Introduction

Nickel (Ni) is included in several medical devices, including prostheses, pace makers, stents, and dental implants, owing to its beneficial properties such as resistance to corrosion and durability. However, Ni ion elutes from Ni-containing materials possibly causing inflammation1,2,3. Actually, the prevention of neointima formation by Ni-free stainless stent was demonstrated4. We also reported that the implantation of an Ni wire subcutaneously into the back of mice induced the elution of Ni2+, the expression of several inflammatory proteins such as cyclooxygenase-2 (COX-2) and neutrophil chemokine macrophage inflammatory protein-2 (MIP-2, CXCL2), and leukocyte infiltration as the initial responses5,6. Importantly, infiltration and activation of neutrophils enhanced further elution of Ni2+5. Thus, inhibition of Ni2+-induced inflammatory cell activation would be one of the strategies to prevent Ni2+ elution.

It was generally accepted that Ni2+ binds to various extracellular proteins to form a novel antigen causing delayed-type hypersensitivity7,8,9. For example, Ni2+ binds to human serum albumin inducing activation of human T cells9. Furthermore, Ni2+ forms different Ni epitopes leading to polyclonal Ni-specific T cell activation. However, Ni2+ directly activates various inflammatory cells5 and induces death of monocytes10. For example, Ni2+ binds to Toll-like receptor 4 (TLR4) on the cell surface, activating the NF-κB pathway11. In addition to cell surface proteins, Ni2+ binds to and modulates intracellular proteins; these ions enter the cells and inhibit prolyl hydroxylases (PHDs), resulting in the activation of a transcription factor called the hypoxia-inducing factor-1α (HIF-1α)4,12. As HIF-1α activation plays crucial roles in cytokine production and angiogenesis, Ni2+ uptake into the cells was one of the important steps in Ni2+-induced damage.

Transporters for Ni2+ uptake have been reported in microorganisms13,14. In contrast, Ni2+ transport systems in human cells have not yet been identified. The uptake of heavy metal ions, such as Cu2+, Fe2+, and Zn2+, occurs via the divalent metal transporter, DMT1, in mammalian cells15,16. The Zn transporter, Zrt- and Irt-like protein (ZIP, SLC39A) family, which consists of over 25 members17, is also involved in the influx of several heavy metal ions. Each of these members exhibits specificity toward a specific metal. However, the metal specificity of the transporter involved in Ni2+ uptake remains unclear.

Ni2+ uptake in cells and nuclei in the human monocytic cell line, THP-1, has already been reported18. THP-1 cells also have the ability to produce IL-8 by treatment with Ni compounds19,20. Therefore, using THP-1 cells, we examined whether the competition between Ni2+ and other ions affected IL-8 production. Especially, to assess the accumulation of metals in the cells and Ni2+ elution in the tissues precisely, we used inductively coupled plasma mass spectrometry (ICP-MS), a highly sensitive and efficient analysis technique for detecting various metal ions. In this study, we found that the physiological concentration of Zn2+ affected the uptake of Ni2+ by THP-1 cells and the sensitivity of mouse to Ni2+.

Results

NiCl2-stimulated increase in Ni2+ content and IL-8 production in THP-1 cells

THP-1 cells were treated with various concentrations of NiCl2 for 24 h and the Ni2+ content in the cells and IL-8 level in the medium were determined. Both Ni2+ content and IL-8 production increased in a NiCl2 concentration-dependent manner (Fig. 1a and b). As IL-8 production was significantly induced by NiCl2 at the concentration of ≥0.2 mM (Fig. 1b), 0.2 mM NiCl2 was used in all the experiments. Ni2+ content in the cells increased in a time-dependent manner (Fig. 1c), consistent with the concentration-dependent increase in the cells, and IL-8 level in the medium increased significantly from the 4-h mark (Fig. 1d). The incubation of THP-1 cells in 0.2 mM NiCl2 for 24 h did not affect the viability as determined by the MTT assay (data not shown).

Figure 1
figure 1

Ni2+ uptake and IL-8 production in THP-1 cells. THP-1 cells were treated with various concentrations of NiCl2 for 24 h (a and b) and 0.2 mM NiCl2 for the indicated times (c and d). The amount of Ni2+ in the cells (a and c) and IL-8 in the supernatant (b and d) were determined using ICP-MS and ELISA, respectively. The vertical lines represent the S.E.M. of 3 samples. ##p < 0.01 vs. 0 mM (a and b) or 0 h (c and d).

