Collagen is increasingly attracting attention for bone tissue engineering applications. However, due to its low mechanical properties, applications including mechanical loads or requiring structural integrity are limited. To tackle this handicap, collagen can be combined with (nanoscale) silica in a variety of composite materials that are attractive for bone tissue engineering. Considering research carried out in the past 15 years, this article reviews the literature discussing the development of silica/collagen composites that have been synthesized by adding silica from different sources as inorganic bioactive material to collagen as organic matrix. Different routes for the fabrication of collagen/silica composites are presented, focusing on nanocomposites. In vitro cell bioactivity studies demonstrated the osteogenic and, in some cases, angiogenic potential of the composites. Relevant in vivo studies discussing integration of the materials in bone tissue are discussed. Due to the understanding of possible interaction between silicon species and collagen, the effect of different silica precursors on the collagen self-assembly process is also discussed. On the basis of literature results and as discussed in this review, collagen/silica nanocomposites and hybrids represent attractive biomaterials for bone regeneration applications.
1.1 Effect of inorganic materials on polymeric composites
The mineralization capacity of a family of materials applied as bone-substitutes is referred to as “bioactivity”, and materials exhibiting this capacity, such as hydroxyapatite (HA), calcium phosphate and bioactive glasses are called bioactive materials . These materials possess the capacity to promote nucleation and subsequent growth of calcium phosphate crystals on their surfaces. Generally, most polymeric materials do not possess this ability, but the addition of inorganic phases can render the resulting composites bioactive by providing nucleation sites for the formation of HA or precipitation of other calcium phosphates . The addition of an inorganic bioactive phase can also provide cell adhesion sites that facilitate integration with surrounding bone tissue [3, 4]. These inorganic materials can create a firm bond with bone at the site of implantation by forming an intermediate layer of HA on their surface [1, 5]. Incorporation of inorganic filler particles into soft polymer matrices affects also the mechanical properties [6–8]. In particular, compression strength and stiffness can be increased by incorporation of inorganic fillers in polymer matrices.
Inorganic fillers can be incorporated into different polymers in the form of micron-sized or nanoscale-sized particles or fibers. The size of the inorganic filler particles is important because it affects the effective properties of the composites. The incorporation of nanosized inorganic filler particles with the desired morphology generally improves the mechanical properties of the resulting composites in comparison with the mechanical properties of the neat polymer and of composites containing micron-sized inorganic filler particles [9, 10]. Nanosized inorganic filler particles provide a higher specific surface area and thus a higher interface area that might enhance the interfacial bonding strength between the inorganic filler particles and the polymer matrix, and thus the overall mechanical properties of the resulting composites could be significantly improved .
One relevant area of application of organic-inorganic composites is in bone tissue scaffolds . Scaffolds for bone tissue engineering should be biocompatible, biodegradable, osteoconductive, osteoinductive and exhibit suitable structural integrity [11, 12].
Silica (SiO2) can be used as a bone-substituting material; it has been reported that silica is biocompatible and osteoconductive . Silicon has been shown to play an important role in bone formation and in the first steps of mineralization on silica/collagen composites [13, 14]. The high density of surface silanol groups (Si-OH) present on the surface of amorphous silica can promote the formation of biologically active bone-like apatite, enhancing the bioactivity of silica-based biopolymer nanocomposites [15, 16]. For example, a crystalline phase precipitates on the surface of silica/collagen composites when soaked in simulated body fluid (SBF), whereas the collagen hydrogel and silica particles alone do not induce this effect. The carboxylate functional group of collagen can efficiently bind calcium ions, and at the same time these calcium ions can associate with the released silicic acid from silica to form a calcium silicate phase that may serve as nucleation site for further calcium phosphate deposition .
Silica can also act as an inorganic phase to enhance the mechanical properties and load-bearing capacity of polymer-based composites [15, 17–20]. An electrostatic interaction between the negatively charged silica species and the positively charged groups of organic polymeric materials, e.g., amine groups of collagen and gelatin , can improve the interaction of SiO2 particles and the organic matrix.
In this context, silica-incorporated proteinous polymeric materials, e.g., forming collagen composite scaffolds, can be considered bone-substituting materials because of their high osteoconductivity, biocompatibility and bioactive properties [18, 21, 22], which has been established in vitro by studies of the adhesion, proliferation and osteogenic differentiation of mesenchymal stem cells [13, 20].
Silica precursors (e.g., sodium silicate, silicon catecholate, orthosilicic acid, etc.) are generally considered for the fabrication of silica-based biopolymer xerogel or hydrogel composite scaffolds for bone tissue engineering . In a recent study, it was also revealed that sol-gel-derived silica has the potential to form new blood vessels surrounding the bone substitute material, confirmed by immunohistochemistry , which is of high relevance for the development of scaffolds for vascularized bone regeneration .
Moreover, mesoporous silica-based biopolymer scaffolds could serve as a controlled drug delivery vehicle for bone tissue engineering  because of interactions taking place between the silanol groups covering the surface of mesoporous channels and the functional groups of the drug . Recently it was found that collagen end-capped mesoporous silica nanoparticles (MSNs) are useful carriers for targeted drug delivery to cancer cells .
Three processes are generally used for conversion of hybrid hydrogels into dry samples, namely, freeze-drying, critical-point drying and ambient drying. Cryogels are obtained by freeze-drying, which can destroy the gel structure due to swelling during freezing. Critical-point drying is the process by which it is possible to obtain a dry sample from a hydrogel without altering its original structure, which is known as aerogel. However, aerogels are fragile and highly susceptible to moisture due to their high porosity. More dense and robust composite xerogels can be obtained by ambient drying [13, 28]. Xerogels are prepared by controlling the evaporation rate of a liquid from the hydrogel. A capillary pressure (P) generates during evaporation of the liquid from the hydrogel, given in Equation (1), where γ is liquid surface tension, θ is contact angle and r is the radius of pore :
The stability of the xerogel strongly depends on the generated capillary force during evaporation of the liquid. If the capillary forces exceed the xerogel strength, cracks are formed and make it unstable. Thus, the capillary forces can be controlled by controlling the surface tension of the liquid, which can be obtained by changing temperature, relative humidity, vapor pressure, etc. .
The increasing number of investigations reported in the literature on innovative applications of silica-collagen nanocomposites has motivated the preparation of this review paper. A comprehensive analysis of the literature on the development, characterization and application of nano silica-incorporated collagen-based composites is presented, with focus on applications in bone tissue engineering.
1.2 Collagen as a biopolymer
Collagen as a major constituent of the extracellular matrix is a very important material for biomedical applications because of its natural abundance, biodegradability and biocompatibility [14, 29–32].
Natural bone is inherently composed of collagen fibrils that are mineralized by calcium phosphate phases (CPPs) analogous to HA . Collagen is the most abundant protein in mammals and forms a structural network of mostly all tissues because of a high degree of polymorphism that leads to the formation of a variety of different structures [18, 33]. Collagen is a useful biomaterial because of its relevant properties, such as high affinity to water, controllable biodegradation, hemostatic properties, low inflammatory response, low cytotoxicity and ability to facilitate cellular attachment [34, 35]. A significant property of collagen is its structural integrity to serve as a template for deposition of calcium phosphate and calcium carbonate in bone, which makes it a very convenient material for bone tissue engineering .
The potential drawback of pure collagenous materials, e.g., to be used as structural scaffold in bone regeneration approaches, is that they do not possess adequate mechanical stability, which is usually required for cell culture tests in vitro, in vivo implantation, and to provide structural functionality and biophysical stimuli to cells for bone regeneration [28, 36]. To overcome this drawback, inorganic bioactive materials are added to collagen for fabrication of collagen/inorganic composites [18, 22, 28, 37]. There are many types of collagen/inorganic materials developed so far, namely, (a) binary systems, such as collagen/HA [38–41], collagen/tricalcium phosphate (TCP) , collagen/silica [14, 21, 30] and collagen/bioactive glass [11, 14, 43, 44]; (b) ternary systems, such as collagen/HA/silica [19, 28]; and (c) quaternary systems like collagen/HA/TCP/silica . Both biological and mechanical properties of such composite materials are essential for their successful use as scaffolds in bone tissue engineering. The next paragraphs will review an important subgroup of such collagen-based composites, e.g., those incorporating silica as the inorganic phase, describing fabrication technologies, properties and applications.
2 Silica/collagen composite scaffolds
2.1 Freeze-drying-based materials
Freeze-drying is one of the most used techniques for fabrication of composite scaffolds with highly porous structure. Silica-collagen hybrid composite scaffolds [19, 22] and membranes  have been produced by freeze-drying technique for bone tissue engineering. Different silica precursors have been used for making such silica-collagen hybrid composites; in some cases the composites also have incorporated calcium and phosphorous.
Heinemann et al.  used tetraethoxysilane (TEOS) as silica precursor that was hydrolyzed to obtain orthosilisic acid as silica source. Initially, in order to induce fibrillogenesis, type I collagen was mixed with phosphate buffer solution at 37°C and pH 7.4 after purification by salt precipitation and dialyzing against deionized water. Then collagen fibrils were extracted and resuspended in 0.01 m Tris-HCl buffer at pH 7.4. Silicic acid was added to the collagen suspension with a silica/collagen ratio of 25:75, which was followed by vigorous vortexing and freeze- drying. The freeze-dried scaffolds were immersed in 1 wt% N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide/N-hydroxysuccinimide in 40% ethanol for 24 h to allow to cross-link. The cross-linked scaffolds were again freeze-dried after rinsing in deionized water.
In a related study, Ghanaati et al.  used freeze-drying process to fabricate highly porous BONITmatrix®-collagen scaffold (collagen-embedded calcium phosphate-silica) to evaluate the osteoblastic behavior in vitro and in vivo. BONITmatrix® is a fully synthetic highly porous biomaterial, based on a biphasic mixture of 87% calcium phosphates and 13% SiO2. The nanoporous BONITmatrix® granules were produced by a sol-gel process. In this method, calcium phosphate was put into alkoxide-based silica sol and stirred homogeneously. During this process, silica was interconnected with calcium phosphate and the gel was dried at 200°C to get the porous structure. Then the hybrid nanoporous materials were mixed with 1% collagen solution followed by freeze-drying to fabricate porous scaffolds.
