Selection, characterization and in vivo evaluation of novel CD44v6-targeting antibodies for targeted molecular radiotherapy.

Molecular radiotherapy combines the advantages of systemic administration of highly specific antibodies or peptides and the localized potency of ionizing radiation. A potential target for molecular radiotherapy is the cell surface antigen CD44v6, which is overexpressed in numerous cancers, with limited expression in normal tissues. The aim of the present study was to generate and characterize a panel of human anti-CD44v6 antibodies and identify a suitable candidate for future use in molecular radiotherapy of CD44v6-expressing cancers. Binders were first isolated from large synthetic phage display libraries containing human scFv and Fab antibody fragments. The antibodies were extensively analyzed through in vitro investigations of binding kinetics, affinity, off-target binding, and cell binding. Lead candidates were further subjected to in vivo biodistribution studies in mice bearing anaplastic thyroid cancer xenografts that express high levels of CD44v6. Additionally, antigen-dependent tumor uptake of the lead candidate was verified in additional xenograft models with varying levels of target expression. Interestingly, although only small differences were observed among the top antibody candidates in vitro, significant differences in tumor uptake and retention were uncovered in in vivo experiments. A high-affinity anti-CD44v6 lead drug candidate was identified, mAb UU-40, which exhibited favorable target binding properties and in vivo distribution. In conclusion, a panel of human anti-CD44v6 antibodies was successfully generated and characterized in this study. Through comprehensive evaluation, mAb UU-40 was identified as a promising lead candidate for future molecular radiotherapy of CD44v6-expressing cancers due to its high affinity, excellent target binding properties, and desirable in vivo distribution characteristics.

Selection of TNF-alpha binding affibody molecules using a beta-lactamase protein fragment complementation assay.

Protein fragment complementation assays (PCAs) based on different reporter proteins have been described as powerful tools for monitoring dynamic protein-protein interactions in living cells. The present study describes the construction of a PCA system based on genetic splitting of TEM-1 beta-lactamase for the selection of proteins specifically interacting in the periplasm of Escherichia coli bacterial cells, and its application for the selection of affibody molecules binding human tumour necrosis factor-alpha (TNF-alpha) from a combinatorial library. Vectors encoding individual members of a naïve 10(9) affibody protein library fused to a C-terminal fragment of the beta-lactamase reporter were distributed via phage infection to a culture of cells harbouring a common construct encoding a fusion protein between a non-membrane anchored version of a human TNF-alpha target and the N-terminal segment of the reporter. An initial binding analysis of 29 library variants derived from surviving colonies using selection plates containing ampicillin and in some cases also the beta-lactamase inhibitor tazobactam, indicated a stringent selection for target binding variants. Subsequent analyses showed that the binding affinities (K(D)) for three selected variants studied in more detail were in the range 14-27 nm. The selectivity in binding to TNF-alpha for these variants was further demonstrated in both a cross-target PCA-based challenge and the specific detection of a low nm concentration of TNF-alpha spiked into a complex cell lysate sample. Further, in a biosensor-based competition assay, the binding to TNF-alpha of three investigated affibody variants could be completely blocked by premixing the target with the therapeutic monoclonal antibody adalimumab (Humira), indicating overlapping epitopes between the two classes of reagents. The data indicate that beta-lactamase PCA is a promising methodology for stringent selection of binders from complex naïve libraries to yield high affinity reagents with selective target binding characteristics.

Decreased immune reactivity towards a knobless, affibody-targeted adenovirus type 5 vector.