Full size image

Effects of metal ions on the uptake of Ni ions

THP-1 cells were treated with 0.2 mM NiCl2 in the presence of various divalent cations (0.03 mM), including Zn2+, Mg2+, Fe2+, Co2+, Cu2+, or Mn2+, added as dichloride salts. The Ni2+ content in the cells after 24 h of incubation was determined by ICP-MS. The increase in the intracellular Ni2+ content was inhibited by ZnCl2, CoCl2, and MnCl2 (Fig. 2a). In contrast, the increase in Ni2+ content was not inhibited by the divalent metal transporter 1 (DMT1) inhibitor, Chlorazol Black (Fig. 2b). Because Ni2+ activates Toll-like receptor 4 (TLR4), the effects of the TLR4 inhibitor, TAK-242, on Ni2+ uptake were determined. The results suggested that TAK-242 did not affect Ni2+ uptake (Fig. 2c), suggesting that TLR4 activation was not involved in Ni2+ uptake. To confirm whether ZnCl2 also inhibits Ni2+ uptake in the other cell lines, a human monocytic cell line, U937 (Fig. 2d), and a human embryonic kidney cell line, HEK293 (Fig. 2e) were treated with 0.2 mM NiCl2 in the presence of 0.03 mM ZnCl2. Ni2+ content in these cells was increased by NiCl2 treatment, and this increase was reduced by ZnCl2. These findings suggested that Ni2+ uptake occurred generally via a Zn2+-sensitive transporter.

Figure 2
figure 2

Effects of divalent cations and inhibitors on Ni2+ uptake. (a,b, and c): THP-1 cells were treated with 0.2 mM NiCl2 in the presence or absence of 0.03 mM metal chlorides (a), 0.1 mM Chlorazol B (b), or 0.01 mM TAK-242 (c) for 24 h. (d and e): U937 (d) and HEK293 (e) cells were treated with NiCl2 in the presence or absence of 0.03 mM ZnCl2 for 24 h. The Ni2+ uptake of the cells was determined using ICP-MS. The vertical lines represent the S.E.M. of 3 samples. ##p < 0.01 vs. Control, **p < 0.01 vs. 0.2 mM NiCl2.

Full size image

Cellular compartmentalization of Ni ions and the effects of ZnCl2

To confirm whether Ni2+ entered the cells or was bound to the cell membrane, the cellular compartmentalization of Ni2+ was determined by the fluorescence indicator, Newport Green. This compound was used to detect Ni2+ in the immune cells in a previous study21. Although Newport Green could bind to both Zn2+ and Ni2+, the concentration of ZnCl2 used in this experiment, 0.01 mM, did not apparently increase the fluorescence. In contrast, treatment with 0.2 mM NiCl2 increased the fluorescence in the cells, indicating that Ni2+ entered the cells. Consistent with the data of ICP-MS, treatment with ZnCl2 inhibited the NiCl2-induced increase in fluorescence (Fig. 3), indicating that even at a low concentration, Zn2+ inhibited Ni2+ uptake.

Figure 3
figure 3

Detection of Ni2+ in the cells by Newport Green. THP-1 cells were treated with 0.2 mM NiCl2 in the presence or absence of 0.01 and 0.03 mM ZnCl2 for 24 h. Intracellular Ni2+ content was detected with Newport Green. The white scale bar indicates 10 μm.

Full size image

Effects of ZnCl2 and MnCl2 on Ni2+-induced IL-8 production

To clarify whether the inhibition of Ni2+ uptake resulted in the inhibition of IL-8 production, the cells were treated with 0.2 mM NiCl2 in the presence of 0.01 and 0.03 mM ZnCl2 and MnCl2. The increase in the Ni2+ content was reduced by ZnCl2 and MnCl2 in a concentration-dependent manner (Fig. 4a and d). Treatment with ZnCl2 did not affect the Zn2+ content in the cells, but that with MnCl2 increased the Mn2+ content. In these conditions, IL-8 production was also inhibited by these cations (Fig. 4c and f). MnCl2 at 0.03 mM concentration slightly induced IL-8 production by itself, both in the presence and absence of NiCl2 (Fig. 4f), indicating that Mn2+ has a weak ability to induce IL-8 production by itself.