Lee et al.  used calcium-containing silica xerogel to fabricate silica-collagen nanohybrid membrane by freeze-drying for guided bone regeneration. Calcium-containing silica xerogel was prepared by adding calcium chloride (CaCl2) and triethylphosphate to a solution of tetramethylorthosilane to obtain the following specific composition (in wt%): 80% SiO2-15% CaO-5% P2O5. The polymerized silica sols were added to the collagen solution in different weight ratios (collagen/silica=90:10, 80:20, 70:30 and 60:40) and mixed homogeneously. The resulting hybrid solutions were freeze-dried followed by warm-pressing at 40°C. Then these membranes were again freeze-dried after washing with phosphate-buffered saline (PBS).
The presence of homogenously distributed nanosized silica xerogel particles was confirmed by transmission electron microscopy (TEM) and energy-dispersive X-ray spectroscopy (EDX) profile . Furthermore, there were no by-products formed in silica-collagen nanohybrid membranes due to the incorporation of silica sols in the collagen matrix, as confirmed by Fourier transform infrared spectroscopy (FTIR).
Silica-collagen nanohybrid membranes showed excellent bioactive characteristics because of the formation of small and numerous apatite crystals on the surface of the nanohybrid membrane after immersion in SBF. The pure collagen membrane did not exhibit such characteristics in SBF, although the carboxyl anionic groups of amino acids of collagen could act as a nucleation site for mineralization .
Heinemann et al.  revealed that 85–95% of seeded human bone marrow stromal cells (hBMSCs) were adhered on silica-collagen nanocomposite scaffolds after 1 day of culture, indicating good initial adherence. Moreover, hBMSCs were seen to significantly proliferate at day 14 and the cell number was increased four to six times during 28 days of cultivation. Surprisingly, it was found that the alkaline phosphatase (ALP) activity of non-induced hBMSCs on hybrid scaffolds was lower compared to that of pure collagen scaffolds at 14 and 21 days of cultivation, but the values of ALP activities of both scaffolds seemed to come to the same value at 28 days of cultivation.
In the study of Lee et al. , round clusters of MC3T3 cells were observed on pure collagen membrane; the cells were dispersed as individual entities with increasing silica xerogel content in the collagen-silica nanohybrid membrane, and no round cell clusters were found in collagen-30% silica membrane, which indicated excellent adhesion and spreading characteristics. There was no significant difference in cell proliferation between pure and hybrid collagen membranes; however, hybrid membranes showed higher ALP activity than pure collagen membrane. Among all hybrid membranes investigated, collagen membrane containing 30% silica exhibited the highest ALP activity, which was twice that of the pure collagen membrane. These results confirmed that the hybridization of silica xerogel with collagen expressed the osteoblastic phenotype more effectively than the pure collagen membrane, and 30% silica xerogel was confirmed as the optimum concentration for hybridization of collagen for effective guided bone regeneration .
To investigate the osteoblast cell behavior on the BONITmatrix®-collagen scaffold, primary human osteoblasts (HOSs) and human osteosarcoma cell line MG63 were seeded to human blood serum-coated porous scaffolds by Ghanaati et al.  according to a previous coating method based on human serum . Scanning electron microscopy (SEM) micrographs in Figure 1  show that both HOS and MG63 cells were adhered and spread individually on the porous scaffolds after 1 day of seeding, and the surface of the scaffolds was completely covered at 10 days of seeding. In reverse transcription-polymerase chain reaction analysis of osteoblastic gene expression on serum-coated scaffolds, regular expression of all osteoblast-specified genes (ALP, collagen-I, osteocalcin, osteonectin, transforming growth factor β1, fibroblast growth factor 2 and β-actin) was exhibited by HOS; besides, MG63 showed similar gene expression except for showing no visible transcription of ALP and osteopontin.
Ghanaati et al.  implanted BONITmatrix®-collagen scaffold in the subscapular region of rat to evaluate the tissue integration within the connective tissue around the implants. Macrophages and fibroblasts containing thin cell layer was observed around each implanted granule on day 3 of implantation. The number of these cells was increased in the following days; in addition, cell-rich and well-vascularized connective tissue was found at day 15. It was revealed that connective tissue had penetrated through the implanted granules after between 15 and 30 days of implantation. Moreover, the number of vessels, total vessel area and the vascularization of implants were significantly enhanced with the period of implantation.
The guided bone regeneration ability of silica-collagen nanohybrid membranes has been also investigated by implanting them in the calvaria defects of rats . New bone was formed around the calvaria defects for all implants, but the regenerated bone area was larger in the nanohybrid membrane than in the pure collagen membrane. The volume fractions of newly generated bone for pure collagen and nanohybrid membranes were calculated as 18.9±4.9% and 34.3±3.6%, respectively, after 3 weeks of implantation. Moreover, the silica-collagen nanohybrid membranes showed higher rate of degradation, which led to rapid bone regeneration than that of pure collagen membrane. The bone healing process is associated with the degradation of the implant materials, and the degradation products must be obviously nontoxic and biocompatible [32, 46]. In addition, the products released (e.g., ions) during degradation may have a major effect to stimulate osteoblastic cells and to accelerate new bone formation [47–50].
2.2 Sol-gel/polycondensation-based materials
The sol-gel process is widely used to fabricate scaffolds for bone regeneration . Silica-collagen nanohybrids have been observed to act as skeletal structure in nature (e.g., marine basal spicules), which were successfully reproduced in vitro by applying sol-gel technique . In fact, sol-gel is the most used technique to fabricate silica-collagen hydrogel and xerogel composites. Different types of silica precursors have been used to produce these types of composites because of the different effects that different precursors have on the self-assembling characteristics of the collagen structure . The most popular silica precursors are TEOS/tetraethylorthosilicate [18, 21, 24, 28, 53], tetramethoxysilane (TMS)/tetramethylorthosilicate [17, 54], potassium silicon tris-catecholate  and sodium silicate [23, 55]. The chemical structures of these silica precursors are presented in Table 1. Silica precursors serve as silica sources that are autopolycondensed into colloidal silica particles. This process facilitates the formation of three-dimensional (3D) gel networks with the interaction of positively charged amine groups of collagen, thus forming silicified collagen matrices . This process is called sol-gel polymerization. Two types of silica-collagen composites, namely, hydrogel and xerogel, have been produced with this technique.
2.2.1 TEOS/tetraethylorthosilicate as silica precursor
TEOS has been mostly used as silica precursor for fabrication of hydrogel and xerogel collagen-based hybrid composite. Silica sols have been prepared in various ways; for example, Brasack et al.  mixed TEOS with 1,4-dioxane, catalyzed with 0.01 m HCl. In another method, 0.25% ammonia instead of 0.01 m HCl was used as catalyst . The most applied method is based on TEOS hydrolyzed with deionized water (TEOS/water molar ratio=1:4) combined with 0.01 m HCl as a catalyst at 4°C for 24 h to produce orthosilicic acid that serves as silica source [13, 18, 24, 28, 56]. Then the orthosilicic acid is added to the homogeneous suspension of collagen solution at pH 7.4 followed by mixing homogeneously. Homogeneous solution of fibrillar bovine collagen can be prepared by dialyzing tropocollagen followed by fibrillation, lyophilization and resuspending in Tris-HCl, according to previously described methods . After proper mixing and stabilization for 3 days, the hydrogels are dehydrated by successive soaking in 30%, 50%, 70%, 80%, 90% and 100% ethanol. Then the hydrogels are gently dried in a climate chamber at 37°C and 95% relative humidity to obtain the monolithic xerogels from the hydrogel [18, 24, 28]. This process mimics the biosilicification process . Gelation time is the key parameter for controlling the interaction between collagen and polymerizing silica in the hybrid composite, which has been investigated by Heinemann et al. . They found that gelation time of the hydrogel increased with increasing volume of silicic acid and with decreasing concentration of collagen solution. This phenomenon was explained considering that the presence of collagen fibrils contributes interparticle bridging sites so that less amount of silica particles is required to achieve percolation of the composite network , which is schematically presented in Figure 2. Moreover, the gel formation kinetics may be affected by electrostatic interactions between the positively charged amine groups of collagen and negatively charged silica species [13, 57]. It has also been reported that silica species have the potential to influence collagen fibrillogenesis, as investigated by Eglin et al. [58, 59]. Silica in the form of negatively charged particles has a strong interionic affinity to positively charged groups, such as amine groups of collagen, as described elsewhere . This is likely the reason for the influence of gelling time with different concentrations of silicic acid and collagen solution.
The effect of incorporation of collagen in the structure of silica xerogel has been analyzed in detail by Heinemann et al. [13, 18, 28]. The dispersed fibrillar collagen was oriented randomly and discontinuously in the silica matrix phase in 85:15 and 70:30 silica-collagen xerogels, and the deformation of fractured surfaces of xerogels increased with increasing percentage of collagen due to pullout of silicified collagen fibrils from the bulk material. Heinemann et al.  observed by SEM that seeded human mesenchymal stem cells (hMSCs) were well spread and seemed to be well adhered to the xerogel surface, whereas the differentiated osteoblast-like cells covered the surface of the xerogel by forming a dense layer after 14 days of cultivation.
The relative amount of collagen has significant impact on the mechanical properties of silica-collagen composite xerogels, as discussed in a previous study . The Young’s modulus of pure silica xerogel (4.2 GPa) slightly increased due to incorporation of 10 wt% collagen but then significantly decreased to about 2 GPa for 60:40 silica-collagen xerogels. Moreover, the compressive strength and strain at fracture increased very rapidly with increasing collagen content in the xerogel. The 80:20 silica-collagen xerogel showed the highest compressive strength, which was double than that for pure silica xerogel (100 MPa). The highest strain at fracture was measured for 70:30 silica-collagen xerogel (∼11%), which was almost quadruple that of pure silica xerogel . In another study of Heinemann et al. , it was revealed that 30% collagen containing silica-based xerogel exhibited compressive strength of 115 MPa and 18% strain at ultimate strength.
