Selective inhibitory effects of zinc on cell proliferation in esophageal squamous cell carcinoma through Orai1

Abstract

Zinc, an essential micronutrient, has a cancer preventive role. Zinc deficiency has been shown to contribute to the progression of esophageal cancer. Orai1, a store-operated Ca²⁺ entry (SOCE) channel, was previously reported to be highly expressed in tumor tissues removed from patients with esophageal squamous cell carcinoma (ESCC) with poor prognosis, and elevation of its expression contributes to both hyperactive intracellular Ca²⁺ oscillations and fast cell proliferation in human ESCC cells. However, the molecular basis of cancer preventive functions of zinc and its association with Orai1-mediated cell proliferation remains unknown. The present study shows that zinc supplementation significantly inhibits proliferation of ESCC cell lines and that the effect of zinc is reversible with N,N,N′,N′-tetrakis (2-pyridylmethyl) ethylenediamine, a specific Zn²⁺ chelator, whereas nontumorigenic esophageal epithelial cells are significantly less sensitive to zinc treatment. Fluorescence live cell imaging revealed that extracellular Zn²⁺ exerted rapid inhibitory effects on Orai1-mediated SOCE and on intracellular Ca²⁺ oscillations in the ESCC cells. Knockdown of Orai1 or expression of Orai1 mutants with compromised zinc binding significantly diminished sensitivity of the cancer cells to zinc treatment in both SOCE and cell proliferation analyses. These data suggest that zinc may inhibit cell proliferation of esophageal cancer cells through Orai1-mediated intracellular Ca²⁺ oscillations and reveal a possible molecular basis for zinc-induced cancer prevention and Orai1-SOCE signaling pathway in cancer cells.—Choi, S., Cui, C., Luo, Y., Kim, S.-H., Ko, J.-K., Huo, X., Ma, J., Fu, L.-W., Souza, R. F., Korichneva, I., Pan, Z. Selective inhibitory effects of zinc on cell proliferation in esophageal squamous cell carcinoma through Orai1.

Esophageal cancer is the sixth leading cause of human cancer deaths worldwide, and the National Cancer Institute of the United States estimates that there were 16,910 new cases and 15,690 deaths from this malignancy in 2016 alone (1). Although there have been notable improvements in cancer survival rates over the past decades, the 5-yr survival rate of patients with esophageal cancer is still under 20%. These poor treatment outcomes have led to extensive research regarding the mechanisms of the carcinogenesis and effective chemoprevention for esophageal cancer.

There are two major forms of esophageal cancer: esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma. Epidemiologic studies suggest that gastroesophageal reflux disease and Barrett’s esophagus are risk factors for esophageal adenocarcinoma, whereas ESCC is strongly related to environmental factors, such as smoking, alcohol consumption, and zinc deficiency (2, 3). Accumulating evidence from both clinical and animal studies has demonstrated that zinc is important for maintaining healthy esophageal epithelium, and zinc deficiency results in abnormal esophageal cell proliferation, promoting tumor development (2, 4). Because zinc is an important cofactor of more than 300 enzymes and contributes to a stable 3-dimensional structure of numerous proteins, it plays multifunctional roles in maintaining normal and healthy esophageal epithelium (5). However, the exact molecular mechanisms underlying its cancer-preventive roles are not well understood.

A previous report from our group demonstrated that Orai1, a store-operated Ca²⁺ entry (SOCE) channel, was highly expressed in tumor tissues removed from patients with ESCC and that its high expression was strongly correlated with poorer prognosis (6). The elevated Orai1 expression contributes to hyperactive intracellular Ca²⁺ oscillations and cell proliferation in ESCC cells; both pharmacological channel inhibitors and knockdown (KD) of Orai1 reduced intracellular Ca²⁺ oscillations and cell proliferation in vitro and inhibited tumor growth in vivo. Zinc has been shown to modulate functions of a number of ion channels, such as the L-type Ca²⁺ channel and K⁺ channels, and zinc as a second messenger is may cross-communicate with Ca²⁺ signaling pathways (7, 8). However, in esophageal cancer, the connection between the cancer-preventive function of zinc and Orai1-regulated cell proliferation as well as the molecular basis for this association have not been established. This study presents molecular and intramolecular targets for zinc cancer-preventive activities and reports selective zinc inhibitory effects on ESCC cells compared with nontumorigenic esophageal epithelial cells.

MATERIALS AND METHODS

Cell culture and transfection

HET-1A, KYSE-150, KYSE-30, and KYSE-790 cell lines were cultured in RPMI-1640/Ham’s F12 medium (1:1 mix) (Sigma-Aldrich, St. Louis, MO, USA) supplemented with 5% fetal bovine serum (FBS) and 1% penicillin/streptomycin at 37°C in a 5% CO₂ humidified incubator. KYSE-150 Orai1-KD cells were created by stable transfection of the short hairpin RNA (shRNA) targeting the 3′-UTR of human orai1 gene (shRNA-Orai1) (6). NES-G4T, an immortalized human normal esophageal squamous epithelial cell line, was maintained in DMEM/Ham’s F12 medium (3:1 mix) (Sigma-Aldrich) supplemented with 1% cosmic calf serum (HyClone; GE Healthcare Life Sciences, Logan, UT, USA), hydrocortisone (0.4 μg/ml), epidermal growth factor (20 ng/ml), transferrin (5 μg/ml), insulin (5 μg/ml), cholera toxin (10⁻¹⁰ M), tri-iodothyronine (2 × 10⁻¹¹ M), adenine (180 μM), and 1% penicillin/streptomycin at 37°C in a 5% CO₂ humidified incubator (9). All transfections were performed using Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA) according to the manufacturer’s protocol.

