- Email: [email protected]
Expression of the zinc transporter ZIP1 in osteoclasts
Bone 37 (2005) 296 – 304 www.elsevier.com/locate/bone
Expression of the zinc transporter ZIP1 in osteoclasts Mohammed A. Khadeer a,1, Surasri N. Sahu a,1, Guang Bai a, Sunia Abdulla b, Anandarup Gupta a,* a
Department of Biomedical Sciences, Dental School, University of Maryland, 666 West Baltimore Street, Baltimore, MD 21201, USA b Department of Endodontics and Periodontics, Dental School, University of Maryland, Baltimore, MD 21201, USA Received 10 March 2005; revised 8 April 2005; accepted 13 April 2005 Available online 6 July 2005
Abstract Zinc has been previously demonstrated to be a potent inhibitor of osteoclastogenesis and osteoclast function. The mechanisms for cellular uptake of zinc into osteoclasts have not been characterized. We have corroborated previous studies on the reduction of osteoclastogenesis in the presence of extracellular zinc. We demonstrate that osteoclasts express a ubiquitous plasma membrane zinc transporter, namely ZIP1, which was diffusely distributed throughout the cytoplasm. Following an adenoviral-mediated overexpression of ZIP1 in murine osteoclasts, ZIP1 was predominantly colocalized with actin at the sealing zone and significantly inhibited osteoclast function, as assessed by resorptive activity. Finally, overexpression of ZIP1 negatively impacted NF-nB binding activity, as assessed by electrophoretic mobility shift assays. In conclusion, these data both corroborate previous studies on regulation of osteoclast formation and activity by zinc and reveal the presence of a zinc uptake mechanism that exerts an important effect on osteoclast activity. D 2005 Elsevier Inc. All rights reserved. Keywords: Osteoclasts; Zinc; ZIP1; Resorption; NF-nB
Introduction Zinc has been demonstrated to be an essential trace element for normal skeletal growth, and zinc deficiency has been associated with retarded growth [1]. Zinc has been implicated in bone mineralization through regulation of alkaline phosphatase and other metalloenzymes [2,3]. Zinc is fairly abundant in bone; the mean zinc content has been estimated as 110– 300 Ag/kg bone [4]. Intracellular zinc homeostasis is maintained by two distinct families of transporters, Zrt-Irt-like proteins (ZIP) and the cation diffusion facilitator (CDF) family [5]. The ZIP family of transporters mediates uptake of zinc into cells, whereas the CDF family of transporters is involved in zinc efflux and intracellular sequestration. These zinc transporters function through either secondary active transport or facilitated diffusion [6]. * Corresponding author. Fax: +1 410 706 0193. E-mail address: [email protected] (A. Gupta). 1 Both authors contributed equally to this work. 8756-3282/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2005.04.035
Osteoclasts are the principal resorptive cells of bone that are derived from the monocyte-macrophage lineage [7]. Most adult bone diseases, such as osteoporosis and periodontal disease, are due to an imbalance in bone remodeling, which favors resorption and loss of bone mass. Zinc has been shown to have a positive effect on skeletal growth by favoring bone formation [8,9]. Zinc has been well established to act as an inhibitor of osteoclast formation and activity [10,11]. Therefore, zinc has been shown to have a positive effect on skeletal growth by stimulation of osteogenesis, accompanied by a parallel inhibition of osteoclastogenesis. In the current study, using a murine osteoclast culture system, we have corroborated previous studies that have demonstrated inhibition of osteoclast formation in the presence of increasing extracellular concentrations of zinc. Next, we have demonstrated the presence of ZIP1, a ubiquitous plasma membrane zinc transporter, which may be responsible for zinc uptake in osteoclasts. Finally, using a recombinant adenovirus, we have successfully overexpressed ZIP1 in osteoclasts, which we found to exert a significant negative impact on both osteoclast activity and
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
the NF-nB pathway, which is essential for osteoclastogenesis and activity. Our data provide an important insight into the molecular mechanisms by which zinc and zinc transporters can regulate osteoclast formation and function.
Materials and methods Isolation of osteoclast-like cells from bone marrow The tibiae and femurs of 7-week-old mice were used to isolate bone marrow stromal cells, as previously described [12]. The protocol (04-08-03) for these animal studies has been approved by the IACUC at the University of Maryland, Baltimore. Bone marrow cells were suspended in aminimal essential medium (a-MEM, Gibco-BRL) supplemented with 10% fetal bovine serum (designated as a-10 MEM) and cultured at 37-C in a 5% CO2 incubator. After 24 h, non-adhered cells were layered on Histopaque-1077 (Sigma, St. Louis, MO) and centrifuged at 300 g for 15 min at room temperature (RT). The cell layer between the Histopaque and the medium was removed and washed with a-10 MEM, centrifuged at 2000 rpm for 7 min. Both recombinant M-CSF-1 (R&D Systems, Minneapolis, Minnesota) and RANKL were added to the cultures at concentrations of 10 ng/ml and 100 ng/ml, respectively, as previously described [12]. Multinucleated osteoclasts were seen to form and mature between days 4 –5. The percentage of tartrate-resistant acid phosphatase (TRAP)-positive cells was assessed to be ¨99%, using the TRAP+ assay, as previously described [12]. RNA isolation and RT-PCR for ZIP1 Total RNA was extracted from mature osteoclasts using TRIZOL Reagent (Invitrogen) as previously described [12]. Using reverse transcriptase (Invitrogen, Carlsbad, CA), cDNA was synthesized and used as a template to amplify the murine ZIP1 fragment. Primers for amplification of murine ZIP1 were as follows: mZIP1 (forward, F): 5V-CCAGAGCTCCTGGTGTGGCG-3V, and the hZIP1 (reverse, R): 5V-GCAGCAGTCCAGGAGACAA-3V. The conditions for PCR-amplification of mZIP1 were as follows: an initial denaturation at 94-C for 4 min followed by 35 cycles of 94-C for 1 min, 50-C for 1 min, and 72-C for 1 min. A final extension was done for 10 min at 72-C. The expected size for the mZIP1 amplicon was 230 kb. Western blot analysis Murine osteoclasts were used after day 5 in culture for preparation of lysates. Following two quick rinses with icecold phosphate-buffered saline (PBS), the cells were lysed in a buffer containing 10 mM Tris – HCl, pH 7.05, 50 mM NaCl, 0.5% Triton X-100, 30 mM sodium pyrophosphate, 5 mM NaF, 0.1 mM Na3VnO4, 5 mM ZnCl2, and 2 mM
297
PMSF. Lysates were pelleted by centrifugation at 15,000 rpm, for 15 min, at 4-C. The supernatant was transferred into a fresh microfuge tube and held on ice. Protein concentrations were measured using the Bio-Rad protein assay reagent kit (Bio-Rad, Hercules, CA), so that equal amounts of protein were loaded per lane (¨25 Ag) and were analyzed by SDS-PAGE on 8% gels. The gel was then electroblotted onto a PVDF membrane by wet – dry transfer. Nonspecific protein binding was blocked with 5% nonfat dry milk powder dissolved in PBS containing 0.1% Tween-20. The blots were incubated overnight with a 1:10,000 dilution of the primary antibody (chicken IgY) to the ZIP1 protein. This was followed by detection with a horseradish-peroxidase (HRP)-conjugated secondary goat anti-rabbit antibody (Sigma, 1:2000 dilution). The secondary antibody (HRP-conjugated goat anti-rabbit antibody) was detected by enhanced chemiluminescence, as previously described [13,14]. Indirect immunofluorescence The subcellular distribution of ZIP1 was examined in osteoclasts by indirect immunofluorescence. Osteoclasts were rinsed twice in ice-cold PBS, fixed with 4% paraformaldehyde, and permeabilized with PBS containing 0.2% Triton X-100. The cells were blocked with PBS containing 3% bovine serum albumin for 2 h at room temperature (RT). The cells were incubated for 2 h in a 1:500 dilution of affinity-purified chicken IgY directed against ZIP1 at room temperature. The primary antibodies were detected with Alexa-488-conjugated secondary antibodies (1:1000 dilution, Molecular Probes). Actin was labeled with rhodamine phalloidin. As negative controls, osteoclasts were incubated for 2 h in a 1:500 dilution with the non-immune serum (NIS, IgY). Confocal microscopy (Carl Zeiss LSM Meta 510) was performed; all images shown were taken with either a 63 or 100 oil-immersion objective lens. Construction of recombinant adenoviral vectors for ZIP1 Recombinant adenovirus for ZIP1 was generated as follows and according to methods previously described [15]. Briefly, the gene for ZIP1 was ligated into pShuttle-IREShrGFP2 (AdEasy-XL Adenoviral Vector system; Stratagene, La Jolla, CA) using the NheI– XhoI multiple cloning sites. This vector contains the CMV promoter and a dicistronic expression cassette in which the multiple cloning site (MCS) is followed by the EMCV-IRES, which directs translation of a humanized recombinant green fluorescent protein (hrGFP). This allows the expression of the gene of interest to be monitored at the single-cell level due to expression of the hrGFP on the same transcript. Recombinant Adeno-ZIP1-IRES-hrGFP2 was confirmed by digestion with PacI. High titer (2 1010 pfu/ml) viral stocks were made as follows: approximately 15 Ag of shuttle
298
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
plasmid and 5 Ag of pVQAd 9.2-100 backbone plasmid were digested with PacI. Both linear fragments were cotransfected into HEK293 cell line in a 60 mm plate at ¨50% confluency by standard CaCl2 method. The initial viral lysate was amplified in HEK293 cells, and the final viral lysate purified over two rounds of CsCl gradient ultracentrifugation. Virus particles were dialyzed against 3% sucrose/PBS buffer, diluted to 1 1012 particles/mL, and then frozen at 80-C. We assayed for infectious particle concentration by a cell-based plaque assay on HEK293 cells. Presence of replication competent adenovirus was checked by plaquing on the wild type permissive cell line A549 for at least 14 days. The adenoviral titer for both adenoviruses was ¨1011 pfu/ml. Recombinant adenovirus expressing EGFP (rAd-EGFP) was prepared as previously described [16] and served as an irrelevant negative control for infection of osteoclasts. Assessment of intracellular zinc Murine osteoclasts were incubated in the zinc-sensitive dye Zinquin ester (20 AM, Biotium, CA) for 30 min and subsequently rinsed in PBS. The cells were viewed under a Nikon TE200 inverted fluorescence microscope and UV filters (absorption 364 nm, emission 485 nm), as previously described [14]. In vitro resorption assays Following transfection or infection of murine osteoclasts with plasmid containing ZIP1 or rAd-ZIP1 for 48 h, respectively, mature murine osteoclasts were plated on CaPO4 (1 Am)-coated quartz discs (Osteologici discs, BD Biosciences). Approximately 48 h later, the effects on osteoclast-mediated resorption were assessed by rinsing the discs in distilled water to lyse the osteoclasts, as previously described [12]. The extent of resorption could be assessed by clear areas on the quartz discs where the CaPO4 film had disappeared. The discs were photographed (20 objective), and the area and number (of pits) were counted using SigmaScan (SPSS Inc., Chicago, IL). Statistical analyses were performed using unpaired t tests, and significance of differences between means was assessed at a minimum P < 0.05. Assay for NF-jB activity in osteoclasts Electrophoretic mobility shift assays (EMSAs) were performed following a protocol described previously [16,17]. Briefly, the osteoclast cultures were rinsed twice with ice-cold PBS. The cells were scraped into hypotonic lysis buffer A (10 mM HEPES [pH 7.8], 10 mM KCl, 1.5 mM MgCl2, 0.5 mM dithiothreitol, plus a protease inhibitor cocktail). Following incubation on ice for 15 min, NP-40 was added to a final concentration of 0.6%. The nuclei were pelleted by centrifugation at 10,000 g for 5 min. The
cytosolic fraction was removed, and the nuclei were resuspended in nuclear extraction buffer B (20 mM HEPES [pH 7.8], 420 mM NaCl, 1.2 mM MgCl2, 0.2 mM EDTA, 25% glycerol, 0.5 mM DTT, plus a protease inhibitor cocktail), vortexed briefly, and rocked for 30 min in 4-C. The insoluble material was removed by centrifugation for 5 min at 12,000 g. Finally, the nuclear proteins in the supernatant quantitated using the Bio-Rad protein assay. The nuclear proteins were dialyzed against a dialysis buffer (20 mM HEPES-Cl [pH 7.9], 0.2 mM EDTA, 100 mM KCL, 20% glycerol, 0.2 mM PMSF, 0.5 mM DTT). After preparation of nuclear proteins, EMSA was performed as follows. A probe for the NF-nB binding site was produced with DNA GB469 (upper strand, 5V-TCTGAGGAGTGGGACTGGCCAGCAT) and GB460 (lower strand, 5V-GCCTAATGCTGGCCAGTCCCACTCC). The oligonucleotide for NF-nB was labeled by Klenow enzyme in the presence of [a-32P]dCTP (>3000 mCi/mmol, ICN Biomedicals, Irvine, CA). The double stranded oligonucleotide probes for NF-nB were end-labeled with [g32P] dATP and T4 DNA polymerase kinase, annealed at 37-C for 30 min, and purified by running through a G-50 Column. Nuclear extracts (2.5 Ag) were incubated in binding buffer (200 mM HEPES, pH 8.3, 600 mM KCl, 50 mM DTT, and 10 Ag/Al BSA), poly (dI-dC), and Ficoll for 18 min at room temperature (RT). The end-labeled probe was added to this mixture and further incubated at RT for 25 min. Protein– DNA complexes were separated in a 4% non-denaturing polyacrylamide gel. Gels were dried and analyzed using a phosphorimager (Typhoon, Amersham BioSciences) or autoradiographic film. For supershift assays to identify components in the DNA – protein complex, the nuclear extracts were preincubated with antibodies (2 Ag) specific for p65, p50, C-Rel, and p52; an antibody specific for Sp1 was used as a negative control, as previously described [16].
