. patients from whom biopsies were obtained to gener- 44 Gentleman RC, Carey VJ, Bates DM, Bolstad B, ate normal and/or tumor organoid cultures. Dettling M, Dudoit S, Ellis B, Gautier L, Ge Y, Gentry Table S2. List of genes regulated by calcitriol J et al. (2004) Bioconductor: open software (q < 0.05) in matched normal and tumor organoids development for computational biology and identified by RNA-seq analyses. bioinformatics. Genome Biol 5, R80. Table S3. GO molecular pathways regulated by cal- 45 Love MI, Huber W & Anders S (2014) Moderated citriol in human normal and tumor organoids accord- estimation of fold change and dispersion for RNA-seq ing to significant target genes identified in the RNA- data with DESeq2. Genome Biol 15, 550. seq analyses. 46 Kutmon M, van Iersel MP, Bohler A, Kelder T, Nunes Table S4. RNA-seq data of most regulated genes by N, Pico AR & Evelo CT (2015) PathVisio 3: an calcitriol in human normal and tumor organoids. extendable pathway analysis toolbox. PLoS Comput Biol 11, e1004085. Table S5. RNA-seq data of calcitriol regulation of 47 van Iersel MP, Kelder T, Pico AR, Hanspers K, Coort selected genes in human normal and tumor organoids. S, Conklin BR & Evelo C (2008) Presenting and Table S6. Humanized genes regulated by calcitriol only exploring biological pathways with PathVisio. BMC in colon organoids from wild-type mice (Microarray, Bioinformatics 9, 399. q-value < 0.05). 48 Langmead B & Salzberg SL (2012) Fast gapped-read Table S7. Assigned genes with VDR-binding sites coin- alignment with Bowtie 2. Nat Methods 9, 357–359. cident in three ChIP-seq assays performed in human 49 Zhang Y, Liu T, Meyer CA, Eeckhoute J, Johnson DS, normal organoid cultures (MACS2 significantly Bernstein BE, Nusbaum C, Myers RM, Brown M, Li enriched with calcitriol treatment). W et al. (2008) Model-based analysis of ChIP-Seq Appendix S1. Supplementary methods. (MACS). Genome Biol 9, R137. 72 The FEBS Journal 287 (2020) 53–72 ª 2019 The Authors. The FEBS Journal published by John Wiley & Sons Ltd on behalf of Federation of European Biochemical Societies.