Mineral-binding peptides for direct, high-affinity protein immobilization, purification and targeting

Abstract

The invention relates to a mineral binding peptide, wherein said mineral binding peptide comprises the consensus sequence TGSGQTFTYINWGPGNPDNSG (SEQ ID NO. 200) and is selected from PHG(653-784) (SEQ ID NO. 7) or a truncated peptide thereof or a tandemized peptide thereof. The invention further relates to the use of said mineral binding peptides and minerals or mineral carriers in non-medical and medical applications. The invention further relates to a method of eluting said mineral binding peptides from said minerals or mineral carriers and to a method for affinity purification of proteins comprising said proteins.

Description

[0001] Mineral-binding peptides for direct, high-affinity protein immobilization, purification and targeting

[0002] Field of the invention

[0003] The invention relates to a mineral binding peptide, wherein said mineral binding peptide comprises the consensus sequence

[0004] TGSGQTFTYINWGPGNPDNSG (SEQ ID NO. 200) and is selected from PHG(653-784) (SEQ ID NO. 7) or a truncated peptide thereof or a tandemized peptide thereof.

[0005] The invention further relates to the use of said mineral binding peptides in non-medical and medical applications. The invention further relates to a method of eluting said mineral binding peptides from minerals or mineral carriers and to a method for affinity purification of proteins comprising said proteins.

[0006] Background of the invention / state of the art

[0007] The discovery of protein tags represented a game changing event in protein chemistry and related disciplines. The term “tag” describes a short peptide sequence with specific properties, which is usually used as a fusion partner to add such properties to a target protein. Such tags facilitate both easy purification of the target protein and its antibody-mediated detection (e.g. by Western blot, pull down assays). In addition, tags can also be used to immobilize the target proteins onto specific carrier materials, which is currently and increasingly used in the field of biocatalysis (e.g. single- or multi-enzyme reactions for the environmentally friendly synthesis of pharmaceuticals or aroma compounds) or in analytical applications (e.g. biosensors, lateral flow assays). A portfolio of different tags is available for these diverse applications. The most common tags currently used in affinity applications, which either bind to proteins, small organic molecules or complexed metal ions immobilized on a carrier. E.g. a polymer. The most relevant examples are summarized in Fig. 1. Noteworthy, the cheapest and thus most widely used tagging system is based on polyhistidine peptides (Fig. 1), which bind to heavy metal ions of polymer-bound ligands (e.g. nitrilotriacetate (NTA) complexes of Ni2+, Co2+, Zn2+) and are used in so-called immobilized metal affinity chromatography (IMAC) applications. However, its heavy metal content makes the IMAC system unattractive for many industrial applications due to the physiological activity of the metals (metabolic and gene toxicity, nickel allergy), which causes additional costs fortheir removal and metal trace analysis. In combination with the high prices for the other carrier materials (Fig. 1) on the market, affinity chromatography is still outcompeted in industry by conventional but more laborious methods for protein separation (a.o. salt precipitation, ion exchange, hydrophobic chromatography, gel permeation).

[0008] In contrast to conventional carrier materials, many minerals are cheap, robust, recyclable and biocompatible substances - features which predestine them as carriers for industrial production processes. Thus, minerals are well established e.g. as scaffolding substances for chemical catalysts. However, the application of minerals for purification of biomacromolecules (e.g. in the challenging field of protein isolation) is much less developed. Especially a technical viable process which applies minerals as affinity matrix in protein purification (i.e. in the isolation of target proteins from complex mixtures such as cell extracts) or immobilization (i.e. the fixation of proteins to surfaces) has not been described yet. The main reason behind this technological gap is the lack of suitable protein sequences which mediate the mineral binding affinity and are applicable as high-affinity tags.

[0009] Biominerals are a class of naturally occurring inorganic compounds, which show a low solubility in water and biological fluids and are therefore used as skeletal structures by many organisms. Well-known examples include calcium salts such as CaCCh (found in exoskeletons of sponges and crustaceans, mussel shells, avian eggshells) (Evans (2008) Chem. Rev. 108, 4455), hydroxyapatite (a basic calcium phosphate, which is the main constituent of vertebrate bones) (Hunter et al. (1996) Biochem. J. 317, 59) and the magnetic iron oxide magnetite (FesOt, found in a few magnetotactic bacteria). The high biocompatibility of these materials already favored their use in a number of biochemical applications in basic research (e.g. easily separable, surface-modified magnetite particles for RNA isolation I pulldown assays) as well as in medicine (z.B. tooth fillings made of synthetic apatites). However, a large-scale implementation of such minerals in industrial biotechnology - especially as a non-toxic, mechanically stable and re-usable stationary phase in protein purification and immobilization - has not been achieved so far.

[0010] This circumstance mainly originates from the fact that till today only a few peptide sequences are known, which can act as "adapters" to mediate a targeted specific interaction of a fused protein of interest with the carrier material. In contrast, a substantial number of proteins have evolved a natural capability to bind to minerals, and therefore became key factors for the integration of inorganic components into all forms of life. These proteins also actively contribute to the formation of mineralized tissues via so-called biomineralization, a developmental process, which also requires a coordinated interaction of proteins with mineral surfaces. Especially the latter fact has contributed to research on peptide sequences which bind specifically to CaCCh and hydroxyapatite. Such peptides can be classified into four main groups according to the biotechnologically relevant features including chemical properties and aggregation behavior: • phosphorylated peptides: Binding to the mineral occurs via phosphorylated amino acid residues (mainly serine or threonine) (Hecker et al. (2004) J. Struct. Biol.146, 310; Deshpande et al. (2011) Biomacromolecules 12, 2933; Sprenger et al. (2018) Langmuir 34).

[0011] • peptides containing non-canonical amino acids: The interaction requires other post- translational modifications such as the formation of y-glutamate (Lee et al. (2017) Biomater. Sci. 5, 663) or 3,4-DOPA residues (Pagel et al. (2016) Angew. Chem. 55, 4826).

[0012] • oligomeric peptides: The binding requires a homo- or hetero-oligomerization of the peptides (He et al. (2003) Nat. Mat. 2, 552; Su et al. (2013) Biochem. J. 453, 179; Jain et al. (2017) Biochemistry 56, 3607).

[0013] • monomeric peptides without any structural modifications: This group consists of peptides containing canonical amino acids only, which are functional in mineral binding as a monomer.

[0014] While applications using chemically modified binders as potential mineral affinity tags are scarce (JP5768025B2), peptides belonging to the last group of binders are most valuable due to practical reasons. Those include the fact that recombinant expression of canonical peptides in heterologous organisms (especially in the widespread bacterial systems) does not require further genetic modification of the host, such as the integration of phosphorylating kinases I the deletion of disturbing endogenous phosphatases or the construction of additional biosynthetic pathways to synthesize non-proteinogenic amino acids. Furthermore, the chance that the tag leads to an unfavorable aggregation of the recombinant target protein is much lower if a monomeric peptide is used for this purpose. Several studies have been performed to generate peptides interacting with selected minerals mainly by combinatorial methods (e.g. phage display), including binders for silica (SiCh) (Zerfass et al. (2015) Protein Expr. Purif. 108, 1 ; Yamagishi et al. (2016) Sci. Rep. 6, 35670; Rodiguez et al. (2017) Colloids Surf. B Biointerfaces 160, 154; Soto-Rodriguez et al. (2017) Protein Expr. Purif. 135, 70; patent WO2015042464A1), CaCCh (patent US7754680B2), hydroxyapatite (Natalio et al. (2010) Acta Biomaterialia 6, 3720; Yarbrough et al. (2010) Calcif. Tissue Int. 86, 58; Huang et al. (2016) Nat. Sci. Rep. 6, 38410, patent US9809633B2), transition metal oxides TiCh / ZrCh (Lemloh et al. (2017) Materials 10, 119; patent WO2015128643A1), clay minerals (patent DK1999140T3), ZnS (patent US20180111968A1), Mo / W chalcogenides (patent US9493513B2), elemental Ti (patent ES2327398T3), metal alloys (patents US8969252B2 and US8201724B2) or semiconducting materials (patents KR100942320B1 and US8058392B2). However, a broad applicability of these published sequences in protein separation and targeting is often not given. The peptides show either (I) a low affinity to the carrier mineral, which interferes with the application conditions (e.g. high-salt conditions, disturbing reagents), (II) a too limited binding specificity (i.e. interaction with impractical minerals only but low affinity for other, technically more suitable carriers), (III) a binding restricted to a certain pH and temperature (i.e. no tolerance for changing acidity or temperature conditions), (IV) are prone to aggregation (i.e. they negatively influence the expression yield of the target protein), or (V) are difficult to remove from the solid support under conditions mild enough to preserve the target peptides integrity.

[0015] Taken together, there is a great need for peptides that are capable of interacting with selected minerals under a broader range of conditions.

[0016] Summary of the invention

[0017] It was thus the purpose of the invention to provide peptides with improved properties (according to one or more of the points l-V above) that bind with high affinity to a broad range of minerals in order to overcome the disadvantages of the peptides known in the prior art.

[0018] In general, the invention includes selected peptides and applications, including a method for developing canonical, high-affinity mineral binding peptides by a combination of screening and mutagenetic improvement, resulting in optimized tags which are capable of binding to a broad variety of minerals. The applicability of the sequences in biotechnology is proven by the successful affinity purification of tagged recombinant target proteins.

[0019] In one aspect, the invention provides a method for preparing mineral binding peptides by identification of binding motifs in natural peptides followed by improvement of their binding affinities by mutagenetic analysis, peptide truncation, and the combination of natural and artificial peptides by subsequent tandemization, tri- and multimerization. The method of this aspect includes the possibility to adjust (I) the affinity to certain minerals (e.g. binding at high stringency), and (II) the stability of the interaction at varying pH I temperature, and (III) the tolerance of the binding against common biochemical reagents (especially thiols, buffers, chelating agents and denaturants).

[0020] In a further aspect, the invention provides mineral-binding peptides that were developed with the method of the invention, and mineral binding peptides that can be derived therefrom.

[0021] In a further aspect, the invention provides minerals (mineral carriers) for their use in the interaction with mineral-binding peptides that were developed with the method of the invention.

[0022] In a further aspect, the invention provides a method for eluting the mineral binding peptides of the invention from minerals using sugars and / or sugar alcohols and / or a salt gradient and / or further eluting agents, and the application of this method in affinity chromatography. In a further aspect, the invention provides the use of the mineral binding proteins of the invention for preparing fusions or conjugates with peptides, proteins, nucleic acids, small molecules, natural and synthetic polymers for certain applications.

[0023] In a further aspect, the invention provides the use of the mineral binding peptides of the invention (I) in affinity purification on mineral surfaces, (II) in immobilization of fused proteins or conjugated molecules to such surfaces, (III) in targeting in biological systems, (IV) for application in carrier-based analytical techniques requiring surface modification, and (V) for application as additives in the food, pharmaceutical and feed sector.

[0024] In a further aspect, the invention provides a method of removal of the mineral binding peptides of the invention by cleavage of integrated cut sites accessible by cleaving reagents or catalysts, especially biocatalysts or enzymes, such as external proteases, or in particular self-processing domains.

[0025] Brief description of the drawings

[0026] Figure 1 shows systems used in the prior art for affinity chromatography and immobilization of proteins. The fused target protein (protein of interest (POI)) , is schematically represented by a light grey ellipsis whereas the polymeric / mineral carriers are depicted in gray spheres / hexagons. CBD, chitin binding domain; CBP, calmodulin-binding protein; GST, glutathione-S-transferase; IMAC, immobilized metal affinity chromatography; MBP, maltose binding protein.

[0027] Figure 2 shows the screening of mCherry proteins tagged with putative binding peptides for affinity to minerals. The binding affinity data were determined after incubation of bacterial lysates containing a standardized amount of mCherry protein (15 pM) with the specified insoluble materials in HEPES buffer (100 mM HEPES / NaOH, 100 mM NaCI, pH 7.5) at 12 °C as described in the examples of the invention. The increasing shades of gray indicate interaction strengths in 20% increments, ranging from 0% to 100%. The salt additives listed in the table were included in the reactions at an optimized concentration to suppress unspecific binding.

[0028] Figure 3 shows the affinity of mCherry fusions carrying truncated variants of the PHG tag to selected minerals. The binding affinity data were determined after incubation of bacterial lysates containing a standardized amount of mCherry protein (15 pM) with the specified carrier minerals carriers in HEPES buffer (100 mM HEPES / NaOH, 100 mM NaCI, pH 7.5) at 12 °C as described in the examples of the invention. The increasing shades of gray indicate interaction strengths in 20% increments, ranging from 0% to 100%. The additives listed in the table were included in the reactions at an optimized concentration to suppress unspecific binding.

[0029] Figure 4 shows binding affinity data that were determined under high-stringency conditions using bacterial lysates containing a standardized amount of mCherry protein (15 pM), which were incubated with the mineral carriers in HEPES buffer (100 mM HEPES / NaOH, 100 mM NaCI, pH 7.5) at the specified temperature (see the examples of the invention). The increasing shades of gray indicate interaction strengths in 20% increments, ranging from 0% to 100%. The stringency was adjusted by including the specified concentration of additives, which was optimized alongside the reaction temperature.

[0030] Figure 5 shows the interaction of untagged mCherry and the fusion protein mCherry- PHG(693-784)-PCT (SEQ ID NO. 63) with insoluble metal oxides, sulfides, olivines and fluorapatite minerals. The binding affinity data were determined using a bacterial lysate containing 15 pM of the mCherry protein, which was incubated with the specified mineral carriers in HEPES buffer (100 mM HEPES / NaOH, 100 mM NaCI, pH 7.5) at 12 °C as specified in the examples of the invention. The increasing shades of gray indicate interaction strengths in 20% increments, ranging from 0% to 100%. The stringency was adjusted by the addition of the specified concentration of sodium phosphate buffer.

[0031] Figure 6 shows the determination of binding parameters for the interaction of mCherry- PHG(693-784)-PCT (SEQ ID NO. 63) with selected minerals. The binding affinity (A) and binding rate (B) of the protein for y-Fe2O3 (open circle), Fe3O4microparticles (average size > 5 pM) (filled circle) and Srs(PO4)3OH (open triangle) were determined in bacterial lysates containing HEPES buffer (100 mM HEPES / NaOH, 100 mM NaCI, pH 7.5) and 350 mM NaH2PO4 / NaOH (pH 7.5) at 12 °C as described in the examples of the invention.

[0032] Figure 7 sows the influence of the salt concentration on the interaction of mCherry- PHG(693-784)-PCT (SEQ ID NO. 63) with magnetite microparticles (average size > 5 pM). Binding of the protein from bacterial lysates was assayed in the presence of different alkali chlorides (A), ammonium-type chlorides (B), sodium halogenides (C) or sodium salts of oxyacids (D). The binding was probed via absorbance measurements, which were carried out after 60 min of incubation in phosphate-containing buffer (100 mM HEPES / NaOH, 350 mM NaH2PO4 / NaOH, pH 7.5) at 12 °C and subsequent centrifugation (see examples of the invention). (E) The eluotropic potential of the different cat- or anions (compared for the respective chlorides or sodium salts) is demonstrated by the half-maximum concentration (given in mM) required to suppress protein binding. The data were extracted from a Hill-type fit of the curves shown in panels A - D.

[0033] Figure 8 shows the purification of model proteins from bacterial lysates via the MinTag system. (A) Influence of NaCI (open symbols) or KCI (filled symbols) on the interaction of mCherry-PHG(693-784)-PCT (SEQ ID NO. 63) with y-Fe2O3 (circles) or Srs PO^sOH (triangles). The binding was probed via absorbance measurements, which were carried out after 60 min of incubation in HEPES buffer (100 mM HEPES / NaOH, 100 mM NaCI, pH 7.5) containing 200 mM (Srs(PO4)3OH) or 350 mM sodium phosphate buffer (y-Fe2O3) at 12 °C and subsequent centrifugation (see examples of the invention). (B) Batch purification of mCherry-PHG(693-784)-PCT (SEQ ID NO. 63) on both types of mineral in 1 ml scale. Srs PO^sOH was sedimented between the individual steps via centrifugation as described in the examples of the invention, whereas y-Fe2O3 was pelleted via magnetic force (magnet can be seen in the lower left corner after binding and during washing and elution). (C) Ratio of immobilized to unbound mCherry-PHG(693-784)-PCT (SEQ ID NO. 63) during the chromatographic steps shown in panel B. (D) SDS-PAGE analysis of protein-containing fractions obtained during the batch purification of the tagged enzyme FDH-PHG(693-784)- PCT (SEQ ID NO. 81)PHG(693-784)-PCT (SEQ ID NO. 63) (arrow) on Y-Fe2O3. The purification was carried out using the same procedure as in panel B but in a larger scale (23 ml). M, molecular markers (molecular weight indicated at left in kDa).