Figure 4
figure 4

Effect of ZnCl2 or MnCl2 on Ni2+ uptake and IL-8 production in THP-1 cells. THP-1 cells were treated with NiCl2 in the presence of 0.01 and 0.03 mM ZnCl2 (a–c) or MnCl2 (d–f) for 24 h and then the amounts of Ni2+ (a and d), Zn2+ (b), and Mn2+ (e) in the cells, and IL-8 in the supernatant (c and f) were determined using ICP-MS and ELISA, respectively. The vertical lines represent the S.E.M. of 3 samples. ##p < 0.01 vs. Control, **p < 0.01 vs. 0.2 mM NiCl2, ††p < 0.01 vs. 0.03 mM MnCl2.

Full size image

Effects of ZnCl2 on CoCl2- and LPS-induced IL-8 production

To confirm the selectivity of the action of ZnCl2, the effects of ZnCl2 on CoCl2- and LPS-induced IL-8 production were examined. Treatment with 0.2 mM CoCl2 increased Co2+ content in the cells and IL-8 production. ZnCl2 (0.01 and 0.03 mM) inhibited this increase in a dose-dependent manner (Fig. 5a and b). In contrast, the same concentrations of ZnCl2 and MnCl2 did not inhibit LPS-induced IL-8 production (Fig. 5c and d), indicating that Zn2+ did not affect the signaling pathway inducing IL-8 expression in this case.

Figure 5
figure 5

Effects of ZnCl2 on Co2+ uptake and IL-8 production induced by CoCl2 and LPS. THP-1 cells were treated with CoCl2 (a and b) and 0.2 μg/ml LPS (c and d) in the presence of 0.01 and 0.03 mM ZnCl2 for 24 h. The amounts of Co2+ (a) in the cells, and IL-8 in the supernatant (b,c, and d) were then determined using ICP-MS and ELISA, respectively. The vertical lines represent the S.E.M. of 3 samples. ##p < 0.01 vs. Control, **p < 0.01 vs. 0.2 mM CoCl2.

Full size image

Enhancement of Ni wire-induced inflammation in a Zn-deficient state

Finally, we examined whether the physiological concentration of Zn2+ affects Ni2+-induced inflammation in low Zn diet-fed mice. Consumption of the low-Zn diet for two weeks reduced Zn2+ levels in the serum to one third of the normal levels (Fig. 6a), but the level in the skin tissues was unchanged (Fig. 6b). As previously reported, implantation of the Ni wire on the back of mice induced inflammation, visible as vasodilation/erythema (Fig. 6c), edema (Fig. 6d), and the expression of inflammatory proteins such as MIP-2 (Fig. 6e) and COX-2 (Fig. 6f). In the mice fed with low-Zn diet for 2 weeks, interestingly, the Ni2+-induced expression of MIP-2 and COX-2 was significantly higher than that in the control group (Fig. 6e and f). The concentration of Ni2+ in the serum and skin tissues was also higher in the low-Zn diet group than in the control group (Fig. 6g and h), indicating that enhanced inflammation promoted Ni2+ elution.

Figure 6
figure 6

Enhancement of Ni2+-induced inflammation in a Zn-deficient state in mice. Mice were fed a low-Zn diet or normal diet for 2 weeks and then an Ni wire was implanted subcutaneously in their dorsa. The mice were sacrificed 0, 8, or 24 h after the implantation. The amounts of Zn2+ in the serum (a) and skin (b) of mice before the implantation were determined using ICP-MS. The skin around the wire was photographed (c) and weighed (d). Ni2+ in the serum (g) and skin (h) were determined using ICP-MS. The expression of MIP-2 (e) and COX-2 (f) was measured by qRT-PCR for the respective times. Values are normalized to those of GAPDH. The vertical lines represent the S.E.M. of the respective values for 3–4 mice. **p < 0.01 vs. 0 h control diet group, #p < 0.05, ##p < 0.01, ###p < 0.001 vs. the corresponding control diet group, ††p < 0.01 vs. 0 h low-Zn diet group.

