Vitamin D receptor expression and associated gene signature in tumour stromal fibroblasts predict clinical outcome in colorectal cancer

Abstract

Objective: Colorectal cancer (CRC) is a major health concern. Vitamin D deficiency is associated with high CRC incidence and mortality, suggesting a protective effect of vitamin D against this disease. Given the strong influence of tumour stroma on cancer progression, we investigated the potential effects of the active vitamin D metabolite 1α,25-dihydroxyvitamin D₃ (1,25(OH)₂D₃) on CRC stroma.

Design: Expression of vitamin D receptor (VDR) and two 1,25(OH)₂D₃ target genes was analysed in 658 patients with CRC with prolonged clinical follow-up. 1,25(OH)₂D₃ effects on primary cultures of patient-derived colon normal fibroblasts (NFs) and cancer-associated fibroblasts (CAFs) were studied using collagen gel contraction and migration assays and global gene expression analyses. Publicly available data sets (n=877) were used to correlate the 1,25(OH)₂D₃-associated gene signature in CAFs with CRC outcome.

Results: High VDR expression in tumour stromal fibroblasts was associated with better overall survival (OS) and progression-free survival in CRC, independently of its expression in carcinoma cells. 1,25(OH)₂D₃ inhibited the protumoural activation of NFs and CAFs and imposed in CAFs a 1,25(OH)₂D₃-associated gene signature that correlated with longer OS and disease-free survival in CRC. Furthermore, expression of two genes from the signature, CD82 and S100A4, correlated with stromal VDR expression and clinical outcome in our cohort of patients with CRC.

Conclusions: 1,25(OH)₂D₃ has protective effects against CRC through the regulation of stromal fibroblasts. Accordingly, expression of VDR and 1,25(OH)₂D₃-associated gene signature in stromal fibroblasts predicts a favourable clinical outcome in CRC. Therefore, treatment of patients with CRC with VDR agonists could be explored even in the absence of VDR expression in carcinoma cells.

Keywords: COLORECTAL CANCER; MYOFIBROBLASTS; VITAMIN D RECEPTOR GENE.

Figure 1

Vitamin D receptor (VDR) expression in stromal fibroblasts of colorectal tumours predicts clinical outcome. (A) Representative immunohistochemical images of VDR protein expression in carcinoma and stromal compartments of colorectal tumours. Bars, 20 μm. (B and C) Kaplan–Meier survival curves depicting overall survival (OS) or progression-free survival (PFS) of patients with colorectal cancer (CRC) (n=658) stratified by VDR protein expression levels in carcinoma cells (B) or in tumour stromal fibroblasts (C).

Figure 2

Establishment of human colon normal fibroblast (NF) and cancer-associated fibroblast (CAF) primary cultures. (A) Representative phase-contrast images showing the outgrowth of fibroblasts from fresh colon biopsies obtained by surgery. Bar, 100 μm. (B) Representative phase-contrast and immunofluorescence images showing NF and CAF phenotypes and expression of fibroblast (vimentin and α-SMA) or epithelial (cytokeratin-18 and E-cadherin) markers. HT29 human colon carcinoma cells have an epithelial origin and were used as a control. Bars, 60 μm.

Figure 3

Patient-derived colon normal fibroblast (NF) and cancer-associated fibroblast (CAF) primary cultures express vitamin D receptor (VDR) and respond to 1α,25-dihydroxyvitamin D₃ (1,25(OH)₂D₃). (A) VDR RNA expression measured by RT-qPCR in 32 paired human colon NF and CAF primary cultures and calculated in relation to that of CCD18Co human normal colon fibroblasts. Horizontal bars indicate the median values. (B) RT-qPCR analysis of VDR expression in the same primary cultures as in (A) treated for 48 h with 100 nM 1,25(OH)₂D₃ or vehicle. Mean±SEM of the fold-change after 1,25(OH)₂D₃ treatment is depicted. (C) CYP24A1 RNA expression measured by RT-qPCR in the same primary cultures as in (A) treated for 48 h with 100 nM 1,25(OH)₂D₃ or vehicle. Data are shown as log₁₀ and horizontal bars indicate the median values. (D) Correlation between VDR RNA expression in vehicle-treated and CYP24A1 RNA expression in 1,25(OH)₂D₃-treated (48 h) human colon NFs and CAFs (n=64). Data are shown as log₁₀.

Figure 4

1α,25-Dihydroxyvitamin D₃ (1,25(OH)₂D₃) represses the protumoural phenotype of colorectal cancer (CRC) patient-derived stromal fibroblasts. (A) Representative collagen gel contraction assay. Normal fibroblasts (NFs) from patient 11 were embedded in collagen gels in the presence of 100 nM 1,25(OH)₂D₃ or vehicle and gel area was measured at the indicated times. Representative stereomicroscope images of collagen gels are shown. Bar, 400 μm. The experiment was performed in triplicate and mean±SD is depicted. (B) Collagen gel contraction assay of 10 paired NF (N) and cancer-associated fibroblast (CAF) (C) primary cultures in the presence of 100 nM 1,25(OH)₂D₃ or vehicle. 1,25(OH)₂D₃ effect (fold-change vs vehicle) in the gel area after 72 h is depicted. The results from each patient (left) and the mean±SEM of all patients (right) are shown. (C) Representative SW480-ADH cell migration assay. Cell migration was assessed after 24 h of Transwell-mediated coculture of SW480-ADH human colon carcinoma cells with NFs from patient 27 pretreated with 100 nM 1,25(OH)₂D₃ or vehicle for 48 h. Representative images of migrating cells and a scheme of the experiment are shown. Bar, 100 μm. The experiment was performed in triplicate and mean±SD is depicted. Transwells without fibroblasts were used as a control. (D) Promigratory action exerted by 11 paired NF (N) and CAF (C) primary cultures pretreated with 100 nM 1,25(OH)₂D₃ or vehicle for 48 h on SW480-ADH cells. The effect of 1,25(OH)₂D₃ pretreatment (percentage of the promigratory action exerted by vehicle-pretreated fibroblasts) is shown. The results from each patient (left) and the mean±SEM of all patients (right) are depicted. (E) Correlation between VDR RNA expression in vehicle-treated NF and CAF primary cultures and their promigratory action on SW480-ADH cells after 1,25(OH)₂D₃ pretreatment (n=22). Data are shown as log₁₀.