[0034] Figure 9 shows the induced auto-proteolytic cleavage of tagged mCherry proteins. (A, B) Schematic representation of the protease fusion proteins used in this study, consisting of either a subunit of mCherry (A, PDB 2H5Q) or eGFP (B, PDB 1CV7), a cleavage motif (scissor symbol) recognized during auto-processing, the protease domain with inducible activity (FrpC protease domain: circle in dark grey; CPD protease domain: circle light grey) and a C-terminal affinity tag (e.g. His6 (SEQ ID NO. 14)). The molecular weight of the fragments arising from proteolytic cleavage is indicated. (C and D) SDS-PAGE analysis of the autoprocessing of the FrpC fusion carrying a hexahistidine tag in solution after addition of 10 mM Ca2+(C) or of the CPD fusion proteins carrying a hexahistidine tag after activation by 250 pM or 500 pM IPe (D). The reactions were incubated for the specified time in the presence (+) or absence (-) of the inducers as described in the examples of the invention. Samples containing each 5 pg of the recombinant proteins were loaded to the gels. In panel D, a catalytically inactive variant (NC) of the CPD domain, in which the active site cysteine was exchanged against serine, was used as a negative control. M, molecular markers (molecular weight indicated at left in kDa).

[0035] Figure 10 shows the influence of sugars and sugar alcohols on the interaction of mCherry- PHG(693-784)-PCT (SEQ ID NO. 63) with hydroxyapatites. Binding to the strontium (A, C) or calcium mineral (B, D) was probed in the presence of sugars and sugar alcohols (compounds and their concentrations are specified in the figures) via absorbance measurements, which were carried out after incubation in HEPES buffer (100 mM HEPES / NaOH, 100 mM NaCI (pH 7.5)) containing NaH2PO4 / NaOH (pH 7.5, strontium hydroxyapatite: 230 mM, calcium hydroxyapatite: 200 mM) at 12 °C. The minerals were sedimented by centrifugation prior to the measurements (see examples of the invention).

[0036] Figure 11 shows a multiple sequence alignment of the PHG motifs comprising the consensus sequence of SEQ ID NO. 200. This consensus sequence is absent from the truncation PHG(653-721) (SEQ ID NO. 201) (top row) only, which proved to be deficient in mineral binding.

[0037] Figure 12 shows a multiple sequence alignment of the PHG containing tandemized mineral binding proteins, which comprise the consensus sequence of SEQ ID NO. 200.

[0038] Detailed description of the invention

[0039] In one aspect, the invention provides a method of screening for mineral binding peptides comprising the steps of a. Providing a set of peptides and measuring their binding affinity to said mineral surfaces; b. Comparing the binding affinity of the peptides measured in step a) to the binding affinity of a control peptide; c. Selecting peptides that show an increased binding affinity, preferably an at least 15% increased binding affinity, for the surface of said at least one mineral, when compared to the binding affinity of the control peptide for said surface of said at least one mineral.

[0040] The terms “peptides” or “mineral binding peptides” or “tags” are interchangeably used throughout the specification, claims and figures of this application. The “peptides” or “mineral binding peptides” or “tags” according to the invention are selected from natural and artificial peptides. In a preferred embodiment, said peptides include peptides originating from vertebrate bones, shells of mollusks, bird’s eggs, magnetotactic bacteria, blue mussel shell peptides, and artificial polyionic peptides. Preferably, said peptides are selected from the group consisting of PXXC (SEQ ID NO. 16), N33 (SEQ ID NO. 17), DM P 1(399-437) (SEQ ID NO. 1), Osteocalcin(16-49) (SEQ ID NO. 2), Ovocleidin-17 (SEQ ID NO. 3), PFMG1 (SEQ ID NO. 4), Pif80(912-923) (SEQ ID NO. 5), bKRMP3 (SEQ ID NO. 6), PHG(653-784) (SEQ ID NO. 7), BMSP(836-957) (SEQ ID NO. 9), BMSP(975-1066) (SEQ ID NO. 10), PAT (SEQ ID NO. 11), Aspein_2-20, Aspein_2-25, E(QD)5, E(EN)5, CBP1 (SEQ ID NO. 12), PCT (SEQ ID NO. 13), Maml (SEQ ID NO. 18) and Mms6 (SEQ ID NO. 19), and a peptide that has a sequence identity of at least 75 % to at least one peptide thereof.

[0041] The “peptides” or “mineral binding peptides” or “tags” according to the invention also relate to fragments, analogs and derivatives of such peptides. The terms "fragment", "derivative" and "analog", when referring to a peptide of SEQ ID NOs 1 to 83, means peptides that retain essentially the same mineral binding affinity to a mineral. An analog might, for example, include a proprotein, which can be activated by cleavage of the proprotein portion to produce an active mature peptide according to the invention. The peptides of the present invention may be recombinant peptides, natural peptides or synthetic peptides. The fragment, derivative or analog of a peptide of SEQ ID NOs 1 to 83, may be (i) one in which one or more of the amino acid residues is substituted with a conserved or non-conserved amino acid residue (preferably a conserved amino acid residue) and such substituted amino acid residue may or may not be one encoded by the genetic code, or (ii) one in which one or more of the amino acid residues includes a substituent group, or (iii) one in which additional amino acids are fused to the mature peptide, such as a leader or secretory sequence, or for substrate or complex binding of the mature peptide or a proprotein sequence, or (iv) one in which one or more amino acid residues are inserted as a spacer. Such fragments, derivatives and analogs are deemed to be within the scope of those skilled in the art to provide upon the basis of the teachings herein.

[0042] “Fusion proteins” according to the invention are proteins created through the joining of two or more genes that originally coded for separate proteins. Translation of this fusion gene results in a single or multiple polypeptides with functional properties derived from each of the original proteins. Recombinant fusion proteins are created artificially by recombinant DNA technology for use in biological research or therapeutics. Fusion proteins of the invention also comprise chimeric proteins, which are usually designated as hybrid proteins made of polypeptides having different functions or physico-chemical patterns.

[0043] A “conjugate” or “conjugated protein” according to the invention is a protein that functions in interaction with other (non-polypeptide) chemical groups attached by covalent bonding. Many proteins contain only amino acids and no other chemical groups, and they are called simple proteins. However, these simple proteins may be joined to peptides, other proteins or non- proteinogenic materials such as nucleic acids, lipids, (poly)saccharides, polymers or low- molecular weight functional entities (acids such as forming phosphates or acylations, saccharides, dyes, chemical probes, natural products, immunogenic epitopes, receptor ligands, drugs or toxins) or to posttranslational modifications by means of chemical or enzymatic reaction between the protein and the said reactant, leading to a covalent conjugate with modified properties.

[0044] The peptides, mineral binding peptides or tags of the present invention include the peptides of SEQ ID NOs 1 to 83, as well as peptides which have at least 50 % sequence identity to a peptide of SEQ ID NOs 1 to 83, such as at least 60 % sequence identity to a peptide of SEQ ID NOs 1 to 83, preferably at least 70 % sequence identity to a peptide of SEQ ID NOs 1 to 83 such as at least 75 % sequence identity to a peptide of SEQ ID NOs 1 to 83, more preferably at least 80 % sequence identity to a peptide of SEQ ID NOs 1 to 83 such as at least 85 % sequence identity to a peptide of SEQ ID NOs 1 to 83, and most preferably at least 90 % sequence identity to a peptide of SEQ ID NOs 1 to 83 such as at least 95 %, 96%, 97%, 95 % or 99 % sequence identity to a peptide of SEQ ID NOs 1 to 83.

[0045] These values and ranges of sequence identities apply individually to all sequences in all aspects and embodiments of the present invention.

[0046] The percentage of "sequence identity" is determined by comparing two optimally aligned nucleic acid or polypeptide sequences over a "comparison window" on the full length of the reference sequence. A "comparison window" as used herein, refers to the optimal alignment between the reference and variant sequence after that the two sequences are optimally aligned, wherein the variant nucleic acid or polypeptide sequence in the comparison window may comprise additions or deletions (i.e., gaps) of 20 percent or less, usually 5 to 15 percent, or 10 to 12 percent, as compared to the reference sequences (which does not comprise additions or deletions) for optimal alignment. Identity percentage is calculated by determining the number of positions at which the identical nucleic acid bases or amino acid residues occurs in both sequences to yield the number of matched positions, dividing the number of matched positions by the total number of positions in the reference sequence (i.e., the full length in amino acid or nucleotide) and multiplying the results by 100 to yield the percentage of sequence identity. Two nucleic acid or polypeptide sequences are said to be "identical" if the sequence of nucleotides or amino acids in the two sequences is the same when optimally aligned as described above.

[0047] In a preferred embodiment, the peptide, mineral binding peptide or tag of the present invention as used throughout all aspects of the invention, is not an already known peptide selected from PAT (SEQ ID NO. 11), PCT (SEQ ID NO. 13), and His6 (SEQ ID NO. 14). However, a peptide comprising PAT (SEQ ID NO. 11), PCT (SEQ ID NO. 13) or multiple His residues, such as His6 (SEQ ID NO. 14) or His8 (SEQ ID NO. 15), may be comprised in a fusion protein or conjugate together with a peptide, mineral binding peptide or tag of the present invention.

[0048] The peptides, mineral binding peptides or tags of the present invention may further comprise a cut site accessible for catalytic cleavage, preferably an external proteases, or a selfprocessing domain.

[0049] Fragments or portions of the peptides of the present invention may be employed as intermediates for producing the corresponding full-length peptides by peptide synthesis, or other chemical synthesis with or without the introduction of additional monomeric units (e.g. non-canonical amino scids).

[0050] The invention further relates to a composition comprising a peptide, mineral binding peptide or tag of the present invention. Such a composition may be a buffer solution used in method of protein purification or the like. In a further embodiment, such composition may be a pharmaceutical composition for use of the peptides, mineral binding peptides or tags of the present invention in medical applications.

[0051] “Minerals” or “mineral carriers” according to the invention include (I) all insoluble carbonates, phosphates and hydrogenphosphates, hydroxyapatites, halogenoapatites such as fluorapatites, sulfates, and halogenides containing cations of the elements Li, Be, Mg, Ca, Sr, Ba, Ra, Al, Ga, Ge, In, Sn, Pb, As, Sb and of all transition metals and lanthanides, and (II) oxides, sulfides, selenides and tellurides of the aforementioned elements , (III) insoluble salts of organic acids (e.g. oxalates, fatty acid salts) of the aforementioned elements, (IV) natural and artificial (alumo)silicates, the latter comprising e.g. ino- and phylloslicates, zeolites and especially ceramics, (V) insoluble minerals of the aforementioned elements which contain more than one cation (such as spinels, LiCoPO4, LiFePO4), more than one anion or inorganic anions based on complex formation (e.g. hexacyanoferrates). Minerals or mineral carriers according to the invention further different physical forms thereof, such as macroscopic crystals, mineral powders, amorph solids, sintered minerals, mineral nanomaterials, aqueous sols and gels, and surface-bound minerals (e.g. oxide films) present on solid metals, metal foams, metal powders and nanoparticles. All these materials can be in principle applied in batch, column or microfluidics format or in the shape of (surface-modified) wires, electrodes etc..

[0052] In one embodiment, the oxides, sulfides, selenides and tellurides of subgroup (II) include silica and TiO2.

[0053] In one embodiment, the oxides, sulfides, selenides and tellurides of subgroup (II) do not include silica and TiO2. In a preferred embodiment, minerals or mineral carriers according to the invention are selected from the group consisting of:

[0054] (I) phosphates and carbonates of earth alkali elements such as Ca, Mg, Sr,

[0055] (II) phosphates and carbonates of biologically well tolerated transition metals possessing a coordination chemistry similar to calcium (Zn, Mn, Ce),

[0056] (III) earth alkali sulfates (Ca, Sr),

[0057] (IV) calcium oxalate, and

[0058] (V) readily available, inert silicates (kaolinite, talcum, florisil, SiCh) and oxides (y- AI2O3).

[0059] In a more preferred embodiment, minerals or mineral carriers according to the invention are selected from the group consisting of MgCCh x H2O, CaCCh, SrCCh, MnCCh x H2O, Ce2CC>3 x H2O, (NH4)MgPO4x 6 H2O, Mg3(PO4)2x H2O, Ca5(PO4)3OH, Sr5(PO4)3OH, KMnPO4x H2O, Zns(PO4)2, CePO4x 0.5 H2O, CaSO4x 2 H2O, SrSO4, AI2O3, Fe3O4, CaC2O4x H2O and kaolinite.

[0060] In a further preferred embodiment, minerals or mineral carriers according to the invention do not include phosphates and carbonates of calcium, such as CaCCh, and Cas(PO4)3OH.

[0061] In a yet more preferred embodiment, minerals or mineral carriers according to the invention are selected from the group consisting of FesCU, y-Fe2C>3, FeS, TiC>2, MgAhO4, LiMn2O4, LiCoPO4, LiFePO4, calcium fluorapatite, and strontium fluorapatite.

[0062] The invention further relates to a composition comprising a mineral or mineral carrier of the invention together with a peptide, mineral binding peptide or tag of the invention. Such a composition may be a buffer solution used in method of protein purification or the like. In a further embodiment, such composition may be a pharmaceutical composition for use of the minerals or mineral carriers of the invention together with the peptides, mineral binding peptides or tags of the present invention in medical applications.

[0063] In order to ensure stable binding of the peptides of the invention to the minerals of the invention, the screening method of the invention is performed under conditions (in the screening composition) selected from

[0064] • a pH value in the range between 4.0 and 9.0, preferred 5 - 9, and / or

[0065] • a temperature in the range between 6 °C and 37 °C, and / or

[0066] • a suitable buffer, which is e.g. selected from HEPES buffer, Tris buffer and buffer compositions comprising imidazole / HCI, Tris / HCI, glycine / NaOH, ethanol, urea, polyethylene glycol and / or low amounts of Ni2+. The pH-value, temperature and buffer may be varied in the screening composition, depending on the specific peptide of the invention and / or the specific mineral of invention, which is used in the screening method.

[0067] Thereby,

[0068] (I) the binding affinity of said peptide to said mineral or mineral carrier; and / or

[0069] (II) the stability of the interaction between of said peptide and said mineral or mineral carrier at varying pH and / or temperature, and / or

[0070] (III) the tolerance of the binding against biochemical reagents such as thiols, buffers, chelating agents and denaturants is adjusted.

[0071] Suitably, the screening method of the invention is performed in HEPES buffer (100 mM HEPES / NaOH, 100 mM NaCI, pH 7.5) at a temperature in the range from 6 °C to 37°C, preferably 12 °C.

[0072] In a further embodiment the method of the invention further comprises in step a) the addition of an agent for the suppression of a non-specific interaction between the peptides and the minerals and / or for the increase of the stringency conditions. Said agent preferably selected from NaH2PC>4 / NaOH (pH 7.5), NaHCCh, Na2SC>4 and urea and is preferably added in a concentration between 20 mM and 2000 mM, depending on the specific peptide sequence and the specific mineral used in the screening method.

[0073] Individual concentrations of the additives Na^PCu / NaOH (pH 7.5), NaHCCh, Na2SC>4 and / or urea used in the screening of specific peptide sequences can be obtained from Figures 2, 3 and 4.

[0074] In step c) of the method of the invention there is suitably selected a peptide that shows an at least 15 % increase in the binding affinity to a mineral of the invention compared to a control peptide. The control peptide is e.g. the mCherry protein. Preferably, there is selected a peptide that shows a binding affinity between 20.0 % and 100 %, or between 40 % and 100% to a mineral of the invention. More preferably, there is selected a peptide that shows a binding affinity between 60 % and 100% to a mineral of the invention. Most preferably, there is selected a peptide that shows a binding affinity between 80.0 % and 100 % to a mineral of the invention.

[0075] The “binding” or “binding affinity” as used herein throughout the specification, figure send the is represented by value of the bound protein in (%) percent and can be calculated according to the following formulae: The amount of protein attached to the carrier is calculated using the following formula: n r c i, r ■ -i nn / mean(mineral-containing sample) \

[0076] Percentage of bound protein = 100 — I - * 100 I

[0077] \ mean (control sample) /

[0078] The scaled errors of the binding constants is calculated according to the formulae: absolute error^inding constant relative error (binding constant) * percentage bound protein

[0079] In order to assay mineral binding of tagged proteins, fusions of a mineral binding peptide of the invention with the mCherry protein (or another non-mineral binding indicator protein moiety) are recombinantly produced with conventional methods. Cell lysates containing the respective mCherry fusion protein are then mixed with dry mineral carrier. The reactions, which are carried out at least in duplicates, are then suspended in HEPES buffer up to a standardized volume. Therefore, the decrease in volume caused by the added mineral is taken into account, which is calculated based on the density of the respective mineral. Depending on the mineral carrier, the HEPES buffer added in this step is supplemented with additives such as Na^PCu / NaOH (pH 7.5), NaHCOs, Na2SC>4 or urea, which are either required to suppress unspecific binding or to increase the stringency conditions. The reaction mixtures are incubated under standard conditions and the mineral carriers are then pelleted by centrifugation (2 min, 20,000 g). The absorbance of the supernatants at 595 nm is determined using HEPES buffer as a blank. Samples which do not contain any mineral are used as a negative control. More details of this mineral binding assay are described in the examples of the invention.

