Anal Bioanal Chem (2012) 403:517–526 DOI 10.1007/s00216-012-5850-9

ORIGINAL PAPER

A hydrogel-based versatile screening platform for specific biomolecular recognition in a well plate format Meike V. Beer & Claudia Rech & Sylvia Diederichs & Kathrin Hahn & Kristina Bruellhoff & Martin Möller & Lothar Elling & Jürgen Groll

Received: 5 October 2011 / Revised: 4 February 2012 / Accepted: 7 February 2012 / Published online: 26 February 2012 # Springer-Verlag 2012

Abstract Precise determination of biomolecular interactions in high throughput crucially depends on a surface coating technique that allows immobilization of a variety of interaction partners in a non-interacting environment. We present a one-step hydrogel coating system based on isocyanate functional six-arm poly(ethylene oxide)-based star polymers for commercially available 96-well microtiter plates that combines a straightforward and robust coating application with versatile bio-functionalization. This system generates resistance to unspecific protein adsorption and cell adhesion, as demonstrated with fluorescently labeled bovine serum albumin and primary human dermal fibroblasts (HDF), and high specificity for the assessment of biomolecular recognition processes when ligands are immobilized on this surface. One particular advantage is the wide range of

biomolecules that can be immobilized and convert the per se inert coating into a specifically interacting surface. We here demonstrate the immobilization and quantification of a broad range of biochemically important ligands, such as peptide sequences GRGDS and GRGDSK-biotin, the broadly applicable coupler molecule biocytin, the protein fibronectin, and the carbohydrates N-acetylglucosamine and Nacetyllactosamine. A simplified protocol for an enzymelinked immunosorbent assay was established for the detection and quantification of ligands on the coating surface. Cell adhesion on the peptide and protein-modified surfaces was assessed using HDF. All coatings were applied using a one-step preparation technique, including bioactivation, which makes the system suitable for high-throughput screening in a format that is compatible with the most routinely used testing systems.

Meike V. Beer and Claudia Rech have contributed equally.

Keywords Bioassays . Immunoassays/ELISA . Interaction screening . Biofunctionalization . Hydrogel coating . 96-Well plate format

M. V. Beer : K. Hahn : J. Groll (*) Department of Functional Materials in Medicine and Dentistry, University of Würzburg, Pleicherwall 2, 97070 Würzburg, Germany e-mail: [email protected] C. Rech : L. Elling (*) Laboratory for Biomaterials, Institute for Biotechnology and Helmholtz Institute for Biomedical Engineering, RWTH Aachen University, Pauwelsstr. 20, 52074 Aachen, Germany e-mail: [email protected] S. Diederichs : K. Bruellhoff : M. Möller Interactive Materials Research Institute (DWI e.V.) and Institute of Technical and Macromolecular Chemistry, RWTH Aachen University, Forckenbeckstr. 50, 52074 Aachen, Germany

Abbreviations BSA ECM ELISA ELLA FN GlcNAc GSII HDF His6CGL2 IPDI

Bovine serum albumin Extracellular matrix Enzyme-linked immunosorbent assay Enzyme-linked lectin assay Fibronectin N-Acetylglucosamine Lectin II from Griffonia simplicifolia Primary human dermal fibroblast Recombinant lectin produced in Escherichia coli BL21 Isophorone diisocyanate

518

LacNAc NCO-sP(EO-stat-PO) POD SA

M.V. Beer et al.

N-acetyllactosamine Isocyanate functionalized six-arm star-shaped prepolymer Peroxidase Streptavidin

Introduction Surfaces used in biosensors and for biomolecular interaction studies need to fulfill two major criteria; minimization of unspecific interactions and high specificity for the recognition of immobilized bioactive ligands [1, 2]. The combination of these two properties ensures control of the interaction of biomolecules on the surface with high signal-to-noise ratio for the specific biomolecular detection system. A number of different systems have been developed that fulfill these criteria. Coatings based on poly(ethylene glycol) (PEG) have been widely used for sensors and interaction screenings due to the profound ability of PEG to resist protein adsorption and cellular adhesion [3]. In most cases, self-assembled monolayers (SAMs) are prepared on ultraflat gold surfaces from thiofunctional PEG or alkane–PEG with hydroxyl or methylether groups pointing towards the solution interface [4–10]. Functionality may be introduced by a fraction of PEG-bearing functionalizable endgroups such as amines or carboxylates for SAM preparation and subsequent chemical derivatization [11, 12]. For example, functionalized 8-amino-3,6dioxaoctan-terminated surfaces with hexamethylenediisocyanate were used by Bradner et al. for immobilization of several ligands in a printed microarray with inter-spot surface quenching by ethylene glycol [13]. Also a variety of carbohydrates were used for functionalization of PEGSAMs on gold surfaces via Michael addition [14]. However, most of the mentioned techniques are restricted to specialized sensors or model substrates and cannot easily be transferred into standard and widely used 96-well microtiter plate format. In contrast to microarrays, hydrogelcoated microtiter plates offer the possibility to combine the advantages of biofunctionalized hydrogel surfaces with simple and low-cost readout equipment. Ideally, a coating procedure would be easy to apply without the need for complex instrumentation and allow the binding of a broad range of different ligands. Automated readouts of processes such as cytotoxicity assays, surface-dependent bacterial and cell growth, as well as screening of active agent concentrations, only to name a few, can be simplified by using such coated well plates. Ligand densities and accessibility on such surfaces is crucial for evaluation of interaction efficiency. Immobilization of biotin on the surface of hydrogel surfaces was optimized and quantified by enzyme-linked immunosorbent assay (ELISA) [15, 16]. It becomes clear that the