In this study, a prototype Adenovirus type 5 (Ad5) vector deleted of the fiber knob domain and carrying an Affibody molecule as the targeting ligand showed decreased susceptibility to human pre-existing antibodies. This vector, Ad5/R7-Z(taq)Z(taq), has short fibers carrying seven shaft repeats, a non-native trimerization signal and an affibody molecule (Z(taq)) reactive to Taq polymerase. Ad5/R7-Z(taq)Z(taq) could be specifically targeted to 293 cells stably expressing membrane-bound anti-Z(taq) idiotypic affibody called Z(ztaq) (293Z(ztaq)). Sera from 50 blood donors were analyzed for neutralization activity (NA) against the parental Ad5/Fiwt vector and knobless Ad5/R7-Z(taq)Z(taq) on 293Z(ztaq) cells. Twenty-three sera had NA titers (> or =1:64) against Ad5/Fiwt (46%) and only two against Ad5/R7-Z(taq)Z(taq) (4%). Characterization of sera with NA titers showed that the knob domain is one of the targets of the antibodies. Neutralization assays using sera pre-adsorbed on knob and hexon proteins showed that the NA of the sera was carried mainly by anti-knob and anti-hexon antibodies, but in certain sera the anti-hexon antibodies represent the major population of the neutralizing antibodies (NAbs). Our results suggested that a combination of knob deletion and hexon switching could be an effective strategy for Ad vectors to better evade the anti-Ad NAbs.

Tumor cell targeted gene delivery by adenovirus 5 vectors carrying knobless fibers with antibody-binding domains.

Most human carcinoma cell lines lack the high-affinity receptors for adenovirus serotype 5 (Ad5) at their surface and are nonpermissive to Ad5. We therefore tested the efficiency of retargeting Ad5 to alternative cellular receptors via immunoglobulin (Ig)-binding domains inserted at the extremity of short-shafted, knobless fibers. The two recombinant Ad5's constructed, Ad5/R7-Z(wt)-Z(wt) and Ad5/R7-C2-C2, carried tandem Ig-binding domains from Staphylococcal protein A (abbreviated Z(wt)) and from Streptococcal protein G (C2), respectively. Both viruses bound their specific Ig isotypes with the expected affinity. They transduced human carcinoma cells independently of the CAR pathway, via cell surface receptors targeted by specific monoclonal antibodies, that is, EGF-R on A549, HT29 and SW1116, HER-2/neu on SK-OV-3 and SK-BR-3, CA242 (epitope recognized by the monoclonal antibody C242) antigen on HT29 and SW1116, and PSMA (prostate-specific membrane antigen) expressed on HEK-293 cells, respectively. However, Colo201 and Colo205 cells were neither transduced by targeting CA242 or EGF-R nor were LNCaP cells transduced by targeting PSMA. Our results suggested that one given surface receptor could mediate transduction of certain cells but not others, indicating that factors and steps other than cell surface expression and virus-receptor interaction are additional determinants of Ad5-mediated transduction of tumor cells. Using penton base RGD mutants, we found that one of these limiting steps was virus endocytosis.

Inhibition of the CD28-CD80 co-stimulation signal by a CD28-binding affibody ligand developed by combinatorial protein engineering.

CD28 is one of the key molecules for co-stimulatory signalling in T cells. Here, novel ligands (affibodies) showing selective binding to human CD28 (hCD28) have been selected by phage display technology from a protein library constructed through combinatorial mutagenesis of a 58-residue three-helix bundle domain derived from staphylococcal protein A. Analysis of selected affibodies showed a marked sequence homology and biosensor analyses showed that all investigated affibodies bound to hCD28 with micromolar affinities (KD). No cross-reactivity towards the related protein human CTLA-4 could be observed. This lack of cross-reactivity to hCTLA-4 suggests that the recognition site on hCD28 for the affibodies resides outside the conserved MYPPPYY motif. The apparent binding affinity for hCD28 could be improved through fusion to an Fc fragment fusion partner, resulting in a divalent presentation of the affibody ligand. For the majority of selected anti-CD28 affibodies, in co-culture experiments involving Jurkat T-cells and CHO cell lines transfected to express human CD80 (hCD80) or LFA-3 (hLFA-3) on the cell surface, respectively, pre-incubation of Jurkat cells with affibodies resulted in inhibition of IL-2 production when they were co-cultured with CHO (hCD80+) cells, but not with CHO (hLFA-3+) cells. For one affibody variant denoted Z(CD28:5) a clear concentration-dependent inhibition was seen, indicating that this affibody binds hCD28 and specifically interferes in the interaction between hCD28 and hCD80.

Genetic modification of adenovirus 5 tropism by a novel class of ligands based on a three-helix bundle scaffold derived from staphylococcal protein A.