[0080] Preferably, in step c) of the method of the invention, a peptide from the group consisting of DMP1 (399-437) (SEQ ID NO. 1), Osteocalcin(16-49) (SEQ ID NO. 2), PFMG1 (SEQ ID NO. 4), Pif80(912-923) (SEQ ID NO. 5), bKRMP3 (SEQ ID NO. 6), PHG(653-784) (SEQ ID NO. 7) BMSP(836-957) (SEQ ID NO. 9), BMSP(975-1066) (SEQ ID NO. 10), PAT (SEQ ID NO. 11), CBP1 (SEQ ID NO. 12), PCT (SEQ ID NO. 13), N33 (SEQ ID NO. 17), and Maml (SEQ ID NO. 18) is selected. More preferably, the peptide selected in step c) is a peptide selected from the group consisting of bKRMP3 (SEQ ID NO. 6), PHG(653-784) (SEQ ID NO. 7), PAT (SEQ ID NO. 11) and PCT (SEQ ID NO. 13). Most preferably, the peptide selected in step c) is PHG(653-784) (SEQ ID NO. 7). Further preferred embodiments of selection step c) of the method of the invention for specific pairs of minerals and peptides of the invention are shown in Table 1 and in Figure 2.

[0081] Table 1 : Preferred embodiments with regard to selection step c) of the method of the invention for specific pairs of minerals and peptides of the invention

[0082] In a further embodiment, the method according to the invention, further comprises the steps d.a.i) truncation of the peptide selected in step c) d.a.ii) Selecting truncated peptides that show an increased binding affinity, preferably an at least 15% increased binding affinity, for the surface of said at least one mineral, when compared to the binding affinity of a control peptide for said surface of said at least one mineral.

[0083] The peptide may be truncated at the N-terminus, at the C-terminus, or both the N-terminus and the C-terminus. In step d.a.ii) of the method of the invention there is suitably selected a truncated peptide that shows an at least 15 % increase in the binding affinity to a mineral of the invention compared to a control peptide. The control peptide used in this embodiment is e.g. the full-length peptide PHG(653-784) (SEQ ID NO. 7). Preferably, there is selected a truncated peptide that shows a binding affinity between 20.0 % and 100 % or between 40 % and 100% to a mineral of the invention. More preferably, there is selected a truncated peptide that shows a binding affinity between 60 % and 100% to a mineral of the invention. Most preferably, there is selected a truncated peptide that shows a binding affinity between 80.0 % and 100 % to a mineral of the invention. Preferably, the truncated peptide is derived from the peptide PHG(653-784) (SEQ ID NO. 7) and is selected from the group consisting of: PHG(677-784) (SEQ ID NO. 22), PHG(693-784) (SEQ ID NO. 30), PHG(708-784) (SEQ ID NO. 24), PHG(722-784) (SEQ ID NO. 25), PHG(733-784) (SEQ ID NO. 26), PHG(693-774) (SEQ ID NO. 23), PHG(693-765) (SEQ ID NO. 28), and PHG(693-754) (SEQ ID NO. 29). More preferably the truncated peptide is a peptide selected from the group consisting of: PHG(653-784) (SEQ ID NO. 7), PHG(693-784) (SEQ ID NO. 30), and PHG(693-765) (SEQ ID NO. 28). Most preferably, the truncated peptide is the peptide PHG(693-784) (SEQ ID NO. 30).

[0084] Further preferred embodiments with regard to selection step d.a.ii) of the method of the invention for specific pairs of minerals and truncated peptides of the invention are shown in Table 2 and in Figure 3.

[0085] Table 2: Preferred embodiments in regard to selection step d.a.ii) of the method of the invention for specific pairs of minerals and truncated peptides of the invention

[0086] In a further embodiment, the method according to invention, further comprises the steps d.b.i) combination of natural and artificial peptides by tandemization, tri- and multimerization of the peptide selected in step c) or the truncated peptide selected in step d.a.ii); d.b.ii) Selecting combination peptides that show an increased binding affinity, preferably an at least 15% increased binding affinity, for the surface of said at least one mineral, when compared to the binding affinity of a control peptide for said surface of said at least one mineral.

[0087] Preferably, in step d.b.i) there is performed a tandemization or multimerization of natural and artificial peptides with the peptide selected in step c) or the truncated peptide selected in step d.a.ii).

[0088] In step d.b.ii) of the method of the invention there is suitably selected a tandemized peptide that shows an at least 15 % increase in the binding affinity to a mineral of the invention compared to a control peptide. The control peptides used in this embodiment are e.g. the individual peptides in their non-tandemized form. Preferably, there is selected a tandemized peptide that shows a binding affinity between 20.0 % and 100 % or between 40 % and 100% to a mineral of the invention. More preferably, there is selected a tandemized peptide that shows a binding affinity between 60 % and 100% to a mineral of the invention. Most preferably, there is selected a tandemized peptide that shows a binding affinity between 80.0 % and 100 % to a mineral of the invention.

[0089] Suitably, the tandemized peptide is selected from the group consisting of PHG(653-784)-PCT (SEQ ID NO. 62), PHG(693-784)-PCT (SEQ ID NO. 63), PHG(693-765)-PCT (SEQ ID NO. 64), PHG(693-784)-bKRMP3 (SEQ ID NO. 66), PHG(653-784)-His8, PHG(653-784)- Pif80(912-923) (SEQ ID NO. 74), PHG(653-784)-PAT (SEQ ID NO. 68), PHG(693-784)-PAT (SEQ ID NO. 69), PHG(693-765)-PAT (SEQ ID NO. 70), PHG(653-784)-Aspein_2-30 (SEQ ID NO. 75), PHG(653-784)-Aspein_2-41 (SEQ ID NO. 76), PHG(653-784)-E(QD)5 (SEQ ID NO. 71) and PHG(653-784)-E(EN)5 (SEQ ID NO. 72).

[0090] Preferred is a tandem peptide selected from the group consisting of: PHG(653-784)-PCT (SEQ ID NO. 62), PHG(693-784)-PCT (SEQ ID NO. 63), and PHG(693-765)-PCT (SEQ ID NO. 64).

[0091] Most preferred is the tandemized peptide PHG(693-784)-PCT (SEQ ID NO. 63).

[0092] Tandemization can be performed by any conventional method, such as genetic engineering and recombinant expression. The tandemized peptides may further comprise a cut site accessible for external proteases or self-processing domain. A suitable method for producing tandemized peptides is described in the examples of the invention. Further preferred embodiments regarding selection step d.b.ii) of the method of the invention for specific pairs of minerals and tandemized peptides of the invention are shown in Table 3 and in Figure 4.

[0093] Table 3: Preferred embodiments in regard to selection step d.b.ii) of the method of the invention for specific pairs of minerals and tandemized peptides of the invention

[0094] In preferred independent embodiments, PHG(693-784)-PCT (SEQ ID NO. 63) binds with > 20 %, preferably > 40 %, more preferably > 80 %, most preferably > 90 % or > 95% to a mineral selected from the group consisting of Fe3O4, y-Fe2O3, FeS, TiO2, MgAhO4, LiCoPO4, LiFePO4, calcium fluorapatite, and strontium fluorapatite under low stringency conditions (in the absence of phosphate), and / or

[0095] PHG(693-784)-PCT (SEQ ID NO. 63) binds with > 20 %, preferably > 40 %, more preferably > 80 %, most preferably > 90 % or > 95% to a mineral selected from the group consisting of Fe3O4, y-Fe2O3, FeS, TiO2, MgAhO4, LiMn2O4 and UCOPO4 under high stringency conditions (in the presence of 350 mM phosphate); wherein the average particle size of the minerals is in the range from 5 nm to 150 pm, preferably in the range from 20 nm to 5 pm, more preferably in the range from 20 nm to 100 nm, most preferably in the range from 20 nm to 40 nm or in the range from 50 nm to 100 nm.

[0096] Further preferred embodiments in regard to the tandemized peptide PHG(693-784)-PCT (SEQ ID NO. 63) for binding to selected minerals under low stringency and high stringency conditions are shown in Figure 5.

[0097] In a further aspect, the invention provides a mineral binding peptide, wherein said mineral binding peptide is selected from the group consisting of: a. a natural or synthetic peptide selected from the group consisting of PXXC (SEQ ID NO. 16), N33 (SEQ ID NO. 17), DM P 1(399-437) (SEQ ID NO. 1), Osteocalcin(16-49) (SEQ ID NO. 2), Ovocleidin-17 (SEQ ID NO. 3), PFMG1 (SEQ ID NO. 4), Pif80(912-923) (SEQ ID NO. 5), bKRMP3 (SEQ ID NO. 6), PHG(653-784) (SEQ ID NO. 7), BMSP(836-957) (SEQ ID NO. 9), BMSP(975- 1066) (SEQ ID NO. 10), PAT (SEQ ID NO. 11), CBP1 (SEQ ID NO. 12), PCT (SEQ ID NO. 13), Maml (SEQ ID NO. 18), and Mms6 (SEQ ID NO. 19); b. a truncated peptide selected from the group consisting of PHG(677-784) (SEQ ID NO. 22), PHG(693-784) (SEQ ID NO. 30), PHG(708-784) (SEQ ID NO. 24), PHG(722-784) (SEQ ID NO. 25), PHG(733-784) (SEQ ID NO. 26), PHG(693- 774) (SEQ ID NO. 23), PHG(693-765) (SEQ ID NO. 28), and PHG(693-754) (SEQ ID NO. 29); and c. a tandemized peptide selected from the group consisting of PHG(653-784)-PCT (SEQ ID NO. 62), PHG(693-784)-PCT (SEQ ID NO. 63), PHG(693-765)-PCT (SEQ ID NO. 64), PHG(693-784)-bKRMP3 (SEQ ID NO. 66), PHG(653-784)- His8, PHG(653-784)-Pif80(912-923) (SEQ ID NO. 74), PHG(653-784)-PAT (SEQ ID NO. 68), PHG(693-784)-PAT (SEQ ID NO. 69), PHG(693-765)-PAT (SEQ ID NO. 70), PHG(653-784)-Aspein_2-30 (SEQ ID NO. 75), PHG(653- 784)-Aspein_2-41 (SEQ ID NO. 76), PHG(653-784)-E(QD)5 (SEQ ID NO. 71) and PHG(653-784)-E(EN)5 (SEQ ID NO. 72).

[0098] Preferably, said mineral binding peptide is selected from the group consisting of a. a natural or synthetic peptide selected from the group consisting of bKRMP3 (SEQ ID NO. 6), PHG(653-784) (SEQ ID NO. 7), PAT (SEQ ID NO. 11) and PCT (SEQ ID NO. 13); b. a truncated peptide of PHG(653-784) (SEQ ID NO. 7) which is selected from the group consisting of PHG(693-784) (SEQ ID NO. 30), and PHG(693-765) (SEQ ID NO. 28); and c. a tandemized peptide selected from the group consisting of PHG(653-784)-PCT (SEQ ID NO. 62), PHG(693-784)-PCT (SEQ ID NO. 63), and PHG(693-765)- PCT (SEQ ID NO. 64).

[0099] Most preferably, the mineral binding peptide of the invention comprises the consensus sequence

[0100] TGSGQTFTYINWGPGNPDNSG (SEQ ID NO. 200) and is selected from PHG(653-784) (SEQ ID NO. 7) or a truncated peptide thereof or a tandemized peptide thereof, i.e. is selected from the group consisting of PHG(653-784) (SEQ ID NO. 7), a truncated peptide of SEQ ID NO: 7, a tandemized peptide comprising the peptide of SEQ ID NO:7, and a tandemized peptide comprising a truncated peptide of SEQ ID NO:7.

[0101] This consensus sequence is absent from the truncation PHG(653-721), which proved to be deficient in mineral binding. PHG(653-721) has the sequence MCEGGWEKFEESCYLFSLSSGTWDSGKTFCEGEGGHLVEISSLAEDNFIRDYVRNRGLIAY ETSDSWIGG (SEQ ID NO. 201).

[0102] Mineral peptides comprising the consensus sequence of SEQ ID NO. 200 show high affinities to certain tested minerals, as it is shown in the working examples hereinbelow.

[0103] In some embodiments, the mineral binding peptide of the invention is a truncated peptide selected from the group consisting of PHG(677-784) (SEQ ID NO. 22), PHG(693-784) (SEQ ID NO. 30), PHG(708-784) (SEQ ID NO. 24), PHG(722-784) (SEQ ID NO. 25), PHG(733-784) (SEQ ID NO. 26), PHG(693-774) (SEQ ID NO. 23), PHG(693-765) (SEQ ID NO. 28), PHG(693-754) (SEQ ID NO. 29), and a peptide that has a sequence identity of at least 75 % to at least one peptide thereof.

[0104] In preferred embodiments, the mineral binding peptide of the invention is a truncated peptide selected from the group consisting of PHG(677-784) (SEQ ID NO. 22), PHG(693-784) (SEQ ID NO. 30), PHG(708-784) (SEQ ID NO. 24), PHG(722-784) (SEQ ID NO. 25), PHG(733-784) (SEQ ID NO. 26), PHG(693-774) (SEQ ID NO. 23), and PHG(693-765) (SEQ ID NO. 28).

[0105] In more preferred embodiments, the mineral binding peptide of the invention is a truncated peptide selected from the group consisting of PHG(677-784) (SEQ ID NO. 22), PHG(693-784) (SEQ ID NO. 30), PHG(693-774) (SEQ ID NO. 23), and PHG(693-765) (SEQ ID NO. 28).

[0106] In some embodiments, the mineral binding peptide of the invention is a tandemized peptide selected from the group consisting of PHG(653-784)-PCT (SEQ ID NO. 62), PHG(693-784)- PCT (SEQ ID NO. 63), PHG(693-765)-PCT (SEQ ID NO. 64), PHG(693-784)-bKRMP3 (SEQ ID NO. 66), PHG(653-784)-His8, PHG(653-784)-Pif80(912-923) (SEQ ID NO. 74), PHG(653- 784)-PAT (SEQ ID NO. 68), PHG(693-784)-PAT (SEQ ID NO. 69), PHG(693-765)-PAT (SEQ ID NO. 70), PHG(653-784)-Aspein_2-30 (SEQ ID NO. 75), PHG(653-784)-Aspein_2-41 (SEQ ID NO. 76), PHG(653-784)-E(QD)5 (SEQ ID NO. 71) PHG(653-784)-E(EN)5 (SEQ ID NO. 72), and a peptide that has a sequence identity of at least 75 % to at least one peptide thereof.

[0107] In preferred embodiments, the mineral binding peptide of the invention is a tandemized peptide selected from the group consisting of PHG(653-784)-PCT (SEQ ID NO. 62), PHG(693-784)- PCT (SEQ ID NO. 63), PHG(693-765)-PCT (SEQ ID NO. 64), PHG(693-784)-bKRMP3 (SEQ ID NO. 66), PHG(653-784)-PAT (SEQ ID NO. 68), PHG(693-784)-PAT (SEQ ID NO. 69), PHG(653-784)-Aspein_2-30 (SEQ ID NO. 75), PHG(653-784)-E(EN)5 (SEQ ID NO. 72).

[0108] In more preferred embodiments, the mineral binding peptide of the invention is a tandemized peptide selected from the group consisting of PHG(653-784)-PCT (SEQ ID NO. 62), PHG(693- 784)-PCT (SEQ ID NO. 63), PHG(693-765)-PCT (SEQ ID NO. 64), and PHG(653-784)-PAT (SEQ ID NO. 68).

[0109] Most preferably, said mineral binding peptide of the invention is selected from the group consisting of a. the natural or synthetic peptide PHG(653-784) (SEQ ID NO. 7); b. the truncated peptide PHG(693-784) (SEQ ID NO. 30); and c. the tandemized peptide PHG(693-784)-PCT (SEQ ID NO. 63).

[0110] The mineral binding peptides of the invention are suitable in certain non-medical applications. Therefore, in further embodiments, the invention relates to methods of using a mineral binding peptide of the invention in (I) affinity purification of target proteins on mineral surfaces, (II) immobilization of proteins fused or conjugated to said mineral binding peptide to a mineral surface, including immobilized enzymes or enzyme cascades and electrobiocatalysis by proteins I transition metal complexes immobilized to conductive minerals, (III) application in carrier-based analytical techniques requiring surface modification, such as ELISA, biosensors, lateral flow assays, surface plasmon resonance, and (IV) application as additives in the food, pharmaceutical and feed sector such as surface-modified minerals as ingredients of nutraceuticals, animal feed products or toothpaste, (V) surface modification with peptides of technical appliances, e.g. enzyme / peptide covered window shields or ceramics, or and (VI) adhesives in biotechnological applications and biological systems.

[0111] The mineral binding peptides of the invention are suitable in certain medical and therapeutic applications. In further embodiments, the invention relates to the mineral binding peptides of the invention for use in methods of targeting in biological systems, i.e. delivery of drugs and therapeutic proteins in the body such as tooth fillings supporting healing, therapy of caries and bone fractures, magnetic delivery of drugs to specific tissues and organs and titanium implants.