Sol-gel-derived silica possesses HA-forming ability in vitro, as previously reported [60, 61]. HA-forming ability and the Ca/P ratios of the particles that were deposited on the surface of silica-collagen composite xerogels of different compositions have been studied by SEM and EDX, respectively, as shown in Figure 3 . The Ca/P ratio of HA deposited on pure silica xerogel was found to be 1.61 by EDX spectroscopy; the value is close to the ratio (Ca/P=1.67) for stoichiometric HA. Incorporation of 15% collagen into the silica-based xerogel had no significant effect on the HA-forming ability. However, HA-forming ability was found to decrease due to incorporation of 30% collagen . Brasack et al.  confirmed the bioactivity of silica-collagen composites by the deposition of mineral spheroids during the conditioning in SBF at pH 7.4. Biocompatibility of silica-collagen xerogel was studied by seeding hMSCs on the surface of the xerogel, as described by Heinemann et al. . Confocal laser microscopy images showed that cells were well spread on the surface of the hybrid xerogels after 24 h of seeding. Cell nuclei were also visible at the same time, and cells proliferated during the culture period, which was confirmed by observing higher cell density after 14 days of seeding. The quantitative determination of cell proliferation was carried out by determining the DNA content during hMSC cultivation on the hybrid xerogels. It was observed that the DNA content of osteogenically non-induced cell fraction increased significantly more than that of the induced fraction. The ALP activity of non-induced cells remained constant at a low level, but osteogenically induced cells seeded on the hybrid xerogels showed an increasing trend of ALP activity that reflected the degree of osteoblastic progression .
Alt et al.  implanted silica-collagen xerogel composite in the area of circumferential defect created in the distal metaphyseal femur of 12-week-old rats to investigate the angiogenesis potential of the implanted material. After 6 weeks of implantation, the defective femora were harvested and scanned for micro-computed tomography (CT) (9 μm)3 and nano-CT (3 μm)3 imaging. It was found that the microvessels were distributed within the defect zone, which was confirmed by histology and immunohistochemistry. The newly formed blood vessels with granulation tissue were formed around the bone-substituted material, which was detected by platelet endothelial cell adhesion molecule 1 (PECAM-1) immunopositive staining of the endothelial cells in immunohistological analysis, as shown in Figure 4 . Moreover, it was revealed that most of the vessels were found near the host bone but a few were found near the implant.
It was also shown that the vascular volume fraction (VVF) increased with decreasing volume of bone-substituted material (BSM), but the volume of the fracture zone had no significant effect on VVF . The mean value of VVF was calculated to be 3.01±0.4%. The newly formed vessels were visualized with nano-CT imaging and analyzed up to 10-μm voxel size using gray scale. From this analysis it was confirmed that silica-collagen xerogel composites possessed angiogenic potential.
Heinemann et al.  used silica-collagen xerogels for the investigation of the influence of embedded HA-like CPPs on physiological as well as biological properties. Silica-collagen xerogels were developed by sol-gel technique. In this study, silicic acid served as the silica source in the xerogels, which was prepared by hydrolysis of TEOS under acidic condition. Hydrogel of silica and collagen was made by mixing silicic acid and collagen suspension in a ratio of 70 wt% and 30 wt%, respectively, and kept for 3 days to stabilize the gel. Xerogels were obtained by drying the hydrogels in a 95% relative humidity chamber at 37°C and cut into disc-like shapes of 5 mm in diameter and 3 mm in height. To enhance the bioactivity and to evaluate the effect on osteoblasts and osteoclasts, calcium phosphate cement (CPC) powder or HA powder was added to the silica-collagen hydrogels. The bioactivity of silica-collagen (biphasic) xerogels was found to be much lower than that of silica-collagen-CPC/HA (triphasic) xerogels, which was measured by determination of calcium concentration in the SBF or in cell culture medium, used as the incubation medium of the xerogels. It was observed that the bioactivity of the xerogels positively affected the proliferation and differentiation behavior of osteoblasts and osteoclasts. It was found that the expression levels of all osteoblast-related markers were higher for the silica-collagen xerogels and decreased with increasing CPC or HA concentration in the triphasic xerogels. Moreover, a comparatively higher density of small multinuclear osteoclasts was observed to be evenly distributed in the monocultivation of human monocytes (hMcs) on triphasic xerogels in comparison to that on biphasic xerogels. Thus, highly bioactive materials favored osteoclasts over the osteoblasts, whereas low bioactive materials favored osteoblasts over osteoclasts, when both cell lineages were combined in co-culture . The high level of bioactivity of the triphasic xerogels reduced the calcium concentration in the medium, which inhibited the proliferation and differentiation of osteoblasts. Therefore, the bioactivity of silica-collagen xerogels can be controlled to make the optimum balance of osteoblasts to osteoclasts, which is essential for healthy bone formation. Moreover, it was revealed by Heinemann et al.  that the differentiation of monocytes to osteoclasts was higher for the biphasic xerogels compared to triphasic xerogels, which was evaluated by the detection of tartrate resistant acid phosphatase (TRAP) 5b activity in the monoculture as well as in the co-culture with hBMSCs/HOS.
Desimone et al.  used 12- and 80-nm silica nanoparticles (labeled as Si12 and Si80, respectively) for fabrication of silica-collagen bionanocomposites. The 80-nm silica nanoparticles were synthesized by hydrolyzation and polymerization of tetraethylorthosilicate with ammonia as base catalysis medium, according to the Stöber method . Silica nanoparticle suspensions were acidified with acetic acid to maintain the pH at 3.0 followed by adding it in different amounts (final concentration, 5 and 10 mm) to the mixture of 0.6 ml of collagen solution (2.8 mg/ml in 17 mm acetic acid) and 0.8 ml of complete culture medium in an ice bath. Then the resulting solution was neutralized with 0.8 ml of 0.1 m NaOH. Subsequently, 0.6 ml of PBS-supplemented medium containing human dermal fibroblast (Promocell) suspension was added. The cell density was maintained at 1×105. Reference collagen hydrogels were fabricated according to the same protocol except that no silica nanoparticles were added. Within the collagen hydrogels, the entrapped fibroblasts pulled the collagen fibrils, resulting in increase of the density and contraction of the hydrogel and leading to enhanced stiffness. However, the contraction process was delayed by incorporation of silica nanoparticles. Addition of 10 mm Si80 showed only 24% of surface contraction at day 1, whereas pure collagen showed 43% of surface contraction at the same time.
Highly porous structures of silica-collagen nanocomposites were observed by SEM and TEM, which also indicated that the silica nanoparticles did not modify the self-assembly of collagen molecules, as shown in Figure 5 . Si12 nanoparticles were observed to be surrounding the collagen molecules. However, Si80 nanoparticles were systematically associated with collagen fibrils within the hydrogel network. Inductively coupled plasma atomic emission spectroscopy analysis showed that the release behavior of silica for Si80-collagen gel was slightly faster during the first 14 days, but all samples showed about the same result at 21 days. Si12-collagen nanocomposites showed the highest storage and loss modulus.
The viability of fibroblasts within the hydrogels was analyzed by MTT assay as shown in Figure 6 . The result showed that the largest silica particles (Si80) exhibited the highest metabolic activity of entrapped fibroblasts in comparison to the silica particles with smaller size (Si12). Besides, all silica-collagen nanocomposites showed much lower matrix metalloproteinase 2 activity than the control collagen gel. Cells were spread, proliferated and homogeneously distributed with a spindle shape in the core region of all silica nanoparticle (Si12, Si80)-containing collagen nanocomposites. Moreover, optical microscopic images confirmed the presence of an interconnected pore network that should play an important role in the transfer of nutrients and waste products, thus enhancing the bioactivity of the nanocomposites.
Tetraethylorthosilicate has been used by Niu et al.  as silica precursor for intrafibrillar silicification of collagen scaffolds. Hydrolyzed tetraethylorthosilicate (40%) was mixed with absolute ethanol, water and 37% HCl in the molar ratios 1.875:396.79:12.03:0.0218 for 1 h to prepare 3% silicic acid solution. Then the 3% silicic acid solution was mixed with 72 mm choline chloride in a volume ratio of 1:1 to obtain 1.5% choline-stabilized silicic acid solution, which was used as silicifying medium. Silicified collagen scaffolds (SCSs) were produced by dipping the rehydrated collagen scaffolds in silicifying medium for 4 days, with daily change of the medium. The rehydrated collagen scaffolds were produced by treating the dehydrated collagen scaffolds with 6.67×10-4m poly(allylamine) hydrochloride for 4 h. SCSs exhibited significantly higher biomechanical properties in comparison to untreated collagen scaffolds. The tangent moduli and the resilience moduli of the SCSs were found to be significantly higher than that of collagen scaffolds. The biomechanical properties of SCSs were decreased in PBS and in PBS+collagenase over the incubation time investigated, however, which was significantly better than in the case of collagen scaffolds, thereby providing improved mechanical stability to the scaffolds during the cell homing. Extrafibrillar apatite deposition was observed along the surface of collagen leaflets after 2 days by TEM and confirmed by X-ray diffraction (XRD) analysis. The deposition of calcium and phosphorus along the surface of collagen leaflets, analyzed by scanning transmission electron microscopy EDX (STEM-EDX) analysis confirmed that SCSs exhibited osteoconductivity in SBF, which should promote cell attachment to the internal region of the scaffolds [64, 65]. It was found that SCSs and collagen scaffolds exhibited similar mitochondrial succinic dehydrogenase activities of MSCs and endothelial progenitor cells (EPCs) as the Teflon negative control. Moreover, there was no significant difference observed of early and late apoptotic cells of MSCs and EPCs among collagen scaffolds, SCSs and the Teflon negative control. The results indicated that intrafibrillar silicification of collagen scaffolds had no adverse effect on the viability of MSCs and EPCs. Nevertheless, differentiated MSCs seeded on SCSs exhibited highly significant upregulation in mRNA expression of genes associated with osteogenesis than that of MSCs seeded on collagen scaffolds. Moreover, the amount of deposited calcium by SCSs was found to be significantly higher than that of collagen scaffolds, which indicated that SCSs had higher osteogenic potential. In vitro proangiogenic properties have been investigated by culturing EPCs on extracellular matrix for 6 h to observe the development of capillary-like tubes and sprouting new capillaries. The number of capillary tubes and branched points induced by SCS extracts were found to be significantly higher than those of Teflon and collagen scaffolds, which indicated that SCSs exhibited better angiogenic potential in EPCs than collagen scaffolds. Moreover, the release of silicic acid from SCSs induces apatite deposition that promotes the osteogenic potential, as well as leads to improvement in angiogenic potential over the collagen scaffolds.
