Introduction

Over the course of its life, the integrity of any metazoan, from the simplest organisms to humans, is compromised to varying degrees by injury or chronic degenerative changes, leading to dysfunction. In humans, when the physiological repair capacity is insufficient, medical intervention is required to prevent a collapse of the fragile cellular homeostasis and the tuned organization of cells and tissues. There is a need for new concepts in regenerative therapy. Animals that have physiologically extensive regenerative abilities, such as sponges, are possible models for copying these concepts from nature. During the evolution of the metazoans from their basal phylum, the sponges (Porifera), with their ~40,000 predicted protein-coding loci,1 to the “crown taxa”, humans, with 20,000 to 25,000 genes,2 a gradual decrease in the genetic complexity is observed. This fact supports the view that the basal animals like sponges have a higher redundancy potential, combined with a variety of alternative metabolic circuits that provide them with a higher evolutionary stability. In turn, the regeneration capability of basal animals is more extensive compared to the “more advanced” taxa, including humans.

In animals, particularly in humans, some defects are not adequately repaired especially when physiological functional restoration is not supported by the inherent regeneration potential. If tissue/cell homeostasis is not maintained, the repair processes are insufficient with the result that the affected, damaged tissue is only replaced by tissue-like entities that lack physiological and biochemical functions. Or, as in present-day solutions, the defective parts are replaced or augmented by largely inert or only partially biological substituting materials. For example, the deficiencies of the bone-supporting implants can be overcome with tissue regenerating materials that restore not only the mechanical but also the biological and functional properties of the damaged tissue in the defect region. It is important to emphasize that any regeneration process requires an optimal physiological functioning of the cells, which depends on molecular and mechanical signals from the environment.3 Both the cells in the tissue surrounding the defect, and particularly the cells within the areas to be repaired, must be supplied with growth factors and nutrients that support the anabolic repair pathways and, importantly, with metabolic energy (ATP or another energy-carrying molecule).4 Not only intracellularly but also extracellularly, a sufficient supply of biochemically useful energy is important for the regeneration of the extracellular matrix (ECM). The number of cells in these regions is often low, for example in cartilage, which contains only ~10% chondrocytes.5 During regeneration, biochemical reactions in the ECM are driven by the cell-based synthesis of fibrous mats (collagen) or hydrogels (hydrogel-forming polypeptides). These reactions are anabolic reactions that require the conversion of chemical fuels like glucose into ATP (energy). In addition, enzymes are needed that break down larger precursor molecules in order to drive anabolic ones in bone, cartilage, and muscle regeneration.6 Accordingly, the two cornerstones, enzymes and energy, contribute significantly to the regeneration pathways. The initial basis for this gain came from studies from the basal animals, the sponges.

This review has a special focus on the regenerative abilities of two inorganic biomaterials in the formation and repair of both soft and hard tissues, amorphous silica (“biosilica”) and amorphous polyphosphate (polyP). Both biomaterials or biopolymers are regeneratively active only in their amorphous form and, in accordance with their physiological function, exhibit properties that the conventional organic biopolymers used in regenerative medicine do not have, such as in the case of polyP, the delivery of metabolic energy required for energy-dependent repair processes, as discussed in this review. In particular, the role that sponges, as the most basal phylum, have played in understanding the critical function of enzymes in the energy-dependent tissue regeneration/repair processes, including the formation of biomineralized structures, is discussed, while previous reviews on these biomaterials primarily focus on either only one these materials7,8 or on their physicochemical properties or more mechanistic aspects.9

Regeneration Capacity Across Animal Taxa

The basic principles of regeneration processes have been enigmatic for a long time. In 1606, it was stressed that bone formation and the regeneration of this organ depend on spiritual forces, virtus naturalis, based and localized in the lower body area, which is characterized by a “warm and humid” local environment.10 Later, especially by Trembley, first evidence-based experiments were performed, which revealed remarkable regenerative abilities especially in the more basal Metazoa such as the phylum Cnidaria.11 The sponges (Porifera), the basal phylum of Metazoa, traditionally served as a model system for regeneration.12 Animal taxa that evolved later from sponges have a reduced regeneration capacity mainly due to the presence of lower amounts of stem cells.13 Thomas Hunt Morgan introduced the perspective of functional genomics by studying the regeneration of segments in earthworms as a model.14

In Table 1, a summary of the regeneration capacity across various animal taxa is shown. The ability to regenerate damaged or lost tissue greatly varies among different species and can include the whole body, the primary body axis, or only certain structures or tissues.15 Whole-body regeneration is particularly found in the basal metazoans such as sponges, cnidarians and ctenophores.16,17 The regeneration of the sponge body or sponge tissue after an injury can either start from small body fragments or occur via aggregation of dissociated sponge cells.16 Sponges as sessile marine organisms are particularly prone to wounding, eg, by grazers or mechanical injury, which may lead to the occurrence of chronic wounds if repair fails and homeostasis is not retained. Investigations of the response of the sponge Aplysina aerophoba to damage caused mechanically or by a spongivorous opisthobranch using differential gene expression analysis based on RNA-sequence data revealed a repair mechanism involving enzymes/proteins such as metalloproteases, transglutaminases, and integrins, and signaling pathways (Wnt and mitogen-activated protein kinase – MAPK), similar to that involved in wound healing in higher animals, including humans.18

