Effects of extracellular magnesium on the differentiation and function of human osteoclasts

Acta Biomaterialia xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Acta Biomaterialia

journal homepage: www.elsevier .com/locate /actabiomat

Effects of extracellular magnesium on the differentiation and function ofhuman osteoclasts

http://dx.doi.org/10.1016/j.actbio.2014.02.0101742-7061/� 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

⇑ Corresponding author. Tel.: +49 4152 87 1292; fax: +49 4152 87 2666.E-mail address: [email protected] (B.J.C. Luthringer).

1 L.W. and B.J.C.L. contributed equally and therefore share first authorship.

Please cite this article in press as: Wu L et al. Effects of extracellular magnesium on the differentiation and function of human osteoclasts. Acta Bi(2014), http://dx.doi.org/10.1016/j.actbio.2014.02.010

Lili Wu a,b,1, Bérengère J.C. Luthringer a,⇑,1, Frank Feyerabend a, Arndt F. Schilling b, Regine Willumeit a

a Institute of Materials Research, Department for Structural Research on Macromolecules, Helmholtz-Zentrum Geesthacht (HZG), Geesthacht, Germanyb Department of Plastic Surgery and Hand Surgery, Klinikum Rechts der Isar, Technical University Munich, Munich, Germany

a r t i c l e i n f o a b s t r a c t

Article history:Received 17 September 2013Received in revised form 25 January 2014Accepted 3 February 2014Available online xxxx

Keywords:OsteoclastMagnesiumBiodegradationPrimary cellsCell culture

Magnesium-based implants have been shown to influence the surrounding bone structure. In an attemptto partially reveal the cellular mechanisms involved in the remodelling of magnesium-based implants,the influence of increased extracellular magnesium content on human osteoclasts was studied. Peripheralblood mononuclear cells were driven towards an osteoclastogenesis pathway via stimulation with recep-tor activator of nuclear factor kappa-B ligand and macrophage colony-stimulating factor for 28 days. Con-comitantly, the cultures were exposed to variable magnesium concentrations (from either magnesiumchloride or magnesium extracts). Osteoclast proliferation and differentiation were evaluated based oncell metabolic activity, total protein content, tartrate-resistant acid phosphatase activity, cathepsin Kand calcitonin receptor immunocytochemistry, and cellular ability to form resorption pits. While magne-sium chloride first enhanced and then opposed cell proliferation and differentiation in a concentration-dependent manner (peaking between 10 and 15 mM magnesium chloride), magnesium extracts (withlower magnesium contents) appeared to decrease cell metabolic activity (�50% decrease at day 28) whileincreasing osteoclast activity at a lower concentration (twofold higher). Together, the results indicatedthat (i) variations in the in vitro extracellular magnesium concentration affect osteoclast metabolismand (ii) magnesium extracts should be used preferentially in vitro to more closely mimic the in vivoenvironment.

� 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.

1. Introduction

Magnesium is the second most abundant intracellular divalentcation. Consequently, Mg and its corrosion products (magnesiumchloride, oxide, sulphate or phosphate in a saline environment)exhibit excellent biocompatibility [1–3]. Most Mg (approximately67%) is found in bone, 30% of which is exchangeable (present onthe crystal surface of the bone). Thus, bone provides a dynamicstore for maintaining the intra- and extracellular concentrationsof Mg [4,5]. Magnesium is involved in diverse mechanisms (e.g.acting over 300 enzymes as a cofactor or in metabolic pathways)and plays a structural role in the cell membrane and chromosomes[6,7]. Its clinical importance is often underestimated (as serum Mglevels are not routinely tested). Magnesium is commonly noted asa calcium antagonist [8], e.g. in muscle contraction/relaxation,neurotransmitter release and action potential conduction in nodaltissue. Additionally, the role of these elements as a ‘‘chronic

regulator’’ (magnesium) or ‘‘acute regulator’’ (calcium) remains adebated question [9]. Nevertheless, their fates and roles are inter-locked. Under cellular magnesium deficiency (hypomagnesaemia),potassium is affected. As magnesium is needed to activate Na/Kpumps, Na and K gradients will not be maintained, leading to pas-sive intracellular loss of K and increases in intracellular Na andhydrogen ions [10]. Subsequently, intracellular calcium overloadwill take place, as calcium transport is magnesium dependent,and Na inhibits Na/Ca exchange. This general electrolyte imbalancewill lead to decreases in cell activity and vitality as well as electri-cal instability (e.g. cardiac arrhythmia). Generally, Mg supplemen-tation alone is sufficient to restore electrolyte homeostasis [11]. Asthe kidney can respond rapidly to an elevated serum level of Mg,hypermagnesaemia is rare (mainly due to advanced chronic kidneydisease and iatrogenic artefacts, e.g. laxatives and anti-acids [12–14]), and it has therefore not seemed important to study the effectsof high Mg concentrations at a cellular level.

Rising life expectancies as well as complications associated withthe long-term presence of implants have increased the interest indegradable materials for bone substitution [15]. This situationhas led to a renaissance of magnesium as a base for alloys for

omater

http://dx.doi.org/10.1016/j.actbio.2014.02.010

mailto:[email protected]


http://www.sciencedirect.com/science/journal/17427061

http://www.elsevier.com/locate/actabiomat


2 L. Wu et al. / Acta Biomaterialia xxx (2014) xxx–xxx

orthopaedic applications. Recent advances in material science andengineering have enabled corrosion rates and mechanical proper-ties to be specifically modulated [16] through the use of alloyingelements. Mg-based implants can now provide temporary struc-tural support. Naturally, degradation of these implants leads to ahigh magnesium concentration around the implant, and it has beenshown that this effect can result in a welcome increase in bonemass around the implant [16,17]. While a contribution of bone-forming osteoblasts to effect has been demonstrated, the impactof potentially decreased osteoclastic resorption around theMg-based implant is still open to discussion.

Therefore, the aim of the present study was to analyse the directeffect of increased magnesium concentrations on the differentia-tion and function of human osteoclasts.

2. Materials and methods

2.1. Magnesium-containing solutions

A 1 M MgCl2 (Sigma-Aldrich Chemie GmbH, Munich, Germany)solution was prepared in double-distilled water (ddH2O) and sub-sequently sterile filtered through a membrane filter (0.2 lm;Merck KGaA, Darmstadt, Germany). Appropriate dilutions (0 (con-trol), 2, 5, 10, 15 and 25 mM) were prepared by diluting the 1 MMgCl2 solution directly in the cell culture medium.