[0112] Preferred is the use of PHG(693-784)-PCT (SEQ ID NO. 63) in the aforementioned items (I) to (IV). PHG(693-784)-PCT (SEQ ID NO. 63) not only binds with a high percentage to the carbonate, phosphate and oxidic minerals listed in Fig. 4 and Fig. 5, but also to magnetite, y- Fe2Os (maghemite), TiO2, MgAhO4, and the conductive lithium compounds LiMn2O4 and UCOPO4. The usability thereof is described in the results section of the examples of the invention hereinbelow. Especially preferred is the use of PHG(693-784)-PCT (SEQ ID NO. 63) for binding to magnetic materials maghemite and magnetite, the phosphate mineral strontium hydroxyapatite, and the calcium mineral calcium apatite. The static binding capacity of these carriers (Fig. 6A) differed by one order of magnitude: While maghemite could bind up to 158 mg of PHG(693-784)-PCT (SEQ ID NO. 63) per cm3of solid mineral the binding capacity of magnetite was lower (69 mg protein cm-3) but still exceeded the binding capability of resins commonly used in affinity purification of proteins (e.g. Ni2+-loaded materials). A protein binding of 20 mg PHG(693-784)-PCT (SEQ ID NO. 63) cm-3was observed for strontium hydroxyapatite (see Results section of the examples of the invention).

[0113] In further embodiments, the invention provides a fusion or conjugate of a mineral binding peptide of the invention to (I) proteins with enzymatic activity, (II) proteins with binding properties for other biomolecules (e.g. antibodies, receptors), small molecules or inorganic ions (e.g. heavy metal binding peptides or protein domains), (III) therapeutic proteins, including growth factors, peptide hormones and antimicrobial peptides, (IV) structural proteins to cover, shield or protect the abovementioned types of mineral surfaces (e.g. hydrophobins) or to be used as adapters for components for synthetic biology applications; and the application of chemical and biological conjugation to attach molecules to the mineral binding peptides of the invention, thereby conferring a mineral binding. Such molecules for attachment to the mineral binding peptides of the invention include peptides, proteins, nucleic acids, small molecules (natural products, drugs and synthetic bioactive molecules) and natural and synthetic polymers.

[0114] Such fusions or conjugates can be prepared by any conventional method, such as peptide synthesis, genetic engineering and recombinant expression. Said fusion or conjugates may further comprise a cut site accessible for external proteases or self-processing domain. A suitable method for producing such fusions or conjugates is described in the examples of the invention.

[0115] In a further aspect, the invention relates to the use of minerals or mineral carriers as identified herein in said methods of using the mineral binding peptides of the invention as described hereinbefore. Preferred in said methods of using the mineral binding peptides of the invention are minerals or minerals carriers selected from the group consisting of

[0116] (I) phosphates and carbonates of earth alkali elements such as Ca, Mg, and Sr,

[0117] (II) phosphates and carbonates of biologically well tolerated transition metals possessing a coordination chemistry similar to calcium (Zn, Mn, Ce),

[0118] (III) earth alkali sulfates (Ca, Sr),

[0119] (IV) calcium oxalate, and

[0120] (V) readily available, inert silicates (kaolinite, talcum, florisil, SiCh) and oxides (y- AI2O3).

[0121] In one embodiment, the earth alkali elements of the phosphates and carbonates of subgroup (I) include calcium.

[0122] In one embodiment, the earth alkali elements of the phosphates and carbonates of subgroup (I) do not include calcium.

[0123] More preferred in said methods of using the mineral binding peptides of the invention are minerals or mineral carriers selected from the group consisting of MgCO3x H2O, CaCO3, SrCO3, MnCO3x H2O, Ce2CO3x H2O, (NH4)MgPO4x 6 H2O, Mg3(PO4)2x H2O, Ca5(PO4)3OH, Sr5(PO4)3OH, KMnPO4x H2O, Zn3(PO4)2, CePO4x 0.5 H2O, CaSO4x 2 H2O, SrSO4, AI2O3, Fe3O4, CaC3O4x H2O and kaolinite.

[0124] In a further preferred embodiment, minerals or mineral carriers according to the invention do not include phosphates and carbonates of calcium, such as CaCO3, and Cas(PO4)3OH.

[0125] Yet more preferred in said methods of using the mineral binding peptides of the invention are minerals or mineral carriers selected from the group consisting of Fe3O4, y-Fe2O3, FeS, TiO3, MgAhO4, LiMn3O4, LiCoPO4, LiFePO4, calcium fluorapatite, and strontium fluorapatite.

[0126] In a further aspect, the invention provides a method for eluting the mineral binding peptides of the invention and / or the fusions and / or conjugates as described hereinbefore from a mineral or mineral carrier as identified herein by using an agent selected from the group consisting of

[0127] • sugars and polyols / sugar alcohols such as glycerol, and sugar acids, e.g. ascorbic acid,

[0128] • a salt gradient,

[0129] • inorganic compounds containing Rb, Cs, Br, I, nitrates, phosphates, ammonium, guanidinium and urea derivatives, salts containing cations of Li, Al, earth alkali and lanthanide metals, fluorides, (per)halogenates, borates, nitrites, (hydrogen)carbonates, sulfates, oxoanions containing Si, Ge, As, Se, Sn and transition metals, organic salts such as acetates, mono-, di-, and polycarboxylates, e.g. citrate, sulfonates, phenolates, ionic liquids, and (poly)ionic polymers, and chelating agents such as organic acids, amines, borates, and polydentate ligands.

[0130] Preferred agents for eluting the mineral binding peptides of the invention and / or the fusions and / or conjugates as described hereinbefore from a mineral or mineral carrier as identified herein are selected from the group consisting of NaCI, KCI, RbCI, CsCI, NH4CI, NaBr, Nal, NaNOs, NaOAc, and Na2SO4. Preferably, these salts are used in the range from 50 mM to 1000 mM, more preferably from 100 mM to 900 mM, most preferably from 200 mM to 800 mM or from 200 mM to 700 mM for the elution for eluting the mineral binding peptides of the invention and / or the fusions and / or conjugates as described hereinbefore from a mineral or mineral carrier as identified herein. Further preferred embodiments in regard to elution profiles obtained for these salts and individual concentrations in mM for each of these salts are shown in Fig. 7 and Fig. 8.

[0131] Further preferred agents for eluting the mineral binding peptides of the invention and / or the fusions and / or conjugates as described hereinbefore from a mineral or mineral carrier as identified herein are selected from the group consisting of sugars or their derivatives, such as different mono- and disaccharides (D-glucose, sucrose) and / or polyols, sugar alcohols (D- sorbitol, glycerol) in particular, when the mineral is strontium or calcium hydroxyapatite (see Fig. 10A, B), or sugar acids like ascorbic acid. Preferably, these agents are used in a concentration range from 5 % (w / v) to 40 % (w / v), preferably from 5 % (w / v) to 30 % (w / v), more preferably from 10 % (w / v) to 30 % (w / v), most preferably from 10 % (w / v) to 20 % (w / v) for eluting the mineral binding peptides of the invention and / or the fusions and / or conjugates as described hereinbefore from a mineral or mineral carrier as identified herein. Glycerol prevents the protein binding at concentrations exceeding 20 % (w / v) in both cases (see Fig. 10C, D). The use of sugars or their derivatives in the elution method of invention has the advantage that sugars are biocompatible and - in contrast to many common eluents such as salts or imidazole - do not severely interfere with the stability of the target protein during storage or its use in e. g. medicinal applications. Further preferred embodiments in regard to elution profiles obtained for these sugars and sugar alcohols and individual concentrations in % (w / v) for each of these sugars and sugar alcohols are shown in Fig. 10.

[0132] The afore described method of eluting the mineral binding peptides of the invention and / or the fusions and / or conjugates as described hereinbefore from a mineral or mineral carrier as identified herein is preferably part of a method for affinity purification of proteins in lab scale as well as in preparative scale. The steps of a method for affinity purification of proteins are generally known in the art. Exemplary, by use to the mineral binding peptides of the invention and the minerals or mineral carriers of the invention, a method for affinity purification of proteins comprises the following steps: a) Providing a fusion protein or a conjugate comprising a mineral binding peptide of the invention, and providing a mineral or mineral carrier of the invention; b) Incubating the fusion protein or the conjugate and the mineral or mineral carrier of step a) under suitable conditions and for a time sufficient to enable binding the fusion protein or the conjugate to the mineral or mineral carrier; c) Washing the resulting fusion protein - mineral complex or conjugate - mineral complex of step b), d) Eluting the fusion protein or conjugate, which comprises a mineral binding peptide of the invention, from the mineral or mineral carrier; and, e) Optionally, determining the yield of said fusion protein or conjugate eluted in step d).

[0133] The fusion protein or conjugate preferably comprises a mineral binding peptide is selected from the group consisting of: a. a natural or synthetic peptide selected from the group consisting of PXXC (SEQ ID NO. 16), N33 (SEQ ID NO. 17), DM P 1(399-437) (SEQ ID NO. 1), Osteocalcin(16-49) (SEQ ID NO. 2), Ovocleidin-17 (SEQ ID NO. 3), PFMG1 (SEQ ID NO. 4), Pif80(912-923) (SEQ ID NO. 5), bKRMP3 (SEQ ID NO. 6), PHG(653-784) (SEQ ID NO. 7), BMSP(836- 957) (SEQ ID NO. 9), BMSP(975-1066) (SEQ ID NO. 10), PAT (SEQ ID NO. 11), CBP1 (SEQ ID NO. 12), PCT (SEQ ID NO. 13), Maml (SEQ ID NO. 18), and Mms6 (SEQ ID NO. 19); b. a truncated peptide selected from the group consisting of PHG(677-784)

[0134] (SEQ ID NO. 22), PHG(693-784) (SEQ ID NO. 30), PHG(708-784)

[0135] (SEQ ID NO. 24), PHG(722-784) (SEQ ID NO. 25), PHG(733-784)

[0136] (SEQ ID NO. 26), PHG(693-774) (SEQ ID NO. 23), PHG(693-765)

[0137] (SEQ ID NO. 28), and PHG(693-754) (SEQ ID NO. 29); and c. a tandemized peptide selected from the group consisting of PHG(653- 784)-PCT (SEQ ID NO. 62), PHG(693-784)-PCT (SEQ ID NO. 63), PHG(693-765)-PCT (SEQ ID NO. 64), PHG(693-784)-bKRMP3 (SEQ ID NO. 66), PHG(653-784)-His8, PHG(653-784)-Pif80(912-923) (SEQ ID NO. 74), PHG(653-784)-PAT (SEQ ID NO. 68), PHG(693-784)-PAT (SEQ ID NO. 69), PHG(693-765)-PAT (SEQ ID NO. 70), PHG(653-784)- Aspein_2-30 (SEQ ID NO. 75), PHG(653-784)-Aspein_2-41 (SEQ ID NO. 76), PHG(653-784)-E(QD)5 (SEQ ID NO. 71) and PHG(653-784)- E(EN)5 (SEQ ID NO. 72).

[0138] In a preferred embodiment, said fusion protein or conjugate used in this method comprises PHG(693-784)-PCT (SEQ ID NO. 63) as mineral binding protein.

[0139] The mineral or mineral carrier used in this method is selected from the group consisting of MgCO3x H2O, CaCO3, SrCO3, MnCO3x H2O, Ce2CO3x H2O, (NH4)MgPO4x 6 H2O, Mg3(PO4)2x H2O, Ca5(PO4)3OH, Sr5(PO4)3OH, KMnPO4x H2O, Zn3(PO4)2, CePO4x 0.5 H2O, CaSO4x 2 H2O, SrSO4, AI2O3, Fe3O4, CaC2O4x H2O, kaolinite, , y-Fe2O3, FeS, TiO2, MgAI2O4, LiMn2O4, LiCoPO4, LiFePO4, calcium fluorapatite, and strontium fluorapatite.

[0140] In a preferred embodiment, said mineral or mineral carrier used in this method is selected from the group consisting of y-Fe2O3(magnetite, which allows for a separation by magnetic force the fusion protein or conjugate comprising a mineral binding protein of the invention) and Srs(PO4)3OH (which allows for a separation by centrifugation from a complex medium such as cell lysate after binding of the fusion protein or conjugate comprising a mineral binding protein of the invention). Such a separation step by magnetic force or centrifugation may be performed after step b) and before step c) of the afore described method.

[0141] Suitable conditions for binding the fusion protein or the conjugate to the mineral or mineral carrier according to step b) of the afore described method are e.g. as follows:

[0142] • Incubation in HEPES buffer (100 mM HEPES / NaOH, 100 mM NaCI, pH 7.5),

[0143] • Incubation for 60 min.

[0144] • Incubation at 12 °C,

[0145] • Said HEPES buffer optionally comprising sodium phosphate buffer in the range from 200 mM to 350 mM.

[0146] The washing step c) can be performed several times, e.g. three times, in suitable buffer, such as the HEPES incubation buffer, optionally comprising sodium phosphate buffer in the range from 200 mM to 350 mM. Elution of the fusion proteins or conjugates from the mineral or mineral carrier is performed with an agent described herein, preferably a salt selected from the group consisting of NaCI, KCI, RbCI, CsCI, NH4CI, NaBr, Nal, NaNCh, NaOAc, and Na2SC>4 or a sugar or sugar alcohol (selected from the group consisting of monosaccharides, disaccharides such as D-glucose and sucrose, and sugar alcohols such as D-sorbitol, glycerol).

[0147] Preferred eluents are NaCI, which is suitably used a concentration of 1 M in the elution buffer, and KCI, which is suitably used in a concentration of 2 M in the elution buffer. The elution buffer is e.g. HEPES buffer (100 mM HEPES / NaOH, 100 mM NaCI, pH 7.5).

[0148] Further preferred embodiments in regard to the affinity purification of proteins are described in the examples of the invention and in Fig. 8.

[0149] The affinity purification method of the invention may further comprise the step of cleaving the mineral binding peptide of the invention from the fusion protein or conjugate after elution from mineral or mineral carrier in step d). For these reasons, the fusion proteins or conjugates comprising a mineral binding peptide of the invention may further comprise a cut site accessible for external proteases, or preferably a self-processing domain. More preferred are self-processing domains which are additionally regulated by the presence of inducer molecules, which have to be bound by the protease in order to induce cleavage. Such cut sites or self-processing domains can be introduced in the fusion proteins or conjugates with any conventional method, such as genetic engineering. Preferably said cut site or self-processing domain is placed between the target protein and the mineral binding protein or mineral binding peptide of the invention.

[0150] In preferred embodiment, said self-processing domain is selected from the group consisting of the protease domain FrpC, which is activated by micromolar concentrations of Ca2+ions; and the protease domain CPD, which strictly requires inositol hexakisphosphate (I Pe) for activity.

[0151] The application of such self-processing and inducible domains has at least two advantages to common tag removal protocols used so far: (I) the protease does not have to be produced in an additional recombinant expression, and (II) an internal protease, which can be activated during affinity purification to elute tag-free target protein exclusively, omits the necessity to remove any separately added proteolytic enzyme via subsequent purification steps.

[0152] Further preferred embodiments regarding the use of cut sites and self-processing domains are described in the examples of the invention and in Fig. 9.

[0153] The invention is further described in a non-limiting manner in the following examples.

[0154] Examples of the invention 1. Materials and Methods

[0155] Materials

[0156] The commercial mineral carriers used in this study were purchased from Sigma-Aldrich (St. Louis, USA) with the following exceptions: CaSCU x 2 H2O and y-AhCh were obtained from Merck (Darmstadt, Germany), CaCOs was a product of Roth (Karlsruhe, Germany), kaolinite and magnesium silicate (florisil) were purchased from Fluka (Munchen, Germany), y-Fe2O3 and FeS were products of Alfa Aesar (Ward Hill, USA), and (ZnCOa)2 ■ (Zn(OH)2)3 was obtained from Applichem (Darmstadt, Germany). All other chemicals and solvents were of the highest purity commercially available in Germany.

[0157] Mineral synthesis

[0158] The synthesis of (NH4)MgPO4 6 H2O from MgSO4 (Strahle and Schweda (1995) Hirzel Verlag Stuttgart), strontium hydroxyapatite (Srs(PO4)3OH) from Sr(NO3)2 (George et al. (1987) J. Mater. Sci. 22, 2274), CePO4 ■ 0.5 H2O from CeCh (Hikichi et al. (1978) Bull. Chem. Soc. Jpn. 51 , 3645) and KMnPO4 ■ H2O from MnSO4 (Koleva et al. (2011) Dalton Trans. 40, 7394) was carried out as described in the respective references. All minerals were washed three times with distilled water before drying, which was carried out in a drying oven at 70 °C. The drying process was continued until the weight of the minerals remained constant.

[0159] SrSO4 was precipitated from 1.625 I of Sr(NOs)2 solution (50 mM) at 65 °C. (NH4)2SO4 solution (50 mM, 1.625 I) was added dropwise under stirring within 90 min. The precipitate (yield: 90 % of theory) was washed and dried as described for the synthetic minerals above.