2.2.2 TMS/tetramethylorthosilicate as silica precursor
Ehrlich et al.  and Heinemann et al.  studied in detail the structural and biochemical characterization of the basal spicules of Monorhaphis chuni, a marine glass sponge, and revealed that a collagenous fibrillar protein acted as a template for mineralization of silica in all silica-containing structural layers of the spicule. SEM micrographs of M. chuni spicule treated with alkali solution confirmed that it was constructed with a multilayer structure and all layers were connected among each other by nanostructured protein fibers. Moreover, mineralized silica was distributed on the surface of nanofibers in the form of nanopearl necklaces. Ehrlich et al.  and Heinemann et al.  fabricated silica-collagen hybrid scaffold by a method that mimics the biosilicification process. In this approach, TMS was used as silica precursor, which was hydrolyzed with water as well as 0.01 m HCl as a catalyst at 4°C for 24 h to form orthosilicic acid. The hybridization of silica and collagen was performed by intensive mixing of homogeneous collagen suspensions and prehydrolyzed TMS (orthosilicic acid) in ambient condition. The suspension was then poured into the cavities of polystyrene well plates. Heinemann et al.  found that stable hybrid hydrogels had been formed by adding higher concentration (0.025–1.000 m) of prehydrolyzed TMS. Silicified porous collagen scaffolds were obtained by freeze-drying, and compact xerogels were obtained after air-drying for 7 days at 20°C in a 30% relative humidity atmospheric chamber .
It has been found that silicification of collagen sponge in vitro occurred by self-assembling and nonenzymatic mechanisms [17, 66]. Silica-collagen biomaterials of rod-like shape several millimeters in diameter demonstrated morphological similarity to M. chuni basal spicules as observed by SEM. By analyzing the morphological structure of self-assembled silica-collagen composites, it was observed that amorphous silica nanoparticles were deposited from silicic acid on the surface of the collagen fibrils in the form of a nanopearl necklace, replicating the nanoparticulate structural morphology of M. chuni basal spicules. To analyze the biocompatibility of the silica-collagen composite, hMSCs were cultivated on the surface of the composites, and it was observed that the cells were well adhered and proliferated after 14 days of cultivation.
Eglin et al.  developed a silicified collagen composite by vapor diffusion of volatile alkoxysilane (TMS), which was hydrolyzed at the hydrated collagen surface and then diffused and condensed in the template according to Carturan et al. . In this process, a vial containing 1 ml of collagen solution and an open glass vial containing 2 ml of TMS were kept in a tightly closed 50-ml glass vessel at 20°C for 6 days. During this experiment, TMS vapor diffused to collagen and hydrolyzed at the hydrated surface of collagen and condensed. TEM micrographs and polarized light microscopic analysis confirmed the cholesteric morphology of the composite; no aggregation of collagen fibrils was observed. The silica-collagen composite was calcined at 120°C for 2 days and at 600°C for 20 h to obtain a porous scaffold.
2.2.3 Sodium silicate as silica precursor
Sodium silicate has been applied as silica source to produce silica-containing collagen-based hydrogel composites by simultaneous polymerization of aqueous silicates and self-assembly of triple helices of collagen in the presence of human dermal fibroblasts [55, 68]. Cellularized hydrated collagen was obtained by neutralization of diluted acid-soluble collagen solution according to the method of Bell et al. . In culture medium containing polystyrene dishes, cold diluted collagen solution (1.2 mg/ml) was mixed with acidified diluted sodium silicate solution (concentration ≤5 mm) of pH 3. Human dermal fibroblasts in culture medium were added to the solution after neutralization with aqueous NaOH. Hydrogel discs of 3.4 cm in diameter and 2 mm in thickness were obtained. This process was developed by Desimone et al. [55, 68].
It was revealed that the mechanical properties were enhanced due to silicification of the collagen matrix, which was confirmed by the increase in both storage and loss moduli for silicified collagen hydrogel in comparison to pure collagen hydrogel. This observation can be explained according to the previous reports [18, 53, 70] on biopolymer-silica hybrid materials considering that silica precursors penetrated into the structure of the biopolymer and developed an interconnected network that improved mechanical properties because of the stiffness of silica and the interaction with the biomineral surface. The thermal stability of the composite was observed to increase due to the silicification of the collagen matrix. The collagen denaturation temperature was measured to be 65°C for silicified collagen hydrogel, which was 10°C higher than that for pure collagen hydrogel. This result was ascribed to the ability of silicates to cover collagen fibrils that restricted the unfolding possibility of the protein, leading to the increase of the protein denaturation temperature, as described in a previous study .
In the study of Desimone et al. , the highest number of fibroblast adhesions was observed on the surface of the silicified collagen hydrogel compared to pure collagen hydrogel after 6 and 24 h of seeding cell in vitro. Moreover, it was observed that the silica-collagen hydrogel provides a better surface for proliferation of fibroblasts than that of pure collagen hydrogel itself. In another study of Desimone et al. , it was confirmed that fibroblasts could survive in the silica-collagen hybrid hydrogel and the highest survival rate was obtained in the 1 mm silicate-containing collagen-based hydrogel, which was analyzed by methyl thiazol tetrazolium (MTT) assay for determination of the number of metabolically active fibroblasts during 2 weeks of seeding. Moreover, Desimone et al.  investigated with MTT assay the relative metabolic activity of entrapped fibroblasts in both collagen hydrogels containing silica nanoparticles and collagen hydrogels containing sodium silicate. They observed that silica nanoparticle-collagen hydrogels induced the highest metabolic activity of fibroblasts. This result can be explained by the silica release rate, which was observed to be faster for sodium silicate-collagen hydrogel, indicating the possible toxicity of released Si species, which could be a reason for the observed low metabolic activity, in agreement with other reports . The morphology of the hydrogels produced by Desimone et al.  was observed by SEM micrographs. The pure collagen hydrogel consisted of a porous structure that became a more entangled network with overcrossing areas due to low silicification of collagen. For highly silicified collagen hydrogel, the morphology consisted of rope-like twisted bundles of collagen fibrils with average diameter of 400 and 700 nm at day 1 and day 14 of incubation, respectively. The data suggested that the silica colloids had the ability to coat the individual collagen fibrils and they favored the aggregation of collagen fibrils, leading to larger fibers [21, 55, 68].
The catabolic activity of entrapped fibroblasts depends on the production of MMP-2 enzyme that indicates the collagen degradation ability of entrapped cells, which was not significantly modified at low silica content but increased progressively for the highest silica content. This result indicated that large silica-collagen fibril bundles, which formed from higher silica concentration in collagen hydrogel, might not favor fibroblast adhesion.
2.2.4 Potassium silicon tris-catecholate as silica precursor
To investigate the influence of different silica precursors on collagen self-assembly, Eglin et al.  used potassium silicon tris-catecholate as silica precursor. In this study, type I collagen solution (5 mg/ml) in 0.005 m acetic acid was mixed with 4.2 ml of cooled deionized water followed by sonication at 5°C for 20 min. The stock solution of 0.05 or 0.2 m potassium silicon tris-catecholate was prepared in deionized water and then diluted with chilled 0.2 m tris buffer solution to make the final solutions of silicon concentrations ranging from 0.0002 to 0.02 m. Then the stock solution of silica precursor was mixed with collagen solution in equivalent volume ratio.
Silicon catecholate dissociated to form the complex ions orthosilicic acid and catechol in low concentration silicone catecholate containing collagen hydrogel. These complex ions modified the water structure in the vicinity of triple helices of collagen to promote the aggregation of fibrils. The silicomolybdate blue assay has been used to monitor silicic acid concentration. The molybdate active silicon concentration was detected due to breakdown of silicon catecholate complex and the condensation of orthosilicic acid in buffer solution of pH 7.4. The lower amount of silicon catecholate containing collagen hydrogel exhibited the concentration of molybdate active silicon, which was much closer to that of pure silicon catecholate itself. However, the hydrogels containing 2.5 and 5 mm silicon catecholate showed significantly lower concentration of molybdate active silicon over the first hour, indicating the breakdown of the complex followed by condensation of silicon species.
2.3 Immersion process
Silica-collagen hybrids have also been fabricated by immersing silica spheres in collagen solution to allow adsorption of collagen on the surface of silica spheres  or by immersing collagen sponges in silicic acid solution to facilitate infiltration of silica inside the fibrillar collagen . In the study of Ye et al. , silica spheres (of grain size 450–500 nm) were prepared by the Stöber method . The supernatant was removed by centrifugation and subsequently washed with pH 5.0 buffer solution. Collagen solutions with different concentrations at pH 5.0 and 0.1 m sodium acetate and 0.1 m KBr as electrolytes were added into the suspension containing silica spheres followed by stirring for 30 min to allow adsorption of collagen onto the silica spheres. Then the collagen-coated silica spheres were separated from the supernatant by centrifuging the mixture at 4000 rpm for 10 min. The resulting particles were redispersed in pH 5.0 buffer and again centrifugated and the supernatant was discarded. The particles were finally washed with ultrapure water followed by drying in vacuum at room temperature. The adsorption behavior of collagen onto silica spheres and the structural and thermal properties of the composite were analyzed. It was confirmed that collagen macromolecules were self-aggregated during adsorption of collagen onto silica spheres. The composites exhibited lower infrared emissivity values that indicated a strong interfacial interaction between the surface of silica spheres and collagen. FTIR spectra showed amide group stretching vibrations of collagen, indicating the formation of hydrogen bonds at the interface between collagen and silica spheres. Interfacial interactions and hydrogen bonding played a key role in adsorption of collagen on the surface of silica spheres.
