S-layer protein covered surfaces

The present invention relates to S-layer protein covered surfaces .

Non-specific adsorption of proteins at surfaces leads to fouling of biosensors, decreased performance and failure of indwelling devices such as implants, stents and electrodes, and decreased sensitivity of medical tests that detect binding of specific proteins. Thus, the ability to resist biofouling is important for the design of biocompatible coatings for implants and for biosensors capable of detecting analytes. Furthermore said non-specific adsorption may also lead to a reduced efficiency in the course of filtrations of proteinaceous solutions and to a blockage of the pores of a filter. Surface characteristics can also play a pivotal role in modulating the behaviour of cellular systems. Chemical modification of surfaces may, for instance, prevent cell adhesion to substrates.

It is well known that reducing the hydrophobicity of a membrane is advantageous, as it reduces the fouling tendencies of a surface. To decrease the hydrophobicity inherent to most polymeric membrane materials, it is known either to chemically modify the surface with hydrophilic groups (e.g. hydroxyl groups) or, alternatively, to coat the surface with a hydrophilic layer, the layer usually being polymeric in nature. The coated hydrophilic layer improves the affinity of the composite material towards water, increasing its wettability and, in some cases, making the surface completely wettable by water .

WO 2003/055906 A relates to modified S-layer proteins, which comprise inserted heterologous peptides. These modified S-layer proteins are able to bind to surfaces of, e.g., silicon wafers.

In Fahmy K. et al. (Biophysical J 91 (2006) : 996-1007 ) S- layer proteins are described which are able to bind heavy metal ions like palladium.

In Vadillo-Rodriguez V. et al. (Colloids and Surfaces, B: Biointerfaces 41 (2005) : 33-41) binding properties of two lactobacillus strains (one coated with S-layer proteins, one not coated with S-layer proteins) to hydrophilic and hydrophobic surfaces .

It is an object of the present invention to provide a surface with anti-fouling properties which inhibits substantially the binding of biomolecules to said surface. Another object relates to a modified surface on the one hand binding specifically molecules and on the other hand exhibiting anti-fouling properties.

Therefore, the present invention relates to an S-layer protein covered surface, wherein the uncovered surface has a biomolecule binding capacity of less than 150 μg/cm ² as measured by the BSA binding test and in that the surface is covered at least partially with an S-layer protein lattice.

It surprisingly turned out that S-layer proteins can bind to surfaces which are regularly considered as exhibiting low protein binding characteristics. Furthermore the S-layer proteins are also able to form the typical two-dimensional crystalline lattice. Since S-layer proteins themselves exhibit anti-fouling properties the substrate or the parts covered with the S-layer lattice show also anti-fouling properties.

S-layer lattices of the present invention are composed of identical protein or glycoprotein subunits aligned in the most accurate orientation.

Pores in the lattice ensure no significant pressure drop in the course of ultrafiltration or reverse osmosis filtrations. This may be associated with the nanopatterned antifouling propertiesof S-layers. The surface of a solid support, for instance, exhibits random arrangement of charged groups or hydrophilized and /or hydrophobic domains.

With S-layers the surface (e.g. antifouling surface) properties are much more defined due to the regular arrangement of the (glyco) protein subunits.

Moreover, the topography of the S-layer lattice provides a more defined "nanocontrolled" surface. On other antifouling supports such accurate conditions and surface properties are not given.

^M S-layer" proteins are the outermost cell envelope component of a broad spectrum of bacteria and archaea. S-layers are composed of a single protein or glycoprotein species (Mw 40-20OkDa) and exhibit either oblique (pi, p2), square (p4) or hexagonal (p3, pβ) lattice symmetry with unit cell dimensions in

the range of 3 to 30 nm. S-layers are generally 5 to 10 nm thick and show pores of identical size (diameter, 2 - 8 nm) and morphology. S-layer proteins are routinely utilised in several technical applications. For instance the EP 0 306 473 Bl discloses the use of S-layer proteins as a carrier for haptens, immunogenic or immunostimulant substances. The immobilisation of molecules, particularly of proteins, on S-layer proteins is described in the EP 0 362 339 Bl. In the WO 02/097118 Al the production of a layer of functional molecules on a substrate using S-layer proteins is disclosed. Therein the protein layer on the surface of a carrier is formed by creating an electrochemical potential difference between the solution and the surface. The EP 0 189 019 Bl discloses the making and the use of an ultrafiltration membrane employing S-layer proteins. The ultrafiltration membranes used therein do not show low binding properties according to the present invention. It turned out that the overall amount of S-layer protein/S-layer fusion protein on the hydrophilic surface is at least 6 times higher than the amount of BSA on the same surface. The functionality of the proteins is fully retained. For instance, the S-layer fusion protein rSbpA-ZZ is still able to bind human IgG after the immobilization on the low protein binding surface. Furthermore, this further shows that the bio-functionality is maintained and the functional group is properly exposed away from to the surface. S-layer proteins as used herein includes also S-layer glycoproteins .

S-layer proteins advantageously show also intermolecular binding (i.e. single S-layer protein molecules bind to other S- layer protein molecules adjacent in an S-layer lattice) . Therefore, also regions of a surface which normally would not be covered by any protein molecule will be covered by the S-layer lattice. This is regularly observed when a protein (e.g. BSA) is used to cover surfaces and some regions of said surface are not covered by any protein molecule. Such behaviour leads to a surface irregularly covered with a protein.

