Inhibition of Mouse Breast Tumor-Initiating Cells by Calcitriol and Dietary Vitamin D

Abstract

The anticancer actions of vitamin D and its hormonally active form, calcitriol, have been extensively documented in clinical and preclinical studies. However, the mechanisms underlying these actions have not been completely elucidated. Here, we examined the effect of dietary vitamin D and calcitriol on mouse breast tumor-initiating cells (TICs, also known as cancer stem cells). We focused on MMTV-Wnt1 mammary tumors, for which markers for isolating TICs have previously been validated. We confirmed that these tumors expressed functional vitamin D receptors and estrogen receptors (ER) and exhibited calcitriol-induced molecular responses including ER downregulation. Following orthotopic implantation of MMTV-Wnt1 mammary tumor cells into mice, calcitriol injections or a vitamin D-supplemented diet caused a striking delay in tumor appearance and growth, whereas a vitamin D-deficient diet accelerated tumor appearance and growth. Calcitriol inhibited TIC tumor spheroid formation in a dose-dependent manner in primary cultures and inhibited TIC self-renewal in secondary passages. A combination of calcitriol and ionizing radiation inhibited spheroid formation more than either treatment alone. Further, calcitriol significantly decreased TIC frequency as evaluated by in vivo limiting dilution analyses. Calcitriol inhibition of TIC spheroid formation could be overcome by the overexpression of β-catenin, suggesting that the inhibition of Wnt/β-catenin pathway is an important mechanism mediating the TIC inhibitory activity of calcitriol in this tumor model. Our findings indicate that vitamin D compounds target breast TICs reducing tumor-initiating activity. Our data also suggest that combining vitamin D compounds with standard therapies may enhance anticancer activity and improve therapeutic outcomes.

Fig. 1. Presence and functional activity of ER and VDR in tissue slice cultures of MMTV-Wnt1 tumors

MMTV-Wnt1 tumor orthografts were used to generate tissue slices. The slices were cultured and the expression of ER and VDR and their functional responses were determined as described in Methods Section. A, Basal ER and VDR levels. [³H]-labeled ligand binding assays revealed the expression of ER and VDR proteins in the tumor slices. B–F, VDR functional responses. Tissue slice cultures were treated with 0.1% ethanol vehicle or 100 nM calcitriol (Cal) for 5 h and the mRNA levels of the Cyp24, Cyp27B1, Vdr, Erα and Cyp19 were determined by qRT-PCR (n = 4; * p<0.05 and ** p<0.01 as compared to the Std group). G, ER functional response. Tissue slice cultures in phenol red-free culture media were exposed to 0.2% ethanol vehicle (control, Con), 100 nM calcitriol (Cal), 10 nM E₂ (E2) or a combination of both (Cal+E2) for 5 h and PR mRNA levels were determined by qRT-PCR (n = 4; * p<0.05 as compared to Con and + p < 0.05 as compared to E2). At least three different orthografts were used to generate the tissue slices and each experiment was conducted in duplicate. Values represent mean ± SEM.

Fig. 2. Effect of calcitriol and dietary vitamin D₃ levels on the appearance and growth of MMTV-Wnt1 tumors

FVB/N mice were fed the standard diet (Std), a vitamin D-deficient diet (Lo-D) or a vitamin D-supplemented diet (Hi-D) for 12 weeks (Experimental weeks 0–12). A parallel experimental group consisted of mice on the standard diet receiving calcitriol injections (Cal) in the last seven weeks of the 12-week period (Experimental weeks 6–12). On Experimental week 12 MMTV-Wnt1 tumor cell suspensions were implanted in the left inguinal mammary fat pads of the mice. Diets and calcitriol treatments were continued for the next 7 weeks (Experimental weeks 12–19) and tumor appearance and growth were monitored. A, Tumor appearance. Kaplan-Meier analysis of the occurrence of palpable tumors in mice in the various experimental groups: Std (red line), LO-D (blue line), Cal (green line) and Hi-D (magenta line). B, End-point tumor volumes. Tumor volumes in the various experimental groups at the end of the study are shown (The experiment was performed twice with half the number of mice each time. Total number of mice in each group was as follows: Std diet =10; Cal=6; Hi-D=6; and Lo-D=6.; * p<0.05 and ** p < 0.001 as compared to the Std group). C, Changes in mRNA expression of Vdr, CYP27B1, Erα, p21, Cox-2 and aromatase (Cyp19) in MMTV-Wnt1 tumors due to calcitriol and dietary vitamin D status. Relative mRNA expression of each gene is shown with the expression in tumors from mice on the standard diet (Std) set at 1. (n = 6–10 determinations; * p < 0.05 as compared to Std). Values represent mean ± SEM.

Fig. 3. Breast cancer tumor-initiating cells express functionally active vitamin D receptor

A, FACS sorting schema for purification of CD49f^hiEpCam^low MMTV-Wnt-1 TIC-enriched cells. B–D, VDR functional responses in breast cancer TICs. Sorted CD49f^hiEpCam^low TIC cells were treated with 1 nM calcitriol (Cal) for 16 h and mRNA levels of Vdr, Cyp24, and Cyp27b1 were determined by qRT-PCR (n = 3; *** p<0.01 as compared to the CON group; ND, not detected). Values represent mean ± SEM.

Fig. 4. Calcitriol inhibits the self-renewal of breast cancer tumor-initiating cells

A, Representative images and bar graphs showing that calcitriol treatment (vehicle or 0.1, 1, 10, and 100 nM) decreased the spheroid forming abilities of TICs in a dose dependent manner (n = 3; ** P < 0.01 and *** P < 0.001 as compared to the control). B, Calcitriol treatment (at 1 nM) in primary culture reduced the spheroid formation in secondary culture (n = 3; *** P < 0.001 as compared to white bar and + P < 0.001 as compared to dotted bar). C, Analysis of combined treatment of TICs with calcitriol (10 nM) and 2 Gy of ionizing radiation (n = 3; P < 0.05). Values represent mean ± SEM.

Fig. 5. Limiting dilution analysis of calcitriol-treated TICs

A, Kaplan-Meier plot of tumor appearance. Time taken for gross tumor appearance was traced after orthotopic implantation of 50 and 100 control- (ethanol) or calcitriol- treated cells.

Fig. 6. Effect of calcitriol on Wnt signaling and downstream target genes

A, Changes in the expression of Wnt target genes in MMTV-Wnt1 mammary tumors harvested from mice on the standard diet (Std) and those receiving calcitriol (Cal). Relative mRNA expression of each gene is shown with the expression in tumors from mice on the standard diet (Std) normalized to 1 (n = 6–10; *, P < 0.05). B, Changes in the expression of Wnt target genes in CD49f^hiEpCAM^low TIC-enriched cells treated with ethanol or calcitriol (1 nM) for 16 hrs. Relative mRNA expression of each gene is shown (n = 3; ***, P < 0.05). C, Effect of constitutively active β-catenin on calcitriol sensitivity of TICs. Spheroid formation by CD49f^hiEpCam^low cells was assayed after treatment with vehicle (ethanol) or calcitriol (1 nM) following transduction with control or constitutively active β-catenin expressing lentiviruses (n = 5–6; P < 0.05). Values represent mean ± SEM.