The use of adenovirus (Ad) as an efficient and versatile vector for in vivo tumor therapy requires the modulation of its cellular tropism. We previously developed a method to genetically alter the tropism of Ad5 fibers by replacing the fiber knob domain by an extrinsic trimerization motif and a new cellular ligand. However, fibers carrying complex ligands such as single-chain antibody fragments did not assemble into functional pentons in vitro in the presence of penton base, and failed to be rescued into infectious virions because of their inability to fold correctly within the cytoplasm of Ad-infected cells. Here we show that the coding sequence for a disulfide bond-independent three-helix bundle scaffold Z, derived from domain B of Staphylococcal protein A and capable of binding to the Fc portion of immunoglobulin (Ig) G1, could be incorporated into modified knobless Ad fiber gene constructs with seven shaft repeats. These fiber gene constructs could be rescued into viable virions that were demonstrated to enter 293 cells engineered for IgG Fc surface expression but not unmodified 293 cells, via a mechanism that could be specifically blocked with soluble Fc target protein. However, the tropism modified viruses showed a slightly impaired cellular entry and a lower infectivity than wildtype (WT) virus. In addition, we generated recombinant fibers containing an IgA binding Affibody ligand, derived from combinatorial specificity-engineering of the Z domain scaffold. Such fiber constructs also showed the expected target specific binding, indicating that the affibody protein class is ideally suited for genetic engineering of Ad tropism.

Generation of metal-binding staphylococci through surface display of combinatorially engineered cellulose-binding domains.

Ni(2+)-binding staphylococci were generated through surface display of combinatorially engineered variants of a fungal cellulose-binding domain (CBD) from Trichoderma reesei cellulase Cel7A. Novel CBD variants were generated by combinatorial protein engineering through the randomization of 11 amino acid positions, and eight potentially Ni(2+)-binding CBDs were selected by phage display technology. These new variants were subsequently genetically introduced into chimeric surface proteins for surface display on Staphylococcus carnosus cells. The expressed chimeric proteins were shown to be properly targeted to the cell wall of S. carnosus cells, since full-length proteins could be extracted and affinity purified. Surface accessibility for the chimeric proteins was demonstrated, and furthermore, the engineered CBDs, now devoid of cellulose-binding capacity, were shown to be functional with regard to metal binding, since the recombinant staphylococci had gained Ni(2+)-binding capacity. Potential environmental applications for such tailor-made metal-binding bacteria as bioadsorbents in biofilters or biosensors are discussed.

Recombinant human factor VIII-specific affinity ligands selected from phage-displayed combinatorial libraries of protein A.

Factor VIII-specific affibodies were selected from phage displayed libraries constructed by combinatorial mutagenesis of an alpha helical bacterial receptor domain derived from staphylococcal protein A. Bead-immobilized recombinant human factor VIII (rVIII) (80 and 90 kDa chains) protein was used during competitive biopannings in the presence of free 80-kDa chain protein, resulting in the selection of several binders that showed dissociation constants (Kd) in the range 100-200 nM as determined by biosensor analyses. One variant (Z[rVIII:3], 90-kDa chain specific) was further characterized in small-scale affinity chromatography experiments, and showed efficient and selective recovery of biologically active rVIII from Chinese hamster ovary cell supernatant-derived feed stocks. The purity of the enriched rVIII was comparable with rVIII material purified by immunoaffinity chromatography using a 90-kDa chain-specific monoclonal antibody. Interestingly, epitope mapping showed that the monoclonal antibody and the affibody ligand competed for the same or at least overlapping epitopes on rVIII. In addition, the Z[rVIII:3] variant was produced by peptide synthesis with a C-terminal cysteine to enable directed coupling to solid supports. This 59-residue protein was analyzed by circular dichroism and showed a secondary structure content similar to that of the parental Z domain used as scaffold. In biosensor studies, the synthetic affibody was immobilized recruiting the C-terminal cysteine residue, and demonstrated to bind both recombinantly produced and plasma-derived factor VIII. From a secondary library, constructed by re-randomization of relevant positions identified after alignment of the first-generation variants, a panel of affinity-improved second-generation affibodies were selected of which one clone showed a dissociation constant (Kd) for rVIII of 5 nM. Several of these variants also showed higher apparent binding efficiencies towards rVIII when analyzed as immobilized ligands in biosensor experiments. Taken together, the results suggest that affibody ligands produced by bacterial or synthetic routes could be of interest as an alternative to monoclonal antibodies in purification processes or as diagnostic or monitoring tools.

