SEARCH

SEARCH BY CITATION

Keywords:

  • adipose-derived stem cells;
  • bioactive glass;
  • mesenchymal stem cells;
  • osteogenic differentiation;
  • submicron particles

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discussion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

Spherical monodispersed bioactive particles are potential candidates for nanocomposite synthesis or as injectable particles that could be internalized by cells for the local sustained delivery of inorganic therapeutic ions (e.g., calcium or strontium). Particles are also likely to be released from porous bioactive glass and sol–gel hybrid scaffolds as they degrade; thus, it is vital to investigate their interaction with cells. Spherical monodispersed bioactive glass particles (mono-SMBG), with diameters of 215 ± 20 nm are synthesized using a modified Stöber process. Confocal and transmission electron microscopy demonstrate that mono-SMBGs are internalized by human bone marrow (MSCs) and adipose-derived stem cells (ADSCs) and located within cell vesicles and in the cytoplasm. Particle dissolution inside the cells is observed. Alamar Blue, MTT and Cyquant assays demonstrate that 50 μg mL−1 of mono-SMBGs did not inhibit significantly MSC or ADSC metabolic activity. However, at higher concentrations (100 and 200 μg mL−1) small decrease in metabolic activity and total DNA is observed. Mono-SMBG did not induce ALPase activity, an early marker of osteogenic differentiation, without osteogenic supplements; however, in their presence osteogenic differentiation is achieved. Additionally, large numbers of particles are internalized by the cells but have little effect on cell behavior.


1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discussion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

Spherical and monodisperse bioactive glass particles have great potential for therapeutic applications1 and for incorporation into nanocomposite medical devices, but it is important to have control of their synthesis and to understand how they interact with cells.

Due to donor site morbidity and other complications that arise from autograft operations, there have been concerted efforts to develop synthetic bone grafts that can reduce the need for transplants. Porous biodegradable scaffolds are needed that can act as temporary templates (scaffolds) for bone regeneration. The scaffolds should be able to share load with the host bone and stimulate new bone growth (bioactivity). Most current commercial synthetic bioactive bone grafts are bioceramics such as synthetic hydroxyapatite or other calcium phosphates.2 However, all bioceramics share the same problem: they are brittle, leaving autograft as the gold standard surgical procedure. To improve toughness there is currently a shift from bioceramics to nanocomposites and hybrid materials3–5 that more closely mimic the nanostructure of bone, which in materials science terms is a nanocomposite of collagen fibrils and nanocrystals of biological hydroxyapatite. A bioactive nanocomposite can be made by dispersion of bioactive nanoparticles in a biodegradable polymer matrix6 and hybrids can be considered as interpenetrating networks of inorganic and organic components that are not distinguishable at the microscale.3, 4, 7 The choice of bioactive component could be bioactive glass, which has been shown to stimulate more bone growth in vivo than calcium phosphate-based ceramics,8 due to the release of soluble silica species and calcium ions that stimulate osteoblasts to produce bone matrix.9

Bioactive glass microparticles are used in a wide range of orthopaedic applications (e.g., BonAlive [BonAlive, Finland] and NovaBone [NovaBone Products LLC, USA]). Monodispersed submicron bioactive glass particles (mono-SMBG) are an attractive alternative to microparticles as their small size and higher specific surface area makes them ideal for injection into the bone defects or incorporation into composite scaffolds.

While producing spherical and monodispersed silica particles is now routine,10–14 synthesizing bioactive glass (in the SiO2–CaO system) submicron particles of controlled morphology is nontrivial15 as adding calcium causes the particles to become irregular.16 In our previous work, we synthesized spherical submicron particles with a composition of 85 mol% SiO2, 15 mol% CaO16 with a broad particle size distribution (150–350 nm). The synthesis method did not allow the addition of more calcium. The first aim of this work was to synthesize mono-SMBGs with higher CaO content. Monodispersity is required for controlled nanocomposite synthesis or for an injectable product.

Scaffolds made from bioactive glass sol–gel foams,17 nanocomposites or hybrids are all likely to release particles during degradation, therefore the second aim was to investigate cellular response to the mono-SMBGs.

Much research has been carried out on the use of silica nanoparticles as delivery vehicles for organic drugs.18 However, inorganic ions have also been found to have therapeutic properties. An example is bioactive glass that releases strontium ions that can inhibit osteoclast activity, which is of benefit to osteoporotic patients.19 Therefore, particles that are internalized by cells and do not inherently affect cell behavior could be doped with therapeutic ions for localized and sustained delivery.