polymer composition of the hydrogels determines the surface concentration of accessible ligand making such surfaces highly specific for recognition of the ligand. The reported hydrogel surfaces need to be optimized separately for each new hydrogel surface/ligand combination. We present here a hydrogel coating which can be applied and bioactivated by simple incubation steps in commercial well plates and is highly flexible for the immobilization of various biomolecules and applicable for ELISA and ELLAbased as well as cell growth assays. In previous work, we introduced NCO-sP(EO-stat-PO) as a coating system for glass, silicon [17], and silicone rubbers [18]. This system relies on six-arm star-shaped molecules with a backbone of statistically copolymerized ethylene oxide (EO) and propylene oxide (PO) in the ratio 4:1 with a molecular weight of 12 kDa (Fig. 1). The distal ends of the star-shaped molecules are functionalized with aliphatic isocyanate groups. We could show that aqueous solutions of these molecules can be used to prepare coatings on amino-functionalized surfaces. Covalent attachment of the molecules to the surface as well as intermolecular cross-linking through hydrolysis of the isocyanates to amines and subsequent reaction with other isocyanates to urea-bridges between molecules leads to a three-dimensionally cross-linked polymer network on the surface with high polymer segment density. After completion of the cross-linking reaction, these surface coatings minimize protein adsorption and resist cell adhesion [19]. Most importantly, one particular advantage of the system is the straightforward possibility to covalently bind ligands without the use of further chemical activation steps or cross-linkers. Solutions of NCO-sP(EO-stat-PO) and freshly prepared coatings up to several hours after coating still contain reactive isocyanate groups. Their concentration decreases by time and vanishes after 8 h due to the water induced cross-linking reaction described above. Using the reactivity of isocyanates towards nucleophilic groups such as thiols, amines, or alcohols, functionalization can be achieved either through mixing of the ligands directly in the solution prior to coating formation or by incubation of freshly prepared coatings with ligand solution [20]. This can be achieved by simple layer incubation or by patterning techniques without the need for additional cross-linkers. Higher reactivity of thiols and amines can also be used for preferential binding, and the system may be used for microarray applications and on chip single base extension for single-nucleotide polymorphism detection [1]. For the latter, another advantage of the system becomes apparent. Since the reactive isocyanate groups are all positioned at the distal ends of the star molecules with a molecular weight of 2 kDa, all immobilized ligands are automatically separated from the polymer network through long spacer molecules that enable highly efficient binding and activity of the polymerase enzyme on the surface-bound sequences.

Hydrogel-based versatile screening platform in well plate format

In this study, we established a robust and reliable onestep coating procedure of 96-well plates with NCO-sP(EOstat-PO) and its possible applications for in vitro screenings and bioassays (Fig. 1). Commercially available aminofunctional 96-well plates were used as substrates, and minimized protein adsorption as well as cell adhesion was demonstrated using fluorescently labeled bovine serum albumin (BSA) and primary human dermal fibroblasts (HDF), respectively. To demonstrate the spectrum of ligands that can be immobilized, a range of molecules were covalently embedded into the layer. The amino-functional biotin derivative biocytin was used as widely used standard linker that specifically recognizes streptavidin (SA) with high affinity. Additionally, the extracellular matrix protein fibronectin (FN), the FN-derived cell adhesion-mediating peptide sequence GRGDS and the GRGDSK-biotin, and the aminofunctionalized carbohydrates N-acetylglucosamine (GlcNAc) as well as N-acetyllactosamine (LacNAc) were immobilized on the hydrogel coatings. ELISA and ELLA protocols were established for the different molecules to determine the ligand density. While GlcNAc serves as mere proof of principle for the immobilization and detection of a sugar ligand, LacNAc was selected because of its important role in vivo as basic component for biological relevant epitopes like blood group ABO or Lex on glycoproteins and glycolipids [21–26]. Furthermore, LacNAc is a ligand for galectins, a family of galactose binding proteins, which mediates cell–cell and cell– matrix interactions [27, 28].