Results Effects of extracellular zinc on osteoclastogenesis In order to test for the effects of zinc on osteoclast formation, osteoclasts were generated from murine bone marrow. Osteoclast precursors were incubated for 6 days either in the absence of, or presence of, increasing concentrations of ambient ZnCl2 (1, 5, and 10 AM), as shown in Fig. 1 (panel A, 1). The formation of osteoclast-like cells was evaluated by staining for TRAP+ cells, as previously described [12]. The total zinc content in culture media prior to supplementation with ZnCl2 was determined to be 3 AM by atomic absorption spectrometry, as previously described [18]. There was no indication of cellular toxicity or loss of cell viability associated with supplementation of culture medium with graded increases in ZnCl2 (ranging from 0 –10 AM), as shown in Fig. 1 (panel A, 2). The optimal concentration of extracellular ZnCl2 supplementation of culture medium for
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
299
Fig. 1. Effects of extracellular zinc on osteoclastogenesis and expression of ZIP1 in osteoclasts. (A) Changes in extracellular zinc and effects on osteoclastogenesis. (A,1) Multinucleated murine osteoclasts were generated from osteoclast precursor cells starting day 1 (of culture) in the presence of graded concentrations of extracellular ZnCl2 (1, 5, and 10 AM ZnCl2). The number of TRAP+ multinucleated osteoclasts (OC) was assessed on day 6 of culture (n = 12, P < 0.05). (A, 2) Representative images of murine osteoclasts in culture that were incubated in the presence of graded supplementation of (culture) medium with ZnCl2 (ranging from 0 – 10 AM). The basal level of zinc present in culture medium prior to supplementation with ZnCl2 was determined to be ¨3 AM using atomic absorption spectrometry. (B) Expression of ZIP1 in osteoclasts. Using RT-PCR, a ZIP1 fragment of 230 bp from murine osteoclasts was amplified from a human osteoclast (OC) cDNA library (lane 1) and primary murine osteoclasts (OC, lane 2). Specificity of species-specific (human, mouse) primers for amplification of ZIP1 was demonstrated by absence of transcript when the cDNA templates were switched (lanes 3 – 4). The full-length ZIP1 cDNA template cloned from a human prostate cancer cell line, PC3 cells served as a positive control (lane 5). Sequence homology between the human and murine ZIP1 transcripts revealed ¨82% nucleotide identity. (C) ZIP1 protein expression in osteoclasts. The ZIP1 protein was detected in human mesenchymal stem cells (MSC) and murine osteoclasts (mOC) with an affinity-purified ZIP1-specific chicken IgY.
significant inhibition (45 –55%) of osteoclastogenesis was determined to be ¨5– 10 AM zinc (n = 12, P < 0.05). Expression of ZIP1 in osteoclasts We next asked the question whether a ubiquitous zinc transporter ZIP1 is expressed in osteoclasts. Previous
studies have shown that the ZIP1 family of zinc transporters is expressed in bone marrow [19]. Using reverse transcriptase-polymerase chain reaction (RT-PCR), a ZIP1 fragment of 230 bp from murine osteoclasts was amplified, as shown in Fig. 1 (panel B); a similar transcript for ZIP1 was also amplified from a human osteoclast cDNA library (panel B). The positive control used for the PCR amplifi-
300
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
cation was the full-length ZIP1 cDNA template cloned from a human prostate cancer cell line, PC3 cells [14]. The sequence homology between the human and murine ZIP1 transcripts revealed ¨82% nucleotide identity. In addition, the ZIP1 protein could be detected in both human mesenchymal stem cells (MSCs) and murine osteoclasts (mOC), as shown in Fig. 1 (panel C). ZIP1 is a 324 amino acid protein with a predicted molecular mass of ¨34 kDa [14] but can be seen as a dimer, presumably because of a leucine zipper motif [14]. In both the human osteoblast and murine osteoclast, the ZIP1 protein can be detected at ¨68 kDa. Expression, overexpression, and cellular distribution of ZIP1 in murine osteoclasts We have examined the developmental expression of ZIP1 in murine osteoclasts, as shown in Fig. 2 (panel A). The protein levels of ZIP1 were not appreciably changed during the transition from pre-osteoclast precursors (pOC,
days 1 –4) to multinucleated osteoclasts (mnOC, days 5– 6). A recombinant adenovirus was generated for ZIP1 for overexpression in murine osteoclasts, as shown in Fig. 2 (panel B). Murine osteoclasts were infected with recombinant adenoviruses (rAd, MOI of ¨100) on day 5 of culture when the majority of cells were multinucleated osteoclasts. First, 48 h later, overexpression of ZIP1 was confirmed by RT-PCR for ZIP1, as shown in Fig. 2 (panel B). Compared to uninfected control (Con) or rAd-EGFP-infected osteoclasts, there was at least a 4-fold increase in the gene expression of ZIP1 following infection rAd-ZIP1; the transcript for GAPDH was amplified for normalization of the ZIP1 signal, as shown in Fig. 2 (panel B). Second, the increase in gene expression was followed by Western blot analysis for the ZIP1 protein (panel C), which paralleled the increase in ZIP1 gene expression. Third, we examined the subcellular distribution of ZIP1 in murine osteoclasts, before and after overexpression of ZIP1, as shown in Fig. 2 (panel D). Murine osteoclasts were infected
Fig. 2. Expression, overexpression, and cellular distribution of ZIP1 in murine osteoclasts. (A) Developmental expression of ZIP1 in murine osteoclasts. The protein levels of ZIP1 during the transition from pre-osteoclast precursors (pOC, days 1 – 3) to multinucleated osteoclasts (mnOC, days 5 – 6) were determined by Western blot analysis. Protein levels of h-actin were used to normalize the signal for ZIP1. (B) Overexpression of the ZIP1 gene in osteoclasts. A recombinant adenovirus was generated for ZIP1 for overexpression in murine osteoclasts. Murine osteoclasts were infected with recombinant adenoviruses (rAd, MOI of ¨100) on day 5 of culture when the majority of cells were multinucleated osteoclasts. First, 48 h later, overexpression of ZIP1 was confirmed by RT-PCR for ZIP1, compared to uninfected control (Con) or rAd-EGFP-infected osteoclasts. The transcript for GAPDH was amplified for normalization of the ZIP1 signal. (C) Overexpression of ZIP1 protein in osteoclasts. Following infection of murine osteoclasts with either rAd-EGFP or rAd-ZIP1, Western blot analysis with a ZIP1-specific chicken IgY was performed. (D) Subcellular distribution of ZIP1 in osteoclasts. Murine osteoclasts were infected either with a replication-deficient adenovirus encoding for the ZIP1 gene (rAd-ZIP1, frame 3) or for EGFP (rAd-EGFP, CON, frame 2). Immunostaining for ZIP1 was performed with a polyclonal chicken IgY for ZIP1, and actin was stained with rhodamine phalloidin. Osteoclasts were also incubated with non-immune serum (NIS, IgY, frame 1) to demonstrate lack of an apparent immunofluorescent signal for ZIP1, using the same settings for confocal microscopy (i.e., pinhole aperture, detector gain, and sensitivity). (E) Accumulation of intracellular zinc in ZIP1-overexpressing osteoclasts. Murine osteoclasts were infected with rAdZIP1 for 48 h and examined separately for EGFP-derived fluorescence from the reporter gene as an estimate of infection efficiency (frame 2) and for accumulation of intracellular zinc with Zinquin ester, a zinc-sensitive dye (frame 4). The images obtained from rAd-ZIP1-infected osteoclasts were compared to EGFP (frame 1)- and Zinquin ester (frame 3)-derived fluorescence in uninfected osteoclasts.
