Inorganic polymeric phosphate, polyphosphate (polyP), can be chemically prepared either in a crystalline or an amorphous state [11,12]; the biogenic polyphosphate (bio-polyP) is amorphous [12,18]. Similar to silica/bio-silica, synthesis of polyP requires high temperature [19], while the biogenically formed bio-polyP is metabolically produced in bacteria and animals at ambient temperatures via kinases (see [12]). Bio-polyP is synthesized in a wide range of organisms, from bacteria, fungi and algae to plants and animals (see [12]). The polymer is readily soluble in water at millimolar concentrations with chain lengths below 100 phosphate units [20,21]. The natural bio-polyP is only found as a linear polymer, in which tens to hundreds of phosphate residues are linked by phosphoanhydride bonds. In the absence of any enzyme, the phosphoanhydride bonds within the polymer are stable over wide temperature and pH ranges [11,12]. PolyP had been and is successfully used as a food additive as well as a base material in the formulation of cosmetic products [22]. The nutritional benefit of polyP in animal food materials is well established [23]. Since bio-polyP exists as a multivalent anion, it binds to essential cations, such as Ca, Mg, Mn, Fe, and Co ions (see [12]). More specifically, bio-polyP acts as a strong Ca²⁺ chelator and also as an antioxidant [24].

The biological function of polyP has been studied to some details in microorganisms (reviewed in [25]) and more recently also in animals (reviewed in [12,26]). Based on experimental data, it has been proposed that bio-polyP acts as a storage substance of energy, as a chelator for metal cations, as an inducer of apoptosis, and—importantly—as a stimulating agent in mineralization of bone tissue [26,27,28] (reviewed in [11]). Likewise important, bio-polyP acts as a modulator of gene expression. Results [29] suggest that in the osteoblast-like cell line, MC 3T3-E1, bio-polyP causes an increased gene expression of osteocalcin, osterix, bone sialoprotein, and tissue non-specific alkaline phosphatase, all proteins known to be crucial for bone formation [30,31]. The gene expression data in MC 3T3-E1 cells have been obtained with 1 mM polyP [29,32].

Bio-polyP can increase skeletal mineralization process, but until today, it is still unknown whether this happens in its polymeric form or in monomeric phosphates which are formed from bio-polyP through hydrolysis by phosphatases [33]. The susceptibility of bio-polyP for enzymes, especially phosphatases is well established [34,35,36]. As one consequence of the enzymatic hydrolysis of polyP, a release of Ca²⁺ ions has been proposed; this cation is metabolically utilized during hydroxyapatite formation [18].

Very recently, it could be demonstrated that bio-polyP displays morphogenetic activity on bone-forming osteoblasts, SaOS-2 cells, and inhibitory activity on RAW 264.7 cells acting as osteoclasts. The osteoblast-like SaOS-2 cells form hydroxyapatite crystals, in response to exposure to bio-polyP, based on their potency to express key molecules known to control hydroxyapatite formation (see [37,38,39]), e.g., the bone morphogenetic protein 2 (BMP-2), an inducer of bone formation [40], osteoprotegerin (OPG), a cytokine that is expressed in osteoblasts with a significant role in the maturation of osteoclasts as well as in the control of bone mineral density [41], and the receptor activator of the NF-κB ligand (RANKL), a mediator that binds to RANK which is a receptor that mediates maturation of osteoclasts [42]. In turn, the relative concentration ratio between OPG and RANKL is crucial for the differentiation and survival of osteoclasts, since OPG can bind to RANKL and by that inactivates its function [7,38]. Hence, the OPG and RANKL ratio controls the osteoinductivity on the level of RANKL, a decisive ligand required for the differentiation of osteoclasts [43]. Likewise, the osteoclast-like RAW 264.7 cells have the potency to readily differentiate into osteoclasts when they are exposed to recombinant RANKL [44] and, by that, have been successfully used as a model for studies of osteoclastogenesis in vitro [45]. It is also important to mention that the activity of bio-polyP can induce alkaline phosphatase, an enzyme which provides inorganic phosphate required for the synthesis of hydroxyapatite [39].