As used herein, "BSA binding test" and "IgG binding test" refers to a method for determining the ability and the extent (quantitatively) of a surface to bind a protein, in particular BSA (bovine serum albumin) or IgG. Both of these proteins are

regularly used model proteins to show to which extent a surface is able to bind proteins (see e.g.: Weigert et al . (1995) Journal of Membrane Science. 106: 147-159).

"Binding capacity of less than 150 μg/cm ² ", as used herein, means that a protein, in particular the model protein BSA, binds to the substrate in an amount of less than 150 μ ^' g per cm ² of the surface area. In particular preferred embodiments of the present invention BSA binds to the surface in an amount of less than 100 μg/cm ² , less than 75 μg/cm ² , less than 50μg/cm ² , less than 30 μg/cm ² , less than 10 μg/cm ² .

The term "surface", as used herein, refers to a surface which is part of a three-dimensional object or body, like a membrane, sphere (i.e. bead), plate (e.g. made of glass, silicone) etc..

"The surface is covered at least partially with an S-layer protein lattice" means that the surface may be covered completely or partially with crystallised S-layer proteins. The partial coverage of a surface (at least 10%, preferably at least 25%, more preferably at least 50%, even more preferably at least 75%) allows giving the surface a shape.

According to a preferred embodiment of the present invention the biomolecule' s unspecific binding capacity is less than 100 μg/cm ² , preferably less than 75 μg/cm ² , preferably less than 50 μg/cm ² , more preferably less than 30 μg/cm ² , as measured by the BSA binding test. S-layer proteins can be bound even on hydrophilic surfaces which exhibit a binding capacity for biomolecules of less than 75 μg/cm ² .

According to another preferred embodiment of the present invention the surface is hydrophilic.

In particular hydrophilic surfaces exhibit low biomolecule (e.g. protein) and organism (e.g. prokaryotic, eukaryotic) binding properties. The surface which according to the present invention is at least partially covered by an S-layer protein lattice comprises hydrophilic functional groups, preferably hydroxyl groups .

Techniques used to study the hydrophilic / hydrophobic character of membrane surfaces are dynamic contact angle (DCA), protein adsorption studies , pulsed force mode - atomic force microscopy and surface plasmon resonance (SPR-QCM-D, Quartz

Crystal Microbalance) (see e.g. Kwock.et al. (1998) Colloids and Surfaces. 142:219-235; Meincken et al. (2005) Applied Surface Science. 252: 1772-1779; Weigert et al. (1995) Journal of Membrane Science. 106: 147-159).

The surface may be covered with a hydrophilic coating (e.g. US 2005/0133441) or being ^' functionalised. According to a preferred embodiment of the present invention the surface is part of a membrane, preferably a porous membrane, a bead, sphere, slide, filament, column, tube, particle or plate.

The surface which according to the present invention is covered by the S-layer protein lattice may be the surface of any three-dimensional object, in particular of a membrane, bead, sphere, slide, filament, column, tube, particle or plate.

In a particular preferred embodiment of the present invention the surface is part of a porous membrane, i.e. of a filter. Porous membranes can be classified as "microporous" membranes or "ultrafiltration" and "nanoporous" membranes on the basis of the size of the pores of the membrane. Generally, the range of pore sizes for microporous membranes is considered to be from approximately 0.05 μm to approximately 10.0 μm, whereas the range of pore sizes for ultrafiltration membranes is considered to be from approximately 0.002 μm to about 0.05 μm. These pore sizes refer to pore diameter for circular or approximately circular pores, or to a characteristic dimension for non-circular pores. The pore size of a membrane can be denominated by the size of the smallest species (particle or molecule) that cannot pass through the membrane above a specified fractional passage. A common rating is below 10% passage, which corresponds to a 90% cut-off or retention.

Other methods are known to those skilled in the art, including image analysis of scanning electron microscopy to obtain pore size distribution characteristics. Microporous membranes are typically used to remove particulates from liquids and gases. An important application of microporous membranes is in sterile filtration of pharmaceutical solutions to remove any bacteria that may be present in the solution. Microporous membranes are also used as sterile gas vents, which allow gas flow but prevent airborne bacteria from passing through the filter. Ultrafiltration membranes are generally used in

applications where retention of smaller species is desired. For example, ultrafiltration membranes are used in the biotechnology industry to concentrate proteins, and in diafiltration applications to remove salt and low molecular weight species from protein solutions. Both ultrafiltration and microporous membranes can be fabricated in several forms, including sheets, tubes, and hollow fibers.

Porous membranes are made from a variety of materials, polymers being the most common. Many commercial membranes are made from engineering plastics, such as polyethersulfone, polysulfone, polyvinylidene fluoride, polyethylene, polytetrafluoroethylene, polypropylene and so forth, to take advantage of their robust thermal, mechanical, and chemical- resistance properties. These materials are hydrophobic and have a high propensity " to adsorb biomolecules and cells (e.g. prokaryotic, eukaryotic) .

In general, a hydrophilic membrane which is readily wet with an aqueous solution is preferred for filtration of aqueous solutions. In contrast, with hydrophobic membranes, contact with a low surface tension pre-wetting liquid followed by water exchange is required to start permeation. This not only introduces added material cost to the process, but any wetting liquid must be exhaustively flushed, which greatly increases the possibility of contamination, process time and cost. A hydrophobic membrane which is not wet with an aqueous solution can be utilized to filter organic solutions or gases.