Fig. 1 NCO-sP(EO-stat-PO) chemistry, cross-linking reaction in water (a) and scheme of the coating procedure and specific ligand presentation for interaction assays in well plate format (b). Chemistry of the NCO-sP(EO-stat-PO) system with statistically copolymerized ethylene oxide and propylene oxide in the polymer backbone and IPDI-derived endgroups and the water-induced cross-linking reaction of the

519

Materials and methods Coating NCO-sP(EO-stat-PO) prepolymers were synthesized as described before [29]. Prepolymers were dissolved in tetrahydrofuran (THF; dried over sodium, VWR, Darmstadt, Germany). After addition of water (9:1 v/v mixture of water: THF, prepolymer concentration 1–10 mg/mL), the solution was kept at room temperature for 5 min. Subsequently, 400 μL of the solution were filled in each well of a CovaLink™NH 96-well microtiter plate (NUNC, Langenselbold, Germany). After 1, 10, or 15 min incubation at room temperature, polymer solution was removed and the plates stored at room temperature. Functionalization Biocytin and human plasma FN were purchased from SigmaAldrich (Steinheim, Germany). The peptides GRGDS (for cell adhesion) and GRGDSK-biotin (for quantification of peptide density on the surface; lysine serves as spacer and linker, biotin is bound to the ε-amino-group of lysine through an amide bond) were obtained from Bachem (Bubendorf, Switzerland). Amino-functionalized GlcNAc and LacNAc structures bearing terminal primary amine groups (for chemical structure see Fig. 5) were synthesized as described earlier [30, 31]. The ligand solutions (50 μL) were incubated on coated

isocyanate functional prepolymers (a) and scheme of well plate coating with a functional hydrogel layer that may be functionalized with a variety of bioactive ligands (peptide, sugar, protein) (b). This coating system enables the detection of specific interactions of surface immobilized ligands with proteins, cells, or peptides

520

wells (1 h after preparation of the coating). Ligand concentrations were varied as follows—0–500 μg/mL (biocytin, GRGDSK-biotin), 0–20 μg/mL (FN), and 0–10 mM (GlcNAc, LacNAc). The concentration of (GRGDS) was set to 100 μg/mL. GRGDS, biocytin, GRGDSK-biotin, and carbohydrates were solubilized in sodium carbonate buffer (0.02 M Na2CO3/NaHCO3, pH 9.4); FN was dissolved in deionized water. After 1 h incubation, wells were washed with 300 μL deionized water.

M.V. Beer et al.

ELISA Wells were coated and functionalized as described above and left at room temperature for at least 12 h to ensure complete hydrogel cross-linking. Incubation of the wells with 300 μL deionized water for 60 min was followed by a washing step (three-times) with 300 μL PBS–Tween (0.05% v/v, Tween-20, Roth, Karlsruhe, Germany). Biocytin/GRGDSK-biotin

Protein adsorption Protein adsorption was tested on uncoated CovaLink™wells as well as NCO-sP(EO-stat-PO) coated and functionalized hydrogel coatings. For functionalization, 100 μg/mL GRGDS and 5 μg/mL FN were used. Wells were left at room temperature for 24 h and incubated in distilled water for 60 min. A solution (50 μg/mL in 0.01 M phosphate buffered saline (PBS) buffer, pH 7.4) of fluorescent-labeled bovine serum albumin (BSA-tetramethylrhodamin, BSATR, Invitrogen, Karlsruhe, Germany) was prepared. The protein solution (100 μL) was then incubated for 60 min in each well followed by washing the plates three times with PBS buffer and distilled water, respectively. Protein adsorption was observed by fluorescence microscopy using an Axioplan 2 imaging microscope (Carl Zeiss, Goettingen, Germany) within the same day. Pictures were taken with an AxioCamMRc digital camera and analyzed using the AxioVisionV4.6 software package. A constant exposure time of 4 s was used to assure comparability. Cell culture Primary HDF (maximum passage, 8) isolated from foreskin were kindly provided by Prof. Baron (Department of Dermatology and Allergology, University Hospital of the RWTH Aachen University, Germany). Cells were cultured in Dulbecco’s modifies Eagle’s medium (DMEM) medium (Invitrogen Darmstadt, Germany) supplemented with 10% fetal bovine serum (Biowest, Nuaillé, France) and 1% penicillin/streptomycin (PAA, Cölbe, Germany) at standard cell culture conditions (37 °C, 5% CO2, 95% humidity). Cells were harvested with accutase (PAA, Cölbe, Germany). Three hundred microliters cell suspension (20,000 cells/ mL in DMEM) were seeded in each well and incubated under standard cell culture conditions for 2 weeks. Cells were seeded on non-functionalized, NCO-sP(EO-stat-PO) coated, and hydrogel coatings functionalized with GRGDS (100 μg/mL) and FN (5 μg/mL), respectively. On NCO-sP (EO-stat-PO)-coated wells, cells were reseeded after 1 and 2 weeks. Cell adhesion and spreading were monitored by optical microscopy using an inverted Axiovert 100A imaging microscope (Carl Zeiss, Goettingen, Germany).