In order to avoid any chelating activity of bio-polyP in our in vitro studies, we applied this polymer together with CaCl₂ (bio-polyP•Ca²⁺ complex). This precaution excludes the possibility that the observed biological effects might be attributed to a depletion of Ca²⁺ ions that are required for hydroxyapatite deposition [46]. The bio-polyP•Ca²⁺ complex was found to be a strong inducer of hydroxyapatite formation in SaOS-2 cells and in particular, to cause an enhanced expression of BMP-2 [47]. In parallel, bio-polyP•Ca²⁺ strongly inhibits the progression of RAW 264.7 cells into osteoclasts, which is reflected by the reduction of cells expressing tartrate-resistant acid phosphatase (TRAP), a well established marker protein for terminally differentiated osteoclasts [48]. As an additional endpoint marker for osteoclast differentiation, the effect of bio-polyP•Ca²⁺ on the function of IκBα kinase (nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, alpha) was determined. This kinase is one key molecule which causes the activation of NF-κB during RANKL-caused (pre)osteoclast differentiation (see [49]). The results revealed that bio-polyP•Ca²⁺ inhibited at low concentrations (10 to 100 μM) the phosphorylation, and by that, the signaling function of IκBα via the respective kinase in RAW 264.7 cells [47].

The effect of bio-polyP on the biomineralization extent (hydroxyapatite formation) had been studied and the light microscopic images are given here (Figure 1). Incubation of SaOS-2 cells was performed for 10 days on plastic coverslips in 24-well plates, using McCoy’s medium/10% FCS (fetal calf serum) in the absence or presence of the activation cocktail, composed of β-glycerophosphate–ascorbic acid–dexamethasone [47]. The cultures which were grown in the presence of the activation cocktail were exposed to 10 μM polyP or 10 μM bio-polyP•Ca²⁺ for 10 days. Then the slips were stained with alizarin red S, a largely sensible indicator dye for hydroxyapatite visualization. An inspection of the cultures on the coverslips revealed, as expected, that the intensity of the staining was lowest in cultures grown in the absence of activation cocktail (Figure 1Ea). In contrast, cultures that were grown in the presence of activation cocktail showed a strong intensive staining, reflecting a higher level of hydroxyapatite mineralization (Figure 1Eb). If the cultures were treated with the activation cocktail and 10 μM polyP, no striking difference in the intensity of alizarin red S staining, compared with cultures grown with activation cocktail only, was observed (Figure 1Ec). However, if the cultures were grown in medium containing the activation cocktail and 10 μM bio-polyP•Ca²⁺, a strong and almost homogeneous staining of the cell layer was seen, a finding that reflects the strong hydroxyapatite formation by SaOS-2 cells (Figure 1Ed).

A closer inspection of the inducing effect of bio-polyP•Ca²⁺ by digital light microscopy revealed that the samples grown in McCoy’s medium/FCS showed only a scattered staining of the cell layer (Figure 1Aa). Likewise, the red staining was low if the samples were examined by red/green emitting fluorescent light (Figure 1Ab). A distinct increase in the red intensities was seen if the cells were cultivated in the presence of activation cocktail and then stained with alizarin red S in order to monitor the hydroxyapatite deposition (Figure 1B). The red patches (Figure 1Ba) as well as the red fluorescence areas (Figure 1Bb) strongly increased and extended over 50% of the visual fields. If the cells were incubated for 7 days in the presence of both the activation cocktail and 10 μM bio-polyP, the red patches/red fluorescence areas were smaller (Figure 1Ca,Cb) compared to the fields stained with alizarin red S in cultures grown with the activation cocktail only (Figure 1B). However, if the cultures were incubated with the activation cocktail together with 10 μM polyP (Ca²⁺ salt), an almost complete red staining was seen, reflecting an intensive hydroxyapatite mineralization (Figure 1D). These data show that the activation cocktail is required for extensive mineralization of SaOS-2 cells. Moreover, it is demonstrated that bio-polyP, if complexed with Ca²⁺ strongly augments the inducing effect of the activation cocktail on hydroxyapatite crystallite formation.