Generation of human Orai1 mutants

The plasmids containing genes encoding green fluorescent protein (GFP) fusion proteins with either wild type (WT) or mutants of human Orai1 have been previously described (6). To generate various Orai1 mutants, we followed the protocol of Quickchange site-directed mutagenesis (Stratagene, La Jolla, CA, USA) with minor modifications. Forward and reverse primers, including mutations in the DNA sequences encoding amino acids of interests, are listed in Table 1. Template DNA (30 ng) and Pfu Ultra polymerase (Stratagene) were used in the PCR, and the PCR products were digested with DpnI (New England Biolabs, Ipswich, MA, USA) to remove the original templates and transformed into competent DH5-α cells. All plasmids were sequenced to confirm the mutation.

TABLE 1.

Primers used in site-directed mutagenesis

Primer, hOrai1	Sequence, 5′–3′
	Forward	Reverse
H113A	GCTGGACGCTGACGCCGACTACCCACCGG	CCGGTGGGTAGTCGGCGTCAGCGTCCAGC
C126A	CGCCTTCAGTGCCGCCACCACAGTGCTGG	CCAGCACTGTGGTGGCGGCACTGAAGGCG
C143A	CATGATCAGCACCGCCATCCTGCCCAACA	TGTTGGGCAGGATGGCGGTGCTGATCATG
C195A	GGTGGTGCTGCTCGCCTGGGTCAAGTTCT	AGAACTTGACCCAGGCGAGCAGCACCACC
3DA	GGAGGTGCAGCTGGCCGCTGCCCACGCCTACCCACCGGGG	CCCCGGTGGGTAGGCGTGGGCAGCGGCCAGCTGCACCTCC

Cell proliferation assay

Each cell line and the Orai1 mutation–transfected cells were seeded in 96-well plates at 2000 cells/well and cultured for 16 h. The cells were then treated with the indicated amounts of ZnSO₄ for the indicated lengths of time. Stock solution, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (Sigma-Aldrich) was prepared at 1 mg/ml and filter sterilized. For the assay, 20 µl of MTT solution was added to each well containing 100 µl of medium and cells, and the plate was placed in the incubator. After 2.5 h, all media, including the MTT solution, were carefully removed, and 120 µl DMSO (Sigma-Aldrich) was added to solubilize the purple precipitates. From each well, including the blanks, the absorbance at 570 nm was measured in a FlexStation 3 microplate reader (Molecular Devices, Sunnyvale, CA, USA). Relative cell number percentages were calculated by comparing the optical density value of each group to the value of the appropriate control group, depending on experiments.

Cell cycle analysis

Cells were seeded in 6-well plates at density of 2 × 10⁵ cells/well and cultured overnight before addition of ZnSO₄ alone or with N,N,N′,N′-tetrakis (2-pyridylmethyl) ethylenediamine (TPEN). Cells were cultured for another 24 h and then collected by trypsin digestion. After gentle washing with PBS, the cell suspensions were transferred to ice-cold 70% ethanol while being gently vortexed to maintain the single cell suspension. The fixed cells were kept in the ethanol for 4 h at 4°C. After centrifugation of the ethanol-suspended cells for 5 min at 300 g, the cell pellets were washed with PBS, resuspended, and incubated in PBS supplemented with 50 μg/ml DNase-free RNase for 30 min. The cells were washed with PBS and stained with propidium iodide staining solution, including both DNase-free RNase A and a permeabilization reagent (Thermo Fisher Scientific, Waltham, MA, USA), according to the manufacturer’s protocol. A FACSAria IIu flow cytometer (BD Biosciences, San Jose, CA, USA) and CellQuest software (BD Biosciences) were used to determine the cell percentage distribution of G₀/G₁, S, and G₂/M phases.

Intracellular Ca²⁺ measurement

SOCE and Ca²⁺ oscillations were monitored following previously published procedures (10, 11). SOCE measurements were performed either with the glass dish–based fluorescent microscope (to visualize single cells) or with a cuvette-based spectrofluorometer (to detect fluorescent signals from cell population). For the microscope method, the cells cultured in glass-bottomed dishes (1.5 coverslip; MatTek, Ashland, MA, USA) were loaded with 5 μM Fura-2 acetoxymethyl ester (Biotium, Fremont, CA, USA) for 45 min at 37°C. Intracellular Ca²⁺ was measured by the fluorescence microscope with a ×40 objective (TE200 Super Fluo; Nikon, Tokyo, Japan) that was connected to a dual-wavelength spectrofluorometer (excitation λ = 350/385 nm and emission λ = 510 nm) (Photon Technology International, Birmingham, NJ, USA). Cellular endoplasmic reticulum (ER) Ca²⁺ stores were depleted by 5 μM thapsigargin (TG) and 0.5 mM EGTA in balanced salt solution (BSS) (mmol 140 NaCl, 2.8 KCl, 2 MgCl₂, 10 HEPES; pH 7.2), and then SOCE was observed upon addition of 2 mM CaCl₂. To examine their effects on SOCE according to the experiments, 25 µM ZnSO₄ and/or 100 µM TPEN were added. SOCE activities are presented with the difference between basal and maximal values of F₃₅₀/F₃₈₅ (ΔF₃₅₀/F₃₈₅) after addition of 2 mM CaCl₂ in BSS. For the cuvette-based spectrofluorometer method, cells were harvested, and 10⁶ cells were prepared in BSS in a quartz cuvette (10 mm) (Starna Cells, Atascadero, CA, USA). The fluorescence signals (excitation λ = 350/385 nm; emission λ = 510 nm) were recorded while the TG, EGTA, CaCl₂, ZnSO₄, and/or DTT were added to the cuvette.