Table 1 Regeneration Capacity Across Various Animal Taxa

Taking Concepts from Nature as a Blueprint for Regeneration

Intense efforts over the past 40 years have uncovered, in a scientific, causal-analytical manner, the strategies of how living systems function with the aim of exploiting them to meet the urgent needs in engineering tissue replacement biomaterials. Biomimetic and bioinspired approaches based on these achievements have contributed considerably to a paradigm change in the field of tissue engineering.20–22 Studies on sponges, the evolutionarily oldest metazoans, showed that these animals produce a range of organic biomolecules that they effectively defend against bacteria and viruses.23–25 Then, the application of molecular biology techniques made it possible to show – what was not expected before – that these basal animals have in their protein toolkit, eg, building blocks even for basic immune defense systems that are successfully used by humans against foreign invaders. The existence of the Rhesus factor26 or the polymorphic immunoglobulin molecules,27 members of the acquired immune system, are mentioned here as examples. Even more, analyses of the formation of the skeletal system of the siliceous sponges, made of amorphous silica (“biosilica”), enabled the discovery of principles underlying tissue regeneration including biomineral formation, which were later found also in higher vertebrates including humans. In particular, it was discovered that biomineral formation is driven by enzymes, just like the biological formation of organic biomolecules and the functioning of metabolic pathways.28,29 Thus, sponges served as a model to progress advances in osteochondral regeneration.30 In addition, due to their excellent mechanical but also light transmitting properties,31 which are particularly evident in the siliceous sponge spicules, sponges can be considered and used as blueprints for the design of functional composite materials not only for nanobiomedical but also for technological applications.

In general, scaffolds formed from natural biopolymers provide an excellent matrix and niche for cell attachment, growth, migration, and differentiation, mimicking the ECM of native tissues (for a recent review, see Ref.32). In these properties, natural biomaterials differ from bioinert synthetic materials used in tissue engineering and regenerative medicine that do not show bioactivity. Only by modifying them according to the design of natural polymers can synthetic bioinspired materials with biological activities be obtained.33 Natural biomimetic biomaterials also have the advantage that they are non-toxic and do not exhibit genotoxic or teratogenic effects. In contrast to the natural biomaterials currently used, which are organic biopolymers such as polysaccharides (eg, hyaluronic acid, alginate, chitosan, and cellulose), polyesters (eg, polylactic acid and polyhydroxyalkanoates), and polypeptides/proteins (eg, collagen, gelatin, and fibroin),32 the two inorganic materials described here, biosilica and polyP, although occurring physiologically, are also readily available by chemical methods, thereby overcoming the problem of batch variability shown by many natural polymers.

Sponges (Porifera): Their Seminal Contributions Towards Understanding Biomineralization

Based on the lessons learned from sponges, the two inorganic polymeric materials amorphous biogenic biosilica (“biosilica”) and inorganic polyphosphate (polyP) have attracted increasing interest in regenerative medicine due to their unique properties. While biosilica forms a sponge skeleton, polyP is a ubiquitous polymer, found from bacteria to men, in sponges particularly in the environment of biosilica. Findings in sponges have been of crucial importance for an understanding of the biological mineralization processes taking place in organisms, as compared to abiotic mineralization.34,35 While abiotic mineralization solely relies on chemical processes, biomineralization is based on chemical reactions that take advantage of genuine biochemical conversion processes. Biochemical reactions are principally catalyzed by enzymes, following the rules of thermodynamics. They often link individual processes together, where endergonic, non-spontaneous processes can be driven by exergonic, energy-releasing reactions.

More specifically, two principles have been revealed in sponges as are outlined in this review: First, that the vast majority of biochemical reactions, including biomineralization reactions, are enzyme driven and dependent. From sponges, the first enzyme that mainly contributes to inorganic mineral formation was discovered: silicatein.28,36,37 While chemical reactions only strive to reach an equilibrium, biological processes proceed in open systems and are attuned to non-equilibria allowing both continuous and discontinuous flux of matter and energy.38

Biosilica, enzymatically synthesized in vivo in sponges (Figure 1), has become a paragon for a mineralic biomaterial. Later, it became increasingly aware that enzymes are also involved in the synthesis and degradation of the inorganic scaffold of other skeletal systems such as bone.39 Of note here is the alkaline phosphatase (ALP), a ubiquitous, membrane-bound tetrameric enzyme that is attached via glycosyl-phosphatidylinositol moieties to the outer cell surface and involved in osteoid formation and mineralization.40 The tartrate-resistant acid phosphatase and cathepsin K, an osteoclast-specific enzyme, are involved in bone resorption.41 Furthermore, it has been suggested that bone formation also involves a carbonic anhydrase that synthesizes amorphous Ca-carbonate bioseeds during the course of Ca-phosphate bone mineral deposition.42

Figure 1 Sponges, the earliest animals on Earth, have proven to be a valuable model system for understanding the basic principles of biomineralization. In these organisms, in siliceous sponges, the first enzyme involved in mineral formation (formation of the sponge skeletal elements, the spicules) was discovered: silicatein. Centers for ATP consumption were found adjacent to the spicules. Subsequently, the polymer inorganic polyphosphate (polyP) was identified as an extracellular storage for metabolic energy, which serves to generate ATP by successive enzymatic degradation via alkaline phosphatase (ALP) and adenylate kinase (ADK). In contrast, mineralization processes require higher temperatures for the chemical reactions instead of enzymes and ATP. The process of biomineralization is facilitated by organic templates/sheets, often collagen, secreted from the cells and undergoing modifications (often glycosylation), allowing the formation of supramolecular assemblies, and finally biomineral deposition.