Wells were incubated with 300 μL glycidol solution (2.23 mg/mL in 0.2 M bicarbonate buffer) for 60 min followed by washing three times with 300 μL PBS–Tween. Ligands were detected by incubation with 50 μL streptavidin–peroxidase (SA-POD; 1:3,000 in PBS, Roche, Mannheim, Germany) for 60 min. Wells were washed again with PBS–Tween. After dissolving one OPD tablet (Dako, Hamburg, Germany) in 3 mL deionized water and 1.25 μL H2O2 (30%, v/v), 100 μL of the OPD solution were added to each well. After 1 min incubation, the reaction was stopped by adding 100 μL 3 M HCl. Optical density (OD) of each well was measured in the microplate reader model Sunrise (Tecan, Maennedorf, Switzerland) at a wavelength of 492 nm. Results are shown on a logarithmic scale. Negative control values were below 0.1 (results not shown). FN Primary rabbit-anti-human FN antibody and goat-anti-rabbit IgG-peroxidase were purchased from Sigma-Aldrich (Steinheim, Germany). After coating, functionalization, and incubation with deionized water as described above, wells were incubated with 50 μL of a 1:1,000 dilution of anti-FN antibody in PBS for 60 min and washed three times with 300 μL PBS–Tween. Fifty microliters of a 1:1,000 dilution of goat-anti-rabbit IgG-peroxidase in PBS was added in each well for 60 min. Afterward, wells were washed with PBS–Tween again. Detection with OPD solution and OD measurement was carried out as described above. GlcNAc/LacNAc Lectin II from Griffonia simplicifolia (GSII) was purchased from Vector Laboratories (Burlingame, USA). The recombinant lectin His6CGL2 was produced in E. coli BL21 (DE3) and purified as described earlier [31] and stored in PBS. Coatings functionalized with the sugar GlcNAc were incubated with glycidol solution as described above. After three times washing with 300 μL PBS–Tween, GlcNAc was detected with 50 μL of the biotinylated lectin GSII (10 μg/ mL in 10 mM HEPES, pH 7.5). Coatings carrying the carbohydrate LacNAc were incubated with 50 μL of the

Hydrogel-based versatile screening platform in well plate format

recombinant galectin His6CGL2 (50 μg/mL, in PBS) with subsequent washing for three times with 300 μL PBS– Tween. The detection of lectins was carried out by incubation with 50 μL streptavidin–peroxidase (1:1,000 in PBS) for biotinylated GSII bound to GlcNAc and 50 μL antiHis6-peroxidase from mouse IgG2a (1:4000 in PBS, Roche, Mannheim, Germany) for the recombinant galectin His6CGL2 bound to LacNAc for 60 min. After washing, the detection of bound peroxidase was carried out as described above with 2 min incubation of the OPD substrate.

Results and discussion Preparation of the coatings Coating of glass slides and silicon substrates [17] or silicone rubbers [18] with NCO-sP(EO-stat-PO) requires surface pretreatment steps such as activation using UV/ozone or oxygen plasma and amino-functionalization, for example, through reaction with aminopropyl–triethoxysilane or the use of ammonia plasma. The aim of this study was to establish a coating protocol that may be applied using standard laboratory equipment in well plate format. Therefore, commercially available amino-functional 96-well plates were used without any further surface treatment of the well plates. NCO-sP(EO-stat-PO) prepolymer was dissolved in THF. After a homogeneous solution was formed, water was added in excess so that the final water content of the solution was 90 vol.%. For preparation of the coatings, aminofunctional wells were filled with the solution directly after addition of water to the prepolymer solution in THF or after 1 and 5 min of water addition for a defined period of time followed by removal of the solution from the wells and storage of the wells at ambient conditions. During optimization of the layer preparation, the concentration of prepolymer in the solution used for coating was varied between 1 and 10 mg/mL, and the incubation time in the well plates was 1, 10, or 15 min. The resulting coatings were assessed regarding protein adsorption and cell adhesion. Protein adsorption and cell adhesion on unmodified coatings For analysis of protein-repellent properties of the NCO-sP (EO-stat-PO) coatings in 96-well plates, uncoated and coated wells were incubated with fluorescently labeled BSA, as BSA is an often used model protein for assessment of non-fouling biomaterials [32]. Integration time was kept constant for all samples to allow compatibility of the results. Strong protein adsorption occurred on uncoated CovaLink™ wells (Fig. 2a). For incubation with NCO-sP(EO-stat-PO) at concentrations lower than 10 mg/mL and incubation times shorter than 10 min, protein adsorption was reduced but not minimal (results not shown). On coatings prepared from 10 mg/mL

