Development of an ELISA-Based Potency Assay for Inactivated Influenza Vaccines Using Cross-Reactive Nanobodies

Inactivated vaccines are the main influenza vaccines used today; these are usually presented as split (detergent-disrupted) or subunit vaccines, while whole-virus-inactivated influenza vaccines are rare. The single radial immune diffusion (SRD) assay has been used as the gold standard potency assay for inactivated influenza vaccines for decades; however, more recently, various alternative potency assays have been proposed. A new potency test should be able to measure the amount of functional antigen in the vaccine, which in the case of influenza vaccines is the haemagglutinin (HA) protein. Potency tests should also be able to detect the loss of potency caused by changes to the structural and functional integrity of HA. To detect such changes, most alternative potency tests proposed to date use antibodies that react with native HA. Due to the frequent changes in influenza vaccine composition, antibodies may need to be updated in line with changes in vaccine viruses. We have developed two ELISA-based potency assays for group 1 influenza A viruses using cross-reactive nanobodies. The nanobodies detect influenza viruses of subtype H1N1 spanning more than three decades, as well as H5N1 viruses, in ELISA. We found that the new ELISA potency assays are sensitive to the nature of the reference antigen (standard) used to quantify vaccine antigens; using standards matched in their presentation to the vaccine type improved correspondence between the ELISA and SRD assays.

Cross-Reactive SARS-CoV-2 Neutralizing Antibodies From Deep Mining of Early Patient Responses

Passive immunization using monoclonal antibodies will play a vital role in the fight against COVID-19. The recent emergence of viral variants with reduced sensitivity to some current antibodies and vaccines highlights the importance of broad cross-reactivity. This study describes deep-mining of the antibody repertoires of hospitalized COVID-19 patients using phage display technology and B cell receptor (BCR) repertoire sequencing to isolate neutralizing antibodies and gain insights into the early antibody response. This comprehensive discovery approach has yielded a panel of potent neutralizing antibodies which bind distinct viral epitopes including epitopes conserved in SARS-CoV-1. Structural determination of a non-ACE2 receptor blocking antibody reveals a previously undescribed binding epitope, which is unlikely to be affected by the mutations in any of the recently reported major viral variants including B.1.1.7 (from the UK), B.1.351 (from South Africa) and B.1.1.28 (from Brazil). Finally, by combining sequences of the RBD binding and neutralizing antibodies with the B cell receptor repertoire sequencing, we also describe a highly convergent early antibody response. Similar IgM-derived sequences occur within this study group and also within patient responses described by multiple independent studies published previously.

Broad Reactivity Single Domain Antibodies against Influenza Virus and Their Applications to Vaccine Potency Testing and Immunotherapy

The antigenic variability of influenza presents many challenges to the development of vaccines and immunotherapeutics. However, it is apparent that there are epitopes on the virus that have evolved to remain largely constant due to their functional importance. These more conserved regions are often hidden and difficult to access by the human immune system but recent efforts have shown that these may be the Achilles heel of the virus through development and delivery of appropriate biological drugs. Amongst these, single domain antibodies (sdAbs) are equipped to target these vulnerabilities of the influenza virus due to their preference for concave epitopes on protein surfaces, their small size, flexible reformatting and high stability. Single domain antibodies are well placed to provide a new generation of robust analytical reagents and therapeutics to support the constant efforts to keep influenza in check.

Nanobodies mapped to cross-reactive and divergent epitopes on A(H7N9) influenza hemagglutinin using yeast display

Influenza H7N9 virus continues to cause infections in humans and represents a significant pandemic risk. During the most recent 5th epidemic wave in 2016/17 two distinct lineages with increased human infections and wider geographical spread emerged. In preparation for any future adaptations, broadly reactive antibodies against H7N9 are required for surveillance, therapy and prophylaxis. In this study we have isolated a panel of nanobodies (Nbs) with broad reactivity across H7 influenza strains, including H7N9 strains between 2013 and 2017. We also describe Nbs capable of distinguishing between the most recent high and low pathogenicity Yangtze River Delta lineage H7N9 strains. Nanobodies were classified into 5 distinct groups based on their epitope footprint determined using yeast display and mutational scanning. The epitope footprint of Nbs capable of distinguishing high pathogenic (HP) A/Guangdong/17SF003/2016 from low pathogenic (LP) A/Hong Kong/125/2017 (H7N9) were correlated to natural sequence divergence in the head domain at lysine 164. Several Nbs binding to the head domain were capable of viral neutralisation. The potency of one nanobody NB7-14 could be increased over 1000-fold to 113 pM by linking two Nbs together. Nbs specific for distinct epitopes on H7N9 may be useful for surveillance or therapy in human or veterinary settings.