In contrast to enzymatic, biotic biomineralization, abiotic mineralization, eg, geothermal mineralization, is driven by heat derived from, eg, cosmic rays/ionizing radiation (Figure 1). In addition, it was shown that, in addition to enzymes, bone biomineralization as well as the organization of the bone architecture also require energy in the form of ATP,43,44 without specifying the extent.45,46 In fact, the energy issue was largely ignored until approaches to regenerate or repair tissue through biochemically based substitution therapy were addressed. Both osteoblasts and osteoclasts release ATP into the extracellular space in amounts that depend on the proliferation and differentiation state of the cells.43 Despite a variety of ATP export channels into the extracellular space, the ATP concentration there is very low at ~10 nM, in contrast to the large intracellular pool at ~100 µM (see Ref.5). Since the level of ATP in human blood is also low at ~100 nM, an additional energy source for ATP generation had to be postulated.

Based on the therapeutic success with polyP, after application of this polymer for the repair in different organs, bones, and chronic wounds,30,47–49 it was proposed and then proved that polyP with its energy-rich acid anhydride linkages could serve as a source for metabolically useful energy in the form of ATP and/or ADP,50,51 reviewed in Ref.5 As reported later, the energy stored in polyP can be converted to ADP/ATP in a stepwise enzymatic reaction chain.

In addition to the two corner stones, (i) enzymes and (ii) metabolic energy, organic templates are usually required for the deposition of minerals during biomineralization. In bone, there are the collagen matrices built from collagen bundles that act as a platform for the apposition of mineral aggregates. Their post-translationally modified side chains, processed by hydroxylation and glycosylation, provide suitable deposition sites.52

Sponge Biosilica

The sponges (Porifera) share a common body plan with the other, more evolved metazoans.53 This fundamental finding, based on extensive studies of the (expressed) sponge genome, together with the discovery that these animals synthesize their mineral skeletons with the help of enzymes, as mentioned above,28,36,37 formed the reason why these animals have become metazoic model taxa for the development of biomimetic materials and processes applicable for both nanobiomedical and engineering purposes for the benefit of humans. Equally important was the discovery that biomineral formation is an energy-consuming process that requires an energy (ATP) source.

Based on their skeletons, the Porifera are subdivided into the classes of the siliceous Demospongiae and Hexactinellida and the calcareous sponges, the class Calcarea.54 The siliceous spicules of most Demospongiae remain individualized and the secreted cellular tissue units are formed around them. In these sponges, the spicules are linked together with organic molecules (collagen-related spongin and lectin) that form a bulky extracellular matrix.55 It is the siliceous scaffold that directs the particular, species-characteristic form of the sponge tissue. The siliceous skeleton of the hexactinellids is composed of discrete spicules, which often fuse to a basket-like scaffold as in the hexactinellid Euplectella aspergillum (Figure 2A).56 In particular, the siliceous skeleton of this class of sponges transmits light along the outer surface of the cage (Figure 2B), but also within the inner fan pocket (Figure 2C). Particularly impressive are the central giant spicules of the deep-sea sponge Monorhaphis chuni with a length of up to 3 m (diameter 1.7 cm), which represent the largest biosilica structures on Earth (Figure 2D).57 Light can pass efficiently along these giant spicules (Figure 2E). These spicules are composed of up to 800 lamellae, each 5–10 μm thick, which are arranged concentrically around the axial canal (Figure 2F). In the center, the axial canal harbors a 10 µm thick proteinaceous filament, the axial filament (Figure 2F).

Figure 2 The siliceous sponges and their silica skeletons. (A–C) The hexactinellid sponge E. aspergillum; in (B), the silica cage is illuminated with a laser beam. (D and E) The hexactinellid Monorhaphis chuni with its up to 3 m large giant basal spicule; in (E), the silica rod is illuminated with green laser light. (F) Cross break through a giant basal spicule showing the central axial canal (ac) and the surrounding lamellae (lam). (G and H) A tylostyle spicule harboring (G) the central axial filament (af) of (H) the demosponge S. domuncula. (I) A broken aster spicule from the demosponge G. cydonium exposing the axial filament (af) in the axial canal (ac). (J) Cross break of a tissue unit from S. domuncula showing the immunostained silica-forming enzyme, silicatein, in the axial filament (af) and also on the surface (su) of the tylostyle. (K) Cryosection through a S. domuncula tissue unit; the transition in color from blue to orange reflects the increase in ATP level around the spicules (sp).

