Activation of pro-survival metabolic networks by 1,25(OH)2D3 does not hamper the sensitivity of breast cancer cells to chemotherapeutics

Abstract

Background: We have previously identified 1,25-dihydroxyvitamin D₃ [1,25(OH)₂D₃], the bioactive form of vitamin D₃, as a potent regulator of energy-utilization and nutrient-sensing pathways in prostate cancer cells. In the current study, we investigated the effects of 1,25(OH)₂D₃ on breast cancer (BCa) cell metabolism using cell lines representing distinct molecular subtypes, luminal (MCF-7 and T-47D), and triple-negative BCa (MDA-MB-231, MDA-MB-468, and HCC-1143).

Methods: 1,25(OH)₂D₃'s effect on BCa cell metabolism was evaluated by employing a combination of real-time measurements of glycolysis/oxygen consumption rates using a biosensor chip system, GC/MS-based metabolomics, gene expression analysis, and assessment of overall energy levels. The influence of treatment on energy-related signaling molecules was investigated by immunoblotting.

Results: We show that 1,25(OH)₂D₃ significantly induces the expression and activity of the pentose phosphate pathway enzyme glucose-6-phosphate dehydrogenase (G6PD) in all BCa cell lines, however differentially influences glycolytic and respiratory rates in the same cells. Although 1,25(OH)₂D₃ treatment was found to induce seemingly anti-oxidant responses in MCF-7 cells, such as increased intracellular serine levels, and reduce the expression of its putative target gene thioredoxin-interacting protein (TXNIP), intracellular reactive oxygen species levels were found to be elevated. Serine accumulation in 1,25(OH)₂D₃-treated cells was not found to hamper the efficacy of chemotherapeutics, including 5-fluorouracil. Detailed analyses of the nature of TXNIP's regulation by 1,25(OH)₂D₃ included genetic and pharmacological inhibition of signaling molecules and metabolic enzymes including AMP-activated protein kinase and G6PD, as well as by studying the ITCH (E3 ubiquitin ligase)-TXNIP interaction. While these investigations demonstrated minimal involvement of such pathways in the observed non-canonical regulation of TXNIP, inhibition of estrogen receptor (ER) signaling by tamoxifen mirrored the reduction of TXNIP levels by 1,25(OH)₂D₃, demonstrating that the latter's negative regulation of ER expression is a potential mechanism of TXNIP modulation.

Conclusions: Altogether, we propose that regulation of energy metabolism contributes to 1,25(OH)₂D₃'s anti-cancer effects and that combining 1,25(OH)₂D₃ with drugs targeting metabolic networks in tumor cells may lead to synergistic effects.

Keywords: AMPK; Breast cancer; Estrogen receptor; G6PD; ITCH; Metabolism; Serine; TXNIP; Vitamin D.

Fig. 1

Analysis of 1,25(OH)₂D₃’s metabolism-regulating effects in BCa cells. a Extracellular acidification, respiration, and impedance rates were monitored in real-time in response to 1,25(OH)₂D₃ (100 nM) over the course of 3 days, followed by a 20-h recovery period, in which cells were exposed to running medium (RM) not containing the molecule. 1,25(OH)₂D₃ clearly induces glycolytic rate in luminal (MCF-7 and T-47D) BCa cells and reduces respiration rate to varying degrees in TNBC (MDA-MB-231, MDA-MB-468, and HCC-1143) cells. Values of each cell line are normalized to measurements obtained from the respective DMSO-treated cells. Data presented are representative of 2 biological replicates. b GC/MS analysis of TCA cycle intermediates and amino acids in select BCa cell lines treated for 72 h with 1,25(OH)₂D₃ (n = 4). A strong accumulation in serine is observed with treatment in MCF-7 cells, whereas levels of citrate were differentially regulated by 1,25(OH)₂D₃ in both cell lines. PEP phosphoenolpyruvate, α-KG alpha-ketoglutarate. c mRNA expression analysis of metabolism-related genes in MCF-7 and MDA-MB-231 cells in response to a 72 h treatment with 1,25(OH)₂D₃. G6PD was found to be induced by 1,25(OH)₂D₃ in both cell lines. Relative expression was calculated using the ∆∆Ct method, with vinculin as the housekeeping gene. Data presented are the average of 2 biological replicates. G6PD protein expression is significantly induced in MCF-7 (d) and MDA-MB-231 (e) cells by 1,25(OH)₂D₃ (72 h), as demonstrated by western blot and densitometric analysis. Statistical comparison between DMSO- and 1,25(OH)₂D₃-treated cells was made using a two-tailed Student’s t test. P values less than or equal to 0.05, 0.01 and 0.001 are depicted by *, **, and ***, respectively. Error bars ± SD; n = 3. G6PD mRNA (f), protein (g), and activity (h) levels are induced by 1,25(OH)₂D₃ (72 h) in all BCa cell lines. i Intracellular ATP levels were assessed in BCa cells in response to 1, 2, and 3 days of treatment with 1,25(OH)₂D₃. Treatment significantly induced ATP levels in MDA-MB-231 cells in 48 and 72 h, however reduced ATP levels in MCF-7 at the latest time point