In vitro selection of enzymatically active lipase variants from phage libraries using a mechanism-based inhibitor.

The 'detergent lipase' Lipolase, from Thermomyces lanuginosa was subjected to a combinatorial protein engineering/phage display approach with the aim of identifying new enzyme variants with improved characteristics in the presence of detergents. First it was demonstrated that wild-type Lipolase could be produced in Escherichia coli retaining full activity and be displayed as an active enzyme fused to coat protein 3 on E. coli phage M13. A phagemid library designed to result in approximately two to three mutations per lipase gene was then constructed. Nine amino acids located in two regions close to the active site were targeted for randomization. Selections using a mechanism-based biotinylated inhibitor showed that phages displaying Lipolase could be specifically enriched from a population of control phages. Selections on a library phage stock in the presence of inhibitor and a commercial powder detergent resulted in a step-wise increase in the proportion of active clones. Analysis of 84 active clones revealed that they all expressed lipase activity, but with lower activities than that of a wild-type Lipolase-producing clone. In six of the seven most active clones a wild-type serine at position 83 had been replaced by threonine, a substitution known to alter the substrate chain length preference of Lipolase variants. Furthermore, the selection had enriched enzyme variants with a high degree of conservatism in one of the variegated regions, suggesting that this region is important for enzymatic activity and that the designed selection procedure was relevant. The selected variants contained primarily basic amino acid residues within the other variegated region. Taken together, the described results show that selection protocols based on enzymatic activity can be designed for this enzyme class which should be of importance for future protein engineering attempts.

Charge engineering of a protein domain to allow efficient ion-exchange recovery.

We have created protein domains with extreme surface charge. These mutated domains allow for ion-exchange chromatography under conditions favourable for selective and efficient capture, using Escherichia coli as a host organism. The staphylococcal protein A-derived domain Z (Zwt) was used as a scaffold when constructing two mutants, Zbasic1 and Zbasic2, with high positive surface charge. Far-ultraviolet circular dichroism measurements showed that they have a secondary structure content comparable to the parental molecule Zwt. Although melting temperatures (Tm) of the engineered domains were lower than that of the wild-type Z domain, both mutants could be produced successfully as intracellular full-length products in E. coli and purified to homogeneity by ion-exchange chromatography. Further studies performed on Zbasic1 and Zbasic2 showed that they were able to bind to a cation exchanger even at pH values in the 9 to 11 range. A gene fusion between Zbasic2 and the acidic human serum albumin binding domain (ABD), derived from streptococcal protein G, was also constructed. The gene product Zbasic2-ABD could be purified using cation-exchange chromatography from a whole cell lysate to more than 90% purity.

Alpha-amylase inhibitors selected from a combinatorial library of a cellulose binding domain scaffold.

A disulfide bridge-constrained cellulose binding domain (CBD(WT)) derived from the cellobiohydrolase Cel7A from Trichoderma reesei has been investigated for use in scaffold engineering to obtain novel binding proteins. The gene encoding the wild-type 36 aa CBD(WT) domain was first inserted into a phagemid vector and shown to be functionally displayed on M13 filamentous phage as a protein III fusion protein with retained cellulose binding activity. A combinatorial library comprising 46 million variants of the CBD domain was constructed through randomization of 11 positions located at the domain surface and distributed over three separate beta-sheets of the domain. Using the enzyme porcine alpha-amylase (PPA) as target in biopannings, two CBD variants showing selective binding to the enzyme were characterized. Reduction and iodoacetamide blocking of cysteine residues in selected CBD variants resulted in a loss of binding activity, indicating a conformation dependent binding. Interestingly, further studies showed that the selected CBD variants were capable of competing with the binding of the amylase inhibitor acarbose to the enzyme. In addition, the enzyme activity could be partially inhibited by addition of soluble protein, suggesting that the selected CBD variants bind to the active site of the enzyme.

Design and production of recombinant subunit vaccines.

