key: cord-0826582-6s2n5apk
authors: E. G. Müller, Werner; Schröder, Heinz C.; Neufurth, Meik; Wang, Xiaohong
title: An unexpected biomaterial against SARS-CoV-2: Bio-polyphosphate blocks binding of the viral spike to the cell receptor
date: 2021-08-02
journal: Mater Today (Kidlington)
DOI: 10.1016/j.mattod.2021.07.029
sha: b0809003da95acb72bea424fac61f34a68bfd7df
doc_id: 826582
cord_uid: 6s2n5apk

No other virus after the outbreak of the influenza pandemic of 1918 affected the world's population as hard as the coronavirus SARS-CoV-2. The identification of effective agents/materials to prevent or treat COVID-19 caused by SARS-CoV-2 is an urgent global need. This review aims to survey novel strategies based on inorganic polyphosphate (polyP), a biologically formed but also synthetically available polyanionic polymeric material, which has the potential of being a potent inhibitor of the SARS-CoV-2 virus-cell-docking machinery. This virus attaches to the host cell surface receptor ACE2 with its receptor binding domain (RBD), which is present at the tips of the viral envelope spike proteins. On the surface of the RBD an unusually conserved cationic groove is exposed, which is composed of basic amino acids (Arg, Lys, and His). This pattern of cationic amino acids, the cationic groove, matches spatially with the anionic polymeric material, with polyP, allowing an electrostatic interaction. In consequence, the interaction between the RBD and ACE2 is potently blocked. PolyP is a physiological inorganic polymer, synthesized by cells and especially enriched in the blood platelets, which releases metabolically useful energy through enzymatic degradation and coupled ADP/ATP formation. In addition, this material upregulates the steady-state-expression of the mucin genes in the epithelial cells. We propose that polyP, with its two antiviral properties (blocking the binding of the virus to the cells and reinforcing the defense barrier against infiltration of the virus) has the potential to be a novel protective/therapeutic anti-COVID-19 agent.

forces between two partially ionized species with opposite charges. Those electrostatic interactions, like between (poly)arginine and (poly)phosphate, can be strong and are sometimes even attributed as "covalent-like" [13] . Polyanions act as receptors for many human viruses [14] . Prominent polycations are the polyamines, which interact with negatively charged nucleic acids, proteins and phospholipids via ionic and/or hydrogen bonds [15] .

In medicine, polyelectrolytes have so far only been used on a large scale as soft particles or as coatings for complex, oppositely charged polyelectrolytes or counterions. In contrast, most pharmaceuticals are not polymers but low-molecularweight natural products, or derivatives of them, with arabinofuranosyladenine or arabinofuranosylcytosine, which inhibit herpes simplex virus outbreak on lips or are used for a treatment of leukemia, as examples [16, 17, 18] . These compounds inhibit either competitively or non-competitively enzyme activities through direct or indirect binding and inhibition as ligand-like compounds at their active sites ( Fig. 1 (a,b) ). More recently, polymeric polyelectrolytes have been introduced into the clinics because of their therapeutic benefits, like the polyanion heparin, which controls coagulation [19] . Those macromolecules inhibit enzyme reactions or interfere with ligand-receptor interactions and thereby prevent receptor dimerization and the subsequent signaling processes through membranes ( Fig. 1 (c) ).

Some polyelectrolytes block binding of viruses to the target cells like sulphoevernan, a polyanionic polysaccharide that inhibits binding of the human immunodeficiency virus-1 (HIV-1) to the host cell [20] . Viruses are promising targets for the development of polyelectrolyte drugs, as most of them are constructed of nucleocapsid structures (enclosing the genetic material of the virus) that carry clusters of positively charged molecules [21] . Most of these charges are contributed by the amino acids (aa) Arg and Lys. Interestingly, many surface (glyco)proteins comprise negatively charged clusters, mostly based on the presence of sialic acid residues which are usually terminally located on the glycan side chains.

In the present review it is highlighted that polyP, a highly anionic inorganic polymeric material is formed intracellularly and from there enters the extracellular space. Via the cardiovascular system, polyP, either packaged in platelets or released from the cells after activation, reaches the tissues and organs. In turn, the polymeric material has access to and can bind to virus particles and to cells harboring or releasing virus particles. In this review, we will focus on SARS-CoV-2, which is neutralized/blocked by polyP. This virus particle carries on the distal tips of its proteinaceous spikes a polycationic groove, which attracts polyP and allows this biomaterial to bind. In addition, polyP has the property to act as a polymeric material during the innate immune defense.

Linear oligomers/polymers of polyP composed of phosphate (Pi) residues with tens to hundreds of high energetic phosphoanhydride bonds are predominantly present in nature ( Fig. 2(a) ), while the other two forms of polyP polyanions, the cyclic polyP (metaphosphates) and the cross-linked polyP (branched inorganic phosphates) are rarely found, only in marine algae ( Fig. 2(b and c) ) [22] . Polyphosphodiesters (Fig. 2(d) ) are polyP-like polyanionic polymers, but without phosphoanhydride bonds, which have a structure similar to the backbone of nucleic acids and are highly water-soluble, and biodegradable and are biocompatible. PolyP and its different forms can also be prepared chemically by condensation of orthophosphate usually at high temperatures. Bio-polyP is a ubiquitously present polymer that provides a flexible polyanionic scaffold, which facilitates the assembly of other macromolecules.

The substrate for the enzymatic synthesis of polyP, ATP, was discovered by Lohmann [23] . The first clue that polyP is an energy-rich phosphate was provided by Ebel [24] who disclosed the structural relationship between ATP and polyP. The evidence that polyP plays significant functions in metabolism was elaborated in yeasts [25] . Especially driven by studies of Kulaev [26] , Sylvia Ruth Kornberg [27] and her later husband Arthur Kornberg [28] polyP was given a prominent place in biochemistry. The synthesis of polyP was co-discovered with the one of DNA [28] .

The megakaryocytes and the platelets, which originate from them, are the dominant cells containing polyP ( Fig. 3 (a) ). Two intracellular organelles are rich in polyP, the mitochondria and the acidocalcisomes [1, 29, 30] . These two organelles are in close contact with each other, suggesting a functional interaction ( Fig. 3 (b) ). Fatty acids, amino acids and pyruvate are taken up by the mitochondria via carriers, e.g. carnitine:acyl carnitine antiporter, or voltage-dependent anion channels/porins. There, in the citric acid cycle (tricarboxylic acid cycle or also termed Krebs cycle) the metabolites undergo end-oxidation under the release of reducing equivalents, NADH and FADH2, and drive the electron transport chain. During these exergonic processes an electrochemical proton gradient is built up that energetically allows the synthesis of ATP. It has been reported that polyP also exists in mitochondria where it might function as an energy reservoir [31] [32] [33] [34] [35] . An inhibition of the electron transport chain at the level of complex I by rotenone causes a drop of the polyP pool [36] . This reaction is paralleled with an effect of polyP on the membrane potential (ΔμH)/proton gradient (ΔpH + ). On the inner mitochondrial membrane, electrons are passed along the electron transport chain and there the protons are pumped out in the intermembrane space. Thereby, a gradient is formed that drives the protons back through this membrane and fuels ATP synthase. This enzyme synthesizes ATP from ADP and Pi. Subsequently, the adenine nucleotide translocator/carrier (ANT), also termed ADP/ATP translocase, the most abundant protein in the inner mitochondrial membrane exchanges the formed free ATP by free ADP through the membrane [37] . Then the ATP is released into the cytosol through the voltage-dependent anion channels (VDAC), which are abundant in the outer mitochondrial membrane ( Fig. 3(b) ) [38] . The VDAC becomes activated during depolarizing potentials.

