Abstract
BACKGROUND:
Glass polyalkenoate cements (GPCs) are bio-adhesives which consist of ionomeric glass particles embedded in a poly-salt matrix. These materials have been used in dentistry and orthodontics extensively but are presently being optimized as bone putties for orthopedic applications.
OBJECTIVE:
This study utilized a patented ionomeric glass (mole fraction: SiO2:0.48, ZnO:0.36, CaO:0.12, SrO:0.04) to formulate two GPCs: GPC A (<45 μm particle size glass) and GPC B (45 μm–63 μm). These formulations were previously assessed for their effect on osteoblast viability and osteogenic function. However, the immunomodulatory effects of GPC A and B have not previously been investigated.
METHOD:
Non-toxic concentrations of (a) GPC dissolution products and (b) fragmented GPC particles were tested for their ability to affect the secretion of cytokines (TNF-α, IL-1β, IL-6 and IL-10) by rat peripheral blood mononuclear cells (PBMCs), in the presence or absence of the stimulant liposaccharide (LPS). Additionally, the ionic concentrations of Sr, Zn, Ca, and Si were measured in GPC ionic extracts, and the size, shape and concentration of fragmented GPC particles in deionized water were characterized using an optical microscope-based particle analyzer.
RESULTS:
The results showed that GPC A ionic products reduced the concentration of TNF-α secreted by stimulated cells compared with cells stimulated in the absence of GPC products. Interestingly, the particles released from GPC A significantly increased the secretion of both TNF-α and IL-6 from unstimulated cells, compared to control cells.
CONCLUSION:
Neither GPC B ionic products nor released particles were found to be biologically active with respect to PBMC cytokine secretion.
Introduction
Macrophages are a specialized cell type derived from circulating monocytes and are critical players in the host’s immune response to a biomaterial [1]. Injury as a result of the implantation procedure initiates an immediate inflammatory reaction, which is a dynamic and multi-step process that serves to recruit immune cells to the site of injury to clear away debris and pathogens [2]. Pro-inflammatory cytokines are released by activated macrophages, which serve to recruit more cells to the site of injury, stimulate angiogenesis and modulate the immune response appropriately [3,4]. These specialized immune cells later modify their phenotype to aid in inflammation resolution by releasing anti-inflammatory factors to facilitate tissue repair as the matrix formation phase is commenced. This controlled wound healing response is expected and inevitable, however, as a foreign body, an implanted biomaterial may potentiate the inflammatory response leading to chronically activated macrophages [5]. This is detrimental to normal tissue regeneration and in the case of bone implants, may stimulate osteoclastogenesis and bone resorption around the prosthesis [6,7]. This underscores the need to characterize the immunomodulatory effects of novel materials.
Glass Polyalkenoate cements (GPCs) are bio-adhesives formed by mixing a bioactive glass with polyacrylic acid (PAA) [8]. This initiates an acid-base setting reaction where hydrogen ions attack the surface of glass particles, releasing metal cations which subsequently cross-link PAA chains [9]. The final cement consists of unreacted bioactive glass particles embedded in a poly-salt matrix. GPCs have been employed in both restorative and orthodontic dentistry since the 1970’s [10], and their ability to bear load and bind to both bone and surgical metal has driven research toward developing them for orthopedic applications [11–13]. Dental GPCs are formulated with fluoro-alumino silicate bioactive glass, which is not suitable for skeletal applications due to the deleterious effects associated with leaching aluminum (Al) into body fluids [14–16]. In the present study, strontium (Sr) and zinc (Zn) are included in a patented (US 7,981,972), Al-free bioactive glass (SiO2:0.48, ZnO:0.36, CaO:0.12, SrO:0.04) which is utilized to formulate GPCs, which have been previously optimized for sternal fixation and radial fixation applications [17–19]. These novel bio-adhesives have been found to be antibacterial and have the ability to release therapeutic ions (Sr and Zn) into the surrounding environment [20–22]. Our previous study found that that low concentrations of dissolution ions from these novel GPCs were able to significantly accelerate the differentiation and mineralization of osteoprogenitor cells above control cultures, and this effect was different depending on the GPC composition [22]. However, we also reported that high concentrations of dissolution ions were shown to have deleterious effects on osteogenesis and cell viability.