While bioactive glass microparticles have been shown to induce differentiation of adult20 and fetal21 osteoblasts and murine mesenchymal stem cells22 in vitro toward the osteogenic lineage, reports in the literature on human stem cell response to bioactive glasses differ.22–25 The aim here was to investigate the response of MSCs and adipose tissue-derived stem cells (ADSCs) to the bioactive mono-SMBGs. The same properties that make nano- and submicron-sized particles promising for technological advances are also the very properties that can induce undesirable effects to biological systems. Nano- and submicron-sized particles access intracellular compartments that are inaccessible to larger particles, potentially contributing to cell injury mainly due to oxidative stress and inflammatory response.18 A thorough understanding of the cellular interactions of particles with controlled size and chemistry will enable predictions to be made about how their physicochemical properties link to cell reactivity. Recently, we reported the effect of agglomerated SMBGs with a broad size distribution, with individual diameters in the range of 150–350 nm, on MSCs. SMBGs were nontoxic and were infrequently internalized by the MSCs.16 The small amount of internalization was attributed to the agglomeration of particles and small number of particles with diameters less than 250 nm. Among the many factors, such as size, shape, and surface charge, particle agglomeration plays a vital role in the interaction of particles with cells.26 In vitro and in vivo studies have shown that the use of agglomerated particles may cause inaccurate assessment of nanoparticles.27–29 In addition, agglomeration can strongly hinder determination of fundamental dose–response relationships.30 The aim was to obtain individual, spherical, uniform particles smaller than 250 nm and examine their uptake and effect on toxicity, growth, and differentiation on adult stem cells. For the synthesis of SMBGs in our previous work, a polymer was used as a template, but the polymer prevented calcium incorporation into the silicate network.16 Here, a modified process using ultrasonic agitation was developed and optimized. The mechanisms of uptake of the SMBGs by the cells and any morphological or chemical changes to the particles following uptake were assessed using a combination of transmission electron microscopy (TEM) and confocal light microscopy techniques. Due to the osteogenic potential of bioactive glasses, the effect of mono-SMBG on osteogenic differentiation of MSCs and ADSCs was also investigated.

2. Results and Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discussion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

2.1. Particle Synthesis and Characterization

Mono-SMBG were successfully produced by adapting a sol–gel processing route (based on the Stöber process)10 using a combination of ultrasonication during the synthesis and by optimizing the calcium concentration. Figure 1 shows TEM images of the particles. Without sonication (Figure 1a) the particles were not dispersed. However, sonication alone did not create monodispersed particles, the molar ratio of Ca(NO3)2 ·4H2O:TEOS was critical. Figure 1b and c shows that increasing the Ca(NO3)2 ·4H2O:TEOS ratio from 0.42:1 to 1.3:1 improved the spherical nature and dispersion of the particles. Increasing the ratio further to 4:1 (Figure 1d) caused agglomeration and led to the formation of many small clusters around the particles (Figure 1d). Therefore, monodispersed, spherical, and dense SMBGs, with a molar ratio of Ca(NO3)2 ·4H2O:TEOS of 1.3:1 (Figure 1c) were regarded as optimized. These results confirm that there is a threshold amount of Ca(NO3)2 ·4H2O that results in the formation of optimized monodispersed SMBGs.

thumbnail image

Figure 1. Bright-field TEM images of mono-SMBG, prepared by the sol–gel process, showing the effect on particle size, morphology, and dispersion of a) no sonication during the process and b–d) sonication with a molar ratio of Ca(NO3)2.4H2O:TEOS of b) 0.42:1, c) 1.3:1, and d) 4:1. Scale bars = 0.4 μm.

Download figure to PowerPoint

Compositional analysis (ICP-OES) showed that the optimized mono-SMBGs had a composition of 80 mol% SiO2 and 20 mol% CaO (mean values were 80.92 mol% SiO2 and 19.18 mol% CaO, ±0.26). The particles were sintered at 680 °C in line with previous studies,16 which showed this was the maximum stabilization temperature that could be used without crystallization of the glass. X-ray diffraction (XRD) of the new particles following sintering at 680 °C confirmed that they were amorphous due to the absence of sharp diffraction peaks (Figure 2a). The mean particle diameter measured using TEM was 215 nm (±20 nm, n = 185). Previous work used a polymer template to improve dispersion of particles (without sonication), but this led to a broad size distribution of agglomerated SMBGs and use of the polymer inhibited calcium incorporation resulting in a maximum of 15 mol% CaO could be incorporated into the SMBGs.16 Here, EDX chemical analysis (Figure 2b) confirmed the presence of Si and Ca in the mono-SMBGs with a composition of 80 mol% SiO2 and 20 mol% CaO.

thumbnail image

Figure 2. Mono-SMBG particle characterization of optimized process a) XRD trace of mono-SMBGs calcined at 680 °C showing amorphous nature; b) EDX analysis of the particles in Figure 1c showing Si, O, and Ca peaks, c) dynamic light scattering (DLS) of the size distribution of the mono-SMBGs, and d) The distribution in the zeta-potential:measurements taken from five mono-SMBGs batches.

Download figure to PowerPoint

Evaluation of the size distribution of mono-SMBG conducted using DLS, revealed a narrow size distribution with a mode of 220 nm (Figure 2c). The particle size obtained from DLS was larger by 5 nm than the mean obtained from TEM analysis (215 ± 20 nm, n = 185), a disparity which could be attributed to the fact that DLS measures the hydrated radius,14 hence the results obtained from TEM images are closer to the actual size of the particles (Figure 1c). In order to acquire a better understanding of the dispersion stability and surface charge of mono-SMBGs, their zeta potential (ζ-potential) was measured in 100% ethanol (100% EtOH) (Figure 2d). A ζ-potential above ±30 mV is an indication of particle stability, as the repulsive charge between the particles is usually enough to prevent agglomeration. Figure 2d shows repeated ζ-potential measurements for the optimized mono-SMBGs and the average ζ-potential of the mono-SMBG was −28 mV (n = 5, ±3.5 mV).