Protection From Influenza by Intramuscular Gene Vector Delivery of a Broadly Neutralizing Nanobody Does Not Depend on Antibody Dependent Cellular Cytotoxicity

Cross-subtype neutralizing single domain antibodies against influenza present new opportunities for immunoprophylaxis and pandemic preparedness. Their simple modular structure and single open reading frame format are highly amenable to gene therapy-mediated delivery. We have previously described R1a-B6, an alpaca-derived single domain antibody (nanobody), that is capable of potent cross-subtype neutralization in vitro of H1N1, H5N1, H2N2, and H9N2 influenza viruses, through binding to a highly conserved epitope in the influenza hemagglutinin stem region. To evaluate the potential of R1a-B6 for immunoprophylaxis, we have reformatted it as an Fc fusion for adeno-associated viral (AAV) vector delivery. Our findings demonstrate that a single intramuscular injection in mice of AAV encoding R1a-B6 fused to Fc fragments of different isotypes equipped either, with or without antibody dependent cellular cytotoxicity (ADCC) activity, was able to drive sustained high-level expression (0.5-1.1 mg/mL) in sera with no evidence of reduction for up to 6 months. R1a-B6-Fc fusions of both isotypes gave complete protection against lethal challenge with both pandemic A/California/07/2009 (H1N1)pdm09 and avian influenza A/Vietnam/1194/2004 (H5N1). This data suggests that R1a-B6 is capable of cross-subtype protection and ADCC was not essential for R1a-B6 efficacy. Our findings demonstrate AAV delivery of cross-subtype neutralizing nanobodies may be an effective strategy to prevent influenza infection and provide long-term protection independent of a host induced immune response.

Cross-Reactive and Lineage-Specific Single Domain Antibodies against Influenza B Hemagglutinin

Influenza B virus (IBV) circulates in the human population and causes considerable disease burden worldwide, each year. Current IBV vaccines can struggle to mount an effective cross-reactive immune response, as strains become mismatched, due to constant antigenic changes. Additional strategies which use monoclonal antibodies, with broad reactivity, are of considerable interest, both, as diagnostics and as immunotherapeutics. Alternatives to conventional monoclonal antibodies, such as single domain antibodies (NanobodiesTM) with well-documented advantages for applications in infectious disease, have been emerging. In this study we have isolated single domain antibodies (sdAbs), specific to IBV, using alpacas immunised with recombinant hemagglutinin (HA) from two representative viruses, B/Florida/04/2006 (B/Yamagata lineage) and B/Brisbane/60/2008 (B/Victoria lineage). Using phage display, we have isolated a panel of single domain antibodies (sdAbs), with both cross-reactive and lineage-specific binding. Several sdAbs recognise whole virus antigens, corresponding to influenza B strains included in vaccines spanning over 20 years, and were capable of neutralising IBV pseudotypes corresponding to prototype strains from both lineages. Lineage-specific sdAbs recognised the head domain, whereas, sdAbs identified as cross-reactive could be classified as either head binding or stem binding. Using yeast display, we were able to correlate lineage specificity with naturally occurring sequence divergence, at residue 122 in the highly variable 120 loop of the HA1 domain. The single domain antibodies described, might have applications in IBV diagnostics, vaccine potency testing and as immunotherapeutics.

International standards for monoclonal antibodies to support pre- and post-marketing product consistency: Evaluation of a candidate international standard for the bioactivities of rituximab