[0160] Fluorapatites were obtained from 1 mmol of the respective hydroxyapatite (502 mg Cas(PO4)3OH or 740 mg Srs(PO4)3OH), which was stirred in a plastic reaction vessel with one equivalent of aqueous NH4F solution (40 mM, 25 ml) at 25 °C. After incubation for 22 hours, the temperature was increased to 65 °C and stirring was continued for further 19 hours. The supernatant from gravity sedimentation was assayed for fluoride by reaction with the colored ferric salicylate (FeSal), using a modification of the assay developed by Dippe and Ulbrich- Hofmann (Dippe and Ulbrich-Hofmann (2009) Anal. Biochem. 392, 169). Therefore, the samples (25 pl) were diluted with an equal amount of water in a microplate, and 200 pl of FeSal reagent (1 mM FeCh and 6 mM sodium salicylate in 500 mM HCOOH / NaOH, pH 4.0) were added. After incubation for 5 min at 25 °C, the absorbance was measured at 490 nm. The fluoride concentration was calculated using a standard curve (0 - 20 mM NH4F) taken under the same conditions, indicating a substitution of hydroxide ions by fluoride of 48.2 % (reaction with Cas(PO4)3OH) I 73.4 % (reaction with Srs(PO4)3OH). The fluorinated minerals were collected by filtration and washed extensively with water. The materials were dried to constant weight at 80 °C, yielding 405 mg of the partially fluorinated calcium mineral 1 674 mg of the partially fluorinated strontium mineral. FeaO4 nanoparticles with an average particle size of 5 nm, which were synthesized according to the procedure described by Antuch and co-workers (Antuch et al. (2019) ChemElectroChem 6, 1567), were a kind gift of Dr. Leslie Reguera (Universidad de la Habana, Cuba).

[0161] 2. Cloning and assembly of expression plasmids

[0162] Codon-optimized DNA fragments encoding mineral-binding peptides, the mCherry protein or protease domains were obtained by gene synthesis (Eurofins Genomics, Luxemburg). Mineralbinding peptides mined from the mollusks Mytilus edulis and Crassostrea gigas were cloned from cDNA obtained from mantle tissues of adult specimens, which were collected in January 2017 at Hooksiel haven (coordinates 53°64’28” N, 8°08’67” E, Germany). cDNA synthesis was carried out using the Revert Aid First Strand cDNA synthesis kit (Thermo Fisher, Waltham, USA) using RNA prepared with Tri reagent (Sigma-Aldrich, St. Louis, USA) as a starting material. The gene encoding formate dehydrogenase (NCBI accession Q08911.1) was amplified from genomic DNA isolated from baker’s yeast (strain pJ69-4A).

[0163] The cloning strategies for assembly of expression constructs I primers used for subcloning into the expression plasmid pET28a(+) (Merck Millipore, Billerica, USA) are summarized in Table 4. All PCR amplifications of DNA were carried out using Phusion high-fidelity polymerase (Thermo Fisher, Waltham, USA) according to the instructions given by the manufacturer. The assembly of plasmids was based on standard protocols for either (I) conventional restriction digest with subsequent ligation by T4 DNA ligase (Thermo Fisher, Waltham, USA), which was performed according to the manufacturer’s instructions, (II) circular plasmid extension cloning (Quan et al. (2009) PLoS One 4, e6441) or (III) Golden Gate based cloning via Bsal-mediated digest followed by ligation (Engler et al. (2008) PLoS One 3, e3647). Site-directed mutagenesis, which was used to introduce I remove stop codons or to modify linker sequences (see Table 4), was achieved via the Quick Change method (https: / / static.igem.Org / mediawiki / 2012 / a / a5 / Site_ Directed_Mutagenesis.pdf).

[0164] Table 4: Plasmids, peptide tags and expression conditions used in this study. Peptide linkers connecting the tag to the C-terminal of the mCherry reporter protein are shown in boldface. Restriction I recombination sites in the oligonucleotide primers used for cloning are underlined. Nucleotides exchanged by site-directed mutagenesis are marked in italics. The synthetic fragments marked with an asterisk were adapted to an optimum codon usage in the expression host E. coli. aa, amino acid residues; Al, auto-inducing medium; CPEC, circular plasmid extension cloning; IPTG, isopropyl-p-D-thiogalacto-pyranoside; LB, Luria-Bertani broth; n.a., not applicable; * codon usage was optimized for an expression in E. coli.

[0165] The protein mCherry has SEQ ID NO. 199.

[0166] 3. Production of cell lysates and polyhistidine-tagged proteins

[0167] The expression of the plasmids was carried out either in E. coli BL21(DE3) or in E. coli Rosetta (DE3) strains (see Table 4). Cells were either grown in Luria-Bertani (LB) medium (10 g I’1peptone, 5 g I’1yeast extract, 10 g I’1NaCI (pH 7.5)) or in auto-inducing medium (Studier (2005) Protein Expr. Purif. 41 , 207). Kanamycin (50 pg ml’1) was used for selection of transformed BL21(DE3) cells and a combination of kanamycin (50 pg ml’1) and chloramphenicol (50 pg ml’1) for selection of transformed Rosetta (DE3) cells. Detailed information on the expression conditions for each individual construct is given in Table 4. In case of cultivation in LB medium, gene expression was induced by addition of isopropyl-p-D-thiogalactoside (final concentrations are given in Table 4 after the cell suspension had reached an optical density at a wavelength of 600 nm of 0.8 to 1 .0.

[0168] Cells were then harvested by centrifugation (10 min, 8,000 g). For production of cell extracts used in binding assays (see section below), the pellet was re-suspended in 4-(2-hydroxyethyl)- 1 -piperazineethanesulfonic acid (HEPES) buffer (100 mM HEPES / NaOH, 100 mM NaCI (pH 7.5)), which was supplemented with DNase I (0.1 mg ml’1, New England Biolabs, Ipswich, USA), lysozyme (10 pg ml’1, Sigma-Aldrich, St. Louis, USA) and MgCh (2 mg I’1). After cell disruption by high-pressure dispersion, the lysate was clarified by centrifugation (30 min, 15,000 g) at 4 °C. The supernatant was then dialyzed three times against a hundredfold volume of HEPES buffer to remove low molecular weight compounds. The solutions were standardized to a concentration of 30 pM of recombinant protein by dilution with HEPES buffer. Cells expressing the polyhistidine-tagged proteins mCherry-Hise, mCherry-FDH-His6 (SEQ ID NO. 80), mCherry-CPD-His6, mCherry-FrpC-His6 (SEQ ID NO. 83), eGFP-CPD-His6 (SEQ ID NO. 82) and eGFP-FrpC-His6 (SEQ ID NO. 83) were re-suspended in an imidazole-containing buffer (50 mM HEPES / NaOH, 300 mM NaCI, 10 mM imidazole, 10 % (w / v) glycerol (pH 7.5)). After high-pressure dispersion or sonication, the recombinant proteins were purified from the lysates by IMAC as described elsewhere (Dippe et al. (2019) Adv. Synth. Catal. 361 , 5363). All protein samples were stored at -80 °C until use.

[0169] Protein determination

[0170] The concentration of the purified proteins mCherry-Hise, mCherry-CPD-Hise, mCherry-FDH- His6 (SEQ ID NO. 80), eGFP-CPD-His6 (SEQ ID NO. 82) and eGFP-FrpC-His6 (SEQ ID NO. 83) and of commercial standards (alcohol dehydrogenase and lysozyme, both obtained from Sigma-Aldrich, St. Louis, USA) was determined using Bradford reagent (Roth, Karlsruhe, Germany) according to the instructions of the manufacturer. Bovine serum albumin (BSA, 0 - 80 pg ml’1) was used for calibration of the assay. The concentration of all other chromogenic proteins was assessed by measuring the absorbance at 595 nm. The amount of recombinant protein was calculated using the extinction coefficient of the mCherry chromophore (see section below) according to the Beer-Lambert Law.

[0171] Determination of the extinction coefficient of the mCherry chromophore

[0172] The absorbance of serial dilutions of purified mCherry-Hise protein (0 - 500 mg ml’1in HEPES buffer, concentration determined by the Bradford assay as described above) was measured at a wavelength of 595 nm. The extinction coefficient of the protein (58,730 M’1cm’1) was calculated from the linear fit of the absorbance plotted against the protein concentration and its molecular weight (28.456 kDa).

[0173] Sodium dodecyl sulfate gel electrophoresis (SDS-PAGE)

[0174] Protein separation was carried out in acrylamide gels with 12.0% or 12.5 % (w / v) crosslinking (Laemmli (1970) Nature 227, 680). Protein bands were visualized by staining with Coomassie brilliant blue R250 (Meyer and Lambert (1965) Biochim. Biophys. Acta 107, 144). Protein markers from Fermentas (Waltham, USA) were used as a molecular weight reference.

[0175] Assay for mineral binding

[0176] First the binding conditions of the mineral binding assay were optimized in order to reduce unspecific protein binding to the minerals. Therefore, cell lysates (50 pl) containing untagged mCherry (30 pM) were mixed with 25 % (w / v) suspension of the mineral carriers (40 pl) in HEPES buffer. In order to increase the stringency conditions, 10 pl of additive solution with increasing amounts of additive (0 - 1 M Nab^PO^NaOH (pH 7.5), NaHCCh, Na2SC>4, MgCh, CaCh, NaNOs or oxalic acid / NaOH (pH 7.5) in HEPES buffer) were added. The reactions, which were performed in duplicates, were incubated under standard conditions (incubation for one hour under rotation (35 rpm) at 10 °C). The mineral carriers were pelleted by centrifugation (2 min, 20,000 g) and the absorbance of the supernatants at 595 nm was determined using HEPES buffer as a blank. Samples which were supplemented with HEPES buffer instead of additive solution were used as a control.

[0177] In order to assay mineral binding of tagged proteins, cell lysates (200 pl) containing the respective mCherry fusion protein (30 pM) were mixed with 48 mg of dry mineral carrier. The reactions, which were carried out at least in duplicates, were then filled up to a total volume of 400 pl by addition of HEPES buffer. Therefore, the decrease in volume caused by the added mineral was taken into account, which was calculated based on the density of the respective mineral. Depending on the mineral carrier, the HEPES buffer added in this step was supplemented with the additives Na^PCu / NaOH (pH 7.5), NaHCCh, Na2SC>4 or urea, which were either reguired to suppress unspecific binding (type of additive and its final concentration are specified in Fig. 2 and 3) or to increase the stringency conditions (specified in Figs. 4 - 8). The reactions were incubated under standard conditions as described above and the mineral carriers were then pelleted by centrifugation (2 min, 20,000 g). The absorbance of the supernatants at 595 nm was determined using HEPES buffer as a blank. Samples which did not contain any mineral were used as a negative control.

[0178] Mineral binding of peptides at different pH values was measured using the standard assay conditions, except that the buffer system was exchanged against 100 mM 2-(N- morpholino)ethanesulfonic acid (MES) / NaOH (pH 5.0 - 7.0), HEPES / NaOH (pH 6.5 - 8.5), or 4-(2-hydroxyethyl)-piperazine-1 -propanesulfonic acid (HEPPS) / NaOH (pH 7.5 - 9.0) in regard to the desired buffer range. All reaction buffers were adjusted to an egual ionic strength of 200 mM by addition of NaCI. The dependence of the binding affinity on the reaction temperature was determined under the same buffer conditions used in the standard assay but with incubation for one hour at non-standard reaction temperatures (6 - 37 °C).

[0179] Binding of commercial tag-free proteins to selected minerals was measured in triplicate reactions (400 pl) containing 30 pM of protein (alcohol dehydrogenase, BSA or lysozyme), 48 mg of mineral (Fe3C>4, kaolinite, KMnPC>4 ■ H2O, Cas(PO4)3OH or Srs(PO4)3OH), and HEPES buffer with selected salt additives (mineral-specific type of additive and its final concentration are listed in Fig. 2). After incubation under standard conditions (see above), the reactions were centrifuged (2 min, 20,000 g) and the protein concentration in the supernatant was determined using the Bradford assay (see above). For calculation of the binding affinity of the proteins, control samples which did not contain any mineral were used.

[0180] Calculation of binding parameters

[0181] The amount of protein attached to the carrier was calculated using the following formula: n A - > -i nn / mean(mineral-containing sample) \

[0182] Percentage of bound protein = 100 — I - * 100 I

[0183] \ mean (control sample) /

[0184] The scaled errors of the binding constants were calculated according to the formulas: absolute crror binding constant') relative error (binding constant * percentage bound protein

[0185] 4. Analytical scale purification of tagged mCherry under batch conditions

[0186] For purification experiments on maghemite, the mineral (20 mg) was equilibrated with 1 ml of maghemite buffer (100 mM HEPES / NaOH, 100 mM NaCI, 350 mM NaH2PO4 / NaOH (pH 7.5)) by 10 min of incubation under rotation (35 rpm). The mineral was subsequently sedimented using a magnetic rack. The buffer was removed and 1 ml of bacterial lysate containing mCherry-PHG(693-784)-PCT (SEQ ID NO. 63) (15 pM in maghemite buffer) was added. After further incubation for 10 min under rotation (35 rpm) at 25 °C, the maghemite was again sedimented magnetically. After removal of the supernatant the material was washed three times with 1 ml of maghemite buffer. The mineral-bound protein was eluted by addition of 1 ml of KCI-based elution buffer (100 mM HEPES / NaOH, 350 mM NaH2PO4 / NaOH, 2 M KOI (pH 7.5)).

[0187] Purification on strontium hydroxyapatite was performed in the same reaction scale, but differed in the sedimentation of the material between the individual purification steps which was performed by centrifugation (10 s, 2,000 g). First, 80 mg of the material was equilibrated with 1 ml of apatite buffer (100 mM HEPES / NaOH, 100 mM NaCI, 200 mM NaH2PO4 / NaOH (pH 7.5)). After 10 min of incubation under rotation (35 rpm), the mineral was pelleted and mixed with 1 ml of bacterial lysate containing mCherry-PHG(693-784)-PCT (SEQ ID NO. 63) (15 pM in apatite buffer). After binding of the protein for one hour at 25 °C, the apatite was washed three times with 1 ml of apatite buffer. Elution was carried out by incubation of the mineral in 1 ml of NaCI-containing elution buffer (100 mM HEPES / NaOH, 200 mM NaH2PO4 / NaOH, 1 M NaCI (pH 7.5)).

[0188] For both types of carrier material, the elution efficiency was probed for four different reaction times (0, 2, 5, 10 or 15 min). For each incubation time, three independent reactions were performed. During each purification step, samples of the supernatants were taken. After a second centrifugation step (1 min, 20,000 g), which was carried out to remove traces of remaining minerals, the samples were analyzed for presence of the mCherry fusion protein by both determination of the absorbance at a wavelength of 595 nm and SDS-PAGE (see above). Prior to the electrophoretic separation, the samples were desalted by extensive dialysis against 10 mM Tris / HCI (pH 7.5). The amount of recovered target protein was calculated using its experimentally determined extinction coefficient (see above).

[0189] 5. Preparative-scale batch purification of tagged FDH on maghemite

[0190] Cells expressing the fusion enzyme FDH-PHG(693-784)-PCT (SEQ ID NO. 81)PHG(693-784)- PCT (SEQ ID NO. 63) were lysed in HEPES buffer (100 mM HEPES / NaOH, 100 mM NaCI (pH 7.5)) as described above. 15 ml of the clarified lysate were adjusted to maghemite buffer conditions by addition of 8.08 ml phosphate stock solution (1 M NaH2PO4 / NaOH (pH 7.5) in HEPES buffer). Maghemite (185 mg), which was equilibrated with maghemite buffer as described in the previous section, was added. Binding of the tagged protein was carried out at 25 °C under vigorous mixing using a glass stirrer. After 30 min of incubation, the maghemite was sedimented by magnetic force using a NdFeB magnet (N35M, 20 x 19 x 4 mm, magnetic flow: 79.5 -10'6Wb, Conrad Electronics, Hirschau, Germany). The supernatant was removed and the mineral was washed three times with each 23 ml of maghemite buffer. Elution was initiated by suspension of the mineral in 11.5 ml of elution buffer(100 mM HEPES / NaOH, 350 mM NaH2PO4 / NaOH, 2 M KCI (pH 7.5)). After 30 min of incubation at 25 °C, the supernatant from magnetic separation was desalted by extensive dialysis against HEPES buffer.

[0191] Samples (2 pl) of the supernatants from each purification step or 5 pg of purified protein were analyzed by SDS-PAGE (see above). The FDH activity of the supernatants was assayed spectrophotometrically in cuvettes with a light path of 1 cm. The reactions (400 pl) contained HEPES buffer, 300 mM sodium formate and 100 pM NAD+. The increase in absorbance at 339 nm was continuously recorded in intervals of 10 s for a total reaction time of 15 min. From the linear part of the progress curve, the enzyme activity was calculated using the absorption coefficient of NADH from literature (E = 6220 M-1cm-1) (see “Photometric detection of nicotinamide adenine dinucleotides" found under http: / / aatbio.com / protocol / A1420d1.pdf).

[0192] Column purification of tagged mCherry on strontium hydroxyapatite A cartridge (5 ml, PureCube compact cartridge, Cube Biotech) was filled with synthetic Srs(PO4)3OH via the slurry method, for which a suspension of the apatite in distilled water was used. The experiment was carried out with a constant flow rate of 1 ml min-1using a peristaltic pump. The column was equilibrated with apatite buffer (see section above) and loaded with 60 ml of a bacterial lysate containing mCherry-PHG(693-784)-PCT (SEQ ID NO. 63) (29.4 pM in apatite buffer). The column was subsequently washed with 20 ml apatite buffer and the mCherry protein was recovered by application of NaCI-containing elution buffer (100 mM HEPES / NaOH, 350 mM Nab^PO^NaOH, 2 M KCI (pH 7.5)). Regeneration of the material was achieved by treatment with aqueous NaOH (100 mM), followed by extensive washing with distilled water.