Fig. 2

Seemingly pro-survival metabolic effects of 1,25(OH)₂D₃ do not hamper the efficacy of metabolism-targeting therapeutic regimens. a Kaplan-Meier plots demonstrating the significant inverse correlation between G6PD mRNA expression and overall survival in ER+ (left) and total (right) BCa patients. b Knocking-down G6PD in MCF-7 cells significantly reduces cell survival, independent of 1,25(OH)₂D₃ (100 nM; 72 h). Statistical comparison between DMSO- and 1,25(OH)₂D₃-treated cells was made using a two-tailed Student’s t test. P values less than or equal to 0.05, 0.01 and 0.001 are depicted by *, **, and ***, respectively. Error bars ± SD; n = 3. Cellular survival was assessed using SRB assay, and % survival is calculated by normalizing the absorbance value obtained with the different experimental conditions to DMSO-treated cells transfected with negative control (NC) siRNA. 1,25(OH)₂D₃ significantly enhances the anti-tumor effects of DHEA assessed by SRB assay. MCF-7 cells were treated for 72 h with increasing concentrations of DHEA in the presence of either DMSO or 1,25(OH)₂D₃, and % survival is calculated by normalizing the absorbance value obtained with the different experimental conditions to DMSO-treated cells. c Schematic overview of cellular serine uptake, synthesis, and metabolism. 3PS 3-phosphoserine. d 72 h treatment of MCF-7 cells with 1,25(OH)₂D₃ differentially influences the mRNA expression of enzymes involved in serine synthesis but upregulates the expression of those involved in serine metabolism. Relative expression was calculated using the ∆∆Ct method, with vinculin as the housekeeping gene. e 1,25(OH)₂D₃ significantly influences cellular survival in response to amino acid deprivation. MCF-7 cells were cultivated in medium lacking serine, glycine, or both, in the presence of either DMSO or 1,25(OH)₂D₃ for 72 h, after which SRB assay was performed. 1,25(OH)₂D₃ was found to mildly however significantly dampen the reduction in survival of MCF-7 cells in response to serine deprivation. 1,25(OH)₂D₃ was also found to enhance the efficacy of CBR-5884 and 5-FU, inhibitors of PHGDH and TYMS, respectively, assessed by SRB assay

Fig. 3

Regulation of AMPK signaling and TXNIP expression in BCa cells by 1,25(OH)₂D₃. a The AMPK-TXNIP signaling axis was found to be differentially regulated by a 72-h treatment with 1,25(OH)₂D₃ (100 nM). AMPK signaling was found to be induced and TXNIP expression reduced in MCF-7 cells in response to treatment. Western blot and densitometric analysis of TXNIP expression in MCF-7 cells treated for 24 (b) and 72 h (c) with 1,25(OH)₂D₃. Statistical significance between DMSO- and 1,25(OH)₂D₃-treated cells is calculated using a two-tailed Student’s t test, where p values less than or equal to 0.05, 0.01, and 0.001 are depicted in the figures by *, **, and ***, respectively. Error bars ± SD; n = 3. d–g 72 h treatment with 1,25(OH)₂D₃ significantly induces intracellular ROS levels. Microscopic analysis of cellular superoxide using DHE staining (d) and associated quantification (e). Hoechst dye was used to stain nuclei. Mitochondrial superoxide was stained using MitoSOX Red, and mitochondria were stained using MitoTracker Green (f). Microscopic images were taken every 15 min for 5 h, and fluorescence intensity of MitoSOX Red was quantified (g)

