Zinc ion dyshomeostasis increases resistance of prostate cancer cells to oxidative stress via upregulation of HIF1α

Abstract

Zinc ions (Zn²⁺) are known to influence cell survival and proliferation. However the homeostatic regulation of Zn²⁺ and their role in prostate cancer (PC) progression is poorly understood. Therefore the subcellular distribution and uptake of Zn²⁺ in PC cells were investigated. Inductively coupled plasma mass spectroscopy and fluorescent microscopy with the Zn²⁺-specific fluorescent probe FluoZin-3 were used to quantify total and free Zn²⁺, respectively, in the normal prostate epithelial cell line (PNT1A) and three human PC cell lines (PC3, DU145 and LNCaP). The effects of Zn²⁺ treatment on proliferation and survival were measured in vitro using MTT assays and in vivo using mouse xenografts. The ability of Zn²⁺ to protect against oxidative stress via a HIF1α-dependent mechanism was investigated using a HIF1α knock-down PC3 model. Our results demonstrate that the total Zn²⁺ concentration in normal PNT1A and PC cells is similar, but PC3 cells contain significantly higher free Zn²⁺ than PNT1A cells (p < 0.01). PNT1A cells can survive better in the presence of high concentrations of Zn²⁺ than PC3 cells. Exposure to 10 µM Zn²⁺ over 72 hours significantly reduces PC3 cell proliferation in vitro but not in vivo. Zn²⁺ increases PC3 cell survival up to 2.3-fold under oxidative stress, and this protective effect is not seen in PNT1A cells or in a HIF1α-KD PC3 cell model. A state of Zn²⁺ dyshomeostasis exists in PC. HIF1α is an integral component of a Zn²⁺-dependent protective mechanism present in PC3 cells. This pathway may be clinically significant through its contribution to castrate-resistant PC survival.

Keywords: castrate resistant; hypoxia inducible factor 1 alpha; iron; prostate cancer; zinc.

Figure 1. CRPC-like cells contain significantly higher basal free Zn²⁺ ions but equal total zinc compared to normal controls

(A) Total zinc concentration (ppb) measured by Inductively Coupled Plasma Mass Spectroscopy (ICP-MS) in untreated PNT1A, LNCaP, DU145 and PC3 cells. (B) Baseline intracellular free zinc (Zn²⁺) concentration (nM) was measured using a FluoZin-3 fluorescent probe in the same 4 prostate cell lines. Zn (nM) = Kd x (F-Fmin)/(Fmax-F) was used to calculate zinc concentration. Intracellular Zn²⁺ uptake following exposure to 10 μM (C) or 50 μM (D) ZnCl₂ for 4 or 24 hours was measured in PNT1A and PC3 cells. ^***p < 0.001 PNT1A vs. PC3 ^##p < 0.05 and ^##p < 0.01. Values are expressed as the mean ± SEM of at least three separate experiments.

Figure 2. Co-localisation of Zn²⁺ to mitochondria in prostate cells

(A) PC3 cells and (B) PNT1A cells were exposed to 10 μM ZnCl₂ for 30, 120 and 240 minutes and stained with Hoechst (DAPI), FluoZin-3 (FITC) and MitoTracker (TRITC) fluorescent dyes. Immunofluorescent microscopy images were acquired at 60× (oil immersion) magnification using a Nikon DS-Qi 1Mc Camera with 250 ms, 120 ms and 35 ms exposure times for DAPI, FITC and TRITC channels respectively. Composite images were created by merging blue (DAPI), green (FITC) and red (TRITC) channels. FITC and TRITC colour intensity for each pixel in the corresponding composite image is plotted on scatter plots. (C) Co-localisation of Zn²⁺ to mitochondria for PNT1A and PC3 cells was estimated using Pearson correlation coefficient between FITC (FluoZin-3) and TRITC (MitoTracker) colours. Values are expressed as fold change compared to time 0 hours. Statistical significance for the PNT1A vs. PC3 comparison was determined using the Bonferroni-Sidak method with α = 0.05. ^***p < 0.001. Values are expressed as the mean ± SEM of at least three separate experiments.

Figure 2. Co-localisation of Zn²⁺ to mitochondria in prostate cells

(A) PC3 cells and (B) PNT1A cells were exposed to 10 μM ZnCl₂ for 30, 120 and 240 minutes and stained with Hoechst (DAPI), FluoZin-3 (FITC) and MitoTracker (TRITC) fluorescent dyes. Immunofluorescent microscopy images were acquired at 60× (oil immersion) magnification using a Nikon DS-Qi 1Mc Camera with 250 ms, 120 ms and 35 ms exposure times for DAPI, FITC and TRITC channels respectively. Composite images were created by merging blue (DAPI), green (FITC) and red (TRITC) channels. FITC and TRITC colour intensity for each pixel in the corresponding composite image is plotted on scatter plots. (C) Co-localisation of Zn²⁺ to mitochondria for PNT1A and PC3 cells was estimated using Pearson correlation coefficient between FITC (FluoZin-3) and TRITC (MitoTracker) colours. Values are expressed as fold change compared to time 0 hours. Statistical significance for the PNT1A vs. PC3 comparison was determined using the Bonferroni-Sidak method with α = 0.05. ^***p < 0.001. Values are expressed as the mean ± SEM of at least three separate experiments.

