An S-Layer Heavy Chain Camel Antibody Fusion Protein for

Apr 20, 2004 - The S-layer protein moiety consisted of the N-terminal part which specifically recognized the SCWP as binding site and the self-assembl...
18 downloads 8 Views 157KB Size
Download PDF
664

Bioconjugate Chem. 2004, 15, 664−671

An S-Layer Heavy Chain Camel Antibody Fusion Protein for Generation of a Nanopatterned Sensing Layer To Detect the Prostate-Specific Antigen by Surface Plasmon Resonance Technology Magdalena Pleschberger,†,‡ Dirk Saerens,§ Stefan Weigert,‡ Uwe B. Sleytr,†,‡ Serge Muyldermans,§ Margit Sa´ra,†,‡ and Eva M. Egelseer*,†,‡ BMT-Biomolecular Therapeutics GmbH, Brunnerstrasse 59, A-1235 Vienna, Austria, Center for Ultrastructure Research and Ludwig Boltzmann-Institute for Molecular Nanotechnology, University of Natural Resources and Applied Life Sciences, A-1180 Vienna, Austria, and Vrije Universiteit Brussel, Vlaams Interuniversitair Instituut voor Biotechnologie, Pleinlaan 2, B-1050 Brussel, Belgium. Received February 13, 2004

The bacterial cell surface layer (S-layer) protein of Bacillus sphaericus CCM 2177 assembles into a square lattice structure and recognizes a distinct type of secondary cell wall polymer (SCWP) as the proper anchoring structure in the rigid cell wall layer. For generating a nanopatterend sensing layer with high density and well defined distance of the ligand on the outermost surface, an S-layer fusion protein incorporating the sequence of a variable domain of a heavy chain camel antibody directed against prostate-specific antigen (PSA) was constructed, produced, and recrystallized on gold chips precoated with thiolated SCWP. The S-layer protein moiety consisted of the N-terminal part which specifically recognized the SCWP as binding site and the self-assembly domain. The PSA-specific variable domain of the camel heavy chain antibody was selected by several rounds of panning from a phage display library of an immunized dromedary, and was produced by heterologous expression in Escherichia coli. For construction of the S-layer fusion protein, the 3’-end of the sequence encoding the C-terminally truncated form rSbpA31-1068 was fused via a short linker to the 5’-end of the sequence encoding cAb-PSA-N7. The S-layer fusion protein had retained the ability to self-assemble into the square lattice structure. According to the selected fusion site in the SbpA sequence, the cAb-PSA-N7 moiety remained located on the outer surface of the protein lattice. After recrystallization of the S-layer fusion protein on gold chips precoated with thiolated SCWP, the monomolecular protein lattice was exploited as sensing layer in surface plasmon resonance biochips to detect PSA.

INTRODUCTION

Crystalline bacterial cell surface layers (S-layers) can be considered one of the most common surface structures in prokaryotic cells (1-5). S-Layers of bacteria and archaea are composed of identical protein or glycoprotein subunits, which completely cover the cell surface during all stages of bacterial growth and division. The S-layer subunits can be aligned in lattices with oblique, square, or hexagonal symmetry with a center-to-center spacing of the morphological units of approximately 3 to 35 nm. In bacteria, the S-layer subunits are linked to each other and to the underlying cell envelope layer by noncovalent interactions. In the case of Bacillaceae, the N-terminal part is involved in anchoring the S-layer subunits to the rigid cell wall layer via a distinct type of secondary cell wall polymer (SCWP) (5-10). The polymer chains are covalently linked to the peptidoglycan backbone. Because many S-layer subunits self-assemble in suspension or recrystallize into regularly structured monolayers on * To whom correspondence should be addressed: Center for Ultrastructure Research, University of Natural Resources and Applied Life Sciences, Gregor Mendelstrasse 33, A-1180 Vienna, Austria. Tel: +43-1-47 654 2208; Fax: +43-1-47 89 112; e-mail: [email protected]. † BMT-Biomolecular Therapeutics GmbH. ‡ University of Natural Resources and Applied Life Sciences. § Vrije Universiteit Brussel.

solid supports, such as noble metals, plastics, or silicon wafers, as well as on Langmuir lipid films, on liposomes, and at the air-water interphase, they can be considered as unique biomaterials with properties most relevant for applications in molecular nanotechnology, nanobiotechnology, and biomimetics (1-3). The S-layer protein of Bacillus sphaericus CCM 2177 assembles into a square lattice with a center-to-center spacing of the morphological units of 13.1 nm. On SDSgels, this S-layer protein migrates as a single band with an apparent relative molecular mass of 127 000. The gene encoding SbpA has been sequenced, cloned in plasmid pET28a, and expressed in Escherichia coli HMS174(DE3) (11). The protein precursor includes a 30 amino acid long typical Gram-positive signal peptide and consists of total of 1268 amino acids. Since the N-terminal part of SbpA carries an S-layer homologous (SLH)-domain which is required for binding of the S-layer subunits to the SCWP, the surface accessibility of selected amino acid positions in the C-terminal part was investigated after fusion of the short affinity peptide Strep-tag I (STI) for binding of streptavidin. In comparison to the entire SbpA (rSbpA31-1268), the C-terminally truncated form rSbpA31-1068 showed a significantly increased accessibility of the C-terminal end. Despite the deletion of 200 C-terminal amino acids, rSbpA31-1068/STI was still capable of selfassembling into the square lattice structure. Since the fused Strep-tag I remained located on the outer surface

10.1021/bc049964w CCC: $27.50 © 2004 American Chemical Society Published on Web 04/20/2004

