Magnesium prevents vascular calcification in vitro by inhibition of hydroxyapatite crystal formation

Magnesium has been shown to effectively prevent vascular calcification associated with chronic kidney disease. Magnesium has been hypothesized to prevent the upregulation of osteoblastic genes that potentially drives calcification. However, extracellular effects of magnesium on hydroxyapatite formation are largely neglected. This study investigated the effects of magnesium on intracellular changes associated with transdifferentiation and extracellular crystal formation. Bovine vascular smooth muscle cells were calcified using β-glycerophosphate. Transcriptional analysis, alkaline phosphatase activity and detection of apoptosis were used to identify transdifferentiation. Using X-ray diffraction and energy dispersive spectroscopy extracellular crystal composition was investigated. Magnesium prevented calcification in vascular smooth muscle cells. β-glycerophosphate increased expression of osteopontin but no other genes related to calcification. Alkaline phosphatase activity was stable and apoptosis was only detected after calcification independent of magnesium. Blocking of the magnesium channel TRPM7 using 2-APB did not abrogate the protective effects of magnesium. Magnesium prevented the formation of hydroxyapatite, which formed extensively during β-glycerophosphate treatment. Magnesium reduced calcium and phosphate fractions of 68% and 41% extracellular crystals, respectively, without affecting the fraction of magnesium. This study demonstrates that magnesium inhibits hydroxyapatite formation in the extracellular space, thereby preventing calcification of vascular smooth muscle cells.

Experimental design. For calcification, medium was supplemented with 5% (v/v) FBS and 10 mM β-glycerophosphate (BGP, Merck Millipore, Massachusetts, USA). BGP requires cellular activity for its cleavage to free Pi, and was chosen in this setup to increase the medium Pi concentration to minimize Mg 2+ -Pi interactions prior to cellular exposure. In the high Mg 2+ treatment medium, MgCl 2 (Merck Millipore) was supplemented to reach a final concentration of 2 mM MgCl 2 . At 80% confluence, cells grown in a 12-wells plate were incubated with designated media for 14 days, which was changed every 2-3 days.
TRPM7 inhibition by 2-APB. Transient receptor potential melastatin 7 (TRPM7) was inhibited by incubation with 10 μM of 2-Aminoethyl diphenylborinate (2-APB, Sigma). At 80% confluence, 2-APB was supplemented in combination with the different culture media as outlined in the previous section during 14 days in a 24-wells plate.
Pi concentration of the cell culture supernatant. Pi was measured in the cell culture supernatant (or culture medium) calorimetrically using the malachite green method as described elsewhere 23 . Briefly, a reaction mix consisting of molybdate and malachite green was added to the samples and standards and incubated for 30 minutes at room temperature. The absorbance was measured at 620 nm using a Benchmark Plus Microplate Spectrophotometer System (Bio-Rad, California, USA).
Quantification of Ca 2+ deposition. Cells were decalcified with 0.1 M HCl for 5 minutes at room temperature with gentle rocking, which effectively dissolved all Ca 2+ deposits present. The Ca 2+ concentration in the supernatant was determined by the o-cresophthalein complexone method. o-cresophthalein color reagent (Sigma) was incubated with the samples and standards and the absorbance was measured immediately at 570 nm, as described previously 24 . Subsequently, the cells were neutralized in PBS and lysed in 0.1 M NaOH/0.1% (w/v) sodium dodecyl sulfate for total protein isolation. Ca 2+ concentrations were normalized for total protein as determined by Pierce BCA protein detection kit according to the manufacturer's instructions (Fisher Scientific, Massachussets, USA). Alizarin Red staining. Calcification  RNA isolation and real-time polymerase chain reaction (RT-qPCR). Total RNA was extracted from VSMC using TRIzol (Invitrogen) and treated with DNAse (1 U/μg RNA, Promega, Wisconsin, USA) to remove genomic DNA. cDNA was synthesized from 1.5 μg total RNA by Moloney Murine Leukemia Virus reverse transcriptase (Invitrogen) for one hour at 37 °C. The primers used for PCR amplification are shown in Table 1 and were equally efficient. RT-qPCR was executed in duplicate using IQ SYBRGreen Mix according to the manufacturers protocol (Bio-Rad), using a Bio-Rad thermos-cycler (Bio-Rad). The expression of target genes was normalized to GAPDH expression levels.