It appears to be indicative that near the mitochondria the acidic acidocalcisomes are located, which are the main organelles for storage, and perhaps synthesis, of polyP. These 50−300 nm sized cell organelles are ubiquitously present from bacteria to animal cells. They are surrounded by an 8 nm thick membrane and, besides of polyP [150 mM] , store large amounts of Ca 2+ [2 M] and ATP [400 mM] [30] . A series of anabolic and catabolic enzymes are present in the acidocalcisomes that are involved in phosphate/polyP metabolism. Among them is the vacuolar transporter chaperone (Vtc) complex, as well as a putative phosphate transporter and an acid phosphatase [39] . Even more, in yeast, the Vtc complex is the endogenous vacuolar polyP polymerase that efficiently imports polyP from the cytosol into the acidocalcisomes and, by that, prevents the accumulation of toxic polyP intermediates in the cytosol [40] . The Vtc exposes its enzymatically active side towards the cytosol. Its subunits Vtc4/Vtc3-Vtc2/Vtc1 are the sites of polyP metabolism, more specific, subunit Vtc4/5 catalyzes the synthesis of polyP (the polyP polymerase) and extrudes the polymer into the acidocalcisomes (Fig. 3 (b) ). So far it is only possible to speculate, but it has not yet been clearly shown in biochemical studies, whether these results on the polyP polymerase, obtained with budding yeast also apply to mammals [41; 42] .

Besides of the platelets (thrombocytes), the mast cells are filled with polyP chains which are encapsulated into acidocalcisomes as well [43] . These cells have a longer half-life with ∼40 d [44] , while the half-life of platelets is only 8 to 9 d. The precursor cells of the platelets are the megakaryocytes (50-100 μm in diameter). It needs to be elaborated if in these precursor cells the polyP polymers are synthesized [45] . About 2000-5000 new platelets are formed from one megakaryocyte ( Fig. 3 (a) ) [46] . In the platelets, two pools of polyP are present. First, a polyP fraction with a chain length of 60-90 Pi units, and second, a pool with polymers of a longer average chain length of 200 to 1000 Pi units [47, 48] . Importantly, short-chain polyP is released in a soluble form, while the long-chain polyP molecules become encapsulated into 100 to 200 nm large particles [49] . Interestingly, only the long-chain polyP fraction binds to the platelets, while short-chain polyP does not and exists in the soluble phase [50] .

It is functionally important that the polyP released from the platelets into the circulating blood has a size of <100 Pi units [47, 48] . Until now it is not yet finally disclosed if those short-chain polyP molecule modulate the blood clotting kinetics. Some studies demonstrate that this size-range fraction of polyP elicits no effect [51] , others showed that this polymer is a potent mediator of contact pathway activation reactions during blood clotting [52] . It is likely but not definitively proven that polyP undergoes enzymatic hydrolysis in the circulating blood by the alkaline phosphatase (ALP) [29, 53] .

Until now, a mammalian enzyme that exclusively hydrolyses the phosphoanhydride bonds of the linear polyP molecule, a polyP-specific polyphosphatase, has not been discovered. In addition to the polyP-degrading (but non-specific) ALP, the insect and vertebrate protein h-prune, a short-chain polyP exopolyphosphatase, and the diadenosine and diphosphoinositol phosphohydrolase, an endopolyphosphatase, also hydrolyze polyP [54, 55] .

The polyP polymers are flexible molecules with a bond angle of 130° at the P-O-P linkage and of 102° at the O-P-O atoms ( Fig. 4 (a) ). The P-O bond length at the oxygen bridges of polyP is 1.62 -1.66 Å [56, 57] . In aqueous solution, the polyanionic polyP chain takes an extended conformation because of the Coulomb repulsion of the negatively charged Pi groups [58] . The orientation of the linked PO4 tetrahedra adapts to either an eclipsed or a staggered conformation ( Fig. 4 (b and c)) [59, 60] . In the Ca 2+ salt form, the polyP chain adopts an eclipsed conformation ( Fig. 4 (b) ), while the monovalent Na + as counterion favors the staggered conformation of polyP ( Fig. 4 (c)) [61] . This transition might have a functional importance. This differential salt formation of polyP provides the polymer with a type of "induced fit" potential. With Na + as a counterion, the P=O π bonds arranged along the polyP chain ( Fig. 4 (c) ) have a longer spacing (≈4.5 Å) compared with the exposing P-O bonds in the salt with Ca 2+ with ≈2.5 Å (Fig. 4 (b) ) [62] . These values imply that for an association of polyP with peptide stretches (spacing of consecutive peptide bonds of around ≈4.5 Å) a staggered conformation with Na + will most likely match better than the eclipsed conformation.

Furthermore, mitochondria have a comparable low concentration of Ca 2+ with 0.2 to 20 µM in comparison to the acidocalcisomes with 2 M [1], the concentration of Na + in the cytoplasm varies between 50 and 60 mM [63] . These concentration differences most likely dictate the type salt formation between the anionic polymer and the respective cations. In the mitochondria and the cytoplasm the polymer should be present in a soluble form with the counterion Na + , while in the acidocalcisomes polyP exists as Ca 2+ salt nanoparticles [NP] [48] . In a biomimetic approach using the particles in the in the acidocalcisomes as a template [48] , amorphous polyP nanoparticles have been prepared by our group, applying a super-stoichiometric ratio of Ca 2+ to phosphate, in an alkaline environment at a pH of 10 [64] . The size of the mesoporous nanoparticles formed varies around 200-500 nm ( Fig. 4 (d-i)); they display morphogenetic activity (reviewed in: [5, 65] ). A similar procedure with Mg 2+ and polyP likewise resulted in the synthesis of amorphous nanoparticles with a slightly smaller size of ∼100 nm (Fig. 4 (d-ii)) [66] .

The process of coacervation proceeds in an aqueous phase with macromolecules, like polyP, and results first in a formation of a dense phase that is in thermodynamic equilibrium with a dilute phase. This kind of "lyophilic colloid" material forms spontaneously in water at neutral pH and constructs stable membrane-like structures. In the salt form with the cation Ca 2+ , polyP forms a coacervate phase at a close to neutral pH, which is most likely the physiological active form of the polymer [67] . Earlier studies have shown that nanoparticles tightly associate with biomolecules under formation of (hard) protein corona particles with modified and changed biomedical properties [68] . By this, the beneficial effect of some nanoparticles become confined [69] .

The polyP coacervation is prevented if, like in the experiments with Ca-polyP nanoparticles, their zeta (ζ) potential increases and in turn interferes with the fusion process [70] . An increase in the negativity of the ζ potential is favorable for implant materials [70] but not for a material with an intimate contact to the target cells in which polyP should elicit its functional activity. The ζ potential of the Ca-polyP nanoparticles is in the range of -42.3 mV for particles with a size between 307 and 859 nm and -33.6 mV for particles of 61-198 nm. In turn, the nanoparticles are stable, also in culture medium, for about 10 h ( Fig. 5 (a) ) [67] . However, since the ζ potential is prone to ionic polyelectrolytes or macromolecules like proteins, adsorption of these molecules to the particles reduces the ζ potential, followed by a concomitant shift of the slipping plane from the solid surface of colloidal particles [72] . The reduction of the ζ potential is paralleled with an increase of biocompatibility and a faster release of morphogenetic signals to the cellular environment [67, 71] . After transfer of the Ca-polyP nanoparticles (oreven more efficient -Mg-polyP nanoparticles) into cell culture medium supplemented with fetal calf serum, the particles attract protein(s)/peptides out of the aqueous environment and integrate them into the dynamic coacervate ( Fig. 5 (b)) [67] . A scanning electron microscopical analysis of the transformation of nanoparticles to a coacervate is shown in Fig. 5 (b-i to iv).

Within the polyP polymer chain the Pi units are linked via energy-rich acid anhydride bonds to non-branched chains [1, 8, 73] . In any biotic environment two enzymes are present that process polyP and its reaction components. At first, polyP is hydrolyzed from the ends of the molecule under release of Pi and polyPn-1 by ALP ( Fig. 6 (a and c)) [74, 75] . The degradation follows a processive mechanism (the enzyme remains attached to the polyP substrate until complete degradation before it dissociates from the substrate). Because of the higher affinity of ALP to the longer polyP molecules (lower Km value) those chains are degraded first before hydrolysis of the shorter molecules can start [74] . The ALPs are promiscuous enzymes [76] since besides of polyP, they hydrolyze ATP, ADP, AMP, PPi, glucose-1-phosphate, glucose-6phosphate, fructose-6-phosphate, β-glycerophosphate and others, at a pH between pH 7.5 and pH 9.4. During the ALP driven hydrolysis, energy (ΔG 0 ) is released that can be at least partially transformed to biochemically utilizable metabolic energy [77] .