Functional S-Layer Fusion Proteins

of the protein lattice, rSbpA31-1068 was used as base form for the construction of various S-layer fusion proteins. The S-layer protein SbpA recognizes a distinct type of SCWP consisting of eight to nine disaccharide repeating units as the proper anchoring structure in the rigid cell wall layer. As described in a previous study, the SCWP can be exploited for an oriented binding of rSbpA-fusion proteins on solid supports to generate regularly structured functional protein lattices (12). The Camelidae is the only taxonomic family known to possess functional heavy chain antibodies lacking light chains. These unique antibody isotypes interact with the antigen by virtue of a single variable domain, termed VHH. The variable domain of a camel heavy chain antibody is the smallest known complete antigen binding fragment from a functional immunoglobulin. The single domain nature of VHHs gives rise to several unique features as compared to antigen-binding derivatives of conventional antibodies. Besides the advantage of single gene cloning and selection from an in vivo matured library, recombinant VHHs have other technological, functional, and physicochemical advantages, such as (i) high expression yields and ease of purification, (ii) a highly soluble and stable single domain immunoglobulin fold, (iii) the generation of antigen-specific, high-affinity binders, (iv) the recognition of unique conformational epitopes with the dominant involvement of its enlarged variable region CDR3, and (v) the close homology to human VH fragments (13, 14). These features can be expected to lead to a number of applications in which VHHs perform better than conventional antibody fragments, e.g., as enzyme inhibitors, as modular building units for multivalent or multifunctional constructs or as immunoadsorbents and detection units in biosensors (13, 14). The prostate-specific antigen (PSA) is a useful marker to screen potential prostate cancer patients. The current diagnostic test systems determine the concentration of total PSA with monoclonal antibodies that recognize free as well as PSA complexed with R-1-anti-chymotrypsin (ACT). For the employment of a VHH in a PSA biosensor, VHHs that recognize free and complexed PSA are required. Moreover, the kinetic requirements in the biosensor impose a high probe density that can probably only be obtained with single domain VHHs. For generating a monomolecular protein lattice capable of binding PSA on the outermost surface in high density, an S-layer fusion protein incorporating the VHH domain of a heavy chain camel antibody directed against PSA was constructed and the sequence encoding the fusion protein heterologously expressed in E. coli. After recrystallization of the S-layer fusion protein on gold chips precoated with thiolated SCWP, the monomolecular protein lattice was exploited as sensing layer in surface plasmon resonance (SPR) biochips. MATERIALS AND METHODS

Immunization of a Dromedary for Generating PSA Binding Antibodies and Fractionation of IgG Subclasses. One male dromedary (Camelus dromedarius) kept at the Central Veterinary Research Laboratories (Dubai, United Arabic Emirates) was injected seven times subcutaneously at weekly intervals with 500 µg of highly purified PSA, mixed in Gerbu adjuvans. Fortyfive days after beginning of immunization, 100 mL of anticoagulated blood was collected, from which plasma and peripheral blood lymphocytes were isolated with a lymphocyte preparation kit from WAK (WAK, Chemie, Steinbach, Germany).

Bioconjugate Chem., Vol. 15, No. 3, 2004 665

Separation of the different plasma immunoglobulin G (IgG) subclasses was performed by differential adsorption on Hitrap-proteinA and Hitrap-proteinG columns (Pharmacia). The IgG3 and IgG1 subclasses were eluted from the proteinG column with sodium acetate buffer (pH 3.5) and glycine-HCl buffer (pH 2.7), respectively. The flowthrough was subsequently loaded onto the Hitrap-proteinA column to recover the IgG2 subclass during elution with acetate buffer (pH 3.5). Construction of a Library and Selection of PSA Specific Antibody Fragments. The mRNA was extracted from the peripheral blood lymphocytes (15). In a subsequent step, the cDNA was prepared using Readyto-Go beads (Pharmacia). The gene fragments encoding the variable domain until the CH2 domain were amplified with the specific primers CALL001 (5′-GTC CTG GCT CTC TTC TAC AAG G-3′) and CALL002 (5′-GGT ACG TGC TGT TGA ACT GTT CC-3′), which anneal at the leader sequence and at the CH2 exon of the heavy chains of all dromedary immunoglobulins, respectively. After reamplification of the VHH gene fragments with nested primers annealing at the framework1 and framework4 regions (15), the final PCR fragments were cloned into the phagemid vector pHEN4, using the restriction sites PstI and NotI. The VHH library was expressed on phages after infection with M13K07 helper phages. Specific VHHs against PSA were enriched by three consecutive rounds of in vitro selection on biotinylated PSA captured on an ELISA plate coated with 10 µg of Neutravidin (PIERCE) per well. Bound phage particles were eluted with 100 mM triethylamine (TEA, pH 10.0). Solutions containing eluted particles were immediately neutralized with 1 M Tris-HCl buffer (pH 7.4) and used to infect exponentially growing E. coli TG1 cells. The enrichment of phage particles carrying the antigenspecific VHHs was assessed by comparing the number of eluted phages from antigen-coated versus non-antigencoated wells. After the second and third round of panning, individual colonies were picked up and expression of genes encoding VHHs was induced by the addition of 1 mM isopropyl-β-D-thiogalactopyranoside (IPTG). The recombinant VHH extract from the periplasm was tested for antigen recognition in ELISA (15). Expression and Purification of Antibody Fragments. The VHH genes of the clones that scored positive in ELISA were recloned into the expression vector pHEN6, using the restriction enzymes PstI and BstEII. The plasmid constructs were transformed into E. coli WK6(su-) cells. Production of recombinant VHHs was performed in shaking flasks by growing the bacterial culture in Terrific Broth supplemented with 0.1% glucose and ampicillin. When the optical density measured at 600 nm was between 0.6 and 0.9, expression was induced by the addition of 1 mM IPTG. Cultures were then incubated for 16 h at 28 °C. After pelleting the cells, the periplasmic proteins were extracted by osmotic shock and the extract was loaded onto a Ni-nitrilotriacetic acid (Ni-NTA) Superflow Sepharose column (Qiagen). After being washed, the bound proteins were eluted with sodium acetate buffer (pH 4.7). The eluted fraction was concentrated on Vivaspin concentrators with a molecular mass cut off of 5 kDa (Vivascience), and the purity was checked by SDSPAGE. Solid-Phase ELISA for the Detection of PSASpecific Phages. Maxisorb 96-well plates (Nunc, Maxisorb) were coated with Neutravidin or the PSA-specific monoclonal antibody 3E6 (DIAMED) at a concentration of 5 µg/mL in phosphate-buffered saline (PBS) overnight at 4 °C. Residual sites were blocked for 2 h at room

