Matter‐tag: A universal immobilization platform for enzymes on polymers, metals, and silicon‐based materials

Enzyme immobilization is extensively studied to improve enzyme properties in catalysis and analytical applications. Here, we introduce a simple and versatile enzyme immobilization platform based on adhesion‐promoting peptides, namely Matter‐tags. Matter‐tags immobilize enzymes in an oriented way as a dense monolayer. The immobilization platform was established with three adhesion‐promoting peptides; Cecropin A (CecA), liquid chromatography peak I (LCI), and Tachystatin A2 (TA2), that were genetically fused to enhanced green fluorescent protein and to two industrially important enzymes: a phytase (from Yersinia mollaretii) and a cellulase (CelA2 from a metagenomic library). Here, we report a universal and simple Matter‐tag–based immobilization platform for enzymes on various materials including polymers (polystyrene, polypropylene, and polyethylene terephthalate), metals (stainless steel and gold), and silicon‐based materials (silicon wafer). The Matter‐tag–based enzyme immobilization is performed at ambient temperature within minutes (<10 min) in an aqueous solution harboring the phytase or cellulase by immersing the targeted material. The peptide LCI was identified as universal adhesion promoter; LCI immobilized both enzymes on all investigated materials. The attachment of phytase‐LCI onto gold was characterized with surface plasmon resonance spectroscopy obtaining a dissociation constant value (KD) of 2.9·10−8 M and a maximal surface coverage of 504 ng/cm².

To show the general applicability of the developed platform, Matter-tags were fused to two industrially relevant enzymes, phytase (YmPh, Yersinia mollaretii,~47 kDa; Shivange & Schwaneberg, 2017;Shivange et al., 2012) and cellulase (CelA2, metagenome library, 69 kDa; Ilmberger et al., 2012). Phytases (myo-inositol hexakisphosphate phosphohydrolases) catalyze the sequential removal of inorganic phosphate from phytic acid by hydrolysis. Phytases have a market value of approximately US $350 million per year representing more than 60% of the total feed enzyme market (Shivange & Schwaneberg, 2017). Cellulases catalyze the depolymerization of cellulose and find industrial applications in, for example, bioethanol production, textile, pulp and paper, food, and laundry (Kuhad, Gupta, & Singh, 2011). In addition to the two selected enzymes, the enhanced green fluorescent protein (eGFP) was fused to the Matter-tags. Intramolecular interactions between enzyme and Matter-tag are avoided by a stiff helix (17 amino acids, AEAAA-KEAAAKEAAAKA; Arai, Ueda, Kitayama, Kamiya, & Nagamune, 2001) that also assists to minimize or avoid interactions of immobilized enzymes with the material surface.
We studied the binding of three Matter-tags to six selected materials by direct visualization of attached eGFP-Matter-tag-fusion proteins with confocal fluorescence microscopy. The general applicability of the Matter-tag-platform was subsequently shown with two selected enzymes based on standard fluorogenic activity determination systems on six selected materials. The binding of YmPh-LCI was further characterized with surface plasmon resonance (SPR) spectroscopy calculating the dissociation constant and maximal surface coverage.

| MATERIALS AND METHODS
All used chemicals were purchased from AppliChem GmbH (Darmstadt, Germany), Carl Roth GmbH (Karlsruhe, Germany) and Sigma-Aldrich Chemie GmbH (Steinheim, Germany) with a purity of analytical-reagent grade or higher. Synthetic genes were acquired from GeneArt AG (Regensburg, Germany), oligonucleotides were obtained from Eurofins Scientific SE (Ebersberg, Germany) in salt-free form and enzymes were purchased from New England Biolabs (Frankfurt am Main, Germany).
Polymerase chain reaction purification and plasmid extraction kits were purchased from Qiagen GmbH (Hilden, Germany) and Macherey-Nagel

| Plasmids and strains
The plasmid pET28a (+) (Novagen, Darmstadt, Germany) was used as expression vector for the eGFP-and cellulase-fusion constructs. The plasmid pALXtreme-5b (Shivange et al., 2012) was used for the expression of phytase fusion constructs. The Escherichia coli strain DH5α was used as cloning strain. For expression of eGFP-and cellulase-fusion constructs, E. coli BL21-Gold (DE3) was used.

| Generation of eGFP-Matter-tag-fusion constructs
The eGFP was fused as reporter protein to Matter-tags to have a direct visual proof of binding to the six selected materials. N-Matter-tag-fusion constructs pET28a::CecA/LCI/TA2-17H-eGFP and C-Matter-tag-fusion constructs pET28a::eGFP-17H-TEV-CecA/LCI/TA2 were generated according to previous reports (Rübsam et al., 2017;Rübsam, Weber et al., 2018). The detailed cloning strategy is described in Supporting Information Material; primers are shown in Table S1.

| Expression of eGFP-Matter-tag-fusion constructs
Protein production in shaking flasks of eGFP-Matter-tag-fusion constructs was performed as previously reported (Rübsam et al., 2017). The cell pellets were stored at −20°C.

| Expression of phytase-Matter-tag-fusion constructs
Precultures were inoculated with the corresponding cryoculture (10 ml LB-media in glass tube; 10 g/L tryptone, 10 g/L NaCl, 5 g/L yeast extract, and 100 μg/ml ampicillin) and incubated ( Protein purity was evaluated with ImageJ 1.48v software, analyzing the intensity of bands of sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) images (Schneider, Rasband, & Eliceiri, 2012; Figure S4). The activities of purified enzymes were determined as described in Supporting Information Material. The purified enzymes were used for SPR spectroscopy.
2.9 | Immobilization analysis of eGFP-Matter-tag-fusion proteins by confocal fluorescence microscopy RT). As control, supernatant containing eGFP-17H-TEV without Mattertag was investigated. According to previous work, target enzymes and proteins were used in excess (Rübsam et al., 2017). Supernatant was removed, materials were rinsed with 10 ml of Tris/HCl buffer (50 mM, pH 8.0), and transferred into 25-well squared petri dishes filled with 2 ml of Tris/HCl buffer (50 mM, pH 8.0). Samples were incubated (2 min), buffer was removed, and new buffer was added (three times). The samples were removed, rinsed with 10 ml of ddH 2 0, and dried with nitrogen flow.
Binding of eGFP-Matter-tag-fusion proteins was determined by confocal fluorescence microscopy (TCS SP8; Leica Microsystems CMS GmbH, Mannheim, Germany). Samples were excited with 488 nm, 10% laser intensity. Detection was performed with a PMT2 detector (emission 500-565 nm, gain was varied for each material; see Table S5).

| Preparation and binding of enzyme-Matter-tag-fusion proteins to selected materials
Frozen cell pellets were suspended in the corresponding buffer (YmPh-fusions: 50 mM Tris/HCl buffer, pH 7.4; CelA2M2-fusions: T A B L E 2 Matter-tag-fusion proteins generated in this study with abbreviations Note: The three Matter-tags, CecA, LCI, and TA2 were genetically fused to the eGFP, phytase (YmPh), cellulase (CelA2M2) and expressed Abbreviations: CecA, Cecropin A; eGFP, enhanced green fluorescent protein; LCI, liquid chromatography peak I; TA2, Tachystatin A2 0.2 M KPi buffer, pH 7.2; 6 ml buffer on 1 g cell pellet 2.11 | Activity determination of cellulase and cellulase-Matter-tag-fusion proteins in solution and after material binding (4-MUC assay) The activity of CelA2M2 and CelA2M2-Matter-tag-fusion proteins was determined with the 4-methylumbelliferyl-β-D-cellobioside (4-MUC) assay (Lehmann et al., 2012 2.12 | Activity determination of phytase and phytase-Matter-tag-fusion proteins in solution and after material binding (4-MUP assay) The activity of YmPh and YmPh-Matter-tag-fusion proteins was determined with the 4-methylumbelliferyl-β-D-phosphate (4-MUP) assay (Shivange et al., 2012   The adsorption process of YmPhy-LCI was quantified by applying the Langmuir isotherm model (Equation (1)) based on the assumption of a dynamically reversible adsorption process of noninteracting solutes with all adsorption sites being of equal adsorption energy leading to a monomolecular layer formation (Latour, 2015).

| Reusability of phytase-LCI on PS and PP
The amount of solute at the surface (Q; mole/m 2 ) is given by the amount of solute attached at surface saturation (Q max , mole/m 2 ) multiplied with the solution concentration of the solute (C, mole/L) divided by the sum of solution concentration of the solute (C, mole/ L) and inverse adsorption equilibration constant, (K eq ; L/mole). The dissociation constant K D is determined to be the solution concentration corresponding to 50% of the maximal amount of solute attached at surface saturation (Q max ). K D is calculated to be K eq −1 (Latour, 2015).

| RESULTS
The

| Enzyme immobilization utilizing Matter-tags
The activity of the two industrially relevant enzymes, YmPh and  Matter-tags to CelA2M2 resulted in improved activities up to 2.4 fold compared to WT CelA2M2 on stainless steel. On gold, all CelA2M2-Matter-tag-fusion proteins led to increased CelA2M2 activity (between 1.6 and 3.3 fold compared to WT CelA2M2). On silicon wafer, significantly increased activities (up to 5.3 fold) of immobilized Mattertag-fusion proteins compared to WT CelA2M2 were observed.

| Reusability of phytase-LCI on selected materials PS and PP
The performance of YmPh-LCI after immobilization and repetitive utilization was investigated on PS ( Figure 2a) and PP (Figure 2b). After one reaction cycle, the residual activity decreased to 81% and 80% on PS and PP, respectively (Figure 2). After eight subsequently performed cycles, phytase activity was >50% compared to its initial activity.

| Adsorption isotherm and kinetics of YmPh-LCI on gold
The adsorption of YmPh-LCI on gold was monitored by SPR in real time.
Solutions of YmPh-LCI in Tris/HCl buffer (50 mM, pH 7.4) were injected over gold at a concentration range of 2.5-3,000 nM, while the surface plasmons were monitored. After about 20 min of contact, no further increase of adsorbed proteins was observed and this value was taken as the equilibrium adsorption for every studied concentration (Table S6).
The amount of adsorbed protein at equilibrium increased monotonically with the concentration of peptides (Figure 3a). The equilibrium adsorption of proteins was utilized to build adsorption isotherms ( Figure   3b). The isotherm could be fitted to a Langmuir model. This means that the mechanism of binding depends on a single type of process, the binding of the peptide and that all the binding sites at the surface have the same probability of binding regardless of the degree of saturation.
This further supports that the LCI is the sole binding element and not the enzyme fragment of the fusion protein. From the Langmuir fit, we derived an equilibrium constant (K eq ) and the dissociation constant (K D ) amounting for 34 ± 7·10 −3 nM −1 (3.4 ± 0.7·10 7 M −1 ) and 29 ± 6 nM F I G U R E 2 Reusability of YmPh-LCI after immobilization on (a) PS and (b) PP. Enzyme activity was determined after each washing cycle with the 4-MUP assay. 4-MUP, 4-methylumbelliferyl-β-D-phosphate; PP, polypropylene; PS, polystyrene (2.9 ± 0.6·10 −8 M), respectively. The maximal surface coverage (Q max ) was determined to be 504 ± 21 ng/cm 2 (Figure 3b). Such value could be obtained for any concentration higher than 200 nM. Remarkably, such amounts of adsorbed protein is close to a full monolayer of protein and represents a very high degree of surface functionalization achievable with very low concentrations. As a control, we measured the adsorption of WT YmPh lacking LCI. We performed the adsorption kinetics at a concentration of 200 nM, a value over which the YmPh-LCI already reached to maximum. The total adsorption of YmPh-LCI amounted for 457 ng/cm 2 , about 2.6-folds more than for the counterpart without Matter-tag (WT YmPh: 176 ng/cm 2 ) for the same concentration ( Figure 3c). This suggests that the main driving force for the surface binding is the LCI peptide. Interestingly, the activity of YmPh-LCI on gold is 2.8 fold increased compared to WT YmPh (Figure 1, immobilization of YmPh-LCI on gold), which indicates that LCI promotes an oriented immobilization.

| DISCUSSION
Being able to universally immobilize enzymes as a dense monolayer in aqueous solutions at ambient temperature on various materials is an important platform for various biotechnological applications (e.g., biosensors, biocatalytic processes, or surface coatings of polymers or medical products). Adhesion-promoting peptides can deliver such immobilization characteristics (Figure 4). Oriented immobilization as a defined monolayer on PP was previously reported by our group (Rübsam et al., 2017) and confirmed for YmPh-LCI on gold by the plotted Langmuir isotherm, which fits well to the experimental data. Figure 1 shows that the three Matter-tags are sufficient to enable efficient immobilization on a broad range of materials comprising synthetic polymers (PP, PS, PET), metals (stainless steel, gold), and silicon-based materials (silicon wafer). As a general trend, one could observe that C-terminal fusion of Matter-tags to eGFP are better binders than N-terminal fusions. The latter was confirmed by C-terminal fusions of all Matter-tags with the YmPh (see Figure 1). Fusion of Matter-tags to phytase and cellulase had no significant influence on protein production (see Figures S2 and S4). Volumetric activities of phytase/cellulase containing cell-free lysates were increased/decreased depending on the fused Matter-tag (see Figure S3). However, the specific activity of purified  (Rübsam, Weber et al., 2018) within a KnowVolution campaign (Rübsam, Davari et al., 2018).
F I G U R E 3 (a) Adsorption kinetics of YmPh-LCI on gold at a range of concentrations between 2.5 and 3,000 nM. (b) Langmuir isotherm of YmPh-LCI on gold at a range of concentrations between 2.5 and 3,000 nM (dotted line). The value Q max represents the maximum surface coverage, K eq is the equilibrium constant, K D denotes the dissociation constant, and R 2 represents the correlation coefficient. material binding spectrum compared to CecA and TA2. The LCI structure is nicely composed of a hydrophobic and polar surface, which enables an excellent balance between hydrophobic and polar interactions ( Figure 4a). LCI therefore represents an ideal adhesion promoter for applications in biocatalysis (e.g., silica), within chromatography/ purification systems (PS, PP), biosensors (gold), and biomaterials used in implants, textiles, or as packaging materials (PET, stainless steel).
How could LCI interact with the polymers PS, PP, and PET? LCI contains four Tyr and is rich in hydrophobic amino acids such as Ile, Val, and Phe. The side chains of these amino acids form many hydrophobic interactions with PS, PP, and PET (e.g., through the benzene ring structure of PS, hydrophobic interactions with CH 2 group in PP, and both π-stacking interactions and H-bonds with PET).
LCI has a stretched structure with aromatic rings of Tyr and Phe as side chains, suggesting that π-π stacking through the phenyl groups may play a major role in the binding to PS and PET. Pyrrole ring in Pro and isopropyl group of Val may also contribute to this interaction. Trp, Phe, Pro, and Val, which exist in LCI sequence, are the most frequent amino acids in PS binding sequences (Adey, Mataragnon, Rider, Carter, & Kay, 1995;Hunter, 1994;Hunter & Sanders, 1990;Nakanishi, Sakiyama, & Imamura, 2001;Yamaguchi, Isozaki, Nakamura, Takaya, & Watanabe, 2016).
How could LCI bind to silicon wafers? Silicon wafers always have a layer of 3-5 nm of native silicon oxide. Recent studies of peptide adsorption onto silica surfaces indicated the importance of polar sidechains for adsorption (Seker & Demir, 2011). LCI contains several basic and acidic amino acids such as Asp and Arg (polar surface, Figure 4a).
These amino acids together with the other glutamines may be ionized and function in the binding by electrostatic adsorption. Apart from electrostatic interaction, the importance of different effects such as π-π interactions and hydrophilic interactions were also reported in silica binding peptides (Notman et al., 2010;Oren et al., 2007;Puddu & Perry, 2012). Upon binding to silica surfaces, several electrostatic interactions between positively charged residues (Lys and Arg) and the negatively charged silica are expected.
How could LCI bind to stainless steel and gold? Studies on the adsorption of peptides on stainless steel surfaces suggest that the adsorption behavior of a peptide can be considered to be largely electrostatic (Imamura, Kawasaki, Awadzu, Sakiyama, & Nakanishi, 2003). Sarikaya, Tamerler, Jen, Schulten, and Baneyx (2003) have also reported that for gold-binding peptide sequences, the main contribution to the adsorption energy comes from the polar residues.
Therefore, the polar surface of LCI (Figure 4a) may be involved in interaction with stainless steel and gold.

| CONCLUSION
In conclusion, the three selected Matter-tags enable the immobilization of proteins and enzymes on various surfaces (synthetic polymers: PS, PP, PET; metals: stainless steel, gold, and silicon wafer). The advantages of the Matter-tag platform lie in its simple single step application at ambient temperature in an aqueous solution containing the Matter-tag-fusion protein. Interestingly, dense monolayers are obtained for investigated materials and LCI proved to be a universal binder due to an excellent distribution of hydrophobic and hydrophilic interactions. The latter is in contrast to small peptides (<8 amino acids) achieved through a defined 3D-structure of β-strands, which enable to "freely" tune the hydrophobicity and hydrophilicity above as well as below the β-strands. Furthermore, properties of adhesion-promoting peptides can be tailored to improve, for instance, its binding strength or proteolytic resistance by directed evolution/rational design. In addition, the linker between Matter-tags and fused enzymes can be varied in terms of its length and flexibility to further improve enzyme performance.