Full size image

Discussion

In this study, we found that Ni2+ entered the THP-1 cells in a Zn2+, Mn2+, and Co2+-sensitive manner, and that Zn2+ inhibited Ni2+ uptake, resulting in reduced IL-8 production. More importantly, we showed that Ni2+-induced inflammation was enhanced in a systemic low-Zn state. Our findings suggest that maintaining a normal level of Zn2+ is important to reduce the incidence of Ni-induced inflammation and allergy.

As expected, the incubation of THP-1 in the presence of NiCl2, elicited an increase in intracellular Ni2+ level and IL-8 production. The accumulation of Ni2+ in THP-1 cells was induced rapidly until 4 h and then it accumulated gradually. The findings, consistent with those in the previous report18, suggested that the increase was regulated by Ni2+ influx and efflux balance. The increase in Ni2+ level in the cells was antagonized by Zn2+, Mn2+, and Co2+, indicating the involvement of transporter(s) sensitive to these divalent cations. The antagonizing effects of ZnCl2 and MnCl2 were observed at concentrations lower than those of NiCl2, indicating that the affinity of Zn2+ and Mn2+ was much higher than that of Ni2+ to the transporter. The putative transporters were DMT1 and ZIPs. Although DMT1 has an affinity to Ni2+16, it was likely to contribute minimally to Ni2+ uptake in THP-1 cells, because the DMT1 inhibitor, Chlorazol Black22,23, did not decrease Ni2+ uptake. The ZIP family consists of several members and some of them have an affinity to Ni2+24,25,26. All ZIPs except for ZIP12 were expressed in THP-1 cells27, and ZIP225,28, ZIP326, ZIP8, and ZIP1424,29,30 have been shown to have an affinity to Zn2+, Mn2+, Co2+. In addition, ZIPs are known to be induced by the stimulation of TLR431. However, the possibility that Ni2+ induced Zn transporters via the stimulation of TLR4 was rejected, because TAK-242 did not affect the increase in Ni content in the cells incubated for 24 h. These findings suggested that the Ni2+ entered via constitutively expressed ZIP-type transporters. However, because several family members might be involved in Ni2+ uptake and because they have no specific inhibitors, it was difficult to identify the one responsible in this case. We started screening the specific inhibitors of Ni2+ influx to identify the transporter.

We, for the first time, also showed that antagonizing Ni2+ uptake by Zn2+ resulted in the inhibition of IL-8 production. Zn2+ also inhibited Co2+ uptake and Co2+-induced IL-8 production whereas Zn2+ did not inhibit LPS-induced IL-8 production, indicating that Zn2+ did not affect the signaling pathway for IL-8 expression. In contrast, although Mn2+ inhibited Ni2+ uptake, Mn2+ itself induced IL-8 production. These findings were consistent with the observation that Mn2+ as well as Ni2+ could activate HIF-1α12. These findings also suggested that Zn2+ has the ability to attenuate Ni2+ and Co2+-induced inflammation.

The protective effects of Zn2+ at physiological concentrations were also observed in an in vivo model. We had reported that Ni2+ elution from the Ni wire induced inflammatory events, such as neutrophil infiltration and prostaglandin and histamine production5,6, and that the initial inflammatory responses induced further elution of Ni2+5. Using the Ni wire-implanted mouse model, we showed that Ni2+-induced inflammation was enhanced in a Zn-deficient state. Additionally, the mice fed with Zn-deficient diet for 2 weeks showed an enhanced Ni wire-induced expression of MIP-2, a neutrophil chemokine, and COX-2. The elution of Ni2+ was also enhanced, probably via augmentation of the inflammation, as consistent with the previous study. The severe Zn deficiency causes various defects in the function of the skin, such as barrier function. However, in our condition, although Zn2+ concentration in the serum was apparently decreased, that in the skin was unchanged, indicating that functions of the skin were not impaired. Even though the Ni2+ elution and Ni2+-induced cytokine expression were enhanced, this suggested that the concentration of Zn2+ in the serum and/or in the intercellular fluids affected the Ni2+ uptake of leukocytes infiltrated from the blood stream. These results suggested that Ni2+-induced inflammatory cell responses were enhanced in the Zn-deficient state, resulting in the increase in Ni2+ elution. As we focused on the initial responses induced by the uptake of Ni2+, whether the changes in these responses affect the induction of Ni allergy remain to be elucidated. The effects of Zn-deficient condition on Ni allergy are under investigation.