A magnesium extract (Mg extract) was prepared according toEN ISO standards 10993:5 [18] and 10993:12 [19] (i.e. the relation-ship of specimen weight to the extraction medium was 0.2 g ml–1).Magnesium specimens were cut from a cast ingot of pure Mg(99.95%; Helmholtz Centre Geesthacht, Geesthacht, Germany) incuboid form (1 cm � 1 cm � 0.5 cm). After sterilization via sonica-tion for 20 min in 100% isopropanol (Merck, Darmstadt, Germany),the samples were incubated in extraction medium (Eagle’sminimum essential medium, alpha modification (a-MEM); LifeTechnologies GmbH, Karlsruhe, Germany) supplemented with10% foetal bovine serum (FBS; PAA Laboratories GmbH, Linz, Aus-tria) for 72 h under physiological conditions (5% CO2, 20% O2, 95%humidity, 37 �C). The Mg extract was diluted 1:5, 1:10 or 1:30 incell culture medium (referred to as 5�, 10�, and 30� solutions,respectively; pure cell culture medium without the extract wasused as a control). The Mg extract solutions were further character-ized by measuring their osmolality, pH and Mg and calcium con-tents. Osmolality and pH were quantified using a Gonotec 030-Dcryoscopic osmometer (Gonotec, Berlin, Germany) and an ArgusXpH Meter (Sentron Europe BV, Roden, the Netherlands), respec-tively. Magnesium concentrations were measured using the metal-lochromic dye xylidyl blue following a colorimetric method. A10 ll aliquot of the Mg extract, extraction medium, or of differentconcentrations of MgCl2 (standard curve) was mixed with 1.5 ml ofa xylidyl blue solution (200 mmol l–1 trishydroxymethylaminome-thane buffer, pH 12, containing 0.12 mmol l–1 xylidyl blue II,0.05 mmol l–1 ethylene glycol tetraacetic acid (Titriplex),69 mmol l–1 potassium carbonate, 2.1 mol l–1 ethanol and 0.05%sodium azide (all chemicals supplied by Sigma-Aldrich ChemieGmbH, Munich, Germany). Following a 10 min incubation at roomtemperature (RT), absorbances of the mixtures were measured at520 nm with an enzyme-linked immunosorbent assay reader (Te-can Sunrise; TECAN Deutschland GmbH, Crailsheim, Germany).As the intensity of the colour is directly proportional to the magne-sium content, unknown concentrations were determined using thestandard curve. Calcium content was colorimetrically determinedusing calcium O-cresolphtalein kit (Futura system Srl, Rome, Italy)and a CaCl2 (Sigma-Aldrich Chemie GmbH, Munich, Germany)standard curve. Fifty microlitres of the samples were mixed with2 ml of reagent. After 5 min of incubation at RT, the absorbances

Please cite this article in press as: Wu L et al. Effects of extracellular magnesiu(2014), http://dx.doi.org/10.1016/j.actbio.2014.02.010

of the mixtures were measured at 580 nm with a Tecan Sunrisemicroplate reader. Unknown concentrations were determinedfrom the standard curve.

2.2. Cell culture

Peripheral blood mononuclear cells (PBMC) were freshly iso-lated from buffy coat specimens from healthy, consenting donors(purchased from the Institute for Clinical Transfusion Medicineand Immunogenetics Ulm, Ulm, Germany) via a density gradientcentrifugation technique using Ficoll Paque Plus (Amersham Bio-sciences, Uppsala, Sweden). As described in greater detail in Ref.[20], heparinized human peripheral blood (approximately 25 ml)was diluted 1:8 in phosphate-buffered saline (PBS; Biochrom AG,Berlin, Germany), carefully layered over Ficoll Paque, then centri-fuged at 350g for 30 min. The monocyte-enriched PBMC fractionaccumulated at the interface between PBS and Ficoll Paque wascarefully collected and washed twice with PBS. PBMC were thencultured in 48-well plates (with or without dentine chips) at a den-sity of 2 � 106 cells ml–1 in a-MEM (Biochrom AG, Berlin, Ger-many) supplemented with 10% FBS (Life Technologies GmbH,Karlsruhe, Germany), 2 mM L-glutamine (Life Technologies GmbH,Karlsruhe, Germany), 1% antibiotic/antimitotic solution (PAALaboratories GmbH, Pasching, Austria) and factors promotingosteoclastic differentiation (i.e. recombinant receptor activator ofnuclear factor kappa-B ligand (RANKL; Peprotech Germany, Ham-burg, Germany; 40 ng ml–1), in combination with a stimulator ofhematopoietic precursors (macrophage colony-stimulating factor,M-CSF; Peprotech Germany, Hamburg, Germany; 20 ng ml–1) at37 �C and 5% CO2 and 95% H2O saturation. Dentine (ivory) waskindly provided by German customs in accordance with the inter-national laws for the protection of species. Non-adherent cellswere discarded after 24 h. To evaluate the effects of MgCl2 andthe Mg extract on osteoclastogenesis, different MgCl2 concentra-tions (0 (control), 2, 5, 10, 15 and 25 mM) and Mg extract dilutions(0 (control), 5�, 10� and 30� in culture medium) were added tothe media. Human osteoclasts were differentiated for a period of28 days by changing 50% of the cell culture medium every twodays. Biochemical tests (to determine the total protein content(see the Supplemental data ‘‘Total protein extraction’’ for detailsof the materials and method), viability/proliferation and tartrate-resistant acid phosphatase (TRAP) activity) were performed ondays 3, 7, 14, 21 and 28 to observe the impact of MgCl2 and theMg extract on osteoclastogenesis. TRAP staining, immunocyto-chemistry detection of cathepsin K (CK) and calcitonin receptor(CTR) and resorption assays (testing the cells’ ability to formresorption pits) were performed at the end of the cultivation peri-od (day 28). For MgCl2, the experiments were performed threetimes with three different donors, while for the Mg extract, theywere conducted twice with two different donors.

2.3. Cell metabolism assay

To assess cellular metabolism, water-soluble tetrazolium (WST-1; Roche Diagnostic GmbH, Mannheim, Germany) was used. WST-1 is reduced at the external surface of the plasma membrane bynicotinamide adenine dinucleotide (NADH) oxidase (mirroringmitochondrial electron transport, i.e. the metabolic activity of thecell), forming highly coloured formazan, which absorbs at450 nm. After 3, 7, 14, 21 or 28 days of culture, the cells were incu-bated with a 10 vol.% WST-1 dye solution (n = 3 wells) under cellculture conditions for 2 h. As blanks, samples of the culture media(with corresponding Mg-containing solutions) without cells weretreated similarly. After 2 h, 200 ll of the culture supernatant (intriplicate) was pipetted into 96-well plates, which were subse-quently read using a microplate reader (Berthold Technologies

m on the differentiation and function of human osteoclasts. Acta Biomater


Table 1Characterization of Mg extract solutions.