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
either with a replication-deficient adenovirus encoding for the ZIP1 gene (rAd-ZIP1, frame 3) or for EGFP (rAd-EGFP, frame 2). Immunostaining for ZIP1 was performed with a polyclonal chicken IgY for ZIP1 (Alexa 488, green) and actin was stained with rhodamine phalloidin (red). In rAdEGFP-infected osteoclasts, which served as controls (CON, frame 2), the signal for the native ZIP1 was mostly absent near the periphery, mostly distributed toward the cell interior, and showed very little colocalization existed with actin. However, after overexpression of ZIP1 (rAd-ZIP1, frame 3), the immunofluorescent signal for ZIP1 was discretely localized at the plasma membrane and showed significant colocalization with actin. The indirect immunofluorescence for the cellular distribution of ZIP1 was deemed to be specific since there was no apparent immunofluorescent signal for ZIP1 when osteoclasts were incubated with the non-immune chicken serum (NIS, IgY, frame 1), using identical settings for confocal microscopy (i.e., pinhole aperture, detector gain, and sensitivity). Finally, we examined the accumulation of intracellular zinc in osteoclasts that were overexpressing the ZIP1 protein, as shown in Fig. 2 (panel E). Similar to experiments detailed above, murine
301
osteoclasts were infected with rAd-ZIP1 for 48 h and examined separately for EGFP-derived fluorescence from the reporter gene as an estimate of infection efficiency (panel E, frame 2) and for accumulation of intracellular zinc with Zinquin ester, a zinc-sensitive dye (panel E, frame 4). The images obtained from rAd-ZIP1-infected osteoclasts were compared to uninfected osteoclasts that showed very low levels of EGFP-derived epifluorescence (panel E, frame 1) and low levels of zinc accumulation (panel E, frame 3). Effects of ZIP1 overexpression on osteoclast function We asked the question whether overexpression of ZIP1 would impact osteoclast function. We hypothesized that an overexpression of ZIP1 would increase the uptake of zinc into the osteoclast since the immunofluorescent data indicated that the ZIP1 protein was primarily localized at the plasma membrane in ZIP1-overexpressing osteoclasts (Fig. 2, panel D, frame 3). Mature murine osteoclasts were infected with either rAd-EGFP (control) or rAd-ZIP1 for 48 h. Subsequently, the infected osteoclasts were replated on CaPO4-coated Osteologici discs for 24 h in vitro resorp-
Fig. 3. Effects of ZIP1 overexpression on osteoclast function. (A) In vitro resorption assay. Murine osteoclasts were infected with either rAd-EGFP (control) or rAd-ZIP1 for 48 h. Subsequently, the infected osteoclasts were replated on CaPO4-coated Osteologici discs for resorption assays in the presence of 10 AM ZnCl2, which was added to the culture medium for 24 h. (A, a) Representative images of in vitro assays for osteoclast-mediated resorption in the absence of supplementation of culture medium with 10 AM ZnCl2 (a, 1), or supplementation with 10 AM ZnCl2 alone (a, 2), or following overexpression of ZIP1 in the absence of supplementation with 10 AM ZnCl2 (a, 3), or following overexpression of ZIP1 in the presence of supplementation with 10 AM ZnCl2 (a, 3). (A, b) The corresponding resorption areas (b, 1) were photographed and quantified using SigmaScan (b, 2). (B) Comparison of changes in resorptive activity by ZIP1overexpressing osteoclasts. Data obtained from quantitation of resorption areas were plotted to reflect changes in resorptive activity (after 24 h of being plated on CaPO4-coated Osteologici discs and represented as ‘‘percentage of area resorbed by controls’’) under the following conditions: control untreated osteoclasts (i.e., 0 AM ZnCl2), osteoclasts cultured in the presence of 10 AM ZnCl2, rAd-EGFP-infected osteoclasts (rAd-EGFP,10 AM ZnCl2), and ZIP1overexpressing osteoclasts (rAd-ZIP1,10 AM ZnCl2). Statistical comparisons were performed with unpaired t tests, and the P values are reported in the Results section.
302
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
tion assays, according to protocols previously published [12,20], in the presence of 10 AM ZnCl2. Representative images of osteoclast-mediated in vitro resorption under various experimental conditions that include control untreated osteoclasts (0 AM ZnCl2, A, a, 1), supplementation of culture medium with 10 AM ZnCl2 (A, a, 2), overexpression of ZIP1 alone (A, a, 3), or overexpression of ZIP1 in the presence of 10 AM ZnCl2 (A, a, 4) are shown in Fig. 3 (panel A, a, 1– 4). Next, the corresponding representative resorption areas were photographed (panel A, b, 1) and quantified using SigmaScan (panel A, b, 2), as described in Materials and methods. Our data indicate that resorption was not significantly altered either in the presence of 10 AM ZnCl2 or by rAd-EGFP-infected osteoclasts (in the presence of 10 AM ZnCl2), compared to control untreated osteoclasts, as shown in Fig. 3 (panel B). However, there was a dramatic reduction in resorption by ZIP1-overexpressing osteoclasts (rAd-ZIP1, 10 AM ZnCl2, 28.5 T 4.40, P = 0.0001), compared to either untreated controls (0 AM ZnCl2, 100 T 9.80) or osteoclasts cultured in the presence of 10 AM ZnCl2 (75.3 T 7.14, P = 0.0005) or rAdEGFP-infected osteoclasts (10 AM ZnCl2, 28.5 T 4.41, P = 0.0004). These data indicate that, first, the presence of 10 AM ZnCl2 in the culture medium is not sufficient to significantly inhibit resorptive activity by mature osteoclasts, and second, the inhibition of resorptive activity in
ZIP1-overexpressing osteoclasts is specific for the ZIP1 protein since there was no impact on resorption by osteoclasts which were infected with the recombinant adenovirus encoding for EGFP (rAd-EGFP). Assay for changes in NF-jB activity following overexpression of ZIP1 in osteoclasts NF-nB is essential for osteoclast formation and activity [17,21,22]. It is known that NF-nB sites interact with dimerized NF-nB factors that include p65, p50, p52, C-Rel, and Rel B. Homo- or heterodimers of NF-nB factors are associated with InB factor in the cytoplasm and remain inactive. In response to activating signals, InB dissociates from the complex, thus releasing NF-nB dimers, which become activated and subsequently translocate to the nucleus for regulation of transcription [23]. We asked the question whether reduced osteoclast activity following overexpression of ZIP1 is associated with reduced NF-nB activity. Following infection of pre-osteoclast precursors (on day 4 of culture) with either rAd-EGFP or rAd-ZIP1 for 72 h, nuclear extracts were prepared, and an electrophoretic mobility shift assay (EMSA) was performed using 32Plabeled NF-nB oligonucleotide. As shown in Fig. 4, NF-nB activity was decreased as evidenced by decreased DNAbinding activity (panel A, lane 3), compared to either
Fig. 4. Assay for changes in NF-nB activity in ZIP1-overexpressing osteoclasts. (A) Electrophoretic mobility shift assay (EMSA) for NF-nB activity. Following infection of pre-osteoclast precursors (on day 4 of culture) with either rAd-EGFP or rAd-ZIP1 for 72 h, nuclear extracts were prepared. An Ig/HIV NF-nB consensus sequence (22 bp) was end-labeled with 32P and used as probe in EMSA experiments. NF-nB binding activity in ZIP1-overexpressing osteoclasts (lane 3), compared to either control untreated osteoclasts (lane 2), rAd-EGFP-infected osteoclasts, or TNF-a treated osteoclasts (10 ng/ml, 24 h, lane 5). As positive controls for the assay, nuclear extracts were prepared from P19 cells, an embryonic carcinoma cell line, and different protein concentrations were tested (lanes 6 – 7, P19-1, P19-2). (B) Constitutive presence of NF-nB protein (p50) in nuclei of osteoclasts (antibody supershift assay). Nuclear proteins were extracted from osteoclasts, and nuclear proteins were preincubated with indicated antibodies (specific for p65, p50, C-Rel, p52, or Sp1) individually before the probe was added. Nuclear proteins were extracted from undifferentiated P19 cells. For antibody supershift assays, the nuclear proteins were preincubated with antibody specific for p65, p50, C-Rel, p52, or Sp1. Band B was supershifted by p50 antibody to form band C. Nuclear proteins extracted from osteoclasts formed two protein – DNA complexes with the NF-nB consensus, as shown by an antibody (a)-induced supershift of the NF-nB factor, p50. The upper slowly migrating band (B) contained the NF-nB factor, p50, as shown by antibody supershift lane 3; the band B was supershifted by the p50 antibody (ap50) to form band C. The antibody supershift experiments revealed the newly translocated p65 factor (ap65, band D). In the same assay, the Sp1 antibody served as a control for non-related factors. The fast migrating band (A) was not interfered with by any of the antibodies tested, and its identity remains unknown.
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
control untreated osteoclasts (CON, lane 2), rAd-EGFPinfected osteoclasts, or TNF-a treated osteoclasts (10 ng/ml, 24 h, lane 5). Nuclear extracts were also prepared from P19 cells, an embryonic carcinoma cell line, and different protein concentrations were tested (lanes 6 – 7, P19-1, P19-2); these nuclear extracts served as additional positive controls for the NF-nB activation assay, as previously described [16]. In order to validate the results of the EMSA and clarify NF-nB factor identity, we performed an antibody supershift assay, as shown in Fig. 4 (panel B). Nuclear proteins were extracted from osteoclasts, and nuclear proteins were preincubated with indicated antibodies (specific for p65, p50, C-Rel, p52, or Sp1) individually before the probe was added. We show that nuclear proteins extracted from osteoclasts formed two protein – DNA complexes with the NF-nB consensus (Fig. 4, panel B), as shown by an antibody (a) supershift of the NF-nB factor, p50. The upper slowly migrating band (B) was supershifted by the p50 antibody (ap50) to form band C (lane 3), indicating that band B contains p50. Furthermore, band B was completely supershifted by the p65 antibody (ap65, band D). This complete supershift suggests that p65 involves formation of all DNA –protein complexes in band B. In comparison, the migration or intensity of band B was not changed by other antibodies tested, including one against Sp1 as a control for non-related factors, as previously described [16]. These data demonstrate that band B is composed of a heterodimer of p50 and p65 and/or a homodimer of p65 with the NF-nB consensus. Considering that overexpression of ZIP1 (in murine osteoclasts) significantly reduced band B, we believe that the heterodimer of p50 and p65 and/or a homodimer of p65 in osteoclasts mediates initiation of the NF-nB signaling pathway. Finally, we noticed that there is a fast migrating band A, which was not interfered with by any of the antibodies tested, and its identity remains unknown, as previously reported [16].