The present in vitro and in vivo findings suggested that Zn2+ modulated Ni2+ uptake and the activation of the inflammatory cells. Our findings also suggested the need to issue a warning that a Zn-deficient state may exacerbate medical device-induced inflammation. A recent report indicated that the prevalence of Zn deficiency in Japanese adult males and females increased with increasing age, and that infants were also susceptible to Zn deficiency32. Therefore, it is important to ascertain whether people with Zn-deficiency are susceptible to Ni allergy, and to determine Zn2+ levels to avoid the induction of Ni-induced inflammation in people implanted with medical devices.

Methods

Nickel chloride (NiCl2), zinc chloride (ZnCl2), cobalt chloride (CoCl2), copper (II) chloride dihydrate (CuCl2·2H2O), iron (II) chloride tetrahydrate (FeCl2·4H2O), magnesium chloride hexahydrate (MgCl2·6H2O), manganese (II) chloride tetrahydrate (MnCl2·4H2O), lipopolysaccharides (LPS) from Escherichia coli O111, and 30% (w/v) H2O2 were purchased from Wako Pure Chemical Industries (Osaka, Japan). Chlorazol Black and TAK-242 were purchased from Sigma-Aldrich Co. (St. Louis, MO) and Calbiochem-Merck Millipore (Darmstadt, Germany), respectively. Newport GreenTM DCF diacetate was purchased from Invitrogen (Carlsbad, CA) and the Ni wire (purity 99.98%, diameter 0.8 mm) from Nilako (Tokyo, Japan). HNO3 (69% (w/w)) was purchased from Kanto Chemical Co., Inc. (Tokyo, Japan).

Cell culture

The human monocytic cell line, THP-1 (Cell Resource Center, Tohoku University) and U937 (JCRB Cell Bank, National Institute of Biomedical Innovation, Health and Nutrition, Japan), and the human epithelial cell line, HEK293 (ATCC, Manassas, VA) were used. Cells were cultured in RPMI 1640 medium (Nissui, Tokyo, Japan) supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS, Biowest, Miami, FL), penicillin G potassium (18 μg/ml), streptomycin sulfate (50 μg/ml), L-glutamine (0.3 mg/ml), and NaHCO3 (1.8 mg/ml), and incubated at 37 °C under a humidified atmosphere containing 5% CO2.

Mice

Four-week-old male ICR mice were purchased from SLC (Shizuoka, Japan). They were fed a standard diet (CE-2, CLEA, Tokyo, Japan) (control diet group, n = 12) or a Zn-deficient diet (CLEA, Tokyo, Japan) (low-Zn diet group, n = 12) for two weeks under a 12-h light/dark cycle in a specific, pathogen-free barrier facility. All animal experiments were approved by the Institutional Animal Care and Use Committee of Tohoku University, and performed in accordance with the Regulations for Animal Experiments and Related Activities at Tohoku University and Guidelines for Proper Conduct of Animal Experiments by the Ministry of Education, Culture, Sports, Sciences, and Technology of Japan.

Treatment of cells with stimulants and inhibitors

NiCl2, ZnCl2, CoCl2, CuCl2, FeCl2, MgCl2, MnCl2, and LPS were dissolved in water. Chlorazol Black and TAK-242 were dissolved in dimethyl sulfoxide. THP-1 cells (5.0 × 105 cells/ml) were seeded into 24-well plates, and stimulated with various concentrations of these reagents. The inhibitors were added with NiCl2.

Implantation of the Ni wire

The Ni wire was cut into 5-mm length, sterilized by ultraviolet irradiation, and then washed with ethanol. Mice were anesthetized using isoflurane (Wako, Osaka, Japan) and then sterilized Ni wires were implanted subcutaneously in their dorsa using a 13 G implant needle (Natsume, Tokyo, Japan). In the control group, mice underwent a similar surgical procedure, but without the implantation of the Ni wire.

ELISA

After incubation of each of the sample, IL-8 in the supernatants was assayed using an ELISA kit (eBioscience, San Diego, CA) according to the manufacturer’s protocol.

Real-time PCR

Total RNA was extracted from the mouse skin tissue surrounding the Ni wire using RNAiso Plus (Takara, Shiga, Japan) according to the manufacturer’s protocol. The total RNA was reverse-transcribed into complementary DNA (cDNA) using the PrimeScript RT reagent kit (Takara, Shiga, Japan). Subsequently, real-time PCR was performed using an SYBR® Premix Ex TaqTM II (Takara, Shiga, Japan) and the Takara PCR Thermal Cycler Dice® real time system (TP800, Takara, Shiga, Japan). The oligonucleotides used for RT-PCR were the following: Mouse GAPDH: (forward) 5′-TGT GTC CGT CGT GGA TCT GA-3′ and (reverse) 5′-TTG CTG TTG AAG TCG CAG GAG-3′, mouse MIP-2: (forward) 5′-CCA CCA ACC ACC AGG CTA CAG GGG C-3′ and (reverse) 5′-AGC CTC CTC CTT TCC AGG TCA GTT AGC-3′, mouse COX-2: (forward) 5′-GAA GTC TTT GGT CTG GTG CCT G-3′ and (reverse) 5′-GTC TGC TGG TTT GGA ATA GTT GC-3′. The normalization and fold changes were calculated using the ΔΔCt method.

Determination of Ni2+, Zn2+, Mn2+, Co2+ concentrations with ICP-MS

THP-1 cells were stimulated by NiCl2 for 24 h in Fig. 1a and b, or for the indicated time in Fig. 1c and d. The cells were stimulated by NiCl2 and/or other metal chlorides for 24 h in Figs 2, 4 and 5. After the incubation, they were collected and washed five times with PBS (phosphate-buffered saline), and then suspended in 150 μl PBS. The cell suspension was sonicated for 30 s and the aliquot was diluted 10-fold with 5% (w/w) HNO3. The concentration of Ni2+ and other metal ions in each sample was determined by Agilent 7500 Series ICP-MS (Agilent Technology, Santa Clara, CA).

To determine the metal concentrations in the mouse skin and serum, circular skin tissue sections (1 cm in diameter) from the region surrounding the Ni wire were excised and the wet weight of skin was measured. The skin tissue sample, approximately 80 mg, was boiled in 3 ml 69% (w/w) HNO3 for 30 min, and then, 300 μl 30% (w/v) H2O2 was added to the samples, on ice. The skin samples were then boiled again for approximately 30 min, and pure water was added to attain a total weight of 10 g. Mouse blood was incubated for 12 h at 4 °C and then centrifuged at 1,200 × g, 4 °C for 30 min. The supernatant was diluted 10-fold with 5% (w/w) HNO3, and centrifuged at 500 × g, 4 °C for 5 min. The supernatant was collected. The Ni2+, Zn2+ concentration of each sample was also determined by ICP-MS.

Bradford determination of protein concentration

The protein contents in the sonicates of cells were determined using the Bio-Rad Protein Assay Dye Reagent Concentrate (Bio-Rad, Tokyo, Japan), according to the manufacturer’s protocol.

Newport green fluorescence staining of intracellular Ni ions

THP-1 cells were stimulated by NiCl2 and/or ZnCl2 for 24 h. After the incubation, the cells were collected and washed five times with 1 × PBS, and then treated for 30 min with 5 μM Newport GreenTM DCF diacetate (Invitrogen, Carlsbad, CA) dissolved in dimethyl sulfoxide. After this treatment, the cells were washed once with 1 × PBS and placed on a Micro Slide Glass (76 × 26 mm, 0.9–1.2 mm thickness, Matsunami-glass, Osaka, Japan), cover-slipped with Fluoromount (DBS, Diagnostic BioSystems, CA). Fluorescence images (excitation at 505 nm and emission at 535 nm) were acquired using a laser scanning confocal microscope LSM 800 (Carl Zeiss, Germany).

Statistical analysis

The statistical significance of the results was analyzed using the unpaired two-tailed Student’s t-test, and the Bonferroni multiple comparison test or Student-Newman-Keuls test for multiple comparisons. For some experiments, a statistical outlier removal was performed using the Smirnov-Grubbs’ rejection test and the Thompson test.