666 Bioconjugate Chem., Vol. 15, No. 3, 2004

temperature with 1% (wt/vol) casein in PBS. After incubation with biotinylated PSA (1 µg/mL), serial dilutions of purified IgG subclasses, phages from different rounds of panning, or soluble VHHs were added. Detection was performed with a rabbit anti-dromedary IgG antiserum, a horseradish peroxidase-anti-M13 conjugate (Pharmacia), a mouse anti-hemagglutinin decapeptide tag (BAbCO), or an anti-His tag antiserum (Serotec), respectively. The absorption at 405 nm was measured 15 min after the addition of the substrate p-nitrophenyl phosphate (p-NPP, Sigma). Further Characterization of cAb-PSA-N7 by Epitope Mapping by Means of ELISA and Affinity Measurements by Surface Plasmon Resonance (SPR). Either Neutravidin (5 µg/mL) or the PSA-specific monoclonal antibody 3E6 (5 µg/mL) were coated on ELISA plates, followed by the addition of 1 µg/mL biotinylated or nonbiotinylated PSA or PSA-ACT complex. Bound VHHs were detected with an anti-His tag monoclonal antibody or Nickel Express Doctor (Killguard Laboratories), respectively. Affinity measurements were assessed by the addition of different concentrations of PSA to purified His-tailed VHHs coated on a Ni-NTA-biochip (Biacore) according to the manufacturer’s descriptions. The kinetic parameters and KD were determined with the BIAevaluation software (Biacore). Cloning of the Chimeric Gene Encoding the Fusion Protein rSbpA31-1068/cAb-PSA-N7. All PCR reactions were performed as described by Jarosch et al. (16). The sequence encoding the C-terminally truncated form rSbpA31-1068 was PCR amplified from chromosomal DNA of B. sphaericus CCM 2177 by using the oligonucleotide primers sbpA/NcoI [5’-CG GAT TCC ATG GCG CAA GTA AAC GAC TAT AAC AAA ATC-3’] and sbpA/ NheI [5’-GAC CGC GCT AGC TTC TGA ATA TGC AGT AGT TGC TGC C-3’], which introduced the restriction sites (bold) NcoI or NheI. The cAb-PSA-N7 sequence was amplified from plasmid cAb-PSA-N7-pHEN6 by using the oligonucleotide primers cAb/NheI [5’-CGG ATT GCT AGC GAT GTG CAG CTG CAG GAG-3’] and cAb/XhoI [5’-GAC CGC TCG AGT TAT GAG GAG ACG GTG ACC TGG G-3’], which introduced the restriction sites NheI or XhoI. To obtain plasmid pET28a-sbpA(93-3204), the gelpurified PCR product sbpA(93-3204) was ligated into the corresponding restriction sites of plasmid pET28a, which was established in E. coli TG1. For generating the chimeric sbpA(93-3204)/cAb-PSA-N7 gene, the gel-purified PCR product cAb-PSA-N7, encoding the VHH domain of the camel antibody directed against PSA, was ligated into the corresponding restriction sites of plasmid pET28asbpA(93-3204). This was used for transformation of E. coli TG1. Expression of the Chimeric sbpA(93-3204)/cAb-PSAN7 Gene and Isolation of the S-Layer Fusion Protein from the Host Cells. The plasmid stability test and heterologous expression of the sbpA(93-3204)/cAb-PSAN7 gene in E. coli One Shot BL21 Star (DE3) (Invitrogen) were performed as described by Jarosch et al. (16). Samples of the E. coli One Shot BL21 Star cultures were taken 1 to 5 h after induction of sbpA(93-3204)/cAb-PSAN7 gene expression by the addition of 1 mM IPTG (Gerbu). Preparation of samples for SDS-PAGE analysis was carried out as described by Laemmli (17). For electron microscopic investigation, whole cells of E. coli One Shot BL21 Star (DE3) were prepared as described (18). Isolation of the S-layer fusion protein from E. coli One Shot BL21 Star (DE3) cells induced to express the

Pleschberger et al.

sbpA(93-3204)/cAb-PSA-N7 gene and purification by gel permeation chromatography (GPC) was performed as described (19) except that DNA was degraded by DNAse treatment. For that purpose, the insoluble pellet obtained after sonification of whole cells and centrifugation of the suspension was treated with a solution of 1 mg/mL DNAseI (Roche) in 100 mM MgCl2‚7H2O in MilliQ-water. For removal of residual DNAseI, the pellet was washed twice with 1% Triton X-100 in 50 mM Tris-HCl buffer (pH 7.2) and 50 mM Tris-HCl buffer (pH 7.2). After the pellet was resuspended in 4 mL of 5 M guanidine hydrochloride (GHCl) in 50 mM Tris-HCl buffer (pH 7.2) and centrifugation, the clear supernatant was subjected to GPC. Fractions eluting at a molecular mass of ∼123 000 Da expected for rSbpA31-1068/cAb-PSA-N7 were pooled, dialyzed against distilled water for 18 h at 4 °C, lyophilized, and stored at -20 °C. Immunoblotting using a polyclonal rabbit antiserum raised against the S-layer protein of B. sphaericus CCM 2177 and a polyclonal rabbit anti-camel antiserum (Sigma C7540) was performed as described (19, 20). Investigation of the Self-Assembly Properties of the S-Layer Fusion Protein and Recrystallization on Peptidoglycan-Containing Sacculi. The production of self-assembly products as well as recrystallization of the S-layer fusion protein on peptidoglycan-containing sacculi was performed as described (19). Peptidoglycancontaining sacculi carrying a monolayer of recrystallized S-layer fusion protein are referred to as recrystallization products in all further experiments. To keep the S-layer fusion protein water-soluble, the GHCl-extract was dialyzed against 1 mM 1,4-dithio-DL-threitol (DTT) in MillQwater. Dot Blot Assays for Investigating the Accessibility of the Fused cAb-PSA-N7. To investigate the accessibility of the fused cAb-PSA-N7 in the watersoluble, self-assembled, or recrystallized S-layer fusion protein, 5 µL samples with an S-layer protein content of 1 mg/mL were dried onto a nitrocellulose membrane. Plain rSbpA (1 mg/mL) was taken as blank and patient serum containing PSA (diluted 1:20 in TBS) was used as positive control. Dot blot assays were carried out as described in a previous study (19). For binding of PSA, the membrane was incubated with 20 ng of PSA from patient serum in 20 mL of blocking solution (2% Top Block (Fluka) in TBS) for 1 h at 20 °C. To detect PSA, the monoclonal mouse antibody 3E6 raised against PSA (1.38 mg/mL; diluted 1:2000 in blocking solution) was used. Binding of cAb-PSA-N7 and the S-Layer Fusion Protein rSbpA31-1068/cAb-PSA-N7 to a CM5 SPR Sensor Chip with Immobilized PSA. SPR experiments were performed using a Biacore 2000 system (Biacore, Sweden). For covalent binding of PSA on a CM5 sensor chip (Biacore, Sweden), the carboxylic groups on the CM5 matrix were activated by injecting a mixture containing 0.1 M N-hydroxysuccinimide (NHS, Fluka) and 0.4 M 1-ethyl-3-[3-(dimethylamino)propyl]carbodiimide hydrochloride (EDC, Sigma) for 7 min at a flow rate of 5 µL/min. Subsequently, a PSA (Scipac, UK) solution was passed over the surface with an appropriate reaction time, followed by a 7 min injection of 1 M ethanolamine-HCl solution (pH 8.5) to block nonreacted activated groups. The amount of covalently bound PSA was expressed in terms of resonance units (RU; 1000 RU correspond to approximately 1 ng/mm2). For investigation of the affinity to immobilized PSA, cAb-PSA-N7 and the S-layer fusion protein, 1 µg/mL and 100 µg/mL, respectively, in HBS buffer (10 mM Hepes, 150 mM NaCl, 3