Experimental data show that this energy is used for the formation of an energy-rich phosphoanhydride bond in ADP, via phosphoryl transfer to AMP ( Fig. 6 (b) ) [78] [79] [80] . During this reaction the Pi units released from polyP are linked in the transition state, as metaphosphate, to an intermediate compound comprising an acid polarized carbon atom, similar to the one which exists in the mesomeric guanidino group in the protonated Arg ( Fig. 6 (b and d) ) [1, 81, 82] . In a subsequent step, the intermediate Pi/metaphosphate unit is transferred to AMP under the formation of ADP. In continuation, ADP is then up-phosphorylated to ATP under the consumption of two moles ADP and the release of one mole AMP via the adenylate kinase (ADK), likewise a ubiquitously distributed extracellularly localized (cell membrane bound) enzyme ( Fig.  6 (b)) [83] .

The generation of ATP, after exposure of polyP to cells, is substantiated by cell culture studies [79, 80] . Addition of the ALP inhibitor levamisole to the system blocks the synthesis of ADP from AMP [84] . Additional exposure with Ap5A, a dinucleoside polyphosphate which is an inhibitor of ADK, abolishes the ATP generation, also in the extracellular space [80, 83] . Therefore, we propose a coupled ALP-ADK reaction as the metabolic chain leading to the generation of extracellular ATP from polyP.

The viruses have developed intricate systems to enter the host cell and to usurp the host cellular replication machinery. As obligate intracellular parasites, all these processes are prerequisites for them to propagate and to produce progeny viruses. If the virus succeeds in penetrating the host cell, serious diseases, like in COVID-19 (coronavirus disease 2019), or chronic diseases, such as AIDS or viral hepatitis, can emerge. An inhibition of binding of SARS-CoV-2 to the respective host cells is one key for a prevention of viral infection. In the following it is outlined that polyP is an effective biomaterial which interacts with the viral binding protein and by this blocks its attachment to the cell surface.

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the etiologic agent of COVID-19 [85] . Already before the outbreak in Wuhan, China, and the subsequent spreading as a pandemic, COVID-19-related infectious disease emerged with the examples of the severe acute respiratory syndrome (SARS), first emerged in 2002 in Guangdong, China, and the Middle East respiratory syndrome (MERS) in 2012 in Jeddah, Saudi Arabia. These viruses are grouped together to the Coronaviridae, a family of enveloped, positive-strand RNA viruses, which are infectious for amphibians, birds, and mammals [86] . The Coronaviridae store their genetic information in as much as 25 to 32 kb [kilobases] of RNA and therefore represent the viruses with the largest known RNA. In comparison, the RNA virus HIV-1 has a small genome size with less than 10 kb [87, 88] . Both viruses even amplify their genetic information by transcription of more than one open reading frame of the primary RNA information carrier [89, 90] .

Coronaviridae have a globular morphology with a diameter of 118 to 136 nm and 9 to 12 nm large surface projections formed by their spike (S) glycoproteins that extend from the virus envelope [91] . Members of the Coronaviridae with the subfamily of Coronavirinae are widespread in animals and cause only mild respiratory or enteric infections, with the exception of SARS-CoV-2.

SARS-CoV-2 is transmitted via an airborne route. In consequence, social distancing and mandated face covering are strongly advised as crucial and effective protections to fight the pandemic [92] . This precaution will control and reduce nebulization of viruscontaining particles as a result of coughing/sneezing, and less frequently by normal breathing, talking and singing of infected individuals. The vehicle of virus transmission are aerosol droplets with a diameter of ∼1-5 μm. They will be exhaled via the nose and mouth and enter the lung after passing the trachea. In the respiratory tract, they come into contact with the mucus and eventually penetrate the periciliary layer covering the airway epithelia and ultimately initiate infection of the epithelium with its goblet cells and ciliated cells [93] . Patients may develop pneumonia and severe dyspnea that require intensive care. Accordingly, an acute respiratory syndrome is considered as the hallmark of severe COVID-19 even though accumulating evidence suggests that SARS-CoV-2 also attacks other organs and various body systems [94] . The polymer polyP, which is discussed later, is expected to prevent attachment in the nasopharyngeal and oropharyngeal airways system. It has been described that thrombotic processes, occurring during the progression of COVID-19, cause coagulopathies, unexpected clotting occurring in veins. Frequently, the clinical picture of COVID-19 is linked with an abnormal low level of blood platelets, with thrombocytopenia [95] . A low platelet count parallels with an increased mortality of patients [96] . This syndrome is characterized by a decrease of cytokines/chemokines and also of polyP, which is primarily stored in the platelets [29] . A low level of polyP is also found in delta-storage-pool-disease, a thrombocytopenia, which is characterized not only by a deficiency in the level of polyP but also a pathological modulation of the processes of coagulation and fibrinolysis and a diathesis for bleeding [97] .

In this review, we focus on the initial phase of the viral replication cycle because of the following rationale. It appears to be evident that an initial inhibition of the attachment of the virus to the corresponding cellular receptors, e.g. by physiological binding molecules/materials like polyanionic polymers, or by components of the immune system, will neutralize the infectious agent [98, 99] . Components both of the innate immune system (e.g., complement system) and the adaptive immune response (antibodies) appear to be promising candidates. Here we highlight that the polyanionic polymer polyP masks the most distal part of the viral spike protein with its cationic domain and blocks the interaction between the viral spike protein and the host cell receptor.

Like with any virus the SARS-CoV-2 life cycle is divided into three stages; entry (attachment, fusion, uncoating), genome replication (replication and protein translation), and exit (assembly, maturation, and release). This virus utilizes two kinds of molecules on the cell surface to attach and bind to the host cell. At first, the prominent protein at the outer surface of the lipid envelope that surrounds the virus particle, the spike S-protein, attaches to the corresponding host cell receptor, the angiotensin-converting enzyme 2 (ACE2) [100] . This initial interaction of SARS-CoV-2 with the cell is flanked or controlled by additional perhaps less specific cell surface molecules, like glycosaminoglycans-heparins that interact with the two primary components (S-protein and ACE2) and might link them together [101] .

Subsequently, more downstream after binding of the viral S-protein with its exposed receptor binding domain (RBD) to ACE2, enzymes are involved that process the Sprotein and, by this, prepare the virus for the fusion process with the host cell. These cellular enzymes are furin, transmembrane protein serine protease 2 (TMPRSS2), and cathepsin L; then, later, further enzymes act, which complete the viral replication cycle, the RNA-dependent RNA polymerase, helicase, exonuclease, papain-like protease, and 3C-like protease [102] . Inhibitors of these enzymes will block the viral reproduction cycle directly [103] [104] [105] [106] .