Discussion The role of zinc in the preservation of bone mass by stimulating bone formation and inhibiting bone resorption has been well established [24]. The zinc content in adult bone is fairly high compared to other tissues [25,26]. We have hypothesized that, during bone resorption, osteoclasts may be exposed to gradual increases in zinc, which upon uptake into cells through previously unidentified mechanisms, can inhibit osteoclast function. Previously, several studies have clearly established that zinc and zinc-chelating compounds (10 6 to 10 4 M) can inhibit in vitro osteoclastogenesis and osteoclast activity [10,11,27 – 31]. We have corroborated these earlier studies by demonstrating that, indeed, exposure of murine osteoclast precursors to graded increases in (ambient) zinc concentration inhibits osteoclastogenesis, with an optimal extracellular zinc concentration of 10 AM.
303
Having established the paradigm for zinc-mediated inhibition of osteoclast formation, we hypothesized that osteoclasts must possess mechanisms or transporters for zinc uptake. In most cells, the intracellular zinc homeostasis is maintained by a combination of several uptake, sequestration, and efflux mechanisms [5]. We have demonstrated gene and protein expression of ZIP1, a plasma membrane zinc transporter and a member of the expanding ZIP superfamily of zinc transporters [19]. We hypothesize that ZIP1 is the major plasma membrane zinc transporter in osteoclasts; we have screened for gene expression of several other members of the ZIP superfamily but have not been able to detect transcripts for other ZIP proteins (data not shown). In the murine osteoclast, the cellular distribution of ZIP1 was largely confined to a diffuse cytoplasmic pool, but when overexpressed, was discretely localized at or near the plasma membrane and was colocalized with the actin ring, which delineates the sealing zone [32]. In the current study, we have not determined whether the cellular distribution of ZIP1 is altered in response to increases or depletion of extracellular zinc levels, as has been previously reported for ZIP1 in other cell types [33]. It has been demonstrated that some members of the ZIP family, namely ZIP1 and ZIP3, can be found in intracellular organelles in zincreplete medium but are capable of trafficking between the plasma membrane and intracellular compartments, mostly a consequence of changes in endocytic rates. Zinc deficiency increased plasma membrane levels of these zinc transporters by decreasing their rates of endocytosis, providing evidence of a conserved mechanism controlling their activity. We have found that an adenovirus-mediated overexpression of ZIP1 led to a profound inhibition of mature osteoclast activity, as determined by in vitro resorption assays. These data indicate that increased surface expression of ZIP1 may enhance the total uptake of zinc into the osteoclast. In future studies, we will determine the kinetics of ZIP1 in osteoclasts, both in native and ZIP1overexpressing osteoclasts. Previous studies in other cell types have established that, generally, overexpression of zinc transporters increases V max without having any effect on the affinity of the transporter for zinc [6,34,35]. In a very recent study, it has been indirectly suggested that zinc negatively impacts osteoclast differentiation by inhibiting the NF-nB pathway in osteoclast precursors [31]. When murine bone marrow stromal cells were cultured in the presence of osteoclastogenic cytokines (such as LPS, TNF-a, and RANKL) in the presence of zinc sulfate, only TNF-a and RANKL-induced osteoclastogenesis was significantly inhibited. This inhibition was relieved in the presence of transcription and translation inhibitors. The authors surmised that zinc affected its inhibition on osteoclastogenesis through the NF-nB pathway that is specifically stimulated by both TNF-a and RANKL [21,36,37]. We have extended these preliminary studies by
304
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
demonstrating that adenovirus-mediated overexpression of the ZIP1 protein in osteoclast precursors led to a very significant decrease in NF-nB binding activity, as determined by electrophoretic mobility shift assays. These data provide the first direct evidence for a zinc transporter negatively impacting osteoclast differentiation and activity. In conclusion, these data provide a preliminary characterization of ZIP1, a ubiquitous zinc transporter, which is also expressed in the osteoclast, and extend previous studies of the negative impact of zinc on osteoclast differentiation and function. Finally, our data also provide the first evidence of a direct inhibition of the osteoclastogenic NF-nB pathway by overexpression of ZIP1 in osteoclast precursors. These studies add new information on the underlying molecular mechanisms by which zinc may impact osteoclast formation and activity.
Acknowledgments We acknowledge the help of Bryan Kitahara, University of Michigan, Ann Arbor, MI, who was supported on a Short Term Research Training Grant for Dental Students T35DE07334-05. This work was partly supported by an NIH grant AR 44792 (to A.G.), and NS 38077 (to G.B.).
References [1] Nishi Y. Zinc and growth. J Am Coll Nutr 1996;15:340 – 4. [2] McCall KA, Huang C, Fierke CA. Function and mechanism of zinc metalloenzymes. J Nutr 2000;130:1437S – 46S. [3] Chesters JK. Biochemical functions of zinc in animals. World Rev Nutr Diet 1978;32:135 – 64. [4] Gomez S, Rizzo R, Pozzi-Mucelli M, Bonucci E, Vittur F. Zinc mapping in bone tissues by histochemistry and synchrotron radiationinduced X-ray emission: correlation with the distribution of alkaline phosphatase. Bone 1999;25:33 – 8. [5] Guerinot ML. The ZIP family of metal transporters. Biochim Biophys Acta 2000;1465:190 – 8. [6] Liuzzi JP, Cousins RJ. Mammalian zinc transporters. Annu Rev Nutr 2004;24:151 – 72. [7] Teitelbaum SL. Bone resorption by osteoclasts. Science 2000;289: 1504 – 8. [8] Calhoun NR, Smith Jr JC, Becker KL. The role of zinc in bone metabolism. Clin Orthop 1974;0:212 – 34. [9] Saltman PD, Strause LG. The role of trace minerals in osteoporosis. J Am Coll Nutr 1993;12:384 – 9. [10] Kishi S, Yamaguchi M. Inhibitory effect of zinc compounds on osteoclast-like cell formation in mouse marrow cultures. Biochem Pharmacol 1994;48:1225 – 30. [11] Moonga BS, Dempster DW. Zinc is a potent inhibitor of osteoclastic bone resorption in vitro. J Bone Miner Res 1995;10:453 – 7. [12] Gupta A, Lee BS, Khadeer MA, Tang Z, Chellaiah M, Abu-Amer Y, et al. Leupaxin is a critical adaptor protein in the adhesion zone of the osteoclast. J Bone Miner Res 2003;18:669 – 85. [13] Khadeer MA, Tang Z, Tenenhouse HS, Eiden MV, Murer H, Hernando N, et al. Na+-dependent phosphate transporters in the murine osteoclast: cellular distribution and protein interactions. Am J Physiol: Cell Physiol 2003;284:C1633 – 44.
[14] Franklin RB, Ma J, Zou J, Guan Z, Kukoyi BI, Feng P, et al. Human ZIP1 is a major zinc uptake transporter for the accumulation of zinc in prostate cells. J Inorg Biochem 2003;96:435 – 42. [15] Anderson RD, Haskell RE, Xia H, Roessler BJ, Davidson BL. A simple method for the rapid generation of recombinant adenovirus vectors. Gene Ther 2000;7:1034 – 8. [16] Liu A, Hoffman PW, Lu W, Bai G. NF-nB site interacts with Sp factors and up-regulates the NR1 promoter during neuronal differentiation. J Biol Chem 2004;279:17449 – 58. [17] Abu-Amer Y. IL-4 abrogates osteoclastogenesis through STAT6dependent inhibition of NF-nB. J Clin Invest 2001;107:1375 – 85. [18] Costello LC, Liu Y, Franklin RB, Kennedy MC. Zinc inhibition of mitochondrial aconitase and its importance in citrate metabolism of prostate epithelial cells. J Biol Chem 1997;272:28875 – 81. [19] Gaither LA, Eide DJ. The human ZIP1 transporter mediates zinc uptake in human K562 erythroleukemia cells. J Biol Chem 2001; 276:22258 – 64. [20] Gupta A, Guo XL, Alvarez UM, Hruska KA. Regulation of sodiumdependent phosphate transport in osteoclasts. J Clin Invest 1997;100: 538 – 49. [21] Abu-Amer Y, Tondravi MM. NF-nB and bone: the breaking point. Nat Med 1997;3:1189 – 90. [22] Boyle WJ, Simonet WS, Lacey DL. Osteoclast differentiation and activation. Nature 2003;423:337 – 42. [23] Boone DL, Lee EG, Libby S, Gibson PJ, Chien M, Chan F, et al. Recent advances in understanding NF-nB regulation. Inflammatory Bowel Dis 2002;8:201 – 12. [24] Relea P, Revilla M, Ripoll E, Arribas I, Villa LF, Rico H. Zinc, biochemical markers of nutrition, and type I osteoporosis. Age Ageing 1995;24:303 – 7. [25] Katranoushov I, Dyankov O. A densitometric method of determining the mineral content of bone. Br J Radiol 1972;45:210 – 2. [26] Lappalainen R, Knuuttila M, Lammi S, Alhava EM, Olkkonen H. Zn and Cu content in human cancellous bone. Acta Orthop Scand 1982;53:51 – 5. [27] Yamaguchi M, Kishi S. Zinc compounds inhibit osteoclast-like cell formation at the earlier stage of rat marrow culture but not osteoclast function. Mol Cell Biochem 1996;158:171 – 7. [28] Yamaguchi M. h-Alanyl-l-histidinato zinc and bone resorption. Gen Pharmacol 1995;26:1179 – 83. [29] Kishi S, Yamaguchi M. Characterization of zinc effect to inhibit osteoclast-like cell formation in mouse marrow culture: interaction with dexamethasone. Mol Cell Biochem 1997;166:145 – 51. [30] Suzuki Y, Morita I, Yamane Y, Murota S. Preventive effect of zinc against cadmium-induced bone resorption. Toxicology 1990;62: 27 – 34. [31] Yamaguchi M, Uchiyama S. Receptor activator of NF-nB ligandstimulated osteoclastogenesis in mouse marrow culture is suppressed by zinc in vitro. Int J Mol Med 2004;14:81 – 5. [32] Vaananen HK, Zhao H, Mulari M, Halleen JM. The cell biology of osteoclast function. J Cell Sci 2000;113(Pt 3):377 – 81. [33] Wang F, Dufner-Beattie J, Kim BE, Petris MJ, Andrews G, Eide DJ. Zinc-stimulated endocytosis controls activity of the mouse ZIP1 and ZIP3 zinc uptake transporters. J Biol Chem 2004;279: 24631 – 9. [34] Gaither LA, Eide DJ. Eukaryotic zinc transporters and their regulation. Biometals 2001;14:251 – 70. [35] Eide DJ. The SLC39 family of metal ion transporters. Pfluegers Arch 2004;447(5):796 – 800. [36] Zhang YH, Heulsmann A, Tondravi MM, Mukherjee A, Abu-Amer Y. Tumor necrosis factor-a (TNF) stimulates RANKL-induced osteoclastogenesis via coupling of TNF type 1 receptor and RANK signaling pathways. J Biol Chem 2001;276:563 – 8. [37] Kong Y-Y, Yoshida H, Sarosi I, Tan H-L, Timms E, Capparelli C, et al. OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 1999;397:315 – 23.