Niu et al.  developed silicified collagen hybrids by immersing polyallylamine-enriched collagen sponges in the solution of choline-stabilized silicic acid to facilitate infiltration of silicic acid inside collagen sponges. SEM micrographs showed that silicified collagen sponges remained highly porous, and STEM-EDX confirmed the presence of silicon within the collagen fibrils. The mechanical properties of silicified collagen sponges were found to be significantly higher than those of nonsilicified collagen sponges. For example, the tangent modulus and work of fracture (energy absorbed per unit volume to rupture) of the silicified collagen were 48,000 and 1500 times higher than the values for nonsilicified collagen, respectively. These data favor the use of silicified collagen sponges as porous scaffolds for bone repair in the regions with minimal to moderate load-bearing requirements. Biosilicification also confirmed that polyallylamine-enriched collagen served as a template for polymerization of the silica precursor phase .
2.4 Other processing techniques
Eglin et al.  fabricated a silica-collagen hybrid hydrogel composite by a gelling process, which was done by allowing ammonia solution to diffuse into a mixture of sol-gel-derived silica particles and collagen. Initially, silica particles were prepared by hydrolysis-condensation reactions of TEOS in ethanol/water acidic solution, which was previously described by Brinker and Scherer . Collagen solution was added to aluminum hydroxide and silica particles (silica/collagen weight ratio=10:1) in a plastic vial that was previously punctured with a needle at the middle portion, followed by mixing homogeneously. Then the tightly closed plastic vial was introduced into a glass vial containing 30% ammonia solution to allow diffusion of ammonia into the silica-collagen mixture through the pinhole of the plastic vial. The resulting hydrogel was taken out after 6 h and soaked in deionized water.
SEM micrographs showed that a significant amount of a crystalline phase, forming grape-like aggregates 2–5 μm in diameter, was homogeneously deposited on the silica-collagen surface after 14 days of immersion in SBF. However, very few crystalline particles were deposited on the silica surface alone after the same period of immersion in SBF. The d-spacing and relative intensities of diffraction peaks of deposited materials on the surface of silica-collagen hydrogel could not be easily ascribed in the XRD patterns, which are shown in Figure 7 .
After 3 days of soaking of silica-collagen hydrogels in SBF, the XRD spectrum exhibits only one diffraction peak at 2θ=32.0°, corresponding to calcium silicate hydroxide or highly silica-substituted HA with calcium to phosphorus molar ratio equal to 1.09. This ratio is different from Ca/P=1.67, which is typical of HA. New signals were detected at 2θ=11.2°, 15.7° and 27.0°, which could suggest formation of calcium phosphate, α-TCP and/or orthocalcium phosphate (OCP) after 7 days, and more distinctly after 14 days of soaking in SBF (Figure 7C).
The released concentration of silicic acid at different soaking periods in SBF was measured using the blue silicomolybdate method to investigate the effect of collagen on dissolution of silica. It was observed that the concentration of released silicic acid reached a plateau at around 4×10-5 mol/l in less than 3 days. In pure silica, calcium and phosphate ions are not present. However, collagen can compensate for the loss of HA precursors in the hydrogel composite. The acidic groups (aspartic acid, glutamic acid, etc.) in collagen that contain carboxylate functional groups can make bonds with Ca2+. Further deposition of calcium phosphate can be facilitated by providing nucleation sites by the calcium silicate phase, which forms by association of silicic acid with calcium ions.
Luo et al.  used MSNs that were end-capped with collagen to fabricate redox-responsive nanoreservoirs for targeted drug delivery. The unique mesoporous structure of MSNs that acts as ideal stimuli-responsive carriers for controlled drug and gene delivery was exploited . In this process, collagen was used as a cap to encapsulate fluorescent probes within the porous channels of the MSNs, where collagen was immobilized on the upper surface of MSNs by making disulfide bonds that could be cleaved with various reducing agents. The galactose group containing lactobionic acid (LA), was incorporated as the targeting moiety. Previously, cell-specific gene  and hormone  transfections were successfully achieved from different substrates via LA molecules. The endocytosis and intreacellular delivery of nanoparticles by HepaG2 and endothelial cells were analyzed and it was observed that the number of HepaG2 cells in the LA-MSNs-collagen was 2 and 2.2 times higher than that of endothelial cells at 2 and 4 h, respectively. From this study, it was confirmed that collagen-capped MSNs could serve as a redox-responsive nanoreservoir for targeted drug delivery, and it showed biocompatibility, cellular uptake properties and the possibility to be used for treatment of liver cancer. Extension of this approach to consider relevant drugs for bone regeneration and to combat bone infections could be also considered.
3 Discussion and areas of application
Different types of biocomposites, e.g., hydrogels, xerogels, hybrid porous scaffolds and nanocomposites, have been derived from silica and collagen, generally for application in bone regeneration, blood vessel formation (vascularization), wound healing and drug delivery. Silica-collagen combinations are attractive biomaterials due to their potential for osteogenesis, angiogenesis, wound healing, and drug delivery ability. They constitute therefore an important group of materials with favorable characteristics for bone tissue engineering.
Osteogenesis describes the formation of new bone by the action of osteoblasts or osteoblast-like cells. Biomaterials used as bone implants or scaffolds should exhibit several properties like biocompatibility, biodegradability, osteoconductivity and osteoinductivity and should possess appropriate mechanical stability . Silica-collagen nanocomposites fulfill these requirements, making them valuable innovative materials in bone tissue engineering . As discussed before, it has been revealed that silica-collagen [14, 18] composite materials had the nucleation potential for deposition of biomimetic hydroxyl carbonate apatite. Silica releases silicic acid upon material dissolution during conditioning in SBF, and the silanol groups can induce apatite formation by taking calcium and phosphate ions [79–81]. This type of release (calcium, sodium and phosphates ions) is absent in the case of pure silica. In silica-collagen composite materials, this lack of calcium and phosphorus ions is counterbalanced by collagen itself . Moreover, collagen serves as a proteineous material that has acidic groups, such as aspartic acid or glutamic acid, which possess carboxylic groups that can easily bind to calcium ions, thus serving as nucleation sites for further deposition of calcium phosphate without altering the triple helical structure of collagen [11, 14].
Osteogenesis strongly depends on the response of osteoblasts or osteoblast-like cells on the composites. Cell adhesion and proliferation are critical cellular processes that depend on both the characteristics of the biomaterials and the expression of cell surface molecules [82, 83]. Initial cell adhesion can be affected by pore size distribution and specific surface area of the material that can affect the subsequent steps such as proliferation, migration and differentiation of cells [19, 84, 85]. It was also revealed that collagen was secreted by osteoblasts, which is relevant for bone healing . The adhesion, proliferation and differentiation of hMSCs into osteoblast-like cells on the surface of silica-collagen xerogel have also been investigated . In another study, human monocytes were directly seeded on the surface of silica-collagen composite xerogel to investigate their potential to differentiate to osteoclast-like cells [18, 52]. Moreover, bioactivity of the implant materials is an important factor for making the balance between the bone-building (osteoblasts) and bone-resorbing (osteoclasts) cells involved in the bone remodeling process, which can be controlled by adhesion of varying amounts of CPC or HA [52, 87, 88]. It has been observed that low levels of bioactivity of biomaterials enhanced the proliferation and differentiation of osteoblasts, and on the contrary, hMc was differentiated to smaller multinuclear osteoclasts on the highly bioactive materials . Moreover, it has been demonstrated that the proliferation of hBMSC, ALP activity and gene expression of osteogenic markers were reduced with increasing CPC concentration in silica-collagen xerogels . The same results have been observed for MC3T3-E1 osteoblasts; especially, the ALP activity decreased with increasing calcium concentration in the silica xerogels . Reduced extracellular calcium concentration was seen to cause significant metabolic stress for primary osteoblasts and mesenchymal stem cells, resulting in reduced proliferation and differentiation [90, 91]. Lower concentration of calcium in the highly bioactive materials incubated culture medium (due to the deposition of calcium phosphate on the surface of highly bioactive materials) could reduce the extracellular calcium concentration. It is known that biomaterials can influence the gap junction communication in osteoblasts by binding or releasing calcium ions, which control the differentiation of osteoblasts . The correlation of osteoblasts and osteoclasts has been evaluated by confocal laser scanning microscopy, showing a high number of multinuclear osteoclasts embedded between a few osteoblast “pylons” on highly bioactive xerogels and a low number of multinuclear osteoclasts embedded in a close “labyrinth” of osteoblasts on comparatively low bioactive xerogels . Here, biphasic silica-collagen xerogels exhibited comparatively low bioactivity, which encouraged proliferation of osteoblasts in comparison to osteoclasts formation. However, the bioactivity of silica-collagen composites could be increased by incorporation of CPP, which favored osteoclast formation but diminished osteoblast generation. It has been observed that triphasic bioactive silica-collagen xerogels with 5% calcium phosphate is an appropriate scaffolding system for controlling a balanced ratio of both cell types, osteoblasts and osteoclasts, which is very important for regeneration of healthy bone .
However, silicified collagen containing the highest concentration of silica exhibited poorer cell intrusion because of lower pore size distribution that hindered the interconnectivity of the composite scaffold . In vivo bone regeneration has been confirmed for silica-collagen biomaterials [22, 30]. The bone regeneration rate was seen to be strongly dependent on the degradation rate of the hybrid materials during in vivo implantation [22, 30]. Degradation of the implanted hybrid material is executed mainly by mononuclear cells, whereas only a few multinucleated giant cells are involved . Scaffolds for bone tissue engineering should possess suitable mechanical properties . In the case of silica-collagen composite xerogel, the concentration of collagen has a significant effect on the mechanical properties [17, 18]. This result can be explained by the tensile strength of collagen  that can enhance the corresponding mechanical characteristic of composite xerogels by embedding the silica phase. Moreover, the brittle characteristic of pure silica xerogel is reduced by incorporation of collagen . In addition, the density of silica-collagen composites decreases with increasing collagen content, which has a significant impact on the compressive strength of the composite [17, 18].