To measure Ca²⁺ oscillations, cells were seeded in 96-well imaging plates (BD Falcon, Franklin Lakes, NJ, USA) and loaded with 3 μM Fluo-4-AM (Invitrogen) in culture medium without serum. After 20 min, the cell culture medium was changed to the normal culture medium without phenol red, and the cellular fluorescence intensity was monitored using the BD Pathway 855 BioImager System (BD Biosciences) at 37°C in the presence of 5% CO₂. Time-lapse live cell imaging was recorded and analyzed with the manufacturer’s software.

Western blot analysis

Cells were lysed in modified RIPA buffer in the presence of protease inhibitors cocktail (Sigma-Aldrich) as previously described (12). Equal amounts of proteins were run in SDS polyacrylamide gels, and the separated proteins were transferred to PVDF membranes (Bio-Rad, Hercules, CA, USA). The blot was incubated with 5% nonfat dry milk blocking buffer (Bio-Rad) for 2 h and probed with specific antibodies for Orai1 (EMD Millipore, Billerica, MA, USA) or Tubulin (Sigma-Aldrich) in 5% milk buffer overnight. After incubating the appropriate horseradish peroxidase–conjugated secondary antibodies (Cell Signaling Technology, Danvers, MA, USA), protein expression was visualized using ECL detection methods (Denville Scientific, Holliston, MA, USA).

Confocal microscopic imaging

Cells were seeded in glass-bottomed dishes and transfected with various plasmids. After 24 h of incubation, the media were replaced with the normal medium containing 1 μg/ml Hoechst 33342 (Thermo Fisher Scientific) for 5 min to stain nuclei. Cells were washed with PBS 2 times and treated with BSS with 2 mM CaCl₂ (Resting) or with 0.5 mM EGTA plus 10 μM TG (+TG) for 20 min. Fluorescent images were taken with a laser scanning confocal microscope (LSM 780; Carl Zeiss, Jena, Germany). GFP, mOrange, and Hoechst 33342 were excited by 488-, 548-, and 405-nm diode lasers, respectively.

Statistical analyses

All experiments were performed at least 3 times, and most results are presented as means ± sd. The data sets with high sample numbers are shown as means ± se as indicated in the figure legends. Statistical analyses were carried out using Student’s t test or 1-way ANOVA with Tukey’s post hoc test.

RESULTS

KYSE-150 cells are more sensitive to extracellular zinc than HET-1A cells

Cell growths were examined in the human ESCC cell lines KYSE-150, KYSE-30, and KYSE-790 and in the nontumorigenic esophageal epithelial cell line HET-1A upon treatment with various concentrations of extracellular ZnSO₄ in culture medium (Fig. 1A, B; Supplemental Fig. 1). The MTT assay demonstrates that ZnSO₄ supplementation at a concentration of 25 μM was not notably effective because it did not change the growth curve of any cell line. Whereas ZnSO₄ supplementation at a concentration of 75 μM significantly affected cell growth of all cell lines [Fig. 1A (right panels), B (dashed lines)], 50 μM ZnSO₄ inhibited cell growth in KYSE-150 but not in HET-1A cells [Fig. 1A (middle panel), B (dotted lines)]. At this concentration or lower, no sign of apoptosis or necrosis was observed in HET-1A and ESCC cells. Furthermore, the relative numbers of cells cultured in various ZnSO₄ concentrations for 24 h clearly demonstrate that KYSE-150 cells were much more sensitive than HET-1A (56 vs. 89.1 or 23.6 vs. 58.6% at 50 or 75 μM of ZnSO_4, respectively) (Fig. 1C). Representative phase-contrast images show that 50 μM ZnSO₄ causes significant reduction of cell numbers in KYSE-150 but not in HET-1A cells (Fig. 1A). These results indicate that the growth of KYSE-150 cells is more affected by zinc supplementation than that of HET-1A cells. Similar results were observed in other ESCC cell lines (e.g., KYSE-30 and KYSE-790) (Supplemental Fig. 1). Compared with NES-G4T, another nontumorigenic esophageal epithelial cell line, KYSE-150 cells were more sensitive to ZnSO₄ treatment in the same cell culture medium for NES-G4T, especially at concentrations of 50 or 75 μM (Supplemental Fig. 2).

Figure 1.

Human esophageal squamous cell carcinoma KYSE-150 cells are more sensitive to zinc treatment than nontumorigenic esophageal epithelial HET-1A cells. A) Representative phase-contrast images of HET-1A and KYSE-150 cells obtained with 0, 50, or 75 µM ZnSO₄ in culture medium. B) Growth curves of HET-1A and KYSE-150 cells in culture medium with various concentrations of ZnSO₄. Relative cell numbers at each indicated time point were evaluated by MTT assay and are presented as a percentage of cell numbers at 0 h. C) Comparison of HET-1A and KYSE-150 cells upon treatments with various concentrations of ZnSO₄. All cells were grown for 24 h in the culture medium containing the indicated concentrations of ZnSO₄ and were analyzed by MTT assay. Relative cell numbers were expressed as a percentage of cell numbers at 24 h in culture medium without ZnSO₄. *P < 0.01.