The SARS-CoV-2 spike S-protein is a glycoprotein composed of three monomeric subunits with a size of 180-200 kDa each. They comprise a large extracellular Nterminal segment, a transmembrane domain, and a shorter intracellular C-terminal segment [107, 108] . Each monomer of this homotrimer consists of two subunits, the S1 subunit which carries the RBD responsible for the cell attachment [109, 110] . The S2 subunit which contains the fusion peptide and the two HR [hydrophobic repeat regions] domains (HR1 and HR2) mediates the fusion of the viral envelope with the host cell membrane [108, [111] [112] [113] . The polysaccharide side chains of the S-protein facilitate the passage through the mucus on the surface of the epithelia (Fig. 7 (a) ). Prior to binding to the cellular receptor ACE2, the RBD has to undergo a conformational change from a closed, "down" state to an opened, "up" conformation which is needed for receptor binding (Fig. 7 (b) ). The cellular receptor used by SARS-CoV-2, the ACE2 protein, is a zinc-containing metalloenzyme (a carboxypeptidase) that converts angiotensin I to angiotensin 1-9, and angiotensin II to angiotensin 1-7. In contrast to its homolog ACE, which forms angiotensin II [114] , ACE2 is a heterodimeric protein, which is primarily expressed in cells of the respiratory tract, in particular of the alveolar epithelial type II. In addition, it is present on cells of the vascular endothelium of lung, heart, intestine, and kidney [115, 116] . The binding of the viral S-protein to ACE2 occurs via the RBD domain of the S1 subunit which binds to the host cell as a trimer [117] ; (Fig. 7 (c) ). The dissociation constant of the SARS-CoV-2 S -ACE2 interaction is high and measures KD = 14.7 nM [117] . This docking reaction is paralleled by a conformational change of the S1 subunit. Only, in the opened, "up" conformation the RBD is accessible for the receptor, while in the closed, "down" state this domain is unable to efficiently bind to the receptor protein ( Fig. 7 (b and c) ) [108, 117] . The subsequent step, membrane fusion/virus entry, requires a proteolytic cleavage of the S-protein by host cell proteases into the S1 and S2 subunits ( Fig. 7 (d) ) [118, 119] . Two cleavage sites exist, the S1/S2 site, at the boundary of the S1 and S2 subunits, and the S2' site, close to the fusion peptide [120] [121] [122] . The major cleavage site is the S1/S2 site, which is characterized by the presence of a cluster of basic (Arg) aa residues [108, 117] . Several host cell proteases catalyze the proteolytic activation/cleavage of the S-protein at the S1/S2 site. The main proteases involved are the serine protease, TMPRSS2, located in the host cell membrane, as well as the serine proteases furin and cathepsin L [123] [124] [125] . The plasma membrane-bound TMPRSS2 and furin are highly expressed in lung epithelial cells [126] . Proteolytic activation of the S-protein by TMPRSS2, after SARS-CoV-2 binding to ACE2, allows the virus to fuse directly with the host cell membrane (Fig. 7 (e and f)). During this process, a trimeric hairpin structure is formed by the two heptad repeats HR1 and HR2. After this step, viral RNA is channeled into the host cell [124] .

During the initial interaction of the viral RBD with the cellular ACE2, the virus particles have to cross not only the mucus protection layer on the outer surface of the epithelial layer but also to penetrate the bulky heparan sulfate (Fig 8 (a) ). Both systems form a strong barrier between the cells and the outer respiratory system, which is in contact with the respiratory air. The glycosaminoglycan polymeric chains control the association of SARS-CoV-2 and the subsequent internalization of the virus [127] . This association has been localized at the glycosaminoglycan binding motif at the S1/S2 proteolytic cleavage site [127] . While the anionic and cell membrane associated heparan sulfate molecules link the virus to the cell surface and modulate infection (Fig  8 (c) ) the anionic polyP, as a soluble, not membrane-associated polymer, blocks the binding of the viral spike protein to the cellular receptor (Fig 8 (b) ). Until now it cannot predicted if, besides of the association step of the viral RBD to the cellular ACE2, polyP also inhibits the endocytosis process. Data are lacking for soluble polyP and polyP nanoparticles, but the potential seems to be there [128] .

The RBD at the S-protein is the initial sensor of the virus, screening for suitable receptors at the host cell membrane, a common feature for all members of the Coronaviridae [119, 129, 130] . In SARS-CoV-2 the RBD sequence has a length of 218 aa, a theoretical molecular weight of 24,507 Da and an isoelectric point of 8.9.

The initial interaction of SARS-CoV-2 with the host cells occurs via the RBD and ACE2. This interaction is specific, since the sequence of the RBD with its characteristic pattern of basic aa, the cationic groove, is exclusively present in SARS-CoV-2. Data bank analysis with the deposited sequences revealed an expect value of 2e -166 [131] . The lower this value, i.e. the closer to zero, the more significant is the result, meaning it is unlikely that this pattern is found again in the data bank [119] .

Furthermore, it is notable that none of the Arg residues arranged at the cationic groove of the SARS-CoV-2 RBD has been changed if compared with other strains of SARS-CoV-2, even to mutated SARS-CoV-2 like the one with the accession number 6XDG_E, expect value 4e -160 . This finding implies that the evolutionary adaptation of the RBD is close to optimal for the present-day environment of the SARS-CoV-2 within the global human population [132] .

The polycationic groove on the RBD surface One prominent aa pattern exists on the surface of the RBD. This domain comprises a continuous stretch of the basic aa Arg and Lys [109, 119] . In more details, the RBD has a surplus of basic aa (with Arg + Lys + His: 11 + 10 + 1) over the acidic aa (Asp + Glu: 15). In consequence, the calculated theoretical isoelectric point (pI) of this domain adjusts to 8.9, reflecting the abundance of basic aa over acidic aa. The pI's of the basic aa Arg (pI of 10.8), Lys (9.5) and His (7.5) indicate that below these pH values the aa are mainly positively charged. Therefore, polyP as a polyanionic inorganic material [131] and the organic polyanionic biopolymer heparin [133] , with their polyanionic surfaces, have been proposed to interact, via electrostatic interactions, with the RBD of the S-protein of coronaviruses in general [134] , and of SARS-CoV-2 in particular [131, 133] .

On the surface of the COVID-19 RBD, five Arg and four Lys units are present among which six are clustered together, forming a continuous trail (cationic groove) ( Fig. 9 (a  and b) ). Interestingly two of the Arg residues Arg457 and Arg466 are spatially connected with Asp467 and Glu465 (anionic pair), two aa that are known to build up a strong intramolecular proton transfer system [135, 136] . Such reactive centers, which are formed by two cationic aa, Arg, that are traversed by the two anionic aa, Asp and Glu, facilitate a covalent-like reaction with addition to the Arg guanidinium group [137] .

Using the algorithm of Kyte and Doolittle [138] a continuous stretch of aa with a hydropathy index of -4,5 exists, reflecting the clusters of Arg (index -4.5) and Lys (index -3.9). This path matches with the cationic groove on the RBD surface ( Fig. 9 (c) ). In parallel, the electrostatic potential has been calculated by applying Poisson-Boltzmann equation [139] . The prediction revealed that the continuous Arg-/Lys-rich patch (cationic groove) coincides with the positively charged areas while the crossing negatively charged stretch matches with the transversing Asp/Glu aa pair, the anionic pair ( Fig. 9 (d) ). This cationic Arg-rich cluster of the RBD matches intriguingly perfect with the spacing of the anionic phosphate units within the polyP chain ( Fig. 9 (e) ). In contrast, in other organic polyanionic molecules, like in heparan sulfate, non-charged building blocks interrupt the charged units, the monosaccharides.

In September 2020 [140] a new mutant of SARS-CoV-2, termed lineage B.1.1.7 emerged in the United Kingdom, in Kent and in Greater London. This mutant accumulated 17 mutations in its genome, eight of them are located in the gene that encodes the S-protein on the virus surface. This mutational change came not unexpected, since this virus is not equipped with a nucleic acid repair system. In turn, a median mutation rate of 1.12 × 10 −3 mutations per site per year has been calculated for the SARS-CoV-2 [141] . However, this mutant turned out to be more virulent, with a higher disease propensity. Among the evolved mutations in the lineage B.1.1.7, one of them, the N501Y mutation, was predicted to have a higher affinity to human ACE2 protein than the earlier strain [142] . This mutation took place at position 501 with asparagine (N) by tyrosine (Y) (Fig. 9 (e)) [140] . While mutations in SARS-CoV-2 are usually attributed to the zoonotic nature of this virus and facilitate a jumping between vertebrate species, the B.1.1.7 lineage emerged in patients and has been implicated with a higher affinity of the RBD to ACE2 and, in parallel, with a more severe disease progression [143, 144] . The modeling experiment suggests that the new mutant N501Y should not affect the topology of the cationic groove, since the location of the tyrosine is at the interface between the RBD and ACE2 (Fig. 9 (e) ).

It is conceivable that for a selective inhibition of the interaction between a ligand (like the RBD of the viral S-protein) and the corresponding receptor, e.g. with ACE2, a unique sequence or pattern must be present on the RBD. As outlined, such as sequence was identified with the cationic groove on the SARS-CoV-2 RBD surface ( Fig. 9 (a) ).