Enzymes Involved in Sponge Biomineralization

It is a distinctive feature of both classes of siliceous sponges, the demosponges and the hexactinellids, that their skeletal elements, composed of amorphous biosilica, are synthesized by an enzyme, silicatein. With silicatein, the first enzyme was discovered that catalyzes the formation of an inorganic “polymeric” material, here biosilica, from an inorganic precursor, here monomeric silicic acid.37,58–60

Silicatein is an enzyme protein found exclusively in Porifera. Three isoforms of silicatein are found in the axial filaments.28,37 Molecular sequence analyses revealed that the silicatein family of proteins originates from the cathepsin family of proteases, more precisely from cathepsin L.61 Like the silicateins, the cathepsins are hydrolytic enzymes.62 Since sponges are suspension/filter feeders, our group had proposed and then identified this major catabolic enzyme, cathepsin L.

Among the siliceous sponges, the demosponge Suberites domuncula (Figure 2H) has been used in most cell biological and molecular studies, because this sponge species can be readily kept in aquaria and allowed the establishment of a cell culture, the primmorphs.63 In the demosponges, the axial filament is voluminous (Figure 2G and I). The silicatein filaments determine the morphology of the spicules, either rod-shaped (as in S. domuncula) or star-shaped (as in Geodia cydonium).

Cross fractures through the spicules revealed that not only the central silicatein rod but also the surface of the spicules reacts with antibodies against silicatein (Figure 2J). Interestingly, when cryosections through S. domuncula are studied, a regional distribution of ATP is measured (Figure 2K). The highest ATP levels are found in the vicinity of the spicules. The latter finding reflects the crucial role of ATP in the biomineralization processes and the dependence of the formation of skeletal elements on metabolic activity.64 Biomineralization requires not only an enzyme that lowers the activation energy of this process but also metabolic energy in the form of ATP.

The other class of sponges, the calcareous sponges, Calcarea, are stabilized by calcareous spicules that are formed of calcite.65 Following the general rule that any physiological crystalline structure is built from amorphous precursors, calcite formation starts with amorphous Ca-carbonate (ACC).66 Our group found that the Ca-carbonate-based spicules are synthesized enzymatically by a carbonic anhydrase,29 a finding that has been suggested earlier.67 Also of interest is the finding that the ACC precursor is stabilized by polyP.68 Consequently, incubation of ACC with the polyP-hydrolyzing enzyme ALP causes the transformation of the ACC phase into calcite.69

The Substrate and Mechanism of Silicatein Reaction

Sponges take up silicon from seawater after conversion of SiO2 to soluble, biological biosilica (SiO2•nH2O) and from there to readily soluble H4SiO4 (aq).70 Silicic acid is transported via a Si transporter into the sponge cells,71 where it serves as a substrate for silicatein for the formation of amorphous biosilica. The synthetic silicon compound tetraethyl orthosilicate (TEOS), as a silicic acid precursor, was used for the functional analyses.28,58,72

Silicatein catalyzes the (poly)condensation/polymerization of its physiological substrate, ortho-silicic acid, at low concentrations (<1 mM).59,73 Higher concentrations are required for chemical, non-enzymatic condensation reactions (>1 mM).73–75 In bioinspired fabrication processes, silicatein acts as a hydrolase (silicic acid esterase) that facilitates the hydrolysis of alkoxysilane compounds (cleavage of Si–O–C “ester” bonds), eg, of TEOS.28,37,76

Among demosponges, the first deduced amino acid sequences of silicateins were published for S. domuncula28,77 and T. aurantium.37 The catalytic triad (catalytic center) of silicatein consists of the amino acids Ser, His, and Asn. The cathepsins have a Cys residue instead of Ser. Another characteristic sequence of silicatein is a Ser-rich cluster.

The mechanism of biosilica formation with silicatein has been outlined.78 In the marine environment, the concentration of silicon is around ~10 mg mL‒1. Silicic acid is taken up by cells via a specific transporter where it acts as a substrate for silicatein (Figure 3).71 The reaction starts with nucleophilic attack of the catalytic triad Ser-OH group at a silicic acid substrate molecule, supported by hydrogen bridge formation to the His imidazole group in the catalytic center (Figure 3). The third amino acid of the catalytic triad of silicatein, the Asn residue (not shown), binds to the OH leaving group of the substrate, which is released as a water molecule. This step enables the formed Ser-bound silicic acid, again facilitated by H-bridge formation to His, to undergo two condensation reactions, resulting in the formation of an enzyme-bound disilicic acid and then a trisilicic acid species, which is then released from the enzyme after cyclization as cyclotrisilicic acid (cyclotrisiloxane). Next, a purely chemical condensation process takes place, initiated by preferential addition of further silicic acid species to this cyclic silicon compound. The subsequent steps of silica formation basically follow the Stöber synthesis, allowing a controlled growth of spherical silica particles to condensed silica deposits by hydrolysis of alkyl silicates with subsequent polycondensation of the produced silicic acid units.79

Figure 3 Physiological silicic acid polymerase reaction and bioinspired hydrolase reaction catalyzed by the biosilica-forming enzyme silicatein. Silicatein mediates the condensation of monomeric ortho-silicic acid [Si(OH)4] to polymeric amorphous silica (n•SiO2; formation of Si–O–Si bonds). In a second further reaction, silicatein acts as a hydrolase, which catalyzes the hydrolytic cleavage of the alkoxysilane compound TEOS [Si(OR)4; R = ethyl] (cleavage of Si–O–C bonds) allowing a bioinspired synthesis of artificial amorphous silica-based (nano)materials for biomedical applications. Shown in the center (enclosed in a ring) are the two amino acids of the catalytic triad, Ser and His, which interact with silicic acid and mediate the condensation. (A) Silicatein core protein of an aster from G. cydonium. (B) Complete sponge aster after incubation of silicatein with TEOS. (C) A spicule from S. domuncula with the axial canal (ac) centered by the silicatein-containing axial filament (af). (D) An axial filament with thorny silica (sil) protrusions. (E) Silicatein deposits layered onto thorny template, after reaction with TEOS. (F) Microspheres lacking silicatein after incubation with TEOS. (G) Silicatein-containing microspheres after TEOS incubation.