Expression of the zinc transporter ZIP1 in osteoclasts Mohammed A. Khadeer a,1, Surasri N. Sahu a,1, Guang Bai a, Sunia Abdulla b, Anandarup Gupta a,* a
Department of Biomedical Sciences, Dental School, University of Maryland, 666 West Baltimore Street, Baltimore, MD 21201, USA b Department of Endodontics and Periodontics, Dental School, University of Maryland, Baltimore, MD 21201, USA Received 10 March 2005; revised 8 April 2005; accepted 13 April 2005 Available online 6 July 2005
Abstract Zinc has been previously demonstrated to be a potent inhibitor of osteoclastogenesis and osteoclast function. The mechanisms for cellular uptake of zinc into osteoclasts have not been characterized. We have corroborated previous studies on the reduction of osteoclastogenesis in the presence of extracellular zinc. We demonstrate that osteoclasts express a ubiquitous plasma membrane zinc transporter, namely ZIP1, which was diffusely distributed throughout the cytoplasm. Following an adenoviral-mediated overexpression of ZIP1 in murine osteoclasts, ZIP1 was predominantly colocalized with actin at the sealing zone and significantly inhibited osteoclast function, as assessed by resorptive activity. Finally, overexpression of ZIP1 negatively impacted NF-nB binding activity, as assessed by electrophoretic mobility shift assays. In conclusion, these data both corroborate previous studies on regulation of osteoclast formation and activity by zinc and reveal the presence of a zinc uptake mechanism that exerts an important effect on osteoclast activity. D 2005 Elsevier Inc. All rights reserved. Keywords: Osteoclasts; Zinc; ZIP1; Resorption; NF-nB
Introduction Zinc has been demonstrated to be an essential trace element for normal skeletal growth, and zinc deficiency has been associated with retarded growth [1]. Zinc has been implicated in bone mineralization through regulation of alkaline phosphatase and other metalloenzymes [2,3]. Zinc is fairly abundant in bone; the mean zinc content has been estimated as 110– 300 Ag/kg bone [4]. Intracellular zinc homeostasis is maintained by two distinct families of transporters, Zrt-Irt-like proteins (ZIP) and the cation diffusion facilitator (CDF) family [5]. The ZIP family of transporters mediates uptake of zinc into cells, whereas the CDF family of transporters is involved in zinc efflux and intracellular sequestration. These zinc transporters function through either secondary active transport or facilitated diffusion [6]. * Corresponding author. Fax: +1 410 706 0193. E-mail address: [email protected] (A. Gupta). 1 Both authors contributed equally to this work. 8756-3282/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.bone.2005.04.035
Osteoclasts are the principal resorptive cells of bone that are derived from the monocyte-macrophage lineage [7]. Most adult bone diseases, such as osteoporosis and periodontal disease, are due to an imbalance in bone remodeling, which favors resorption and loss of bone mass. Zinc has been shown to have a positive effect on skeletal growth by favoring bone formation [8,9]. Zinc has been well established to act as an inhibitor of osteoclast formation and activity [10,11]. Therefore, zinc has been shown to have a positive effect on skeletal growth by stimulation of osteogenesis, accompanied by a parallel inhibition of osteoclastogenesis. In the current study, using a murine osteoclast culture system, we have corroborated previous studies that have demonstrated inhibition of osteoclast formation in the presence of increasing extracellular concentrations of zinc. Next, we have demonstrated the presence of ZIP1, a ubiquitous plasma membrane zinc transporter, which may be responsible for zinc uptake in osteoclasts. Finally, using a recombinant adenovirus, we have successfully overexpressed ZIP1 in osteoclasts, which we found to exert a significant negative impact on both osteoclast activity and
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
the NF-nB pathway, which is essential for osteoclastogenesis and activity. Our data provide an important insight into the molecular mechanisms by which zinc and zinc transporters can regulate osteoclast formation and function.
Materials and methods Isolation of osteoclast-like cells from bone marrow The tibiae and femurs of 7-week-old mice were used to isolate bone marrow stromal cells, as previously described [12]. The protocol (04-08-03) for these animal studies has been approved by the IACUC at the University of Maryland, Baltimore. Bone marrow cells were suspended in aminimal essential medium (a-MEM, Gibco-BRL) supplemented with 10% fetal bovine serum (designated as a-10 MEM) and cultured at 37-C in a 5% CO2 incubator. After 24 h, non-adhered cells were layered on Histopaque-1077 (Sigma, St. Louis, MO) and centrifuged at 300 g for 15 min at room temperature (RT). The cell layer between the Histopaque and the medium was removed and washed with a-10 MEM, centrifuged at 2000 rpm for 7 min. Both recombinant M-CSF-1 (R&D Systems, Minneapolis, Minnesota) and RANKL were added to the cultures at concentrations of 10 ng/ml and 100 ng/ml, respectively, as previously described [12]. Multinucleated osteoclasts were seen to form and mature between days 4 –5. The percentage of tartrate-resistant acid phosphatase (TRAP)-positive cells was assessed to be ¨99%, using the TRAP+ assay, as previously described [12]. RNA isolation and RT-PCR for ZIP1 Total RNA was extracted from mature osteoclasts using TRIZOL Reagent (Invitrogen) as previously described [12]. Using reverse transcriptase (Invitrogen, Carlsbad, CA), cDNA was synthesized and used as a template to amplify the murine ZIP1 fragment. Primers for amplification of murine ZIP1 were as follows: mZIP1 (forward, F): 5V-CCAGAGCTCCTGGTGTGGCG-3V, and the hZIP1 (reverse, R): 5V-GCAGCAGTCCAGGAGACAA-3V. The conditions for PCR-amplification of mZIP1 were as follows: an initial denaturation at 94-C for 4 min followed by 35 cycles of 94-C for 1 min, 50-C for 1 min, and 72-C for 1 min. A final extension was done for 10 min at 72-C. The expected size for the mZIP1 amplicon was 230 kb. Western blot analysis Murine osteoclasts were used after day 5 in culture for preparation of lysates. Following two quick rinses with icecold phosphate-buffered saline (PBS), the cells were lysed in a buffer containing 10 mM Tris – HCl, pH 7.05, 50 mM NaCl, 0.5% Triton X-100, 30 mM sodium pyrophosphate, 5 mM NaF, 0.1 mM Na3VnO4, 5 mM ZnCl2, and 2 mM
297
PMSF. Lysates were pelleted by centrifugation at 15,000 rpm, for 15 min, at 4-C. The supernatant was transferred into a fresh microfuge tube and held on ice. Protein concentrations were measured using the Bio-Rad protein assay reagent kit (Bio-Rad, Hercules, CA), so that equal amounts of protein were loaded per lane (¨25 Ag) and were analyzed by SDS-PAGE on 8% gels. The gel was then electroblotted onto a PVDF membrane by wet – dry transfer. Nonspecific protein binding was blocked with 5% nonfat dry milk powder dissolved in PBS containing 0.1% Tween-20. The blots were incubated overnight with a 1:10,000 dilution of the primary antibody (chicken IgY) to the ZIP1 protein. This was followed by detection with a horseradish-peroxidase (HRP)-conjugated secondary goat anti-rabbit antibody (Sigma, 1:2000 dilution). The secondary antibody (HRP-conjugated goat anti-rabbit antibody) was detected by enhanced chemiluminescence, as previously described [13,14]. Indirect immunofluorescence The subcellular distribution of ZIP1 was examined in osteoclasts by indirect immunofluorescence. Osteoclasts were rinsed twice in ice-cold PBS, fixed with 4% paraformaldehyde, and permeabilized with PBS containing 0.2% Triton X-100. The cells were blocked with PBS containing 3% bovine serum albumin for 2 h at room temperature (RT). The cells were incubated for 2 h in a 1:500 dilution of affinity-purified chicken IgY directed against ZIP1 at room temperature. The primary antibodies were detected with Alexa-488-conjugated secondary antibodies (1:1000 dilution, Molecular Probes). Actin was labeled with rhodamine phalloidin. As negative controls, osteoclasts were incubated for 2 h in a 1:500 dilution with the non-immune serum (NIS, IgY). Confocal microscopy (Carl Zeiss LSM Meta 510) was performed; all images shown were taken with either a 63 or 100 oil-immersion objective lens. Construction of recombinant adenoviral vectors for ZIP1 Recombinant adenovirus for ZIP1 was generated as follows and according to methods previously described [15]. Briefly, the gene for ZIP1 was ligated into pShuttle-IREShrGFP2 (AdEasy-XL Adenoviral Vector system; Stratagene, La Jolla, CA) using the NheI– XhoI multiple cloning sites. This vector contains the CMV promoter and a dicistronic expression cassette in which the multiple cloning site (MCS) is followed by the EMCV-IRES, which directs translation of a humanized recombinant green fluorescent protein (hrGFP). This allows the expression of the gene of interest to be monitored at the single-cell level due to expression of the hrGFP on the same transcript. Recombinant Adeno-ZIP1-IRES-hrGFP2 was confirmed by digestion with PacI. High titer (2 1010 pfu/ml) viral stocks were made as follows: approximately 15 Ag of shuttle
298
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
plasmid and 5 Ag of pVQAd 9.2-100 backbone plasmid were digested with PacI. Both linear fragments were cotransfected into HEK293 cell line in a 60 mm plate at ¨50% confluency by standard CaCl2 method. The initial viral lysate was amplified in HEK293 cells, and the final viral lysate purified over two rounds of CsCl gradient ultracentrifugation. Virus particles were dialyzed against 3% sucrose/PBS buffer, diluted to 1 1012 particles/mL, and then frozen at 80-C. We assayed for infectious particle concentration by a cell-based plaque assay on HEK293 cells. Presence of replication competent adenovirus was checked by plaquing on the wild type permissive cell line A549 for at least 14 days. The adenoviral titer for both adenoviruses was ¨1011 pfu/ml. Recombinant adenovirus expressing EGFP (rAd-EGFP) was prepared as previously described [16] and served as an irrelevant negative control for infection of osteoclasts. Assessment of intracellular zinc Murine osteoclasts were incubated in the zinc-sensitive dye Zinquin ester (20 AM, Biotium, CA) for 30 min and subsequently rinsed in PBS. The cells were viewed under a Nikon TE200 inverted fluorescence microscope and UV filters (absorption 364 nm, emission 485 nm), as previously described [14]. In vitro resorption assays Following transfection or infection of murine osteoclasts with plasmid containing ZIP1 or rAd-ZIP1 for 48 h, respectively, mature murine osteoclasts were plated on CaPO4 (1 Am)-coated quartz discs (Osteologici discs, BD Biosciences). Approximately 48 h later, the effects on osteoclast-mediated resorption were assessed by rinsing the discs in distilled water to lyse the osteoclasts, as previously described [12]. The extent of resorption could be assessed by clear areas on the quartz discs where the CaPO4 film had disappeared. The discs were photographed (20 objective), and the area and number (of pits) were counted using SigmaScan (SPSS Inc., Chicago, IL). Statistical analyses were performed using unpaired t tests, and significance of differences between means was assessed at a minimum P < 0.05. Assay for NF-jB activity in osteoclasts Electrophoretic mobility shift assays (EMSAs) were performed following a protocol described previously [16,17]. Briefly, the osteoclast cultures were rinsed twice with ice-cold PBS. The cells were scraped into hypotonic lysis buffer A (10 mM HEPES [pH 7.8], 10 mM KCl, 1.5 mM MgCl2, 0.5 mM dithiothreitol, plus a protease inhibitor cocktail). Following incubation on ice for 15 min, NP-40 was added to a final concentration of 0.6%. The nuclei were pelleted by centrifugation at 10,000 g for 5 min. The
cytosolic fraction was removed, and the nuclei were resuspended in nuclear extraction buffer B (20 mM HEPES [pH 7.8], 420 mM NaCl, 1.2 mM MgCl2, 0.2 mM EDTA, 25% glycerol, 0.5 mM DTT, plus a protease inhibitor cocktail), vortexed briefly, and rocked for 30 min in 4-C. The insoluble material was removed by centrifugation for 5 min at 12,000 g. Finally, the nuclear proteins in the supernatant quantitated using the Bio-Rad protein assay. The nuclear proteins were dialyzed against a dialysis buffer (20 mM HEPES-Cl [pH 7.9], 0.2 mM EDTA, 100 mM KCL, 20% glycerol, 0.2 mM PMSF, 0.5 mM DTT). After preparation of nuclear proteins, EMSA was performed as follows. A probe for the NF-nB binding site was produced with DNA GB469 (upper strand, 5V-TCTGAGGAGTGGGACTGGCCAGCAT) and GB460 (lower strand, 5V-GCCTAATGCTGGCCAGTCCCACTCC). The oligonucleotide for NF-nB was labeled by Klenow enzyme in the presence of [a-32P]dCTP (>3000 mCi/mmol, ICN Biomedicals, Irvine, CA). The double stranded oligonucleotide probes for NF-nB were end-labeled with [g32P] dATP and T4 DNA polymerase kinase, annealed at 37-C for 30 min, and purified by running through a G-50 Column. Nuclear extracts (2.5 Ag) were incubated in binding buffer (200 mM HEPES, pH 8.3, 600 mM KCl, 50 mM DTT, and 10 Ag/Al BSA), poly (dI-dC), and Ficoll for 18 min at room temperature (RT). The end-labeled probe was added to this mixture and further incubated at RT for 25 min. Protein– DNA complexes were separated in a 4% non-denaturing polyacrylamide gel. Gels were dried and analyzed using a phosphorimager (Typhoon, Amersham BioSciences) or autoradiographic film. For supershift assays to identify components in the DNA – protein complex, the nuclear extracts were preincubated with antibodies (2 Ag) specific for p65, p50, C-Rel, and p52; an antibody specific for Sp1 was used as a negative control, as previously described [16].