Taken these data together we show that bio-polyP affects the tuned balance between the osteoblasts and osteoclasts in the anabolic direction, implying that hydroxyapatite synthesis is favored at the expense of hydroxyapatite degradation/dissolution. A summary of the action modes of bio-polyP is given in the schematic presentation shown in Figure 2.

The formation of the skeletal system from the earliest metazoans, the sponges (phylum Porifera) [50] to the crown taxa, the mammalians [51] and the insects [52], is dominated by a tuned communication between cells controlling anabolic processes and cells executing catabolic reactions. Basically, two kinds of inorganic scaffold materials had been applied in the metazoan kingdom to form skeletons: calcium (calcium-based skeletal systems) and—only found in siliceous sponges—silica (silica-based skeletons).

A breakthrough in the understanding of the siliceous spicule formation of the demosponges and the hexactinellid sponges came from the discovery that the axial filaments of the spicules, the skeletal elements of these sponges (demosponges [53]; hexactinellids [54]) contain an enzymatically active protein which synthesizes polymeric silicate, the bio-silica. This enzyme, termed silicatein, has been found to catalyze/polycondensate bio-silica during the axial and radial growth of the spicules. In contrast to plant phytoliths and diatom frustules, where bio-silica is deposited from a super-saturated solution onto organic templates, the siliceous spicules of sponges are formed in a hypo-saturated intraorganism environment following an enzymatic mechanism by lowering the activation energy of the polycondensation reaction. In in vitro systems, orthosilicic acid or TEOS (tetraethyl orthosilicate) has been used as substrate/bio-silica precursor for the enzymatic reaction. During the latter process, ethanol is released [53]. The silicatein-mediated formation of silica proceeds at silica substrate concentrations of around 200 μM, far below the concentrations which are required for the chemical condensation of 1 mM or higher at neutral pH [55] (reviewed in [56,57]).

The silicateins were first identified in the axial filament of the demosponge Tethya aurantium [58]. They comprise a family of related protein sequences which are consisting of three isoforms, silicatein-α, silicatein-β, and silicatein-γ, in a molar ratio of 12:6:1. The silicateins belong to the papain-like cysteine protease superfamily and are the most closely related to the cathepsin family [59]. The first cathepsin in sponges was identified and cloned from the demosponge Geodia cydonium [60].

Due to their excellent tissue compatibility and property, the physiologically bio-degradable substitution materials from natural origin, e.g., collagen [77], or silk fibroin [78], or derived and processed natural biomaterials/biopolymers, e.g., chitosan [79], starch [80], or poly(3-hydroxybutyrate) [81], and finally also totally synthetic polymers such as poly(lactide) [82], poly(lactide/glycolide) [83], and polycaprolactone [84], have been developed. These biopolymers mimic the supra-molecular structures of the natural extracellular matrix of tissues and have the very suitable biocompatibility characteristics of human bone. Also their porosity and fibrous network allow a “homing” for both the progenitor and the differentiated bone cells.

The substitution materials based on demineralized bone matrix derived from human tissue provide a suitable advancement of the grafting material for repairing bone defect, since they retain the ability to act not only osteoconductively, allowing the circulating bone cells to attach and to proliferate but also to be osteoinductive [85], because they still contain bone growth factors, e.g., BMP-2, TGF-β (transforming growth factor beta), IGF-II (insulin-like growth factor 2), or PDGF (platelet-derived growth factor), even though in different concentrations [86]. Recently, a collagen microencapsulation technology had been described, by which bone marrow-derived mesenchymal stem cells had been entrapped into a biomimetically fabricated collagen fiber meshwork. This injectable material (microspheres) had been studied in vitro and found to display not only osteoconductive but also osteoinductive properties [87].