Extract pH Osmolality(osmol kg�1)

Mg content(mM)

Ca content(mM)

a-MEM+10%FBS

7.80 0.33 0.93 1.76

1� 8.35 0.51 26.67 0.085� 8.04 0.46 6.08 1.4210� 7.90 0.38 3.50 1.5930� 7.80 0.34 1.46 1.70

L. Wu et al. / Acta Biomaterialia xxx (2014) xxx–xxx 3

GmbH, Bad Herrenalb, Germany) at 450 nm with a 620 nm refer-ence wavelength.

2.4. TRAP activity assay

After 3, 7, 14, 21 and 28 days of culture, osteoclast differentia-tion was evaluated based on TRAP activity (three biologicalreplicates per magnesium concentration). Enzyme activity was as-sessed extracellularly (‘‘extracellular TRAP’’, TRAP secreted into themedium, measured in cell culture supernatants). A 150 ll aliquotof the TRAP reaction solution (100 mM sodium acetate (Carl RothGmbH, Karlsruhe, Germany), 50 mM disodium-tartrate dihydrate(Carl Roth GmbH, Karlsruhe, Germany) and 7.6 mM p-nitrophenylphosphate disodium hexahydrate (pNPP disodium hexahydrate;Sigma-Aldrich Chemie GmbH, Munich, Germany)) was added to50 ll of the samples in 96-well plates, followed by incubation for1 h at 37 �C. The enzymatic reaction was then stopped by adding50 ll of a stop solution (3 M NaOH; Carl Roth GmbH, Karlsruhe,Germany), and the absorbance was measured at 405 nm (referencewavelength, 620 nm) using a microplate reader (Berthold Technol-ogies GmbH, Bad Herrenalb, Germany).

2.5. TRAP staining

To confirm the generation of multinucleated osteoclasts, TRAPcytochemical staining was carried out after 28 days (three repli-cates for each magnesium concentration). Adherent cells wererinsed with PBS and then fixed/permeabilized in a solution con-taining 3.7% formaldehyde (Carl Roth GmbH, Karlsruhe, Germany)and 0.2% Triton X-100 (Carl Roth GmbH, Karlsruhe, Germany) inPBS for 5 min at RT. After removal of the fixation/permeabilizationsolution, the cells were dyed to detect acid phosphatase for 10–20 min at 37 �C (staining solution prepared just prior to use:0.1 mg ml–1 naphthol AS-MX phosphate (Sigma-Aldrich ChemieGmbH, Munich, Germany), 0.5 mg ml–1 fast red violet LB salt (Sig-ma-Aldrich Chemie GmbH, Munich, Germany), 10 mM disodiumtartrate dihydrate (Carl Roth GmbH, Karlsruhe, Germany) and40 mM sodium acetate (pH 5.0; Carl Roth GmbH, Karlsruhe, Ger-many)). The staining solution was then removed and replaced withPBS. Red-stained TRAP-positive cells containing at least three nu-clei were scored as osteoclasts. For quantification, the means ofat least five representative fields of view per condition were deter-mined (formulated as the number of positive cells per mm2).

2.6. Cathepsin K and calcitonin receptor immunocytochemistry

After 28 days, CK and CTR were detected via immunocytochem-istry (two replicates for each magnesium concentration). Cellswere washed with PBS, fixed with 3.7% formaldehyde (Carl RothGmbH, Karlsruhe, Germany) for 10 min at RT and then rinsed thor-oughly with PBS. The fixed cells were subsequently permeabilized(only for CK staining) with 0.1% Triton X-100 (Carl Roth GmbH,Karlsruhe, Germany) and 3% H2O2, 1:1 (v/v; Carl Roth GmbH, Kar-lsruhe, Germany), in PBS for 5 min at RT, followed by two PBSwashing steps. Nonspecific binding sites were blocked by immers-ing the cells in 10% (w/v) BSA (Carl Roth GmbH, Karlsruhe,Germany) in PBS for 1 h at 37 �C. Next, the cells were incubatedwith a polyclonal anti-CK or polyclonal anti-CTR primary IgGantibody (1:450 in PBS containing 1% (w/v) BSA; Santa Cruz Bio-technology, Inc., Heidelberg, Germany) for 1 h at 37 �C. After two5 min PBS washing steps, the samples were incubated with theappropriate secondary antibodies (Alexa Fluor 568 Goat Anti-Rab-bit IgG or Alexa Fluor 488 Goat Anti-Rabbit IgG (Life TechnologiesGmbH, Karlsruhe, Germany) and goat anti-rabbit IgG-Texas Red(Santa Cruz Biotechnology, Inc., Heidelberg, Germany) for theanti-CK and anti-CTR primary antibodies, respectively; 1:450 in


PBS containing 1% (w/v) BSA). Staining omitting the primary anti-bodies was carried out as a negative control to identify nonspecificstaining. To visualize cell nuclei, cells were counterstained with1.5 lg ml–1 4-6-diamidino-2 phenylindole (DAPI; Sigma-AldrichChemie GmbH, Munich, Germany) diluted in PBS for 5 min at RT.

2.7. Resorption assay

Osteoclast activity was determined in a lacunar resorptionassay using cells cultured on dentine chips (at least 3 chips percondition). At the end of the culture period (28 days), the cellswere removed with 30% H2O2 (Carl Roth GmbH, Karlsruhe,Germany) and by wiping the surface of the chips with tissue tow-els. To visualize resorption lacunae, dentine chips were treatedwith a 1% toluidine blue (Carl Roth GmbH, Karlsruhe, Germany)solution (in ddH2O) for at least 5 s at RT. The chips were thenrinsed twice with tap water, and the resorbed areas (pits) were de-tected microscopically. Five representative images were randomlycaptured from each dentine piece, and the area of pit resorptionwas quantified using the ImageJ analysis software [21]. In an assayto counteract or minimize the intrinsic artefacts that can be intro-duced by dentine chip variation (e.g. composition), the resorptionactivity (reported as a percentage of the resorbed area) per TRAP-positive cell is presented here.

2.8. Statistical analysis

Statistical analyses were performed using the SigmaStat pack-age (Systat software GmbH, Erkrath, Germany; version 11.0).Standard analyses comparing more than two treatments were con-ducted via one-way repeated measures analysis of variance (ANO-VA). Depending on the data distribution, either one-way repeatedmeasures ANOVA or one-way repeated measures ANOVA on rankswas performed with the Dunn or Holm–Sidak post-hoc test. Statis-tical significance was accepted if the significance level was p < 0.05.

3. Results

3.1. Magnesium extract-containing solutions

To characterize the Mg extract and Mg extract/medium dilu-tions, magnesium and calcium contents were measured, as weretheir pH and osmolalities (Table 1).