Results Effects of extracellular zinc on osteoclastogenesis In order to test for the effects of zinc on osteoclast formation, osteoclasts were generated from murine bone marrow. Osteoclast precursors were incubated for 6 days either in the absence of, or presence of, increasing concentrations of ambient ZnCl2 (1, 5, and 10 AM), as shown in Fig. 1 (panel A, 1). The formation of osteoclast-like cells was evaluated by staining for TRAP+ cells, as previously described [12]. The total zinc content in culture media prior to supplementation with ZnCl2 was determined to be 3 AM by atomic absorption spectrometry, as previously described [18]. There was no indication of cellular toxicity or loss of cell viability associated with supplementation of culture medium with graded increases in ZnCl2 (ranging from 0 –10 AM), as shown in Fig. 1 (panel A, 2). The optimal concentration of extracellular ZnCl2 supplementation of culture medium for
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
299
Fig. 1. Effects of extracellular zinc on osteoclastogenesis and expression of ZIP1 in osteoclasts. (A) Changes in extracellular zinc and effects on osteoclastogenesis. (A,1) Multinucleated murine osteoclasts were generated from osteoclast precursor cells starting day 1 (of culture) in the presence of graded concentrations of extracellular ZnCl2 (1, 5, and 10 AM ZnCl2). The number of TRAP+ multinucleated osteoclasts (OC) was assessed on day 6 of culture (n = 12, P < 0.05). (A, 2) Representative images of murine osteoclasts in culture that were incubated in the presence of graded supplementation of (culture) medium with ZnCl2 (ranging from 0 – 10 AM). The basal level of zinc present in culture medium prior to supplementation with ZnCl2 was determined to be ¨3 AM using atomic absorption spectrometry. (B) Expression of ZIP1 in osteoclasts. Using RT-PCR, a ZIP1 fragment of 230 bp from murine osteoclasts was amplified from a human osteoclast (OC) cDNA library (lane 1) and primary murine osteoclasts (OC, lane 2). Specificity of species-specific (human, mouse) primers for amplification of ZIP1 was demonstrated by absence of transcript when the cDNA templates were switched (lanes 3 – 4). The full-length ZIP1 cDNA template cloned from a human prostate cancer cell line, PC3 cells served as a positive control (lane 5). Sequence homology between the human and murine ZIP1 transcripts revealed ¨82% nucleotide identity. (C) ZIP1 protein expression in osteoclasts. The ZIP1 protein was detected in human mesenchymal stem cells (MSC) and murine osteoclasts (mOC) with an affinity-purified ZIP1-specific chicken IgY.
significant inhibition (45 –55%) of osteoclastogenesis was determined to be ¨5– 10 AM zinc (n = 12, P < 0.05). Expression of ZIP1 in osteoclasts We next asked the question whether a ubiquitous zinc transporter ZIP1 is expressed in osteoclasts. Previous
studies have shown that the ZIP1 family of zinc transporters is expressed in bone marrow [19]. Using reverse transcriptase-polymerase chain reaction (RT-PCR), a ZIP1 fragment of 230 bp from murine osteoclasts was amplified, as shown in Fig. 1 (panel B); a similar transcript for ZIP1 was also amplified from a human osteoclast cDNA library (panel B). The positive control used for the PCR amplifi-
300
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
cation was the full-length ZIP1 cDNA template cloned from a human prostate cancer cell line, PC3 cells [14]. The sequence homology between the human and murine ZIP1 transcripts revealed ¨82% nucleotide identity. In addition, the ZIP1 protein could be detected in both human mesenchymal stem cells (MSCs) and murine osteoclasts (mOC), as shown in Fig. 1 (panel C). ZIP1 is a 324 amino acid protein with a predicted molecular mass of ¨34 kDa [14] but can be seen as a dimer, presumably because of a leucine zipper motif [14]. In both the human osteoblast and murine osteoclast, the ZIP1 protein can be detected at ¨68 kDa. Expression, overexpression, and cellular distribution of ZIP1 in murine osteoclasts We have examined the developmental expression of ZIP1 in murine osteoclasts, as shown in Fig. 2 (panel A). The protein levels of ZIP1 were not appreciably changed during the transition from pre-osteoclast precursors (pOC,
days 1 –4) to multinucleated osteoclasts (mnOC, days 5– 6). A recombinant adenovirus was generated for ZIP1 for overexpression in murine osteoclasts, as shown in Fig. 2 (panel B). Murine osteoclasts were infected with recombinant adenoviruses (rAd, MOI of ¨100) on day 5 of culture when the majority of cells were multinucleated osteoclasts. First, 48 h later, overexpression of ZIP1 was confirmed by RT-PCR for ZIP1, as shown in Fig. 2 (panel B). Compared to uninfected control (Con) or rAd-EGFP-infected osteoclasts, there was at least a 4-fold increase in the gene expression of ZIP1 following infection rAd-ZIP1; the transcript for GAPDH was amplified for normalization of the ZIP1 signal, as shown in Fig. 2 (panel B). Second, the increase in gene expression was followed by Western blot analysis for the ZIP1 protein (panel C), which paralleled the increase in ZIP1 gene expression. Third, we examined the subcellular distribution of ZIP1 in murine osteoclasts, before and after overexpression of ZIP1, as shown in Fig. 2 (panel D). Murine osteoclasts were infected
Fig. 2. Expression, overexpression, and cellular distribution of ZIP1 in murine osteoclasts. (A) Developmental expression of ZIP1 in murine osteoclasts. The protein levels of ZIP1 during the transition from pre-osteoclast precursors (pOC, days 1 – 3) to multinucleated osteoclasts (mnOC, days 5 – 6) were determined by Western blot analysis. Protein levels of h-actin were used to normalize the signal for ZIP1. (B) Overexpression of the ZIP1 gene in osteoclasts. A recombinant adenovirus was generated for ZIP1 for overexpression in murine osteoclasts. Murine osteoclasts were infected with recombinant adenoviruses (rAd, MOI of ¨100) on day 5 of culture when the majority of cells were multinucleated osteoclasts. First, 48 h later, overexpression of ZIP1 was confirmed by RT-PCR for ZIP1, compared to uninfected control (Con) or rAd-EGFP-infected osteoclasts. The transcript for GAPDH was amplified for normalization of the ZIP1 signal. (C) Overexpression of ZIP1 protein in osteoclasts. Following infection of murine osteoclasts with either rAd-EGFP or rAd-ZIP1, Western blot analysis with a ZIP1-specific chicken IgY was performed. (D) Subcellular distribution of ZIP1 in osteoclasts. Murine osteoclasts were infected either with a replication-deficient adenovirus encoding for the ZIP1 gene (rAd-ZIP1, frame 3) or for EGFP (rAd-EGFP, CON, frame 2). Immunostaining for ZIP1 was performed with a polyclonal chicken IgY for ZIP1, and actin was stained with rhodamine phalloidin. Osteoclasts were also incubated with non-immune serum (NIS, IgY, frame 1) to demonstrate lack of an apparent immunofluorescent signal for ZIP1, using the same settings for confocal microscopy (i.e., pinhole aperture, detector gain, and sensitivity). (E) Accumulation of intracellular zinc in ZIP1-overexpressing osteoclasts. Murine osteoclasts were infected with rAdZIP1 for 48 h and examined separately for EGFP-derived fluorescence from the reporter gene as an estimate of infection efficiency (frame 2) and for accumulation of intracellular zinc with Zinquin ester, a zinc-sensitive dye (frame 4). The images obtained from rAd-ZIP1-infected osteoclasts were compared to EGFP (frame 1)- and Zinquin ester (frame 3)-derived fluorescence in uninfected osteoclasts.