Functional S-Layer Fusion Proteins

mM EDTA, 0.005% Tween) pH 7.2, were conducted over the chip with a flow rate of 2 µL/min. Afterward, four steps of regeneration with 10 mM glycine-HCl buffer (pH 2.2) were performed to ensure the removal of noncovalently bound proteins. To generate reference surfaces, all experiments were performed as described above with the exception that no PSA was supplied. Recrystallization of rSbpA31-1068/cAb-PSA-N7 on Gold Chips Precoated with Thiolated SCWP and Evaluation by Atomic Force Microscopy (AFM). Coating of gold chips (silicon 〈100〉 , coated with 1 nm Ti and 60 nm Au) with an area of 1 cm2 with thiolated SCWP and recrystallization of the S-layer fusion protein was done as described (19). A Nanoscope III AFM (Digital Instruments, Santa Barbara, CA) equipped with a liquid cell and a 120 µm scanner and oxide-sharpened Si3N4 tips (NanoProbes, Digital Instruments, Santa Barbara, CA) with a nominal spring constant of 0.06 N/m was used for all AFM investigations. Imaging of recrystallized rSbpA31-1068/cAb-PSA-N7 was carried out in contact mode in liquid (100 mM NaCl). Height and deflection signals were recorded simultaneously (error signal mode). The applied force was kept at a minimum in order not to destroy the sample. The scan rate was 5.1 Hz at a scan size of 1 by 1 µm. Investigation of the PSA Binding Capacity of the rSbpA31-1068/cAb-PSA-N7 Monolayer Obtained by Recrystallization of this S-Layer Fusion Protein on SCWP-Coated Gold Chips by SPR. The S-layer fusion protein was recrystallized on SCWP-coated gold chips (Biacore, Sweden) at a concentration of 100 µg/mL for 25 min at a flow rate of 1 µL/min. For reference studies, rSbpA31-1068 was used. Patients’ sera containing PSA (diluted 1:5 in HBS buffer to obtain a final concentration of 20 ng/mL) were injected. To verify specific binding of PSA, a solution of female blood serum (200 µL) was injected. During the reaction time, a continuous flow of HBS buffer (10 µL/min) was maintained over the sensor surface. The contact and dissociation time for PSA was 40 min and 800 s, respectively. RESULTS

Immunization of a Dromedary with PSA for Provoking a Humoral Response. A dromedary received seven injections with purified PSA. Forty-five days after the first injection, blood was drawn and the immunoglobulins were purified from its plasma by differential adsorption on proteinA and proteinG columns. The immune response in subclasses IgG1, IgG2, and IgG3 was assessed by ELISA. Starting from 5 µg/mL, serial dilutions of purified IgG1, IgG2, and IgG3 were used in a solid-phase ELISA with biotinylated PSA captured onto a Neutravidin coat. Bound IgG was subsequently detected with rabbit anti-dromedary IgG antiserum (Figure 1). An antigen-specific response was observed for heavy chain antibodies (IgG2 and IgG3) and conventional antibodies (IgG1) (21). Construction of a Library and Selection of PSA Specific Antibody Fragments. A library was constructed starting from 1.5 × 107 lymphocytes. After extraction of mRNA, cDNA was prepared and the genes coding for the variable domains of the immunoglobulins were amplified. The final VHH gene product was ligated into the phagemid vector pHEN4. A VHH library of 1.5 × 107 individual transformants was obtained when 3 µg of vector was transformed. As determined by PCR, 90% of the clones within the library carried a VHH gene insert

Bioconjugate Chem., Vol. 15, No. 3, 2004 667

Figure 1. Analysis of antigen-specific antibodies. Solid-phase ELISA for detection of serial dilutions of fractionated IgG against PSA. Biotinylated PSA (1 µg/mL) was captured onto 5 µg/mL Neutravidin coating. Bound IgG1 (9), IgG2 (b), and IgG3 (2) was detected with rabbit anti-dromedary IgG antiserum.