The mentioned hydrophilicity/hydrophobicity analyses of the surface of RBD, together with the electric charge distribution pattern and the presence of aa with negative hydropathy indices suggested that the highly charged polyanionic polymer polyP with its high charge distribution shell could tightly interact with the RBD. In the first studies on the effect of polyP on the binding affinity between the viral RBD and cellular ACE2 receptor a polymer with a chain length of ∼40 Pi units has been tested [131] . For the docking experiments polyP with 15 Pi units is used since this length can span the entire RBD.

PolyP provides the steric prerequisites to tightly interact with the RBD due to the rotational flexibility of the polymer and its ability to bend at the anhydride linkages with 130° and 102° [1] . Longer chains or overhangs of the polymer, after binding to the RBD and to the RBD : ACE2 complex, could even bind to the ACE2 receptor as well (Fig.  10) . Interestingly, at the interface/docking site between RBD and ACE2 a likewise cluster of the basic aa Arg (R393), Lys (K31 and K68) and His (H34) is present which would allow longer polyP chains, after binding to the RBD, to turn back and also to bind at this interface. Such bending would be facilitated by a partial interaction of the polyP chain with cations. At a higher resolution polyP binds to the RBD at the Arg/Lys (R/K) residues at position 457, 462, 466, 355 and 357 ( Fig. 9 (b) and Fig. 10) . There, the association of the polyP with the RBD surface is tight, at the binding sites with a distance of ∼3Å. This number matches with the published distances between Arg and phosphate [145, 146] . The attachment sites within the polyP chain to the RBD have a spacing of 5 Pi units at one end (between Arg residues R457 and R466) and 3 Pi units in the middle and again 5 Pi units at the terminus (Arg residues R355 and R357). This arrangement suggests an odd number of the RBD-binding Pi units within the polyP, suggesting a staggered conformation of the polymer between these attachment points and excluding Ca 2+ as binding counterions, at least at the binding area. These experimental data also imply that polyP molecules of different chain lengths interfere with the interaction of the virus with the cells surface and modulate the infection process.

The Arg residues at the RBD have been modified in two directions, first by covalent addition of a cyclohexanedione ring, and second by Pi produced in vitro from polyP. Firstly, the RBD has been incubated with 1,2-cyclohexanedione (CHD) in a sodium borate buffer (pH 8-9) to derivatize the guanidinium group under formation of N 7 ,N 8 -(1,2-dihydroxycyclohex-1,2-ylene)-L-arginine [DHCH-Arg] [147, 148] . Prior to this modification the charged guanidinium group has a planar structure and is largely unreactive due to the resonance stability in this group. After derivatization of the Arg moieties, DHCH-Arg, become strong electrophiles [149] and can react with its vacant orbitals to electron-rich centers ( Fig. 11 (a) ). Especially if Arg is localized adjacent to an acidic aa, like with the two Arg residues 457 and 466 which are spatially associated with Asp-467 and Glu-465, a site of strong intramolecular proton transfer is formed [135] allowing a strong "covalent"-like binding to phosphate. Arg and especially exposed derivatized Arg moieties can even passively penetrate across cellular membranes [150] . After modification, the DHCH-Arg units comprise with their DHCH groups at physiological pH a strong electrostatic potential. In turn, the amino acid can invade into the first dipole water layer that forms a bulky, 1 to 1.5 nm thick, hydration shell [151] . In addition, water molecules surround the modified Arg and build a gradient of dynamic dipoles providing the aa with an increased ligand reactivity [152] . Additionally, DHCH-Arg is bending the guanidinium group out of the overall surface plane. Therefore, Arg by itself and especially the modified DHCH-Arg is penetrating further into the hydration layer (Fig. 11 (c) ).

A second, physiological derivatization of Arg proceeds with Pi groups along the pathway seen in cells ( Fig. 11 (b) ); there a stable linkage of the Pi to the carbon atom of the guanidinium group occurs. In intact cells a physiological synthesis of phosphoarginine, exposing a phosphoramidyl amino acid residue (N ω -position) is known [153, 154] . Projecting this finding to the situation with the RBD, especially in the physiological environment, the added polyP will be enzymatically hydrolyzed by the ALP under the release of Pi. This Pi unit will at least transitionally bind to the polarized carbon atom in the Arg guanidinium group (Fig. 11 (b) ). Also this change at Arg will not only increase the electrostatic potency of the aa but also generate an Arg derivative that protrudes into the hydration shell surrounding the protein, here the RBD (Fig. 11  (c) ).

Binding assay for the polyanionic polyP to the cationic groove The mentioned modeling studies disclosed that on the surface of the RBD a cationic groove, formed of the basic aa Arg and Lys, as well as His, exists that could interact with polyanionic polymers. In order to test the efficacy of potential inhibitors of the RBD : ACE2 binding, an assay system based on recombinant proteins has been applied [131, 155] . The recombinant ACE2 was attached to a well plate and the recombinant RBD of the S-protein (labeled with biotin) was used as the interacting ligand. The binding strength between the RBD to the ACE2 is sensitively measured after incubation with labeled streptavidin in a chemiluminescence assay (Fig. 12 (a) ) [131, 156] .

Incubation of the RBD with polyP causes a reduction of the binding propensity to the ACE2 protein. Under controlled conditions with non-modified polyP a significant inhibition, a reduction of binding to 72%, is measured with 1 µg mL -1 [131] . The inhibition increases, reflected by a decrease of binding between RBD and ACE2, to 28% at the high concentration of 100 µg mL -1 of polyP ( Fig. 12 (b) ). While this series of experiments was performed with a polyP sample of 40 Pi units (Na-polyP40), a smaller polymer with a chain length of only 3 Pi residues (Na-polyP3) was selected in comparison. Surprisingly, a likewise strong inhibition was measured Fig. 12 (b) . Two conclusions can be drawn; either the strength of inhibition is saturable with Na-polyP40 due to a steric hindrance of Na-polyP40 at the Arg residues, or the short phosphate oligomers act inhibitory but cover only a portion of the cationic groove. To proof this proposition the Arg residues at the RBD were chemically modified [156] .

An increase in the inhibitory potency of polyP could be achieved with the RBD which exposes the derivatized Arg amino acids. The modification of Arg was performed with CHD to obtain DHCH-Arg [156] . This change turns Arg to an aa with a strong electrostatic potential around the Arg guanidinium group and allows a stronger access of the anionic polyP to the Arg moieties at the RBD. This was confirmed experimentally by showing that the inhibitory potency of polyP in the RBD : ACE2 binding assay increases. A comparative inhibition analysis between Arg and DHCH-Arg showed that by application of Na-polyP40 at a concentration of 0.1 µg mL -1 the binding efficiency was only reduced to 88% (non-modified Arg at the RBD), while a much stronger reduction to 23% was measured with the DHCH-Arg-modified RBD in the binding system ( Fig. 12 (c) ). After a further increase in the concentration to 10 µg mL -1 only a reduction of the binding to 60% was seen for the non-modified RBD, but a much stronger inhibition to 3% with the DHCH-Arg-modified viral binding protein [156] .

For a molecular stoichiometric analysis, the strength of inhibition has been assessed on a molecular basis, correlating the concentrations of the RBD molecules in the assay (applying the DHCH-Arg-modified RBD) with Na-polyP40. At a concentration of 0.05 μg mL -1 of polyP a 50% inhibition was reached in the binding assay with DHCH-Arg-modified RBD. Based on the molecular weight of the RBD with ∼24,500 Da, a ratio between 6.85 • 10 -8 mmol polyP and 8.16 • 10 -8 mmol RBD was calculated to be present in the binding assay at 50% inhibition. This calculation implies that every molecule of RBD is associated with one molecule of Na-polyP40, also reflecting that the binding of polyP to the DHCH-Arg-modified RBD and the following interaction with the cellular ACE2 receptor is highly sensitive and likewise selective.