Collagen-silica composites have been shown to possess angiogenic potential, which was assessed by investigating the formation of new blood vessels surrounding the bone-substituting implant and confirmed by immunohistochemistry . It has also been shown that SCSs possess in vitro proangiogenic properties, which was investigated by culturing EPCs, followed by observing the development of capillary-like tubes, sprouting of new capillaries and the formation of cellular networks. Moreover, silicified collagen exhibits better angiogenic potential in EPCs than non-SCSs, confirming a previous hypothesis that angiogenesis is enhanced by silicification .
3.3 Therapeutic drug delivery
In recent years, organic-inorganic hybrid materials have been developed as suitable matrices for therapeutic approaches because porous inorganic materials have the potential to host organic molecules, such as drugs, and subsequently they can act as sustained release systems [26, 95]. Among the family of bioceramic materials, mesoporous silica-based materials are drawing enormous interest to be used as sustained release materials because of their ability to host different guest molecules [96–98]. A host-guest interaction takes place between the functional groups of the drug and the silanol groups covering the surface of mesoporous channels of mesoporous silica-based biomaterials used in controlled drug delivery systems [26, 96, 99]. This host-guest interaction between the inorganic matrix and organic guest materials, e.g., drugs, is based on the intrinsic morphological properties (mesoporous structure, large surface area, tunable pore size) and its chemical characteristics [27, 98]. Moreover, silica-based bioceramics exhibit biocompatibility characteristics that are essential for applications as local drug delivery systems in bone tissue engineering .
Surface-functionalized, end-capped MSNs have been used in different studies [100–102] as ideal stimuli-responsive carriers for controlled drug and/or gene delivery because of their unique mesoporous structure. The nanoreservoirs based on MSNs, which are end-capped with collagen, possess strong potential of redox-responsive controlled drug release and also cell-specific targeting, as reported by Luo et al. . It has been observed that MSNs end-capped with collagen through disulfide linkage exhibited around 12% leaching of drug within 14 h, whereas MSNs/collagen materials (collagen was physically absorbed on MSNs) released around 65% of the drug during the same time [27, 76]. The redox-response character and controlled-release behavior of the nanoreservoirs of MSNs end-capped with collagen have been analyzed by using dithiothreitol (DTT) . DTT is generally used as an external stimulus to trigger the redox-responsive release behavior of the nanoresorviors. It has been observed that 80% of the drug was released from the nanoreservoirs within 2 h after addition of DTT, thus suggesting a good response to DTT. This result can be explained as breakage of the disulfide linkages between collagen and MSN due to addition of DTT, which facilitates the drug-releasing behavior .
4 Conclusion and outlook
Organic-inorganic composites and hybrids can be considered “third-generation” bioactive materials that are generally used as appropriate scaffolding systems to support and encourage proliferation and differentiation of bone cells as well as to promote new bone formation. The reviewed literature shows that using silica/collagen composites (or hybrid biomaterials) represents a successful approach to developing suitable bone tissue engineering scaffolds because of their attractive osteogenic, angiogenic and controlled drug delivery potential. Studies demonstrated that collagen plays a significant role for the deposition of HA layer on the surface of the composites, leading to a suitable mineralization potential of the materials. Denaturation of collagen could be possible during immersion of silica/collagen composites in SBF; however, this effect has not been studied yet. The evaluation of this possible effect on the in vitro degradation behavior of collagen-silica composites could be an interesting topic for further research. This knowledge could be helpful to fabricating appropriate silica-collagen composites with tuned degradability for application in bone tissue engineering. For the fabrication of silica/collagen composites, different precursors have been used as silica sources that mimic, for example, the biosilicification process of marine basal spicules. It has been observed that collagen-silica interactions are strongly dependent on the silica source. Mechanical properties and thermal stability of silica/collagen composites and hybrids were found to increase due to silicification of the collagen matrix. Silicification leads to an interpenetrated network of bio-organic (collagen) and mineral phases (silica). Silica/collagen composites and hybrids have shown promising potential for bone regeneration and connective tissue vessel formation in vivo. Further research should concentrate on the specific effect of the degradation behavior of the materials (silica/collagen composites based on different silica precursors) and on angiogenesis, e.g., to assess neovascularization in vivo. Moreover, collagen end-capped mesoporous silica nanoreservoirs have shown great potential in controlled therapeutic drug delivery. This type of nanoreservoirs is a promising system for designing bioactive and multifunctional biomaterials for bone regeneration.
Within the application in bone tissue engineering, special emphasis must be given for developing osteoregenerative silica/collagen composites by incorporation of osteoinductive agents (hormones, peptides, growth factors) or through the control of their morphological structure as well as their chemical composition. The hierarchical porosity of the scaffold and incorporation of osteoinductive agents must be tailored to enhance cell adherence, proliferation, differentiation, bone tissue ingrowth and, finally, vascularization. In this context silica-collagen composites with appropriate 3D porosity and precise architecture must be developed to be applied in the filling of specific bone defects. Additive manufacturing fabrication methods are promising techniques that can design a fully interconnected 3D structure with predetermined dimensions and porosity . Moreover, controlling the bioactive properties of silica-collagen composites or hybrid biomaterials is essential in order to maintain a balanced ratio of osteoblast-to-osteoclast activities given the significant influence of this parameter in the bone remodeling process.
We would like to thank the Emerging Fields Initiative (EFI) of the University of Erlangen-Nuremberg (project TOPbiomat) and the German Academic Exchange Service (DAAD) for financial support.
 Hench LL. Bioactive materials: the potential for tissue regeneration. J. Biomed. Mater. Res. 1998, 41, 511–518.10.1002/(SICI)1097-4636(19980915)41:4<511::AID-JBM1>3.0.CO;2-FSearch in Google Scholar
 Gkioni K, Leeuwenburgh SC, Douglas TE, Mikos AG, Jansen JA. Mineralization of hydrogels for bone regeneration. Tissue Eng., Part B 2010, 16, 577–585.10.1089/ten.teb.2010.0462Search in Google Scholar
 Cao W, Hench LL. Bioactive materials. Ceram. Int. 1996, 22, 493–507.Search in Google Scholar
 Rea SM, Best SM, Bonfield W. Bioactivity of ceramic-polymer composites with varied composition and surface topography. J. Mater. Sci.: Mater. Med. 2004, 15, 997–1005.Search in Google Scholar
 Kamitakahara M, Ohtsuki C, Miyazaki T. Review paper: behavior of ceramic biomaterials derived from tricalcium phosphate in physiological condition. J. Biomater. Appl. 2008, 23, 197–212.Search in Google Scholar
 Rezwan K, Chen QZ, Blaker JJ, Boccaccini AR. Biodegradable and bioactive porous polymer/inorganic composite scaffolds for bone tissue engineering. Biomaterials 2006, 27, 3413–3431.10.1016/j.biomaterials.2006.01.039Search in Google Scholar
 Ara M, Watanabe M, Imai Y. Effect of blending calcium compounds on hydrolytic degradation of poly(dl-lactic acid-co-glycolic acid). Biomaterials. 2002, 23, 2479–2483.Search in Google Scholar
 Koo JH. Polymer nanocomposites: processing, characterization, and applications. McGraw-Hill: New York, USA, 2006.Search in Google Scholar
 Boccaccini AR, Erol M, Stark WJ, Mohn D, Hong Z, Mano JF. Polymer/bioactive glass nanocomposites for biomedical applications: a review. Compos. Sci. Technol. 2010, 70, 1764–1776.Search in Google Scholar
 Marelli B, Ghezzi CE, Barralet JE, Boccaccini AR, Nazhat SN. Three-dimensional mineralization of dense nanofibrillar collagen-bioglass hybrid scaffolds. Biomacromolecules 2010, 11, 1470–1479.10.1021/bm1001087Search in Google Scholar
 Yunos DM, Bretcanu O, Boccaccini AR. Polymer-bioceramic composites for tissue engineering scaffolds. J. Mater. Sci. 2008, 43, 4433–4442.Search in Google Scholar
 Heinemann S, Heinemann C, Ehrlich H, Meyer M, Baltzer H, Worch H, Hanke T. A novel biomimetic hybrid material made of silicified collagen: perspectives for bone replacement. Adv. Eng. Mater. 2007, 9, 1061–1068.Search in Google Scholar