Figure 2. Enrichment of PSA-specific virons by consecutive rounds of in vitro selection. Solid-phase ELISA for detection of antigen-specific phages from different rounds of panning. Biotinylated PSA (1 µg/mL) was captured on 5 µg/mL Neutravidin coating. Virons (109) of the library (1), after the first (2), after the second (3) and the third round (4) of panning were loaded onto the antigen. Bound phages were detected with anti-M13 horse-radish peroxidase conjugate.

of proper size. The VHH repertoire from the library was expressed on phages after infection with helper phages M13K07. Three rounds of in vitro selection or panning were performed on biotinylated PSA captured by a Neutravidin coat. After being extensively washed, bound phages were eluted with TEA. The enrichment factors were calculated in two different ways. First, it was calculated as the ratio of the number of PSA specific colonies to the number of nonspecific colonies. An enrichment was observed in the second round, with the maximum enrichment after the last round of selection. Second, the enrichment was also assessed in solid-phase phage ELISA adding 109 virons per well (Figure 2). From the corrected signals (ODPSA - ODBlank), the best ratio (specific to nonspecific) was found after the second round of panning. In total, 25 different VHHs binding specifically to PSA were isolated. For the present study, the specific binder cAb-PSA-N7 was selected. Further Characterization of cAb-PSA-N7 by Epitope Mapping and Affinity Measurements. The epitope region recognized by cAb-PSA-N7 was mapped in ELISA. Binding of the VHH was tested on different forms of PSA, such as free PSA and PSA complexed with ACT. Both forms were captured onto either the 3E6

668 Bioconjugate Chem., Vol. 15, No. 3, 2004

Pleschberger et al.

Table 1. Different Characteristics of the PSA-Specific Single Domain Antibody, cAb-PSA-N7a binding to binding to binding to free PSA PSA-ACT biotinylated PSA KD (nM) cAb-PSA-N7

+

+

+

0.09

a

Recognition of free PSA, PSA complexed with ACT (PSAACT), and biotinylated PSA by cAb-PSA-N7 was assessed by solidphase ELISA. The dissociation constant (KD) was calculated from the kinetic constants measurements by SPR.

Figure 3. SDS-PAGE patterns (lanes 1-3) of SDS-extracts of whole cells of E. coli One Shot BL21 Star cells harboring the vector pET28a containing the sbpA(93-3204)/cAb-PSA-N7 gene construct, before (lane 1) and 5 h (lane 2) after induction of gene expression. Lane 3: SDS-extracts of GPC purified rSbpA31-1068/ cAb-PSA-N7. Immunoblot analysis (lanes 4 and 5) of SDSextracts of GPC purified rSbpA31-1068/cAb-PSA-N7. The presence of the SbpA-portion was detected with anti-SbpA antiserum and anti-rabbit IgG-alkaline phosphatase conjugate (lane 4). The presence of the cAb-PSA-N7 moiety was checked with antidromedary IgG antiserum and anti-rabbit IgG-alkaline phosphatase conjugate (lane 5).

monoclonal antibody or after biotinylation onto Neutravidin. Since binding of cAb-PSA-N7 was observed on PSA captured by 3E6 as well as on the PSA-ACT complex (Table 1), the results indicated binding onto a region of PSA different from its interaction site with ACT and from the epitope region recognized by 3E6. To further characterize its binding properties, the kinetic constants of cAb-PSA-N7 for PSA were measured by SPR on Biacore 3000 and the dissociation constant (KD) was calculated accordingly (Table 1). Because of its high affinity, the specific binder cAb-PSA-N7 was selected out of 25 different binders (Saerens et al., unpublished data). Cloning and Expression of the Gene Encoding rSbpA31-1068/cAb-PSA-N7 and Isolation of the SLayer Fusion Protein from the Host Cells. The PCR product obtained by amplification of the sbpA gene using the primers sbpA/NcoI and sbpA/NheI which encoded the truncated form of the S-layer protein SbpA (rSbpA31-1068) was ligated into the pET28a vector. After transformation of E. coli TG1, amplification and isolation of the plasmid pET28a-sbpA(93-3204), the gene encoding cAb-PSA-N7 was ligated into this plasmid via the corresponding restriction sites. The resulting vector pET28a-sbpA(93-3204)/cAb-PSAN7 was established in E. coli One Shot BL21 Star and expression was induced by the addition of IPTG. Biomass samples were harvested at various points of time and subjected to SDS-PAGE analysis and ultrathin sectioning. In comparison to E. coli cells harvested before the addition of IPTG (Figure 3; lane 1), an additional high molecular mass protein band with an apparent molecular mass of 123 000 Da was observed on SDS-gels in samples from E. coli cultures induced to express the chimeric sbpA(93-3204)/cAb-PSA-N7 gene (Figure 3; lane 2). As derived from SDS-PAGE, the S-layer fusion protein had accumulated in the insoluble fraction of the lysed E.