The concentration of the two fractions of polyP, existing in the circulating blood, the relatively short polymers with 60-100 phosphate units and the long chain polymers were determined in human serum/plasma. For the short chain fraction a concentration of around 10 µM (with respect to orthophosphate units; equivalent to 0.8 µg mL -1 ) and for the long chain fraction ~40 µM (3.2 µg mL -1 ) were quantitated [157] . It must be stressed that at present the quantification methods are not very precise. In addition, polyP levels in blood are mostly estimates based on the soluble, short-chain polyP fraction released after platelet activation [47] , neglecting the insoluble, particulate fraction of polyP on the platelet surface [48] . Nevertheless, the concentrations found to be inhibitory in the assay with the viral RBD, derivatized DHCH-Arg, reach values determined for polyP within the physiological range.

PolyP is an inorganic biomaterial, which reduces the interaction between the viral RBD and the corresponding cellular ACE2. Therefore, it was interesting to clarify if also organic polyanions cause this inhibitory effect, like heparin. Heparin is an established anticoagulant and inhibitor of enzymes that mediate blood clotting [158] [159] [160] [161] . This polymer has been reported to hold anti-viral activity [162] . Recently three groups published an interaction of heparin, and related compounds like heparan sulfate, with the RBD of the SARS-CoV-2 S-protein [101, 163, 164] . Based on these data an application for patients has been proposed [165, 166] . Applying circular dichroism spectroscopy, these authors recorded a change in the conformation of the RBD after exposure to heparin [163, 164] . In turn, an inhibitory effect of heparin on binding of the virus to the ACE2 receptor has been implicated. Furthermore, experiments showed that heparin has not only the propensity to interact with the RBD but additionally also with the S1/S2 proteolytic cleavage site in the S-protein and even further to the glycosaminoglycan binding domains in this viral sensor.

Using the sensitive RBD : ACE2 binding assay, no reduction of the binding of the RBD to the ACE2 receptor was measured for heparin/heparan sulfate within the concentration range of 0.2 to 200 µg mL -1 of the organic polymer ( Fig. 12 (d) ) [156] .

In the respiratory airways system, a two-phase mucus/mucin layer is covering and protecting the epithelial cell layer. This layer is particularly bulky on the surface of the oropharyngeal airways epithelia. From there the viruses enter the pulmonary circulatory system. The mucociliary transport layer, exteriorly present on the epithelial cells, acts as a self-clearing and self-cleaning mechanism of the airways in the respiratory system. It directs the mucus to the pharynx from where it is either swallowed or coughed up; the clearance velocity is fast with ∼5.5 mm min -1 [167] .

The virus is entering the nasal lumen [168] from where it spreads via an airborne transmission route. Then viral particles reach the surface region of the respiratory epithelia and either resist there or invade through the mucus/mucin layers the epithelial cells via the entry molecules ACE2 and TMPRSS2 [169] .

In healthy individuals the mucus is an efficient protection system against an immediate infection of cells with the virus [170] . The mucus is produced in the epithelial cells as slippery aqueous secretion. This viscous colloidal fluid is supplemented with inorganic salts, antimicrobial proteins/enzymes (lysozymes, lactoferrin), immunoglobulins, and the mucins. By this, the mucus acts as a physical barrier against foreign materials, both of physical (e.g., dust particles) and (bio)chemical origin, including bacteria or viruses [169, 171, 172] . The mucins, the major functional components of the mucus, are a large family of heavily glycosylated proteins, containing both N-linked and O-linked glycans.

Until now 22 mucin-type glycoproteins have been identified, which are collectively termed MUC [172] . The family of mucins are grouped into the secreted mucins, gelforming components of viscoelastic mucus gels protecting the epithelia (like MUC5AC and MUC5B) and the membrane-bound mucins with the prevalent MUC1 [172] . The epithelium is compost of the goblet cells, which produce the mucins, and the ciliated cells which are decorated with ∼200 cilia each of 7 µm in length ( Fig. 13 (a) ). The cells secrete the membrane-tethered mucins (MUC1, MUC4, and MUC16) deposited in the periciliary layer, while the outer layer of the mucus of the respiratory tract contains the two gel-forming mucins (MUC5AC and MUC5B) [172] .

Extracellularly, the different mucins with their functional groups form a tight meshwork. The termini of the glycans are usually built by charged sialic acid units [173] [174] [175] and by this have the propensity to form salt bridges. From there the hierarchical organization to a network starts by building up Ca 2+ -mediated links from monomeric mucins, stabilized with hydrogen bonding. Finally, disulfide bonds link the monomeric mucins to a solid scaffold into which host cells and foreign bacteria can be entrapped ( Fig. 13 (b) ) [176] . Also additional non-mucus proteins, like cytokines can be included [177] .

Due to their small sizes viruses can penetrate the mucus barrier and reach the epithelium. Some virus, like the influenza viruses, are secreting their own neuraminidase, which facilitates their migration through the extracellular glycoprotein shield and also through the mucus [178] . In contrast, larger virus particles like the herpes simplex virus (size of 180 nm) remain immobilized within the mucus layer. As a self-defense of the epithelial cells against influenza viruses they induce an upregulation of mucins, especially in human nasal epithelial cells [179] .

Only little is known about the presence of potential antivirally acting factors/compounds in the mucus. Some intrinsic antiviral activity in mucins has been measured which has been attributed to the adhesive and enveloping property of the heavily glycosylated proteins [180] .

As outlined the polyanionic polymer, the inorganic polyP, which abolishes the binding affinity of the RBD to the ACE2 receptor, also remains active in the mucin environment [181] . Even more, the inhibition is even found at polyP concentrations which are close to those measured in the circulating blood. This finding has therapeutic relevance, since for COVID-19 patients it has frequently been shown that they display clinical symptoms of thrombocytopenia, which is characterized by abnormally low levels of platelets [96] . From this finding it can be deduced that not only the release of cytokines from platelets but also of polyP from them into the mucus layer is impaired. In order to elucidate if the anionic polyP interferes with the anionic mucins combinatorial experiments have been performed. Commercial mucin (from bovine submaxillary glands) was selected. Electron micrographs of the mucin added to the system show that during the incubation in the binding assay (supplemented with MgCl2) small polyP nanoparticles of a size of ∼20 nm are formed in the mucin matrix ( Fig. 14 (a-i) versus   (a-ii) ). After addition of mucin together with polyP (10 µg mL -1 and 100 µg mL -1 ) no reduction of the polyP inhibitory activity was measured in the RBD : ACE2 binding assay (Fig. 14 (b) ). The physiological concentration of mucin in the airway surface liquid is ~100 µg mL -1 [182] . These findings suggest that mucin does not interfere with the inhibitory potency of polyP in the RBD : ACE2 binding assay.

To elucidate the potential benefit of polyP as a protective material against virus infections, gene expression studies with the alveolar A549 cells, which are positive for the ACE2 receptor [183] , were performed. Like determined in other cell systems also here polyP, even in the presence of mucin, elicits morphogenetic potential [5, 181] . In the steady-state-expression system using alveolar A549 cells both the MUC1 gene and the MUC5AC gene became significantly upregulated (Fig. 14 (c) ). Additionally, it was confirmed that in this cell system ATP is produced extracellularly after exposure to polyP and mediated by the cell-bound ALP and ADK [1] .

As long as no efficient drugs are available for the treatment of COVID-19 patients and the immunization of the population is not sufficient, preventive protective measures are indicated. Such protective measures can be based not only on the use of nose and mouth masks in combination with a suitable lifestyle, as well as disinfectants for contaminated surfaces or skin, but also on materials that show antiviral activity against specific viral molecules or components. In this review it is highlighted that such preventive/protective tools against the initial phase of SARS-CoV-2 infection are at the horizon or beyond. The materials/polymers in focus are polyanionic inorganic polyP and perhaps also organic polysaccharides (heparin/heparan sulfate) that are directed against distinct patterns on the target virus necessary for host cell recognition and infection.