3.2. Cell metabolism assay

To study the effect of Mg on cell metabolism, WST-1 assayswere performed after 3, 7, 14, 21 and 28 days of culture. For MgCl2

(Fig. 1A), on days 3 and 7, increasing WST-1 activity was associatedwith increased MgCl2 concentrations. This trend changed on day14, and a bell-shaped pattern was observable, showing a maximumat 5 or 10 mM (statistically significant difference detected between10 and 25 mM at day 21). However (as observed in the protein con-tent results, please refer to Supplemental data and Supplemental



Fig. 1. Cellular metabolic activity. WST-1 measurements from donors treated with MgCl2 (A) and donors treated with magnesium extract (B). For each time point, 0 mM wasset as the control (100%), to which the WST-1 activities for the other concentrations were normalized. Statistical analyses were performed for all MgCl2 concentrations and Mgextract dilutions at each time point (right panel). Symbols indicate significant differences. Grey blocks represent the difference within one time point.


Fig. S1), on day 3, the Mg extract (Fig. 1B) enhanced WST-1 activity(with a statistically significant difference detected between 0 mMand 5�). At the following time points, the Mg extract dilutionsdecreased WST-1 activity (e.g. statistically significant differenceswere detected on days 14, 21 and 28 between the control and5� Mg extract dilution; see Fig. 1B, right panel).

3.3. TRAP activity

Tartrate-resistant acid phosphatase activity is an importantcytochemical marker of osteoclasts. TRAP serum levels are usedas a biochemical marker reflecting osteoclast differentiation.Therefore, extracellular TRAP activity was measured over time tomonitor the effect of Mg on osteoclast activity. TRAP release wascalculated as a percentage of the control release. MgCl2 (Fig. 2A)had almost no influence on TRAP activity (see the statistical anal-yses, Fig. 2A right panel). However, for the Mg extract (Fig. 2B), aclearer trend was observed. A constant increase in activity associ-ated with decreased dilutions was observable over days (e.g. a sta-tistically significant difference was detected on day 14 between thecontrol and 5� Mg extract dilution (statistically significant differ-ences detected between control and 5� on days 7, 14, 21 and 28– control and 10� on days 3, 7, 14 and 21 – control and 30� onday 7 as well as between 30� and 5� on days 7, 14 and 21 –30� and 10� on days 14 and 21; see Fig. 2B right panel).


3.4. TRAP staining

TRAP is released in the surrounding medium but is also found athigh quantities in active osteoclasts. Therefore, TRAP staining wasperformed on day 28, when cells were expected to display all of thehallmarks of fully differentiated cells. For MgCl2 (Fig. 3A), a bell-shaped distribution was recognizable. An increase in the numberof TRAP-positive cells was observed up to the addition of 15 mMMgCl2, followed by a decrease at 25 mM (falling below the control;e.g. a statistically significant difference was detected between the25 mM and 15 mM treatments; see Fig. 3A5). However, in the Mgextract treatments, a decrease in the number of TRAP-positive cellswas observed at decreased dilutions (Fig. 3B; statistically signifi-cant differences between 0 mM and 10� and between 0 mM and5�).

3.5. Cathepsin K immunocytochemistry

This protease, which is predominantly expressed in activeosteoclasts (lysosome), is a well-known osteoclast marker andwas assessed in cyto at day 28. The number of CK-positive cells ini-tially decreased under treatment with 2 mM MgCl2 (statisticallysignificant differences between 2 mM and 0, 5, 10, 15 and25 mM; see Fig. 4A5). The number of positive cells then remainedstable up to 25 mM, when even fewer positive cells were observedthan in the control. For the Mg extract (Fig. 4B), a continuous



Fig. 2. Extracellular TRAP release. Measurements TRAP release for donors treated with MgCl2 (A) and donors treated with magnesium extract (B). For each time point, 0 mMwas set as the control (100%), to which the TRAP release for other concentrations were normalized. Statistical analyses were performed for all MgCl2 concentrations and Mgextract dilutions at each time point (right panel). Symbols indicate significant differences. Grey blocks represent the difference within one time point.


decrease in the CK-positive cell count was observed at decreaseddilutions of the extract (statistically significant decreases betweenthe control (pure cell culture medium) and 30�, 10� and 5� treat-ments; see Fig. 4B5).

3.6. Calcitonin receptor immunocytochemistry

The calcitonin receptor is a high-affinity receptor for the pep-tide hormone calcitonin and contributes to calcium homeostasisand the regulation of osteoclast-mediated bone resorption. Asobserved for CK staining, under treatment with MgCl2, an in-creased magnesium content first decreased (2 mM), then increased(up to 15 mM) and, finally, decreased (25 mM) the number of CTR-positive cells (for statistical significances, see Fig. 5A5). For the Mgextract, no significant variation was observed.

3.7. Resorption assay

To study the effect of magnesium on the osteoclast resorptioncapability, resorption activity was assessed on day 28 on dentinechips. Taking the results (Fig. 6A) together, it can be stated thatresorption activity increased up to a certain MgCl2 concentrationand then decreased (statistically significant differences between10 and 25 mM; see Fig. 6A5). For the magnesium extract,


resorption activity surprisingly increased up to a 5� dilution(statistically significant differences; Fig. 6B5).

4. Discussion

Magnesium chloride and a magnesium extract were chosen tostudy the in vitro effects of high/non-physiological extracellularmagnesium contents on the ability PBMC to form fully differenti-ated and active osteoclasts. PBMC were concomitantly driven alongan osteoclastic pathway by supplementing their culture mediumwith RANKL and M-CSF, which are well-known inducers of osteo-clastogenesis [22,23]. The determining feature of osteoclasts istheir ability to resorb a mineralized matrix [24,25]. It has beenshown previously that it takes at least 3 weeks for human osteo-clasts to differentiate and develop these features under the condi-tions applied here and that the resorption of the surface takes placemainly between days 21 and 28 [20]. Therefore, to study the effectof Mg on this differentiation process, classical osteoclast markers,such as tartrate-resistant acid phosphatase, matrix metalloprotein-ase cathepsin K, calcitonin receptor and resorption pits, wereanalysed over the course of 28 days.

The M-CSF and (especially) RANKL cytokines are key players inosteoclastogenesis. Binding of M-CSF to its cell surface receptor(colony-stimulating factor 1 receptor, or c-Fms) induces cell prolif-eration, survival and the expression of RANK (the RANKL receptor).



Fig. 3. TRAP staining at day 28: (A) MgCl2 and (B) Mg extract assays. (1–3) Micrographs of TRAP staining for dentine (A: 0, 15 and 25 mM, respectively; B: 0 mM, 30� and 5�,respectively; scale bar, 100 lm). (4) The number of TRAP-positive cells recorded for the 0 mM treatment was set as the control (100%), to which the numbers of TRAP-positivecells under the other concentrations (A) or dilutions (B) were normalized. (5) The respective statistical analyses are presented in multiple comparison graphs.