coli One Shot BL21 Star cells. This was in agreement with data from ultrathin-sectioned preparations of whole cells, which revealed the presence of inclusion bodies in the cytoplasm of the host cells (not shown). The insoluble fraction of the lysed E. coli cells was extracted with 5 M GHCl and subjected to GPC. After unifying the fractions from the second peak and removing GHCl by dialysis, only a single major protein band with an apparent molecular mass of 123 000 Da was observed on SDS-gels (Figure 3; lane 3). On immunoblots, a strong cross reaction was observed between the high molecular mass protein band and the polyclonal rabbit antiserum raised against the S-layer protein SbpA of B. sphaericus CCM 2177, as well as with the polyclonal rabbit antiserum raised against dromedary IgG (Figure 3; lanes 4 and 5, respectively). Investigation of the Self-Assembly and Recrystallization Properties of the S-Layer Fusion Protein and of the Accessibility of the Fused cAb-PSAN7 Moiety by Dot Blot Immunoassay. To investigate the self-assembly properties, purified and lyophilized rSbpA31-1068/cAb-PSA-N7 was denatured in 5 M GHCl, and the solution was dialyzed against 10 mM CaCl2 in 1 mM DTT. Negative-staining confirmed that the S-layer fusion protein had self-assembled into flat double layer sheets, which clearly exhibited the square lattice structure (not shown). The S-layer fusion protein recognized peptidoglycan-containing sacculi of B. sphaericus CCM 2177 as binding site and recrystallized into the square lattice type (Figure 4 A). Dot blot assays were performed in order to investigate the accessibility of the fused cAb-PSA-N7 moiety in the water-soluble, self-assembled, and recrystallized S-layer fusion protein. Patient’s serum containing PSA used as positive control (Figure 5; spot 1) led to a strong reaction with a monoclonal mouse antibody raised against PSA, whereas no reaction was observed with plain rSbpA taken as negative control (Figure 5; spot 2). As shown in Figure 5 (spots 3-5), the reaction toward PSA was independent of the state of the S-layer fusion protein and comparable for all samples. Assuming an identical functionality of cAb-PSA-N7 in all S-layer fusion protein samples, these findings indicated that the accessibility of the fused functional sequence for the antigen was not reduced in the selfassembled or recrystallized state of the S-layer fusion protein. Binding of cAb-PSA-N7 and the S-Layer Fusion Protein rSbpA31-1068/cAb-PSA-N7 to a CM5 SPR Sensor Chip with Immobilized PSA. To compare the affinity of cAb-PSA-N7 and the chimeric S-layer fusion protein to PSA by SPR, PSA was covalently bound to the activated carboxylic acid groups of a CM5 sensor chip (Figure 6). This lead to an increase in 916 RU which corresponded to 3.05 × 10-5 nM/mm2. By passing cAbPSA-N7 over the sensor surface, an increase in 197 RU was measured which corresponded to 1.29 × 10-5 nM/ mm2, or to a molar ratio between bound cAb-PSA-N7 and PSA of 0.43. In the case of rSbpA31-1068/cAb-PSA-N7, an increase in 570 RU was measured. This value corresponded to 4.6 × 10-6 nM/mm2, or to a molar ratio between bound S-layer fusion protein and immobilized PSA of 0.16. The lower molar ratio as determined for the binding of the S-layer fusion protein to immobilized PSA can be explained by the 8-fold higher molecular mass of rSbpA31-1068/cAb-PSA-N7 in comparison to cAb-PSA-N7. Investigation of the PSA Binding Capacity of the rSbpA31-1068/cAb-PSA-N7 Monolayer Obtained by Recrystallization of This S-Layer Fusion Protein

Functional S-Layer Fusion Proteins

Bioconjugate Chem., Vol. 15, No. 3, 2004 669

Figure 4. (A) Electron micrograph of a negatively stained preparation of recrystallization products obtained with rSbpA31-1068/ cAb-PSA-N7 exhibiting the square S-layer lattice type. (B) AFM image of the S-layer fusion protein rSbpA31-1068/cAb-PSA-N7 recrystallized on a gold chip precoated with thiolated SCWP.

Figure 5. Dot blot assay indicating the accessibility of the cAbPSA-N7 moiety in the water-soluble, self-assembled, and recrystallized S-layer fusion protein. Spot 1: patient serum containing PSA (diluted 1:20 in TBS) used as positive control; spot 2: rSbpA used as negative control; spot 3: water-soluble rSbpA31-1068/cAb-PSA-N7; spot 4: self-assembly products formed by rSbpA31-1068/cAb-PSA-N7; spot 5: recrystallization products obtained with rSbpA31-1068/cAb-PSA-N7.

on SCWP-Coated Gold Chips by SPR. As shown by AFM analysis, the S-layer fusion protein was capable of recrystallizing into the square lattice structure on gold chips precoated with thiolated SCWP (Figure 4 B). For generating a sensing layer with well defined arrangement of the ligand on the outer surface of a monomolecular protein lattice, the S-layer fusion protein was conducted over an SPR gold chip precoated with thiolated SCWP. Recrystallization of the S-layer fusion protein led to an increase in the SPR signal to 5770 RU, which corresponded to 4.6 × 10-5 nM/mm2, and therefore to a monolayer of the S-layer fusion protein. When a solution of PSA containing patient serum diluted 1:5 in HBS buffer was conducted over the rSbpA31-1068/cAb-PSA-N7 monolayer, specific binding of the antigen was observed. For female blood serum, no specific binding was observed (data not shown). In Figure 7, the sensorgram of a rSbpA31-1068 monolayer used as the reference was subtracted from that obtained with the S-layer fusion protein. By this subtraction, a value of 1074 RU could be calculated for bound PSA, which corresponded to 3.6 × 10-5 nM/mm2. When this value was correlated to the value of 4.6 × 10-5 nM/mm2 calculated for the rSbpA31-1068/ cAb-PSA-N7 monolayer, a molar ratio of 0.78 PSA molecules per S-layer fusion protein was obtained. These measurements reveal that at least three PSA molecules were bound per morphological unit of the square S-layer lattice consisting of four S-layer fusion protein subunits. Since PSA showed no affinity to rSbpA31-1068 at all, it was confirmed that binding of PSA to the S-layer fusion protein was due to specific interactions with the cAb-PSAN7 moiety. DISCUSSION

Studies on the structure-function-relationship of the S-layer protein SbpA revealed that the N-terminal part