The intrinsic complexity and heterogeneity of therapeutic monoclonal antibodies is built into the biosimilarity paradigm where critical quality attributes are controlled in exhaustive comparability studies with the reference medicinal product. The long-term success of biosimilars will depend on reassuring healthcare professionals and patients of consistent product quality, safety and efficacy. With this aim, the World Health Organization has endorsed the need for public bioactivity standards for therapeutic monoclonal antibodies in support of current controls. We have developed a candidate international potency standard for rituximab that was evaluated in a multi-center collaborative study using participants' own qualified Fc-effector function and cell-based binding bioassays. Dose-response curve model parameters were shown to reflect similar behavior amongst rituximab preparations, albeit with some differences in potency. In the absence of a common reference standard, potency estimates were in poor agreement amongst laboratories, but the use of the candidate preparation significantly reduced this variability. Our results suggest that the candidate rituximab standard can support bioassay performance and improve data harmonization, which when implemented will promote consistency of rituximab products over their life-cycles. This data provides the first scientific evidence that a classical standardization exercise allowing traceability of bioassay data to an international standard is also applicable to rituximab. However, we submit that this new type of international standard needs to be used appropriately and its role not to be mistaken with that of the reference medicinal product.

Cross-Neutralising Nanobodies Bind to a Conserved Pocket in the Hemagglutinin Stem Region Identified Using Yeast Display and Deep Mutational Scanning

Cross-neutralising monoclonal antibodies against influenza hemagglutinin (HA) are of considerable interest as both therapeutics and diagnostic tools. We have recently described five different single domain antibodies (nanobodies) which share this cross-neutralising activity and suggest their small size, high stability, and cleft binding properties may present distinct advantages over equivalent conventional antibodies. We have used yeast display in combination with deep mutational scanning to give residue level resolution of positions in the antibody-HA interface which are crucial for binding. In addition, we have mapped positions within HA predicted to have minimal effect on antibody binding when mutated. Our cross-neutralising nanobodies were shown to bind to a highly conserved pocket in the HA2 domain of A(H1N1)pdm09 influenza virus overlapping with the fusion peptide suggesting their mechanism of action is through the inhibition of viral membrane fusion. We also note that the epitope overlaps with that of CR6261 and F10 which are human monoclonal antibodies in clinical development as immunotherapeutics. Although all five nanobodies mapped to the same highly conserved binding pocket we observed differences in the size of the epitope footprint which has implications in comparing the relative genetic barrier each nanobody presents to a rapidly evolving influenza virus. To further refine our epitope map, we have re-created naturally occurring mutations within this HA stem epitope and tested their effect on binding using yeast display. We have shown that a D46N mutation in the HA2 stem domain uniquely interferes with binding of R2b-E8. Further testing of this substitution in the context of full length purified HA from 1918 H1N1 pandemic (Spanish flu), 2009 H1N1 pandemic (swine flu) and highly pathogenic avian influenza H5N1 demonstrated binding which correlated with D46 whereas binding to seasonal H1N1 strains carrying N46 was absent. In addition, our deep sequence analysis predicted that binding to the emerging H1N1 strain (A/Christchurch/16/2010) carrying the HA2-E47K mutation would not affect binding was confirmed experimentally. This demonstrates yeast display, in combination with deep sequencing, may be able to predict antibody reactivity to emerging influenza strains so assisting in the preparation for future influenza pandemics.

The breadth of cross sub-type neutralisation activity of a single domain antibody to influenza hemagglutinin can be increased by antibody valency

The response to the 2009 A(H1N1) influenza pandemic has highlighted the need for additional strategies for intervention which preclude the prior availability of the influenza strain. Here, 18 single domain VHH antibodies against the 2009 A(H1N1) hemagglutinin (HA) have been isolated from a immune alpaca phage displayed library. These antibodies have been grouped as having either (i) non-neutralising, (ii) H1N1 restricted neutralising or (iii) broad cross-subtype neutralising activity. The ability to neutralise different viral subtypes, including highly pathogenic avian influenza (H5N1), correlated with the absence of hemagglutination inhibition activity, loss of binding to HA at acid pH and the absence of binding to the head domain containing the receptor binding site. This data supports their binding to epitopes in the HA stem region and a mechanism of action other than blocking viral attachment to cell surface receptors. After conversion of cross-neutralising antibodies R1a-B6 and R1a-A5 into a bivalent format, no significant enhancement in neutralisation activity was seen against A(H1N1) and A(H5N1) viruses. However, bivalent R1a-B6 showed an 18 fold enhancement in potency against A(H9N2) virus and, surprisingly, gained the ability to neutralise an A(H2N2) virus. This demonstrates that cross-neutralising antibodies, which make lower affinity interactions with the membrane proximal stem region of more divergent HA sub-types, can be optimised by bivalency so increasing their breadth of anti-viral activity. The broad neutralising activity and favourable characteristics, such as high stability, simple engineering into bivalent molecules and low cost production make these single domain antibodies attractive candidates for diagnostics and immunotherapy of pandemic influenza.