The biosilica synthesized by silicatein is deposited onto a structured organic template, the enzyme (Figure 3A and B) which also forms a mature siliceous spicule. Consequently, silicatein existing in the axial canal (the axial filament) predetermines the morphology of the spicules (Figure 3C and D). If the spicules are thorned, the surface of the silicatein axial filament shows thorny protrusions (Figure 3D). Using TEOS as a synthetic substrate, silicatein has been used to coat various surfaces/templates (Figure 3E), such as microspheres (Figure 3F and G).

Biosilica: Bioinspired Application as a Morphogenetic Polymer

The field of application of (bio)silica and the bioinspired/biomimetic materials and biomineral structures based on it is wide, ranging from the use in regenerative medicine, improved bacterial fermentation to optical fibers and even as a construction model for architectural buildings.

Biosilica has been considered as a paradigm for biological mineral morphogenesis and evolution not only in sponges80 but also in diatoms.81 In both systems, the morphology of the structures formed is driven by genetically controlled mineralization. The morphogenetic biosilica effect is attributed to its gene-inducing effect.82 In a series of applied experiments, it could be demonstrated that biosilica is a potent material for osteogenic stimulation and differentiation.

Biosilica: A Generic Template for Bone and Tooth Repair

The property of biosilica to be morphogenetically active and supports and likewise accelerates regeneration/repair processes in humans was proven in vitro and in animals (rabbits).80,83 In a first bioinspired approach, the sponge biosilica was applied in in vitro cell systems to evaluate and then to document an increased mineralization potential, first by measuring the expression of respective genes, like the structural proteins within the tooth matrix proteins, amelogenin, ameloblastin, and enamelin, which were strongly upregulated.80 The rationale for this line of study came from Carlisle in her report in 1972.84 Carlisle showed that chicks fed a silicon-rich diet grew and developed substantially faster than control animals. These data were confirmed later85 by demonstrating that the bone strength of broilers is strongly enhanced by dietary supplementation with bioavailable silicon. In continuation, in a series of in vitro studies, it was measured that both biosilica and silicatein induce the growth of mineralizing cells on silicatein/biosilica-coated matrices and cause under these conditions a strong increase in hydroxyapatite (HA) mineral formation;86 in the absence of biosilica, no HA mineral nodules were detected.

Consequently, animal studies were performed using New Zealand White rabbits by inserting 600-µm-large microspheres pressed into pellets (Figure 4A) and implanted into the anterior patella (between the medial and lateral femoral condyles). As an active ingredient, either silicatein (0.8 µg g‒1) or biosilica (100 µg g‒1) was added (Figure 4B). The biosilica pellets, added into PLGA (poly(d,l-lactide-co-glycolide))-based microspheres, were implanted.83 After termination of the study, the implants reacted strongly with OsteoImage (Figure 4C), reflecting the increased regeneration/mineralization activity caused by biosilica.87 The PLGA-based microspheres, supplemented with biosilica, were placed into the drilled space of the patellar grooves of the rabbits (Figure 4D).83 After a total healing period of 100 days, the progression of bone regeneration was inspected by in vivo staining the samples with oxytetracycline dihydrate. The intensity of the blue fluorescence along a bone slide reflects the degree of regeneration.88 The images showed that in the control samples with microspheres, supplemented with β-TCP (β-tricalcium phosphate), only a homogenous tissue is visible around the microspheres that did not brightly stain with oxytetracycline dihydrate and was not flashing under ultraviolet light (Figure 4E). In contrast, in the sections with the implanted biosilica-supplemented spheres, the microspheres are surrounded by bone tissue and this area lights brightly up almost homogenously in blue (Figure 4F).

Figure 4 Morphogenetic activity of biosilica in animal experiments (rabbits) - regeneration of bone in holes drilled into patellar grooves. (A) Microspheres embedded in pellets and (B) biosilica prepared by silicatein for the implant experiments. (C) The microspheres were fabricated together with biosilica. Staining of the mineral nodules formed in rabbits after insertion for 5 days. Staining with OsteoImage (fluorescence flashing). In the (D) femoral implant experiments, new bone formation was detected after in vivo staining with (E and F) oxytetracycline dihydrate under UV light. A striking difference was found between β-TCP controls (E) and biosilica-supplemented microspheres (F). Adapted from Bone, volume 67, Wang SF, Wang XH, Draenert FG, et al. Bioactive and biodegradable silica biomaterial for bone regeneration. 292–304. © 2014, with permissions from Elsevier Inc.83