Alkaline phosphatase activity assay. Cells grown in 12-well plates and cultured in designated media for 2, 8 and 14 days were lysed with 1% (v/v) Triton X-100 in PBS containing protease inhibitors. ALP activity was determined in the total lysate as the hydrolysis of p-nitrophenyl phosphate (sigma) into p-nitrophenol in a basic buffer by ALP by p-nitrophenol production. The reaction was incubated for 30 minutes at 37 °C and the absorbance for p-nitrophenol was measured at calorimetrically at 410 nm for both p-nitrophenol (Sigma) standards and samples. One unit (U) was defined as the production of 1 µmol p-nitrophenol per minute per gram protein. immunofluorescence microscopy using a FITC and Texas Red filter on an Axio Imager M (Zeiss). Apoptosis was quantified using ImageJ software (NIH, Maryland, USA) by calculating the ratio of the area positive for FITC signal versus total area, as the mean of multiple captures in 3 replicates per treatment per time-point.

Detection of apoptosis.
Crystal isolation for X-ray diffraction, scanning electron microscopy and energy-dispersive spectroscopy analysis. Cell culture supernatants of BGP-treated cells in 6-well plates with and without 2 mM MgCl 2 were collected and purified. Supernatants were centrifuged for one hour at 16 000 × g 19 . The pellets containing the nanocrystals were washed with demineralized H 2 O and then re-centrifuged. Subsequently, the crystal pellets were dried and used for analysis. One measurement represents the crystals formed in a total of 7 wells containing 2 ml of culture medium in the BGP treated cultures. As less material was formed in the BGP cultures supplemented with Mg 2+ , one measurement represents the crystals formed in a total of 21 wells containing 2 ml of culture medium in order to reach sufficient amounts to detect by X-ray diffraction. For X-ray diffraction analysis, diffractograms were measured on a PANalytical Empyrean (PANalytical, Almelo, the Netherlands) in transmission mode with fine-focus sealed tube, focusing mirror and PIXcel3D detector, using CuKα radiation. The samples were measured in a capillary, using 0.5 mm soda glass capillaries with a wall thickness of 0.01 mm. For scanning electron microscopy (SEM) (GeminiSEM, Zeiss) in combination with energy-dispersive spectroscopy (EDX) for elemental analysis (QUANTAX 200, Bruker) the crystal pellets were transferred onto copper tape and coated with carbon. High-resolution pictures were obtained using an Everhart-Thornley SE detector. Accelerating voltage was 5 kV for morphological observations and 15 kV for micro-elemental analyses. Due to the use of BGP as calcification inducer, a cell-free control could not be included as cellular presence is necessary to cause Pi accumulation in the medium (data not shown).
Statistics. Parametric data were analyzed by One-Way ANOVA with Tukey's post-hoc test to correct for multiple comparisons using PRISM software (GraphPad, San Diego, CA). Non-parametric data as identified by Shapiro-Wilk test for normality were analyzed using Kruskall-Wallis analysis with Dunn's correction for multiple analysis. Time-course data was analyzed using a Two-Way ANOVA. All data are shown as mean ± SEM. P < 0.05 was considered statistically significant.

β-glycerophosphate supplementation resulted in increased medium Pi concentration. BGP
is a Pi-donor that requires enzymatic cleavage in order to release Pi. As Pi exposure is one of the decisive factors in the calcification process, the Pi concentration of the cell culture supernatant was assessed after 2, 8 and 14 days of treatment ( Fig. 1). 10 mM BGP treatment resulted in gradual increase in Pi concentration over time, reaching 4.6 ± 0.3 mM after 14 days. 2 mM Mg 2+ supplementation led to significantly higher Pi concentrations of 7.6 ± 0.8 mM after 14 days. In contrast, BGP treatment in cell-free conditions did not lead to increased Pi concentrations under the same conditions (data not shown).