Concomitantly, the RANKL/RANK interaction leads to the recruit-ment of tumour necrosis factor receptor-associated factor 6(TRAF6), which will, in turn, activate the mitogen-activated proteinkinase (MAPK) pathway(s). MAPK then participates in signal trans-duction pathways resulting in downstream activation of transcrip-tion factors such as the nuclear factor of kappa light polypeptidegene enhancer in B-cells (NFjB) and microphthalmia-associatedtranscription factor (MITF) [26]. MITF will then control the


expression of genes involved in osteoclast differentiation such asTRAP and CK [27,28]. TRAF6 recruitment also leads to the produc-tion of reactive oxygen species (ROS) via ras-related C3 botulinumtoxin substrate 1 (Rac1) and NADPH oxidase 1 (Nox1). ROS play apositive regulatory role in triggering the MAPK pathway [29]. Nox-1 is a membrane-localized NADPH oxidase. Plasma membraneNADPH oxidases are also efficient WST-1 reducers [30]. Therefore,WST-1 can be used to assess cell metabolic activity via NADH/



Fig. 4. CK immunocytochemistry at day 28: (A) MgCl2 and (B) Mg extract assays. (1–3) Micrographs of CK staining (A: red, 0, 15 and 25 mM, respectively; B: green, 0 mM, 30�and 5�, respectively; scale bar, 100 lm). Nuclei were DAPI stained and appear blue. (4) The number of CK-positive cells recorded for the 0 mM treatment was set as thecontrol (100%), to which the numbers of CK-positive cells under the other concentrations (A) or dilutions (B) were normalized. (5) The respective statistical analyses arepresented in multiple comparison graphs.


NAD+ (which can be directly linked to cell metabolism/viability). Inthe experiments involving MgCl2, it was observed for all donorsthat a burst of formazan production was detected prior to anincrease in extracellular TRAP activity (as well as total protein con-tent). Based on the WST-1 results, it can be assumed that MgCl2


does not hinder cell viability, as WST-1/formazan production isalways at least in the same range as observed for the controlsamples.

Considering the results regarding total protein contents, TRAPactivity and staining, and CTR and CK immunocytochemistry, it



Fig. 5. CTR immunocytochemistry at day 28: (A) MgCl2 and (B) Mg extract assays. (1–3) Micrographs of CTR (red) staining (A: 0, 5 and 25 mM, respectively; B: 0 mM, 30� and5�, respectively; scale bar 100 lm). Nuclei were DAPI-stained and appear blue. (4) The number of CTR-positive cells recorded for the 0 mM treatment was set as the control(100%), to which the numbers of CTR-positive cells for other concentrations (A) or dilutions (B) were normalized. (5) The respective statistical analyses are presented inmultiple comparison graphs.


appears that, up to a certain concentration, MgCl2 increases theformation of osteoclasts, and further concentration increases thenreverse this phenomenon. Additionally, based on analysis ofresorption activity divided by the number of TRAP-positive osteo-clasts (i.e. resorption per osteoclast), it can be stated again that, up


to a certain MgCl2 concentration, the resorption activity per posi-tive cell is increased, followed by a decrease.

Furthermore, a positive effect of Mg on cell proliferation wasreported in the 1970s [31], and more recently in endothelial andepithelial cells [32,33]. Mg is involved in different steps of the cell



Fig. 6. Resorption assay at day 28: (A) MgCl2 and (B) Mg extract assays. (1–3) Micrographs of dentine resorption (A: 0, 15 and 25 mM, respectively; B: 0 mM, 30� and 5�,respectively; scale bar 200 lm). (4) The resorbed area measured for the 0 mM treatment was set as the control (100%), to which the resorbed areas measured for the otherconcentrations (A) or dilutions (B) were normalized. (5) The respective statistical analyses are presented in multiple comparison graphs.


cycle (e.g. initiation of DNA synthesis and modification of thecytoskeleton [34,35]), and it has been postulated to be a keyregulator (in ‘‘the membrane, magnesium, mitosis (MMM) modelof cell proliferation control’’ [31,35]). Up-regulation of cell cycleinhibitors (e.g. p53) has also been detected during Mg deprivation[9]. The key role of Mg most likely occurs because (i) Mg is an


allosteric modulator of numerous processes and (ii) adenosine tri-phosphate (ATP) must be bound to Mg to be biologically active [36](transphosphorylation is a crucial mechanism in numerousprocesses, such as signal transduction pathways or energy metab-olism). However, this positive effect of Mg is not indefinite, as it isdependent on the available pool of ATP.




It has recently been reported that magnesium deficiency in-creases osteoclast formation but inhibits the activity of these cells[37]. Increased osteoclastogenesis was also observed in the presentstudy, but an increased activity was detected as well. How canthese results be compatible? The beginning of an explanation canbe found by examining TRAP enzyme biochemistry. TRAP contains1.7 mol of magnesium per mol of enzyme, and it has been foundthat 1 mM MgCl2 increases the enzymatic activity of TRAP byapproximately 20% [38]. Moreover, Mg directly increases the num-ber of TRAP-positive cells (chondroclasts and osteoclasts) and pos-itively influences the activity of bone-resorbing cells in vivo[39,40]. However, concomitant treatment with calcitonin abolishesthe effect of Mg.

Furthermore, after extensive studies on the effect magnesiumon cell proliferation, Harry Rubin revealed the central role of mTOR(mechanistic target of rapamycin [35]). mTOR is a phosphatidylin-ositol-3-kinase (PI-3K)-related kinase and plays a role as a centralregulator of cellular metabolism, growth and survival (e.g. regulat-ing lipid synthesis, mitochondrial biogenesis and organization ofthe cytoskeleton). An increase in free Mg2+ will lead to increasedMgATP2� levels and activate mTOR, which displays a high Michae-lis constant (Km) for MgATP2� (i.e. mTOR enzyme kinetics are di-rectly affected by the concentration of the substrate MgATP2�).PI-3K is activated downstream the M-CSF/c-Fms pathway [41].mTOR plays an essential role in osteoclast survival (via M-CSFand RANKL signalling, [42,43]). Increased Mg levels would there-fore lead to a higher survival rate of the cells.

Osteoclast adhesion and migration are two mechanisms in-volved in resorption [44]. Adhesion is essential for the initiationof bone resorption (for formation of the sealing zone), mainly viaav b3 integrin, which is able to recognize matrix proteins such asosteopontin [45]. Osteoclasts express at least two a-integrins (a2and av) and three b-integrins (b1, b3 and b5) [46]. Integrin–sub-strate (or cell-to-cell) binding requires divalent cations [47]. Sev-eral publications have indicated that Mg2+ increases the bindingof integrins to a greater extent than Ca2+, which can even be irre-versible [48,49]. Therefore, primary adhesion may be facilitatedbut can be permanent, thereby inhibiting further osteoclast migra-tion and function.