In general, crystalline biominerals such as bone HA are formed from the amorphous precursors.66 The mechanism underlying the morphogenetic activity of biosilica is not fully understood. The release of orthosilicic acid from amorphous biominerals could contribute to the stimulating effect on the mineralization of bone-forming cells as in bioglass. The mechanism proposed for bioglass involves an exchange of Na+ and Ca2+ with H+ ions, leading to the formation of silanol groups and an increase in the surrounding concentration of OH– ions.89 The cleavage of Si–O–Si bonds by OH– then results in the release of orthosilicic acid and the formation/re-condensation of further silanol groups, which – similar to biosilica – leads to a hydrated, silica-rich layer on the glass surface, which is a suitable matrix for mixed carbonated HA deposition by invading Ca2+, PO43–, OH– and CO32– ions. It is known that low silicate concentrations (0.05–0.5 mM) promote HA nucleation.90 Silica has also been shown to stabilize and prevent the crystallization of ACC, the precursor of amorphous calcium phosphate (ACP), and of crystalline HA.91 In addition, modeling studies showed that the silanol groups of cyclic trisilicic acid motifs on the silica surface stereochemically mimic the Ca2+-binding HA nucleation site on bone sialoprotein (BSP).92–94

These data underscore that biosilica shows morphogenetic activity on bone-forming cells not only in vitro but also in vivo, in animal experiments.95

Inorganic Polyphosphate

The second physiological inorganic polymer, identified in sponges,96 that has attracted increasing attention, is polyphosphate (polyP). Originally, polyP was identified in bacteria and yeast (for a review, see Ref.97), and later in animals, like in sponges96 and higher vertebrates/humans.98 This apparently ubiquitous polymer99,100 exhibits a unique property that no other inorganic material useful for human therapy possesses - the delivery of metabolic energy. PolyP (Figure 5A) has emerged as a prime example of a physiological polymer that not only fulfills the structural but also the energy-supplying requirements for a successful biomedical regeneration process.

Figure 5 (A) Chemical structure of polyphosphate (polyP). (B) Intracellular synthesis of polyP. Glucose is the body’s main source of energy. After cellular uptake, glucose is metabolized to pyruvate during glycolysis, which is then channeled via the voltage-dependent anion channel (VDAC) to the mitochondrial intermembrane space and then to the matrix, where it undergoes oxidative metabolism. The generated reduced coenzymes (NADH and FADH2) drive redox reactions and the electron transport chain builds an electrochemical transmembrane proton gradient whose energy is converted to ATP. ATP is released into the cytoplasm via the adenine nucleotide translocator-2 (ANT2) and VDAC. Subsequently, ATP is channeled through the vacuolar transporter-chaperone complex (VTC) in the acidocalcisomal membrane, which functions in yeast as a polyP polymerase. From the acidocalcisomes in the megakaryocyte-platelets, polyP is released to the extracellular space either as soluble Na-polyP or as membrane-associated Ca-polyP nanoparticles (Ca-polyP-NP). Adapted from Müller WEG, Schröder HC, Wang XH. Inorganic polyphosphates as storage for and generator of metabolic energy in the extracellular matrix. Chem Rev. 2019;119:12337–12374. © 2019 American Chemical Society. Creative Commons.5

Physiologically, polyP is stored intracellularly in organelles, now termed acidocalcisomes, which have been intensively studied in trypanosomatids, protozoan parasites.101,102 Earlier they have been identified as metachromatic granules103 or volutin granules.104 In the blood platelets, polyP is accumulated in the dense granules, as identified by Ruiz et al.105 The discovery that ATP and polyP harbor metabolic energy came from Meyerhof and Lohmann (cited in Ref.106). The unequivocal identification of the chemical structure was described by Lohmann and Langen.107,108

Recently, as described later, the proof-of-concept of the therapeutic benefits in the clinic for polyP could even be successfully provided. As a hydrogel, the natural polymer not only provides a platform for cell proliferation, cell differentiation, and cell migration but also provides the metabolic energy required for maintaining the molecular and supramolecular organization of the extracellular matrix and cell function.48,49

Polyphosphate: Cell-Based Synthesis

The reaction chain by which polyP is synthesized in bacterial cells is fairly well understood (reviewed in Ref.109). Less is known about polyP synthesis in higher vertebrate cells.98,110 For the synthesis of polyP in mammals, ATP is required, which is generated intracellularly in the mitochondria (Figure 5B).111–113 There, in the respiratory chain, ATP formation is linked with complex V, the F1Fo-ATPase. For the chemical synthesis of the phosphoanhydride bonds in ATP, heating to >100°C for ~10 h is required.114 Accordingly, for an intracellular, physiological synthesis of ATP at 37°C, with its high-energy anhydride bonds, an activation energy of 110 kJ mol‒1 has to be expended, as determined using Arrhenius plot.115 Therefore, the reactions in the mitochondria must be mediated by enzymes. It is the ATP synthase, localized in the inner mitochondrial membrane, which catalyzes the synthesis of ATP from ADP. In turn, the biosynthetic pathway for the formation of polyP with its >30 high-energy anhydride bonds must involve enzymes as well. In mammalian systems, the genuine polyP-synthesizing enzyme has not yet been discovered. Experimental evidence in yeast suggests that polyP is formed enzymatically from ATP during import into the acidocalcisomes (Figure 5B).116,117 Indicative is the fact that, in the dense granules of the platelets that correspond to the yeast acidocalcisomes, the concentrations of ADP (600 mM), ATP (400 mM), and pyrophosphate (300 mM) are exceedingly high,118 while the polyP content is comparably low (130 mM; based on Pi).119 Therefore, it might be assumed that phosphatases or phosphotransferases present in platelets120,121 could be involved in polyP synthesis through backward reactions from ATP, driven by these enzymes. From the platelets, polyP is released in two forms, either in a soluble form, as a chain with Na+ as counterion, or in an “insoluble” form, as a NP with Ca2+ as counterion (Figure 5B).118,122 Since polyP compartmentalized into the acidic dense granules (pH 5.4) is released in a controlled manner,123,124 it is likely that the Ca2+ gradient (Ca2+ concentration in the dense granules is 2.2 M)119 determines the formation of the two forms of polyP as proposed.118