Focusing on processed, natural biomaterials, e.g., chitosan, those biopolymers are naturally lacking of these bone growth factors which have to be added after fabrication [88]. Likewise biologically or morphogenetically inactive materials are the totally synthetic polymers. This fact is not plausible since those polymers are lacking of the ligands required for the interaction with the receptors on the bone cell surfaces in order to induce the intracellular signaling cascade(s). Those stimulations by specific transcription factors are the clue for any cell- and tissue-specific differentiation process resulting in a functional and spatial interaction of cells to produce the extracellular bone structure.

Pure chemically prepared and fabricated bone materials, e.g., calcium phosphate, calcium sulfate or coralline carbonate and phosphate grafts, are suitable as osteoconductive implants, since they provide stability to the damaged bone. Hence they act osteoconductively and to some extent also osteointegratively, but they are lacking of any osteoinductive properties (see [89]). To take advantage of their excellent mechanical properties, these materials have to be biologically functionalized (reviewed in [6]).

A similar functionalization with biological ligands has to be performed with titanium and its alloys that are widely used as orthopedic and dental implant materials. The major challenge is the mismatch which exists between the mechanical properties of the implant material with the bone tissue [90]. One established solution is the development of adjusted Ti/Nb/Zr/Sn titanium alloys whose Young’s modulus is close to the physiological Young’s modulus of approximately 50 GPa [91]. Two attractive functionalization procedures have been described, first the immobilization of the surface with RGD-ligands [92] and second, tailoring/patterning the surface of the titanium material with nano-spikes which fit in the architectural arrangement of the integrin receptors of the bone cells by which a functional morphology and an expression of growth factors in bone cells are elicited [93].

The functional interaction of osteoblasts with osteoclasts can be impressively described with reference to the widespread disease, osteoporosis; this degenerative bone disease causes loss of bone tissue and reduction of bone density which are reflected by a micro-architectural deterioration of the bone [94,95]. The cellular basis for this disorder is an imbalance between the bone-forming osteoblasts and the bone-resorbing osteoclasts. Even though the bone forming and bone resorbing cells differentiate from different cell lineages, their functions in vivo are intimately linked and their differentiation levels are reciprocally controlled [96]. The major transcription factor involved in the differentiation and proliferation of osteoprogenitor cells is Runx2, a factor that is expressed in the mesenchymal stem cells and along the different stages of the osteoblast lineage [97]. Runx2 itself is under the control of BMP-2 [98]. These inducer factors cause a stage-correlated and increasing expression of a series of genes, for example, of those encoding the bone-specific alkaline phosphatase (b-ALP), collagen type I (COL-I), osteopontin (OP) and—at a later stage—of RANKL, asialoprotein (ASP), bone sialoprotein (BSP) and osteocalcin (OC). The osteoblasts that produce hydroxyapatite, finally differentiate to osteocytes that remain entrapped in the hydroxyapatite deposits or undergo apoptosis [99]. In a feed-back loop, the osteocytes express sclerostin which functions as a potent antagonist of BMP-2. This effect can be counteracted by the parathyroid hormone [100] (Figure 5).

In contrast to osteoblasts, osteoclasts are multi-nucleated cells that originate from the hematopoietic lineage [101,102]. Those stem cells undergo differentiation and maturation in the presence of the macrophage colony-stimulating factor (M-CSF) and of RANKL. As markers for the multi-nucleated osteoclasts, the high expression levels of TRAP, of the calcitonin receptor (CTR) binding protein as well as of the expression of integrin a_vb₃ have been used [103] (Figure 2 and Figure 5). The above mentioned cytokine/receptor triad crucially controls bone formation and bone remodeling, RANKL with its receptor RANK and the endogenous decoy receptor OPG [104,105]. RANKL is synthesized by the osteoblastic lineage cells and is essential for the differentiation of those cells which are involved in bone resorption, the osteoclasts. RANKL is expressed on osteoblasts, T cells, dendritic cells, and their precursors, from where it can be released by specific proteases [106]. After binding of RANKL to RANK, the osteoclasts become activated and resorb bone mineral. During this process, the cells have close contact to the bone surface [107]. At this interphase, the bone vesicles are formed, via integrin (a_vb₃) which contain proton pumps and acid hydrolases (cathepsin K) (Figure 2). Those enzymes and vesicles are inserted into the cells at the bone-apposed area under formation of a “ruffled border”. A “resorptive hemivacuole” is formed between cell and bone, allowing the protons to dissolve hydroxyapatite of the bone (Figure 2). The intracellular pH is kept at a near-neutral level by chloride/bicarbonate exchange and the help of carbonic anhydrase [108].