Affinity maturation and functional dissection of a humanised anti-RAGE monoclonal antibody by ribosome display

The pursuit of more potent, safe, and cost-effective drugs has placed a greater emphasis on antibody optimisation within the drug discovery process. Technologies to rapidly improve antibody drug performance, such as phage display, ribosome display, and yeast display, are playing a key role in this effort. Among these ribosome display is a particularly powerful technology and has recently been applied to the affinity optimisation of a humanised anti-receptor for advanced glycation end products (anti-RAGE) antibody (Finlay et al., J Mol Biol 388:541-558, 2009). By using a combination of error-prone PCR with ribosome display each amino acid position within this humanised antibody was scanned for both its functional importance and its capacity to increase affinity resulting in both affinity-matured antibody variants and a functional map of the antibody paratope.

Dissection of the IgNAR V domain: molecular scanning and orthologue database mining define novel IgNAR hallmarks and affinity maturation mechanisms

The shark antigen-binding V(NAR) domain has the potential to provide an attractive alternative to traditional biotherapeutics based on its small size, advantageous physiochemical properties, and unusual ability to target clefts in enzymes or cell surface molecules. The V(NAR) shares many of the properties of the well-characterised single-domain camelid V(H)H but is much less understood at the molecular level. We chose the hen-egg-lysozyme-specific archetypal Type I V(NAR) 5A7 and used ribosome display in combination with error-prone mutagenesis to interrogate the entire sequence space. We found a high level of mutational plasticity across the V(NAR) domain, particularly within the framework 2 and hypervariable region 2 regions. A number of residues important for affinity were identified, and a triple mutant combining A1D, S61R, and G62R resulted in a K(D) of 460 pM for hen egg lysozyme, a 20-fold improvement over wild-type 5A7, and the highest K(D) yet reported for V(NAR)-antigen interactions. These findings were rationalised using structural modelling and indicate the importance of residues outside the classical complementarity determining regions in making novel antigen contacts that modulate affinity. We also located two solvent-exposed residues (G15 and G42), distant from the V(NAR) paratope, which retain function upon mutation to cysteine and have the potential to be exploited as sites for targeted covalent modification. Our findings with 5A7 were extended to all known NAR structures using an in-depth bioinformatic analysis of sequence data available in the literature and a newly generated V(NAR) database. This study allowed us to identify, for the first time, both V(NAR)-specific and V(NAR)/Ig V(L)/TCR V(alpha) overlapping hallmark residues, which are critical for the structural and functional integrity of the single domain. Intriguingly, each of our designated V(NAR)-specific hallmarks align precisely with previously defined mutational 'cold spots' in natural nurse shark cDNA sequences. These findings will aid future V(NAR) engineering and optimisation studies towards the development of V(NAR) single-domain proteins as viable biotherapeutics.

Affinity maturation of a humanized rat antibody for anti-RAGE therapy: comprehensive mutagenesis reveals a high level of mutational plasticity both inside and outside the complementarity-determining regions

Antibodies that neutralize RAGE (receptor for advanced glycation end products)-ligand interactions have potential therapeutic applications in both acute and chronic diseases. We generated XT-M4, a rat anti-RAGE monoclonal antibody that has in vivo efficacy in an acute sepsis model. This antibody was subsequently humanized. To improve the affinity of this antibody for the treatment of chronic indications, we used random and targeted mutagenesis strategies in combination with ribosome and phage-display technologies, respectively, to generate libraries of XT-M4 variants. We identified a panel of single-chain Fv antibody fragments (scFv's) that was improved up to 110-fold in a homogeneous time-resolved fluorescence competition assay against parental XT-M4 immunoglobulin G (IgG). After reformatting to bivalent scFv-Fc fusions and IgGs, we observed similar gains in potency in the same assay. Further analysis of binding kinetics as IgG revealed multiple variants with subnanomolar apparent affinity that was dictated primarily by improvements in the off-rate. All variants also had improved binding to cell surface-expressed human RAGE, and all retained, or had improved, apparent affinity for mouse RAGE. F100bL in V(H) (variable region of the heavy chain) complementarity-determining region 3 (CDR3) was one of a number of key mutations that correlated with affinity improvements and was independently identified by both mutagenesis strategies. Random mutagenesis coupled with ribosome display and high-throughput screening revealed an unexpectedly high level of mutational plasticity across the whole length of the humanized scFv, suggesting greater scope for structural optimization outside of the primary antigen-combining site defined by V(H) CDR3 and V(kappa) CDR3. In summary, our comprehensive mutagenesis approach not only achieved the desired affinity maturation of XT-M4 but also defined multiple mutational hotspots across the antibody sequence, provided an insight into the specificity-determining residues of the antibody paratope, and identified additional sites within the CDR loops where human germ-line amino acids may be introduced without affecting function.