It has been widely reported that immune responses can be modulated by altering the physical and chemical characteristics of bioactive glasses. Additionally, both the physical interaction of bioactive glass particles with cells and the chemical effect of dissolution ions on cells have been shown to alter cytokine secretion by macrophages. For example, 45S5 Bioglass particles reduced the secretion of pro-inflammatory cytokines, such as tumor necrosis factor-α (TNF-α) and interleukin-6 (IL-6), from stimulated macrophages and monocytes in vitro [23]. Bioglass extracts decreased the expression of TNF-α and interleukin 1-beta (IL-1β), while increasing the expression of interleukin-10 (IL-10), a cytokine involved in inflammation resolution [24]. Conversely, high concentrations of bioactive glass particles (10 μg/ml) have also been shown to stimulate macrophages and increased TNF-α secretion if the concentration of particles is increased and the particle size of the glass is decreased [25,26]. The addition of Sr to bioactive glass significantly attenuates the pro-inflammatory response (IL-6 production) from RAW 264.7 macrophages compared to bioactive glass alone [27]. Lastly, zinc is a well-known for its anti-inflammatory effects and ability to regulate the production of cytokines [28,29]. Zn incorporation into hydroxyapatite (HA) has similarly been shown to modulate immune responses from macrophages and monocytes, notably by decreasing TNF-α production while simultaneously increasing IL-10 production [30].
Considering the important role of macrophages and monocytes in the biocompatibility response to a biomaterial, and the potential ability of released ions and particles to either supress or prolong this reaction, it is important to characterize biomaterial-immune cell interactions for novel materials. The present study utilizes two different GPCs formulated from the same glass precursor: GPC A utilizes <45 μm glass while GPC B utilizes 45–63 μm glass. Additionally, GPC B is formulated with a higher molecular weight PAA and altered powder: liquid ratio to produce a cement with decreased ion release, higher strength and reduced working time [22]. We previously reported that GPC A dissolution products (Sr, Zn) were significantly elevated compared to GPC B, and in turn were able to significantly increase the expression of collagen, ALP and deposition of mineralized nodules in primary osteoblasts. The first aim of this study is to assess how the dissolution products from GPC A and GPC B affect cytokine secretion from stimulated and unstimulated primary peripheral blood mononuclear cells (PBMCs), which contain monocytes and lymphocytes. Cells will be stimulated with lipopolysaccharide (LPS), which is normally found in the outer membrane of gram-negative bacteria and is known stimulator of inflammatory cytokine release by PBMCs [23].
The second aim of this study is to assess how the fragmented particles released from GPC A and GPC B affect cytokine secretion from PBMC’s. Recent studies in sheep and rabbits indicate that GPCs implanted into bone perform better when a larger particle size glass is used [31]. Small particles released from biomaterials can be engulfed by macrophages and potentiate the pro-inflammatory response, a process which is a major contributor to aseptic loosening of implants and clinical implant failure [5,32]. A topic that is seldom explored in GPC research is the possibility of particulate debris release during cement application. GPCs utilized in orthopaedic applications would be applied as directly to bone as unset putties [33]. Once applied, the cement will eventually set into a hard material, however there is the possibility of unreacted bioactive glass leaching off the surface of this unset cement when interacting with body fluids before the material is completely solidified. Here, bioactive glass particles and fragmented debris will likely interact with macrophages and monocytes and promote the secretion of pro-inflammatory cytokines. To investigate this response, we measured the production of cytokines from PBMCs cultured on surfaces coated with subtoxic concentrations of bioactive glass particles released from the two different GPC compositions. Bioactive glass and HA powders have been investigated extensively for their ability to modulate immune responses; however, the in vitro immunomodulatory effects of GPC debris have never been investigated before to our knowledge.