521

prepolymer solutions with a minimum of 10 min incubation time, no fluorescence could be observed, indicating minimized protein adsorption (Fig. 2b). Addition of the solutions into the well plates directly after mixing of water with the prepolymer solution in THF or after 1 or 5 min, respectively, did not result in significant changes. An incubation time of 10 min directly after preparation of the coating solution was thus chosen for assessment of cell adhesion. Primary HDF are cells that exhibit strong ability to adhere to surfaces and were therefore chosen as cells to test the coatings in well plates for their ability to resist cell adhesion. HDFs could adhere on uncoated CovaLink™ surfaces (Fig. 2c). On hydrogel-coated wells, HDFs could not adhere after 24 h (Fig. 2d). To demonstrate long-term cell resistance, media was changed every 2 days, and new cells were re-seeded twice on the same coatings after 1 and 2 weeks. This allows assessment of whether proteins from the medium or proteins excreted from cells are able to adsorb on the coatings with time, thereby creating a cell adhesive environment. Life images were taken 24 h after each re-seeding (Fig. 2e, f). Over the whole period of time, no cell adhesion was observed on the coatings, even after the second reseeding of new cells after 2 weeks of incubation with constantly renewed fully supplemented cell culture medium. Hence, coatings with minimized adsorption of the model protein serum albumin that suppress cell adhesion were obtained with a prepolymer concentration of 10 mg/mL and an incubation time of 10 min which was used for further experiments as standard parameters.These results show that our hydrogel platform is a robust and reliable platform that minimizes unspecific interaction as crucial step toward interaction screening on biofunctionalized coatings that is presented in the next paragraph. Specific interaction on biofunctionalized coatings For the assessment of NCO-sP(EO-stat-PO) coatings functionalized with GRGDS and FN were respectively tested for protein adsorption and cell adhesion (Fig. 3). GRGDS and FN functionalized coatings did not exhibit higher unspecific BSA adsorption (Fig. 3a, b) than non-functionalized NCOsP(EO-stat-PO) coatings. Hence, modification of the coatings with peptide sequences or even whole proteins does not hamper the ability of the hydrogel coating to minimize unspecific protein adsorption. Cell adhesion experiments on NCO-sP(EO-stat-PO) coatings functionalized with GRGDS and FN showed cell adhesion, spreading, and proliferation of HDFs after 24 h (Fig. 3c, d), proving the presence of cell adhesion possibilities in comparison to the non-modified coatings. With the protein adsorption data presented above, we conclude that cell adhesion is solely mediated by the immobilized cell adhesion ligands without additional interaction of the cells with the hydrogel

522

M.V. Beer et al.

Fig. 2 BSA-TR adsorption and HDF adhesion on non-coated and coated wells. Protein adsorption of BSA-TR on uncoated (a) and NCO-sP(EOstat-PO) coated (b) CovaLink™ wells. HDF adhesion on uncoated (c) and NCO-sP(EOstat-PO) coated (d–f) CovaLinkTM wells was observed via life cell microscopy after 24 h. In case of coated wells, cells were re-seeded after 1 and 2 weeks and pictures taken after 24 h. Coatings were prepared with optimal parameters (10 mg/mL prepolymer, 10 min incubation in the wells)

background. For GRGDS coatings, cells were observed over a longer period of time up to 2 weeks, showing that cells proliferate and survive on peptide-functionalized coatings (Fig. 3e, f). These results demonstrate the potential of the biofunctionalized coatings in well plates to use the inert properties of the hydrogel for cell–ligand interaction screenings (here human dermal fibroblasts) by introduction of cell adhesion-mediating ligands. Consequently, this method can be used for sensitive and automated readouts of processes as cytotoxicity tests, impact of ligands on cell growth, and screening for active agent concentrations.

observed unspecific binding of SA-POD to the coatings. This could however completely be prevented by incubation with glycidol to convert amino groups into hydrophilic diols. One explanation could be the structural similarity of the biotin molecule and the isophorone ring of the isophoronediisocyanate (IPDI) as part of the hydrogel coating which may be recognized by the SA. In case of ELISAs using antibodies instead of SA, glycidol incubation was not necessary.

Biocytin, GRGDSK-biotin, and FN Quantification of interactions using ELISA and ELLA techniques Besides cell adhesion experiments, the NCO-sP(EO-statPO) coating is also compatible with ELISA- and ELLAbased assays. Due to the high functionality of the hydrogel, specific ligands can easily be introduced. In addition, these assays on coated well-plates do not afford blocking of unspecific binding sites because coatings resist unspecific protein adsorption (Fig. 3a, b). Only in assays involving avidin or streptavidin, such as ELISAs using SA-POD, we

With our coating system that can be easily functionalized with different types of ligands, such as small molecules, peptides, or proteins, a variety of specific interactions can be detected. Biocytin was chosen as a model and standard molecule representing the ability to detect even small molecules immobilized on the coatings. With its NH2 group, it can covalently bind to fresh NCO-sP(EO-statPO) coatings still containing NCO groups. Figure 4a shows that a maximum biocytin binding to the coating is reached at 100 μg/mL.