Optimization of the sol–gel processing was necessary to obtain monodispersed particles: Sonication during the sol–gel process, the point at which the calcium nitrate was added and the ratio of calcium nitrate:TEOS were all critical for mono-SMBG synthesis. Ultrasonication helped in the preparation of a stable sol. During the early stages of sol–gel synthesis using TEOS, silica nanoparticles form as a result of hydrolysis of the TEOS (creating –Si–OH groups) and some condensation of the Si–OH groups. As condensation continues the particles grow.31 Under acidic catalysis the particles coalesce and condensation reactions occur between Si–OH groups on the surfaces of the particles. Under basic catalysis, the nanoparticles are maintained due to repulsion between the particles, which is the principle of the Stöber process. However, physical interactions can still occur between the particles and this can lead to condensation. Carrying out the sol–gel process under ultrasonic treatment is thought to lead to the formation of a stable sol with dispersed spherical particles. More surprising was the effect of the quantity of calcium precursor and the point at which it was added to the sol on monodispersity. The molar ratio of calcium nitrate:TEOS during the synthesis of mono-SMBG had a significant effect on the morphology and formation of the particles.

The role of calcium nitrate in the sol–gel process has only recently been understood.31 The use of calcium salts, such as calcium nitrate, is limited by the high solubility of the salt. Calcium nitrate is soluble in the sol and in the condensation by-products; therefore, it remains in the form of calcium nitrate in sol until the particles are dried. The silica network forms at room temperature (RT). During drying, the calcium nitrate deposits onto the surfaces of the silica particles. The calcium only enters the glass network, by diffusion, when a temperature of 450 °C is reached during calcination.31 In terms of particle dispersion, the presence of calcium nitrate at the surface of the particles during drying may interrupt the usual charge repulsion of silica nanoparticles by changing the surface chemistry of the particles, introducing some positive charge which will attract the usually negatively charged silica particles. The amount of calcium that can be incorporated is therefore dependent on calcium diffusion into the particles. This was inhibited in previous work by the polymer template, but even without the polymer template, the maximum CaO content was 20 mol%. The amount of calcium nitrate used here to produce the optimized particles would give a nominal composition of 57 mol% SiO2, 43 mol% CaO. However, the compositional analysis showed the composition to be 80 mol% SiO2 and 20 mol% CaO.

The final composition of BG is affected if the pore liquor is removed before drying. During the processing of mono-SMBGs, it is necessary to centrifuge the sol before drying. However, centrifugation of the suspension may result in the removal of the soluble calcium nitrate, leading to reduced calcium content. This is a reason why the maximum amount of CaO that could be incorporated into the particles was 20 mol%. Increasing the amount of calcium nitrate in the sol had no affect on composition instead calcium-rich precipitates formed, likely to be calcium hydroxide due to excess calcium at high pH.32 The high pH is required to produce monodispersed submicron particles as it provides repulsion between the particles, preventing gelation and agglomeration. Other authors suggest that careful control of pH6, 15, 33 and the use of aerosol techniques34 may help increase the calcium content, however, the compositions of the particles were either not reported in the literature after synthesis or only nonquantitative EDX data were provided.

Despite these challenges, monodispersed submicron particles were synthesized with a composition of 80 mol% SiO2 20 mol% CaO and a mean particle size of 215 nm. Sol–gel glasses containing 80 mol% SiO2 are still considered bioactive.35

2.2. MSC and ADSC Viability: Alamar Blue, MTT and Cyquant Assays

To examine the biocompatibility of mono-SMBGs, cell viability was examined by the Alamar blue, MTT, and Cyquant assays (Figure 3). MSCs and ADSCs were exposed to three concentrations (50, 100, and 200 μg ml−1) of mono-SMBG for a 24 h (pulse) and the remaining noninternalized particles were subsequently removed by washing and as the media was changed. Cell behavior was then monitored in culture (chase) compared with cells that had not been exposed to particles.

thumbnail image

Figure 3. Effect of mono-SMBG on MSCs and ADSCs viability was measured by measuring cell metabolic activity using a) Alamar blue and b) MTT assay, and cell number using c) Cyquant assay to measure total DNA. All assays were performed after MSCs and ADSCs were treated with mono-SMBG (50, 100, 200, μg ml−1) for a 24 h pulse and followed after 1, 4, and 7 d in culture. All data values are expressed as percentage of the control samples (no mono-BGP) at day 1. Values represent the mean ± SD. The samples in each experiment were in sextuplicate and the experiment was repeated twice. (*) indicates the statistical significant difference P < 0.01, between the marked bar and the control sample (no mono-BGP) at the same time point.

Download figure to PowerPoint

Both the Alamar blue and MTT assays measure enzymatic activity; the first assay relies on the cytoplasmic reduction potential of metabolically active/viable cells, whereas the latter measures the mitochondrial activity of the cells. It is evident from the Alamar blue and MTT assays that ADSCs were overall more metabolically active, regardless of particle exposure, demonstrating almost four times higher fluorescence (Figure 3a ADSCs) and seven times higher absorbance (Figure 3b ADSCs) than MSCs after 7 d.

From the Alamar blue and MTT assays, it is evident that both cell types were still viable at 7 d with the exception of the ADSCs exposure to 200 μg ml−1 mono-SMBG for 7 d (post-pulse), for which cell activity significantly (P < 0.01 vs control at day 1, no mono-SMBG) decreased from 1046% to 931% viability (Figure 3a). This result was only observed in the Alamar blue assay.