When the extracellular ZnSO₄ concentration was higher than 75 μM, many round-shaped and floating dead KYSE-150 and HET-1A cells were found as early as 10 h after treatment (Fig. 1A, right panel), suggesting that zinc induced cytotoxicity at these concentrations. Because the focus of this study was to examine the impact of zinc supplementation on cell proliferation of ESCC, 50 μM ZnSO₄ was chosen for the following experiments unless otherwise stated.

Zinc induces cell cycle arrest at G₂/M phase in KYSE-150 cells

Flow cytometry was used to analyze the cell cycle distribution of KYSE-150 cells cultured in medium containing 0, 30, or 50 μM ZnSO₄. The percentage of cells in G₂/M phase increased to 30% in the cell culture treated with 50 μM ZnSO₄, compared with 13% in control cells without zinc treatment (Fig. 2A). At concentrations of 50 μM or lower, ZnSO₄ did not significantly increase the sub-G₁ population, indicating no induction of apoptosis. This result was consistent with the observations of normal cell morphology (Fig. 1A).

Figure 2.

Zinc induces cell cycle arrest at the G₂/M phase. A) Cell cycle analysis of KYSE-150 cells by flow cytometry. The cells were grown in culture medium for 24 h with the indicated concentrations of ZnSO₄ (0, 30, or 50 µM) or ZnSO₄ (50 µM) plus TPEN (5 µM). Relative cell numbers (%) in the sub-G₁, -G₀/G₁, S, and -G₂/M phases for each sample were calculated (left panel). To highlight the changes in different ZnSO₄-treated samples, the relative cell numbers in G₂/M phase were replotted for each group (right panel). B) Relative cell numbers of KYSE-150 cells in the presence of 50 µM ZnSO₄ and TPEN (0, 5, 10 µM). All cell samples were cultured for 24 h prior to MTT assay. Relative cell numbers were expressed as a percentage of cell numbers in normal culture medium. Phase-contrast images were obtained from the cells grown in either normal culture medium (control) or medium containing 50 µM ZnSO₄ or with 5 µM TPEN (ZnSO₄ + TPEN), respectively. *P < 0.01, **P < 0.001.

To determine whether the effects of ZnSO₄ supplementation on ESCC cells were due to zinc ions (Zn²⁺), TPEN, a specific Zn²⁺ chelator, was introduced (Fig. 2A, B). The addition of 5 μM of TPEN in culture medium reduced the G₂/M phase from 30 to 14% and recovered cell proliferation nearly to 100% in KYSE-150 cells, even in the presence of 50 μM ZnSO₄. These data suggest that TPEN could completely abolish the effects of ZnSO₄ treatment. It is noteworthy that 10 μM of TPEN produced the same rescue effects, but higher concentrations of TPEN (e.g., 25 μM) caused significant cell death, likely due to severe zinc deficiency in KYSE-150 cells (data not shown).

Zinc inhibits store-operated calcium entry and intracellular Ca²⁺ oscillations in KYSE-150 cells

To determine the association between zinc inhibitory effects on cancer cell proliferation and the Orai1-mediated intracellular Ca²⁺ signaling pathway, SOCE was measured by real-time fluorescent microscopy in KYSE-150 cells in extracellular bath solutions with or without zinc (Fig. 3). The addition of 10 μM TG in Ca²⁺-free bath solution induced the release of Ca²⁺ from the ER into cytosol and resulted in Ca²⁺ store depletion; the addition of 2 mM Ca²⁺ outside resulted in the second peak of elevation of intracellular Ca²⁺ (Fig. 3A, left panel), which was previously confirmed as Orai1-mediated SOCE (6). However, when 25 μM ZnSO₄ was added to extracellular bath solution together with 2 mM Ca²⁺, the amplitude of SOCE was reduced by more than 80% compared with application of 2 mM Ca²⁺ alone. A further test demonstrated that the combination of 1 mM DTT with Ca²⁺ and Zn²⁺ in bath solution recovered SOCE to even a slightly higher extent (Fig. 3A, right panel). To specifically confirm the role of Zn²⁺, TPEN was applied in extracellular bath solution. The inhibited SOCE in Ca²⁺/Zn²⁺ bath solution was almost instantaneously restored when TPEN was added (Fig. 3B), suggesting that Zn²⁺ may directly regulate the SOCE channel.

Figure 3.