[0193] 6. Cleavage of the tag by self-processing protease domains

[0194] The protease fusions mCherry-CPD-HiSe, mCherry-FrpC-His6 (SEQ ID NO. 83), eGFP-CPD- His6 (SEQ ID NO. 82) and eGFP-FrpC-His6 (SEQ ID NO. 83) were produced as described above. The integrity and functionality of the fusions was confirmed for purified proteins in buffer (100 mM HEPES / NaOH (pH 7.5)). To activate the self-processing activity of the protease domains, D-myo-inositol hexakisphosphate (0 - 500 pM, dodecasodium salt, Sigma-Aldrich, St. Louis, USA) was added in case of mCherry-CPD-Hise and eGFP-CPD-His6 (SEQ ID NO. 82), whereas CaCh (0 - 250 mM) was added in case of mCherry-FrpC-His6 (SEQ ID NO. 83) and eGFP-FrpC-His6 (SEQ ID NO. 83). All cleavage processes were carried out at 23 °C for defined reaction times (0.25, 0.5, 1 , 2 and 18 h) and analyzed by SDS-PAGE (see above).

[0195] Results

[0196] Initial screening to identify peptides binding to (bio)mineral surfaces

[0197] A set of peptides with potential affinity for mineral surfaces (Fig. 2) was selected to perform an initial screening for such an activity. The rationale behind this strategy was to identify candidates for a further optimization and to compare their affinity under standardized conditions. A first subset of sequences - including peptides originating from vertebrate bones (PXXC (SEQ ID NO. 16), N33 (SEQ ID NO. 17), DMP1 (399-437) (SEQ ID No.1), osteocalcin), shells of mollusks (PFMG1 (SEQ ID NO. 4), Pif80-11), bird’s eggs (ovocleidin 17), or magnetotactic bacteria (Maml (SEQ ID NO. 18), Mms6 (SEQ ID NO. 19)) - was chosen from literature studies due to their involvement in biomineralization in the source organisms. All peptides which were known to be inactive without any post-translational modifications were excluded from the present study. A second type of candidate peptides was obtained by bioinformatic analysis. Many mollusk proteins with mineral-binding properties are too large to be used as a tag, as they usually contain multiple additional domains with different function. Thus, two of these proteins were analyzed (including the blue mussel shell protein (BMSP) from M. edulis (Suzuki et al. (2011) ChemBioChem 12, 2478) and a homologue of perlucin (Lin et al. (2013) Gene 526, 210) from the invasive oyster C. gigas found by genome mining) for the presence of mineral binding subdomains. A total number of four novel peptide candidates (named BMSP(613-837), BMSP(836-957) (SEQ ID NO. 9), BMSP(975-1066) (SEQ ID NO. 10) and perlucin homologous gene (PHG)(653-784)) could be identified from these proteins by a comparison with published sequences. The aspeins (Tsukamoto et al. (2004) Biochem. Biophys. Res. Commun. 320, 1175) were also analyzed for such motifs, yielding two acidic peptides (aspein(2-20) and (2-36)) which were included in the screen. A third subset of sequences comprised artificial tags, which were taken from literature (CBP1 (SEQ ID NO. 12)) or were added due to their polyanionic (PAT (SEQ ID NO. 11)) or -cationic (PCT (SEQ ID NO. 13)) charge. The latter peptides consisted of stretches of eleven evenly charged amino acid residues (glutamates and aspartates in PAT (SEQ ID NO. 11), arginines and lysines in PCT (SEQ ID NO. 13)). As a control, a common octahistidine tag was also included in the set of artificial tags.

[0198] In order to measure binding, the selected peptides were fused to mCherry, a monomeric protein which carries an imidazolone chromophore (Subach and Verkhusha (2012) Chem. Rev. 112, 4308). The binding of such chromogenic mCherry fusion proteins to a mineral surface can be simply determined due to their inherent absorbance at a wavelength of 595 nm: the recombinant protein is incubated with the mineral, which is then separated by centrifugation. Subsequently, the supernatant is used to access the concentration of free (unbound) mCherry spectrophotometrically. A further advantage of mCherry proteins is their efficient production in the heterologous host E. coli, certainly because the maturation of their internal imidazolone chromophore proceeds similarly fast as in related chromoproteins such as green fluorescent protein (Balleza et al. (2018) Nat. Methods 15, 47). Indeed, the heterologous production of most of the chromogenic proteins fused to the selected potential mineral-binding peptides resulted in high amounts of recombinant protein. However, three of the fused peptides (BMSP(613-839) (SEQ ID NO. 8), Aspein(2-30) (SEQ ID NO. 20), Aspein(2- 36) (SEQ ID NO. 21)) did not result in sufficient quantities of soluble fusion proteins in repeated expression experiments. A strong formation of inclusion bodies was observed in these cases (data not shown). This effect suggests that these peptides cause irreversible aggregation of the fusion partner, and are therefore only poor candidates for the development of a universal mineral-binding tag. Another criterion for selection of the tags for further optimization was their proteolytic stability. Thus, all mCherry fusion proteins were applied as crude bacterial lysates in the binding assays. This experimental strategy directly excludes tags which are not functional after proteolytic degradation, which may happen during recombinant protein production or cell disruption and might probably not be detected via analysis of the protein integrity by SDS-PAGE due to the low molecular weight of the tags.

[0199] The cell lysates containing a standardized amount of mCherry-peptide fusion proteins were then tested for interaction with a set of minerals. Apart from common biominerals found in nature (namely calcium hydroxyapatite, CaCOs and FeaO^, further inorganic carriers were also included in this selection which were chosen due to both a low solubility in water and no or marginal toxicity. In particular, the following minerals were tested: (I) phosphates and carbonates of earth alkali elements other than calcium (Mg, Sr), (II) phosphates and carbonates of biologically well tolerated transition metals possessing a coordination chemistry similar to calcium (Zn, Mn, Ce), (III) earth alkali sulfates (Ca, Sr), (IV) calcium oxalate, and (V) readily available, inert silicates (kaolinite, talcum, florisil, SiCh) and oxides (Y-AI2O3). However, a number of these different materials already showed a distinguished binding of recombinant mCherry without any fused peptide. As this strong non-specific interaction might have led to an improper selection of mineral-binding peptides in the screening, a suppression of this effect by adjustment of stringency conditions was necessary. As the attachment of many peptides to mineral surfaces seems to be mainly driven by ionic interactions (Hunter et al. (2010) Langmuir 26, 8639; Fukuta et al. (2014) Colloids Surf. B Biointerfaces 118, 25; Mao et al. (2019) Langmuir 35, 5911), different salt additives were tested for their suitability as stringencyadjusting agents. Therefore, the additives were selected according to their chemical compatibility with each mineral used in this study, i.e. for their inability to react with the respective carrier. For example, sodium phosphate buffers were used in case of phosphate minerals. In contrast, NaHCCh or Na2SO4 were added to carbonate minerals which otherwise tend to form the more insoluble phosphates as an undesired surface modification. All additives were added at different concentrations to the reactions, which facilitated the selection of an optimum concentration to be used in the screening procedure (Fig. 2). Noteworthy, the carbonates of Mg and Sr showed a significant protein binding (40.9 ± 0.9 and 42.4 ± 0.2 % bound mCherry) under stringent conditions (Fig. 2). This effect was even more pronounced in case of silica and the silicates talcum and florisil. A suppression of this unspecific protein binding was not possible for these minerals, which showed strong interaction with mCherry even at high concentrations of all additives tested (data not shown). The three silicates were therefore excluded from further studies.

[0200] After optimization of the reaction conditions, the affinity of the mCherry fusion proteins for the remaining eighteen minerals which are potentially suitable as carriers was determined (Fig. 2). Interestingly, fusions of mCherry to the peptides from vertebrate source, the bacterial peptides Maml (SEQ ID NO. 18) and Mms6 (SEQ ID NO. 19) or the two mollusk peptides mined from M. edulis (BMSP(836-957) (SEQ ID NO. 9) and BMSP(975-1066) (SEQ ID NO. 10)) did not show a significantly increased affinity (defined here as an increase by at least 15.0 % relative to the untagged mCherry control) to all tested minerals (Fig. 2)#. An exception was the interaction of the DMP1 peptide with CaCCh, SrCCh and (NH4)MgPO4 ■ 6 H2O (which was slightly increased by 29.2, 19.9 and 17.7 %), and of the Maml (SEQ ID NO. 18) peptide with kaolinite, which was increased by 17.1% (Fig. 2). In contrast, the peptides obtained from oyster shells proved to be highly potent mineral binders. Especially the two candidates bKRMP3 (SEQ ID NO. 6) and PHG(653-784) (SEQ ID NO. 7) showed an excellent affinity for a number of carriers (Fig. 2). In particular, bKRMP3 (SEQ ID NO. 6) was capable of binding eight of the eighteen tested minerals with increased affinity. PHG(653-784) (SEQ ID NO. 7) showed an even broader scope for inorganic materials and did bind to all the chemically diverse carriers tested in this study, exceeding protein binding values of 80 % in case of hydrated MgCOa, three phosphates (strontium hydroxyapatite, KMnPO4 ■ H2O and CePO4 ■ 0.5 H2O), magnetite (Fe3O4) and kaolinite (Fig. 2). A comparably high binding affinity was observed for the artificially designed polycationic peptide PCT (SEQ ID NO. 13), which bound to six of the tested minerals with more than 80 % binding affinity but showed a more pronounced selectivity for the mineral carrier (Fig. 2). Interestingly, its anionic counterpart PAT (SEQ ID NO. 11) was far less effective in mineral binding and showed comparable interaction values only with SrCOa and calcium hydroxyapatite (Fig. 2). This effect was even more pronounced in case of the hexahistidine tag, which is only slightly positively charged under the assay conditions. Fusion of mCherry to the hexahistidine tag resulted in an increased affinity for SrCOa, kaolinite and Zna(PO4)2 only (Fig. 2). The binding of the hexahistidine tag to Zn minerals might be explained by its high affinity for transition metals, which had been also described for IMAC resins loaded with Zn2+ions (Riguero et al. (2020) J. Chromatogr. A 1629, 461505). Thus, the amount and charge of ionic amino acid residues seems to play a distinguished role in the interaction of peptides with mineral surfaces - a mechanism which was already proposed in earlier studies on mineralbinding proteins.

[0201] In order to proof that the observed mineral binding was a specific feature of the fused peptides, it was additionally tested if three commercial proteins which do not possess any tag were capable of binding to selected minerals under the same conditions. BSA, yeast alcohol dehydrogenase (ADH) and chicken lysozyme were chosen for this purpose due to their different isoelectric points (BSA: 4.7, ADH: 5.4, lysozyme: 9.7), which might influence surface interaction via charge-dependent association. As electrophoretically homogeneous preparations of the proteins did not show a significant binding to hydroxyapatites, KMnPO4 ■ H2O, magnetite and kaolinite, it was concluded that the interaction found for some of the mCherry-peptide fusions (Fig. 2) is caused by the affinity of the respective tags for specific minerals. Dissection of the mineral-binding motif of the PHG peptide by truncation

[0202] The PHG peptide, which exhibited the best mineral-binding properties among all candidates tested in the aforementioned screening (Fig. 2), consists of 132 amino acid residues. However, it is unknown if the entire peptide is involved in the mineral binding process or if a shorter, specific binding motif is sufficient for the interaction with the mineral surface. In order to probe a potential mineral-binding site, the peptide PHG(653-784) (SEQ ID NO. 7) (in the following designated “full-length PHG”) was dissected by truncation. First, the peptide was truncated stepwise from its N-terminus. While binding to all tested mineral carriers was not significantly reduced for the variant PHG(693-784) (SEQ ID NO. 30) which was devoid of 40 N-terminal residues further depletions increasingly reduced the peptides mineral binding affinities with strongest effects observed for the variant PHG(733-784) (SEQ ID NO. 26) which was depleted by the first 80 residues. (Fig. 3). As a second step of the dissection procedure, the truncated but still fully functional variant PHG(693-784) (SEQ ID NO. 30) was successively shortened from its C-terminus. The mineral binding was not significantly influenced by depletion of 19 residues (PHG 693-765) but was dramatically reduced when 30 of the C-terminal residues were deleted (PHG 693-754) (Fig. 3). Thus, it was concluded that only the central part of the full-length PHG peptide (namely the residues 693 to 765) is sufficient to exert mineral binding. As a truncation of the peptide also did not influence its efficient production in recombinant E. coli, variants of full-length PHG consisting of the mineral-binding motif only are optimum for an application in biotechnology.

[0203] Improvement of the binding affinity of the peptides by tandemization

[0204] Subsequent to the initial identification of mineral-binding peptides by screening (Fig. 2), it was aimed to further improve the binding properties of the best binders. The rationale behind this strategy was to obtain optimized peptides with superior affinity, which is a prerequisite for an application as tags in protein purification or immobilization. Such applications often include binding orwashing steps at high-salt (i.e. high-stringency) conditions, in which literature-known peptides but also the initially identified candidates cannot be used due to insufficient affinity or salt tolerance. For the affinity improvement, the five peptides PHG(653-784) (SEQ ID NO. 7), bKRMP3 (SEQ ID NO. 6), Pif80-11, PCT (SEQ ID NO. 13) and PAT (SEQ ID NO. 11) were chosen due to their broad affinity scope for different mineral carriers. In addition, the common hexahistidine tag was included in the study. These different peptides were fused to the C- terminus of mCherry by tandemization in a combinatorial manner. Therefore, two of the candidates were connected by a short linker consisting of two additional amino acids (sequences are listed in Table 4). However, the heterologous production of a number of these triple fusion proteins proved to be problematic. Especially most of the tandem peptides including bKRMP3 (SEQ ID NO. 6) were prone to aggregation, yielding only very low amounts of recombinant protein and excluded from further studies. The same effect was observed for tandems involving aspein sequences, which already proved to be difficult in recombinant production even as unmodified peptides (see Section “Initial screening to identify peptides binding to (bio)mineral surfaces”) but were additionally included in the tandemization studies due to a possible solubility enhancement.

[0205] In order to properly determine the mineral binding of the tandemized tags, the stringency conditions had to be increased. Therefore, the binding achieved with the parent peptides in the initial screen (Fig. 2) was reduced by addition of higher amounts of salt additives in a second cycle of stringency optimization. For five of the selected minerals (SrCCh, hydroxyapatites, KMnPC>4 ■ H2O, magnetite), a higher amount of the salt components used in the initial screen proved to be efficient. However, the strong binding to kaolinite could only be suppressed by the presence of urea. Using these more selective reaction conditions, the binding of the tandem tags and their parent peptides on selected minerals (Fig. 4) was compared. In general, tandemization of the full-length PHG peptide and its truncated variants increased its affinity if fused to a polyionic sequence. Especially tandemization to the polycationic tag PCT (SEQ ID NO. 13) strongly increased the affinity towards phosphate-based minerals and magnetite when compared to both parent peptides while no significant changes were observed in their binding affinity towards earth alkali carbonates and kaolinite (which were similar to those of the unmodified PCT (SEQ ID NO. 13) tag (Fig. 4)). To a lesser extent, fusion of PHG(693-784) (SEQ ID NO. 30) to bKRMP3 (SEQ ID NO. 6) resulted in an increased affinity for two of the phosphate minerals, magnetite and kaolinite (Fig. 4). In contrast, addition of a hexahistidine tag to full-length PHG did not result in a significantly changed affinity of the PHG subpart, whereas the tandem PHG(653-784)-Pif-80-11 showed binding properties intermediate of the two parent peptides (Fig. 4). Finally, tandemization of full-length PHG with the polyanionic peptides (PAT (SEQ ID NO. 11) and aspein sequences) led to specific changes in the selectivity for certain minerals: Fusion to PAT (SEQ ID NO. 11) reduced the ability to interact with SrCOa but strongly increased the affinity to magnetite and kaolinite, while fusion especially to the extended aspein(20-25) peptide slightly increased the binding of PHG to CaCOs, SrCOa and kaolinite (Fig. 4). In order to gain insight into the charge dependence of these particular interactions, an additional mutagenetic screen of the tandem PHG(653-784)-PAT (SEQ ID NO. 68) was performed. Reduced negative charge implemented by exchange of its glutamate residues against glutamine (tandem PHG(653-784)-E(QD)5 (SEQ ID NO. 71)) or by exchange of the aspartates against asparagine (tandem PHG(653-784)-E(EN)5 (SEQ ID NO. 72)) caused a strongly decreased affinity for all minerals except for kaolinite in case of the latter tandem (Fig. 4). Taken together, tandemization with a second peptide proved to be an efficient strategy to engineer mineral-binding peptides such as PHG for either higher affinity or better selectivity for a certain mineral carrier. Characterization of the binding of the optimized tag PHG(693-784)-PCT (SEQ ID NO. 63) to selected minerals

[0206] After optimization of the mineral affinity of the PHG peptide by assignment of the mineralbinding motif and tandemization (see sections above), the superior tandem PHG(693-784)- PCT (SEQ ID NO. 63) was characterized comprehensively for its affinity scope to mineral interaction partners. Apart from the carbonate, phosphate and oxidic minerals listed in Fig. 4, which were found to be strongly bound by the tag, binding to a set of chemically more diverse minerals was tested in order to provide insight into the scope of applications accessible through the peptide. In particular, a number of metal oxides (ferric and ferrous oxides and sulfides, alumina, TiO2), two spinel-type compounds (MgAhO4, LiMn2O4), and mixed phosphate minerals (LiCoPO4, LiFePO4 and partially fluorinates apatites) were included in the selection. The binding of mCherry-PHG(693-784)-PCT (SEQ ID NO. 63) to these minerals, which was measured at low and high stringency as described in the sections above, was compared with the association properties of the untagged protein determined under the same reaction conditions (Fig. 5). Similar to its affinity for magnetite, the tagged protein was able to strongly interact with y-Fe2O3 (maghemite), TiO2 and MgAhO4 under high-stringency conditions (Fig. 5). The first two minerals are of particular interest in biotechnology due to their inherent magnetism (maghemite) or presence on the surface of titanium and its alloys (TiO2), which are widely used as implant materials in modern medicine. To a lesser extent, the optimized tag was also capable to specifically interact with the sulfide FeS and with the lithium-containing minerals LiMn2O4 and LiCoPO4. These conductive lithium compounds are well known for their use as electrode materials in lithium accumulators, paving the way for application of the tags in electrobiocatalysis (i.e. attachment of tagged redox enzymes to electrodes coated with such minerals in order to enable functional electron transfer) in the future. On the other hand, the two fluorinated apatites tested in the screen did not interact at high stringency (Fig. 5), which is in contrast to the homologous hydroxyapatites (Fig. 4) and may be explained by the different complexation propensities of fluoride and hydroxide ions.