is required for anchoring the S-layer subunits via a distinct type of SCWP to the rigid cell wall layer and that 200 C-terminal amino acids could be deleted without interfering with the self-assembly properties (11, 12). Since the SCWP could be exploited to achieve an oriented binding of the S-layer subunits via their N-terminal region to solid supports, amino acid positions in the C-terminal part of the SbpA sequence were screened in order to identify potential fusion sites that were located on the outer surface of the protein lattice. For that purpose, the sequence encoding Strep-Tag I was fused to the 5’-end of the sequence encoding various C-terminally truncated rSbpA forms. Since after recrystallization of rSbpA31-1068/STI on peptidoglycan-containing sacculi the fused Strep-Tag I was accessible to streptavidin binding on the outer surface of the protein lattice to a significantly higher extent than in the case of the entire rSbpA (11), the sequence encoding rSbpA31-1068 was exploited for the construction of various S-layer fusion proteins. Such fusion proteins comprised either the sequence of the major birch pollen allergen (11), the Fcbinding ZZ-domains (22), or the VHH of a heavy chain camel antibody directed against lysozyme (19). Depending on the fused functional sequence, the S-layer fusion proteins were either used for the development of an antiallergic vaccine (23, 24), or as immunoadsorbent to remove autoantibodies from patients’ sera suffering from autoimmune disease (22). Due to the stability properties of heavy chain camel antibodies, appropriate S-layer fusion proteins were constructed for the development of novel diagnostic test systems (19). The S-layer fusion protein carrying the heavy chain camel antibody directed against lysozyme (rSbpA31-1068/cAb-Lys3) was recrystallized on ELISA plates to provide a surface with a welldefined number and distance of the ligand. According to the construction principle of this S-layer fusion protein, each subunit carried one cAb-Lys3 moiety. The monolayer generated by recrystallization of the S-layer fusion protein on ELISA plates showed the same performance as the rSbpA monolayer to which cAb-Lys3 had been immobilized by chemical coupling. In this case, eight to nine cAb-Lys3 molecules were available per morphological unit of the square S-layer lattice, which was twice as much as in the case of the monolayer consisting of the S-layer fusion protein (19). Although several studies indicated that the S-layer fusion protein had attached with the inner surface to the ELISA plates, this orientation was guaranteed when solid supports carrying the SCWP were used for recrystallization. Besides peptidoglycan-containing sacculi of B. sphaericus CCM 2177, gold chips precoated with thiolated SCWP fulfilled such

670 Bioconjugate Chem., Vol. 15, No. 3, 2004

Pleschberger et al.

Figure 6. Sensorgrams obtained by binding of cAb-PSA-N7 (1) and the chimeric S-layer protein rSbpA31-1068/cAb-PSA-N7 (2) to PSA covalently bound to a CM5 sensor chip. The sensorgrams were obtained by subtracting the respective blank sensorgram from flow cell 1.

Figure 7. Sensorgram showing association (s) and dissociation (- - -) between PSA and rSbpA31-1068/cAb-PSA-N7 recrystallized on a gold chip precoated with thiolated SCWP. The sensorgram indicated specific binding of PSA to the S-layer fusion protein.

requirements. The latter were exploited for the recrystallization of the S-layer fusion protein rSbpA31-1068/cAbPSA-N7 which comprised the sequence of a heavy chain camel antibody that specifically recognized PSA. For isolation of genes encoding appropriate heavy chain camel antibodies, a dromedary was immunized with purified PSA. This proved to be successful since all IgG subclasses reacted positive on PSA. The genes encoding VHHs were amplified from the purified lymphocytes and cloned into a phage vector resulting in a VHH immune library. After panning on biotinylated PSA, we identified more than 20 different binders, characterized by ELISA, and the kinetic parameters were investigated by SPR. The VHH cAb-PSA-N7 with the highest affinity to PSA (KD 0.09 nM) was chosen for further analysis to enable highly sensitive detection of the analyte. The S-layer fusion protein rSbpA31-1068/cAb-PSA-N7 formed self-assembly products and recrystallized on peptidoglycan-containing sacculi of B. sphaericus CCM 2177. In both cases, the S-layer lattice exhibited the square structure with a center-to-center spacing of the morphological units of 13.1 nm. The accessibility of the fused cAb-PSA-N7 moiety for PSA was proved for the water-soluble, self-assembled, and recrystallized S-layer fusion protein, which independent of its state led to comparably strong reactions in dot blot immunoassays. To generate a PSA-specific sensing layer for SPR measurements, the S-layer fusion protein rSbpA31-1068/ cAb-PSA-N7 was recrystallized on gold chips precoated with thiolated SCWP. The formation of the monomolecular protein lattice was confirmed by AFM analysis, as

well as by the level of the measured SPR signal. As derived from the response levels measured for binding of PSA to the rSbpA31-1068 /cAb-PSA-N7 monolayer, the molar ratio between bound PSA and the S-layer fusion protein was 0.78, which means that at least three PSA molecules were bound per morphological unit of the square S-layer lattice. Comparative studies with an rSbpA monolayer confirmed that the measured response levels were specific for PSA binding. To summarize, by using the SCWP as biomimetic linker to gold chips, a sensing layer for SPR could be generated by recrystallization of an S-layer fusion protein. Due to the crystalline structure of the S-layer lattice, the fused ligand sequences show a well defined distance in the protein lattice, and according to the selected fusion site in the S-layer protein, they are located on the outermost surface, which will reduce diffusion limited reactions. A further advantage can be seen in the constant and low distance of the ligands from the optically active gold layer, which is exclusively determined by the thickness of the S-layer and lies in the range of only 10 to 15 nm. Thus, S-layer fusion proteins can be considered as key element for the development of label free detection systems such as SPR, surface acoustic wave (SAW), or quartz crystal microbalance (QCM-D), in which the binding event can be measured directly by a mass increase without the need of any labeled molecule. ACKNOWLEDGMENT

This work was supported by the Austrian Science Foundation, project P14689-MOB, by the Competence Center “Biomolecular Therapeutics” and by the EU project PAMELA. We thank Jacqueline Friedmann for excellent technical assistance concerning the AFM studies and Prof. Dr. Georg Steiner from the AKH for providing us with PSA-containing serum samples from patients. This work was supported by a predoctoral grant from the Institute for the promotion of Innovation by Science and Technology in Flanders (IWT-Flanders) to Dirk Saerens. LITERATURE CITED (1) Sleytr, U. B., Egelseer, E. M., Pum, D., and Schuster, B. (2004) S-Layers. In NanoBiotechnology (Niemeyer, C. M., and Mirkin, C. A., Eds.) pp 77-92, Wiley-VCH, Weinheim. (2) Sleytr, U. B., Sa´ra, M., Pum, D., Schuster, B., Messner, P., and Scha¨ffer, C. (2003) Self-Assembly Protein Systems: Microbial S-Layers. In Biopolymers (Steinbu¨chel, A., and Fahnestock, S., Eds.) Vol. 7, pp 285-338, Wiley-VCH, Weinheim, Germany.