The metabolic activity of ADSCs, measured by the Alamar blue assay, increased significantly (P < 0.01 vs control, no mono-SMBG) only after exposure to 100 μg mL−1 of mono-SMBGs at day 1 (from 100.0% to 198.0%) and to both 100 and 200 μg mL−1 mono-SMBG at day 4 (from 557.5% to 701.5% and to 617.0%, respectively). These results were not consistent with the MTT data (Figure 3b), which demonstrated no significant increase in ADSC metabolic activity at all concentrations of mono-SMBG (P < 0.01). In the MTT assay for ADSC cells, metabolic activity decreased from 1142.1% to 922.9% by the 100 μg mL−1 mono-SMBG at day 4 and from 2428.4% to 1714.2% by the 100 μg mL−1 mono-SMBG at day 7, however this decrease was not significant for P < 0.01. Overall our MTT data indicated no significant effect (decrease or increase) on metabolic activity up to 7 d for either MSCs or ADSCs, after exposure to mono-SMBG. The differences in data obtained from the two assays can be explained by the fact that the two assays test for intrinsic cell metabolism using different mechanisms. While the Alamar blue assay measures the reduction potential of the cells, and the MTT assay evaluates the activity of mitochondrial enzymes.36 Mono-SMBGs could affect these processes via different mechanisms, which would then result in the variation observed.

In contrast, in the total DNA assay (Cyquant), which was used to assess how cell number and consequently cell proliferation was affected by the mono-SMBGs, none of the concentrations of mono-SMBG tested caused any significant (P < 0.01) increase in MSC or ADSC number compared with control (Figure 3c). The highest concentration of mono-SMBG (200 μg mL−1) significantly (P < 0.01) hindered MSC proliferation, shown by a total DNA decrease (from 133.5% to 103.3%) after 4 d, whereas both 100 and 200 μg mL−1 resulted in a significant decrease (from 206.3% to 172.3% and 170.7%, respectively) in ADSC total DNA after 7 d.

This discrepancy between the total DNA and the metabolic activity data suggests that the mono-SMBG encourage somewhat the cells to be more active but not necessarily or to the same extent cell growth, indicating a delay in cell proliferation.

2.3. Cellular Uptake of Mono-SMBG

2.3.1. Confocal Microscopy

In order to distinguish between the internalized particles and those adhering to the outer surface of the plasma membrane, mono-SMBGs were labeled with FITC before incubation with the cells. The fluorescence of the extracellular FITC-labeled mono-SMBG was quenched with Trypan blue. Confocal microscopy (Figure 4) revealed that mono-SMBG, at the concentration of 100 μg mL−1, were internalized by both MSCs and ADSCs and appeared to localize in the cell cytoplasm. Green fluorescence (FITC) was not observed anywhere in the nonexposed cells (Figure 4a and e), indicating the staining in the exposed cells did indeed indicate the location of the particles (Figure 5 and Supplementary Video 1).

thumbnail image

Figure 4. Confocal images of a–d) MSCs and e–h) ADSCs cultured without and with FITC-labeled mono-SMBGs for 6 h following endocytosis inhibition with genistein and dynasore. MSCs and ADSCs incubated respectively a,e) without mono-SMBG and b,f) with 100 μg mL−1 FITC-labeled mono-SMBGs (appear as green), c,g) 100 μg mL−1 of FITC-labeled mono-SMBGs after inhibition of clathrin-mediated endocytosis (genistein treatment), and d,h) 100 μg mL−1 of FITC-labeled mono-SMBG after inhibition of caveolae-mediated endocytosis (Dynasor treatment). Scale bar = 36 μm.

Download figure to PowerPoint

thumbnail image

Figure 5. Confocal images of a) MSCs and b) ADSCs incubated with mono-SMBGs. 3D reconstructions and views of the xz and yz planes showing the FITC-labelled mono-SMBG particles internalised by the cells and associated with actin cytoskeleton. (Red = actin fibres, blue = nucleus and green = mono-SMBG).

Download figure to PowerPoint

To further probe which endocytic mechanism was involved in the uptake of mono-SMBGs into MSCs and ADSCs, two endocytotic pathways were blocked by treating the cells with two pharmacological inhibitors, genistein and dynasore, before exposure to mono-SMBGs. Each uptake pathway fulfills unique functions in the cell and varies mechanistically not only in how the vesicles are formed, but in terms of which cargo molecules they transport, to what intracellular destination their cargo is delivered, and how their entry is regulated.37 By identifying the entry route of mono-SMBG particles in MSCs and ADSCs, prediction or control of the final destination of the particles may be possible. Genistein inhibits protein tyrosine kinases and has been shown to block internalization by caveolae,38 while dynasore was used to block vesicular endocytosis by selectively inhibiting dynamin 1 and dynamin 2 GTPases, which are responsible for vesicle scission during both clathrin- and caveolin-mediated endocytosis and also play a role in some lipid raft-mediated processes.39 Confocal images of both MSCs (Figure 4c and d) and ADSCs (Figure 4g,h) following treatment with the endocytosis inhibitors genistein and dynasore revealed that internalization of the mono-SMBG was not altered, suggesting that cell uptake in both cell types is not caveolae or dynamin mediated. In reports on fluorescent latex beads of sizes ranging from 50 to 1000 nm utilizing non phagocytic cells, particles with sizes larger than 200 nm in diameter (but less than 1 μm) were taken up by the cells preferentially by caveolae-mediated endocytosis, while particles that were 200 nm or smaller were processed along the pathway of clathrin-mediated endocytosis.40 In addition, particles with size of 500 nm were seen to escape lysosomes, since lysosomal localization was no longer apparent.40, 41 Given, this size dependence internalization route, the 215 nm mono-SMBG particles would be expected to be taken up predominantly by caveolae-mediated endocytosis. However, no evidence for that was seen here, making further investigation necessary by looking more extensively in other routes of particle endocytosis used by the cells. In some cases, however size may play a less important role on cellular uptake, which could rather be orchestrated by surface charge properties, something that should be further investigated for the mono-SMBG particles. Confocal microscopy also revealed that mono-SMBG uptake affected the actin cytoskeleton of the MSC's differently than that of the ADSCs. In control, for both cells types, F-actin cytoskeleton appears well organized with defined microfilaments organized in thick bundles around the cell periphery (Figure 4a and e). Mono-SMBG increased the level of cortical actin filaments and often altered the membrane morphology. MSCs with internalized mono-SMBGs demonstrated disrupted and less organized F-actin cytoskeleton (Figure 4b), indicated by poorly formed, less defined microfilaments accumulated near the cell membrane and in the lamellopodia and filopodia edges compared with the control MSCs (Figure 4a). While the actin cytoskeleton of the ADSCs following mono-SMBG uptake appeared organized with defined filaments spread through the cell body in a long fibrillar pattern (Figure 4f), demonstrating an effect on cytoskeleton dependent on cell type. MSCs and ADSCs treated with genistein and dynasore, also demonstrated analogous changes in their F-actin networks (Figure 4c,d,g,h) strongly suggesting that the disruption and disorganization of the cytoskeleton was induced by the mono-SMBG uptake. This is not surprising, since the endocytic machinery is directly connected and functionally integrated to the actin cytoskeleton.42 Internalization of mono-SMBG at the plasma membrane will most likely require actin rearrangement and consequently disrupted cytoskeleton, therefore the difference in the actin cytoskeleton organization could suggest that ADSCs utilized a different mechanism to internalize the mono-SMBG particles than that used by the MSCs.