Mg 2+ prevents vascular smooth muscle cell mineralization. The effect of BGP on the development of calcifications was studied using cellular Ca 2+ measurements and visualized by Alizarin Red staining (Fig. 2). 10 mM BGP treatment of bVSMC resulted in variable but pronounced calcification after 14 days (146 ± 93 versus 2.5 ± 0.2 µg Ca 2+ per gram protein in the control condition, p < 0.05). Ca 2+ deposition was completely prevented by 2 mM Mg 2+ in our model (3.5 ± 0.3 µg/g protein Ca 2+ ).

β-glycerophosphate supplementation upregulated OPN gene expression but did not result in changes in mRNA expression of calcification activators.
To assess the effects of Mg 2+ involved in the prevention of VSMC calcification, gene expression levels of ACTA2 and osteogenic transcription factors RUNX2 and BMP2 were assessed after 2, 8 and 14 days (Fig. 3a-c). 10 mM BGP treatment did not result in expression changes of ACTA2, RUNX2 and BMP2 over the time course of calcification. 2 mM Mg 2+ supplementation did not affect these expression levels. Both ACTA2 and BMP2 mRNA expression significantly increased over time in all conditions, independent of BGP treatment (Fig. 3a-c). In addition, mRNA expression levels of calcification inhibitors OPG, OPN and MGP were assessed ( Fig. 3d-f). OPG expression was significantly downregulated by Mg 2+ supplementation compared all other treatments after 8 days. No effect of BGP was observed at all time-points. MGP gene expression showed a significant increase over time in all treatment groups, while no effect of BGP was Gene (Bos Tauros) Forward primer sequence Reverse primer sequence  (Fig. 5a). Furthermore, 10 mM BGP treatment did not affect TRPM7 mRNA expression levels in cultured VSMC (Fig. 5b).   To investigate the potential role of Mg 2+ in crystal growth and formation, the cell culture supernatants of BGP-treated cells in the presence or absence of Mg 2+ were analyzed for the incidence of crystals using X-ray powder diffraction. In the BGP-treated samples, the X-ray diffraction patterns revealed the presence of a considerable amount of hydroxyapatite crystals (Fig. 6). The broadening of the hydroxyapatite diffraction peaks, compared to the NaCl peaks, indicate that the hydroxyapatite crystals are nano-sized. Both crystals isolated from the cell culture supernatants and a synthetic hydroxyapatite standard, that was used as positive control, matched with a reference diffraction pattern specific for hydroxyapatite crystals (see Supplemental Fig. 2). Hydroxyapatite diffraction peaks were absent in the Mg 2+ -supplemented BGP supernatants. Of note, no crystals other than hydroxyapatite and NaCl were identified. As X-ray diffraction exclusively detects crystalline material and not amorphous material, isolated particles were analyzed by SEM-EDX for morphology and elemental composition (Fig. 7). EDX analysis revealed a reduced Ca 2+ and Pi fraction in crystal clusters of 68% and 41% after Mg 2+ supplementation, respectively, without increasing the fraction of Mg 2+ present in the crystal clusters (Fig. 7b,c).