Zinc inhibits Orai1-mediated SOCE in KYSE-150 cells. A) Representative traces of changes in intracellular Ca²⁺ in cells loaded with the fluorescent Ca²⁺ indicator Fura-2. The ratio of fluorescence (emission at 510 nm) with excitation at 350 and 385 nm (F₃₅₀/F₃₈₅) was used to represent the changes in intracellular Ca²⁺. After being loaded with Fura-2 AM, KYSE-150 cells were trypsinized, suspended in BSS, and immediately subjected to measurement by cuvette-based spectrofluorometer. After initial 30 s of baseline recording, TG (10 µM) in BSS (with EGTA at final concentration of 0.1 mM) was used to deplete ER Ca²⁺ stores (first peak). After ER Ca²⁺ stores were completely depleted, the following compounds were added to the cell suspensions: CaCl₂ (2 mM) (left), CaCl₂ plus ZnSO₄ (25 µM) (middle), or CaCl₂ plus ZnSO₄ (25 µM) and DTT (1 mM) (right). In the left panel, the addition of CaCl₂ to the suspension solution resulted in a significant elevation of intracellular Ca²⁺ (second peak), indicating SOCE-mediated Ca²⁺ influx. To clearly compare the SOCE in different groups, traces of the TG-induced first peak were truncated (indicated by double diagonal lines), and only the second peaks are shown. B) Average traces of SOCE in microscope-based spectrofluorometry. KYSE-150 cells were cultured in glass-bottomed dishes, and the intracellular Ca²⁺ in each individual cell was monitored with a charge coupled device camera. After the ER Ca²⁺ stores were depleted by TG (10 µM), SOCE was measured upon 2 mM CaCl₂ treatment followed by an additional 50 µM TPEN (left). The addition of 25 µM ZnSO₄ (Ca²⁺ + Zn²⁺) completely blocked SOCE, which could be reversed by 50 µM TPEN (right). Data are means ± se (n > 20).

Elevation of Orai1 expression in ESCC was previously reported to be associated with hyperactive intracellular Ca²⁺ oscillations (6). Thus, the inhibitory effect of zinc on Ca²⁺ oscillations was examined in KYSE-150 cells using time-lapse live cell imaging (Fig. 4). Loaded with Ca²⁺-specific fluorescent indicator Fluo-4 AM, more than 70% of KYSE-150 cells demonstrated active intracellular Ca²⁺ oscillations in the cell culture medium containing 1.8 mM Ca²⁺. The addition of 50 μM ZnSO₄ in culture medium almost completely abolished the intracellular Ca²⁺ oscillations but did not affect the baseline fluorescence intensity.

Figure 4.

Zinc inhibits intracellular Ca²⁺ oscillations. A) Representative changes in intracellular Ca²⁺ by fluorescent imaging in KYSE-150 cells. The time-lapse live cell fluorescence imaging was performed in the cells loaded with Fluo-4 AM in a phenol-red free culture medium (control) or medium supplemented with 50 µM ZnSO₄. Snapshots at 10, 20, and 30 s from each group are shown. Red and green circles indicate the single cells that were analyzed in B. B) Representative traces of intracellular Ca²⁺ revealed by relative fluorescent intensity. Red (control) and green (ZnSO₄) traces were obtained from the cells circled in A, and yellow arrowheads indicate the time points at which the images in A were captured.

KD of Orai1 expression reduces the inhibitory effects of zinc on esophageal squamous cell carcinoma cell proliferation

To further test whether Orai1 is a target of zinc-induced growth inhibitory effects in ESCC cells, KYSE-150 cells were transfected with plasmids containing shRNA, which was previously shown to specifically target the 3′-UTR of orai1 (KYSE-150 Orai1-KD cells) (6). Using Western blot analysis, the expression level of Orai1 in KYSE-150 Orai1-KD cells was estimated to be ∼20% of the parent KYSE-150 cells (Fig. 5A). Then, cell proliferation rates of these 2 cell lines and HET-1A were compared in normal and ZnSO₄-supplemented culture medium (Figs. 1B and 5B). The cell number decreased by 43% in KYSE-150 cells in ZnSO₄-supplemented medium as compared with the control medium but was reduced by only 19 and 8% in KYSE-150 Orai1-KD cells and HET-1A cells, respectively (Fig. 5C). These results suggest that Orai1 is, at least partially, a molecular target of zinc inhibitory functions on cell proliferation in ESCC.

Figure 5.

KD of Orai1 reduces zinc inhibitory effects on cell proliferation. A) Western blot analysis of Orai1 expression in HET-1A, KYSE-150, and KYSE-150 Orai1-KD cells. Expression of tubulin is shown as a loading control. B) Growth curves of KYSE-150 Orai1-KD and parental KYSE150 cells in culture medium with or without 50 µM ZnSO₄. Relative cell numbers at each indicated time point were evaluated by MTT assay for 48 h and are presented as a percentage of cell numbers at 0 h. C) Comparison of cell growth of HET-1A, KYSE-150, and KYSE-150 Orai1-KD cells in culture medium with or without 50 µM ZnSO₄. MTT assays were performed on the cells cultured for 24 h, and relative cell numbers were expressed as a percentage of cell numbers in control medium (without ZnSO₄) for each cell line. Data are means ± sd of 3 independent experiments, and statistical significances are shown in the bar graphs. *P < 0.01, **P < 0.001.