Innate immunity is the initial response of animals/humans to microbes that prevents, controls and eliminates infection of the host with bacteria, viruses and other pathogens [184] . It is an evolutionary old system and evolved together with the Porifera ∼800 million years ago [185, 186] . This first-line defense system is efficient and complex and provides an immediate response to microbial assaults. The innate immune response to SARS-CoV-2 has been elucidated to some extent [187] . The single-stranded RNA of this virus is interacting with extracellular and endosomal pattern recognition receptors, the Toll-like receptors (TLRs). After their activation the secretion of cytokines starts with the interferons as the most important molecules of the antiviral defense arsenal. Subsequently other cytokines, like the proinflammatory tumor necrosis factor α and interleukins -1, -6 and -18 are released that potentiate the adaptive immune response [187] . In the first phase of infection the SARS-CoV-2 Sprotein with its RBD has to bind to the cells and, therefore, has to break the barrier of the protective covering mucus layer on the epithelial surfaces. At the beginning of the infection the airway epithelial cells with their mucus layer function as an innate sensor and act as a first mechanical barrier defending against the virus. The epithelial cells can suppress proximate inflammatory processes by inhibition of the pattern recognition receptors signaling chain and the secretion of inhibitory cytokines, eicosanoids, or glucocorticoids [188] .

The innate immune defense system provides not merely a passive protection against the pathogen, like in a passive immunization treatment, but includes also an active response of the body through self-production of defense molecules [189] . Passive immunization using immunoglobulin therapy is indicated for individuals with B cell immunodeficiencies to support the antibody-antigen neutralization steps. This passive treatment is usually supplemented with adjuvants that are additionally used to improve the immune status of the patient and boost the host immune responses [190] .

As summarized in this review polyP, inhibits the RBD : ACE2 interaction even at around physiological concentrations. This passive protection is flanked by an active, a morphogenetic induction of mucin gene expression in cells, exposing the ACE2 receptor [181] . The upregulation of the mucin genes is seen in A549 cells, which are related to the respiratory epithelial cells [191] . In these cells, the expression of the mucin genes is in cis, meaning at the site from where polyP is reaching the cells. More specific, polyP induces the steady-state-expression of the MUC1 gene. MUC1 contributes to the formation of the periciliary layer and contributes to the antiviral activity of the mucus. In addition, polyP induces an activation of MUC5AC. This mucin is localized on top of the periciliary layer. There, in the MUC5AC layer the virus particles become trapped into the mucus (Fig. 15 (a) ) [181] . In turn, polyP will surely contribute to restoration of the cellular integrity and function of the epithelial cell layer, perhaps also in vivo in SARS-CoV-2 infected patients ( Fig. 15 (b) ). In order to apply the polyP formulation to the respiratory system, the region where SARS-CoV-2 first hits the human target tissue, the anionic polymer in the form of soluble Na-polyP and of Ca-polyP nanoparticles, can be dissolved/suspended in a buffered and preserved water-based solution and used as a nose spray. Buffering of the solution to a pH higher than 6.5 is recommended in order to prevent an acid-driven hydrolysis [192] .

The solvent is sprayed into the nostrils a process during which the fluid is nebulized (Fig. 15 (b) ). The aerosol particles formed reach the depths of the pharynx [193] . During this passage the aerosol particles, which contain polyP and are formed by nebulization (step 1), reach the mucus layer that covers the respiratory epithelium (step 2); there the polyP in the aerosol is transferred into the coacervate phase (step 3), which enwraps the virus particles (step 4). During this latter step, the polyP molecules bind to the spikes of the virus and physically prevent the virus to bind to the target cells in the respiratory epithelium.

In conclusion, the polyanionic polyP material acts on three levels as antiviral or antiviral protective material. When inhaled through the airways, the SARS-CoV-2 particles face the bulky mucin layer which frames the airway epithelium ( Fig. 15 (b) ) [194] . When the polymer comes into contact with this mucous, a glyco-proteinaceous fluid, polyP undergoes coacervation and embeds virus and virus-like particles [71] . In turn, polyP strongly binds to the RBD at the S-protein of SARS-CoV-2, as summarized here. Finally, polyP provides metabolic energy to the epithelial cells, amplifying their natural defense properties [195] .

The results highlight that the inorganic biomaterial polyP, as the only polyanionic polymer to date, is a promising candidate for the rational design and the further development of a targeted antiviral defense against SARS-CoV-2. This material is physiological and is metabolized in the body and eliminated via physiological routes.

There are no conflicts to declare.

The authors declare no data availability.

Inhibitors of enzyme reactions and of ligand-to-receptor interactions act differently. (a) During enzyme reactions the substrates (-1 and -2) are covalently modified under formation of products. These processes can be inhibited at the active site of an enzyme (competitive inhibitionlike shown in the sketch) or at at a place adjacent to this position (non-competitive inhibition). As results, the chemical reactions (covalent bond formation between two substrates) are downscaled or even blocked. (b) The ball-andstick models show arabinofuranosyl-cytosine (ara-C; right), which acts as a competitive inhibitor of DNA polymerases in the form of ara-CTP, while the normal substrate dCTP, the phosphorylated form of deoxycytidine (dCyt; left), is incorporated into DNA. The inhibitor carries an OH group at position 2 in trans configuration to the hydroxyl at position 3, while in the deoxyribose at position 2 no hydroxyl group is present (circled). (c) A ligand interacts with a membrane-bound receptor and initiates a transmembrane signaling process. Prior to this, the receptors often form dimers via controlled lateral diffusion of the receptor monomers within the membrane. If the respective interaction between the ligand and the receptor is blocked, e.g. by polyP, the receptor dimerization stops and the signal transduction chain is disconnected. 

Intracellular synthesis and accumulation of polyP; partially hypothetical scheme. (a) Blood platelets are the major stores of polyP. This polymer is presumably synthesized in megakaryocytes, from where the platelets originate by cytoplasmic fragmentation. (b) In mitochondria ATP is produced from reducing equivalents that accumulate during the metabolism of glucose, fatty acids, amino acids and pyruvate, like in the citric acid cycle. The reducing equivalents formed (NADH, FADH2) drive the electron transport chain at the inner mitochondrial membrane, followed by an export of the protons into the intermembranous space. From there the protons are rechanneled via the inner mitochondrial membrane and thereby drive the ATP production by ATP synthase. Finally, ATP releases the mitochondrion via the adenine nucleotide translocase (ANT; inner mitochondrial membrane) and the voltage-dependent anion channels (VDAC; outer mitochondrial membrane) into the cytoplasm. Adjacent to the mitochondria are the acidocalcisomes. In the membranes of these organelles, the Vtc complex is located with its subunits 1-4, where, in budding yeast, polyP is polymerized and subsequently pressed into the lumen of the organelle. In vertebrates this pathway is unknown. Adapted with permission [1] . Copyright 2019 American Chemical Society.

Chemical structure and conformation of polyP. (a and b) Eclipsed and staggered conformation of polyP. The eclipsed conformation is enforced by Ca 2+ , while the staggered conformation is formed with Na + . The chemical structures and the structural models are given. (c) Molecular geometry of polyP. The bond angles, the atomic radii (ratom) and ionic radii (rion) at a phosphate unit are given. (d) Amorphous polyP nanoparticles prepared as (i) Ca-polyP salt from Na-polyP and CaCl2 or (ii) Mg-polyP salt from Na-polyP and MgCl2.

Transition of Ca-polyP nanoparticles to the corresponding coacervate. (a) At pH 10 Ca-polyP, or even faster Mg-polyP, nanoparticles are formed. These particles show a relatively long stability in culture medium without serum. (b) After transfer into medium and serum of pH 7, the nanoparticles undergo coacervation, under concomitant inclusion of serum peptides or other morphogenetically active supplements. After addition of peptides to the nanoparticles, with a high ζ potential, the potential decreases and the coacervate is formed (i to iv). The coacervate phase has a lower stability. 