The development of subunit vaccines is presently the main strategy being evaluated for prevention of infectious diseases. The use of recombinant-DNA techniques has facilitated the development of new principles for design and production of subunit vaccines. First of all, the properties of a target protein immunogen can be improved by the use of gene-fusion technology or by the creation of specific changes, to generate 'second-generation protein vaccines'. Properties that can be modified include protein solubility, protein stability, in vivo half-lives, etc. In addition, for subunit protein vaccine candidates, the immunogenic properties can be significantly augmented by the addition of immunopotentiating tags or by means of targeting to immunoreactive sites. The recombinant subunit vaccine can furthermore be adapted by gene-fusion technology, to be efficiently incorporated into immunopotentiating adjuvant systems. Also in passive vaccination strategies, i.e. the use of antibodies or antibody fragments for prevention of infectious diseases, the recombinant strategies have become increasingly important. Humanized antibodies and antibody fusion proteins represent common present anti-infectious-disease agents. The selected examples will indicate that recombinant strategies will indeed have an impact on the design, selection and production of recombinant proteins to be used in the prevention of infectious diseases.

Partitioning of peptides and recombinant protein-peptide fusions in thermoseparating aqueous two-phase systems: effect of peptide primary structure.

Genetic engineering has been used for fusion of peptides, with different length and composition, on a protein to study the effect on partitioning in an aqueous two-phase system. The system was composed of dextran and the thermoseparating ethylene oxide-propylene oxide random copolymer, EO30P070. Peptides containing tryptophan, proline, arginine or aspartate residues were fused at the C-terminus of the recombinant protein ZZ-cutinase. The aim was to find effective tags for the lipolytic enzyme cutinase for large-scale extraction. The target protein and peptide tags were partitioned separately and then together in the fusion proteins in order to gain increased understanding of the influence of certain amino acid residues on the partitioning. The salt K2SO4 was used to reduce the charge dependent salt effects on partitioning and to evaluate the contribution to the partition coefficient from the hydrophobic-hydrophilic properties of the amino acid residues. The effect of Trp on peptide partitioning was independent of the difference in primary structure for (Trp)n, (Trp-Pro)n, (Ala-Trp-Trp-Pro)n and was only determined by the number of Trp. The effect of the charged residues, Arg and Asp, was dependent on the surrounding residues, i.e. if they were situated next to Trp or not. The partitioning behaviour observed for the peptides was qualitatively and in some cases also quantitatively the same as for the fusion proteins. The effect of the salts sodium perchlorate and triethylammonium phosphate on the partitioning was also studied. The salt effects observed for the peptides were qualitatively similar to the effects observed for the fusion proteins.

Ligands selected from combinatorial libraries of protein A for use in affinity capture of apolipoprotein A-1M and taq DNA polymerase.

Here we show that robust and small protein ligands can be used for affinity capture of recombinant proteins from crude cell lysates. Two ligands selectively binding to bacterial Taq DNA polymerase and human apolipoprotein A-1(M), respectively, were used in the study. The ligands were selected from libraries of a randomized alpha-helical bacterial receptor domain derived from staphylococcal protein A and have dissociation constants in the micromolar range, which is typical after primary selection from these libraries consisting of approximately 40 million different members each. Using these ligands in affinity chromatography, both target proteins were efficiently recovered from crude cell lysates with high selectivities. No loss of column capacity or selectivity was observed for repeated cycles of sample loading, washing and low pH elution. Interestingly, column sanitation could be performed using 0. 5 M sodium hydroxide without significant loss of ligand performance. The results suggest that combinatorial approaches using robust protein domains as scaffolds can be a general tool in the process of designing purification strategies for biomolecules.

Genetic engineering of the Fusarium solani pisi lipase cutinase for enhanced partitioning in PEG-phosphate aqueous two-phase systems.