To highlight again, the activity and function of RANKL is under control of OPG, which is secreted by stroma cells and also by osteoblasts [109]. OPG scavenges RANKL by binding to it and neutralizes its function. From these results, it is pressing to conclude that any deregulation of the tuned expression of the RANKL/RANK/OPG system causes a dysregulation of the differentiation pathways of the osteoblasts and the osteoclasts and in turn impairs bone remodeling [104]. More specific, OPG abolishes the activation of the osteoclast via inhibition of the RANK pathway and by that prevents bone matrix from excessive resorption. Hence, the relative concentration of OPG and RANKL in bone is the major morphogenetic determinant of bone mass and strength.

In vitro and in vivo studies have been performed to counteract the age-correlated OPG-RANKL imbalance [110]. It has been found that the OPG-RANKL ratio is influenced by a series of substances, including hormones and cytokines. In the course of these studies, it could be established that treatment of osteoblasts with stimulators of osteoblast formation (e.g., vitamin D, PTH, prostaglandin E2 or interleukin-11) upregulates the expression of RANKL. Addition of BMP-2, interleukin-1b or TNF-α to those cells increases the OPG mRNA steady-state level and subsequently the OPG protein synthesis in fetal osteoblasts. Exactly here, at this regulatory step, the two biopolymers interfere with the relative RANKL/RANK/OPG levels. In response to bio-silica, the relative expression level of OPG increases [37] and as deduced from the existing data in the literature reduces the functional activity of RANKL.

The biopolymer bio-polyP inhibits the phosphorylation of the IκBα kinase in RAW 264.7 cells, and by that prevents the progression/differentiation of the pre-osteoclasts to the mature osteoclasts [47]. In addition, bio-polyP inhibits the expression of TRAP, amplifying the effect of this polymer on the kinase [47]. Finally it should be mentioned that a new inhibitory cross-talk effect is caused by bio-silica, which is centered in a hitherto not yet characterized inhibitory factor.

These modes of inhibiting/activating interventions of the two biopolymers, bio-polyP and bio-silica, are sketched in Figure 5.

It is the aim of a fabricated scaffold to provide the surrounding cells a three-dimensional (3D) platform onto which bone cells can grow, differentiate and finally deposit hydroxyapatite. Ideally the scaffold is composed of natural fibers which build structures to meet this requirement; those compositions are found in the extracellular matrix (ECM). The dominant structural protein in the ECM is collagen [111]. It builds the 3D structure with nanofibers which have a diameter between 50 and 500 nm [112], and forms the founding architecture for the growing bone. It is collagen type I that forms the structural, organic building of the bone to 90% [113]. There, collagen acts as the nucleation site for the growth of the highly ordered crystals, formed of carbonated apatite [Ca₅(PO₄,CO₃)₃(OH)]. Initially they are only a few nm thick [39]. As the skeletal structure, acting in nature as support for soft tissue [71], bones represent a hybrid material which makes them mechanically flexible, strong, stiff, tough and also lightweight [114].