Figure 5

1α,25-Dihydroxyvitamin D₃ (1,25(OH)₂D₃) regulates the gene expression profile of human colon normal fibroblasts (NFs) and cancer-associated fibroblasts (CAFs). (A) Heat maps showing microarray results for the 100 genes most differentially regulated (false discovery rate (FDR)<0.05 and the highest fold-changes) by 100 nM 1,25(OH)₂D₃ treatment (48 h) in seven paired NF and CAF primary cultures. See online supplementary tables S6 and S7 for the complete list of differentially regulated genes (FDR<0.05). (B) Validation by RT-qPCR of 10 1,25(OH)₂D₃ target genes identified in the microarray study in an independent series of seven paired NFs and CAFs (patients 9, 17, 20, 24, 25, 34 and 35) treated with 100 nM 1,25(OH)₂D₃ or vehicle for 48 h. Mean±SEM of the fold-change after 1,25(OH)₂D₃ treatment is depicted. (C) Venn diagram showing the overlap between 1,25(OH)₂D₃-regulated genes in NFs and CAFs. The number of genes included in each group is depicted and the complete list of genes can be found in online supplementary table S8.

Figure 6

1α,25-Dihydroxyvitamin D₃ (1,25(OH)₂D₃) imposes a gene signature in human colon cancer-associated fibroblasts (CAFs) that is associated with longer survival of patients with colorectal cancer (CRC). (A) Kaplan–Meier survival curves showing the impact of the 1,25(OH)₂D₃-associated gene signature on the disease-free survival (DFS) of patients with CRC from three unrelated publicly available external data sets (GSE33113, n=89, American Joint Committee on Cancer (AJCC) stage II; GSE14333, n=226, AJCC stage I–II–III; GSE39582, n=497, AJCC stage I–II–III). (B) Kaplan–Meier survival curves showing the impact of the 1,25(OH)₂D₃-associated gene signature on the overall survival (OS) of patients with CRC from GSE39582 data set (n=562, AJCC stage I–II–III–IV). (C) Representative immunohistochemical images of CD82 and S100A4 protein expression in human CRC samples. Bars, 20 μm. (D) Box plots showing CD82 or S100A4 protein expression in stromal fibroblasts located in colorectal tumours with low or high vitamin D receptor (VDR) protein expression levels in stromal fibroblasts (n=658). (E) Kaplan–Meier survival curves depicting OS of patients with CRC (n=658) stratified by CD82 or S100A4 protein expression levels in tumour stromal fibroblasts.

Figure 7

Human and mouse fibroblast cell lines express vitamin D receptor (VDR) and respond to 1α,25-dihydroxyvitamin D₃ (1,25(OH)₂D₃). (A) Western blot analysis of VDR protein levels in IMR90, BJ-hTERT and NIH3T3 fibroblasts treated with 100 nM 1,25(OH)₂D₃ or vehicle for the indicated times. β-Actin was used as loading control. Images of a representative experiment and the quantification of three independent experiments (mean±SEM) are shown. (B–E) RT-qPCR analysis of VDR (B), CYP24A1 (C), OPN (D) and S100A4 (E) RNA levels in IMR90, BJ-hTERT and NIH3T3 fibroblasts treated with 100 nM 1,25(OH)₂D₃ or vehicle for the indicated times. Mean±SEM of three independent experiments is shown.

Figure 8

1α,25-Dihydroxyvitamin D₃ (1,25(OH)₂D₃) inhibits the protumoural properties of several fibroblast cell lines of different origin. (A) Proliferation of IMR90 and NIH3T3 cells treated with 100 nM 1,25(OH)₂D₃ or vehicle for the indicated times was estimated by 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assays. 1,25(OH)₂D₃ effect in cell proliferation is shown as a percentage versus vehicle-treated cells. Mean±SEM of three independent experiments is depicted. (B) Migratory capacity of IMR90, BJ-hTERT and NIH3T3 fibroblasts treated with 100 nM 1,25(OH)₂D₃ or vehicle. Representative images of migrating cells and the quantification (mean±SEM) of three independent experiments are shown. Bar, 100 μm. (C) Invasive capacity of IMR90 and NIH3T3 fibroblasts treated with 100 nM 1,25(OH)₂D₃ or vehicle. Representative images of invading cells and the quantification (mean±SEM) of three independent experiments are shown. Bar, 50 μm. (D) Collagen gel contraction assay. IMR90, BJ-hTERT and NIH3T3 fibroblasts were embedded in collagen gels in the presence of 100 nM 1,25(OH)₂D₃ or vehicle and gel area was measured at the indicated times. Representative stereomicroscope images of collagen gels and the quantification (mean±SEM) of three independent experiments are shown. Bar, 400 μm.