The Fusarium solani pisi lipase cutinase has been genetically engineered to investigate the influence of C-terminal peptide extensions on the partitioning of the enzyme in PEG-salt based aqueous two-phase bioseparation systems. Seven different cutinase lipase variants were constructed containing various C-terminal peptide extensions including tryptophan rich peptide tags ((WP)(2) and (WP)(4)), positively ((RP)(4)) and negatively ((DP)(4)) charged tags as well as combined tags with tryptophan together with either positively ((WPR)(4)) or negatively ((WPD)(4)) charged amino acids. The modified cutinase variants were stably produced in Escherichia coli as secreted to the periplasm from which they were efficiently purified by IgG-affinity chromatography employing an introduced N-terminal IgG-binding ZZ affinity fusion partner present in all variants. Partitioning experiments performed in a PEG 4000/sodium phosphate aqueous two-phase system showed that for variants containing either (WP)(2) or (WP)(4) peptide extensions, 10- to 70-fold increases in the partitioning to the PEG rich top-phase were obtained, when compared to the wild type enzyme. An increased partitioning was also seen for cutinase variants tagged with both tryptophans and charged amino acids, whereas the effect of solely charged peptide extensions was relatively small. In addition, when performing partitioning experiments from cell disintegrates, the (WP)(4)-tagged cutinase showed a similarly high PEG-phase partitioning, indicating that the effect from the peptide tag was unaffected by the background of the host proteins. Taken together, the results show that the partitioning of the recombinantly produced cutinase model enzyme could be significantly improved by relatively minor genetic engineering and that the effects observed for purified proteins are retained also in an authentic whole cell disintegrate system. The results presented should be of general interest also for the improvement of the partitioning properties of other industrially interesting proteins including bulk enzymes.

Integrated bioprocess for production of human proinsulin C-peptide via heat release of an intracellular heptameric fusion protein.

An integrated bioprocess has been developed suitable for production of recombinant peptides using a gene multimerization strategy and site-specific cleavage of the resulting gene product. The process has been used for production in E. coli of the human proinsulin C-peptide via a fusion protein BB-C7 containing seven copies of the 31-residues C-peptide monomer. The fusion protein BB-C7 was expressed at high level, 1.8 g l(-1), as a soluble gene product in the cytoplasm. A heat treatment procedure efficiently released the BB-C7 fusion protein into the culture medium. This step also served as an initial purification step by precipitating the majority of the host cell proteins, resulting in a 70% purity of the BB-C7 fusion protein. Following cationic polyelectrolyte precipitation of the nucleic acids and anion exchange chromatography, native C-peptide monomers were obtained by enzymatic cleavage at flanking arginine residues. The released C-peptide material was further purified by reversed-phase chromatography and size exclusion chromatography. The overall yield of native C-peptide at a purity exceeding 99% was 400 mg l(-1) culture, corresponding to an overall recovery of 56%. The suitability of this process also for the production of other recombinant proteins is discussed.

Display of active subtilisin 309 on phage: analysis of parameters influencing the selection of subtilisin variants with changed substrate specificity from libraries using phosphonylating inhibitors.

Many attempts have been made to endow enzymes with new catalytic activities. One general strategy involves the creation of random combinatorial libraries of mutants associated with an efficient screening or selection scheme. Phage display has been shown to greatly facilitate the selection of polypeptides with desired properties by establishing a close link between the polypeptide and the gene that encodes it. Selection of phage displayed enzymes for new catalytic activities remains a challenge. The aim of this study was to display the serine protease subtilisin 309 (savinase) from Bacillus lentus on the surface of filamentous fd phage and to develop selection schemes that allow the extraction of subtilisin variants with a changed substrate specificity from libraries. Subtilisins are produced as secreted preproenzyme that mature in active enzyme autocatalytically. They have a broad substrate specificity but exhibit a significant preference for hydrophobic residues and very limited reactivity toward charged residues at the P4 site in the substrate. Here, we show that savinase can be functionally displayed on phage in the presence of the proteic inhibitor CI2. The free enzyme is released from its complex with CI2 upon addition of the anionic detergent LAS. The phage-enzyme can be panned on streptavidin beads after labelling by reaction with (biotin-N-epsilon-aminocaproyl-cystamine-N'-glutaryl)-l-Ala-l-Ala-l-P ro-Phe(P)-diphenyl ester. Reactions of libraries, in which residues 104 and 107 forming part of the S4 pocket have been randomised, with (biotin-N-epsilon-aminocaproyl-cystamine-N'-glutaryl)-alpha-l-Lys-l-A la-l-Pro-Phe(P)-diphenylester allowed us to select enzymes with increased specific activity for a substrate containing a lysine in P4. Parameters influencing the selection as for instance the efficiency of maturation of mutant enzymes in libraries have been investigated.