 Eglin D, Maalheem S, Livage J, Coradin T. In vitro apatite forming ability of type I collagen hydrogels containing bioactive glass and silica sol-gel particles. J. Mater. Sci.: Mater. Med. 2006, 17, 161–167.Search in Google Scholar
 Yan S, Yin J, Cui L, Yang Y, Chen X. Apatite-forming ability of bioactive poly(l-lactic acid)/grafted silica nanocomposites in simulated body fluid. Colloids Surf., B 2011, 86, 218–224.10.1016/j.colsurfb.2011.04.004Search in Google Scholar
 Heinemann S, Ehrlich H, Knieb C, Hanke T. Biomimetically inspired hybrid materials based on silicified collagen. Int. J. Mater. Res. 2007, 98, 603–608.Search in Google Scholar
 Heinemann S, Heinemann C, Bernhardt R, Reinstorf A, Nies B, Meyer M, Worch H, Hanke T. Bioactive silica-collagen composite xerogels modified by calcium phosphate phases with adjustable mechanical properties for bone replacement. Acta Biomater. 2009, 5, 1979–1990.Search in Google Scholar
 Heinemann S, Heinemann C, Jäger M, Neunzehn J, Wiesmann HP, Hanke T. Effect of silica and hydroxyapatite mineralization on the mechanical properties and the biocompatibility of nanocomposite collagen scaffolds. Appl. Mater. Interfaces 2011, 3, 4323–4331.10.1021/am200993qSearch in Google Scholar
 Si X, Wang Y, Ren L, Zhao N, Gong Y, Wang D-A. Novel mesoporous silica-based antibiotic releasing scaffold for bone repair. Acta Biomater. 2009, 5, 1697–1707.Search in Google Scholar
 Desimone MF, Hélary C, Rietveld IB, Bataille I, Mosser G, Giraud-Guille M-M, Livage J, Coradin T. Silica-collagen bionanocomposites as three-dimensional scaffolds for fibroblast immobilization. Acta Biomater. 2010, 6, 3998–4004.Search in Google Scholar
 Ghanaati SM, Thimm BW, Unger RE, Orth C, Kohler T, Barbeck M, Müller R, Kirkpatrick CJ. Collagen-embedded hydroxylapatite-β-tricalcium phosphate-silicon dioxide bone substitute granules assist rapid vascularization and promote cell growth. Biomed. Mater. 2010, 5, 25004.Search in Google Scholar
 Eglin D, Shafran KL, Livage J, Coradin T, Perry CC. Comparative study of the influence of several silica precursors on collagen self-assembly and of collagen on ‘Si’ speciation and condensation. J. Mater. Chem. 2006, 16, 4220–4230.Search in Google Scholar
 Alt V, Kögelmaier DV, Lips KS, Witt V, Pacholke S, Heiss C, Kampschulte M, Heinemann S, Hanke T, Schnettler R, Langheinrich AC. Assessment of angiogenesis in osseointegration of a silica-collagen biomaterial using 3D-nano-CT. Acta Biomater. 2011, 7, 3773–3779.Search in Google Scholar
 Vallet-Regí M, Colilla M, González B. Medical applications of organic-inorganic hybrid materials within the field of silica-based bioceramics. Chem. Soc. Rev. 2011, 40, 596–607.Search in Google Scholar
 Luo Z, Cai K, Hu Y, Zhao L, Liu P, Duan L, Yang W. Mesoporous silica nanoparticles end-capped with collagen: redox-responsive nanoreservoirs for targeted drug delivery. Angew. Chem. Int. Ed. 2011, 50, 640–643.Search in Google Scholar
 Heinemann S, Coradin T, Worch H, Wiesmann HP, Hanke T. Possibilities and limitations of preparing silica/collagen/hydroxyapatite composite xerogels as load-bearing biomaterials. Compos. Sci. Technol. 2011, 71, 1873–1880.Search in Google Scholar
 Lee CH, Singla A, Lee Y. Biomedical applications of collagen. Int. J. Pharm. 2001, 221, 1–22.Search in Google Scholar
 Lee E-J, Jun S-H, Kim H-E, Koh Y-H. Collagen-silica xerogel nanohybrid membrane for guided bone regeneration. J. Biomed. Mater. Res., Part A 2012, 100, 841–847.10.1002/jbm.a.34019Search in Google Scholar
 Behring J, Junker R, Walboomers XF, Chessnut C, Jansen J. Toward guided tissue and bone regeneration: morphology, attachment, proliferation, and migration of cells cultured on collagen barrier membranes. A systematic review. Odontology 2008, 96, 1–11.10.1007/s10266-008-0087-ySearch in Google Scholar
 Bunyaratavej P, Wang H-L. Collagen membrane: a review. J. Periodontol. 2001, 72, 215–229.Search in Google Scholar
 Lin A, Goh M. Investigating the ultrastructure of fibrous long spacing collagen by parallel atomic force and transmission electron microscopy. Proteins 2002, 49, 378–384.10.1002/prot.10224Search in Google Scholar
 Stenzel K, Miyata T, Rubin A. Collagen as a biomaterial. Annu. Rev. Biophys. Bioeng. 1974, 3, 231–453.Search in Google Scholar
 Miyata T, Taira T, Noishiki Y. Collagen engineering for biomaterial use. Clin. Mater. 1992, 9, 139–148.Search in Google Scholar
 Wahl D, Czernuszka J. Collagen-hydroxyapatite composites for hard tissue repair. Eur. Cells Mater. 2006, 11, 43–56.Search in Google Scholar
 Lickorish D, Ramshaw J, Werkmeister J, Glattauer V, Howlett C. Collagen-hydroxyapatite composite prepared by biomimetic process. J. Biomed. Mater. Res., Part A 2004, 68, 19–27.10.1002/jbm.a.20031Search in Google Scholar
 Kikuchi M, Itoh S, Ichinose S, Shinomiya K, Tanaka J. Self-organization mechanism in a bone-like hydroxyapatite/collagen nanocomposite synthesized in vitro and its biological reaction in vivo. Biomaterials 2011, 22, 1705–1711.10.1016/S0142-9612(00)00305-7Search in Google Scholar
 Kikuchi M, Ikoma T, Itoh S, Matsumoto HN, Koyama Y, Takakuda K, Shinomiya K, Tanaka J. Biomimetic synthesis of bone-like nanocomposites using the self-organization mechanism of hydroxyapatite and collagen. Compos. Sci. Technol. 2004, 64, 819–825.Search in Google Scholar
 Oprita E, Moldovan L, Craciunescu O, Buzgariu W, Tardei C, Zarnescu O. A bioactive collagen-β tricalcium phosphate scaffold for tissue engineering. Cent. Eur. J. Biol. 2006, 1, 61–72.Search in Google Scholar
 Kim HW, Song JH, Kim HE. Bioactive glass nanofiber-collagen nanocomposite as a novel bone regeneration matrix. J. Biomed. Mater. Res., Part A 2006, 79, 698–705.10.1002/jbm.a.30848Search in Google Scholar
 Marelli B, Ghezzi CE, Mohn D, Stark WJ, Barralet JE, Boccaccini AR, Nazhat SN. Accelerated mineralization of dense collagen-nano bioactive glass hybrid gels increases scaffold stiffness and regulates osteoblastic function. Biomaterials. 2011, 32, 8915–8926.Search in Google Scholar
 Thimm BW, Unger RE, Neumann H-G, Kirkpatrick CJ. Biocompatibility studies of endothelial cells on a novel calcium phosphate/SiO2-xerogel composite for bone tissue engineering. Biomed. Mater. 2008, 3, 15007.Search in Google Scholar
 Lee E-J, Teng S-H, Jang T-S, Peng Wang S-WY, Kim H-E, Koh Y-H. Nanostructured poly(ɛ-caprolactone)-silica xerogel fibrous membrane for guided bone regeneration. Acta Biomater. 2010, 6, 3557–3565.Search in Google Scholar
 Li P, Ye X, Kangasniemi I, de Blieck-Hogervorst JM, Klein CP, de Groot K. In vivo calcium phosphate formation induced by sol-gel-prepared silica. J. Biomed. Mater. Res. Part A 1995, 29, 325–328.10.1002/jbm.820290307Search in Google Scholar
 Hamadouche M, Meunier A, Greenspan DC, Blanchat C, Zhong JP, Torre GPL, Sedel L. Long-term in vivo bioactivity and degradability of bulk sol-gel bioactive glasses. J. Biomed. Mater. Res., Part A 2001, 54, 560–566.10.1002/1097-4636(20010315)54:4<560::AID-JBM130>3.0.CO;2-JSearch in Google Scholar
 Balas F, Pérez-Pariente J, Vallet-Regí M. In vitro bioactivity of silicon-substituted hydroxyapatites. J. Biomed. Mater. Res. Part A 2003, 66A, 364–375.10.1002/jbm.a.10579Search in Google Scholar
 Hoppe A, Güldal NS, Boccaccini AR. A review of the biological response to ionic dissolution products from bioactive glasses and glass-ceramics. Biomaterials 2011, 32, 2757–2774.10.1016/j.biomaterials.2011.01.004Search in Google Scholar
 Coradin T, Boissiere M, Livage J. Sol-gel chemistry in medicinal science. Curr. Med. Chem. 2006, 13, 99–108.Search in Google Scholar
 Heinemann S, Heinemann C, Wenisch S, Alt V, Worch H, Hanke T. Calcium phosphate phases integrated in silica/collagen nanocomposite xerogels enhance the bioactivity and ultimately manipulate the osteoblast/osteoclast ratio in a human co-culture model. Acta Biomater. 2013, 9, 4878–4888.Search in Google Scholar
 Brasack I, Böttcher H, Hempel U. Biocompatibility of modified silica-protein composite layers. J. Sol-Gel Sci. Technol. 2000, 19, 479–482.Search in Google Scholar
 Ehrlich H, Heinemann S, Heinemann C, Simon P, Bazhenov VV, Shapkin NP, Born R, Tabachnick KR, Hanke T, Worch H. Nanostructural organization of naturally occurring composites—Part I: silica-collagen-based biocomposites. J. Nanomater. 2008, 2008, Article ID 623838.Search in Google Scholar
 Desimone MF, Hélary C, Quignard S, Rietveld IB, Bataille I, Copello GJ, Mosser G, Giraud-Guille M-M, Livage J, Meddahi-Pellé A, Coradin T. In vitro studies and preliminary in vivo evaluation of silicified concentrated collagen hydrogels. Appl. Mater. Interfaces. 2011, 3, 3831–3838.Search in Google Scholar
 Patwardhan SV, Clarson SJ, Perry CC. On the role(s) of additives in bioinspired silicification. Chem. Commun. 2005, 7, 1113–1121.Search in Google Scholar
 Coradin T, Lopez PJ. Biogenic silica patterning: simple chemistry or subtle biology? ChemBioChem. 2003, 4, 251–259.Search in Google Scholar
 Eglin D, Coradin T, Guille M-M G, Helary C, Livage J. Collagen-silica hybrid materials: sodium silicate and sodium chloride effects on type I collagen fibrillogenesis. Biomed. Mater. Eng. 2005, 15, 43–50.Search in Google Scholar