Fig. 4

TXNIP regulation by 1,25(OH)₂D₃ is independent of metabolic reprogramming. a Schematic overview of treatment-induced metabolic changes that may reduce levels of glycolytic intermediates capable of inducing the nuclear translocation of MondoA/MLX and hence TXNIP expression. Genetic and/or pharmacological strategies for inhibiting key metabolic players are depicted. b Time-dependent increase in G6PD mRNA expression in MCF-7 cells with 1,25(OH)₂D₃ (100 nM) is associated with an attenuation of the initial increase in TXNIP mRNA levels with treatment. Upper and lower ends of floating bars represent maximum and minimum values, respectively, whereas lines in the middle represent the means. The ∆∆Ct method was used to calculate the relative expression with vinculin as the housekeeping gene. The data are obtained from two biological replicates. c Similar to the trend observed on the mRNA level, the induction in G6PD protein level in response to 1,25(OH)₂D₃ first occurs after 12 h, which coincides with the first reduction in TXNIP protein expression. d Co-treatment with DHEA, but not knocking-down G6PD levels, elevates TXNIP mRNA levels in 1,25(OH)₂D₃-treated MCF-7 cells. Cells transfected with either anti-G6PD siRNA or a NC were treated with 1,25(OH)₂D₃ for 72 h, whereas increasing concentrations of DHEA were added to the conditioned medium of MCF-7 cells treated for 48 h with either DMSO or 1,25(OH)₂D₃, for an additional 24 h. e G6PD inhibition using siRNA or DHEA reduces TXNIP protein expression independent of 1,25(OH)₂D₃. f–h Western blot analysis of TXNIP expression in MCF-7 cells treated with a combination of either DMSO or 1,25(OH)₂D₃ and various metabolic inhibitors. Cells were initially treated with DMSO or 1,25(OH)₂D₃ for 48 h, after which increasing concentrations of the different inhibitors were added to the conditioned medium for an additional 24 h. All inhibitors were found to reduce TXNIP protein levels in 1,25(OH)₂D₃-treated cells. In the absence of 1,25(OH)₂D₃, AZD 3965 and CBR-5884 were found to mildly induce TXNIP levels

Fig. 5

1,25(OH)₂D₃ induces proteasomal degradation of TXNIP in MCF-7 cells. a The reduction in TXNIP protein levels by 1,25(OH)₂D₃ (100 nM) in MCF-7 cells is rescued by MG-132 (5 μM) or 2-deoxyglucose (10 mM), but not leupeptin (20 μM). The various molecules were added to the conditioned medium of DMSO- and 1,25(OH)₂D₃-treated (66 h) MCF-7 cells, for an additional 6 h. b, c ITCH mRNA and protein expression is not markedly influenced by 1,25(OH)₂D₃. Relative expression was calculated using the ∆∆Ct method with vinculin as the housekeeping gene. Error bars ± SD; n > 3. d Overall protein ubiquitination in MCF-7 cells was not changed by 1,25(OH)₂D₃ treatment. e Co-immunoprecipitation studies illustrate that the TXNIP-ITCH interaction is not altered by 1,25(OH)₂D₃ treatment of MCF-7 cells. f Negative regulation of TXNIP protein expression by 1,25(OH)₂D₃ is observed in MCF-7 cells with knocked-down AMPKα1 levels. g The non-calcemic 1,25(OH)₂D₃ analogue, calcipotriol (100 nM; 72 h) induces similar effects on TXNIP expression as 1,25(OH)₂D₃. The cell permeable Ca²⁺ chelator BAPTA-AM (20 μM) does not hamper 1,25(OH)₂D₃’s effects on TXNIP expression. BAPTA-AM was added to the conditioned medium of DMSO- and 1,25(OH)₂D₃-treated MCF-7 cells, 2 h prior to the end of the treatment period (72 h)

Fig. 6

Reduction of TXNIP expression in MCF-7 cells by 1,25(OH)₂D₃ is possibly ER-dependent. a–c 1,25(OH)₂D₃ (100 nM) treatment significantly reduces ER⍺ mRNA (72 h) and protein expression in MCF-7 cells. Relative expression was calculated using the ∆∆Ct method, with vinculin as the housekeeping gene. Statistical significance between DMSO- and 1,25(OH)₂D₃-treated cells is calculated using a two-tailed Student’s t test, where p values less than or equal to 0.05, 0.01, and 0.001 are depicted in the figures by *, **, and ***, respectively. Error bars ± SD; n = 3. Tamoxifen (10 μM) treatment (d), but not estradiol (100 nM) (e), reduces TXNIP protein expression in a time-dependent manner. f Non-significant regulation of TXNIP mRNA levels is observed in MCF-7 cells treated for 24 h with either tamoxifen or estradiol, alone and in combination with 1,25(OH)₂D₃. g Glucose uptake in MCF-7 cells is significantly reduced by various concentrations of tamoxifen (24 h), an effect that is significantly ablated in the presence of 1,25(OH)₂D₃. Estradiol reduces glucose uptake in a concentration-dependent manner, with 1,25(OH)₂D₃ only influencing regulation of glucose uptake in response to 10 nM treatment with estradiol