In addition to permeability and retentive properties, porous membranes have other requirements that are dictated by the nature of the application. For example, they must have sufficient mechanical strength to withstand operating pressures and temperatures. In applications where cleanliness is a major requirement, as in the pharmaceutical or microelectronics wafermaking industry, the amount of material that extracts from the membrane in use must be very small. In applications where, the membrane comes in contact with biomolecules, the membrane surface must be resistant to biomolecule adsorption (e.g. low protein binding) . A biomolecule resistant surface is a surface that adsorbs or binds minimal amounts of biomolecules, such as proteins and nucleic acids. It is particularly preferred that a

membrane surface is biomolecule resistant, to reduce permeation loss from fouling by surface adsorption or pore blockage and to prevent product loss by irreversible adsorption or associated biomolecule denaturization. S-layer coated membranes reduce biomolecule denaturization and thus fouling.

Porous media and membranes are used with functional surfaces. Functional surfaces have chemical groups or moieties which react with, or adsorb or absorb specific species in the fluid contacting the media or membrane. Examples of such groups are positively or negatively charged groups, affinity ligands such as antibodies or antigens, metal affinity ligands, and hydrophobic interaction groups.

To impart the aforementioned properties to porous media or membranes, the membrane surface is typically modified to make the surface hydrophilic and biomolecule resistant. This is done by a variety of procedures that coat, attach to, or otherwise cover the surface of the bulk matrix material with a hydrophilic polymer or hydrophilic group or with a hydrophobic polymer or hydrophobic group. While such modification can solve some problems related to the hydrophobic or high biomolecule binding of the bulk matrix material surface, it also can add other problems. For example, such modifications increase the amount of material able to be extracted from the membrane during use, and the modification material can have low tolerance to exposure to alkaline solutions. In addition, in many applications membranes are heated, either by wet heat as in autoclaving or steam sanitization, or by dry heat, as in a drying step. It has been observed that such heating will reduce hydrophilicity and decrease biomolecule resistance of some modified surfaces to the extent that they cannot be used for their intended purpose.

Some membranes of the prior art are made by modifying the surface of preformed porous membranes with cross-linked hydroxyacrylates, where the crosslinking monomer is a difunctional acrylate ("Case A membranes"). Such membranes have excellent resistance to biomolecule adsorption, excellent heat stability of the biomolecule resistance, and acceptable flow loss (i.e., the reduction in flow or permeability compared to the unmodified membrane) .

Much of the prior art describes the use of hydroxyl containing monomers, usually carbonyl ester containing acrylate polymers, to produce membrane surface modifications having hydrophilic character and high resistance to protein binding.

However, it is known that polymers from such monomers are not resistant to strong alkaline solutions. For example, a solution of 1.0 normal sodium hydroxide will hydrolyze the carbonyl containing acrylate polymers to acrylic acid containing polymers. Such acrylic acid containing polymers are ionically charged under certain pH conditions, and will attract and bind oppositely charged proteins or biomolecules, thus increasing sorption and membrane fouling. In addition, acrylic acid containing polymers swell in water to an extent that they constrict pore passages, thus reducing membrane permeability and productivity. Moreover, polymers from hydroxyl containing monomers, such as hydroxy acrylates, further react in strong alkaline solutions and degrade into soluble low molecular weight fragments, which dissolve away and expose the underlying substrate porous media or membrane. In contrast thereto, it could be shown that the S-layer lattice keeps its porosity and thus permeability under a large variety of physicochemical conditions.

The filter of the present invention is desirably made of a low protein binding material, such as derivatized nylon, polysulfone and the like. As used herein, a low protein binding material of a filter or porous membrane binds less than about 0.1% of protein that passes through it. This amount can be determined, for instance, by the following method. 0.4 ml of phosphate buffered saline (PBS) with 3% polyethylene glycol and 100 μl(48 μg) of radioactively labelled BSA are forced through the filter or porous material having a diameter of about 13 mm of which about 10 mm contacts the protein solution and a thickness of about 0.15 mm. The filter is then washed by forcing 5 ml of a wash solution (PBS with 0.05%, polyoxyethylene sorbitan monolaurate (Tween 20)) through it. The filter is removed and placed in 4 ml of scintillation cocktail. The amount of radioactively labelled BSA bound to the filter is calculated using disintegrations per minute measured over five minutes. The

filter material is also desirably deformable or compressible in the direction of fluid flow.

The pore size of the filter used in this invention will vary according to the analyte being assayed for and the type of sample. The pores should be sized according to the particular analyte for which the assay is being performed so that when the analyte is bound to the labelled ligand receptor pair member the resulting complex is prevented from passing through the filter. Although the pore size of the filter may be larger than the average diameter of the biomolecule or organism being assayed in order to obtain adequate sensitivity of the assay the pore size is desirably no more than 2.5 times as large as the analyte being assayed. Filters with pore sizes ranging from 0. lμm in diameter to 5 mm, preferably 0.1 μm to 1 mm, more preferably 0.1 μm to 100 μm in diameter have been used with this invention.

Preferably, filters having pores sized no larger than about 15 μm are used. Preferred filters according to the present invention are for instance, described in Elias Klein (Journal of Membrane Science 179 (2000) : 1-27) .

The membrane and/or the surface of the membrane comprises preferably a polymer selected from the group consisting of polyvinyl chloride, cellulose actetate, polyvinyl alcohol, polyvinyl pyrrolidone (PVP) , polyethylene sulfone and polysulfone.

Increasing hydrophilicity of membranes can be achieved not only through physical coating but also through blending (polyethylene glycol, PVP) , chemical modifications (HEMA grafting) , photochemical modifications, irradiation and plasma polymerization (e.g. Shudong et al. (2003), J. Membr. Sci. 222: 3-18. )

The S-layer protein to be used according of the present invention is selected preferably from the group consisting of SbpA (NCBI AAF22978) .S-layer proteins to be used according to the present invention are from gram-positive and gram-negative prokaryotic organisms and preferably from bacteria such as Bacillus or Lactobacillus Sp. (such as B. subtilis) .

However, the low protein binding surface may be covered by any S-layer protein known in the art.

The S-layer protein is preferably a natural or recombinant wild-type S-layer protein, an S-layer fusion protein, a truncated S-layer protein, a full or truncated S-layer protein with an added functionality at the N- or at the C-terminal end of the S-layer portion, an S-layer glycoprotein, or an S-layer protein mutant (e.g. Cystein mutant (see e.g. Howorka et al. (2000). J. Biol. Chem. 275 (48) : 37876-37886) ).

The S-layer protein to be used to cover a low protein binding surface may vary depending on the purpose to be served. For instance, an S-layer fusion protein is used when the surface has to be modified in order to comprise specific functionalities like binding/capturing molecules from a solution. S-layer glycoproteins may be used when molecules have to be bound to the surface recognizing specifically the glycan structure of said glycoprotein. Truncated S-layer proteins may only be truncated to such an extent which still allows the formation of crystalline S-layer protein lattices.

Natural wild-type S-layer protein as used herein refers to S-layer proteins isolated from an organism, naturally carrying said S-layer proteins on their surface.

One of the advantages of using S-layer protein covered substrates is the possibility to functionalize the surface with a wide range of molecules and nanoparticles (e.g. metals, SiOx, semiconductors, etc.). These molecules may be bound to the S- layer proteins chemically or, if the molecules are proteins, polypeptides or peptides, by recombinant fusion techniques (see e.g. Ilk et al., Appl . Environ. Microbiol. 68, 3251-3260; Pleschberger et al., Bioconjug. Chem. 14, 440-448; Vδllenkle et al., Appl. Environ. Microbiol. 70, 1514-1521; Huber et al., Small 2, 142-150; Moll et al., Proc. Natl. Acad. Sci . 99(23), 14646-14651) .

Therefore, the S-layer fusion protein is an S-layer protein fused to a molecule selected from the group consisting of streptavidin, antibodies, antigens, affinity bodies, enzymes, receptors, ligands, peptides, carbohydrates and the like.

The functional groups bound or fused to the S-layer lattice allow isolating binding partners from a liquid to which the S- layer protein covered surface is exposed to allow binding. The functionality or functionalities of the S-layer proteins may be

selected from the person skilled in the art depending on the purpose for which such a surface is needed. Of course it is also possible to use S-layer proteins of more than one species or S- layer proteins with more than one functionality. The various S- layer proteins may be crystallised at different sites on the surface resulting in a structured surface coating having areas with different coating properties. These S-layer proteins may also be mixed and then crystallised. This allows manufacturing S-layer covered surfaces with differing properties (e.g. binding properties) or areas with defined surface characteristics.

According to a further embodiment of the present invention the S-layer protein lattice on the surface has an unspecific biomolecule binding capacity of less than 150 μg/cm ² as measured by the BSA binding test.

According to a further embodiment of the present invention the S-layer protein lattice on the surface has an unspecific biomolecule binding capacity of less than 150 μg/cm ² as measured by the BSA binding test and one or more specific binding activities towards given biomolecules or organisms (achieved via S-layer fusion proteins or chemical coupling of active groups to the S-layer) (see Vδllenkle et al. 2004. Appl . Environ. Microbiol. 70:1514-1521.; Pleschberger et al. 2004. Bioconj . Chem. 15:664-671.; Pleschberger et al. 2003. Bioconj. Chem. 14:440-448.; Breitwieser et al. 2002. Protein Eng. 15:243-249.)

Another aspect of the present invention relates to a filter membrane comprising a protein covered surface according to the present invention as outlined above. S-layer proteins may be employed to cover at least partially, e.g. on predefined areas (see WO 98/39688) . Yet another aspect of the present invention relates to a method for manufacturing an S-layer protein covered surface comprising the steps of:

• providing an uncovered surface having a biomolecule binding capacity of less than 150 μg/cm ² as measured by the BSA binding test,

• contacting and incubating said surface with an S-layer protein and

• optionally washing the S-layer protein covered surface from unbound S-layer proteins.

The S-layer covered substrate may be produced by the simple crystallisation of S-layer proteins of the present invention on a surface exhibiting reduced biomolecules binding capacity.

According to a preferred embodiment of the present invention the biomolecules binding capacity is less than 100 μg/cm ² , preferably less than 75 μg/cm ² , preferably less than 50 μg/cm ² , more preferably less than 30 μg/cm ² , as measured by the BSA binding test.

The surface is preferably hydrophilic.

According to another preferred embodiment of the present invention the surface comprises hydrophilic functional groups, preferably hydroxyl groups.

The surface is preferably part of a membrane, preferably a porous membrane, a bead, sphere, slide, filament, column, tube, particle, or plate.

The membrane and/or the surface of the membrane comprises preferably a polymer selected from the group consisting of polyvinyl chloride, cellulose actetate, polyvinyl alcohol, polyvinyl pyrrolidone (PVP) , polyethylene sulfone and polysulfone.

According to a preferred embodiment of the present invention the S-layer protein is selected from the group consisting of S- layer proteins from gram-positive and gram-negative prokaryotic organisms and preferably from bacteria such as Bacillus or Lactobacillus Sp. (such as B. subtilis SbpA (NCBI AAF22978) .

The S-layer protein used in the method of the present invention is preferably a natural or recombinant wild-type S- layer protein, a chemically modified S-layer protein, an S-layer fusion protein, a truncated S-layer protein, an S-layer glycoprotein a full or truncated S-layer protein with an added functionality at the N- or at the C-terminal end of the S-layer portion, an S-layer glycoprotein, or an S-layer protein mutant (e.g. Cystein mutant, etc.).

According to a preferred embodiment of the present invention a molecule selected from the group consisting of streptavidin, antibodies, antigens, affinity bodies, enzymes, receptors, ligands, peptides, carbohydrates and the like is fused to the S-layer protein or an S-layer protein derivative (e.g. truncated form) on the S-layer protein's C- or N-terminus or internally.

The S-layer protein lattice on the surface has preferably a biomolecule binding capacity of less than 150 μg/cm ² as measured by the BSA binding test.

Another aspect of the present invention relates to a kit for manufacturing an S-layer protein covered surface, in particular an S-layer protein covered filter, comprising:

• surface with biomolecule binding capacity of less than 150 μg/cm ² as measured by the BSA binding test, and

• S-layer protein.

The components of the kit of the present invention are those described above.

The present invention is further illustrated by the following figures and example without being restricted thereto.

Fig. 1 shows SbpA re-assembled on a silicon wafer. AFM image.

Fig. 2 shows human IgG bound by rSbpA-ZZ and Protein A (Pierce) .

Fig. 3 shows self-assemblies of rSbpA-eGFP under the Fluorescence microscope: lOOx-magnified.

Fig. 4 shows SDS page, silver stained; lane 1: Marker: All blue (BioRad #:1610373), lane 2 to lane 4: Durapore membrane incubated with coating buffer (2), 0.8 nmol/ml BSA (3) and 0.8 nmol/ml wtSbpA (4); lane 5 and lane 6: wtSbpA in solution and BSA in solution, lane 7 to lane 9: Micron-PES membrane incubated with 0.8 nmol/ml wtSbpA (7), 0.8 nmol/ml BSA (8) and coating buffer only (9).

Fig. 5 shows the protein amount per mg Kisker glass beads versus the amount of protein in the coating solution.

Fig. 6 shows fluorescence microscopy images of rSbpA-ZZ coated Kisker glass beads with bound IgG (left) and without bound IgG (right ). Exposure time: 1 sec.

Fig. 7 shows fluorescence microscopy images of rSbpA-eGFP coated Kisker glass beads (left) and sole Kisker glass beads (left) .Exposure time: 20 sec.

EXAMPLE:

In the present example the ability of S-layer (fusion-) proteins and glycoproteins to bio-functionalize surfaces with low protein binding properties is shown.

1. Materials

1.1 S-layer protein SbpA of Bacillus sphaericus CCM 2177 The isolated S-layer protein SbpA from B. sphaericus (wtSbpA) have the intrinsic capability to assemble on various interfaces (air-liquid, liquid-solid, .. ) and in suspension into a 2 dimensional protein lattices (mono or double layers) termed self-assemblies. "Self assemblies" can reveal flat, tubular or vesicular structures, The S-layer fusion proteins described below were constructed to maintain the ability to self-assemble and to add a biological functionality.

1.2 Fluorescent rSbpA-eGFP S-layer fusion protein The S-layer fusion protein rSbpA-eGFP is a functional chimaeric S-layer-enhanced green fluorescent protein. Therefore, the success of the coating process can be directly visualized via fluorescence microscopy when using rSbpA-eGFP. This Slayer fusion protein is only fluorescent if correctly folded. A surface fluorescence will therefore indicate that the protein is immobilized onto the surface and that the bio-functionality is retained.

1.3 rSbpA-ZZ: Fc-Bindinq domain of protein A containing Slayer fusion protein rSbpA-ZZ is a S-layer fusion protein consisting of a C- terminally truncated form of the S-layer protein from Bacillus sphaericus CCM 2177 and two copies of the Fc-binding Z-domain. The Z-domain is a synthetic analogoue of the B. domain of protein A, capable of binding the Fc part of immunoglobulin G (IgG) . IgG binding activity of this S-layer fusion protein can be detected by ELISA technology after the coating process, which will prove that the bio-functionality is retained. Only well exposed functional groups are able to bind IgG out of solution.

Therefore, the binding of IgG indicates a non-upside-down coating of the correctly folded protein.

1.3 Surfaces

The following surfaces are known to have low-protein binding properties.

Membranes :

1) Osmonics-PES membrane (Osmonics Inc., Cat# : SO4SP02500), 0.45 μm pore size; hydrophilic, low protein binding and stable in alkaline pH.

2) Millipore Durapore® membrane filters (Millipore, Cat #: HVLP 02500), 0.45 μm pore size,; made of polyvinylidene fluoride, provides high flow rates and throughput, low extractables and broad chemical compatibility; hydrophilic Durapore membranes bind far less protein than nylon, nitrocellulose or PTFE membranes.

3) Chemical Resistant RC-membrane Filters (Sartorius, Cat# : 18406-25), 0.45 μm pore size, are hydrophilic membrane filters, made of regenerated cellulose, reinforced with non-woven cellulose.

4) Sartobind Membrane Adsorber (Sartorius, Cat .No. : 18706) ; This membrane is made of stabilized reinforced cellulose and has a overall pore size of 0.45 μm.

Beads :

1) Glass Beads (Kisker, Cat#: PGB-05, Size: 10-50 μm, package size: 50 g with 2.4 x 10 ⁷ beads/g) ; manufactured from high pure raw materials with a pure and shining surface. The density is 2.5 g/cm ³ , chemical characterization is 72% SiO ₂ , 13% Na ₂ O, 9% CaO. The glass beads are PbO free.

2. Expression, isolation and purification of proteins

2.1 rSbpA-ZZ and wtSbpA

These two proteins were produced, isolated and purified as described previously (Sleytr et al. (1986) Arch. Microbiol. 146:19-24.; Pum et al. (1997) Coll. Surf. B: Biointerfaces . 8:157-162.; Vollenkle et al. (2004) Appl. Environ. Microbiol. 70, 1514-1521).

2 . 2 Protein production : rSbpA-eGFP/ eGFP

Production of recombinant S-layer proteins in E.coli was carried out using the T7 expression system. pET-28 (+) (Novagen) served as expression vector. The expression host E.coli B121 was grown in shaking flasks in 250 ml M9ZB-medium (pH = 7.0) . At an optical density of 1 (measured at a wavelength of 600 nm) , IPTG was added to a final concentration of 1 mM and the proteins were expressed at 28°C for 5 h. Afterwards, the cells were harvested by centrifugation .

2.3 Cell disruption: rSbpA-eGFP

1 mg of biomass pellet was re-suspended in 10 ml B-Per Bacterial Protein Extraction Reagent containing 1 mM DTT (dithiothreitol) . After incubation for 10 minutes at room

temperature the suspension was centrifuged at 20000 rpm (4 ⁰ C) for 20 minutes. The supernatant contained the soluble protein fraction. Subsequently, the insoluble protein pellet was re- suspended in 10 ml B-Per containing 1 mM DTT. 200 μl of a 10 mg/ml Lysozyme stock were added and the mixture was incubated for 5 minutes at room temperature. Afterwards, 10 μl of a 1 mg/ml DNAse solution and 500 μl of a 0.1 M MgS04-solution were added and the suspension was incubated for another 30 minutes at room temperature. Hence, the suspension was kept on ice. 50 ml of a 1:10 diluted B-Per containing 1 mM DTT was added and the suspension was centrifuged at 20000 rpm (4°C) for 20 minutes. After two washing steps with 30 ml of diluted B-Per containing 1 mM DTT the suspension was centrifuged under the previous conditions and the pellet (inclusion bodies containing the S- layer fusion protein rSbpA-eGFP) was kept at 4 ⁰ C.

2.4 Cell disruption: eGFP

1 mg of biomass pellet was re-suspended in 10 ml B-Per Bacterial Protein Extraction Reagent containing 1 mM DTT (dithiothreitol) . After incubation for 10 minutes at room temperature the suspension was centrifuged at 20000 rpm (4 ⁰ C) for 20 minutes. The supernatant contained the soluble protein fraction with the eGFP in it. The protein eGFP was separated via ammonium sulfate precipitation. The protein solution was kept on ice, continuously stirred. Ammonium sulfate was grounded in a mortar and added in small portions. Saturation from 0-30%, 30-60% and 60-90% were adjusted. The precipitate was isolated via centrifugation at 25000 rpm (4°C) for 20 minutes. The protein of interest was found in the fraction with a saline saturation of 30-60%. The precipitate was dissolved in 2 ml MiIIiQ water (deionized water) . The solution was then dialyzed against RO water containing 1 mM DTT at 4°C for 24 h.

2.5 Purification of the S-laver fusion rSbpA-eGFP

S-layer protein purification was performed on FPLC equipment (AKTAprime) . A maximum of 0.2 g wet pellet (inclusion bodies) was suspended in 2.5 x the amount of extraction buffer (150 mM NaCl, 50 mM Tris-HCl, 5M GHCl) containing 1 mM DTT at room temperature for 30 min. Afterwards, the suspension was centrifuged for 30 min at 25.000 rpm at 4 ⁰ C (Beckman, Avanti JA25TM) . The supernatant was kept and an SDS-PAGE of the pellet

and the supernatant were prepared. The supernatant was diluted 1:1.5 with dilution buffer (150 rtiM NaCl, 50 mM Tris-HCl) containing 1 mM DTT. Finally, the whole amount was applied to a Superdex 200TM (GE Healthcare) column using the pump at a flow rate of 1.2 ml/min. The S-layer protein containing fractions were collected, freeze-dried and lyophilized.

2.6 Purification of eGFP

Protein purification was performed on FPLC equipment (AKTAprime) . 2 ml of the dissolved ammonium sulfate participate were dialyzed for 16 hours against 150 mM NaCl and 50 mM Tris/HCl containing ImM DTT (running buffer) . Finally, the ~5 ml were applied to a Superdex 75TM (GE Healthcare 19-0146-01) column using the pump at a flow rate of 1.2 ml/min. The protein containing fractions were collected, freeze-dried and lyophilized. The protein could be easily dissolved in water before usage and the protein concentration was stored at 4 ⁰ C.

2.7 Preparation of a monomeric protein solution of the S- layer fusion protein rSbpAeGFP/rSbpA-ZZ

1 mg of protein lyophilizate was completely dissolved in ImI of a chaotropic agent solution (5 M GHCl + 50 mM Tris/HCl, pH 7.2). The S-layer protein solution was dialyzed against 3 1 distilled water containing 1 mM DTT (only in the case of rSbpA- eGFP) to remove the chaotropic agent at room temperature. The S- , layer fusion protein had to remain in solution. The distilled water was exchanged after 30 minutes, 60 minutes and 90 minutes. After 3 hours of dialyzation, the S-layer fusion protein solution was retrieved. An additional centrifugation of the solution at 14000rpm for 5 minutes removed all precipitations (e.g. self assemblies). The final protein concentration was determined via a Lowry assay/ spectrometric at 280 nm. The protein solution was stored at 4°C.

3. S-layer as Adhesive layer

The S-layer proteins property to reassemble on surfaces with a low protein binding character (hydrophilic surfaces) was investigated. The amount of S-layer protein per defined square unit was determined by a protein assay (quantitative) and via SDS page analysis (qualitative) . To quantify the S-layer protein amount on certain surfaces, a modified Lowry assay was performed.

3.1 Coating of membranes

To draw the coating solution into the membrane, the membranes were integrated in an Amicon aperture. Primarily, 2.5 ml of MiIIiQ water and then 2.5 ml of coating buffer (0.5 mM Tris/HCl, 10 mM CaC12. pH = 9) were applied onto the membrane for the reassembling of the S-layer proteins. The solutions were run through the membrane once and then discarded. 2 ml of coating buffer containing 0.8 nmol/ml protein were applied onto the membrane, drawn in shortly and incubated for 4 hours. The protein solution was drawn in every 30 minutes putting moderate pressure on the membrane. The discharge was collected and refilled into the Amicon aperture. After the coating, the membrane was washed 5 times with 5 ml MiIIiQ water. The membrane was kept in MiIIiQ water at 4 ⁰ C until usage.

3.2 Coating of glass microspheres

0.8 nmol protein was added to 10 mg microspheres and the suspension was shaken intensely to suspend the microspheres properly. The suspension was kept on the Heidolph REAX 2 for 4 hours at room temperature. Afterwards, the microspheres were washed 3 times with MiIIiQ water and kept at 4 ⁰ C until usage.

4. Proof of functionality maintenance

These test series proved that the S-layer fusion proteins keep their biological functionality after the coating procedure. The wild type S-layer protein wtSbpA and the S-layer fusion proteins rSbpA-ZZ and rSbpA-eGFP were investigated according to standard operation procedures. The formation of the typical S- layer lattice was investigated either re-assembled on solid surfaces (Silicon wafer) via atomic force microscopy (AFM) and in solution (formation of self assemblies) via TEM techniques. Moreover, Human IgG was offered to the immobilized and the free S-layer- fusion protein rSbpA-ZZ. Afterwards, the hlgG molecules were released again via a pH-shift (down to pH = 3.5) and were detected using ELISA techniques and protein assays. The fluorescence of the rSbpA-eGFP was investigated via .Fluorescence microscopy.

5. Results

5.1 Proof of concept 5.1.1 wtSbpA

The S-layer _protein_ of Bacillus sphaericus CCM 2177 SbpA is able to re-assemble on solid supports. This was proven by coating tests under standardized conditions (Gyδrvary et al. J. Microsc-Oxford. 212: 300-306.) The typical S-layer pattern (square lattice) on Silicon wafers was observed via AFM techniques (Fig. 1) .

5.1.2 rSbpA-ZZ

The activity of the S-layer fusion protein rSbpA-ZZ in solution was investigated in comparison to Protein A (Pierce; Reacti-Bind Protein A) to guarantee the proteins quality. To both proteins Human IgG was offered which was then detected with an anti-human IgG POX conjugate (Sigma, A 0293) using ELISA techniques (Fig. 2) .

Fig. 2 shows a typical ELISA curve for both proteins. It predicated that the rSbpA-ZZ used for these test series showed a good hlgG binding activity.

5.1.3 rSbpA-eGFP

To guarantee the ability of rSbpA-eGFP to reassemble on certain interfaces (e.g. air-liquid), the formation of self- assemblies in solution was investigated via TEM techniques. Furthermore, the fluorescence of the conjugated eGFP group on self assemblies was observed by fluorescence microscopy (Fig. 3) .

Fig. 3 shows that the self-assembled S-layer fusion protein was able to emit light under UV radiation.