 Eglin D, Mosser G, Giraud-Guille M-M, Livage J, Coradin T. Type I collagen, a versatile liquid crystal biological template for silica structuration from nano- to microscopic scales. Soft Matter 2005, 1, 129–131.10.1039/b503019fSearch in Google Scholar
 Li P, Nakanishi K, Kokubo T, de Groot K. Induction and morphology of hydroxyapatite, precipitated from metastable simulated body fluids on sol-gel prepared silica. Biomaterials 1993, 14, 963–968.10.1016/0142-9612(93)90186-6Search in Google Scholar
 Peltola T, Jokinen M, Rahiala H, Levanen E, Rosenholm JB, Kangasniemi I. Calcium phosphate formation on porous sol-gel-derived SiO2 and CaO-P2O5-SiO2 substrates in vitro. J. Biomed. Mater. Res. 1999, 44, 12–21.Search in Google Scholar
 Stöber W, Fink A. Controlled growth of monodisperse silica spheres in the micron size range. J. Colloid Interface Sci. 1968, 26, 62–69.Search in Google Scholar
 Niu L-N, Jiao K, Qi Y-P, Nikonov S, Yiu CKY, Arola DD, Gong S-Q, El-Marakby A, Carrilho MRO, Hamrick MW, Hargreaves KM, Diogenes A, Chen J-H, Pashley DH, Tay FR. Intrafibrillar silicification of collagen scaffolds for sustained release of stem cell homing chemokine in hard tissue regeneration. FASEB J. 2012, 26, 4517–4529.Search in Google Scholar
 Shirtliff VJ, Hench LL. Bioactive materials for tissue engineering, regeneration and repair. J. Mater. Sci. 2003, 38, 4697–4707.Search in Google Scholar
 Ehrlich H, Worch H. Collagen, a huge matrix in glass-sponge flexible spicules of the meter-long Hyalonema sieboldi. In Handbook of Biomineralization. Vol.1. The Biology of Biominerals Structure Formation, Bäuerlein, E, Ed., Wiley VCH: Weinheim, Germany, 2007.Search in Google Scholar
 Carturan G, Toso RD, Boninsegna S, Monte RD. Encapsulation of functional cells by sol-gel silica: actual progress and perspectives for cell therapy. J. Mater. Chem. 2004, 14, 2087–2098.Search in Google Scholar
 Desimone MF, Hélary C, Mosser G, Giraud-Guille M-M, Livage J, Coradin T. Fibroblast encapsulation in hybrid silica/collagen hydrogels. J. Mater. Chem. 2010, 20, 666–668.Search in Google Scholar
 Bell E, Ivarsson B, Merrill C. Production of a tissue-like structure by contraction of collagen lattices by human fibroblasts of different proliferative potential in vitro. Proc. Natl. Acad. Sci. USA 1979, 76, 1274–1278.10.1073/pnas.76.3.1274Search in Google Scholar
 Yano S, Iwata K, Kurita K. Physical properties and structure of organic-inorganic hybrid materials produced by sol-gel process. Mater. Sci. Eng., C 1998, 6, 75–90.10.1016/S0928-4931(98)00043-5Search in Google Scholar
 Avnir D, Coradin T, Lev O, Livage J. Recent bio-applications of sol-gel materials. J. Mater. Chem. 2006, 16, 1013–1030.Search in Google Scholar
 Zhang Y, Hu L, Yu D, Gao C. Influence of silica particle internalization on adhesion and migration of human dermal fibroblasts. Biomaterials 2010, 31, 8465–8474.10.1016/j.biomaterials.2010.07.060Search in Google Scholar PubMed
 Ye X, Zhou Y, Sun Y, Chen J, Wang Z. Structure and infrared emissivity of collagen/SiO2 composite. Appl. Surf. Sci. 2008, 254, 5975–5980.Search in Google Scholar
 Niu L-N, Jiao K, Qi Y-P, Yiu CKY, Ryou H, Arola DD, Chen J-H, Breschi L, Pashley DH, Tay FR. Infiltration of silica inside fibrillar collagen. Angew. Chem. Int. Ed. Engl. 2011, 50, 11688–11691.Search in Google Scholar
 Brinker CJ, Scherer GW. Sol-Gel Science: The Physics and Chemistry of Sol-Gel Processing. Academic: Boston, 1990.Search in Google Scholar
 Zhao Y, Trewyn BG, Slowing II, Lin VS-Y. Mesoporous Silica nanoparticle-based double drug delivery system for glucose-responsive controlled release of insulin and cyclic AMP. J. Am. Chem. Soc. 2009, 131, 8398–8400.Search in Google Scholar
 Cai KY, Hu Y, Luo Z, Kong T, Lai M, Sui XJ, Wang YL, Yang L, Deng LH. Cell-specific gene transfection from a gene-functionalized poly(d,l-lactic acid) substrate fabricated by the layer-by-layer assembly technique. Angew. Chem. Int. Ed. 2008, 47, 7479–7481.Search in Google Scholar
 Hu Y, Cai K, Luo Z, Jandt KD. Layer-by-layer assembly of β-estradiol loaded mesoporous silica nanoparticles on titanium substrates and its implication for bone homeostasis. Adv. Mater. 2010, 22, 4146–4150.Search in Google Scholar
 Porter AE, Patel N, Skepper JN, Best SM, Bonfield W. Effect of sintered silicate-substituted hydroxyapatite on remodelling processes at the bone-implant interface. Biomaterials 2004, 25, 3303–3314.10.1016/j.biomaterials.2003.10.006Search in Google Scholar PubMed
 Li P, Ohtsuki C, Kokubo T, Nakanishi K, Soga N, de Groot K. The role of hydrated silica, titania, and alumina in inducing apatite on implants. J. Biomed. Mater. Res., Part A 1994, 28, 7–15.10.1002/jbm.820280103Search in Google Scholar PubMed
 Li P, Ohtsuki C, Kokubo T, Nakanishi K, Soga N, Nakamura T, Yamamuro T. Effects of ions in aqueous media on hydroxyapatite induction by silica gel and its relevance to bioactivity of bioactive glasses and glass-ceramics J. Appl. Biomater. 1993, 4, 221–229.Search in Google Scholar
 Shur I, Zilberman M, Benayahu D, Einav S. Adhesion molecule expression by osteogenic cells cultured on various biodegradable scaffolds. J. Biomed. Res., Part A 2005, 75, 870–876.10.1002/jbm.a.30507Search in Google Scholar PubMed
 Li J, Yun H, Gong Y, Zhao N, Zhang X. Effects of surface modification of poly (3-hydroxybutyrate-co-3-hydroxyhexanoate) (PHBHHx) on physicochemical properties and on interactions with MC3T3-E1 cells. J. Biomed. Mater. Res., Part A 2005, 75, 985–998.10.1002/jbm.a.30504Search in Google Scholar
 Berry CC, Campbell G, Spadiccino A, Robertson M, Curtis ASG. The influence of microscale topography on fibroblast attachment and motility. Biomaterials 2004, 25, 5781–5788.10.1016/j.biomaterials.2004.01.029Search in Google Scholar
 Andrade AL, Valerio P, de Goes AM, de Fátima Leite M, Domingues RZ. Influence of recovering collagen with bioactive glass on osteoblast behavior. J. Biomed. Mater. Res., Part B 2007, 83, 481–489.10.1002/jbm.b.30820Search in Google Scholar
 Heinemann C, Heinemann S, Worch H, Hanke T. Development of an osteoblast/osteoclast co-culture derived by human bone marrow stromal cells and human monocytes for biomaterials testing. Eur. Cells Mater. 2011, 21, 80–93.Search in Google Scholar
 Vallés G, Gil-Garay E, Munuera L, Vilaboa N. Modulation of the cross-talk between macrophages and osteoblasts by titanium-based particles. Biomaterials 2008, 29, 2326–2335.10.1016/j.biomaterials.2008.02.011Search in Google Scholar
 Zhou H, Wei J, Wu X, Shi J, Liu C, Jia J, Dai C, Gan Q. The bio-functional role of calcium in mesoporous silica xerogels on the responses of osteoblasts in vitro. J. Mater. Sci: Mater. Med. 2010, 21, 2175–2185.Search in Google Scholar
 Knabe C, Driessens FCM, Planell JA, Gildenhaar R, Berger G, Reif D, Fitzner R, Radlanski RJ, Gross U. Evaluation of calcium phosphates and experimental calcium phosphate bone cements using osteogenic cultures. J. Biomed. Mater. Res. 2000, 52, 498–508.Search in Google Scholar
 Hempel U, Reinstorf A, Poppe M, Fischer U, Gelinsky M, Pompe W, Wenzel KW. Proliferation and differentiation of osteoblasts on Biocement D modified with collagen type I and citric acid. J. Biomed. Mater. Res., Part B 2004, 71, 130–143.10.1002/jbm.b.30082Search in Google Scholar
 Romanello M, D’Andrea P. Dual mechanism of intercellular communication in hobit osteoblastic cells: a role for gap-junctional hemichannels. J. Bone Miner. Res. 2001, 16, 1465–1476.Search in Google Scholar
 Xu C, Su P, Chen X, Meng Y, Yu W, Xiang AP, Wang Y. Biocompatibility and osteogenesis of biomimetic bioglass-collagen-phosphatidylserine composite scaffolds for bone tissue engineering. Biomaterials 2011, 32, 1051–1058.10.1016/j.biomaterials.2010.09.068Search in Google Scholar
 Hum J, Boccaccini AR. Bioactive glasses as carriers for bioactive molecules and therapeutic drugs: a review. J. Mater. Sci: Mater. Med. 2012, 23, 2317–2333.Search in Google Scholar
 Vallet-Regí M. Ordered mesoporous materials in the context of drug delivery systems and bone tissue engineering. Chem. – Eur. J. 2006, 12, 5934–5943.10.1002/chem.200600226Search in Google Scholar PubMed
 Vallet-Regí M, Balas F, Arcos D. Mesoporous materials for drug delivery. Angew. Chem. Int. Ed. 2007, 46, 7548–7558.Search in Google Scholar
 Vallet-Regí M, Balas F, Colilla M, Manzano M. Bone-regenerative bioceramic implants with drug and protein controlled delivery capability. Progr. Solid State Chem. 2008, 36, 163–191.Search in Google Scholar
 Tu H-L, Lin Y-S, Lin H-Y, Hung Y, Lo L-W, Chen Y-F, Mou C-Y. In vitro studies of functionalized mesoporous silica nanoparticles for photodynamic therapy. Adv. Mater. 2009, 21, 172–177.Search in Google Scholar
 Slowing II, Vivero-Escoto JL, Wu C-W, Lin VS-Y. Mesoporous silica nanoparticles as controlled release drug delivery and gene transfection carriers. Adv. Drug Delivery Rev. 2008, 60, 1278–1288.Search in Google Scholar
 Zhao W, Gu J, Zhang L, Chen H, Shi J. Fabrication of uniform magnetic nanocomposite spheres with a magnetic core/mesoporous silica shell structure. J. Am. Chem. Soc. 2005, 127, 8916–8917.Search in Google Scholar
©2013 by Walter de Gruyter Berlin Boston
This article is distributed under the terms of the Creative Commons Attribution Non-Commercial License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.