The fabricated scaffold designed to mimic the natural inorganic/organic 3D bone structure must follow the nanofibrous architecture and must be engineered with high porosity in order to allow an ingrowth of cells and—in addition—an efficient transport of morphogens, cytokines, growth factors and also nutrients, oxygen as well as waste products (see [115]). A further characteristics and prerequisite for an efficient and functional scaffold is the vascularization of the inserted materials in order to avoid necrotic processes in the center of larger constructs. In turn the materials to be used as a biomimetic scaffold, either formed of an organic ground substance or of metal (titanium), must meet the acting mechanical stresses that occur during tissue neogenesis. Besides of being mechanically suitable, the materials must be osteoconductive, e.g., by using ceramics [116]—or ideally even osteoinductive. Furthermore, the material must be (minimally) biodegradable in order to circumvent the development of immune response and to allow new physiological bone cells to invade, grow and differentiate followed by the formation of an organic fiber network, and finally the full bone tissue regeneration.

Besides of using nanofibrous scaffolds, prepared by molecular self-assembly, bacteria derived hydrogels, or by thermally-induced phase separation (reviewed in [115]), bone substitution materials can be prepared (1) by electrospinning or (2) by three-dimension (3D) printing. The electrospinning process allows the fabrication of 3D materials by using nanofibrous matrices that comprise some structural similarities to the physiological ECM at low cost (reviewed in [117,118,119,120,121]). This process takes advantage of electrostatic forces to generate polymer fibers. Typically the fibers have diameters of 100 to 500 nm to form nanofibers. Often the synthetic organic material contains PLA (polyglycolic acid) and PEG [poly(ethylene glycol)] also together with natural organic polymer e.g., chitin derivatives [121].

3D printing is a—relatively straightforward—technique by which a binding solution is printed into layers of powder, a process that is computer-controlled and imitates a sliced virtual model [122,123]. This technique had been successfully used to develop bone substitution materials with optimized integration and functionality characteristics [124]. Those implants can be tailored to a given individual defect 3D geometry, following the anatomical data information obtained from the patient. Especially calcium phosphate and bioactive glasses have been used as suitable starting materials for the fabrication of the 3D structures [125]. Usually the fabrication of 3D printed bone substitution materials involves a sintering step after an initial aggregate formation of the calcium phosphate and bioactive glass granules. An improvement, with respect to the (partial) elimination of the higher temperature step, was the development of the advanced microdispensor Ultimus technology, by which both fibrin scaffold and PLA meshes were prepared [126]. Likewise, also the layer-by-layer printing of 3D tri-calcium phosphate required a post sintering phase [127]. Despite this disadvantage, those scaffolds are promising materials allowing the osteoclastic cells to differentiate properly; in addition, this material is resorbable.

In order to fabricate 3D scaffolds containing bio-organic polymers, e.g., collagen as supporting fiber, and avoiding any denaturating temperatures, the solid freeform fabrication (SFF) technique [128] coupled to the indirect 3D printing technique [129] was developed. In an exemplary procedure, a collagen scaffold was prepared by using the coupled SFF/indirect 3D printing technique [129]. Tendon collagen type I was dispersed in acetic acid buffer (pH about 3). Using this material, the scaffold was prepared following the SFF strategy [129,130,131]. At first, a computer-simulated negative mold was designed into which the collagenous solution was poured, followed by freezing at −30 °C. Finally, a dehydration and in turn a dissolution process was followed and by that a negative mold with a predefined microchannel network was created. After a critical point drying and cross-linking, the scaffolds was obtained.