5.2 Coating of membranes

The hydrophilic membranes Durapor (Millipore) and Micron-PES (Osmonics) were incubated for a period of 4 hours at room temperature. The membranes were exposed to a protein concentration of 0.08 nmol/ml and were washed afterward with 3 time 5ml MiIIiQ water/per 5 cm ² .

To determine the amount of protein reassembled/ adhesively bound to the membrane, the pre-treated (coated) membranes were cooked in SDS solution and the supernatants were applied on an SDS gel for electrophoresis. Fig. 4 presents a qualitative analyze of the protein content of each membrane.

Fig. 4 shows an SDS PAGE indicating that there was more protein on the Micron-PES membrane than on the Durapore membrane

and that the amount of the S-layer protein wtSbpA was much higher than the amount of BSA attached to the membrane surface.

Subsequently, a protein assay based on Lowry was used to quantify the exact protein amount on each membrane. In this test series an S-layer fusion protein rSbpA-ZZ (presenting a humanlgG binding site) was also investigated for further functionality studies.

Table 1 Quantitative determination of the protein content of different membrane systems.

The amount of S-layer protein clearly exceeded the amount of BSA on the membrane. Furthermore, the functionality of the active group was investigated. Thus, rSbpA-ZZ coated membranes, cross-linked with DMP, were incubated with human IgG (at pH = 9.5) . The membranes had to be stabilized with 40 mM DMP to withstand the further reaction conditions. Subsequently, the bound hlgG was eluted by a pH shift (pH = 3.5) and the hlgG in solution was determined by ELISA techniques.

Table 2 Quantitative determination of the human IgG bound to membranes pre-coated with rSbpA-ZZ

Around 60% of the stabilized rSbpA-ZZ molecules were able to bind hlgG. The functionality of the human IgG binding site is

maintained after the immobilisation, stabilization and reaction procedure.

5.3 Coating of Kisker microspheres

The amount of S-layer protein which is needed in the coating solution to guarantee complete protein coverage of the microspheres (Fig. 5) was determined.

The ability of S-layer fusion proteins to reassemble on this glass surface in comparison to the wild type protein wtSbpA and the sole protein BSA was investigated. The microspheres were incubated with 0.008 nmol protein. After the wash steps, the protein content per mg Kisker glass beads, was determined by a Lowry based assay.

Table 3 Quantification of the amount of proteins which were re-assembled / absorbed on Kisker glass beads

The S-layer protein wtSbpA showed the same amount per mg glass beads. 3 times more S-layer (fusion) proteins were on the glass beads in comparison to BSA. To prove that the S-layer fusion proteins retained their functional activity after the coating process, the IgG binding of rSbpA-ZZ and the fluorescence of rSbpA-eGFP were investigated.

Human IgG was bound to rSbpA-ZZ coated glass beads, which was visualized with an Anti-IgG (goat) FITC conjugate. The fluorescence microscopy images below showed Kisker glass microspheres with and without rSbpA-ZZ coating 20x-magnified. The ability of rSbpA-ZZ to bind human IgG is retained after the coating procedure.

The S-layer fusion protein rSbpA-eGFP has the ability to emit fluorescence. The maintenance of this ability after the coating procedure was also investigated via fluorescence microscopy (Nikon Eclipse E400 with IOOW Hg-lamp for excitation

and integrated digital camera: DS Camera Control Unit DS-Ll and DS Camera head DS-5M) .

Figure 7 Fluorescence microscopy images of rSbpA-eGFP coated Kisker glass beads (left) and sole Kisker glass beads (left) . Exposure time: 20 sec.

In comparison to non-coated glass microspheres, the rSbpA- eGFP coated ones showed a fluorescence signal. As expected, the signal is weak (long exposure time is needed) due to monolayer coverage.

A camera (DS Camera Control Unit DS-Ll and DS Camera head DS-5M) was used to save copies of the microscope images.

The designation of the exposure time of the camera was done because this factor depends on the layer thickness of the protein reassembled on the glass beads thus on the amount of molecules emitting fluorescence per defined square unit. A high exposure time indicates a monolayer formation of the S-layer fusion protein rSbpA-eGFP in comparison to sole glass beads which do not show any fluorescence even at very long exposure times and FITC marked IgG which are not bound in a monolayer (IgG cluster) and therefore emitting more fluoresce light.

The ability of rSbpA-ZZ to bind IgG is retained after the coating procedure. The S-layer fusion protein rSbpA-eGFP has the ability to emit fluorescence. The maintenance of this ability after the coating procedure was also investigated via fluorescence microscopy.

6. Conclusion

The S-layer proteins are able to coat hydrophilic membranes with low protein binding characteristics. The overall amount of S-layer protein/S-layer fusion protein on the hydrophilic membrane surface is 6 times higher than the amount of BSA on the same surface. The functionality of the proteins is retained. rSbpA-ZZ is still able to bind Human IgG after the immobilization procedure. Furthermore, this indicates that the functional group is properly exposed away from to the surface (non-upside down orientation) . S-layer proteins can reassemble on pure glass beads to a higher extent than BSA. The wtSbpA and the S-layer fusion proteins showed 2.5 times more protein on the glass beads surface compared to BSA. In all cases it could be proven that the bio-functionality of the S-layer fusion proteins

is maintained and that the functional groups are correctly exposed.