Functional S-Layer Fusion Proteins (3) Sleytr, U. B., Messner, P., Pum, D., and Sa´ra, M. (1999) Crystalline bacterial cell surface layers (S-layers): from supramolecular cell structure to biomimetics and nanotechnology. Angew. Chem., Int. Ed. 38, 1034-1054. (4) Sleytr, U. B., Sara, M., and Pum, D. (2000) Crystalline bacterial cell surface layers (S-layers): a versatile selfassembly system. In Supramolecular Polymers (Ciferri, A., Ed.) pp 177-213, Marcel Dekker, Inc., New York, Basel. (5) Sa´ra, M., and Sleytr, U. B. (2000) S-Layer proteins. J. Bacteriol. 182, 859-868. (6) Brechtel, E., and Bahl, H. (1999) In Thermoanaerobacterium thermosulfurigenes EM1 S-layer homology domains do not attach to peptidoglycan. J. Bacteriol. 181, 5017-5023. (7) Chauvaux, S., Matuschek, M., and Beguin, P. (1999) Distinct affinity of binding sites for S-layer homologous domains in Clostridium thermocellum and Bacillus anthracis cell envelopes. J. Bacteriol. 181, 2455-2458. (8) Lemaire, M., Miras, I., Gounon, P., and Beguin, P. (1998) Identification of a region responsible for binding to the cell wall within the S-layer protein of Clostridium thermocellum. Microbiology 144, 211-217. (9) Mesnage, S., Fontaine, T., Mignot, T., Delepierre, M., Mock, M., and Fouet, A. (2000) Bacterial SLH domain proteins are noncovalently anchored to the cell surface via a conserved mechanism involving wall polysaccharide pyruvylation. EMBO J. 19, 4473-4484. (10) Sa´ra, M. (2001) Conserved anchoring mechanisms between crystalline cell surface S-layer proteins and secondary cell wall polymers in Gram-positive bacteria? Trends Microbiol. 9, 47-49. (11) Ilk, N., Vo¨llenkle, C., Egelseer, E. M., Breitwieser, A., Sleytr, U. B., and Sa´ra, M. (2002) Molecular characterization of the S-layer gene, sbpA, of Bacillus sphaericus CCM 2177 and production of a functional S-layer fusion protein with the ability to recrystallize in a defined orientation while presenting the fused allergen. Appl. Environ. Microbiol. 68, 32513260. (12) Ilk, N., Kosma, P., Puchberger, M., Egelseer, E. M., Mayer, H. F., Sleytr, U. B., and Sa´ra, M. (1999) Structural and functional analyses of the secondary cell wall polymer of Bacillus sphaericus CCM 2177 that serves as an S-layerspecific anchor. J. Bacteriol. 181, 7643-7646. (13) Muyldermans, S. (2001) Single domain camel antibodies: current status. Mol. Biotechnol. 74, 277-302. (14) Nguyen, V. K., Desmyter, A., and Muyldermans, S. (2001) Functional heavy-chain antibodies in Camelidae. Adv. Immunol. 79, 261-296. (15) Vu, K. B., Ghahroudi, M. A., Wyns, L., and Muyldermans, S. (1997) Comparison of llama VH sequences from conventional and heavy chain antibodies. Mol. Immunol. 34, 11211131.

Bioconjugate Chem., Vol. 15, No. 3, 2004 671 (16) Jarosch, M., Egelseer, E. M., Mattanovich, D., Sleytr, U. B., and Sa´ra, M. (2000) S-layer gene sbsC of Bacillus stearothermophilus ATCC 12980: molecular characterization and heterologous expression in Escherichia coli. Microbiology 146, 273-281. (17) Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. (18) Messner P., Hollaus F., and Sleytr, U. B. (1984) Paracrystalline cell wall surface layers of different Bacillus stearothermophilus strains. Int. J. Syst. Bacteriol. 34, 202-210. (19) Pleschberger, M., Neubauer, A., Egelseer, E. M., Weigert, S., Lindner, B., Sleytr, U. B., Muyldermans, S., and Sa´ra, M. (2003) Generation of a functional monomolecular protein lattice consisting of an S-layer fusion protein comprising the variable domain of a camel heavy chain antibody. Bioconjugate Chem. 14, 440-448. (20) Egelseer, E. M., Schocher, I., Sleytr, U. B., and Sa´ra, M. (1996) Evidence that an N-terminal S-layer protein fragment triggers the release of a cell-associated high-molecular-weight amylase in Bacillus stearothermophilus ATCC 12980. J. Bacteriol. 178, 5602-5609. (21) Lauwereys, M., Arbabi Ghahroudi, M., Desmyter, A., Kinne, J., Holzer, W., De Genst, E., Wyns, L., and Muyldermans, S. (1998) Potent enzyme inhibitors derived from dromedary heavy-chain antibodies. EMBO J. 17, 3512-3520. (22) Vo¨llenkle, C., Weigert, S., Ilk, N., Egelseer, E. M., Weber, V., Loth, F., Falkenhagen, D., B., Sleytr, U. B., and Sa´ra, M. (2004) Construction of a functional S-layer fusion protein comprising an immunglobulin G-binding domain for development of specific adsorbents for extracorporeal blood purification. Appl. Environ. Microbiol. 70, 1514-1522. (23) Breitwieser, A., Egelseer, E. M., Moll, D., Ilk, N., Hotzy, C., Bohle, B., Ebner, C., Sleytr, U. B., and Sa´ra, M. (2002) A recombinant bacterial cell surface (S-layer)-major birch pollen allergen-fusion protein (rSbsC/Bet v1) maintains the ability to self-assemble into regularly structured monomolecular lattices and the functionality of the allergen. Protein Eng. 15, 243-249. (24) Bohle, B., Breitwieser, A., Zwo¨lfer, B., Jahn-Schmid, B., Sa´ra, M., Sleytr, U. B., and Ebner, C. (2004) A novel approach to specific allergy treatment: The recombinant fusion protein of a bacterial surface (S-layer) protein and the major birch pollen allergen Bet v1 (rSbsC-Bet v1) combines reduced allergenicity and immunomodulating capacity. J. Immunol., in press.

BC049964W