The inorganic bio-silica scaffold of siliceous sponge spicules is distinguished by its origin, enzymatically formed compared to all other bio-silica deposits. This origin implies that the nanoparticles formed by silicatein are spatially arranged and deposited via an enzyme-guided pathway [132]. Even more, the enzyme itself, silicatein is not only a catalytically active protein, but also a structure-giving backbone for the primordial/growing bio-silica fibrils [133]. Studying the maturation process of silicatein on a molecular level if it became overt that the enzyme acquires these properties (to be enzymatically active and to be structure-guiding) during the maturation, from the pro-silicatein to the active enzyme, a process during which the propeptide is split off from the inactive precursor [134]. During this hydrolytic cleavage, the active enzyme undergoes a conformational change that the molecule becomes less soluble and precipitates. This effect, which is seen at higher temperature (around 20 °C) results in the formation of self-assembled structures of silicatein, passes a fractal stage. The precipitated silicatein molecules can be re-dissolved in the presence of urea, without losing the biological activity. Those self-assembly formations are stabilized by addition of silintaphin-1 [135], a natural interactor of silicatein, and of PEG, a synthetic polymer. Adding the latter two components (silintaphin-1 and PEG) to the silicatein-mediated enzymic reaction results in the formation of biosilica cubes; those deposits become flat/planar and hard. The bio-silica product increases its compactness if silicatein is supplemented with silintaphin-1 or PEG. The elastic modulus of the silicatein-mediated biosilica product increases in parallel with the addition of silintaphin-1 and/or PEG from 17 MPa (silicatein) via 61 MPa (silicatein:silintaphin-1) to 101 MPa (silicatein:silintaphin-1 and PEG). This increase in hardness, in the presence of PEG, makes the structure-guiding enzyme, silicatein, suitable for the fabrication of controlled, pattern-forming structures.

We are applying these inherent properties of silicatein for 3D bio-silica/silicatein scaffolds fabrication via an indirect printing approach, a SFF/indirect 3D printing technique. The mold for the bio-silica scaffold was designed using computer in order to fabricate the negative mold; bone specimen from a vertebrate femur was used as a template (Figure 6A). In the initial experiments, we did not apply this new technique for the detailed osteon structures (Figure 6B) in order first to establish the proof of principle. In this new process, bio-silica was synthesized first in the presence of silicatein [133]. After the 1 h incubation period, the enzymatic reaction was terminated and the bio-silica product formed was supplemented with PEG and bio-silica at a molar ratio of 1:0.1. The bio-silica/PEG reaction mixture was inspected after 10 min (Figure 6C), 30 min (Figure 6D) and 60 min (Figure 6E) with an optical microscope. It is apparent that the initially formed fluffy bio-silica material increases in stiffness (Figure 6C,D) and the bio-silica polymer formed became finally homogenous and plane-structured (Figure 6E). This process—the hardening of the loosely formed initial bio-silica to the structured PEG/bio-silica—was applied to form the 3D printed bio-silica scaffold (Figure 6F).

Parallel to this SFF/indirect 3D printing technique, we have also successfully applied the direct 3D printing technology. We followed the strategy that had been described for starch [136]. In brief, the starch scaffold was prepared and hardened. Then the 3D fabricated starch scaffold was incubated both with silicatein and its substrate ortho-silicate. Finally the framework was hardened by spraying with ortho-silicate together with PEG at a molar ratio of 1:0.1. For printing, the apparatus ZPrinter 450 (ZCorporation, Rock Hill, SC 29730; USA) had been used. The compressive strength and the stiffness of the sample were tested, and values of approximately 27.4 MPa (average compressive strength) were determined.

The biocompatibility of the bio-silica is excellent and no toxicity was determined in vitro, applying the MTT-viability assay. In addition, the requirement with respect to the porosity allowing a suitable assembly and growth of bone cells within the cavities of the scaffold was analyzed. The histological examination revealed that the cells within the scaffold showed proliferation and differentiation and—after addition of the mineralization cocktail—formed hydroxyapatite crystals. Furthermore, the scaffolds displayed a degradation at the physiochemical-biomimetic environment, in the cell culture medium/serum after 12 days. These analyses supported the view that the porosity of the material is suitable for the infiltration of cells and for a physiological interaction of the cells within the cavities. The pore sizes are within the range of 80–140 μm.

In initial studies, we could demonstrate that the biogenically and morphogenetically active bio-polyP is a likewise ideal polymer to be used for building of scaffold materials. Especially the chemical properties to be soluble at physiological pH conditions as a salt with the cations Na⁺ and K⁺, while becoming insoluble with cation Ca²⁺, provides this material an essentially suitable property to undergo hardening after the 3D printing process. Furthermore, the Ca²⁺ salt of polyP is—like the Na⁺ and K⁺ salts—biologically active and causes an increased release of the cytokine BMP-2.