Discussion
Here, we demonstrate that Mg 2+ inhibits bVSMC mineralization through inhibition of Ca-apatite formation in the extracellular space, independent of VSMC transdifferentiation. Our most important finding is the absence of  hydroxyapatite crystals in the medium of BGP-treated bVSMCs supplemented with Mg 2+ . Characterization by SEM-EDX confirmed the reduction of Ca-apatite crystals in Mg 2+ -supplemented supernatants, without incorporation of Mg 2+ in the formed crystals. Underlining the strong capacity of Mg 2+ to block crystal formation, 2 mM Mg 2+ was sufficient to prevent the calcification process, even though extracellular Pi levels rise to 7 mM in Mg 2+ -BGP treated bVSMCs. Moreover, intracellular action of Mg 2+ is not likely in our setup, because when cellular Mg 2+ uptake was impaired using TRPM7-blocker 2-APB, Mg 2+ still prevented vascular calcification 11 . Accordingly, we propose that Mg 2+ prevents VSMC mineralization through the inhibition of hydroxyapatite formation in the extracellular space, blocking its deposition on VSMC. Hydroxyapatite (Ca 10 (PO 4 ) 6 (OH) 2 ) is the most abundant type of crystal in uremic arterial calcifications and its formation has been shown to be essential for VSMC transdifferentiation and vascular calcification 19,27 . Although it has been suggested that the potential incorporation of Mg 2+ in Ca-apatite crystals (whitlockite, Ca 9 Mg(HPO 4 ) (PO 4 ) 6 ), may reduce crystal pathogenicity and increase solubility, we did not identify any whitlockite after Mg 2+ treatment. Our data suggest that Mg 2+ most likely prevents crystal nucleation, rather than affecting crystal content. These findings are in line with previous studies that exclusively identified hydroxyapatite, and not whitlockite, in deposits on calcifying VSMC supplemented with Mg 2+ 28 . Indeed, high concentrations of Mg 2+ led to less hydroxyapatite deposition in a study by Louvet et al. 29 . Given the lack of crystals after Mg 2+ supplementation, our results indicate that Mg 2+ inhibits the early phases of crystal assembly in high Pi-media. Moreover, we hypothesize that the incorporation of Pi in hydroxyapatite in the BGP condition explains the lower free Pi concentration in BGP-treated compared to the Mg 2+ -BGP-treated culture media. Initial Ca-Pi particle formation in response to elevated Pi-levels has shown to occur in a cell-independent manner, subsequently initiating VSMC transdifferentiation when native VSMC inhibitory capacities diminish 19,30 . Though poorly studied in the context of VSMC mineralization, Mg 2+ is known to stabilize amorphous Ca-Pi particles and therefore inhibit Ca-apatite maturation in acellular systems [31][32][33][34][35][36] . While the exact mechanisms remain unknown, evidence suggests that Mg 2+ may stabilize extracellular ATP. Hydrolysis of ATP is necessary for hydroxyapatite nucleation 21 . The importance of crystal maturation in the initiation of VSMC transdifferentiation and vascular calcification has been frequently emphasized 18,19,30,37 . Recently, Ca-Pi-containing soluble nanoparticles or calciprotein particles (CPP) were shown to stimulate calcification 37 . Interestingly, Mg 2+ delays CPP maturation in uremic serum 38 . These findings support that Mg 2+ prevents mineralization by directly inhibiting Ca-apatite crystal formation or maturation in the extracellular space.
Mg 2+ supplementation has repeatedly shown to prevent osteogenic gene expression. As a result, osteogenic gene expression has been repeatedly considered as Mg 2+ target to prevent osteoblastic transdifferentiation 39 . In line with this hypothesis, previous research showed that Mg 2+ concentrations as low as 0.8 mM reversed established calcification in human VSMC, which could be abrogated by 2-APB treatment 17 . These results suggest that cellular Mg 2+ uptake via TRPM7 prevents VSMC calcification 7,11,17 . However the role of TRPM7 is controversial, as recent evidence suggests that interleukin-18 enhanced VSMC calcification through TRPM7 activation 40 . In our model, TRPM7 inhibition by 2-APB did not affect the Mg 2+ rescue.