Figure 2. Co-localisation of Zn²⁺ to mitochondria in prostate cells

(A) PC3 cells and (B) PNT1A cells were exposed to 10 μM ZnCl₂ for 30, 120 and 240 minutes and stained with Hoechst (DAPI), FluoZin-3 (FITC) and MitoTracker (TRITC) fluorescent dyes. Immunofluorescent microscopy images were acquired at 60× (oil immersion) magnification using a Nikon DS-Qi 1Mc Camera with 250 ms, 120 ms and 35 ms exposure times for DAPI, FITC and TRITC channels respectively. Composite images were created by merging blue (DAPI), green (FITC) and red (TRITC) channels. FITC and TRITC colour intensity for each pixel in the corresponding composite image is plotted on scatter plots. (C) Co-localisation of Zn²⁺ to mitochondria for PNT1A and PC3 cells was estimated using Pearson correlation coefficient between FITC (FluoZin-3) and TRITC (MitoTracker) colours. Values are expressed as fold change compared to time 0 hours. Statistical significance for the PNT1A vs. PC3 comparison was determined using the Bonferroni-Sidak method with α = 0.05. ^***p < 0.001. Values are expressed as the mean ± SEM of at least three separate experiments.

Figure 3. Zinc chloride treatment inhibits in vitro cell proliferation in CRPC-like PC3 cells

(A) Cell viability in PNT1A, LNCaP and PC3 cells after 48 hours of exposure to various concentration of ZnCl₂ was analysed by MTT assay. Proliferation of cells exposed to serum-free medium only (Ctrl) or to ZnCl₂ was assessed by MTT assay at 0, 24, 48 and 72 hours respectively in (B) PC3, (C) PNT1A and (D) LNCaP cells. Statistical analysis using one-way ANOVA (Ctrl vs. Zinc) was performed ^*p < 0.05, ^**p < 0.01 and ^***p < 0.001. Values are expressed as the mean ± SEM of at least three separate experiments. (E) PC3 xenograft tumour volume in SCID mice in three treatment arms: control (n = 7), ZnCl₂ 3mg/Kg (n = 3) and ZnCl₂ 10mg/Kg (n = 7). Arrows represent the days of ZnCl₂ or saline (control) IP injection. The percentage increase compared to day 0 was calculated and the mean ± SEM plotted. Statistical analysis was calculated by two-way ANOVA. ^#p < 0.05, ZnCl₂ 3 mg/Kg versus control.

Figure 4. Zn²⁺ mediated protection against oxidative stress injury in PC3 cells

(A) Cell survival/proliferation was measured by MTT assay. The data demonstrates increased resilience in PC3 cells compared to PNT1A under oxidative stress induced by increasing concentrations of H₂O₂. Cell survival was measured by MTT assay in (B) PC3 and (C) PNT1A cells preconditioned with the indicated concentrations of ZnCl₂ for 4 hours followed by oxidative stress (75 μM H₂O₂ for 24 hours). Values are expressed as the mean ± SEM of at least three separate experiments.

Figure 5. Zn²⁺ induces HIF1α protein expression in PC3 cells in a time- and dose-dependent manner

(A) Baseline normoxic HIF1α expression is high and (B) correlates with free Zn²⁺ concentration in PC cells. HIF1α protein expression following treatment with increasing concentrations of ZnCl₂ for 24 hours was analysed by Western blot in (C) PC3 and (D) PNT1A cells. Zinc ion-stimulated HIF1α protein expression was measured at the indicated times by Western blot in (E) PC3 and (F) PNT1A cells. ^*p < 0.05, ^**p < 0.01 and ^***p < 0.001. Values are expressed as the mean ± SEM of at least three separate experiments.

Figure 6. The protective role of Zn²⁺ against oxidative injury in PC3 cells is HIF1α dependent

A HIF1α knock-down PC3 cell line (HIF1α-KD) was created by transfection of a shRNA vector expressing HIF1α into wild-type (WT) PC3 cells resulting in (A) suppression of HIF1α expression on Western blot. (B) HIF1α protein expression in PC3-WT and PC3-HIF1α-KD cells following pre-treatment with 10μM ZnCl₂ for 4 hours was measured by Western blot. (C) Survival/proliferation was measured by MTT assay in HIF1α-KD cells following preconditioning with ZnCl₂ for 4 hours then exposure to oxidative stress (75 μM H₂O₂ for 24 hours). Values are expressed as the mean ± SEM of at least three separate experiments.

Figure 7. Zn²⁺ ions competitively inhibit HIF1α degradation by displacing Fe²⁺ ions

(A) HIF1α protein was degraded in the presence of ammonium ferric citrate (AFC) in DU145 and PC3 cells. (B) The reduction in HIF1α expression in the presence of iron was partially reversed by Zn²⁺ in PC3 cells. Values are expressed as the mean ± SEM of at least three separate experiments. (C) In a normal prostate epithelial cell (PNT1A) under normoxic conditions, the pathway for HIF1α degradation pathway is activated. Proteasomal degradation is achieved by binding to the pVHL-E3-ubiquitin complex mediated by prolyl-hydroxylase (PHD) enzymes which require the co-factors iron (Fe²⁺), ascorbate and 2-oxoglutarate. (D) In CRPC cells (PC3) under the same normoxic conditions the HIF1α pathway is inhibited by Zn²⁺ ions, which substitute for Fe²⁺ ions at the PHD binding site, and also potentially reduce the co-factor 2-oxoglutarate via mAC inhibition in the citric acid cycle. Ultimately HIF1α is overexpressed in CRPC leading to increased transcription of genes responsible for glucose metabolism, proteolysis, cell survival, erythropoiesis and angiogenesis.