Considering these arguments together, it can be suggested thatMgCl2 enhances proliferation and osteoclast function up to a cer-tain concentration, which is exactly what we observed for MgCl2.

However, the biphasic effect of MgCl2 was not found for the Mgextract. On the contrary, WST-1, protein contents, TRAP stainingand CK staining were consistently decreased under an increasedMg extract content. However, at a high Mg extract content, theresorption per osteoclast was increased. Additionally, the numberof CTR-positive cells was not significantly affected by the Mg ex-tract. The dissimilar effects measured for MgCl2 and Mg extractare intriguing. Such differences have previously been highlightedin primary human osteoblasts, where extracts have been foundto increase the expression of genes involved in bone metabolism(e.g. osteocalcin) in a more effective manner than MgCl2 (datanot shown). One of the main differences between MgCl2 and theMg extract which could explain these different effects was the cal-cium depletion in the pure Mg extract (see Table 1). The Mg/Ca bal-ance is very important for bones and for homeostasis in general.Circulating Ca2+ plays an important role in bone turnover. WhenCa2+ levels are high, calcitonin (primarily released by the thyroid)will inhibit osteoclast activity in bones. In contrast, when theCa2+ level decreases, calcitriol (the hormonally active form of vita-min D) and parathyroid hormone promote bone resorption in vivo.Rubin and co-workers [50] have extensively studied the effects ofMg2+ and Ca2+ omission, depletion and excess. Omission of Ca2+

from the medium results in a striking increase in cell permeability.Mg2+ can apparently substitute for Ca2+ in maintaining normal per-


meability, but in a less-efficient manner, as 5–10 mM Mg2+ is re-quired to maintain Na2+ and K+ at normal levels in the absence ofCa2+. Ca2+ depletion also causes a shift of Mg2+ away from ATP to-ward other binding sites in the cell [51]. Magnesium may thereforenot exert the effect observed under MgCl2 supplementation. Theeffect of calcium depletion on ATP activity may be followed bymonitoring ATP contents and mitochondrial activity in differentcell culture conditions.

The observed differences could also be explained by the osmo-lality and pH, which were both high (Table 1) and are expected tobe two negative parameters for osteoclast formation and activity(for the two Mg extract donor experiments, 1� and 2� dilutionswere tested, but the cells had all died on day 3). In vitro studieshave highlighted the important roles of pH (i.e. acidic conditionscan stimulate osteoclast activity in vitro) [52,53]. The work ofShibutani and Heersche [54], conducted using osteoclasts fromneonatal rabbits, shows that osteoclast differentiation and prolifer-ation are optimal at pH 7.0–7.5, while at pH 6.5–7 resorptive activ-ity is enhanced. In the present study, the pH ranged from 8.35 to7.8. A negative effect on metabolic activity could be seen in theWST-1 results, whereas protein contents and cell differentiationwere less affected. It has been reported that transient receptor po-tential melastin 7 (TRPM7, an inward transporter of Mg) senses os-motic gradients rather than ionic strength, and hypertonicconditions inhibit TRPM7 [55]. The positive effects of Mg describedabove under MgCl2 treatment would then be abolished (resultingin no or a reduced increase in the intracellular Mg content) untilthe extracellular osmolality was restored. Indeed, in our experi-ments, we also observed that, when osmolality was returned tonormal, osteoclast activity was enhanced. TRPM7 specific knockoutor inhibition (e.g. with non-specific lipoxygenase inhibitors likenordihydroguaiaretic acid or NDGA [56]) may help to uncoverthe role of this channel.

Voltage-dependent Ca channels (VDCC) could also explain thedifferences observed for the two Mg-containing solutions. For cellsof the nervous system, it has been shown that VDCC can be blockedby external magnesium ions [57]. The VDCC are a group of voltage-gated ion channels (activated by changes in electrical potentialnear the channel) that are generally found in excitable cells.Long-lasting calcium channels (L-type) are also found in osteo-clasts and may promote movement [58]. Magnesium blockagewould then explain the increased resorption observed at lowermagnesium concentrations in the case of the Mg extract (com-pared to MgCl2). OC migration pattern measurements could be per-formed to validate this hypothesis.

Another interesting aspect which can be discussed is thein vitro–in vivo correlation. It is conceivable that, during in vivoMg-based implant degradation, a decreasing Mg concentrationgradient exits between the implant surface and the farther tissues.This gradient could explain the bell-shaped distribution (MgCl2)and the constant decrease (Mg extract) measured for the in vitroosteoclastic activity. Therefore a depletion of osteoclasts may ap-pear next to the implant, while osteoclasts may form further away.It may be interesting to analyse in vivo experiments in this respect.In fact, this may partly explain why previous studies reported dif-fering osteoclast behaviours in the peri-implantation site of mag-nesium based implants. On the one hand, for example, Janninget al. [59] described reduced osteoclast surface and fewer osteo-clasts, while on the other hand Huehnerschulte et al. [60] showeda large number of osteoclasts. Another important issue in this re-spect is the coupling of osteoblasts to osteoclasts. It may be possi-ble that, by activation of osteoblasts via an osteoclastic relay, thedeleterious effect of increased osteoclastic activity may bematched or even overcome. To test this hypothesis, co-cultureexperiments will be necessary. Indeed, Witte et al. [17], amongothers, referred to a high mineral apposition rates and increased




bone mass/volume around degrading Mg implants in bone. An-other contributing factor could be the formation of implant parti-cles in the degradation process. It has been shown that particlesof different materials can influence osteoclastogenic cytokine pro-duction [61] and that the reaction of the cells is dependent on thesize of the particles [62].

Finally, the solvent used to prepare extract (i.e. a-MEM supple-mented with 10% FBS) is complex, and the chemical reactions tak-ing place during degradation are even more so. More detailedanalyses of Mg extract and its differences to MgCl2, e.g. by induc-tively coupled plasma mass spectrometry or liquid chromatogra-phy–tandem mass spectrometry, may help to understand theobserved differences in cell behaviour.

With respect to the in vivo degradation of magnesium-basedimplants, magnesium extracts are believed to be the closestin vitro model [18,19].

5. Conclusion

MgCl2 and a magnesium extract exhibited different directeffects on osteoclast proliferation and differentiation. MgCl2 wasable to enhance proliferation and osteoclast function up to aconcentration of approximately 15 mM. The magnesium extractappeared to reduce cell metabolism, while protein contents andcell differentiation were less affected. Both substances activatedosteoclastic resorption activity, but the magnesium extract exertedits positive effect at a lower magnesium content.

6. Disclosure

The authors state that they have no conflicts of interest.