[0207] In addition, three of the materials which are bound by the optimized tag (namely the magnetic materials maghemite and magnetite and the phosphate mineral strontium hydroxyapatite) were chosen as model minerals to study the association of the tag in more detail. The static binding capacity of these carriers (Fig. 6A) differed by one order of magnitude: While maghemite could bind up to 158 mg of the tagged protein per cm3of solid mineral the binding capacity of magnetite was lower (69 mg protein cm-3) but still exceeded the binding capability of resins commonly used in affinity purification of proteins (e.g. Ni2+-loaded IMAC materials from Cytiva (https: / / de.vwr.com / store / product / 10646437 / affinity-chromatography-media-imac- sepharosetm-6-fast-flow) or BioRad (https: / / www.bio- rad.com / webroot / web / pdf / lsr / literature / 10001677B.pdf). The lowest protein binding was observed for strontium hydroxyapatite (20 mg protein cm-3). For the two magnetic carriers, binding proceeded fast, i.e. the carriers were saturated with protein after an incubation time of five minutes (Fig. 6B). In contrast, binding to the hydroxyapatite material was slower and saturation was achieved after 30 minutes (Fig. 6B). Interestingly, the affinity of the tag was also dependent on the particle size of the mineral. For example, preparations of the magnetite with high (5 pM) and intermediate particle size (50 - 100 nm) were efficiently bound by the tag, whereas low nanoscale material (5 nm particles) were not accepted for binding. Presumably, the small size of the latter type of particles does not provide enough surface area to interact with a tagged protein which features a similar diameter.

[0208] Influence of the buffer composition on mineral binding of the tags

[0209] Another prerequisite for a general applicability of affinity tags in biotechnology is a stable binding to its interaction partner under a variety of reaction conditions, including different temperatures and pH values, and the tolerance to reagents which are commonly present in biochemical reactions. Thus, the temperature and pH dependence of the binding of four of the tags (the truncated PHG variant PHG(693-784) (SEQ ID NO. 30), PCT (SEQ ID NO. 13), bKRMP3 (SEQ ID NO. 6) and the tandem PHG(693-784)-PCT (SEQ ID NO. 63)) was determined for binding to a set of four minerals (strontium hydroxyapatite, KMnPO4 ■ H2O, magnetite and kaolinite) under standardized conditions (see sections above). Strikingly, the affinity to all minerals was only marginally influenced by a change in temperature in the physiological range (i.e. from 6 to 37 °C) - except for the interaction of the PHG(693-784) (SEQ ID NO. 30) with kaolinite which showed a slight decrease with increasing temperature.

[0210] The influence of different pH values was tested in three different buffers. The buffers covered the pH values typically used in biological systems (pH 5.0 to 9.0) and were adjusted to the same ionic strength. In general, the pH dependence of the binding was strongly dependent on both the tag and the mineral used as a carrier. For example, binding of the PCT (SEQ ID NO. 13) peptide to the phosphate minerals (strontium hydroxyapatite, KMnPO4 ■ H2O) and magnetite decreased at alkaline pH, but its interaction with kaolinite was not significantly influenced by a changed pH. The lower affinity of the tag at high pH, in particular evident in case of strontium hydroxyapatite, provides the opportunity to use this tag in specialized affinity purification procedures such as elution from the carrier by a pH shift. On the other hand, a stable binding across all tested pH values, which is desirable for many other applications, could be found for the interaction of the optimized tag PHG(693-784)-PCT (SEQ ID NO. 63) with KMnPO4 ■ H2O and magnetite. Noteworthy, such a pH-independent binding was not observed for both its parent peptides PHG(693-784) (SEQ ID NO. 30) and PCT (SEQ ID NO. 13).

[0211] In order to gain insight into the tolerance of the mineral binding towards different biochemical reagents, the binding of PHG(693-784)-PCT (SEQ ID NO. 63) to both magnetite and strontium hydroxyapatite was measured in the presence of the different compounds. The measurements were performed under standard conditions, applying commonly used concentrations of the additives. Strikingly, the association of the tag to both carriers was tolerant against reducing agents (such as thiols and ascorbate) and substances used in protein purification (Tris buffer, ethanol, polyethylene glycol, low amounts of Ni2+). However, high concentrations of salts (e.g. 1 M NaCI or KCI, 500 mM (NH4)2SO4) caused a strong reduction of the binding, probably due to a disturbance of ionic interactions between the protein and the mineral surface. Other types of reagents showed a different influence on the two tested minerals. In contrast to the reactions with magnetite, the binding to strontium hydroxyapatite was inhibited by a high concentration of the denaturing agent urea, sodium acetate, glycine and - to a certain extent - also by the complexing reagents EDTA and citrate. Unexpectedly, also sugars and sugar alcohols (such as glycerol) were effective inhibitors of the binding to strontium hydroxyapatite.

[0212] The strong eluting properties of the salt additives deemed us to study this effect in more detail, especially as it can be used for the recovery of mineral-bound protein in purification experiments. Therefore, a variety of salts including different alkali and ammonium chlorides, sodium halides or sodium salts of common oxyacids were tested as in the model system mCherry-PHG(693-784)-PCT (SEQ ID NO. 63) / magnetite in a concentration-dependent manner (Fig. 7). Interestingly, all of the alkali halides (Fig. 7A, C), NH4CI (Fig. 7B) and the tested oxyacid salts (Fig. 7D) suppressed binding of the tagged protein to magnetite at high concentrations. Also (NH4)2SO4 and sodium phosphate, which are frequently used additives in protein purification procedures, were potent suppressors. In contrast the two chaotropic ammonium derivatives tetramethylammonium and guanidinium chloride (Fig. 7B) were not similarly effective suppressing agents. The half-maximum of guanidinium chloride required to suppress binding of PHG(693-784)-PCT (SEQ ID NO. 63) to magnetite was 1.72 ± 0.06 M. The non-ionic denaturant urea was even less effective in inhibition of this interaction. This low sensitivity for chaotropic agents facilitates binding and purification of tagged proteins under mildly denaturing conditions, which is not possible in the purification of tagged proteins via most other affinity techniques (e.g. antibody-based strategies).

[0213] A comparison of the elutropic strength of the different compounds (Fig. 7E) led to the conclusion that a most efficient elution could be achieved by application of salts containing the alkali metal ions Rb+ / K+or the anions I7SO42'. Finally, the influence of sugar derivatives on the binding of PHG(693-784)-PCT (SEQ ID NO. 63)-coupled proteins was analyzed. This effect can also be applied for elution of proteins bound to hydroxyapatites, and thus for the development of a yet unprecedented purification strategy. Sugars or their derivatives are biocompatible and - in contrast to many common eluents such as salts or imidazole - do not interfere with the stability of the target protein during storage or its use in e. g. medicinal applications. Indeed, different mono- and disaccharides (D-glucose, sucrose) and sugar alcohols (D-sorbitol, glycerol) were active as suppressors of binding of the tagged protein to both strontium and calcium hydroxyapatite. Glycerol prevented the protein binding at concentrations exceeding 20 % (w / v) in both cases.

[0214] Application of the optimized tag in protein purification

[0215] The stable interaction of the optimized tandem tag with a number of minerals (Figs. 4, 5) tempted us to use this affinity for the purification of proteins in preparative scale. Therefore, strontium hydroxyapatite and maghemite were chosen as carriers. Both minerals showed a high stability under the buffer conditions used in the purification process (data not shown). Similar to magnetite (Fig. 7A), an elution of tagged proteins bound to the carriers was easily possible using increased concentrations of sodium or potassium chloride (Fig. 8A). The tagged model protein mCherry was then purified in analytical scale in a batch procedure (Fig. 8B). 1 ml of a bacterial lysate was incubated with either the hydroxyapatite or maghemite. After binding of the tagged protein, the carriers were either separated by centrifugation (strontium hydroxyapatite) or magnetic force (maghemite). After three steps of washing, the recombinant protein was eluted by administration of a high-salt buffer containing either 1 M NaCI (strontium hydroxyapatite) or 2 M KCI (maghemite). The efficiency of this batch purification procedure was followed by measurement of the mCherry concentration in the supernatants obtained in each individual purification step. The data (Fig. 8C) indicated that there was only minimal loss of protein during the washing steps, and that the recombinant protein could be recovered from the carriers in good yields (strontium hydroxyapatite: 64 %, maghemite: 74 %).

[0216] Subsequently, a preparative scale batch purification was carried out in larger scale. Therefore, a tagged version of the enzyme FDH was used. The enzyme was first purified from 15 ml of bacterial crude lysate using maghemite as a carrier. The success of purification was monitored by SDS-PAGE analysis, which showed that the enzyme was greatly enriched in the eluted protein fractions (Fig. 8D). The purity of the isolated protein was similar to this obtained by common IMAC protocols. Simultaneously, the enzyme activity was measured during the course of the purification. The activity data (Tab. 5) verified that the enzyme could be successful enriched by the one-step affinity chromatography on maghemite. In a similar purification experiment, in which strontium hydroxyapatite was used as an alternative carrier, an even higher purification factor and recovery yield was observed (Tab. 5).

[0217] Noteworthy, the specific activity of the enzyme tagged with the PHG(693-784)-PCT (SEQ ID NO. 63) peptide was in the range of FDH carrying a hexahistidine tag (Tab. 5), indicating that the novel tag has no significant influence on the functionality of the protein.

[0218] The purification of tagged mCherry by mineral affinity was also possible using column chromatography on strontium hydroxyapatite. The mineral was loaded with bacterial lysate until a saturation was observed, and then eluted with a one-step gradient of NaCI. The elution yielded electrophoretically homogeneous protein with a total protein recovery of 63 %. Advantageously, an efficient regeneration of the column was achieved by a pH shift to strongly alkaline conditions, i.e. by washing with 100 mM NaOH solution (data not shown).

[0219] Table 5: Purification of FDH via the MinTag system. Enzyme carrying the PHG(693-784)-PCT (SEQ ID NO. 63) tag was purified from bacterial cell lysates in a scale of 15 ml as specified in the Experimental Section. The specific activity / kcat of FDH having a C-terminal hexahistidine tag, which was purified by IMAC affinity chromatography, was 756 ± 8 nmol min-1mg-11 39.8 ± 0.4 min-1, n.a., not applicable.

[0220] Cleavage of the tag by fusion to auto-processing protease domains

[0221] A certain disadvantage of affinity tags is their possible influence on the biological activity of the fused target protein, which might arise from steric (e.g. blockage of the active site in enzymes or repulsion during protein-protein interactions) or charge-related reasons (e.g. change of isoelectric point, influence on ionic interactions). Therefore, methods to remove such tags from the target protein have been developed. The majority of these methods are based on proteolytic cleavage via specific cut motifs, which are introduced between the tag and its fusion partner and are not found in the target protein (Wood (2014) Curr. Opin. Struct. Biol. 26, 54). In order to enable such a removal in frame of the mineral affinity system, auto-processing protease domains were installed in in the tagged model protein mCherry-Hise and eGFP-Hise. Such protease domains have been described as part of different large bacterial proteins, which are processed into smaller subunits by protease action. Unlike for other proteases, the activity of these auto-processing domains is additionally regulated by the presence of inducer molecules, which have to be bound by the protease in order to induce cleavage. For the cleavage of the model protein, the two protease domains FrpC, which is activated by micromolar concentrations of Ca2+ions (Osicka et al. (2004) J. Biol. Chem. 279, 24944; Sadlikova et al. (2008) Protein Sci. 17, 1834), and CPD, which strictly requires inositol hexakisphosphate (I Pe) for activity (Sheahan et al. (2007) EMBO J. 26, 2552; Shen et al. (2009) EMBO J. 4, e8119) were chosen. The application of such auto-processing and-inducible proteases has two advantages to common tag removal protocols used so far: (I) the protease does not have to be produced in an additional recombinant expression, and (II) an internal protease, which can be activated during affinity purification to elute tag-free target protein exclusively, omits the necessity to remove any separately added proteolytic enzyme via subsequent purification steps.

[0222] The chosen protease domains, including their N-terminal cut motifs, were individually incorporated in between the mCherry or eGFP protein and the tag of the model proteins. Thus, activation of the fusion proteins, which were produced as purified recombinant proteins, should result in liberation of non-tagged mCherry or eGFP, respectively. Strikingly, the addition of Ca2+to the hexahistidine-tagged FrpC fusion resulted in the expected cleavage in an overnight reaction (Fig. 90). In contrast, the construct mCherry-CPD-Hise was proteolytically processed within a few minutes upon addition of its inducer IPe (with overall yield up to 65 %, Fig. 9D). As the fusion protein eGFP-CPD- PHG(693-784)-PCT (SEQ ID NO. 63) containing a mineralbinding instead of a hexahistidine tag was also cleaved with similar efficiency, an elegant strategy for a removal of the mineral-binding tag was provided.

[0223] Conclusions

[0224] In this application, peptides with high affinity for minerals were developed. The development process was based on two steps: First, a screening for mineral binding identified candidates which were capable of specific interaction with minerals under optimized stringency conditions. In subsequent optimization steps, the affinity of the best binders was increased up to a technical applicability by mutagenetic analysis of the binding motif and peptide tandemization. The obtained peptides strongly bound to a large set of natural and synthetic minerals, including insoluble phosphates, carbonates, clay minerals, magnetic iron oxides and conductive lithium compounds. Strikingly, the association of these binders to two selected carriers was tolerant against different temperatures I pH values and a number of biochemical reagents commonly applied in biochemistry (e.g. thiols, buffers). In contrast, the interaction proved to be reversible in the presence of high concentrations of salts and - as shown for the first time for mineral binders - of sugars and polyhydric alcohols such as glycerol. These effects were exploited in the purification of tagged model proteins and enzymes in batch and column format, which showed a selectivity and protein enrichment comparable to common IMAC methods. The removal of the tags proved to be possible by using self-processing, inducible protease domains. References

[0225] Antuch, M., Matos-Peralta, Y., Llanes, D., Echevarria, F., Rodriguez-Hernandez, J., Hernandez Marin, M., Diaz-Garcia, A., Reguera, L. (2019). Bimetallic Co2+and Mn2+hexacyanoferrate for hydrogen peroxide electrooxidation and its application in a highly sensitive cholesterol biosensor. ChemElectroChem 6, 1567-1573.

[0226] Deshpande, A. S., Fang, P.-A., Zhang, X., Jayaraman, T., Sfeir, C., Beniash, E. (2011). Primary structure and phosphorylation of dentin matrix protein 1 (DMP1) and dentin phosphophoryn (DPP) uniquely determine their role in biomineralization. Biomacromolecules 12, 2933-2945.

[0227] Dippe, M., Ulbrich-Hofmann, R. (2009). Spectrophotometric determination of phosphatidic acid via iron(lll) complexation for assaying phospholipase D activity. Anal. Biochem. 392, 169- 173.

[0228] Dippe, M., Bauer, A.-K., Porzel, A., Funke, E., Muller, A. O., Schmidt, J., Beier, M., Wessjohann, L. (2019). Coenzyme A-conjugated cinnamic acids - enzymatic synthesis of a CoA ester library and application in biocatalytic cascades to vanillin derivatives. Adv. Synth. Catal. 361 , 5363-5350.

[0229] Evans, J. S. (2008). ‘Tuning in’ to mollusk shell nacre- and prismatic-associated protein terminal sequences. Implications for biomineralization and the construction of high performance inorganic-organic composites. Chem. Rev. 108, 4455-4462.