Histidine and cysteine residues of Orai1 are involved in zinc inhibitory effects on Orai1 activity

Zn²⁺ has been shown to affect the functions of a number of ion channels and transporters, such as transient receptor potential (TRP)A1, TRPM5, and NMDA receptors, via its direct interaction with histidine, cysteine, aspartate, or glutamate residues in proteins (7). To identify the potential zinc-interacting sites in Orai1, site-directed mutagenesis was conducted to replace the histidine and aspartate residues with alanine in the flexible outer vestibule region between transmembrane (TM)1 and TM2 of Orai1 (i.e., H113, D110, D112, and D114) as well as 3 cysteine residues (i.e., C126, C143, and C195) (Fig. 6A). Simultaneous 3-point mutations on D110, D112, and D114 were generated (designated as Orai1 3DA). The plasmids containing GFP-tagged orai1 WT or mutants were transfected into KYSE-150 Orai1-KD cells. The confocal microscope images demonstrate that Orai1 WT and various mutants are expressed at plasma membrane (Fig. 6B; Supplemental Fig. 3). Upon TG treatment to deplete ER Ca²⁺ stores, the Orai1 WT can form puncta and colocalize with translocated STIM1 that was tagged with mOrange (Fig. 6B, lower panels). These data indicate that Orai1 WT and mutants are expressed properly in KYSE-150 Orai1-KD cells.

Figure 6.

The critical amino acid residues in Orai1 are involved in zinc-induced growth inhibition. A) Schematic illustration of Orai1 structure. Red, orange, and green rectangles represent histidine, cysteine, and aspartate residues, respectively. Plasma membrane (PM), TM domains, amino acid sequence numbers (numbers in TMs), and TM-linker regions (black bent lines) are shown. B) Representative confocal live cell imaging of the cells expressing GFP-Orai1 and mOrange-STIM1 in KYSE-150 Orai1-KD cells. Resting: cells in BSS containing 2 mM Ca²⁺; +TG: cells treated with 10 µM TG to deplete ER Ca²⁺ stores. C) Comparison of growth of KYSE-150 Orai1-KD cells expressing Orai1 WT or mutants in culture medium containing ZnSO₄ at the indicated concentrations. All transfected cells were collected by cell sorting and cultured in medium containing 0, 25, or 50 µM ZnSO₄ for 24 h. Using MTT assay, relative cell numbers for each group were analyzed and are presented as a percentage of cell numbers in control medium. Data are means ± sd from a representative experiment and statistical significance compared with cells expressing Orai1 WT. **P < 0.001.

The cells transfected with Orai1 3DA did not survive throughout the experiments likely due to dysfunction of the Orai1 channel in culture medium (with ∼0.89 mM Ca²⁺). Therefore, cell proliferation with or without ZnSO₄ treatment was compared in KYSE-150 Orai1-KD cells transfected with Orai1 WT and 4 mutants with single-point mutations (i.e., H113A, C126A, C143A, and C195A). KYSE-150 Orai1-KD cells transfected with Orai1 WT were sensitive to 50 μM ZnSO₄ treatment showing a decrease in the relative cell number up to 49% as compared with the cells in control medium (Fig. 6C). However, KYSE-150 Orai1-KD cells transfected with Orai1 mutants maintained 80–85% of cell proliferation index in the medium containing 50 μM ZnSO₄, thus displaying a decreased sensitivity to zinc. The Orai1-mediated SOCE activities in KYSE-150 Orai1-KD cells transfected with Orai1 WT and mutant forms are summarized in Fig. 7. KYSE-150 Orai1-KD cells transfected with Orai1 WT displayed almost the same level of SOCE as parental KYSE-150 cells, and the SOCE was inhibited by zinc treatment by 43.7% (Fig. 7A–C). However, SOCE activity in KYSE-150 Orai1-KD cells transfected with Orai1 H113A, C126A, C143A, or C195A did not reveal significant differences with or without treatment (Fig. 7A, B).

Figure 7.

H113 and 3 cysteine residues in Orai1 are involved in zinc-mediated SOCE inhibition. A) Representative traces of intracellular Ca²⁺ in KYSE-150 Orai1-KD cells expressing Orai1 WT or mutants treated with or without ZnSO₄. To clearly compare the SOCE in different groups, the traces for TG-induced first peak were truncated (indicated by double diagonal lines), and only the SOCE-induced second peaks are shown. B) Analysis of SOCE activities in KYSE-150 Orai1-KD cells expressing Orai1 WT or mutants. SOCE activity is presented by the difference between basal and maximal values of F₃₅₀/F₃₈₅ (ΔF₃₅₀/F₃₈₅) after adding CaCl₂ in BSS. White bars represent SOCE activities in the cells upon 2 mM CaCl₂ treatment; black bars represent those upon 2 mM CaCl₂ plus 25 µM ZnSO₄. Bar graphs show mean ± se and statistical significances compared with the CaCl₂-treated cells in each cell group (n > 20). C) Relative SOCE resulted from CaCl₂ plus ZnSO₄ vs. from CaCl₂ only. Bar graphs show mean ± se and statistical significances compared with the relative percentages of SOCE in the cells expressing Orai1 WT. *P < 0.05, **P < 0.005; ***P < 0.001.

DISCUSSION

Dietary zinc deficiency is linked to a broad spectrum of diseases that includes immune dysfunction, diabetes, cardiovascular disorders, and cancers (7). For several decades, zinc-deficient populations have shown increased susceptibility to the development of esophageal cancer (13). Mechanisms underlying the inverse association between zinc and esophageal cancer have been linked to zinc deficiency–induced single- and double-strand DNA breaks, increased oxidative stress, and/or impaired immune function (14, 15). This association, however, may involve other mechanisms because Zn²⁺ can modulate a number of ion channels and transporters and, in its role as a second messenger, can cross-communicate with other signal transduction pathways, including Ca²⁺ signaling (7). The possibility of other mechanisms and novel molecular targets has not been well studied, limiting current knowledge in the development of cancer prevention and management. The present study is thus the first report, to our knowledge, that identifies specific intramolecular targets within Orai1 protein (e.g., histidine and cysteine) that are affected by Zn²⁺, resulting in the inhibition of Orai1-mediated Ca²⁺ entry, intracellular Ca²⁺ oscillations, and cell proliferation in esophageal cancer cells.