Initial steps of the SARS-CoV-2 replication cycle. a. Passage of the virion across the mucus barrier. b. Sensing of the viral envelope with its homotrimeric spike S-proteins, consisting of the S1-subunit with the receptor binding domain (RBD), and the S2subunit, for a suitable surface receptor, for ACE2. c. Binding of the RBD of S1 to the dimeric host cell ACE2 receptor. d. Enzymatic/proteolytic cleavage of S-protein at the S1/S2 site. e. Post-cleavage phase and prehairpin formation. f. Fusion of the virus envelope with the host cell membrane. Further details are given in the text. The hurdle-rich migration of the SARS-CoV-2 to the host cell. (a) During the travel of the virus particles with its RBD at the S-protein to the cellular ACE2 receptor two functional layers have to be crossed; the mucus protection layer and the heparan sulfate shield. The virus particles (diameter 80-140 nm) approach the target cells via the airways and there through a mucus layer as well as the cell membrane-associated heparan sulfate shield. The bulky mucus layer, measuring ∼5 µm, covers the epithelial cells. In addition to mucin, the cell surface is decorated with up to 170 nm long heparan sulfate macromolecules. (b) In contrast, to the heparan sulfate molecules the polyP chains are freely present outside of the epithelial cells and block virus binding the cells. (c) The heparan sulfate polymers interact both with the SARS-CoV-2 spike protein and the cell surface and modulate infection.

Surface mapping of the S-protein RBD from SARS-CoV-2, adopted from PDBID:6M0J. In this view a polyP chain, with a length of 12 phosphate units, was used as a polyanionic model for molecular docking. (a) On the surface of the RBD space-filling model, the aa residues Arg (R) and Lys (K) are marked with their respective sequence positions. They form the cationic groove (colored in green). The crossing anionic pair of Asp (D) and Glu (E) is shown in purple, and the crossing area is circled (yellow). (b) Alignment of the polyP chain (red and purple) with 12 Pi units (P-1 to P-12) on the surface of the RBD (transparent), following the cationic groove (green). The interatomic distances of the Pi units to selected Arg residues are given for R457 and P-1, R466 and P-5, R355 and P-7, as well as for R357 and P-10. (c) Hydrophobicity distribution pattern of the surface of the RBD is highlighted between the indices -4.5 (hydrophilic; blue) to +4.5 (hydrophobic; red). (d) Electrostatic potential on the RBD within -10 (red) to +10 (blue). (e) Surface topology of the RBD in the newly emerged mutant SARS-CoV-2 B.1.1.7. The spatial aa arrangement of the cationic groove is not changed. The location of the mutation asparagine (N) by tyrosine (Y) it is circled in black. Adapted with permission [156] . Copyright 2020 The Royal Society of Chemistry.

Diagrammatic sketch of the interaction of polyP with the cellular ACE2 and the viral RBD. The docking interface of RBD and ACE2 is depicted. The basic aa of ACE2 at the RBD : ACE2 interface that potentially interact with polyP are labeled with position numbers and marked in green, like the basic aa on the RBD. Adapted with permission [131] . Copyright 2020 Elsevier.

Increased electrostatic interaction potential of Arg after its modification. (a) The Arg unit on the surface of the RBD undergoes covalent modification with 1,2cyclohexanedione (CHD) in a borate buffer, forming DHCH-arginine which has a negative charge. After this, the modified Arg units can interact with polyP via divalent cations. (b) Similarly, if polyP is enzymatically cleaved by ALP, releasing a Pi unit and a proton (or an intermediate metaphosphate species), a likewise modification of the guanidinium group of Arg proceeds transitionally. An overall negative charge is built at the guanidinium group. (c) Penetration of Arg into the hydration (water dipole) layer at the protein surface. Molecular models of arginine (middle), DHCH-arginine (left) and phospho-arginine (right); the guanidinium group at Arg is marked. Both the modification of the guanidinium terminus of Arg with CHD and with a phosphoryl group increases the reactivity of Arg. The modified aa penetrates with ∼2 Å through the 10 -15 Å thick hydration layer, formed around Arg on the protein surface.

Inhibition of binding of the RBD to the ACE2 by polyP. (a) The RBD : ACE2 binding assay. The recombinant cellular ACE2 is bound to the solid phase and the RBD, labeled with biotin, is present in the fluid phase. The RBD exposing the cationic groove is incubated with polyP. Then the amount of RBD molecules, bound to the ACE2, is quantified by using the streptavidin-biotin amplification/detection system. The protein molecular structures were taken from [109] . Adapted with permission from 2020 Springer Nature. Inhibition of the RBD : ACE2 interaction by polyP. (b) In the binding assay the two polyP fractions, Na-polyP40 (blue bars) and Na-polyP3 (yellow), inhibit the association between the two reaction partners. The maximal reduction of binding is achieved with ∼10 to 30 µg mL -1 . (c) The strength of inhibition is strongly amplified after modification of the Arg residues with CHD. Using those experiments with DHCH--Arg RBD as a target for Na-polyP40 a significant increase of inhibition is measured.

The inhibition values with Na-polyP40 for the modified RBD are given in cross-striped blue bars, while the values for the non-modified RBD are in plain blue. n = 3; * p < 0.05. (d) Effect of heparin and heparan sulfate on the interaction RBD : ACE2 using the two components binding assay. Only at the high concentration of 200 µg mL -1 a slight ( # ) but not significantly enhanced binding of the two reaction partners is measured. Adapted with permission [156] . Copyright 2020 The Royal Society of Chemistry.

The mucus layer of the nasal, pharyngeal cavity. (a) The columnar epithelial cells of the airways comprise two types, the ciliated cells and the secretory (goblet) cells. The secretory, goblet cells release soluble MUC5AC, in which foreign particles can be captured while MUC1 remains in the periciliary layer and bathes there the epithelial cells. The ciliated cells drive the mucus out of the airway system. Against this direction polyP, as nanoparticles (Ca-polyP-NP), is potentially sprayed into the nasal, pharyngeal cavity. (b) Hierarchical organization of the mucin network. The monomeric mucin units are assembled together via Ca 2+ links. Into this network foreign particles, like bacteria and viruses, become entrapped. PolyP added as Me 2+ salt/nanoparticle undergoes enzymatic hydrolysis by ALP, resulting in the release of the energy-rich bound Pi residues, which are used for ATP formation through the concerted action of ALP and ADK.

Preparation of a mucin/collagen-based hydrogel (HG) to study the inhibitory potency of polyP in a bioinspired environment. (a) Morphology of (i) dried collagen/mucin (5 mg mL -1 , 100 µg mL -1 ) as well as of the (ii) collagen/mucin/polyP HG (5 mg mL -1 , 100 µg mL -1 , 10 µg mL -1 ); scanning electron microscopy. NP are marked (> <). (b) Coincubation of polyP with mucin in the RBD : ACE2 binding assay. The studies were performed in parallel in the absence or presence of mucin (100 µg mL -1 ) and two concentrations of polyP (10 µg mL -1 and 100 µg mL -1 ). (c) Influence of collagen alone (1 mg mL -1 ), collagen (1 mg mL -1 ) and mucin (100 µg mL -1 ) together, and mucin with polyP (10 μg mL -1 ) on the expression of the MUC1 and MUC5AC gene in the collagenbased HG during a 6 d incubation period. n = 5; * p < 0.05.

Proposed application of the polyanionic polymeric material inorganic polyP as a physical and biochemical barrier against SARS-CoV-2 infection. (a) In the respiratory tract the polyanionic polymer binds "passively" to the RBD at the S-protein of the virus, followed by a passive neutralization. The epithelium functions as a receiver for polyP and acts in a second, active effector phase, as producer for ATP and mucus/mucins. In the goblet, the secretory, cells, mucin is produced that entraps the virus and by this encapsulation removes the virus from the respiratory tract through the clearance of the airways. (b) The polyanionic polymeric material, applied as soluble Na-polyP and as particle-associated Ca-polyP-NP, in a nose-to-lung (N2L) aerosol delivery system, is (step 1) nebulized in the nasopharyngeal tract. There, the aerosol particles fuse (step 2) with the mucus layered on top of the respiratory epithelium under formation (step 3) of the polyP-formed coacervate. Finally, (step 4) the virus particles are enwrapped by the coacervate. During this process SARS-CoV-2 becomes physically separated from the cells and an infection is suppressed. Adapted with permission [194] . Copyright 2021 Ivyspring International Publisher.

Crystal Structure, Morphology, Defects

Biochemistry and Technology

Viruses: From Understanding to Investigation

Activation of Viruses by Host Proteases

Proc. Natl. Acad. Sci. USA

Proc. Natl. Acad. Sci. USA

Chemistry and Biology of Serpins

☒ The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.☐The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: