Acta Biomaterialia 11 (2015) 459–466
Contents lists available at ScienceDirect
Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat
Scavenging effect of Trolox released from brushite cements Gemma Mestres a, Carlos F. Santos a, Lars Engman b, Cecilia Persson a, Marjam Karlsson Ott a,⇑ a b
Division of Applied Materials Science, Department of Engineering Sciences, Uppsala University, Sweden Department of Chemistry, BMC, Uppsala University, Sweden
a r t i c l e
i n f o
Article history: Received 5 April 2014 Received in revised form 30 July 2014 Accepted 5 September 2014 Available online 16 September 2014 Keywords: Brushite cement Drug release Antioxidant Trolox Macrophages
a b s t r a c t In this study a brushite cement was doped with the chain-breaking antioxidant Trolox. The effect of the antioxidant on the physical properties of the cement was evaluated and the release of Trolox was monitored by UV spectroscopy. The ability of the Trolox set free to scavenge reactive oxygen species (ROS) released by macrophages was determined in vitro using a luminol-amplified chemiluminescence assay. Trolox did not modify the crystalline phases of the set cement, which mainly formed crystalline brushite after 7 days in humid conditions. The setting time, compressive strength and morphology of the cement also remained unaltered after the addition of the antioxidant. Trolox was slowly released from the cement following a non-Fickian transport mechanism and nearly 64% of the total amount was released after 3 days. Moreover, the capacity of Trolox to scavenge the ROS released by macrophages increased in a dose-dependent manner. Trolox-loaded cements are expected to reduce some of the first harmful effects of acute inflammation and can thus potentially protect the surrounding tissue during implantation of these as well as other materials used in conjunction. Ó 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
1. Introduction Implantation of a biomaterial in a living tissue causes post-surgical inflammation [1,2]. The origin of this inflammation is partially due to the surgical injury, which interferes with the homeostatic mechanisms that lead to cellular cascades of wound healing [1]. The typical healing response occurs in a very efficient manner and is characterized by four distinct, but overlapping phases: hemostasis, inflammation, proliferation and remodeling [3]. A biomaterial itself can also affect inflammation. Certain chemical or physical properties of biomaterials may trigger a potent and uncontrolled inflammatory reaction. This may cause damage to the surrounding cells and tissue due to prolonged exposure to proinflammatory cytokines, reactive oxygen species (ROS) and destructive enzymes. For instance, ROS such as hydrogen peroxide (H2O2), superoxide (O 2 ) and hydroxyl radicals ( OH) released by immune cells are known to damage proteins and lipids within the cell membrane and to cause strand breaks and nucleobase modifications in DNA due to their powerful oxidant potency [4– 8]. A very potent inflammatory phase could cause a chronic inflammation that might lead to implant failure [9]. By physically or chemically modifying a material (surface and/ or bulk) it is possible to enhance specific cell responses, beneficial ⇑ Corresponding author. Tel.: +46 18 471 79 11; fax: +46 18 471 35 72. E-mail address: [email protected] (M. Karlsson Ott). http://dx.doi.org/10.1016/j.actbio.2014.09.007 1742-7061/Ó 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
for biomaterial integration, and thus improve implant performance [10]. This knowledge can be combined with the concept of drug delivery systems [11]. In order to cope with inflammation and protect the surrounding tissue during implantation, antioxidants can be loaded into the biomaterial to delay or inhibit the oxidation of lipids, proteins or DNA [12]. Trolox (Fig. 1) is a synthetic compound and a derivative of the vitamin E family of chain-breaking tocopherols and tocotrienols. The hydrophobic side-chain attached to the 2-position of the chromanol entity in the naturally occurring compounds has been replaced with a carboxylic acid. Trolox can therefore be viewed as a more water-soluble form of vitamin E [13]. Trolox has been proven to prevent DNA fragmentation of cells grown in the presence of H2O2 [14] and Satoh et al. showed that the antioxidant property of Trolox can surpass that of a-tocopherol [15]. However, to the authors’ knowledge, no clinical studies using this compound have yet been performed. Calcium phosphate cements (CPCs) are commonly used as bone void fillers [16], but also to enhance screw fixation in poor-quality bone [17]. The two most commonly encountered end products are precipitated hydroxyapatite at pH > 4.2 and brushite/monetite at pH < 4.2 [16–18]. Brushite cements are resorbed quicker than apatite cements [19] due to the higher solubility of brushite compared to that of calcium-deficient hydroxyapatite [20], which may be an advantage for bone regrowth. CPCs are used as bone fillers due to their chemical similarity to the mineral phase of bone and excellent biocompatibility. However, for brushite cements, integration
460
G. Mestres et al. / Acta Biomaterialia 11 (2015) 459–466
with the surrounding tissue appears to depend on the b-tricalcium phosphate (TCP)/monocalcium phosphate (MCPM) ratio, and hence the resulting pH surrounding the cement, with a lower ratio giving higher acidity and possibly fibrous tissue surrounding the implant rather than osteointegration [21]. Limiting the inflammatory reaction may especially be desirable when using brushite cements of a certain composition to optimize properties such as mechanical strength [22]. This may be of even higher interest when used in conjunction with other implants, e.g. metallic screws, and/or in the case of open surgery, since these procedures increase the chances of an infection and subsequently inflammation. It should also be noted that biomaterials loaded with vitamin E may reduce bacterial adhesion [23]. CPCs are able to set at body temperature and result in an intrinsically porous material that allows drug diffusion [24]. It is important to study thoroughly each drug–carrier couple since, on the one hand, the drug may cause physicochemical modifications to the cement paste (e.g. setting time, porosity and mechanical properties) [25]; on the other hand, the cement may modify the active principle of the drug due to, for example, local changes of pH or ionic concentration [24]. However, several biologically active molecules or drugs are able to retain their activity within the cement paste [24], making these materials potentially good drug carriers [11]. The aim of this work was to produce and characterize a brushite cement loaded with Trolox. The release of Trolox with time was monitored and the capacity of the released drug to scavenge reactive oxygen species produced by activated inflammatory cells was determined through an in vitro study.