In addition to our previous findings on the increased Orai1 expression in ESCC cells (6), this study demonstrates that cancer cells, compared with nontumorigenic esophageal epithelial cells that contain low expression of Orai1, show significantly higher sensitivity to zinc treatment (Fig. 1; Supplemental Fig. 1). This implies that the overexpressed Orai1 channel may be associated with the inhibitory effects of zinc. Both knocking down Orai1 and mutating the potential zinc-binding amino acid residues reduced the zinc inhibitory effects on SOCE and cell proliferation (Figs. 6C and 7). Uncontrolled proliferation is a hallmark of cancer cells and is involved in both carcinogenesis and cancer progression. Therefore, results from this study suggest that Orai1-mediated intracellular Ca²⁺ signaling can be targeted by zinc to prevent cancer development and/or progression (Fig. 8).

Figure 8.

Hypothetical scheme illustrating zinc inhibitory functions in ESCC cells. Compared with normal esophageal epithelial cells, ESCC cells heavily rely on Orai1-mediated SOCE and intracellular Ca²⁺ oscillations to support their increased need for rampant cell proliferation. In zinc-sufficient conditions, Zn²⁺ can bind to Orai1 protein likely through H113 and/or cysteine residues to block the channel activity. Subsequently, zinc inhibits SOCE-mediated intracellular Ca²⁺ oscillations and cell proliferation in ESCC cells. Therefore, by targeting Orai1-mediated Ca²⁺ signaling, zinc represents selective cell growth inhibition in ESCC cells. The hexametric structure of Orai1 was adopted here and is shown as red, yellow, green, yellow-green, cyan, and blue cylinders with extracellular loops between TM1 and TM2. C195 (red star) is a redox sensor on TM3 in each monomer. STIM1 is presented as brown cylinders coupled with Orai1. The orange and purple circles represent Ca²⁺ and Zn²⁺, respectively.

Our data show that the addition of ZnSO₄ induces cell cycle arrest at the G₂/M phase (Fig. 2A). Intracellular Ca²⁺ plays important roles throughout the mammalian cell cycle; Ca²⁺ oscillations are especially indispensable at the G₁/S boundary (centrosome duplications) and G₂/M transition (centrosome separation) (16). Because zinc also inhibited intracellular Ca²⁺ oscillations (Fig. 4), these data suggest that zinc causes cell cycle arrest in ESCC cells at the G₂/M phase and inhibits cell proliferation at least partially through intracellular Ca²⁺ signaling.

As a specific Zn²⁺ chelator, TPEN restored cell proliferation in ZnSO₄-treated KYSE-150 cells. This result suggests that the antiproliferation effect of ZnSO₄ is due to Zn²⁺. Interestingly, although the stoichiometry of zinc-TPEN is known to be 1:1 (17, 18), the effects of 50 μM ZnSO₄ supplement on cell proliferation and cell cycle were completely rescued by 10 or even 5 μM TPEN. The absolute concentration of free Zn²⁺ in the zinc-supplemented cell culture medium does not correspond to added zinc because the medium contains FBS and other zinc-binding ligands, such as histidine, cysteine, and sodium phosphate dibasic (19, 20). FBS contains a large amount of albumin that binds zinc ions with relatively high affinity (K_d ∼0.1 µM). Upon measuring the free Zn²⁺ concentrations in DMEM containing 100 µM total zinc, a previous study estimated free Zn²⁺ to be <50 nM (20). It also suggested that typical cell culture medium contains free Zn²⁺ levels 2–3 orders of magnitude lower than total zinc. In accordance with these reports, the free Zn²⁺ concentrations in the present study are likely to be much less than the concentrations of added ZnSO₄ because our cell culture medium contains 5% FBS and other zinc-binding ligands (Fig. 3). Much lower concentrations of ZnSO₄ were needed to block SOCE because SOCE measurement was conducted in a balanced buffer solution without FBS or amino acids. In the human body, plasma zinc turns over rapidly to meet the needs of tissues and has to be replenished daily from the diet. Thus, the fluctuation of free Zn²⁺ in the range of a few micromolars likely occurs in plasma or tissue depending upon zinc sufficiency or deficiency in the diet. Measuring the exact free Zn²⁺ concentration in esophagus tissues in zinc-sufficient or zinc-deficient individuals remains a technical challenge and requires the development of zinc-specific probes for biologic measurement.

A number of studies have shown that zinc can be a modulator, and often an inhibitor, of Ca²⁺ channels, such as purinergic receptors P2X2 and L-type or T-type Ca²⁺ channels (7). Early work from Hoth et al. (21) found that Ca²⁺ release–activated Ca²⁺ current (I_CRAC) is highly specific for Ca²⁺ and that Zn²⁺ is the most potent divalent ion, second only to trivalent ion La³⁺, in mast cells to block I_CRAC. Zn²⁺ reduced 24% of I_CRAC at 10 μM, and 1 mM of the metal ion completely inhibited I_CRAC. Later, Gore et al. (22) reported that Ca²⁺ release—activated channels (CRAC) are not permeable for Zn²⁺ and speculated that the inhibitory effects of Zn²⁺ occur via its binding to the extracellular site of channels. Although it is not clear how closely the channel properties are shared in Orai1-mediated SOCE in ESCC cells and I_CRAC in mast cells, our data suggest that zinc does not permeate through Orai1-mediated SOCE (Supplemental Fig. 4). Considering that TPEN was able to rescue SOCE in zinc-treated KYSE-150 cells in an almost instantaneous fashion (Fig. 3), Zn²⁺ may directly bind to the Orai1 channel and inhibit its channel activity.

Zn²⁺ is coordinated within proteins mainly through imidazole rings of histidine and sulfhydryl groups of cysteine and, to a lesser extent, through aspartate and glutamate residues (23–25). According to the topology information of Orai1 protein, H113 is located in the extracellular loop between TM1 and TM2 that forms a flexible vestibule at the mouth of the channel pore (26). Although there is considerable debate over whether stoichiometry structure of Orai1 is tetramer or hexamer, the H113 residue from each monomer may be closely located to each other. In agreement with this finding, in a previous study, the recombinant Orai1 containing a mutation of H113 to cysteine (H113C) was shown to be a nonfunctional channel (27). In our study, however, H113A Orai1 mutant was still functional in spite of its much lower SOCE activity (Fig. 7). This may be due to the smaller size and lesser reactivity of alanine in comparison to sulfhydryl group–contained cysteine.

The H113A mutant, in addition to its lower SOCE activity, became insensitive to zinc treatment. Considering its location within the vestibule of Orai1 channel and its capability for zinc binding, our data suggest that the proximate H113 residue from each monomer may interact with Zn²⁺ to form an imidazole ring-like structure coordinating the central metal ion. The critical roles of histidine in zinc-Orai1 interaction are consistent with a recent study that examined the frequency of amino acid residues serving as zinc ligands in a variety of ion channels and proteins with known structures (28). According to this study, histidine is a highly prevalent residue in catalytic sites of protein, whereas cysteine is prevalent in the zinc coordination sphere of structural proteins. The study also noted that, in agonist-gated receptors and channels, ∼70% of the Zn²⁺ coordination shell relies essentially on histidine. This report underlines the key role of H113 in Orai1 channel for zinc inhibitory effects on SOCE.

The mechanisms underlying how the 3 cysteine mutations of Orai1 hinder the inhibitory effects of zinc on SOCE remain unclear. The exact topological locations of the C126, C143, and C195 residues in human Orai1 have not been fully determined, although they are known to be arranged in TM2 and TM3 (29). C195, based on its role in Orai1 as a redox sensor and its accessibility to hydrogen peroxide at the exit of TM3 (30, 31), may be accessible by extracellular Zn²⁺. In KYSE-150 cells, the zinc-induced SOCE inhibition was completely restored by both DTT and β-mercaptoethanol (Supplemental Fig. 5). These data suggest that C195 may function as a redox sensor and/or as a zinc-binding site in Orai1 protein. Meanwhile, how zinc directly regulates the other 2 cysteine residues located in transmembrane domains remains unclear. Although we cannot rule out the possibility that zinc can still reach C126 and C143 through hydrophilic microenvironments between transmembrane domains to directly interact with the 2 cysteine residues, it is possible that C126 and C143 participate in zinc inhibitory function through mechanisms other than direct binding. The cysteine mutations may slightly distort the conformational structure of Orai1, leading to the overall resistance of SOCE to zinc supplementation.

In summary, the present study demonstrates that zinc obstructs Orai1-mediated SOCE and inhibits intracellular Ca²⁺ oscillations and cell proliferation in human ESCC cells. Esophageal cancer cells have more abundant Orai1 proteins than normal epithelial cells, and ESCC cells are more sensitive to the zinc inhibitory function. The findings of this study shed light on the chemoprevention and treatment of esophageal cancer by means of targeting zinc-Orai1 interaction. Specific dietary recommendations that aim to reach an optimal zinc status can also be considered to develop modalities for esophageal cancer prevention and treatment.

Supplementary Material

This article includes supplemental data. Please visit http://www.fasebj.org to obtain this information.

Glossary

BSS

balanced salt solution

CRAC

calcium release–activated channel

ER

endoplasmic reticulum

ESCC

esophageal squamous cell carcinoma

FBS

fetal bovine serum

GFP

green fluorescent protein

I_CRAC

calcium release–activated channel–activated calcium current

KD

knockdown

MTT

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide

shRNA

short hairpin RNA

SOCE

store-operated calcium entry

TG

thapsigargin

TM

transmembrane

TPEN

N,N,N′,N′-tetrakis (2-pyridylmethyl) ethylenediamine

TRP

transient receptor potential

WT

wild type

AUTHOR CONTRIBUTIONS

Z. Pan conceived and designed the experiments, analyzed the data, and wrote the paper; S. Choi and C. Cui performed the experiments, acquired and analyzed the data, and wrote the paper; Y. Luo and S.-H. Kim performed intracellular calcium measurement; J.-K. Ko performed initial mutagenesis work; X. Huo and R. F. Souza provided NES-G4T cells and participated in the assay using the cells; and J. Ma, L.-W. Fu, and I. Korichneva advised on experimental design and participated in the discussion and manuscript preparation.