bVSMC are characterized by high basal expression levels of ALP, which makes them prone to calcification. Despite their susceptibility to calcify, the bVSMCs are contractile and do not present any signs of osteoblastic transdifferentiation, as high levels of ACTA2 expression and α-SMA protein expression were preserved in response to BGP supplementation. Although our bVSMCs strongly calcified, BGP treatment did not result in osteogenic conversion as demonstrated by stable expression of BMP2, RUNX2 and ALPL among treatments 14 . Interestingly, both mRNA expression of BMP2 and ACTA2 increased over time. However, these observations were irrespective of treatment and are therefore not related to osteoblastic transdifferentiation of the bVSMC. The only transcriptional response observed during BGP-induced calcification was upregulation of the OPN gene after 14 days, which was prevented by Mg 2+ . Increased OPN expression is associated with calcification 19,41,42 . OPN is an inhibitor of calcification and potently inhibits hydroxyapatite growth and OPN upregulation has been shown to reflect a protective mechanism in response to the phosphate-and hydroxyapatite-rich environment by VSMC [43][44][45] . The absence of OPN upregulation in Mg 2+ -supplemented BGP cultures may therefore be explained by the lack of Ca-Pi formation. Moreover, OPN is only increased at 14 days after calcification was already manifested, suggesting it to be resulting from calcification rather than causing. In addition to osteoinductive signaling, apoptosis has been shown to induce the progression of calcification 25 . Our results indicate that calcification precedes apoptosis, as apoptosis was only detected after 14 days of BGP treatment when calcification was already manifested. In our setup, apoptosis is likely the result of exposure to Ca-Pi crystals, rather than a causative factor for calcification 46 .
In human, rodent and bovine calcification models evidence strongly suggests that calcification is a result of VSMC undergoing osteogenic transdifferentiation and that Mg 2+ effectively abrogates this through upregulation of calcification inhibitors and downregulation of osteogenic genes [6][7][8]11,16,17,47,48 . Indeed, we show the effective inhibition of Mg 2+ in VSMC calcification. However, in contrast to previous studies, our results suggest that calcification is driven by extracellular hydroxyapatite formation independent of osteogenic transdifferentiation in bVSMCs. While many studies show the association between osteogenic transdifferentiation and vascular calcification, it remains debatable whether this transdifferentiation is an undisputable prerequisite for the development of mineralization 49 . Calcification represents the final common pathway of multiple pathological vascular processes 50 . Our results do not contradict intracellular Mg 2+ effects on osteoblastic transdifferentation. However, they do highlight the presence of alternative extracellular effects on crystal formation. Overall, it is important to note that potential intracellular and extracellular pathways involved in the calcification-inhibiting capacity of Mg 2+ are not mutually exclusive. Given the strong effect of Mg 2+ on calcification independent of osteogenic pathways, however, the relative contribution of crystal inhibition compared to any intracellular targets may be considerable and underestimated to date.
An important strength of this study is that our model favored to study the effects of Mg 2+ on extracellular crystal growth, independent of genetic VSMC transdifferentiation. Although it has been reported previously that Mg 2+ inhibits calcification, most study set-ups do not allow to discriminate between extracellular reduction of crystal formation and intracellular inhibition of osteogenic conversion [6][7][8]11,17 . A limitation of this study is that while we show that TRMP7 seems not to be involved in calcification, we cannot exclude that other Mg 2+ channels than TRPM7 facilitate Mg 2+ entry into the bVSMCs. However, it was shown previously that TRPM7 is the main Mg 2+ channel in VSMCs 26 . Therefore, the contribution of other transporters is likely minor. In addition, we show the effectiveness of Mg 2+ primarily through extracellular mechanisms involving hydroxyapatite. In contrast to other studies, the driving force of calcification in our model is not osteoblastic transdifferentiation but mainly hydroxyapatite formation and deposition. Therefore, any intracellular effects of Mg 2+ involving modification of osteogenic genes such as BMP2 and RUNX2 cannot be excluded.
In conclusion, our findings demonstrate a role for Mg 2+ in preventing VSMC mineralization involving direct extracellular Ca-apatite crystal inhibition. An increasing body of studies now report that Mg 2+ prevents vascular calcification by extracellular Pi binding. Mg 2+ has been shown to reduce both vascular and non-vascular calcifications and to improve calcification propensity, favoring a non-cellular mechanism of action 38,51 . Therefore, Mg 2+ may be considered an important and realistic approach to potentially reduce the risk for vascular calcification and subsequent cardiovascular complications in CKD patients. Clinical trials are warranted to further assess the clinical relevance of Mg 2+ in relation to vascular calcifications.