Hydrogel-based versatile screening platform in well plate format

523

Fig. 3 BSA-TR adsorption and HDF adhesion on functionalized coatings. Protein adsorption of BSA-TR on NCO-sP(EO-stat-PO) coatings functionalized with GRGDS (a) and FN (b). HDF adhesion on NCO-sP(EO-stat-PO) functionalized with GRGDS and FN was observed via life cell microscopy after 24 h (c, d). In case of GRGDS-functionalized coatings, cells were left for 2 weeks (e, f). Coatings were prepared with optimal parameters (10 mg/mL prepolymer, 10 min incubation in the wells)

We utilized the biotin technology also for the detection of peptide sequences immobilized on the hydrogel coating. Biotin was coupled to the cell adhesion-mediating peptide GRGDS via an additional C-terminal lysine residue (GRGDSK-biotin) that could be easily detected by SA. In principle, peptide biotinylation, which can also be achieved via terminal cysteins and Michael addition to avoid interference with lysines in the functional sequence of longer peptides, can be used for any peptide sequence of interest allowing the screening of a variety of peptides using the same ELISA procedure. Figure 4b shows a maximal binding of GRGDSK-biotin to the coating at 200 μg/mL. This resembles the results of functionalization with biocytin indicating that the binding mechanism is similar. The ECM protein FN was also immobilized on the hydrogel coatings and detected via ELISA. A maximum binding capacity was reached at a FN concentration of 5 μg/mL (Fig. 4c). In summary, our results demonstrate that coating microtiter plate with NCO-sP(EO-stat-PO) allows covalent immobilization of any small ligands, peptides, and proteins.

Carbohydrates Hydrogel coating of microtiter plates with NCO-sP(EO-statPO) can also be utilized for the immobilization of aminofunctionalized carbohydrates. We chose the sugar GlcNAc and LacNAc to demonstrate proof of principle for lectin recognition and interaction with a galectin, respectively. LacNAc was chosen because of its importance for galectin-mediated cell–cell and cell–matrix interactions in vivo. Immobilizing LacNAc glycan structures on an inert surface would allow a more natural presentation of ECM proteins mediated by galectins. Coupling of the sugars to NCO-sP(EO-stat-PO) coatings was accomplished by deprotection of the amino group at a hydrophobic linker [31]. Detection of the immobilized carbohydrates was performed by ELLA (enzyme-linked lectin assay) using lectins, namely, the GlcNAc-specific lectin II from G. simplicifolia (GSII) and the LacNAc-specific recombinant galectin His6CGL2 from Coprinus cinereus [31, 33]. Figure 5 demonstrates that binding of both lectins to the corresponding carbohydrates increases with higher sugar concentrations. In the case of GlcNAc saturation, binding

524

Fig. 4 ELISA on functionalized NCO-sP(EO-stat-PO) coatings. Relative optical density of ELISAs on NCO-sP(EO-stat-PO) coatings functionalized with different concentrations of biocytin (a), GRGDSK-biotin (b), and FN (c). Biocytin and GRGDSK-biotin was detected by streptavidin–peroxidase conjugate and FN by rabbit-antihuman FN antibody and goat-anti-rabbit IgG-peroxidase

of the lectin to the immobilized monosaccharide on the coating could not be reached up to 10 mM sugar concentration (Fig. 5a). These first assays show that the binding of Fig. 5 ELLA on NCO-sP(EOstat-PO) coatings functionalized with GlcNAc and LacNAc. Modified carbohydrates GlcNAc and LacNAc used in this work and relative optical density of ELLAs on NCO-sP (EO-stat-PO) coatings functionalized with different concentration of GlcNAc detected by GSII (10 μg/mL) (a) and functionalized with different concentration of LacNAc detected by His6CGL2 (50 μg/mL) (b)

M.V. Beer et al.

GlcNAc detected by GSII on the coated well plates is at least as good as optimized assays in commercial microtiter plates [31]. The disaccharide LacNAc interacting with the galectin His6CGL2 was chosen as a more complex and relevant carbohydrate ligand. A maximal binding of the galectin His6CGL2 was reached at incubation with 5 mM LacNAc solution in a reproducible manner (Fig. 5b). Our results demonstrate that immobilization and subsequent detection of the carbohydrates on hydrogel-coated microtiter plates is possible and opens the possibility to use this platform as for the screening of a variety of synthesized glycan structures. In addition, this system possesses the advantage of further construction of more complex structures on the hydrogel surface such as binding of proteins followed by cell adhesion studies which is not possible with the systems that are available so far. In summary, we have established the hydrogel coating of 96-well plates with NCO-sP(EO-stat-PO) and proved its protein- and cell-repellent properties. The coated wells can be utilized for the screening of ligands with biochemical and cellular systems. As an example, functionalization with the cell adhesion-mediating peptide sequence GRGDS and the protein FN resulted in specific cell adhesion of HDFs confirming data of coatings on glass from earlier studies [19]. Additionally, the coatings were functionalized with different ligands: a low-molecular structure (biocytin), a peptide (GRGDSK-biotin), a protein (FN), and carbohydrates (GlcNAc, LacNAc). Due to the small scale of the system, detection and quantification of the ligands with streptavidin, antibodies, and lectins using ELISA/ELLA techniques were possible. In all cases, ligands presented on NCO-sP(EO-statPO) hydrogel surfaces could also be quantified. The great advantage of functionalized NCO-sP(EO-stat-PO) coatings is their inert surface and therefore the specific binding to the immobilized ligands without the need of a blocking step. ELISAs showed no background signal by unspecific binding to the coatings. Only SA was found to slightly bind to untreated coatings, which could be eliminated by glycidol treatment.

Hydrogel-based versatile screening platform in well plate format

525

Conclusions The aim of this study was the development of a hydrogel coating system for 96-well plates that can be applied by simple incubation and washing steps. A key demand was the ability to functionalize the system with a variety of chemically different ligands with the same straightforward techniques. To achieve this, we established a procedure to apply the NCO-sP(EO-stat-PO) hydrogel coating system in commercially available amino-reactive 96-well plates. The coatings were shown to minimize adsorption of BSA and adhesion of primary HDF. Functionalization of the per se inert coatings in the well plates is possible with a variety of ligands. Immobilization of cell adhesion-mediating peptides such as GRGDS and the protein FN introduced specific adhesion of HDFs on the coatings through exclusive interaction of the cells with the ligands. Diverse bioactive ligands, such as biotin, FN, and the sugars LacNAc and GlcNAc that were covalently bound to inert hydrogel coatings, were specifically detected and quantified via ELISA. Consequently, we demonstrated an easy applicable method that has the potential to serve as a sensitive detection method with the ability to create an on-demand surface design directly by the end-user. With the ability to present variable ligands, the system can serve as a screening platform for automated readouts of ligands and binding partners, cytotoxicity tests, impact on cell growth, and screening for active agent concentrations. Interaction studies on this variable coating system can be easily established in laboratories, as the 96-well plate format is compatible for most established testing systems and may be used as basis for the screening of biomolecular interaction in high throughput. Acknowledgment The authors acknowledge funding by the Deutsche Forschungsgemeinschaft (DFG) within the Research Training Group 1035 “BioInterface” (M.V.B., C.R., M.M., L.E.) and the SPP 1257 “Intelligent Hydrogels” (M.M., J.G.). The authors thank Sarah Krauthausen (Interactive Materials Research Institute (DWI e.V.), Aachen) for excellent technical assistance.

References 1. Geistlinger J, Du W, Groll J, Liu F, Hoegel J, Foehr KJ, Pasquarelli A, Schneider EM (2011) P2RX7 genotype association in severe sepsis identified by a novel multi-individual array for rapid screening and replication of risk SNPs. Clin Chim Acta. doi:10.1016/j. cca.2011.05.023 2. Zhang M, Desai T, Ferrari M (1998) Proteins and cells on PEG immobilized silicon surfaces. Biomaterials 19:953–960 3. Feldman K, Haehner G, Spencer ND, Harder P, Grunze M (1999) Probing resistance to protein adsorption of oligo(ethylene glycol)terminated self-assembled monolayers by scanning force microscopy. J Am Chem Soc 121:10134–10141 4. Edinger K, Golzhauser A, Demota K, Woll C, Grunze M (1993) Formation of self-assembled monolayers of n-alkanethiols on gold—

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

18.

19.

20.

21.

a scanning tunneling microscopy study on the modification of substrate morphology. Langmuir 9:4–8 Harder P, Grunze M, Dahint R, Whitesides GM, Laibinis PE (1998) Molecular conformation in oligo(ethylene glycol)-terminated self-assembled monolayers on gold and silver surfaces determines their ability to resist protein adsorption. J Phys Chem B 102:426–436 Mehne J, Markovic G, Proll F, Schweizer N, Zorn S, Schreiber F, Gauglitz G (2008) Characterisation of morphology of selfassembled PEG monolayers: a comparison of mixed and pure coatings optimised for biosensor applications. Anal Bioanal Chem 391:1783–1791 Mrksich M, Whitesides GM (1996) Using self-assembled monolayers to understand the interactions of man-made surfaces with proteins and cells. Annu Rev Biophys Biomol Struct 25:55–78 Peppas NA, Hilt JZ, Khademhosseini A, Langer R (2006) Hydrogels in biology and medicine: from molecular principles to bionanotechnology. Adv Mater 18:1345–1360 Prime KL, Whitesides GM (1991) Self-assembled organic monolayers—model systems for studying adsorption of proteins at surfaces. Science 252:1164–1167 Wang RLC, Kreuzer HJ, Grunze M (1997) Molecular conformation and solvation of oligo(ethylene glycol)-terminated selfassembled monolayers and their resistance to protein adsorption. J Phys Chem B 101:9767–9773 Anderson AS, Dattelbaum AM, Montano GA, Price DN, Schmidt JG, Martinez JS, Grace WK, Grace KM, Swanson BI (2008) Functional PEG-modified thin films for biological detection. Langmuir 24:2240–2247 Kawaguchi T, Shankaran DR, Kim SJ, Gobi KV, Matsumoto K, Toko K, Miura N (2007) Fabrication of a novel immunosensor using functionalized self-assembled monolayer for trace level detection of TNT by surface plasmon resonance. Talanta 72:554–560 Bradner JE, McPherson OM, Mazitschek R, Barnes-Seeman D, Shen JP, Dhaliwal J, Stevenson KE, Duffner JL, Park SB, Neuberg DS, Nghiem P, Schreiber SL, Koehler AN (2006) A robust smallmolecule microarray platform for screening cell lysates. Chem Biol 13:493–504 Cheng F, Shang J, Ratner DM (2011) A versatile method for functionalizing surfaces with bioactive glycans. Bioconjugate Chem 22:50–57 Behravesh E, Sikavitsas VI, Mikos AG (2003) Quantification of ligand surface concentration of bulk-modified biomimetic hydrogels. Biomaterials 24:4365–4374 Sung D, Park S, Jon S (2009) Facile method for selective immobilization of biomolecules on plastic surfaces. Langmuir 25:11289– 11294 Groll J, Ameringer T, Spatz JP, Moeller M (2005) Ultrathin coatings from isocyanate-terminated star PEG prepolymers: layer formation and characterization. Langmuir 21:1991–1999 Salber J, Grater S, Harwardt M, Hofmann M, Klee D, Dujic J, Huang JH, Ding JD, Kippenberger S, Bernd A, Groll J, Spatz JP, Moller M (2007) Influence of different ECM mimetic peptide sequences embedded in a nonfouling environment on the specific adhesion of human-skin keratinocytes and fibroblasts on deformable substrates. Small 3:1023–1031 Groll J, Fiedler J, Engelhard E, Ameringer T, Tugulu S, Klok HA, Brenner RE, Moeller M (2005) A novel star PEG-derived surface coating for specific cell adhesion. J Biomed Mat Res Part A 74A:607–617 Gasteier P, Reska A, Schulte P, Salber J, Offenhausser A, Moeller M, Groll J (2007) Surface grafting of PEO-based star-shaped molecules for bioanalytical and biomedical applications. Macromol Biosci 7:1010–1023 Cummings RD, Kornfeld S (1984) The distribution of repeating [Galbeta1,4GlcNAcbeta1,3] sequences in asparagine-linked