Kinetic characterization of the interaction of the Z-fragment of protein A with mouse-IgG3 in a volume in chemical space.

The kinetic rate parameters for the interaction between a single domain analogue of staphylococcal protein A (Z) and a mouse-IgG3 monoclonal antibody (MAb) were measured in Hepes buffer with different chemical additives. Five buffer ingredients (pH, NaCl, DMSO, EDTA, and KSCN) were varied simultaneously in 16 experiments following a statistical experimental plan. The 16 buffers thus spanned a volume in chemical space. A mathematical model, using data from the buffer composition, was developed and used to predict apparent kinetic parameters in five new buffers within the spanned volume. Association and dissociation parameters were measured in the new buffers, and these agreed with the predicted values, indicating that the model was valid within the spanned volume. The pattern of variation of the kinetic parameters in relation to buffer composition was different for association and dissociation, such that pH influenced both association and dissociation and NaCl influenced only dissociation. This indicated that the recognition mechanism (association) and the stability of the formed complex (dissociation) involve different binding forces, which can be further investigated by kinetic studies in systematically varied buffers.

Staphylococcal surface display of immunoglobulin A (IgA)- and IgE-specific in vitro-selected binding proteins (affibodies) based on Staphylococcus aureus protein A.

An expression system designed for cell surface display of hybrid proteins on Staphylococcus carnosus has been evaluated for the display of Staphylococcus aureus protein A (SpA) domains, normally binding to immunoglobulin G (IgG) Fc but here engineered by combinatorial protein chemistry to yield SpA domains, denoted affibodies, with new binding specificities. Such affibodies, with human IgA or IgE binding activity, have previously been selected from a phage library, based on an SpA domain. In this study, these affibodies have been genetically introduced in monomeric or dimeric forms into chimeric proteins expressed on the surface of S. carnosus by using translocation signals from a Staphylococcus hyicus lipase construct together with surface-anchoring regions of SpA. The recombinant surface proteins, containing the IgA- or IgE-specific affibodies, were demonstrated to be expressed as full-length proteins, localized and properly exposed at the cell surface of S. carnosus. Furthermore, these chimeric receptors were found to be functional, since recombinant S. carnosus cells were shown to have gained IgA and IgE binding capacity, respectively. In addition, a positive effect in terms of IgA and IgE reactivity was observed when dimeric versions of the affibodies were present. Potential applications for recombinant bacteria with redirected binding specificity in their surface proteins are discussed.

An in vitro selected binding protein (affibody) shows conformation-dependent recognition of the respiratory syncytial virus (RSV) G protein.

Using phage-display technology, a novel binding protein (Z-affibody) showing selective binding to the RSV (Long strain) G protein was selected from a combinatorial library of a small alpha-helical protein domain (Z), derived from staphylococcal protein A (SPA). Biopanning of the Z-library against a recombinant fusion protein comprising amino acids 130-230 of the G protein from RSV-subgroup A, resulted in the selection of a Z-affibody (Z(RSV1)) which showed G protein specific binding. Using biosensor technology, the affinity (K(D)) between Z(RSV1) and the recombinant protein was determined to be in the micromolar range (10(-6) M). Interestingly, the Z(RSV1) affibody was demonstrated to also recognize the partially (54%) homologous G protein of RSV subgroup B with similar affinity. Using different recombinant RSV G protein derived fragments, the binding was found to be dependent on the presence of the cysteinyl residues proposed to be involved in the formation of an intramolecular disulfide-constrained loop structure, indicating a conformation-dependent binding. Results from epitope mapping studies, employing a panel of monoclonal antibodies directed to different RSV G protein subfragments, suggest that the Z(RSV1) affibody binding site is located within the region of amino acids 164-186 of the G protein. This region contains a 13 amino acid residue sequence which is totally conserved between subgroups A and B of RSV and extends into the cystein loop region (amino acids 173-186). The potential use of the RSV G protein-specific Z(RSV1) affibody in diagnostic and therapeutic applications is discussed.