Acknowledgements

The authors wish to sincerely thank colleagues from theDepartment of Plastic Surgery and Hand Surgery (TUM) and theDepartment for Structural Research on Macromolecules (HZG) fortheir generous help in the laboratory and for providing useful guid-ance. Financial support from the CSC council and the HelmholtzAssociation is gratefully acknowledged. Study design: L.W., B.L.,F.F., A.S. and R.W. Study conduct: L.W. Data collection: L.W. Dataanalysis: L.W. and B.L. Data interpretation: L.W. and B.L. Draftingmanuscript: B.L. and L.W. Revising manuscript content andapproving final version of manuscript: L.W., B.L., F.F., A.S. andR.W. B.L. takes responsibility for the integrity of the data analysis.

Appendix A. Figures with essential colour discrimination

Certain figures in this article, particularly Figs. 3–6 are difficultto interpret in black and white. The full colour images can be foundin the on-line version, at http://dx.10.1016/j.actbio.2014.02.010.

Appendix B. Supplementary data

Supplementary data associated with this article can be found, inthe online version, at http://dx.doi.org/10.1016/j.actbio.2014.02.010.

References

[1] Zeng R, Dietzel W, Witte F, Hort N, Blawert C. Progress and challenge formagnesium alloys as biomaterials. Adv Biomater 2008;10:B3–B14.

[2] Peuster M, Beerbaum P, Bach FW, Hauser H. Are resorbable implants about tobecome a reality? Cardiol Young 2006;16:107–16.

[3] Gu X, Zheng Y, Cheng Y, Zhong S, Xi T. In vitro corrosion and biocompatibilityof binary magnesium alloys. Biomaterials 2009;30:484–98.


[4] Peacock M. Calcium metabolism in health and disease. Clin J Am Soc Nephrol2010;5(Suppl. 1):S23–30.

[5] Wallach S. Availability of body magnesium during magnesium deficiency.Magnesium 1988;7:262–70.

[6] Altura BM. Basic biochemistry and physiology of magnesium: a brief review.Magnes Trace Elem 1991;10:167–71.

[7] Rude RK, Shils ME. Magnesium. In: Shils ME, editor. Modern nutrition in healthand disease. Philadelphia, PA: Lippincott, Williams & Wilkins; 2006.

[8] Bygrave FL. Cellular calcium and magnesium metabolism. In: Williams DL,editor. An introduction to bio-inorganic chemistry. Thomas; 1976. p. 171–84.

[9] Wolf FI, Trapani V. Cell (patho)physiology of magnesium. Clin Sci2008;114:27–35.

[10] Schroll A. Advances in magnesium research. London: John Libbey & Co.; 2002.[11] Durlach J, Durlach V, Bac P, Bara M, Guiet-Bara A. Magnesium and

therapeutics. Magnes Res 1994;7:313–28.[12] Cunningham J, Rodriguez JM, Messa P. Magnesium in chronic kidney disease

stages 3 and 4, and in dialysis patients. Clin Kidney J 2012;5:i39–51.[13] Herzog P, Holtermuller KH. Antacid therapy – changes in mineral metabolism.

Scand J Gastroenterol Suppl 1982;75:56–62.[14] Xing JH, Soffer EE. Adverse effects of laxatives. Dis Colon Rectum

2001;44:1201–9.[15] Zhang Z, Egaña JT, Reckhenrich AK, Schenck TL, Lohmeyer JA, Schantz JT, et al.

Cell-based resorption assays for bone graft substitutes. Acta Biomater2012;8:13–9.

[16] Staiger MP, Pietak AM, Huadmai J, Dias G. Magnesium and its alloys asorthopedic biomaterials: a review. Biomaterials 2006;27:1728–34.

[17] Witte F, Kaese V, Haferkamp H, Switzer E, Meyer-Lindenberg A, Wirth CJ, et al.In vivo corrosion of four magnesium alloys and the associated bone response.Biomaterials 2005;26:3557–63.

[18] 10993-5:2009 I. Biological evaluation of medical devices. Part 5. Tests forin vitro cytotoxicity 2009.

[19] 10993-12:2012 I. Biological evaluation of medical devices. Part 12. Samplepreparation and reference materials 2012.

[20] Schilling AF, Linhart W, Filke S, Gebauer M, Schinke T, Rueger JM, et al.Resorbability of bone substitute biomaterials by human osteoclasts.Biomaterials 2004;25:3963–72.

[21] Collins TJ. ImageJ for microscopy. Biotechniques 2007;43:25–30.[22] Arai F, Miyamoto T, Ohneda O, Inada T, Sudo T, Brasel K, et al. Commitment and

differentiation of osteoclast precursor cells by the sequential expression of c-Fms and receptor activator of nuclear factor kappaB (RANK) receptors. J ExpMed 1999;190:1741–54.

[23] Suda T, Kobayashi K, Jimi E, Udagawa N, Takahashi N. The molecular basis ofosteoclast differentiation and activation. Novartis Found Symp2001;232:235–47.

[24] Winkler T, Hoenig E, Gildenhaar R, Berger G, Fritsch D, Janssen R, et al.Volumetric analysis of osteoclastic bioresorption of calcium phosphateceramics with different solubilities. Acta Biomater 2010;6:4127–35.

[25] Zhang Z, Egana JT, Reckhenrich AK, Schenck TL, Lohmeyer JA, Schantz JT, et al.Cell-based resorption assays for bone graft substitutes. Acta Biomater2012;8:13–9.

[26] Nakashima T, Hayashi M, Takayanagi H. New insights into osteoclastogenicsignaling mechanisms. Trends Endocrinol Metab 2012;23:582–90.

[27] Mansky KC, Sulzbacher S, Purdom G, Nelsen L, Hume DA, Rehli M, et al. Themicrophthalmia transcription factor and the related helix–loop–helix zipperfactors TFE-3 and TFE-C collaborate to activate the tartrate-resistant acidphosphatase promoter. J Leukoc Biol 2002;71:304–10.

[28] Troen BR. The regulation of cathepsin K gene expression. Ann NY Acad Sci2006;1068:165–72.

[29] Lee NK, Choi YG, Baik JY, Han SY, Jeong DW, Bae YS, et al. A crucial role forreactive oxygen species in RANKL-induced osteoclast differentiation. Blood2005;106:852–9.

[30] Berridge MV, Tan AS. Trans-plasma membrane electron transport: a cellularassay for NADH- and NADPH-oxidase based on extracellular, superoxide-mediated reduction of the sulfonated tetrazolium salt WST-1. Protoplasma1998;205:74–82.

[31] Rubin H. Central role for magnesium in coordinate control of metabolism andgrowth in animal cells. Proc Natl Acad Sci USA 1975;72:3551–5.