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
either with a replication-deficient adenovirus encoding for the ZIP1 gene (rAd-ZIP1, frame 3) or for EGFP (rAd-EGFP, frame 2). Immunostaining for ZIP1 was performed with a polyclonal chicken IgY for ZIP1 (Alexa 488, green) and actin was stained with rhodamine phalloidin (red). In rAdEGFP-infected osteoclasts, which served as controls (CON, frame 2), the signal for the native ZIP1 was mostly absent near the periphery, mostly distributed toward the cell interior, and showed very little colocalization existed with actin. However, after overexpression of ZIP1 (rAd-ZIP1, frame 3), the immunofluorescent signal for ZIP1 was discretely localized at the plasma membrane and showed significant colocalization with actin. The indirect immunofluorescence for the cellular distribution of ZIP1 was deemed to be specific since there was no apparent immunofluorescent signal for ZIP1 when osteoclasts were incubated with the non-immune chicken serum (NIS, IgY, frame 1), using identical settings for confocal microscopy (i.e., pinhole aperture, detector gain, and sensitivity). Finally, we examined the accumulation of intracellular zinc in osteoclasts that were overexpressing the ZIP1 protein, as shown in Fig. 2 (panel E). Similar to experiments detailed above, murine
301
osteoclasts were infected with rAd-ZIP1 for 48 h and examined separately for EGFP-derived fluorescence from the reporter gene as an estimate of infection efficiency (panel E, frame 2) and for accumulation of intracellular zinc with Zinquin ester, a zinc-sensitive dye (panel E, frame 4). The images obtained from rAd-ZIP1-infected osteoclasts were compared to uninfected osteoclasts that showed very low levels of EGFP-derived epifluorescence (panel E, frame 1) and low levels of zinc accumulation (panel E, frame 3). Effects of ZIP1 overexpression on osteoclast function We asked the question whether overexpression of ZIP1 would impact osteoclast function. We hypothesized that an overexpression of ZIP1 would increase the uptake of zinc into the osteoclast since the immunofluorescent data indicated that the ZIP1 protein was primarily localized at the plasma membrane in ZIP1-overexpressing osteoclasts (Fig. 2, panel D, frame 3). Mature murine osteoclasts were infected with either rAd-EGFP (control) or rAd-ZIP1 for 48 h. Subsequently, the infected osteoclasts were replated on CaPO4-coated Osteologici discs for 24 h in vitro resorp-
Fig. 3. Effects of ZIP1 overexpression on osteoclast function. (A) In vitro resorption assay. Murine osteoclasts were infected with either rAd-EGFP (control) or rAd-ZIP1 for 48 h. Subsequently, the infected osteoclasts were replated on CaPO4-coated Osteologici discs for resorption assays in the presence of 10 AM ZnCl2, which was added to the culture medium for 24 h. (A, a) Representative images of in vitro assays for osteoclast-mediated resorption in the absence of supplementation of culture medium with 10 AM ZnCl2 (a, 1), or supplementation with 10 AM ZnCl2 alone (a, 2), or following overexpression of ZIP1 in the absence of supplementation with 10 AM ZnCl2 (a, 3), or following overexpression of ZIP1 in the presence of supplementation with 10 AM ZnCl2 (a, 3). (A, b) The corresponding resorption areas (b, 1) were photographed and quantified using SigmaScan (b, 2). (B) Comparison of changes in resorptive activity by ZIP1overexpressing osteoclasts. Data obtained from quantitation of resorption areas were plotted to reflect changes in resorptive activity (after 24 h of being plated on CaPO4-coated Osteologici discs and represented as ‘‘percentage of area resorbed by controls’’) under the following conditions: control untreated osteoclasts (i.e., 0 AM ZnCl2), osteoclasts cultured in the presence of 10 AM ZnCl2, rAd-EGFP-infected osteoclasts (rAd-EGFP,10 AM ZnCl2), and ZIP1overexpressing osteoclasts (rAd-ZIP1,10 AM ZnCl2). Statistical comparisons were performed with unpaired t tests, and the P values are reported in the Results section.
302
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
tion assays, according to protocols previously published [12,20], in the presence of 10 AM ZnCl2. Representative images of osteoclast-mediated in vitro resorption under various experimental conditions that include control untreated osteoclasts (0 AM ZnCl2, A, a, 1), supplementation of culture medium with 10 AM ZnCl2 (A, a, 2), overexpression of ZIP1 alone (A, a, 3), or overexpression of ZIP1 in the presence of 10 AM ZnCl2 (A, a, 4) are shown in Fig. 3 (panel A, a, 1– 4). Next, the corresponding representative resorption areas were photographed (panel A, b, 1) and quantified using SigmaScan (panel A, b, 2), as described in Materials and methods. Our data indicate that resorption was not significantly altered either in the presence of 10 AM ZnCl2 or by rAd-EGFP-infected osteoclasts (in the presence of 10 AM ZnCl2), compared to control untreated osteoclasts, as shown in Fig. 3 (panel B). However, there was a dramatic reduction in resorption by ZIP1-overexpressing osteoclasts (rAd-ZIP1, 10 AM ZnCl2, 28.5 T 4.40, P = 0.0001), compared to either untreated controls (0 AM ZnCl2, 100 T 9.80) or osteoclasts cultured in the presence of 10 AM ZnCl2 (75.3 T 7.14, P = 0.0005) or rAdEGFP-infected osteoclasts (10 AM ZnCl2, 28.5 T 4.41, P = 0.0004). These data indicate that, first, the presence of 10 AM ZnCl2 in the culture medium is not sufficient to significantly inhibit resorptive activity by mature osteoclasts, and second, the inhibition of resorptive activity in
ZIP1-overexpressing osteoclasts is specific for the ZIP1 protein since there was no impact on resorption by osteoclasts which were infected with the recombinant adenovirus encoding for EGFP (rAd-EGFP). Assay for changes in NF-jB activity following overexpression of ZIP1 in osteoclasts NF-nB is essential for osteoclast formation and activity [17,21,22]. It is known that NF-nB sites interact with dimerized NF-nB factors that include p65, p50, p52, C-Rel, and Rel B. Homo- or heterodimers of NF-nB factors are associated with InB factor in the cytoplasm and remain inactive. In response to activating signals, InB dissociates from the complex, thus releasing NF-nB dimers, which become activated and subsequently translocate to the nucleus for regulation of transcription [23]. We asked the question whether reduced osteoclast activity following overexpression of ZIP1 is associated with reduced NF-nB activity. Following infection of pre-osteoclast precursors (on day 4 of culture) with either rAd-EGFP or rAd-ZIP1 for 72 h, nuclear extracts were prepared, and an electrophoretic mobility shift assay (EMSA) was performed using 32Plabeled NF-nB oligonucleotide. As shown in Fig. 4, NF-nB activity was decreased as evidenced by decreased DNAbinding activity (panel A, lane 3), compared to either
Fig. 4. Assay for changes in NF-nB activity in ZIP1-overexpressing osteoclasts. (A) Electrophoretic mobility shift assay (EMSA) for NF-nB activity. Following infection of pre-osteoclast precursors (on day 4 of culture) with either rAd-EGFP or rAd-ZIP1 for 72 h, nuclear extracts were prepared. An Ig/HIV NF-nB consensus sequence (22 bp) was end-labeled with 32P and used as probe in EMSA experiments. NF-nB binding activity in ZIP1-overexpressing osteoclasts (lane 3), compared to either control untreated osteoclasts (lane 2), rAd-EGFP-infected osteoclasts, or TNF-a treated osteoclasts (10 ng/ml, 24 h, lane 5). As positive controls for the assay, nuclear extracts were prepared from P19 cells, an embryonic carcinoma cell line, and different protein concentrations were tested (lanes 6 – 7, P19-1, P19-2). (B) Constitutive presence of NF-nB protein (p50) in nuclei of osteoclasts (antibody supershift assay). Nuclear proteins were extracted from osteoclasts, and nuclear proteins were preincubated with indicated antibodies (specific for p65, p50, C-Rel, p52, or Sp1) individually before the probe was added. Nuclear proteins were extracted from undifferentiated P19 cells. For antibody supershift assays, the nuclear proteins were preincubated with antibody specific for p65, p50, C-Rel, p52, or Sp1. Band B was supershifted by p50 antibody to form band C. Nuclear proteins extracted from osteoclasts formed two protein – DNA complexes with the NF-nB consensus, as shown by an antibody (a)-induced supershift of the NF-nB factor, p50. The upper slowly migrating band (B) contained the NF-nB factor, p50, as shown by antibody supershift lane 3; the band B was supershifted by the p50 antibody (ap50) to form band C. The antibody supershift experiments revealed the newly translocated p65 factor (ap65, band D). In the same assay, the Sp1 antibody served as a control for non-related factors. The fast migrating band (A) was not interfered with by any of the antibodies tested, and its identity remains unknown.
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
control untreated osteoclasts (CON, lane 2), rAd-EGFPinfected osteoclasts, or TNF-a treated osteoclasts (10 ng/ml, 24 h, lane 5). Nuclear extracts were also prepared from P19 cells, an embryonic carcinoma cell line, and different protein concentrations were tested (lanes 6 – 7, P19-1, P19-2); these nuclear extracts served as additional positive controls for the NF-nB activation assay, as previously described [16]. In order to validate the results of the EMSA and clarify NF-nB factor identity, we performed an antibody supershift assay, as shown in Fig. 4 (panel B). Nuclear proteins were extracted from osteoclasts, and nuclear proteins were preincubated with indicated antibodies (specific for p65, p50, C-Rel, p52, or Sp1) individually before the probe was added. We show that nuclear proteins extracted from osteoclasts formed two protein – DNA complexes with the NF-nB consensus (Fig. 4, panel B), as shown by an antibody (a) supershift of the NF-nB factor, p50. The upper slowly migrating band (B) was supershifted by the p50 antibody (ap50) to form band C (lane 3), indicating that band B contains p50. Furthermore, band B was completely supershifted by the p65 antibody (ap65, band D). This complete supershift suggests that p65 involves formation of all DNA –protein complexes in band B. In comparison, the migration or intensity of band B was not changed by other antibodies tested, including one against Sp1 as a control for non-related factors, as previously described [16]. These data demonstrate that band B is composed of a heterodimer of p50 and p65 and/or a homodimer of p65 with the NF-nB consensus. Considering that overexpression of ZIP1 (in murine osteoclasts) significantly reduced band B, we believe that the heterodimer of p50 and p65 and/or a homodimer of p65 in osteoclasts mediates initiation of the NF-nB signaling pathway. Finally, we noticed that there is a fast migrating band A, which was not interfered with by any of the antibodies tested, and its identity remains unknown, as previously reported [16].
Discussion The role of zinc in the preservation of bone mass by stimulating bone formation and inhibiting bone resorption has been well established [24]. The zinc content in adult bone is fairly high compared to other tissues [25,26]. We have hypothesized that, during bone resorption, osteoclasts may be exposed to gradual increases in zinc, which upon uptake into cells through previously unidentified mechanisms, can inhibit osteoclast function. Previously, several studies have clearly established that zinc and zinc-chelating compounds (10 6 to 10 4 M) can inhibit in vitro osteoclastogenesis and osteoclast activity [10,11,27 – 31]. We have corroborated these earlier studies by demonstrating that, indeed, exposure of murine osteoclast precursors to graded increases in (ambient) zinc concentration inhibits osteoclastogenesis, with an optimal extracellular zinc concentration of 10 AM.
303
Having established the paradigm for zinc-mediated inhibition of osteoclast formation, we hypothesized that osteoclasts must possess mechanisms or transporters for zinc uptake. In most cells, the intracellular zinc homeostasis is maintained by a combination of several uptake, sequestration, and efflux mechanisms [5]. We have demonstrated gene and protein expression of ZIP1, a plasma membrane zinc transporter and a member of the expanding ZIP superfamily of zinc transporters [19]. We hypothesize that ZIP1 is the major plasma membrane zinc transporter in osteoclasts; we have screened for gene expression of several other members of the ZIP superfamily but have not been able to detect transcripts for other ZIP proteins (data not shown). In the murine osteoclast, the cellular distribution of ZIP1 was largely confined to a diffuse cytoplasmic pool, but when overexpressed, was discretely localized at or near the plasma membrane and was colocalized with the actin ring, which delineates the sealing zone [32]. In the current study, we have not determined whether the cellular distribution of ZIP1 is altered in response to increases or depletion of extracellular zinc levels, as has been previously reported for ZIP1 in other cell types [33]. It has been demonstrated that some members of the ZIP family, namely ZIP1 and ZIP3, can be found in intracellular organelles in zincreplete medium but are capable of trafficking between the plasma membrane and intracellular compartments, mostly a consequence of changes in endocytic rates. Zinc deficiency increased plasma membrane levels of these zinc transporters by decreasing their rates of endocytosis, providing evidence of a conserved mechanism controlling their activity. We have found that an adenovirus-mediated overexpression of ZIP1 led to a profound inhibition of mature osteoclast activity, as determined by in vitro resorption assays. These data indicate that increased surface expression of ZIP1 may enhance the total uptake of zinc into the osteoclast. In future studies, we will determine the kinetics of ZIP1 in osteoclasts, both in native and ZIP1overexpressing osteoclasts. Previous studies in other cell types have established that, generally, overexpression of zinc transporters increases V max without having any effect on the affinity of the transporter for zinc [6,34,35]. In a very recent study, it has been indirectly suggested that zinc negatively impacts osteoclast differentiation by inhibiting the NF-nB pathway in osteoclast precursors [31]. When murine bone marrow stromal cells were cultured in the presence of osteoclastogenic cytokines (such as LPS, TNF-a, and RANKL) in the presence of zinc sulfate, only TNF-a and RANKL-induced osteoclastogenesis was significantly inhibited. This inhibition was relieved in the presence of transcription and translation inhibitors. The authors surmised that zinc affected its inhibition on osteoclastogenesis through the NF-nB pathway that is specifically stimulated by both TNF-a and RANKL [21,36,37]. We have extended these preliminary studies by
304
M.A. Khadeer et al. / Bone 37 (2005) 296 – 304
demonstrating that adenovirus-mediated overexpression of the ZIP1 protein in osteoclast precursors led to a very significant decrease in NF-nB binding activity, as determined by electrophoretic mobility shift assays. These data provide the first direct evidence for a zinc transporter negatively impacting osteoclast differentiation and activity. In conclusion, these data provide a preliminary characterization of ZIP1, a ubiquitous zinc transporter, which is also expressed in the osteoclast, and extend previous studies of the negative impact of zinc on osteoclast differentiation and function. Finally, our data also provide the first evidence of a direct inhibition of the osteoclastogenic NF-nB pathway by overexpression of ZIP1 in osteoclast precursors. These studies add new information on the underlying molecular mechanisms by which zinc may impact osteoclast formation and activity.
Acknowledgments We acknowledge the help of Bryan Kitahara, University of Michigan, Ann Arbor, MI, who was supported on a Short Term Research Training Grant for Dental Students T35DE07334-05. This work was partly supported by an NIH grant AR 44792 (to A.G.), and NS 38077 (to G.B.).
References [1] Nishi Y. Zinc and growth. J Am Coll Nutr 1996;15:340 – 4. [2] McCall KA, Huang C, Fierke CA. Function and mechanism of zinc metalloenzymes. J Nutr 2000;130:1437S – 46S. [3] Chesters JK. Biochemical functions of zinc in animals. World Rev Nutr Diet 1978;32:135 – 64. [4] Gomez S, Rizzo R, Pozzi-Mucelli M, Bonucci E, Vittur F. Zinc mapping in bone tissues by histochemistry and synchrotron radiationinduced X-ray emission: correlation with the distribution of alkaline phosphatase. Bone 1999;25:33 – 8. [5] Guerinot ML. The ZIP family of metal transporters. Biochim Biophys Acta 2000;1465:190 – 8. [6] Liuzzi JP, Cousins RJ. Mammalian zinc transporters. Annu Rev Nutr 2004;24:151 – 72. [7] Teitelbaum SL. Bone resorption by osteoclasts. Science 2000;289: 1504 – 8. [8] Calhoun NR, Smith Jr JC, Becker KL. The role of zinc in bone metabolism. Clin Orthop 1974;0:212 – 34. [9] Saltman PD, Strause LG. The role of trace minerals in osteoporosis. J Am Coll Nutr 1993;12:384 – 9. [10] Kishi S, Yamaguchi M. Inhibitory effect of zinc compounds on osteoclast-like cell formation in mouse marrow cultures. Biochem Pharmacol 1994;48:1225 – 30. [11] Moonga BS, Dempster DW. Zinc is a potent inhibitor of osteoclastic bone resorption in vitro. J Bone Miner Res 1995;10:453 – 7. [12] Gupta A, Lee BS, Khadeer MA, Tang Z, Chellaiah M, Abu-Amer Y, et al. Leupaxin is a critical adaptor protein in the adhesion zone of the osteoclast. J Bone Miner Res 2003;18:669 – 85. [13] Khadeer MA, Tang Z, Tenenhouse HS, Eiden MV, Murer H, Hernando N, et al. Na+-dependent phosphate transporters in the murine osteoclast: cellular distribution and protein interactions. Am J Physiol: Cell Physiol 2003;284:C1633 – 44.
[14] Franklin RB, Ma J, Zou J, Guan Z, Kukoyi BI, Feng P, et al. Human ZIP1 is a major zinc uptake transporter for the accumulation of zinc in prostate cells. J Inorg Biochem 2003;96:435 – 42. [15] Anderson RD, Haskell RE, Xia H, Roessler BJ, Davidson BL. A simple method for the rapid generation of recombinant adenovirus vectors. Gene Ther 2000;7:1034 – 8. [16] Liu A, Hoffman PW, Lu W, Bai G. NF-nB site interacts with Sp factors and up-regulates the NR1 promoter during neuronal differentiation. J Biol Chem 2004;279:17449 – 58. [17] Abu-Amer Y. IL-4 abrogates osteoclastogenesis through STAT6dependent inhibition of NF-nB. J Clin Invest 2001;107:1375 – 85. [18] Costello LC, Liu Y, Franklin RB, Kennedy MC. Zinc inhibition of mitochondrial aconitase and its importance in citrate metabolism of prostate epithelial cells. J Biol Chem 1997;272:28875 – 81. [19] Gaither LA, Eide DJ. The human ZIP1 transporter mediates zinc uptake in human K562 erythroleukemia cells. J Biol Chem 2001; 276:22258 – 64. [20] Gupta A, Guo XL, Alvarez UM, Hruska KA. Regulation of sodiumdependent phosphate transport in osteoclasts. J Clin Invest 1997;100: 538 – 49. [21] Abu-Amer Y, Tondravi MM. NF-nB and bone: the breaking point. Nat Med 1997;3:1189 – 90. [22] Boyle WJ, Simonet WS, Lacey DL. Osteoclast differentiation and activation. Nature 2003;423:337 – 42. [23] Boone DL, Lee EG, Libby S, Gibson PJ, Chien M, Chan F, et al. Recent advances in understanding NF-nB regulation. Inflammatory Bowel Dis 2002;8:201 – 12. [24] Relea P, Revilla M, Ripoll E, Arribas I, Villa LF, Rico H. Zinc, biochemical markers of nutrition, and type I osteoporosis. Age Ageing 1995;24:303 – 7. [25] Katranoushov I, Dyankov O. A densitometric method of determining the mineral content of bone. Br J Radiol 1972;45:210 – 2. [26] Lappalainen R, Knuuttila M, Lammi S, Alhava EM, Olkkonen H. Zn and Cu content in human cancellous bone. Acta Orthop Scand 1982;53:51 – 5. [27] Yamaguchi M, Kishi S. Zinc compounds inhibit osteoclast-like cell formation at the earlier stage of rat marrow culture but not osteoclast function. Mol Cell Biochem 1996;158:171 – 7. [28] Yamaguchi M. h-Alanyl-l-histidinato zinc and bone resorption. Gen Pharmacol 1995;26:1179 – 83. [29] Kishi S, Yamaguchi M. Characterization of zinc effect to inhibit osteoclast-like cell formation in mouse marrow culture: interaction with dexamethasone. Mol Cell Biochem 1997;166:145 – 51. [30] Suzuki Y, Morita I, Yamane Y, Murota S. Preventive effect of zinc against cadmium-induced bone resorption. Toxicology 1990;62: 27 – 34. [31] Yamaguchi M, Uchiyama S. Receptor activator of NF-nB ligandstimulated osteoclastogenesis in mouse marrow culture is suppressed by zinc in vitro. Int J Mol Med 2004;14:81 – 5. [32] Vaananen HK, Zhao H, Mulari M, Halleen JM. The cell biology of osteoclast function. J Cell Sci 2000;113(Pt 3):377 – 81. [33] Wang F, Dufner-Beattie J, Kim BE, Petris MJ, Andrews G, Eide DJ. Zinc-stimulated endocytosis controls activity of the mouse ZIP1 and ZIP3 zinc uptake transporters. J Biol Chem 2004;279: 24631 – 9. [34] Gaither LA, Eide DJ. Eukaryotic zinc transporters and their regulation. Biometals 2001;14:251 – 70. [35] Eide DJ. The SLC39 family of metal ion transporters. Pfluegers Arch 2004;447(5):796 – 800. [36] Zhang YH, Heulsmann A, Tondravi MM, Mukherjee A, Abu-Amer Y. Tumor necrosis factor-a (TNF) stimulates RANKL-induced osteoclastogenesis via coupling of TNF type 1 receptor and RANK signaling pathways. J Biol Chem 2001;276:563 – 8. [37] Kong Y-Y, Yoshida H, Sarosi I, Tan H-L, Timms E, Capparelli C, et al. OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymph-node organogenesis. Nature 1999;397:315 – 23.