Methods
Glass and cement preparation
The patented bioactive glass used for this study (mole fraction: SiO2:0.48, ZnO:0.36, CaO:0.12, SrO:0.04) was prepared by a glass manufacturer (Mo-Sci, Rolla, MO) using specified mole fractions. The resulting glass was ground using a ball-mill and sieved to achieve two ranges of particle sizes: <45 μm and 45–63 μm. The resulting glasses were then annealed for 6 hours at 640°C in a furnace (Zircar Hot Spot 110, Florida, New York, USA) and the annealing temperature was reached in 3 h. Subsequently, the glasses were cooled in the furnace to room temperature. GPCs were formulated as specified in Table 1, by mixing the respective glass powders with deionized water and PAA powder on a glass plate using a spatula. GPC A also contained trisodium citrate, which was originally added to improve the rheological properties. The mechanical properties of GPC A and GPC B are reported in [17,18,31].
Cement compositions.
Cement compositions.
The ion extracts were prepared according to the protocols reported previously [22,34,35]. The components of each respective GPC were mixed on a glass plate using a spatula, and unset adhesive putties were added to serum-free RPMI 1640 media (GIBCO) after 3 minutes with the ratio of 1 g bioactive glass powder to 10.0 ml solution, and incubated for 7 days at 37°C. After the incubation period, the supernatants were collected and sterilized by filtration using a 0.22 μm polyethersulfone (PES) membrane (Millipore, Sigma) and stored at 4°C until further use (ISO10993).
The resulting GPC ion extracts (n = 3) from GPC A and GPC B were utilized for ion release analysis. First, they were prepared by adding nitric acid to obtain a final concentration of 0.25% v/v. The concentrations of zinc (Zn), strontium (Sr), calcium (Ca) and silicon (Si) in each sample was measured using inductively coupled plasma–optical emission spectrometry (ICP-OES, 5110, Agilent, Santa Clarita, CA). Calibration standards (0 ppm, 1 ppm, 5 ppm, 10 ppm, 30 ppm) were prepared from 1000 ppm stock solutions (Sigma-Aldrich, Oakville, Canada) for the five elements measured in this study.
For cell culture analysis (cytotoxicity, cytokine secretion), GPC ion extracts were supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, 100 μg/ml streptomycin, 0.25 μg/ml amphotericin and 2 Mm L-glutamine. GPC extracts were added directly to cells either neat or diluted by a factor of 10. Extracts were diluted because our previous study suggested that GPC extracts have significant effects on various cell types at a 1/10 dilution factor [19,22]. Therefore, GPC ion extracts were diluted with RPMI 1640 medium and supplemented as described previously.
Coating of tissue culture surfaces with GPC particles and particle characterization
To obtain GPC particles, GPC components were mixed as previously described in Section 2.2. and unset cement putties were added to deionized water (with the ratio of 1 g glass powder to 10.0 ml solution) and incubated for 7 days at 37°C, with agitation. After the incubation period, the supernatant containing GPC particulate debris was removed and sterilized via repeated centrifugation and washing with 70% ethanol. Finally, the pellets were resuspended in deionized water (10.0 ml) to produce stable suspensions of particulate debris.
For cell culture analysis (cytotoxicity, cytokine secretion), tissue culture surfaces were coated with particles as previously described by Day and Boccaccini [23]. Particle suspensions were added directly to individual wells of a 24-well polystyrene cell culture plate either neat or diluted by a factor of 10. The plates were air dried in a laminar flow hood, leaving a coating of particles on the surface of the tissue culture plates. Subsequently, the plates were sterilized by UV exposure for 4 hours.
Particle suspensions (n = 3) from GPC A and GPC B were characterized using an automated optical particle sizing system (Clemex PSA Research Unit, Clemex Technologies Inc., Longueuil, Quebec, Canada) to determine particle size, shape, and concentration. A 1 ml aliquot of solution was dropped onto a glass slides, placed on the microscope stage, and digital images were obtained which were then analyzed using the Clemex image analysis software.
Blood sample collection and PBMC isolation
PBMCs containing monocytes and lymphocytes were isolated from whole rat blood, according to the protocol described by Böyem [36]. Blood samples were collected from 3-to5-month-old Sprague-Dawley rats according to protocols approved by the St. Michael’s Hospital Animal Care Committee (vivarium protocol #894/895, St. Michael’s Hospital, Toronto, ON). Briefly, the animals were sacrificed under anesthesia and whole blood was collected from the dorsal aorta using a syringe and promptly transferred to EDTA-coated vacutainer tubes (BD fisher, ON, Canada) to prevent coagulation. Whole blood (4 ml) was immediately used for PBMC isolation. First, the blood was diluted in PBS (1:1 ratio) and processed using density gradient centrifugation with Histopaque® 1077. After centrifugation, the mononuclear cell layer (white layer) was carefully extracted, and the resulting cells were washed with PBS three times. PBMCs were then counted using a hemocytometer and presented at least 95% viability. Cells were subsequently suspended in RPMI 1640 media supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin, 100 μg/ml streptomycin, 0.25 μg/ml amphotericin and 2Mm L-glutamine.
Cytotoxicity of GPCs
Cell viability in response to GPC conditioned medium (containing dissolution ions) and GPC particulate debris was assessed after 24 hours using a formazan calorimetric assay. For culture with GPC ion extracts, PBMC culture medium was exchanged for GPC conditioned medium (undiluted or 1/10), and cells were seeded (1 ×106 cells/ml) onto 24-well plates and allowed to incubate for 24 h under standard conditions (5% CO2, 100% humidity, 37°C). For culture with GPC particles, PBMCs were seeded directly on wells coated with GPC particles. Supplemented RPMI media alone was used as a control. This cytotoxicity assay is based on the conversion of a tetrazolium compound (MTT) to insoluble formazan crystals, providing a quantitative metric for cell metabolism, and reflecting the number of viable cells. Briefly, 100 μl of MTT reagent (5 mg/ml in PBS) was be added to each well and re-incubated for 4 hours under standard conditions. This was followed by the addition of 1 ml of dimethyl sulfoxide to each well to dissolve formazan crystals, which was kept for 12 h at 37°C. After the samples were transferred to a new 96-well plate in triplicate, their absorbance was read at 570 nm and 650 nm for reference.
Cytokine release test
The release of IL-6, IL-10, TNF-α and IL-1β from PBMC’s cultured with GPC conditioned medium or on wells coated with GPC particles in the presence or absence of 5 ug/mL LPS was measured using a commercially available enzyme-linked immunosorbent assay (R&D Systems, Mississauga, ON, USA). LPS was used as an in vitro model for immune cell stimulation and represents a positive control for inflammatory activation. Additionally, the ability of GPCs to modulate the inflammatory response was assessed by treating cells with LPS and observing whether cytokine release was altered after the addition of GPC products. Only subtoxic concentrations of GPC products (1/10 dilutions) were utilized for cytokine secretion assays. After 24 hours of culture, cell culture media was removed from wells, centrifuged to remove debris, and stored at −80°C for future analysis. Subsequently, ELISA assays were employed to detect the concentration of TNF-α, IL-1β, IL-6 and IL-10 in cell culture supernatants, which was calculated using standard curves for each cytokine generated by plotting the absorbance values (450 nm–570 nm) of known quantities of each cytokine.
Statistical analysis
All results were expressed as means +∕− standard deviation, and all experiments were repeated in triplicate. Prism GraphPad 9.0 (GraphPad Software Inc.) was used to analyze the data. The Shapiro–Wilk test was used to test normality, and the parametric one-way ANOVA was used to analyze all data, with the Dunnett’s test to correct for multiple comparisons. Results were considered statistically significant when p values were <0.05.
Results
Ion release
The release of dissolution ions from GPC A and GPC B into serum-free RPMI 1640 media was measured using ICP analysis and the results are shown in Fig. 1. Conditioned media from GPC A contained higher concentrations of Ca2+(10.3 ± 3.8 ppm), Sr2+ (3.9 ± 1.3 ppm) and Zn2+(17.3 ± 5.2 ppm) compared to conditioned medium from GPC B (Ca2+ 3.9 ± 0.6 ppm; Sr2+2.1 ± 0.2 ppm; Zn2+3.1 ± 1.1 ppm). Notably, GPC A released 5 times more Zn2+ than GPC B, which is in agreement with our previously published report which assessed ion release in deionized water and Dulbecco’s Modified Eagle Medium [22]. Regarding Si4+ release, the ion extracts from GPC B contained higher concentrations of this ion (22.0 ± 2.8 ppm) compared to GPC A (13.2 ± 2.1 ppm).
Ion release profiles of ionic extracts of GPC A and GPC B in RPMI media, considering calcium, strontium, zinc and silicon. Data shown are mean of n = 3 replicates ± SD.
The GPC particle size distributions were measured optically on microscope slides and results are shown in Fig. 2(A) and (B). The debris released from both compositions appears to be irregularly shaped glass particles, with sphericity values of 0.785 and 0.867 for GPC A and GPC B, respectively (Table 2). For GPC A, we found a size range of particles from 0.75 μm to 34.6 μm, the most frequent size range was between 1–10 μm and 50% of all the particles in the sample were smaller and larger than 1.98 μm (d 50 = 1.98 μm). In the GPC B suspension, we found a size range from 10.13 μm to 82.07 μm, the most frequent size range was between 40–50 μm and 50% of all the particles in the sample were smaller and larger than 49.55 μm (d 50 = 49.55 μm). The particle size analysis also revealed that the particles in the GPC A suspensions were 20 times more concentrated than the particles in the GPC B samples (Table 2).
Characterization of particles released from unset GPCs in deionized water (1:10 dilutions). Representative microscope images demonstrate the structure and size of particles released from (A) GPC A and (B) GPC B. The relative frequency of particles present within stable suspensions taken from (C) GPC A and (D) GPC B are also presented.
Characterization of particles released from GPC A and GPC B in deionized water.
The cytotoxicity of GPC ionic products and released particles were assessed using an MTT assay after culture for 24 hours, where a reduction in the normalized optical density indicates a reduction in the number of viable cells. As shown in Fig. 3(A), both GPC A and B particles resulted in a significant reduction in PBMC viability, as evidence by significantly reduced absorbance values compared to the growth control. When the particle suspensions were diluted by a factor of 10, and then used to coat tissue culture surfaces, no change in cell viability was observed compared to controls, and therefore this concentration was selected for the cytokine release tests. As shown in Fig. 3(B), GPC dissolution ions had a similar effect on cell viability: while total extracts from both GPC compositions significantly reduced cell proliferation after 24 hours, the diluted ionic extracts did not have an effect compared to growth controls. This indicates that the concentration of ions in diluted GPC extracts are in the subtoxic range and can therefore be utilized for cytokine release experiments.
Effects of (A) GPC particles and (B) GPC conditioned medium on PBMC viability after 24 hours, compared to growth control. Complete extracts and 1/10 dilutions were tested for each composition. *Corresponds to a statistically significant difference compared to growth control (p < 0.05).
Cytokine secretion by PBMCs cultured in GPC conditioned medium or on surfaces coated with GPC particles was assessed for TNF-α, IL-6, IL-1β and IL-10 using ELISA assays. The concentration of specific cytokines was extrapolated using standard curves of absorbance generated for each assay and are shown in Fig. 4. For each group (control or test group), the secretion of cytokines was assessed in the presence (+LPS) or absence (control) of LPS. As expected, the addition of LPS induced significant increases in the secretion of pro-inflammatory cytokines (TNF-α, IL-6, IL-1β) compared to control, for all the groups tested. Regarding IL-10 secretion, the statistical analysis did not find that LPS induced a significant increase in this cytokine.
No significant changes in IL-1β secretion were detected between any of the GPC groups and the medium control, for unstimulated or stimulated cells (Fig. 4A). Regarding TNF-α secretion from unstimulated PBMCs, a significant increase was observed in response to particles from GPC A, compared to control cells (p = 0.0011) (Fig. 4B). The results also demonstrate that TNF-α secretion was significantly suppressed in stimulated cells cultured with dissolution ions from GPC A, compared to the control cells exposed to LPS (p = 0.027). None of the other GPC groups had significant effects on TNF-α secretion. Cells cultured on surfaces coated with particles from GPC A were also found to exhibit significantly increased IL-6 production compared to the control cells (p = 0.0045) (Fig. 4C), however no significant changes occurred in the amount of IL-6 secreted in the presence of LPS. Lastly, no significant differences in IL-10 secretion were detected for any of the GPC groups compared to growth control as determined by one-way ANOVA. Neither the particles from GPC B nor ionic products had any effect on cytokine secretion, for both stimulated and unstimulated cells.
Cytokine release by PBMC’s incubated with GPC A or GPC B ionic products/particles. Titration of IL-1β (A), TNF-α (B), IL-6 (C), IL-10 (D) released after 24-hours of culture in the absence (control) and presence of LPS, an inflammatory stimulus. ∗Corresponds to a statistically significant difference compared to growth control (p < 0.05) and 𝜙 corresponds to a statistical significance compared to LPS stimulated controls (p < 0.05).
GPCs have potential as orthopedic bioadhesives, as they are injectable, bioactive and have suitable mechanical properties for fracture fixation [17,37]. Moreover, the ability of GPCs to release cell stimulating ions after implantation into bone gives them a significant clinical advantage over bioinert adhesives. We previously reported that osteogenic performance and in vitro biocompatibility were largely dependent on GPC composition (particle size, PAA molecular weight, powder: liquid ratio), due to differential ion release. However, the ability of GPC ionic products to modulate immune function has not yet been investigated until now. Additionally, given the presence of bioactive glass in GPCs, it is also of great importance to investigate whether unreacted glass particles released from these cements interact with immune cells to modulate their function. Indeed, the present study indicates that unreacted glass can be released from unset cements. This is notable given that the release of particles from HA coatings during wear can become engulfed by macrophages, promoting an inflammatory reaction associated with the aseptic loosening and implant failure [5].
This study was designed to assess the inflammatory response of PBMCs to two GPCs formulated from the same glass precursor, and to determine if cytokine release was dependent on (a) GPC ion release or (b) GPC particle release. Additionally, the ability of these GPC products to manipulate the inflammatory response induced by stimulation with LPS was also assessed. The current study indicates that significant differences occur depending on the composition of the GPC, and that both ion release and particle release play a role in immunomodulation. The most pronounced response observed in the current study was the significant reduction in TNF-α released by cells stimulated with LPS in the presence of GPC A ionic products (compared to stimulated cells alone). This indicates the potential ability of strontium/zinc-doped GPCs to inhibit the secretion of inflammatory cytokines in the presence of an inflammatory stimulus. We also found that GPC A particles significantly enhanced both TNF-α and IL-6 production in unstimulated cells when compared to growth controls. This suggests that the unreacted bioactive glass particles released from GPCs can trigger an inflammatory response, which may be detrimental to the later stages of bone healing and remodeling. TNF-α, IL-6 and IL-1β are inflammatory cytokines produced by macrophages/monocytes during acute inflammation and are known to play an essential role in osteoclast differentiation and recruitment [38]. IL-10, a inhibitor of IL-6 and TNF-α synthesis, is also produced following the phagocytosis of particles and plays a homeostatic role to resolve inflammation and prevent extracellular matrix damage [5,39].
GPC A ionic products significantly reduced TNF-α secretion from PBMC’s stimulated with LPS, suggesting that these products can act as anti-inflammatory compounds in the presence of an inflammatory stimulus. To our knowledge, this work is the first study that shows GPC products have an anti-inflammatory effect in vitro. ICP analysis determined that GPC A released higher concentrations of Ca, Sr, and Zn than GPC B, which may explain enhanced immunomodulatory effects. Sr-substituted bioactive glass extracts were recently found to reduce IL-6 secretion and promote the M2 phenotype in macrophages, which is associated with inflammation resolution and the promotion of osteogenesis [27]. Zinc has known anti-inflammatory properties and dietary Zn deficiency has been shown to lead to impaired immune function, increasing susceptibility to infections and disease. The importance of Zn for proper immune function is highlighted in the autosomal recessive disease acrodermatitis enteropathica, where intestinal Zn intake is low due to a mutated transporter [28]. The symptoms include very low lymphocyte numbers, atrophy of the thymus, and susceptibility to infections. Elevated Zn levels in the extracts from GPC A may have contributed to its significant anti-inflammatory effect regarding TNF-α secretion. Indeed, a previous study found that zinc treatment (20 μM) significantly suppressed the production of TNF-α from PBMC’s stimulated with LPS, specifically via a protein-kinase A dependent mechanism [40].
PBMCs cultured on surfaces coated with particles from GPC A exhibited elevated secretion of pro-inflammatory cytokines (IL-6 and TNF-α), which was found to be significant compared to cells cultured in media alone. These findings support former studies using particulate bioactive glass and HA [23,25,26]. Bosetti et al. reported that 45S5 Bioglass powders significantly increased the secretion of TNF-α and increased the expression of TNF-α mRNA relative to unstimulated peritoneal macrophages and monocytes [25]. Bendall et al. tested three concentrations of Bioglass particles (1.0, 10, 100 μg/ml) and found that all three increased TNF-α secretion in human rheumatoid synoviocytes above control cultures [26]. In contrast to the present study, Lima et al. reported no effect of silver doped bioactive glass particles on IL-6 or TNF-α [41]. Neither particles from GPC A or GPC B had any effect on cytokine secretion from stimulated PBMCs, which contrasts with a report by Day and Boccaccini who found that surfaces coated with 45S5 glass particles produced a significant reduction in IL-6 and TNF-α from monocytes [23].
The size and concentration of released particles was found to differ significantly between the two compositions, which likely explain the current findings. The size of particles released from GPCs was found to vary between the two compositions: GPC A released particles between 1 and 10 μm and GPC B released particles between 40 and 50 μm. The relationship between particle size and inflammatory response has not been widely explored for bioactive glasses, however, the physical characteristics of HA, polymethylmethacrylate (PMMA) and polyethylene particles have been shown to greatly influence their immunogenic response. For example, Laquerriere et al. reported that the smallest HA particles (∼2 μm) induced the largest increase in TNF-α and IL-6 when incubated with human monocytes, in contrast to larger sized particles (44 μm, 140 μm, 217 μm) [32]. In an in vivo study, Malard et al. implanted rat bones with HA particles in different particle size ranges (10–20, 80–100, and 200-400 μm) [42] and found the smallest HA particles induced the largest acute inflammatory reaction after implantation into a bone defect. In vitro studies utilizing macrophage cultures have also demonstrated PMMA and polyethylene particles to be most biologically active at particles sizes below 10 μm, as evidenced by significantly increased TNF-α, IL-6 and matrix metalloproteinases [43,44]. Gonzàlez et al. determined that macrophage activation was highly dependent on both the size and concentration of PMMA particles, with the concentrations of 1010–1011 particles/ml having the most significant effect on cytokine secretion from primary human monocytes (particle size of 0.325 μm) [43].
These findings are also in line with clinical outcomes. It is well documented that the majority of particulate debris retrieved from periprosthetic tissue of aseptically loosened implants is smaller than 5 μm and irregularly shaped [45–49]. Debris size impacts aseptic loosening given that phagocytosis is one of the main cellular responses to released particles. Indeed, several reports have found that macrophages undergo maximum particle internalization when the size range of particles is between 1–3 μm, which also corresponds to the size of most bacteria [50]. While the exact cellular pathways remain unclear, it has been shown that HA-mediated TNF-α secretion is dependent on the activation of toll-like receptors on macrophages [51]. It is important to note that phagocytosis is not the only mechanism involved to explain macrophage activation in response to particles. One study found that the release of TNF-α and IL-6 from macrophages was not affected when cell phagocytosis was inhibited, suggesting that particles can activate the secretion of cytokines via another mechanism, notably solely by binding to extracellular receptors [52].
Differential ion release between the two GPCs can be explained by the formulations of the putties that were used since the bioactive glass was not changed. Ions elute from the surface of bioactive glass particle during the ‘acid attack phase’ of the GPC setting reaction. This process results in the liberation of a large quantity of ions, which serve to cross-link PAA chains, or become released from the surface of GPCs immediately after immersion/implantation. GPC A most likely released a higher concentration of ions (Ca, Sr, Zn) than GPC B due to the reduced particle size and lower molecular weight PAA used to formulate this cement. The chemical mechanism for this effect is described in greater detail in our previously published report [22], but briefly, the higher surface area available for attack results in more ions being released, and the shorter molecular weight PAA chains result in less chain entanglements, producing a more viscous adhesive with longer setting time. The reverse applies to GPC B, which utilized a larger particle size bioactive glass and higher molecular weight PAA, resulting in lower ion release and shorter setting time. The size of released unreacted glass particles corresponds to the differently sized glasses used to formulate the GPCs: <45 μm and 45–63 μm. The faster setting time of GPC B may explain the decreased concentration of particles released relative to GPC A. Indeed, given that GPC B would have been more set at the time of submersion in deionized water, it would be more difficult for particles to be released.
The increased ion release and smaller size of particles released from GPC A most likely explain its ability to alter immune cells cytokine secretion. Although ion release from GPC A may aid to subside an inflammatory reaction after implantation in vivo, the release of small particles may lead to a prolonged, and unwanted chronic inflammatory state. Recently, Mehrvar et al. implanted GPCs based on the GPC B formulation (45–63 μm) into tibial defects in a sheep model and reported good implant/bone contact and no bone resorption or inflammatory reaction after 16 weeks, compared to PMMA which elicited a mild inflammatory response at this time point [31]. Promisingly, increased bone density was also observed around the GPC implants at 16 weeks. Although GPC A has not been utilized for an in vivo study, Hasandoost et al. implanted a tantalum-containing GPC with a glass particle size range of <45 μm into a sheep model and found extensive bone resorption around the implant after both the 6- and 12-week time points [53]. Given our current results, this may be explained by the release of small particles, leading to increased pro-inflammatory cytokine release triggering osteoclast bone resorption. Although GPC B had neither a positive nor negative effect on cytokine secretion in our study, this may be a favorable characteristic given recent in vivo findings. However, further pre-clinical experimentation is required to determine whether particle release occurs after implantation, and if this release coincides with increased inflammation/ bone resorption.
This study does present several limitations, notably, the concentrations of ionic products and released particles determined to be subtoxic in this study may differ greatly in an in vivo model, given the natural buffering systems present in blood. Moreover, this study does not consider the release of wear particles from set cements, which may similarly interact with immune cells. Nevertheless, the early release of particles during the initial setting reaction would impact bone healing and could potentially prolong inflammation and impair bone-implant bonding.
Conclusion
This study attempted to model the clinical scenario as best as possible by utilizing PBMC’s which consist of monocytes and lymphocytes, cells known to interact with implanted biomaterials and have been observed in proximity to implants. In summary, based on our data, cytokine modulation from GPCs was dependent on GPC composition. We found that while GPC B was not immunogenic, GPC A ionic products and released particles had divergent effects on cytokine secretion. GPC A ions reduced the secretion of TNF-α in stimulated cells while GPC A particles enhanced TNF-α secretion and IL-6 secretion in unstimulated cells. These effects corresponded with differential ion release and particle size/concentration. While many previous studies have focused on cytotoxicity to determine a materials biocompatibility, very few have explored immunomodulation with respect to GPCs. This study represents the first to measure and characterize debris release from unset GPCs and suggests that small particles released during GPC setting may produce an inflammatory response in vivo.
Footnotes
Conflict of interest
The authors declare no potential conflict of interest.
Data availability
The data that support the findings of this study are available from the corresponding author upon reasonable request.
Funding
This work was funded by the Canadian Institute of Health Research (CIHR) Project [appl. #399463].
Author contributions
DM conceived the presented idea, performed the methodological evaluation and subsequent data collection and analysis. DM wrote the first draft of the manuscript. MT and MP supervised the project and critically revised the manuscript.