[32] Sgambato A, Faraglia B, Ardito R, Torsello A, Boninsegna A, Cittadini A, et al.Isolation of normal epithelial cells adapted to grow at nonphysiologicalconcentration of magnesium. Biochem Biophys Res Commun 2001;286:752–7.

[33] Maier JA, Bernardini D, Rayssiguier Y, Mazur A. High concentrations ofmagnesium modulate vascular endothelial cell behaviour in vitro. BiochimBiophys Acta 2004;1689:6–12.

[34] Walker GM. Magnesium and cell cycle control: an update. Magnesium1986;5:9–23.

[35] Rubin H. The membrane, magnesium, mitosis (MMM) model of cellproliferation control. Magnes Res 2005;18:268–74.

[36] Saylor P, Wang C, Hirai TJ, Adams JA. A second magnesium ion is critical forATP binding in the kinase domain of the oncoprotein v-Fps. Biochemistry1998;37:12624–30.

[37] Belluci MM, Schoenmaker T, Rossa-Junior C, Orrico SR, de Vries TJ, Everts V.Magnesium deficiency results in an increased formation of osteoclasts. J NutrBiochem 2013;24:1488–98.

[38] Hayman AR, Warburton MJ, Pringle JA, Coles B, Chambers TJ. Purification andcharacterization of a tartrate-resistant acid phosphatase from humanosteoclastomas. Biochem J 1989;261:601–9.


http://dx.10.1016/j.actbio.2014.01.025



http://refhub.elsevier.com/S1742-7061(14)00062-2/h0005


























































































[39] Marie PJ, Hott M. Effect of calcitonin on the magnesium-induced boneresorption in the mouse. Magnesium 1987;6:100–8.

[40] Marie PJ, Travers R, Delvin EE. Influence of magnesium supplementation onbone turnover in the normal young mouse. Calcif Tissue Int 1983;35:755–61.

[41] Shinohara M, Nakamura M, Masuda H, Hirose J, Kadono Y, Iwasawa M, et al.Class IA phosphatidylinositol 3-kinase regulates osteoclastic bone resorptionthrough protein kinase B-mediated vesicle transport. J Bone Miner Res2012;27:2464–75.

[42] Sugatani T, Hruska KA. Akt1/Akt2 and mammalian target of rapamycin/Bimplay critical roles in osteoclast differentiation and survival, respectively,whereas Akt is dispensable for cell survival in isolated osteoclast precursors. JBiol Chem 2005;280:3583–9.

[43] Glantschnig H, Fisher JE, Wesolowski G, Rodan GA, Reszka AA. M-CSF,TNFalpha and RANK ligand promote osteoclast survival by signaling throughmTOR/S6 kinase. Cell Death Differ 2003;10:1165–77.

[44] Novack DV, Faccio R. Osteoclast motility: putting the brakes on boneresorption. Ageing Res Rev 2011;10:54–61.

[45] Holt I, Marshall MJ. Integrin subunit beta3 plays a crucial role in themovement of osteoclasts from the periosteum to the bone surface. J CellPhysiol 1998;175:1–9.

[46] Minkin C, Marinho VC. Role of the osteoclast at the bone-implant interface.Adv Dent Res 1999;13:49–56.

[47] Shankar G, Davison I, Helfrich MH, Mason WT, Horton MA. Integrin receptor-mediated mobilisation of intranuclear calcium in rat osteoclasts. J Cell Sci1993;105(1):61–8.

[48] Takeichi M, Okada TS. Roles of magnesium and calcium ions in cell-to-substrate adhesion. Exp Cell Res 1972;74:51–60.

[49] Elices MJ, Urry LA, Hemler ME. Receptor functions for the integrin VLA-3:fibronectin, collagen, and laminin binding are differentially influenced by Arg-Gly-Asp peptide and by divalent cations. J Cell Biol 1991;112:169–81.

[50] Rubin AH, Terasaki M, Sanui H. Major intracellular cations and growth control:correspondence among magnesium content, protein synthesis, and the onsetof DNA synthesis in BALB/c3T3 cells. Proc Natl Acad Sci USA 1979;76:3917–21.

[51] Bowen-Pope DF, Vidair C, Sanui H, Rubin AH. Separate roles for calcium andmagnesium in their synergistic effect on uridine uptake by cultured cells:significance for growth control. Proc Natl Acad Sci USA 1979;76:1308–12.


[52] Arnett TR. Extracellular pH regulates bone cell function. J Nutr2008;138:415S–8S.

[53] Goldhaber P, Rabadjija L. H+ stimulation of cell-mediated bone resorption intissue culture. Am J Physiol 1987;253:E90–8.

[54] Shibutani T, Heersche JN. Effect of medium pH on osteoclast activity andosteoclast formation in cultures of dispersed rabbit osteoclasts. J Bone MinerRes 1993;8:331–6.

[55] Bessac BF, Fleig A. TRPM7 channel is sensitive to osmotic gradients in humankidney cells. J Physiol 2007;582:1073–86.

[56] Chen H-C, Xie J, Zhang Z, Su L-T, Yue L, Runnels LW. Blockade of TRPM7channel activity and cell death by inhibitors of 5-lipoxygenase. PLoS One2010;5:e11161.

[57] Hartzell HC, White RE. Effects of magnesium on inactivation of the voltage-gated calcium current in cardiac myocytes. J Gen Physiol 1989;94:745–67.

[58] Miyauchi A, Hruska KA, Greenfield EM, Duncan R, Alvarez J, Barattolo R, et al.Osteoclast cytosolic calcium, regulated by voltage-gated calcium channels andextracellular calcium, controls podosome assembly and bone resorption. J CellBiol 1990;111:2543–52.

[59] Janning C, Willbold E, Vogt C, Nellesen J, Meyer-Lindenberg A, Windhagen H,et al. Magnesium hydroxide temporarily enhancing osteoblast activity anddecreasing the osteoclast number in peri-implant bone remodelling. ActaBiomater 2010;6:1861–8.

[60] Huehnerschulte TA, Reifenrath J, von Rechenberg B, Dziuba D, Seitz JM,Bormann D, et al. In vivo assessment of the host reactions to thebiodegradation of the two novel magnesium alloys ZEK100 and AX30 in ananimal model. Biomed Eng Online 2012;11:14.

[61] Lange T, Schilling AF, Peters F, Haag F, Morlock MM, Rueger JM, et al.Proinflammatory and osteoclastogenic effects of beta-tricalciumphosphateand hydroxyapatite particles on human mononuclear cells in vitro.Biomaterials 2009;30:5312–8.

[62] Lange T, Schilling AF, Peters F, Mujas J, Wicklein D, Amling M. Size dependentinduction of proinflammatory cytokines and cytotoxicity of particulate beta-tricalciumphosphate in vitro. Biomaterials 2011;32:4067–75.








































































Documents

Effects of extracellular magnesium on the differentiation and function of human osteoclasts