2.3.2. TEM

TEM was performed to study the distribution and morphology of the mono-SMBGs inside the cells (Figure 6). The number of the mono-SMBGs internalized into the stem cells was much higher than was observed when MSCs were exposed to heterogeneous and agglomerated SMBGs in previous studies under similar conditions.16 In bright-field TEM images, the mono-SMBG particles appeared to be more readily taken up and internalized by ADSCs (Figure 6b) compared with MSCs (Figure 6a). Strikingly, for both cell types the predominant localization of the mono-SMBGs was in the cell cytoplasm. Intracellular SMBGs were also found inside cell vesicles. In Figure 6a clusters of mono-SMBGs were also observed on the outer surface of the plasma membrane of MSCs, apparently being engulfed by membrane protrusions indicating endocytosis.

thumbnail image

Figure 6. TEM images and EDX chemical analysis of mono-SMBG's taken up by MSCs (panels (a) and (b), respectively), and ADSCs (panels (c) and (d), respectively) following incubation with 100 μg mL−1 of mono-SMBG for 24 h. Mono-SMBG were found distributed within the cell cytoplasm, encapsulated inside possible endosomes (white arrows) but also within the cytoplasm (black arrows). The square shows mono-SMBG particles being endocytosed. Particles in both MSCs and ADSCs show reduction in size after 24 h, indicating intracellular dissolution (quantitative data presented on table insets in (b,d)). Scale bar = 2 μm.

Download figure to PowerPoint

The mono-SMBGs in both MSCs and ADSCs showed a reduction in size (average reduction from 215 ± 20 nm, (n = 185) (before exposure) to 162 ± 40 nm inside MSCs (n = 60) and to 158 ± 39 nm for ADSCs, (n = 80) after 24 h, suggesting that the particles may have dissolved inside the cells (Figure 6a and c). EDX of the intracellular SMBGs confirmed the presence of Ca, O and Si (Figure 6b and d).

TEM demonstrated particles localizing in the cytoplasm with no apparent lysosomal vesicle membrane engulfing them (Figure 6). This could suggest that mono-SMBG particles could be internalized through macropinocytosis, an actin-dependent form of bulk uptake of fluid and solid particles, occurring through surface ruffling and is followed by sealing of the opening with the formation of discrete vacuoles, the macropinosomes.42 Macropinocytosis represents a well-organized approach for nonselective cellular uptake of particles with size >1 μm.40 It is likely that these different pathways have evolved so that macropinocytosis can be coordinated with more complex aspects of cell physiology, such as signal transduction, development and modulation of the cell's responses to and interaction with its environment. Our findings on the pathway by which mono-SMBG particles are internalized by MSC and ADSC cells could be used to further determine the application of these novel biomaterials.

The observation of mono-SMBGs inside the cell cytoplasm indicated that they can elude or escape the endocytic pathway (Figure 6a and d, black arrows) which may explain why the genistein and dynasore inhibitors could not identify an endocytic pathway of SMBG uptake into the cells. Elevated intracellular calcium has been reported as an important factor allowing transfecting agents to escape the phagocytic pathway.43 One assumption is that the particle dissolution inside the cells and the release of Ca and Si could be responsible for the presence of the particles in the cell cytoplasm. This pattern of cytoplasmic localization is similar to what has been observed previously with mesoporous silica particles and SMBG particles.16, 30 Escape from the endocytic pathway, is a prerequisite for effective therapeutic effect as mono-SMBG should release their load on the cytoplasm if they are to be used for drug delivery application. Further work is now required to study the mechanism by which the cells process and traffic these particles.

2.4. Alkaline Phosphatase Enzymatic Activity as Marker of Osteogenic Differentiation

If mono-SMBG particles stimulate differentiation toward the osteogenic pathway, they could be used as bone regeneration inducers. Alternatively, if they do not affect cellular response in any way they could be used as intracellular drug delivery devices, where the glasses provide a sustained release of therapeutic ions. An example would be delivery of strontium ions. Strontium can be substituted into a glass composition for calcium and release of critical amounts of strontium from bioactive glasses has been shown to stimulate osteoblasts and inhibit osteoclasts, which would be of benefit to patients with osteoporosis.19, 44 For this reason, it was important to assess whether biodegradable mono-SMBGs (initially without additional therapeutic ions) were inherently able to stimulate osteogenic differentiation of stem cells. In vitro cell survival and expansion levels have previously been shown to increase when the cells are cultured on bioactive glasses of similar composition to mono-SMBG, or in the presence of their dissolution products, until the cells start to differentiate toward the osteogenic lineage, at which point proliferation decreases at the onset of mineralization.22, 45 Therefore, due to the similarity between the compositions that were previously investigated and the composition of mono-SMBG (80S20C), it is crucial to investigate the effect of these particles on MSC and ADSC differentiation.

Alkaline phosphatase (ALPase) enzymatic activity is a widely used biochemical marker of osteoblast phenotype in vivo and stem cells and primary cells differentiating to mature osteoblasts in vitro.46 ALPase enzymatic activity of the cells cultured post-pulse with the mono-BGP in the presence or absence of the osteogenic supplements (β-glycerophosphate, L-ascorbate-2-phosphate, and dexamethasone) was measured as an indication of osteogenic differentiation. Both MSCs and ADSCs presented similar ALPase enzymatic activity profiles both with mono-SMBG and without, demonstrating similar capacities for osteogenic differentiation (Figure 7). In the presence of osteogenic supplements, ALPase activity increased steadily from day 4 to 21 for both cell types (Figure 7a). Indeed, osteogenic supplements significantly increased ALPase enzymatic activity of the cells with mono-BGP and without, strongly demonstrating that the particles did not affect the ability of MSCs and ADSCs to differentiate to osteoblasts (the bone forming cells). However, in the absence of the osteogenic supplements the mono-SMBG did not affect MSC or ADSC ALPase activity producing comparable amounts as the control. This could be due to their uptake by the cells. In previous experiments, where differentiation was observed, ionic dissolution products of the bioactive glass were administered to cells each time the media was changed, or cells were grown directly on the glass.22 In this case, a pulse-chase experiment was performed. After the first change of media, most of the particles should have been removed from the well plates, which could result in lower amounts of extracellular soluble silica and Ca ions present than was previously reported to stimulate osteogenesis20–22 or direct contact of the cells with the material. Conversely, intracellular dissolution may change upon uptake by the cells depending on whether the particles localize within the lysomes or in the cytoplasm. The environment in these sites is different than that in a solution or in vivo. Consequently, we cannot conclude with certainty that the degradation process of the mono-SMBG occurs in the same sequence of reactions as a bioactive glass in media. This could again alter the glass' degradation and ionic dissolution profile and consequently its osteostimulatory effect.

thumbnail image

Figure 7. ALP enzymatic activity of a) MSCs and b) ADSC incubated with 100 and 200 μg mL−1 mono-SMBG for 24 (pulse) and then washed well to remove noninternalized particles and cultured for up to 28 d (chase) in the presence and in the absence of β-glycerophosphate, L-ascorbate-2-phosphate, and dexamethasone. ALPase activity is expressed as nmols of pnp hydrolyzed and normalized to total protein (μg) per hour. Values represent the mean ± SD. The samples in each experiment were in quadruplicate and the experiment was repeated twice. (*) (P < 0.05) and (¥) (P < 0.001) indicate the statistical significant difference between the marked bar and the control sample (no mono-BGP) at the same time point.

Download figure to PowerPoint

3. Conclusions

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discussion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

Monodispersed spherical bioactive glass submicron particles were synthesized by a modified sol–gel process with the use of sonication. The calcium to TEOS ratio and use of sonication were critical for dispersion and for obtaining the spherical shape. The particles were not significantly cytotoxic at the concentration of 50 μg mL−1 and had minimal effects on primary MSCs and ADSC viability at concentrations of 100 and 200 μg mL−1. Doses of greater than 100 mg mL−1 are high and probably not realistic of the concentrations that the cells would see when the particles are administered to the body; nevertheless even at the higher doses, only minimal effects on cell metabolism and proliferation were measured which emphasizes the potential of using the mono-SMBGs for inorganic delivery applications. Confocal microscopy and TEM demonstrated that the particles translocated frequently into the cell cytoplasm and were also partially degraded by both MSCs and ADSCs. Comparison with our previous work on agglomerated heterogeneous SMBGs revealed a different uptake profile, with the mono-SMBG uptake being more frequent, indicating that dispersion of the particles, and finer control of particle size, dramatically increased the amount of the particles taken up by the cells. The uptake mechanism of mono-SMBGs into the cells remains unclear because the endocytosis was not dramatically affected by any of the inhibitors for clathrin- or caveolin-mediated endocytosis. However, the TEM images clearly show uptake and presence of the particles in the cytoplasm.

ALPase enzymatic activity was not affected by the presence of the mono-SMBG even after 21 and 28 d for the MSCs and ADSCs, respectively, indicating that the particles had no effect on the potential of the MSCs and ADSCs to differentiate toward the osteogenic lineage. Further work is required to study the pathway of the mono-SMBGs into the cells and the mechanisms of endosomal escape of these particles.

4. Experimental Section

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discussion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

Synthesis of Mono-SMBG particles: The methodology applied by Rao et al.14 for the synthesis of 100S (100 mol% SiO2) particles was modified to obtain bioactive compositions of mono-SMBGs. The mono-SMBGs were prepared using the sol–gel process, by hydrolysis of tetraethyl orthosilicate (TEOS, (Si(OCH2CH3)4) (98%, Sigma, UK) in ethanol in the presence of an ammonia catalyst in an ultrasonication bath. Initially, 4.5 mL of ethanol and 0.5 mL of H2O were mixed in the sonication bath for 10 min. Then, 100 μL of TEOS was added and left for a further 20 min before 5 mL of ammonium hydroxide (N

H4OH) (28% NH3 in water, 99%) (Sigma) was added drop-wise and then left to mix for 60 min. Calcium nitrate is the common precursor for introducing calcium into the sol–gel process and is usually added 1 h after adding TEOS to water. The effect of the addition calcium nitrate tetrahydrate (Ca(NO3)2 ·4H2O) (Sigma) on particle formation and composition was investigated by using different molar ratios of Ca(NO3)2 ·4H2O to TEOS of 0.42:1 (the ratio that is commonly used for 70S30C sol–gel bioactive glasses,17 1.3:1 and 4:1. The white suspension was then centrifuged to obtain white solid deposit. The deposited solid was dried to remove excess water from the solid particles, producing a white powder, which was then culminated at 680 °C to produce mono-SMBG particles (Figure 8). The sintering temperature was chosen as it was the optimal sintering temperature for the previous study on SMBGs.16 For cell culture experiments particles were sterilized by dry heat at 120 °C for 2 h.

thumbnail image

Figure 8. Schematic representation of the synthesis route of mono-SMBG.

Download figure to PowerPoint

Characterization of Mono-SMBG: Acid digestion compositional analysis was applied to measure the composition of SMBGs. 0.1 g of finely ground glass was mixed carefully with 0.5 g of anhydrous lithium metaborate in a clean dry platinum crucible using a glass rod. The mixture was fused for 1 h at 1400 °C and then left to cool. The crucible was then immersed in 80 mL of 10 vol% nitric acid to completely dissolve the flux and then transferred to a 100 mL polypropylene volumetric flask and made up to the mark. The elemental concentration in the solution was measured using inductively coupled plasma optical emission spectroscopy (ICP-OES, Thermo Scientific Icap 6000 series).

SMBGs were dispersed in ethanol and then collected on 300 mesh copper TEM grids, coated with lacey carbon film. In order to quantify particle size, the diameter of individual particles (n = 185) was measured from TEM images in Image J software. Images were taken after short exposure times to the electron beam to ensure no beam damage to the particles during analysis. Bright field TEM imaging and EDX analysis were performed on a JEOL 2000 microscope operating at 120 kV. Sol–gel glass of known (70 mol% SiO2, 30 mol% CaO) composition were used as standards for EDX analysis.

In order to investigate the size distribution of mono-SMBGs dynamic light scattering (DLS, Malvern instrument 2000) was used. To obtain a better understanding of the stability and surface charge of the mono-SMBGs, zeta-potential measurements (n = 5) were also applied on Malvern instrument 2000. For both size and zeta-potential analysis, mono-SMBGs were suspended in 100% EtOH before measurements.

XRD spectra were collected on a Philips PW1700 series automated powder diffractometer using Cu Ka radiation at 40 KV/40 mA. Data were collected between 5° and 80° 2θ with a step of 0.04° 2θ and a dwell time of 1.5 s to identify any crystallization of the particles.

Cell Source: MSCs and ADSCs were purchased from Cambrex (Cambrex, UK) and maintained in low glucose, phenol red free, Dulbecco's Modified Eagle Medium, supplemented with 10% (v/v) batch tested fetal bovine serum, 50 U mL−1 penicillin and 50 μg mL−1 streptomycin, 1% (v/v) L-glutamine (all from Invitrogen, UK) (which will be referred to as complete medium).

Cytotoxicity of Mono-SMBG. Metabolic Activity and Total DNA: The effect of mono-SMBG on cell activity and total DNA was investigated using pulse-chase exposure, where cells were exposed to a pulse of mono-SMBG for 24 h followed by chase periods of 1, 4, and 7 d. Cells were seeded at a density of 10 000 cells cm−2 and maintained in complete medium. The next day the medium was changed with complete medium containing mono-SMBG at concentrations of 0, 50, 100, and 200 μg mL−1. After 24 h, the media was replaced with complete medium (not containing any particles) which was changed at days 4 and 7. MSC and ADSC viability and proliferation was assessed by measuring their metabolic activity with Alamar Blue® (Invitrogen) and (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) (Sigma) assays and their total DNA with the Cyquant (Invitrogen) assay according to the manufacturers' instructions. Fluorescence and absorbance was measured on a SpectraMAX GeniumXS plate reader and the results represent the mean values ± SD of two individual experiments each in quadruplicate.

Mono-SMBG FITC Labeling: Mono-SMBGs were labeled with FITC (Sigma) according to Rosenholm et al.47 to allow them to be observed using confocal microscopy. 25 mg of heat-sterilized mono-SMBG particles were suspended in a filter sterilized carbonated buffer and all procedures were carried out in aseptic conditions. For the staining, 250 μL of a FITC solution (1 mg mL−1 in 100% ethanol) was added into the buffer containing the particles and the final solution was incubated at RT for 1 h with continuous stirring. Particles were then collected by centrifugation and washed with deionized water.

Confocal Microscopy: In order to determine whether uptake of the mono-SMBGs was mediated by one of the common mechanisms of clathrin or caveolae-mediated endocytosis, the MSCs and ADSCs were cultured in the presence of inhibitors of each mechanism. MSCs and ADSCs were seeded at 30 000 cells cm−2 in complete medium and the next day were treated with 400 × 10−6 M Genistein19 or 80 × 10−6 M Dynasore for 30 min at 37 °C.44 At the end of the incubation period, the medium containing the inhibitor was removed. 100 μg mL−1 of FITC-labeled mono-SMBG dispersed in complete medium was then added for 6 h. After 6 h, the medium was aspirated and cells were washed twice with PBS, detached from the chamber slides by trypsinization and then replated to new four-well chamber slides in fresh complete medium. To confirm that the FITC-labeled particles had been internalized by the cells, and were not attached to the cell surface, Trypan blue solution, which cannot penetrate the membranes of living cells, was added for 15 min at RT to quench the fluorescence of the noninternalized particles on the exterior surface of the cells. Then, cells were fixed with 4% (w/v) paraformaldehyde for 2 min at RT and phalloidin staining was performed as previously described.16 Stained cells were viewed under an inverted confocal microscope (Leica SP5 MP, Leica, Germany). The procedure is schematically outlined in Figure 9.

thumbnail image

Figure 9. Schematic representation of the procedure followed for the mono-SMBG uptake study using confocal microscopy.

Download figure to PowerPoint

TEM: The location of the mono-SMBGs following uptake into cells was investigated using TEM. The MSCs and ADSC were seeded on sterile glass cover slips (Invitrogen) at a density of 60 000 cells cm−2. The cells were treated with 100 μg mL−1 of mono-SMBG for 24 h. A fixation and embedding procedure for TEM analysis was carried out as previously described.16 Sections were cut with an ultramicrotome using a 35° wedge angle diamond knife and immediately collected on uncoated 300 mesh copper grids (Agar Scientific, UK) and dried for 5 min at 37 °C. Bright-field TEM imaging and EDX analysis (Oxford Instruments, UK) were performed on a JEOL 2000 microscope operating at 120 kV. Multiple cells from three different cell exposures (for both the MSCs and ADSCs) were surveyed.

Osteogenic Differentiation: Alkaline phosphatase enzymatic activity. To examine whether mono-SMBGs induce osteoblast differentiation, MSCs and ADSCs were cultured with mono-SMBGs with and without osteogenic supplements and alkaline phosphatase activity was measured. As a positive control of osteogenic differentiation, 10 × 10−3 M β-glycerophosphate (β-g-P), 50 μg mL−1 L-ascorbate-2-phosphate (L-a-2-P) and 50 × 10−6 M dexamethasone (Dex) (all from Sigma) was added to complete medium. Six medium conditions were investigated and are outlined in Table 1.

Table 1. List of conditions tested for osteogenic differentiation of MSCs and ADSCs. FBS is Fetal Bovine Serum, L-Glut is L-Glutamine and p/s is penicillin/streptomycin added to complete medium. β-glycerophosphate (β-g-P), 50 μg ml–1 L-ascorbate-2-phosphate (L-a-2-P) and dexamethasone (Dex) are the osteogenic supplements.
Media Tested for Osteogenic Differentiation
 Mono-SMBG [μg mL–1]FBS [v/v %]L-Glut [v/v %]p/s [v/v %]L-a-2-P [μg mL–1]β-g-P [mM]Dex [μM]
Control1011
Supp1011501050
Supp+100 μg ml–11001011501050
Supp+200 μg ml–12001011501050
100 μg ml–11001011
200 μg ml–12001011

Mono-SMBG particles were sterilized by dry heat. MSCs and ADSCs were seeded at a density of 10 000 cells cm−2 in complete medium. The next day, cells were exposed to the media conditions outlined in Table 1 for 24 h (pulse). After the incubation time of 24 h the cells were washed well to remove any particles which had not been taken up by the cells and fresh medium, either complete or osteogenic, was added with subsequent media changes every 3–4 d. Cells were cultured for 4, 8, 12, 15, and 21 d. Cellular ALP enzymatic activity was measured colorimetrically according to the manufactures instructions (Sigma). Absorbance was then read at a plate reader (SpectraMAX GemimXS plate reader) at 405 nm. ALP activity was calculated using a p-nitrophenol (pNP) standard curve and normalized to total protein measured using DCTM detergent compatible protein assay (Bio-Rad, UK). Alkaline phosphatase activity was expressed as nmol pNP μg protein−1 h−1.

Statistics: All experimental data shown are expressed as mean ± SD. Statistical analysis was performed using the Sigma Stat software using the Student's t-test. P values of <0.05 and <0.001 were considered as statistically significant.

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discussion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

Supporting Information is available from the Wiley Online Library or from the author.

Acknowledgements

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discussion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

The project was funded by the EPSRC grant number EP/E057098/1. JRJ thanks the Philip Leverhulme Prize for Engineering. AEP thanks European Union FP7 project number 257182 (CNTBBB). MMS thanks ERC grant “Naturale”. The authors thank Mr R. Sweeney and Dr M. Ardakani for XRD and TEM assistance, respectively.

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Results and Discussion
  5. 3. Conclusions
  6. 4. Experimental Section
  7. Supporting Information
  8. Acknowledgements
  9. Supporting Information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
adhm_201300126_sm_Movie001.avi5013KMovie001
adhm_201300126_sm_Movie002.avi2135KMovie002
adhm_201300126_sm_suppl.pdf302Ksuppl

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.