[0230] Fukuta, M., Zheng, B., Uenurna, M., Okamoto, N., llraoka, Y., Yamashita, I., Watanabe, H. (2014). Controlled charged amino acids of Ti-binding peptide for surfactant-free selective adsorption. Colloids Surf. B Biointerfaces 118, 25-30.

[0231] George, G., Gupta, S. K., Rao, P. V. R., Narasaraju, T. S. B. (1987). Preparation and characterization of phosphate and arsenate apatites of strontium and their solid solutions. J. Mater. Sci. 22, 2274-2276.

[0232] He, G., Dahl, T., Veis, A., George, A. (2003). Nucleation of apatite crystals in vitro by selfassembled dentin matrix protein 1. Nat. Mater. 2, 552-558.

[0233] Hecker, A., Quennedey, B., Testeniere, O., Quennedey, A., Graf, F., Luquet, G. (2004). Orchestin, a calcium-binding phosphoprotein, is a matrix component of two successive transitory calcified biomineralizations cyclically elaborated by a terrestrial crustacean. J. Struct. Biol. 146, 310-324.

[0234] Hikichi, Y., Hukuo, K., Shiokawa, J. (1978). Syntheses of rare earth orthophosphates. Bull. Chem. Soc. Jpn. 51 , 3645-3646.

[0235] Huang, Z., Shi, X., Mao, J., Gong, S. (2016). Design of a hydroxyapatite-binding antimicrobial peptide with improved retention and antibacterial efficacy for oral pathogen control. Nat. Sci. Rep. 6, 38410.

[0236] Hunter, G. K., Hauschka, P. V, Poole, A. R., Rosenberg, L. C., Goldberg, H. A. (1996). Nucleation and inhibition of hydroxyapatite formation by mineralized tissue proteins. Biochem. J. 317, 59-64.

[0237] Hunter, G. K., O'Young, J., Grohe, B., Karttunen, M., Goldberg, H. A. (2010). The flexible polyelectrolyte hypothesis of protein-biomineral interaction. Langmuir 26, 18639-18646.

[0238] Jain, G., Pendola, M., Huang, Y.-C., Juan Colas, J., Gebauer, D., Johnson, S., Evans, J. S. (2017). Functional prioritization and hydrogel regulation phenomena created by a combinatorial pearl-associated two-protein biomineralization model system. Biochemistry 56, 3607-3618.

[0239] Koleva, V., Zhecheva, E., Stoyanova, R. (2011). Facile synthesis of LiMnPO4 olivines with a plate-like morphology from a dittmarite-type KMnPC>4 H2O precursor. Dalt. Trans. 40, 7385-7394.

[0240] Laemmli, II. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685.

[0241] Lee, J. S., Yu, X., Wagoner Johnson, A. J., Murphy, W. L. (2017). Mineral binding peptides with enhanced binding stability in serum. Biomater. Sci. 5, 663-668.

[0242] Lemloh, M.-L., Altintoprak, K., Wege, C., Weiss, M. I., Rothenstein, D. (2017). Biogenic and synthetic peptides with oppositely charged amino acids as binding sites for mineralization. Materials (Basel) 10, 119.

[0243] Lin, J.-Y., Ma, K.-Y., Bai, Z.-Y., Li, J.-L. (2013). Molecular cloning and characterization of perlucin from the freshwater pearl mussel, Hyriopsis cumingii. Gene 526, 210-216.

[0244] Mao, C. M., Sampath, J., Sprenger, K. G., Drobny, G., Pfaendtner, J. (2019). Molecular driving forces in peptide adsorption to metal oxide surfaces. Langmuir 35, 5911-5920.

[0245] Meyer, T. S., Lambert, B. L. (1965). Use of Coomassie brilliant blue R250 for the electrophoresis of microgram quantities of parotid saliva proteins on acrylamide-gel strips. Biochim. Biophys. Acta 107, 144-145.

[0246] Natalio, F., Link, T., Muller, W. E. G., Schroder, H. C., Cui, F.-Z., Wang, X., Wiens, M. (2010). Bioengineering of the silica-polymerizing enzyme silicatein-alpha for a targeted application to hydroxyapatite. Acta Biomater. 6, 3720-3728.

[0247] Osicka, R., Prochazkova, K., Sulc, M., Linhartova, I., Havlicek, V., Sebo, P. (2004). A novel "clip-and-link" activity of repeat in toxin (RTX) proteins from gram-negative pathogens. Covalent protein cross-linking by an Asp-Lys isopeptide bond upon calcium-dependent processing at an Asp-Pro bond. J. Biol. Chem. 279, 24944-24956.

[0248] Pagel, M., Hassert, R., John, T., Braun, K., WieBler, M., Abel, B., Beck-Sickinger, A. (2016). Multifunctional coating improves cell adhesion on titanium by using cooperatively acting peptides. Angew. Chem. 55, 4826-4830.

[0249] Riguero, V., Clifford, R., Dawley, M., Dickson, M., Gastfriend, B., Thompson, C., Wang, S.-C., O'Connor, E. (2020). Immobilized metal affinity chromatography optimization for polyhistidine tagged proteins. J. Chromatogr. A 1629, 461505.

[0250] Rodiguez, G. M., Bowen, J., Grossin, D., Ben-Nissan, B., Stamboulis, A. (2017). Functionalisation of Ti6AI4V and hydroxyapatite surfaces with combined peptides based on KKLPDA and EEEEEEEE peptides. Colloids Surf. B Biointerfaces 160, 154-160.

[0251] Sadlikova, L., Osicka, R., Sulc, M., Linhartova, I., Novak, P., Sebo, P. (2008). Single-step affinity purification of recombinant proteins using a self-excising module from Neisseria meningitidis FrpC. Protein Sci. 17, 1834-1843.

[0252] Sheahan, K.-L., Cordero, C. L., Fullner Satchell, K. J. (2007). Autoprocessing of the Vibrio cholerae RTX toxin by the cysteine protease domain. EMBO J. 26, 2552-2561.

[0253] Shen, A., Lupardus, P. J., Morell, M., Ponder, E. L., Masoud Sadaghiani, A., Garcia, K. C., Bogyo, M. (2009). Simplified, enhanced protein purification using an inducible, autoprocessing enzyme tag. EMBO J. 4, e8119. Soto-Rodriguez, J., Coyle, B. L., Samuelson, A., Aravagiri, K., Baneyx, F. (2017). Affinity purification of Car9-tagged proteins on silica matrices: Optimization of a rapid and inexpensive protein purification technology. Protein Expr. Purif. 135, 70-77.

[0254] Sprenger, K., Prakash, A., Drobny, G., Pfaendtner, J. (2018). Investigating the role of phosphorylation in the binding of silaffin peptide R5 to silica with molecular dynamics simulations. Langmuir 34, 1199-1207.

[0255] Su, J., Liang, X., Zhou, Q., Zhang, G., Wang, H., Xie, L., Zhang, R. (2013). Structural characterization of amorphous calcium carbonate-binding protein: An insight into the mechanism of amorphous calcium carbonate formation. Biochem. J. 453, 179-186.

[0256] Suzuki, M., Iwashima, A., Tsutsui, N., Ohira, T., Kogure, T., Nagasawa, H. (2011). Identification and characterisation of a calcium carbonate-binding protein, blue mussel shell protein (BMSP), from the nacreous layer. ChemBioChem 12, 2478-2487.

[0257] Tsukamoto, D., Sarashina, I., Endo, K. (2004). Structure and expression of an unusually acidic matrix protein of pearl oyster shells. Biochem. Biophys. Res. Commun. 320, 1175-1180.

[0258] Web protocol “Photometric detection of nicotinamide adenine dinucleotides”, https: / / www.aatbio.com / resources / application-notes / photometric-detection-of- nicotinamide-adenine-dinucleotides, accession date 1stof September 2024.

[0259] Web protocol “Site directed mutagenesis (QuickChange method)”, https: / / static.igem.Org / mediawiki / 2012 / a / a5 / Site_Directed_Mutagenesis.pdf, accession date 1stof September 2024.

[0260] Website https: / / www.bio-rad.com / webroot / web / pdf / lsr / literature / 10001677B.pdf, accession date 1stof August 2022.

[0261] Website https: / / de.vwr.com / store / product / 10646437 / affinity-chromatography-media-imac- sepharosetm-6-fast-flow, accession date 1stof August 2022.

[0262] Wood, D. W. (2014). New trends and affinity tag designs for recombinant protein purification. Curr. Opin. Struct. Biol. 26, 54-61.

[0263] Yamagishi, A., Narumiya, K., Tanaka, M., Matsunaga, T., Arakaki, A. (2016). Core amino acid residues in the morphology-regulating protein, Mms6, for intracellular magnetite biomineralization Sci. Rep. 6, 35670.

[0264] Yarbrough, D. K., Hagerman, E., Eckert, R., He, J., Choi, H., Cao, N., Le, K., Hedger, J., Qi, F., Anderson, M., Rutherford, B., Wu, B., Tetradis, S., Shi, W. (2010). Specific binding and mineralization of calcified surfaces by small peptides. Calcif. Tissue I nt. 86, 58-66.

[0265] Zerfass, C., Braukmann, S., Nietzsche, S., Hobe, S., Paulsen, H. (2015). High yield recombinant production of a self-assembling polycationic peptide for silica biomineralization. Protein Expr. Purif. 108, 1-8.

Claims

Claims1 . A mineral binding peptide, wherein said mineral binding peptide comprises the consensus sequenceTGSGQTFTYINWGPGNPDNSG (SEQ ID NO. 200) and is selected from PHG(653-784) (SEQ ID NO. 7) or a truncated peptide thereof or a tandem ized peptide thereof.

2. The mineral binding peptide of claim 1 , wherein said mineral binding peptide is a truncated peptide selected from the group consisting of PHG(677-784) (SEQ ID NO. 22), PHG(693-784) (SEQ ID NO. 30), PHG(708-784) (SEQ ID NO. 24), PHG(722-784) (SEQ ID NO. 25), PHG(733-784) (SEQ ID NO. 26), PHG(693-774) (SEQ ID NO. 23), PHG(693-765) (SEQ ID NO. 28), PHG(693- 754) (SEQ ID NO. 29), and a peptide that has a sequence identity of at least 75 % to at least one peptide thereof.

3. The mineral binding peptide of claim 1 or 2, wherein said mineral binding peptide is a truncated peptide selected from the group consisting of PHG(677- 784) (SEQ ID NO. 22), PHG(693-784) (SEQ ID NO. 30), PHG(708-784) (SEQ ID NO. 24), PHG(722-784) (SEQ ID NO. 25), PHG(733-784) (SEQ ID NO. 26), PHG(693-774) (SEQ ID NO. 23), and PHG(693-765) (SEQ ID NO. 28).

4. The mineral binding peptide of any of claims 1 to 3, wherein said mineral binding peptide is a truncated peptide selected from the group consisting of PHG(677-784) (SEQ ID NO. 22), PHG(693-784) (SEQ ID NO. 30), PHG(693- 774) (SEQ ID NO. 23), and PHG(693-765) (SEQ ID NO. 28).

5. The mineral binding peptide of claim 1 , wherein said mineral binding peptide is PHG(653-784) (SEQ ID NO. 7).

6. The mineral binding peptide of claim 1 , wherein said mineral binding peptide is a tandem ized peptide selected from the group consisting of PHG(653-784)-PCT (SEQ ID NO. 62), PHG(693-784)-PCT (SEQ ID NO. 63), PHG(693-765)- PCT (SEQ ID NO. 64), PHG(693-784)-bKRMP3 (SEQ ID NO. 66), PHG(653- 784)-His8, PHG(653-784)-Pif80(912-923) (SEQ ID NO. 74), PHG(653-784)- PAT (SEQ ID NO. 68), PHG(693-784)-PAT (SEQ ID NO. 69), PHG(693-765)- PAT (SEQ ID NO. 70), PHG(653-784)-Aspein_2-30 (SEQ ID NO. 75), PHG(653-784)-Aspein_2-41 (SEQ ID NO. 76), PHG(653-784)-E(QD)5 (SEQ ID NO. 71 ) PHG(653-784)-E(EN)5 (SEQ ID NO. 72), and a peptide that has a sequence identity of at least 75 % to at least one peptide thereof.

7. The mineral binding peptide of claim 1 or 6, wherein said mineral binding peptide is a tandemized peptide selected from the group consisting of PHG(653-784)-PCT (SEQ ID NO. 62), PHG(693-784)-PCT (SEQ ID NO. 63), PHG(693-765)-PCT (SEQ ID NO. 64), PHG(693-784)-bKRMP3 (SEQ ID NO. 66), PHG(653-784)-PAT (SEQ ID NO. 68), PHG(693-784)-PAT (SEQ ID NO. 69), PHG(653-784)-Aspein_2-30 (SEQ ID NO. 75), PHG(653-784)-E(EN)5 (SEQ ID NO. 72).

8. The mineral binding peptide of claim 1 , 6 or 7, wherein said mineral binding peptide is a tandemized peptide selected from the group consisting of PHG(653-784)-PCT (SEQ ID NO. 62), PHG(693-784)-PCT (SEQ ID NO. 63), PHG(693-765)-PCT (SEQ ID NO. 64), and PHG(653-784)-PAT (SEQ ID NO. 68).

9. The mineral binding peptide of claim 1 , 6, 7 or 8, wherein said mineral binding peptide is the tandemized peptide PHG(693-784)-PCT (SEQ ID NO. 63).

10. The mineral binding peptide according to any one of claims 1 to 9, wherein said mineral binding peptide binds to at least mineral selected from the group consisting of CaCOs, SrCOs, Cas(PO4)3OH, S rs(P 04)30 H, KMnPO4 x H2O, Fe3O4, and kaolinite.

11. A fusion protein or conjugate comprising a mineral binding peptide of any one of claims 1 to 10 and a further agent selected from the group consisting of(I) proteins with enzymatic activity,(II) proteins with binding affinities for other biomolecules (e.g. antibodies), small molecules or inorganic ions (e.g. heavy metal binding peptides or protein domains);(III) therapeutic proteins, including growth factors, peptide hormones and antimicrobial peptides;(IV) structural proteins to cover, shield or protect the abovementioned types of mineral surfaces (e.g. hydrophobins) or to be used as adapters for components for synthetic biology applications; and(V) peptides, proteins, nucleic acids, small molecules (natural products, drugs and synthetic bioactives) and natural and synthetic polymers, wherein said fusion protein or conjugate optionally comprise a cut site accessible for external proteases, or a self-processing domain.

12. Use of a mineral binding peptide as claimed in any one of claims 1 to 10 in(I) affinity purification of target proteins on mineral surfaces,(II) immobilization of proteins fused or conjugated to said mineral binding peptide to a mineral surface, including immobilized enzyme cascades and electrobiocatalysis by proteins I transition metal complexes immobilized to conductive minerals,(III) application in carrier-based analytical techniques requiring surface modification, such as ELISA, biosensors, lateral flow assays, surface plasmon resonance, (IV) targeting in biological systems,(V) application as additives in the food, pharmaceutical and feed sector such as surface-modified minerals as ingredients of nutraceuticals, animal feed products or toothpaste, and(VI) adhesives in biotechnological applications and biological systems.

13. A mineral binding peptide as claimed in any one of claims 1 to 10 for use in methods of targeting in biological systems, i.e. delivery of drugs and therapeutic proteins in the body such as tooth fillings supporting healing,therapy of caries and bone fractures, magnetic delivery of drugs to specific tissues and organs and titanium implants.

14. A method for affinity purification of proteins comprises the following steps: a) Providing a fusion protein or a conjugate of claim 11 , and providing a mineral or mineral carrier selected from the group consisting of solid carbonates such as MgCOs x H2O, CaCOs, SrCOs, MnCOs x H2O, Ce2CO3 X H2O; insoluble phosphates and insoluble sulfates such as (NH4)MgPO4x 6 H2O, Mgs(PO4)2 x H2O, Ca5(PO4)3OH, Sr5(PO4)3OH, KMnPO4x H2O, Zn3(PO4)2, CePO4x 0.5 H2O, CaSO4x 2 H2O, SrSO4; oxides, sulfides, selenides and tellurides of Li, Be, Mg, Ca, Sr, Ba, Ra, Al, Ga, Ge, In, Sn, Pb, As, Sb and of all transition metals and lanthanides, such as AI2O3, Fe3O4, CaC2O4x H2O; (Alumo)silicates, y-Fe2O3, FeS, TiO2, MgAl2O4, LiMn2O4, LiCoPO4, LiFePO4, calcium fluorapatite, and strontium fluorapatite.; b) Incubating the fusion protein or the conjugate and the mineral or mineral carrier of step a) under suitable conditions and for a time sufficient to enable binding the fusion protein or the conjugate to the mineral or mineral carrier; c) Washing the resulting fusion protein - mineral complex or conjugate - mineral complex of step b), d) Eluting the fusion protein or conjugate from the mineral or mineral carrier; and, e) Optionally, determining the yield of said fusion protein or conjugate eluted in step d), wherein said method optionally comprises the step of cleaving the mineral binding peptide from the fusion protein or conjugate at a cut site accessible for external proteases, or by a self-processing domain, which are optionally comprised in the fusion protein or conjugate.

Images