after the start of mixture of the powder and liquid phases until no indentation could be observed on the cement surface when the needle exerting less pressure was used; the final setting time was obtained following the same procedure but applying a heavier and thinner needle. Setting times were evaluated in triplicate. When testing mechanical properties, cement paste was molded in cylindrical silicon molds (diameter = 6 mm, h = 12 mm) and soaked in PBS solution for 7 days. After this time, cement cylinders were manually polished by fitting them in a stainless steel mold and using a silicon carbide paper, P#1200 (Struers, ref. no. 40400023, Cleveland, OH). The mechanical testing was performed in a universal materials testing machine (AGS-X, Shimadzu, Kyoto, Japan) equipped with a load cell of 5 kN at a crosshead speed of 1 mm min1. Eight specimens of each formulation were tested. Specimens molded in silicon molds (diameter = 12 mm, h = 2 mm) were set in 100% humidity conditions. After 7 days a few specimens were sputtered with a mixture of gold and palladium and their microstructure was evaluated by means of a scanning electron microscope (Tabletop Microscope TM-1000, Hitachi, Tokyo, Japan). Another group of samples was finely crushed with a mortar and analyzed by powder X-ray diffraction (XRD, D5000, Siemens) using Bragg–Brentano geometry and Cu Ka radiation. Scanning was performed with a time step of 1.0 s and a scan step of 0.02° min1 between 20° and 50°. The diffraction patterns were compared with the Joint Committee on Powder Diffraction Standards for CaHPO42H2O (brushite, JCPDS No. 009-0077), CaHPO4 (monetite, JCPDS No. 009-0080), b-TCP (JCPDS No. 009-0169) and MPCM (JCPDS No. 009-0347). 2.3. Release of Trolox
2. Materials and methods 2.1. Cement preparation Cement powder was prepared daily by mixing 54 wt.% b-TCP (b-Ca3(PO4)2, Sigma Aldrich, ref. no. 21218, St Louis, MO), 44 wt.% MCPM (Ca(H2PO4)2H2O, Scharlau, ref. no. CA0211005P, Port Adelaide) and 2 wt.% SPP (sodium pyrophosphate, Na4P2O7, Fluka, ref. no. 71499, St Louis, MO) in a powder mixer (TurbulaÒ, Willy A. Bachofen AG Maschinenfabrik, Muttenz, Switzerland) for at least 20 min. SPP was included as a retardant of the reaction. The particle size of b-TCP was 13.6 ± 0.10 lm. MCPM was previously sieved (Retsch, ref. no. 60131000075, Haan, Germany) and only particles 1.0 indicates zero-order release [27]. The model is restricted to a diffusion limit of 60%. 2.4. In vitro cell response A mouse leukemic macrophage cell line (Raw 264.7) was used as the immune cell model. The cells were maintained in cell culture flasks in an incubator with a humidified atmosphere of 5% CO2 in air at 37 °C. DME/F-12 medium (Thermo Scientific HyClone, ref. no. SH300023.01, Logan, UT) supplemented with 10% fetal bovine serum (PAA Laboratories, ref. no. A15-101, Pasching, Austria) and 1% L-glutamine/penicillin/streptomycin (PAA Laboratories, ref. no. P11-013) was used as culture medium. The culture medium was exchanged every second day. Upon 80% confluence, cells were detached by scratching them in a single direction using a cell scraper. (MidSci, ref. no. 99003, St Louis, MO). These cells were then used for the experimental study or seeded again in new flasks. 2.4.1. Cytotoxicity of Trolox A cytotoxicity study was performed using different concentrations of Trolox. In the current study a range of concentrations between 5 and 1000 lM were used. Since Trolox has limited solubility in water, a stock of 100 000 lM was prepared by dissolving it in DMSO (SERVA, ref. no. 39757.02, Heidelberg, Germany). This stock was further diluted with supplemented medium. Macrophages were seeded in a 96-well plate with a cell density of 30 000 cells cm2. After 5 h, the media was removed and fresh media containing different concentrations of Trolox were added. After 1 and 3 days of incubation, cell viability was tested by using the AlamarBlueÒ assay (Invitrogen, ref. no. DAL1100, Carlsbad, CA). Cells were washed once with PBS and afterwards 200 ll of AlamarBlue (5%) diluted in MEM (Life Technologies, Gibco, ref. no. 51200, Carlsbad, CA) was added to each well. After 1 h at 37 °C, fluorescence was read on a microplate reader (Infinite M200, Tecan, Männedorf, Switzerland) at 560 nm excitation and 590 nm emission. The results were converted to cell numbers by using a calibration curve. Quintuplets of each sample were included and the whole experiment was performed three times. 2.4.2. Scavenging effect of Trolox released by the cement The capacity of Trolox released by the cement to scavenge ROS produced by activated macrophages was evaluated through a luminol-amplified chemiluminescence assay [28]. The scavenging effect of three different types of samples (cement alone, Trolox alone and Trolox-loaded cement) collected during the drug release study at different time points was evaluated. Since the luminolamplified chemiluminescence assay is highly sensitive to pH, aliquots from Trolox-loaded cement and cement alone were neutralized with NaOH (40 ll, 0.1 M) to reach a pH between 7.2 and 7.4. The same volume of PBS was added to the aliquots containing Trolox alone as their pH was already within the desired range. To determine the antioxidant effect of Trolox of each aliquot, 50 ll of the aforementioned samples were added to a white 96well plate. 50 ll of a suspension of 2 106 cells ml1 in a protein-free media (4PBS:1DMEM/100 mM glucose) activated with 1 lM of phorbol-12-myristate-13-acetate (PMA, Sigma Aldrich, ref. no. P1585) were also added to each well. As controls, both 50 ll of media alone (blank) as well as non-activated cells were included. The well plate was then placed in a microplate reader (Infinite M200, Tekan, Männedorf, Switzerland) set at 37 °C.
100 ll of a luminol solution was automatically dispensed in every well. The luminol solution (500 lM) was prepared by adding 1% luminol (from stock solution) and 0.2% HRP (1 mg ml1) (Jackson Immuno Research, ref. no. 016-030-084, West Grove, PA) in a 4PBS:1DMEM/100 mM glucose solution. The luminol stock solution was previously prepared by dissolving 50 mM of luminol (3-aminophthalhydrazide, Fisher Scientific, ref. no. A/3150/44, Waltham, MA) in 0.2 M NaOH. Luminescence was measured every 2 min for a total of 60 min, using an integration time of 1000 ms and a settle time of 150 ms. To prevent exposure to light the experimental procedure was performed in a dark room. Triplicates were included in each experiment and the whole experiment was performed using three sets of samples from independent drug release experiments. 2.5. Statistics Statistical analysis was done in IBM SPSS Statistics 19 (IBM, Chicago, IL) using one-way ANOVA at a significance level of a = 0.05. Scheffé’s post hoc test was used and in the event that equal variances could not be confirmed, Tamhane’s post hoc test was applied. 3. Results 3.1. Cement characterization The initial and final setting times of brushite cements were approximately 8.5 and 11 min, respectively. When Trolox was included in the powder, the resulting cement had similar setting times as the cement without drug (Table 1). The crystalline phases present in the set cement after 7 days were evaluated by XRD (Fig. 2). The results showed that the main phase formed was brushite both when Trolox was added and not. Traces of monetite and b-TCP were also detected in both cement types. In contrast, no MCPM was detected. The compressive strength of brushite cement was also tested after 7 days (Fig. 3). Brushite cement had a compressive strength of 25.9 ± 4.1 MPa. This value remained similar when the cement was prepared in the presence of Trolox (25.5 ± 6.6 MPa). The morphology of brushite cement after 7 days was constituted by rhombohedral crystals between 1 and 40 lm in size precipitated in random directions. The cement’s microstructure was similar for the samples containing Trolox (Fig. 4). 3.2. Release of Trolox Fig. 5 a shows the concentration of Trolox detected at each time point by UV spectroscopy for Trolox-loaded cement, cement alone and Trolox alone. To determine the degradation pattern of Trolox, the drug was dissolved in PBS (control) and its concentration was quantified at regular time intervals. As shown in Fig. 5b, the concentration of Trolox alone remained constant for the first 7 h, thereafter decreasing linearly for 3 days, after which the concentration was once again stabilized. After 3.5 days, 50% of the Trolox was degraded. The degradation profile was used to correct the Trolox released from the brushite cement.
Table 1 Initial (tI) and final (tF) setting time of brushite cement and Trolox-loaded cement (0.4 wt.% Trolox). Data are indicated as average ± standard deviation.
Cement Trolox-loaded cement
tI (min)
tF (min)
8.4 ± 0 4 8.3 ± 0.5
11.2 ± 0.7 11.3 ± 0.9
462
G. Mestres et al. / Acta Biomaterialia 11 (2015) 459–466
After correcting for degradation, the percentage of Trolox released from loaded cements immersed in a PBS solution is shown in Fig. 5c. 64% of the total amount of Trolox loaded into the cements was released after 5 days. To determine the release mechanism, the percentage of released Trolox was plotted following Eq. (2) (Fig. 5d). The concentration of Trolox showed a good fit (r2 = 0.989) with time. As shown in Table 2, the diffusional exponent n was between 0.45 and 1, which indicated that Trolox was released through an anomalous (non-Fickian) transport.
Fig. 2. X-ray diffraction after setting the cements in 100% humidity for 7 days. AU, arbitrary units.
Fig. 3. Compressive strength after setting the cements in PBS for 7 days. Error bars indicate the standard deviation (n = 8).