After full platelet activation, the concentration of polyP in the blood is relatively high with 0.5 to 3 μg mL‒1.118,125 There, the polymer has a physiological chain length of ~50 phosphate (Pi) units. In the particulate Ca2+ form, the polyP chains are longer with ~250 Pi units. However, this value is very variable since there are high levels of ALP in the blood, which hydrolyzes the polymer from the chain termini as an exopolyphosphatase.126

A biocompatible polymer can be particularly beneficial to patients when transported within the body to the target site for tissue repair. Here, too, polyP follows the route of a biomimetic medical compound. In vivo, under physiological conditions, polyP is efficiently distributed to the injured sites in the body with the blood circulation. There, the polymer is delivered by the blood platelets to the damaged tissue regions, where the polymer initiates the regeneration process.127 In fact, platelets with the stored polyP are hallmarks of regeneration.128 It was Julius Bizzozero who discovered platelets as small medicinal pellets of a size between 1 and 2 µm.129 He described them as the initiators of blood clotting. PolyP is a major constituent of platelets.105 Using these vehicles, polyP is distributed throughout the body in order to onset regeneration/repair. The distribution of polyP is flanked by macrophages, which bind or internalize polyP, independently of their signaling roles, via their P2Y1 and RAGE receptors.130 Besides acting on the clotting cascade and enhancing hemostasis, platelets are activated and release polyP into the extracellular space when certain growth factors, such as epidermal growth factor or platelet-derived growth factor, are released.122 In this environment, platelets bind directly to exposed collagen fibers, as well as to von Willebrand factor, fibronectin, and other adhesive proteins.131 This efficient distribution mechanism of polyP in the body contributes to the prominent position that polyP has achieved in regenerative medicine, as described later.

Polyphosphate: Chemical Synthesis

Chemically, polyP can be prepared in sufficiently large quantities.132 Na-polyP with its energy-rich phosphoanhydride bonds (Figure 6A) is obtained by melting of NaH2PO4 at temperatures up to 700°C (Figure 6B). For the fabrication of Ca-polyP, both an enzymatic/wet chemical approach133 and a calcination process have been introduced.134 By shortening the high-temperature processing protocol, it is possible to prepare both amorphous Na-polyP and Ca-polyP-NP almost in parallel (Figure 6B and C). The process runs at 700°C. At the end of the polycondensation reaction, polyP, with a chain length of ~50–100 Pi units, is supplemented with CaCl2 and heated for an additional period of time to obtain the amorphous Ca-polyP-NP with a diameter of ~100 nm (Figure 6D). When Na-polyP is synthesized, the material is ground to 50 µm powder (Figure 6B). Different steps of the procedure, starting from NaH2PO4, are summarized in Figure 6E-1–E-4.

Figure 6 Chemical preparation of polyP. PolyP is prepared by heating the starting material NaH2PO4 to 700°C. Both amorphous (A and B) Na-polyP, and (C and D) Ca-polyP-NP can be prepared in parallel. (A) Na-polyP is formed during heating to 700°C. The glass-like melt is (B) ground to Na-polyP powder. Ca-polyP-NP is prepared from Na-polyP by addition of CaCl2 (C and D). By this, polyP of a physiological chain length of ~50–100 Pi units (Na-polyP50–100) is obtained. (E-1–E-4) Preparation process of Na-polyP and Ca-polyP-NP. (E-1) Start of preparation; (E-2) sliding in the furnace; (E-3) The molten glass poured from the crucible onto (E-4) a steel plate (images with infrared camera).

Polyphosphate: A Biomimetic Molecule for Human Therapy

PolyP as a polyanion can be present in tissues as salt with various cations as counterions. The biological regeneration function of polyP differs depending on the counterion chosen. In addition, the functional activity of polyP depends on the form of the polyP salts, which can exist both in a soluble form, such as the sodium salt (Na-polyP), and in an insoluble nanoparticulate form that can be used as a storage (depot) form, such as salts with various divalent cations, eg, Ca2+ ions. The latter salts can also form a gel-like coacervate phase, a physiologically active form, as outlined later.