526

22.

23.

24.

25.

26. 27.

M.V. Beer et al. oligosaccharides of the mouse lymphoma cell-lines BW5147 and PHAR2.1—binding of oligosaccharides containing these sequences to immobilized datura–stramonium agglutinin. J Biol Chem 259:6253–6260 Fukuda M, Carlsson SR, Klock JC, Dell A (1986) Structures of Olinked oligosaccharides isolated from normal granulocytes, chronic myelogenous leukemia-cells, and acute myelogenous leukemiacells. J Biol Chem 261:2796–2806 Krusius T, Finne J, Rauvala H (1978) Poly(glycosyl) chains of glycoproteins—characterization of a novel type of glycoprotein saccharides from human erythrocyte-membrane. Eur J Biochem 92:289–300 McEver RP, Moore KL, Cummings RD (1995) Leukocyte trafficking mediated by selectin–carbohydrate interactions. J Biol Chem 270:11025–11028 Ujita M, McAuliffe J, Suzuki M, Hindsgaul O, Clausen H, Fukuda MN, Fukuda M (1999) Regulation of I-branched poly-N-acetyllactosamine synthesis—concerted actions by i-extension enzyme, Ibranching enzyme, and β1,4-galactosyltransferase I. J Biol Chem 274:9296–9304 Zhou DP (2003) Why are glycoproteins modified by poly-N-acetyllactosamine glycoconjugates? Curr Prot Pept Sc 4:1–9 Elola MT, Wolfenstein-Todel C, Troncoso MF, Vasta GR, Rabinovich GA (2007) Galectins: matricellular glycan-binding

28. 29.

30.

31.

32.

33.

proteins linking cell adhesion, migration, and survival. Cell Mol Life Sci 64:1679–1700 Hughes RC (2001) Galectins as modulators of cell adhesion. Biochimie 83:667–676 Goetz H, Beginn U, Bartelink CF, Grunbauer HJM, Moeller M (2002) Preparation of isophorone diisocyanate terminated star polyethers. Macromol Mater Eng 287:223–230 Rech C, Rosencrantz RR, Křenek K, Pelantová H, Bojarová P, Römer C, Hanisch F-G, Křen V, Elling L (2011) Combinatorial one-pot synthesis of poly-N-acetyllactosamine oligosaccharides with leloir-glycosyltransferases. Adv Synth Catal 353:2492– 2500 Sauerzapfe B, Křenek K, Schmiedel J, Wakarchuk WW, Pelantová H, Křen V, Elling L (2009) Chemo-enzymatic synthesis of poly-N-acetyllactosamine (poly-LacNAc) structures and their characterization for CGL2-galectin-mediated binding of ECM glycoproteins to biomaterial surfaces. Glycoconjugate J 26:141–159 Lee SH, Ruckenstein E (1988) Adsorption of proteins onto polymeric surfaces of different hydrophilicities—a case-study with bovine serum-albumin. J Colloid Interface Sci 125:365–379 Walser PJ, Haebel PW, Kunzler M, Sargent D, Kues U, Aebi M, Ban N (2004) Structure and functional analysis of the fungal galectin CGL2. Structure 12:689–702