3.2.1. Cytotoxicity of Trolox The cytotoxicity of Trolox was evaluated by incubating macrophages with different concentrations of this compound (0– 1000 lM) for 1 and 3 days (Fig. 6). At both time points, no difference in cell viability was seen for cells incubated with 0–100 lM of Trolox, and only concentrations of 250 and 1000 lM caused a significant decrease (P < 0.05) in cell number as compared to normal media. The number of cells when exposed to 0–250 lM of Trolox increased between 4- and 6-fold from day 1 to day 3 (P < 0.05). However, cell division was slowed down for cells exposed to 1000 lM of Trolox, and only a very slight increase in cell number was observed at day 3 (P > 0.05). 3.2.2. Scavenging capacity of Trolox released from cement The ability of Trolox-loaded cements to scavenge ROS produced by macrophages was evaluated. For this purpose, aliquots obtained from the drug release study were used. Three different aliquots were evaluated: cement alone, Trolox alone and cement loaded with Trolox. Fig. 7 shows the kinetic evolution of luminescence resulting from oxidation of luminol by ROS, when aliquots from four representative time points (30 min, 2 h, 4 h and 22 h) were used. When exposing the macrophages to the aliquots of cement alone, an intense peak with a maximum intensity between 5 and 10 min was observed. In contrast, the aliquot of Trolox alone caused a very low luminescence signal, indicating the ability of Trolox to scavenge ROS, thus avoiding the oxidation of luminol. Most importantly, the aliquots obtained from Trolox-loaded
Fig. 4. Representative images of the morphology of brushite cement (a and b) and the cement loaded with Trolox (c and d), shown at two different magnifications (1000 and 5000).
G. Mestres et al. / Acta Biomaterialia 11 (2015) 459–466
463
Fig. 5. (a) Concentration of Trolox (corrected for sampling). (b) Degradation profile of Trolox (corrected for sampling), with an initial concentration of 215 ± 35 lM (100%) and 100.5 ± 14 lM (49%) after 4.25 days. (c) Percentage of Trolox released (corrected for both sampling and degradation). (d) Korsmeyer–Peppas plot of the Trolox released (corrected for both sampling and degradation). Error bars indicate the standard deviation (n = 3).
Table 2 Fitting parameters for the release curves of Trolox assuming the Korsmeyer–Peppas model. Drug release (%)
k
n
R2
64.4
0.254
0.805
0.989
Fig. 6. Cytotoxicity assay performed incubating macrophages with different concentrations of Trolox (0–1000 lM) for 1 and 3 days. Error bars indicate the standard deviation (n = 5). Statistically significant differences indicated with an asterisk (P < 0.05).
cements also showed a decrease in the peak intensity in comparison with the pristine cement, thus indicating that Trolox still had antioxidant capacity after being mixed into and released from brushite cement. Fig. 8 shows the kinetic evolution of luminescence only from aliquots of Trolox-loaded cement collected at different time points. As seen in Fig. 8, the longer the time of drug release, the higher the scavenger effect.
4. Discussion The aim of this work was to load a calcium phosphate cement with an antioxidant compound and study the scavenging ability of this compound when released. Trolox was the model compound selected; it is known to protect cells from reactive molecules that could cause DNA fragmentation or damage to cell membranes [14]. The drug carrier used in this study was brushite cement. Many methods of producing this cement have been previously described using different compounds as their main reagents, e.g. b-TCP and MCPM [29,30], b-TCP and H3PO4 [30], or tetracalcium phosphate, MCPM and CaO [31]. Although brushite cements are considered per se biocompatible materials, the low pH at the initial stages of the reaction could impair this property [30]. The pH has been reported to drop to 2.5 for b-TCP + MCPM compositions, or even to lower values for the first 30 s when using H3PO4 instead of MCPM [30]. In vivo, osteointegration of the cement has a greater chance of occurring if this low pH is rapidly buffered by physiological solution. However, more acidic formulations (e.g. b-TCP/MCPM in a low ratio) may produce fibrosis around the implant if the initial inflammatory response is not controlled [21]. Other limitations commonly noted for brushite cements are their low injectability, low mechanical strength [29] and short setting times [32]. In this study, the initial and final setting times of the brushite cements were around 8.5 and 11 min, respectively (Table 1). 2 wt.% of SPP was added in the powder phase as a retardant [30,33] in order to delay the initial reaction and thereby increase the cements’ working time. The initial and final setting times were within the suggested range for clinical applications [34]. Trolox, which is an antioxidant compound, was selected for its ability to scavenge ROS released by activated inflammatory cells, e.g. after implantation. A preliminary study based on exposure of inflammatory cells to different concentrations of Trolox (data not shown) indicated that a Trolox concentration between 1 and 20 lM was needed in order to effectively scavenge ROS without harming the cells. In the current work, brushite cement was therefore loaded with a fixed amount of Trolox (0.4 wt.%), and the
464
G. Mestres et al. / Acta Biomaterialia 11 (2015) 459–466
Fig. 7. Kinetic luminescence reaction caused by the oxidation of luminol by ROS (luminescence signal is inversely proportional to the amount of active Trolox). Aliquots from solutions containing Trolox alone, cement alone and Trolox-loaded cement, retrieved at four representative time points: (a) 30 min, (b) 2 h, (c) 4 h and (d) 22 h.
Fig. 8. Kinetic luminescence reaction caused by the oxidation of luminol by ROS (luminescence signal is inversely proportional to the amount of active Trolox). Aliquots from solutions containing Trolox-loaded cements withdrawn at different time points.
scavenging ability of aliquots released at different time points was then evaluated. The addition of Trolox did not have any significant effect on the setting time of the cement. After 7 days in 100% humidity, set cements were mainly constituted by brushite (Fig. 2) and small traces of monetite and b-TCP were found. No MCPM was detected due to its high solubility. After MCPM dissolution, the low pH of the cement paste initially assisted the solubilization of b-TCP [35], after which Ca and P ions released by both reagents caused a supersaturation in the media and, finally, brushite precipitated. The formation of brushite indicated that the pH of the cement paste was
Contents lists available at ScienceDirect
Acta Biomaterialia journal homepage: www.elsevier.com/locate/actabiomat
Scavenging effect of Trolox released from brushite cements Gemma Mestres a, Carlos F. Santos a, Lars Engman b, Cecilia Persson a, Marjam Karlsson Ott a,⇑ a b
Division of Applied Materials Science, Department of Engineering Sciences, Uppsala University, Sweden Department of Chemistry, BMC, Uppsala University, Sweden
a r t i c l e
i n f o
Article history: Received 5 April 2014 Received in revised form 30 July 2014 Accepted 5 September 2014 Available online 16 September 2014 Keywords: Brushite cement Drug release Antioxidant Trolox Macrophages
a b s t r a c t In this study a brushite cement was doped with the chain-breaking antioxidant Trolox. The effect of the antioxidant on the physical properties of the cement was evaluated and the release of Trolox was monitored by UV spectroscopy. The ability of the Trolox set free to scavenge reactive oxygen species (ROS) released by macrophages was determined in vitro using a luminol-amplified chemiluminescence assay. Trolox did not modify the crystalline phases of the set cement, which mainly formed crystalline brushite after 7 days in humid conditions. The setting time, compressive strength and morphology of the cement also remained unaltered after the addition of the antioxidant. Trolox was slowly released from the cement following a non-Fickian transport mechanism and nearly 64% of the total amount was released after 3 days. Moreover, the capacity of Trolox to scavenge the ROS released by macrophages increased in a dose-dependent manner. Trolox-loaded cements are expected to reduce some of the first harmful effects of acute inflammation and can thus potentially protect the surrounding tissue during implantation of these as well as other materials used in conjunction. Ó 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
1. Introduction Implantation of a biomaterial in a living tissue causes post-surgical inflammation [1,2]. The origin of this inflammation is partially due to the surgical injury, which interferes with the homeostatic mechanisms that lead to cellular cascades of wound healing [1]. The typical healing response occurs in a very efficient manner and is characterized by four distinct, but overlapping phases: hemostasis, inflammation, proliferation and remodeling [3]. A biomaterial itself can also affect inflammation. Certain chemical or physical properties of biomaterials may trigger a potent and uncontrolled inflammatory reaction. This may cause damage to the surrounding cells and tissue due to prolonged exposure to proinflammatory cytokines, reactive oxygen species (ROS) and destructive enzymes. For instance, ROS such as hydrogen peroxide (H2O2), superoxide (O 2 ) and hydroxyl radicals ( OH) released by immune cells are known to damage proteins and lipids within the cell membrane and to cause strand breaks and nucleobase modifications in DNA due to their powerful oxidant potency [4– 8]. A very potent inflammatory phase could cause a chronic inflammation that might lead to implant failure [9]. By physically or chemically modifying a material (surface and/ or bulk) it is possible to enhance specific cell responses, beneficial ⇑ Corresponding author. Tel.: +46 18 471 79 11; fax: +46 18 471 35 72. E-mail address: [email protected] (M. Karlsson Ott). http://dx.doi.org/10.1016/j.actbio.2014.09.007 1742-7061/Ó 2014 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
for biomaterial integration, and thus improve implant performance [10]. This knowledge can be combined with the concept of drug delivery systems [11]. In order to cope with inflammation and protect the surrounding tissue during implantation, antioxidants can be loaded into the biomaterial to delay or inhibit the oxidation of lipids, proteins or DNA [12]. Trolox (Fig. 1) is a synthetic compound and a derivative of the vitamin E family of chain-breaking tocopherols and tocotrienols. The hydrophobic side-chain attached to the 2-position of the chromanol entity in the naturally occurring compounds has been replaced with a carboxylic acid. Trolox can therefore be viewed as a more water-soluble form of vitamin E [13]. Trolox has been proven to prevent DNA fragmentation of cells grown in the presence of H2O2 [14] and Satoh et al. showed that the antioxidant property of Trolox can surpass that of a-tocopherol [15]. However, to the authors’ knowledge, no clinical studies using this compound have yet been performed. Calcium phosphate cements (CPCs) are commonly used as bone void fillers [16], but also to enhance screw fixation in poor-quality bone [17]. The two most commonly encountered end products are precipitated hydroxyapatite at pH > 4.2 and brushite/monetite at pH < 4.2 [16–18]. Brushite cements are resorbed quicker than apatite cements [19] due to the higher solubility of brushite compared to that of calcium-deficient hydroxyapatite [20], which may be an advantage for bone regrowth. CPCs are used as bone fillers due to their chemical similarity to the mineral phase of bone and excellent biocompatibility. However, for brushite cements, integration
460
G. Mestres et al. / Acta Biomaterialia 11 (2015) 459–466
with the surrounding tissue appears to depend on the b-tricalcium phosphate (TCP)/monocalcium phosphate (MCPM) ratio, and hence the resulting pH surrounding the cement, with a lower ratio giving higher acidity and possibly fibrous tissue surrounding the implant rather than osteointegration [21]. Limiting the inflammatory reaction may especially be desirable when using brushite cements of a certain composition to optimize properties such as mechanical strength [22]. This may be of even higher interest when used in conjunction with other implants, e.g. metallic screws, and/or in the case of open surgery, since these procedures increase the chances of an infection and subsequently inflammation. It should also be noted that biomaterials loaded with vitamin E may reduce bacterial adhesion [23]. CPCs are able to set at body temperature and result in an intrinsically porous material that allows drug diffusion [24]. It is important to study thoroughly each drug–carrier couple since, on the one hand, the drug may cause physicochemical modifications to the cement paste (e.g. setting time, porosity and mechanical properties) [25]; on the other hand, the cement may modify the active principle of the drug due to, for example, local changes of pH or ionic concentration [24]. However, several biologically active molecules or drugs are able to retain their activity within the cement paste [24], making these materials potentially good drug carriers [11]. The aim of this work was to produce and characterize a brushite cement loaded with Trolox. The release of Trolox with time was monitored and the capacity of the released drug to scavenge reactive oxygen species produced by activated inflammatory cells was determined through an in vitro study.
after the start of mixture of the powder and liquid phases until no indentation could be observed on the cement surface when the needle exerting less pressure was used; the final setting time was obtained following the same procedure but applying a heavier and thinner needle. Setting times were evaluated in triplicate. When testing mechanical properties, cement paste was molded in cylindrical silicon molds (diameter = 6 mm, h = 12 mm) and soaked in PBS solution for 7 days. After this time, cement cylinders were manually polished by fitting them in a stainless steel mold and using a silicon carbide paper, P#1200 (Struers, ref. no. 40400023, Cleveland, OH). The mechanical testing was performed in a universal materials testing machine (AGS-X, Shimadzu, Kyoto, Japan) equipped with a load cell of 5 kN at a crosshead speed of 1 mm min1. Eight specimens of each formulation were tested. Specimens molded in silicon molds (diameter = 12 mm, h = 2 mm) were set in 100% humidity conditions. After 7 days a few specimens were sputtered with a mixture of gold and palladium and their microstructure was evaluated by means of a scanning electron microscope (Tabletop Microscope TM-1000, Hitachi, Tokyo, Japan). Another group of samples was finely crushed with a mortar and analyzed by powder X-ray diffraction (XRD, D5000, Siemens) using Bragg–Brentano geometry and Cu Ka radiation. Scanning was performed with a time step of 1.0 s and a scan step of 0.02° min1 between 20° and 50°. The diffraction patterns were compared with the Joint Committee on Powder Diffraction Standards for CaHPO42H2O (brushite, JCPDS No. 009-0077), CaHPO4 (monetite, JCPDS No. 009-0080), b-TCP (JCPDS No. 009-0169) and MPCM (JCPDS No. 009-0347). 2.3. Release of Trolox
2. Materials and methods 2.1. Cement preparation Cement powder was prepared daily by mixing 54 wt.% b-TCP (b-Ca3(PO4)2, Sigma Aldrich, ref. no. 21218, St Louis, MO), 44 wt.% MCPM (Ca(H2PO4)2H2O, Scharlau, ref. no. CA0211005P, Port Adelaide) and 2 wt.% SPP (sodium pyrophosphate, Na4P2O7, Fluka, ref. no. 71499, St Louis, MO) in a powder mixer (TurbulaÒ, Willy A. Bachofen AG Maschinenfabrik, Muttenz, Switzerland) for at least 20 min. SPP was included as a retardant of the reaction. The particle size of b-TCP was 13.6 ± 0.10 lm. MCPM was previously sieved (Retsch, ref. no. 60131000075, Haan, Germany) and only particles 1.0 indicates zero-order release [27]. The model is restricted to a diffusion limit of 60%. 2.4. In vitro cell response A mouse leukemic macrophage cell line (Raw 264.7) was used as the immune cell model. The cells were maintained in cell culture flasks in an incubator with a humidified atmosphere of 5% CO2 in air at 37 °C. DME/F-12 medium (Thermo Scientific HyClone, ref. no. SH300023.01, Logan, UT) supplemented with 10% fetal bovine serum (PAA Laboratories, ref. no. A15-101, Pasching, Austria) and 1% L-glutamine/penicillin/streptomycin (PAA Laboratories, ref. no. P11-013) was used as culture medium. The culture medium was exchanged every second day. Upon 80% confluence, cells were detached by scratching them in a single direction using a cell scraper. (MidSci, ref. no. 99003, St Louis, MO). These cells were then used for the experimental study or seeded again in new flasks. 2.4.1. Cytotoxicity of Trolox A cytotoxicity study was performed using different concentrations of Trolox. In the current study a range of concentrations between 5 and 1000 lM were used. Since Trolox has limited solubility in water, a stock of 100 000 lM was prepared by dissolving it in DMSO (SERVA, ref. no. 39757.02, Heidelberg, Germany). This stock was further diluted with supplemented medium. Macrophages were seeded in a 96-well plate with a cell density of 30 000 cells cm2. After 5 h, the media was removed and fresh media containing different concentrations of Trolox were added. After 1 and 3 days of incubation, cell viability was tested by using the AlamarBlueÒ assay (Invitrogen, ref. no. DAL1100, Carlsbad, CA). Cells were washed once with PBS and afterwards 200 ll of AlamarBlue (5%) diluted in MEM (Life Technologies, Gibco, ref. no. 51200, Carlsbad, CA) was added to each well. After 1 h at 37 °C, fluorescence was read on a microplate reader (Infinite M200, Tecan, Männedorf, Switzerland) at 560 nm excitation and 590 nm emission. The results were converted to cell numbers by using a calibration curve. Quintuplets of each sample were included and the whole experiment was performed three times. 2.4.2. Scavenging effect of Trolox released by the cement The capacity of Trolox released by the cement to scavenge ROS produced by activated macrophages was evaluated through a luminol-amplified chemiluminescence assay [28]. The scavenging effect of three different types of samples (cement alone, Trolox alone and Trolox-loaded cement) collected during the drug release study at different time points was evaluated. Since the luminolamplified chemiluminescence assay is highly sensitive to pH, aliquots from Trolox-loaded cement and cement alone were neutralized with NaOH (40 ll, 0.1 M) to reach a pH between 7.2 and 7.4. The same volume of PBS was added to the aliquots containing Trolox alone as their pH was already within the desired range. To determine the antioxidant effect of Trolox of each aliquot, 50 ll of the aforementioned samples were added to a white 96well plate. 50 ll of a suspension of 2 106 cells ml1 in a protein-free media (4PBS:1DMEM/100 mM glucose) activated with 1 lM of phorbol-12-myristate-13-acetate (PMA, Sigma Aldrich, ref. no. P1585) were also added to each well. As controls, both 50 ll of media alone (blank) as well as non-activated cells were included. The well plate was then placed in a microplate reader (Infinite M200, Tekan, Männedorf, Switzerland) set at 37 °C.
100 ll of a luminol solution was automatically dispensed in every well. The luminol solution (500 lM) was prepared by adding 1% luminol (from stock solution) and 0.2% HRP (1 mg ml1) (Jackson Immuno Research, ref. no. 016-030-084, West Grove, PA) in a 4PBS:1DMEM/100 mM glucose solution. The luminol stock solution was previously prepared by dissolving 50 mM of luminol (3-aminophthalhydrazide, Fisher Scientific, ref. no. A/3150/44, Waltham, MA) in 0.2 M NaOH. Luminescence was measured every 2 min for a total of 60 min, using an integration time of 1000 ms and a settle time of 150 ms. To prevent exposure to light the experimental procedure was performed in a dark room. Triplicates were included in each experiment and the whole experiment was performed using three sets of samples from independent drug release experiments. 2.5. Statistics Statistical analysis was done in IBM SPSS Statistics 19 (IBM, Chicago, IL) using one-way ANOVA at a significance level of a = 0.05. Scheffé’s post hoc test was used and in the event that equal variances could not be confirmed, Tamhane’s post hoc test was applied. 3. Results 3.1. Cement characterization The initial and final setting times of brushite cements were approximately 8.5 and 11 min, respectively. When Trolox was included in the powder, the resulting cement had similar setting times as the cement without drug (Table 1). The crystalline phases present in the set cement after 7 days were evaluated by XRD (Fig. 2). The results showed that the main phase formed was brushite both when Trolox was added and not. Traces of monetite and b-TCP were also detected in both cement types. In contrast, no MCPM was detected. The compressive strength of brushite cement was also tested after 7 days (Fig. 3). Brushite cement had a compressive strength of 25.9 ± 4.1 MPa. This value remained similar when the cement was prepared in the presence of Trolox (25.5 ± 6.6 MPa). The morphology of brushite cement after 7 days was constituted by rhombohedral crystals between 1 and 40 lm in size precipitated in random directions. The cement’s microstructure was similar for the samples containing Trolox (Fig. 4). 3.2. Release of Trolox Fig. 5 a shows the concentration of Trolox detected at each time point by UV spectroscopy for Trolox-loaded cement, cement alone and Trolox alone. To determine the degradation pattern of Trolox, the drug was dissolved in PBS (control) and its concentration was quantified at regular time intervals. As shown in Fig. 5b, the concentration of Trolox alone remained constant for the first 7 h, thereafter decreasing linearly for 3 days, after which the concentration was once again stabilized. After 3.5 days, 50% of the Trolox was degraded. The degradation profile was used to correct the Trolox released from the brushite cement.
Table 1 Initial (tI) and final (tF) setting time of brushite cement and Trolox-loaded cement (0.4 wt.% Trolox). Data are indicated as average ± standard deviation.
Cement Trolox-loaded cement
tI (min)
tF (min)
8.4 ± 0 4 8.3 ± 0.5
11.2 ± 0.7 11.3 ± 0.9
462
G. Mestres et al. / Acta Biomaterialia 11 (2015) 459–466
After correcting for degradation, the percentage of Trolox released from loaded cements immersed in a PBS solution is shown in Fig. 5c. 64% of the total amount of Trolox loaded into the cements was released after 5 days. To determine the release mechanism, the percentage of released Trolox was plotted following Eq. (2) (Fig. 5d). The concentration of Trolox showed a good fit (r2 = 0.989) with time. As shown in Table 2, the diffusional exponent n was between 0.45 and 1, which indicated that Trolox was released through an anomalous (non-Fickian) transport.
Fig. 2. X-ray diffraction after setting the cements in 100% humidity for 7 days. AU, arbitrary units.
Fig. 3. Compressive strength after setting the cements in PBS for 7 days. Error bars indicate the standard deviation (n = 8).
3.2.1. Cytotoxicity of Trolox The cytotoxicity of Trolox was evaluated by incubating macrophages with different concentrations of this compound (0– 1000 lM) for 1 and 3 days (Fig. 6). At both time points, no difference in cell viability was seen for cells incubated with 0–100 lM of Trolox, and only concentrations of 250 and 1000 lM caused a significant decrease (P < 0.05) in cell number as compared to normal media. The number of cells when exposed to 0–250 lM of Trolox increased between 4- and 6-fold from day 1 to day 3 (P < 0.05). However, cell division was slowed down for cells exposed to 1000 lM of Trolox, and only a very slight increase in cell number was observed at day 3 (P > 0.05). 3.2.2. Scavenging capacity of Trolox released from cement The ability of Trolox-loaded cements to scavenge ROS produced by macrophages was evaluated. For this purpose, aliquots obtained from the drug release study were used. Three different aliquots were evaluated: cement alone, Trolox alone and cement loaded with Trolox. Fig. 7 shows the kinetic evolution of luminescence resulting from oxidation of luminol by ROS, when aliquots from four representative time points (30 min, 2 h, 4 h and 22 h) were used. When exposing the macrophages to the aliquots of cement alone, an intense peak with a maximum intensity between 5 and 10 min was observed. In contrast, the aliquot of Trolox alone caused a very low luminescence signal, indicating the ability of Trolox to scavenge ROS, thus avoiding the oxidation of luminol. Most importantly, the aliquots obtained from Trolox-loaded
Fig. 4. Representative images of the morphology of brushite cement (a and b) and the cement loaded with Trolox (c and d), shown at two different magnifications (1000 and 5000).
G. Mestres et al. / Acta Biomaterialia 11 (2015) 459–466
463
Fig. 5. (a) Concentration of Trolox (corrected for sampling). (b) Degradation profile of Trolox (corrected for sampling), with an initial concentration of 215 ± 35 lM (100%) and 100.5 ± 14 lM (49%) after 4.25 days. (c) Percentage of Trolox released (corrected for both sampling and degradation). (d) Korsmeyer–Peppas plot of the Trolox released (corrected for both sampling and degradation). Error bars indicate the standard deviation (n = 3).
Table 2 Fitting parameters for the release curves of Trolox assuming the Korsmeyer–Peppas model. Drug release (%)
k
n
R2
64.4
0.254
0.805
0.989
Fig. 6. Cytotoxicity assay performed incubating macrophages with different concentrations of Trolox (0–1000 lM) for 1 and 3 days. Error bars indicate the standard deviation (n = 5). Statistically significant differences indicated with an asterisk (P < 0.05).
cements also showed a decrease in the peak intensity in comparison with the pristine cement, thus indicating that Trolox still had antioxidant capacity after being mixed into and released from brushite cement. Fig. 8 shows the kinetic evolution of luminescence only from aliquots of Trolox-loaded cement collected at different time points. As seen in Fig. 8, the longer the time of drug release, the higher the scavenger effect.
4. Discussion The aim of this work was to load a calcium phosphate cement with an antioxidant compound and study the scavenging ability of this compound when released. Trolox was the model compound selected; it is known to protect cells from reactive molecules that could cause DNA fragmentation or damage to cell membranes [14]. The drug carrier used in this study was brushite cement. Many methods of producing this cement have been previously described using different compounds as their main reagents, e.g. b-TCP and MCPM [29,30], b-TCP and H3PO4 [30], or tetracalcium phosphate, MCPM and CaO [31]. Although brushite cements are considered per se biocompatible materials, the low pH at the initial stages of the reaction could impair this property [30]. The pH has been reported to drop to 2.5 for b-TCP + MCPM compositions, or even to lower values for the first 30 s when using H3PO4 instead of MCPM [30]. In vivo, osteointegration of the cement has a greater chance of occurring if this low pH is rapidly buffered by physiological solution. However, more acidic formulations (e.g. b-TCP/MCPM in a low ratio) may produce fibrosis around the implant if the initial inflammatory response is not controlled [21]. Other limitations commonly noted for brushite cements are their low injectability, low mechanical strength [29] and short setting times [32]. In this study, the initial and final setting times of the brushite cements were around 8.5 and 11 min, respectively (Table 1). 2 wt.% of SPP was added in the powder phase as a retardant [30,33] in order to delay the initial reaction and thereby increase the cements’ working time. The initial and final setting times were within the suggested range for clinical applications [34]. Trolox, which is an antioxidant compound, was selected for its ability to scavenge ROS released by activated inflammatory cells, e.g. after implantation. A preliminary study based on exposure of inflammatory cells to different concentrations of Trolox (data not shown) indicated that a Trolox concentration between 1 and 20 lM was needed in order to effectively scavenge ROS without harming the cells. In the current work, brushite cement was therefore loaded with a fixed amount of Trolox (0.4 wt.%), and the
464
G. Mestres et al. / Acta Biomaterialia 11 (2015) 459–466
Fig. 7. Kinetic luminescence reaction caused by the oxidation of luminol by ROS (luminescence signal is inversely proportional to the amount of active Trolox). Aliquots from solutions containing Trolox alone, cement alone and Trolox-loaded cement, retrieved at four representative time points: (a) 30 min, (b) 2 h, (c) 4 h and (d) 22 h.
Fig. 8. Kinetic luminescence reaction caused by the oxidation of luminol by ROS (luminescence signal is inversely proportional to the amount of active Trolox). Aliquots from solutions containing Trolox-loaded cements withdrawn at different time points.
scavenging ability of aliquots released at different time points was then evaluated. The addition of Trolox did not have any significant effect on the setting time of the cement. After 7 days in 100% humidity, set cements were mainly constituted by brushite (Fig. 2) and small traces of monetite and b-TCP were found. No MCPM was detected due to its high solubility. After MCPM dissolution, the low pH of the cement paste initially assisted the solubilization of b-TCP [35], after which Ca and P ions released by both reagents caused a supersaturation in the media and, finally, brushite precipitated. The formation of brushite indicated that the pH of the cement paste was