Cation-Specificity of Polyphosphate: A Smart Nano/Micro Biomaterial

Besides the thermal method described above, polyP can be fabricated in a biomimetic way as nanoparticles (NP) as in vivo in human cells using a wet chemical procedure.118,135 The biological activity of the polyP nanoparticles is determined, at least in part, by the counter-cations used for a particular scaffold (Figure 7). For cartilage repair, Ca2+- and also Mg2+-polyP complexes are more suitable,136 while for bone regeneration, the Ca-polyP forms might be favored due to their higher stability. In addition, the Sr2+-polyP complex shows strong regenerative activity both in vitro and in animal experiments.137 Drugs that are effective, eg, against bone tumors/metastases, such as the bisphosphonate zoledronic acid, can be conveniently integrated into the Ca-polyP-NP and used as implant particles.138 For another application target, chronic wound healing,48,49 a combination of Na-polyP and Ca-polyP-NP acts most efficiently on the regeneration process. In addition, polyP particles, when applied in an aerosol form, have a protective effect on the respiratory epithelium, after conversion into the coacervate form upon contact with mucin, the protein of the airway mucus.139 As a result, viral particles such as SARS-CoV-2 are entrapped and inactivated, in addition to a direct inhibitory effect on this virus.140,141

Figure 7 PolyP is a genuine, smart nano/micro biomaterial whose properties and applications depend on the selected counterion. PolyP-NP, polyphosphate nanoparticle. Reproduced with permission from Wang XH, Schröder HC, Müller WEG. Amorphous polyphosphate, a smart bioinspired nano-/bio-material for bone and cartilage regeneration: Towards a new paradigm in tissue engineering. J Mat Chem B. 2018;6:2385–2412. © 2018 The Royal Society of Chemistry. Creative Commons.142

Polyphosphate Coacervate Formation

It is an exceptional feature of polyP (as Na-polyP) to undergo coacervation in the presence of Ca2+ or other divalent cations, at physiological pH. In principle, polyP nanoparticles, including non-processed Ca-polyP-NP, are biologically inert. They have to be converted into a biologized form, a polyP-coacervate. During this phase transition, polyP provides its physiological functions. At pH 7, Ca2+ together with polyP forms a viscous aqueous phase, the coacervate (Figure 8A and C). These aggregates are formed through liquid–liquid phase separation, resulting in a denser phase and a dilute phase that are in thermodynamic equilibrium.143 In the coacervate phase, polyP shows morphogenetic activity as well as its function to generate ATP.144 In our approach, the two phases and their interconversion were studied using in silico simulation studies, as well as (physico)chemical analyses. The data showed that Ca-polyP coacervate formation occurs at pH 7 and is slower, compared to Ca-polyP-NP formation at pH 10 (Figure 8E and F). Interestingly, if CaCl2 is dropped into a Na-polyP solution at pH 7, the Ca-polyP coacervate initially forms and becomes then converted to NP at pH 10. Conversely, when Na-polyP is added to a CaCl2 solution, only the coacervate phase is obtained at both pH values.144

Figure 8 PolyP / Ca-polyP phases. (A–D) At pH 7, the polyP coacervate is formed from Ca2+ and Na-polyP. During this process, the aqueous liquid-liquid gelatinous phase envelops bacteria such as (B) E. coli and/or (D) attracts mesenchymal stem cells (MSC) and allows them to nest and differentiate there. (E and F) NP formed from Ca2+ and Na-polyP at pH 10.

The coacervate is biocompatible and allows the infiltration of mesenchymal stem cells (MSC) into the gel matrix, where polyP promotes cell proliferation and differentiation;143 the MSC become completely embedded in the matrix (Figure 8D). When the coacervate (Ca2+-polyP) forms during the liquid–liquid phase separation, different local densities arise causing turbulences during which bacteria, such as E. coli, are enwrapped (Figure 8B) and then killed.49

PolyP particles are taken up by endocytosis, as determined by inhibition studies with trifluoperazine dihydrochloride.51 Intracellularly, the particles begin to transform into a coacervate. The free polyanionic polyP (not in a salt form) is not able to traverse cell membranes. Only after caging the polymer, eg, into a guanidinium/oligocarbonate vehicle, polyP can be channeled into the cells.145

Supramolecular Extracellular Matrix Organization: Importance of Polyphosphate Coacervation and Energy Requirement

Biomimetic materials used for soft or hard tissue regeneration can be classified according to the four tissue categories: covering tissue, connective/and supporting tissue, muscle tissue, and nervous tissue.146 The ECM of the different organs has a different percentage of cells, such as the liver with a high (>90%) and cartilage with a low cell density (<10%). A common feature of their ECMes is that the cells are embedded into a fibrillar, often collagenous, hydrogel, which allows architectural stability and simultaneous diffusion of nutrients as well as cell movement. To ensure biocompatibility, a physiological or close-to-physiological hydrogel is preferably used in regenerative medicine. Surely, such hydrogels can elicit signals stimulating tissue regeneration. Normally, hydrogels lack the ability to directly supply the cells with metabolic energy, preferably in the form of ATP.5

Supramolecular hydrogels have been introduced into regenerative medicine because of their unique dynamic properties for self-healing and injectability based on the non-covalent crosslinking organization of their macromolecules. Hydrogels based on this fabrication technology are held together by forces arising from non-covalent bonds such as electrostatic interactions, hydrogen bonds, metal coordination, aromatic stacking, and hydrophobic and van der Waals forces (Figure 9A).147,148 The individual molecules are synthesized by covalent bond formation (“Molecular chemistry”) and become subsequently organized i