Identification and characterization of ErbB-3-binding protein-1 as a target for immunotherapy

Based on immune reactivity in response to a whole-cell colon tumor vaccine and using serological identification of Ags by recombinant cDNA expression cloning, we here describe the molecular and functional identification of a novel human tumor Ag. By screening a cDNA expression library derived from the coloncarcinoma cell line HT-29 with pooled colorectal cancer patients' sera, 26 clones reactive with IgG Abs could be identified. Characterization of these cDNA clones by sequence analysis and alignment, and detailed serological analysis revealed cancer-related immunoreactivity for the ErbB-3-binding protein-1 (Ebp1). Immunohistochemical staining of colorectal tumors and neighboring normal colon tissue indicated the observed cancer-related immunogenicity of Ebp1 to be related to overexpression. Via reverse immunology, five potential HLA-A2-restricted T cell epitopes were identified, of which two (Ebp1(45-54) and Ebp1(59-67)) bound HLA-A2 with intermediate and high affinity, respectively. Analysis of their immunogenicity in vitro indicated that only the high-affinity Ebp1(59) epitope gave rise to CD8(+) T cells capable of recognizing both exogenously loaded Ebp1 peptide and endogenously expressed Ebp1 on target cells. In addition, in vivo CD8(+) T cell responsiveness against the Ebp1(59) epitope could be detected in two of nine and three of six cancer patients PBMC and tumor draining lymph nodes, respectively, but not in nine of nine healthy donors tested. These data confirm that Ebp1 is an immunogenic protein, capable of eliciting CD8-mediated responses in vivo and in vitro, providing a rationale for further exploration of Ebp1 as a possible target for anticancer immunotherapy.

Generation of high-affinity human antibodies by combining donor-derived and synthetic complementarity-determining-region diversity

Combinatorial libraries of rearranged hypervariable V(H) and V(L) sequences from nonimmunized human donors contain antigen specificities, including anti-self reactivities, created by random pairing of V(H)s and V(L)s. Somatic hypermutation of immunoglobulin genes, however, is critical in the generation of high-affinity antibodies in vivo and occurs only after immunization. Thus, in combinatorial phage display libraries from nonimmunized donors, high-affinity antibodies are rarely found. Lengthy in vitro affinity maturation is often needed to improve antibodies from such libraries. We report the construction of human Fab libraries having a unique combination of immunoglobulin sequences captured from human donors and synthetic diversity in key antigen contact sites in heavy-chain complementarity-determining regions 1 and 2. The success of this strategy is demonstrated by identifying many monovalent Fabs against multiple therapeutic targets that show higher affinities than approved therapeutic antibodies. This very often circumvents the need for affinity maturation, accelerating discovery of antibody drug candidates.

Construction and diversification of yeast cell surface displayed libraries by yeast mating: application to the affinity maturation of Fab antibody fragments

Yeast display is a powerful technology for the affinity maturation of human antibody fragments. However, the technology thus far has been limited by the size of antibody libraries that can be generated, as using current transformation protocols libraries of only between 10(6) and 10(7) are typically possible. We have recently shown that Fab antibodies can be displayed on the cell surface of Saccharomyces cerevisiae [van den Beucken, T., Pieters, H., Steukers, M., van der Vaart, M., Ladner, R.C., Hoogenboom, H.R., Hufton, S.E., 2003. Affinity maturation of Fab antibody fragments by fluorescent-activated cell sorting of yeast-displayed libraries. FEBS Lett. 546, 288-294]. This discovery and the knowledge that Fab antibodies are heterodimeric suggest that independent repertoires of heavy chain (HC) and light chain (LC) can be constructed in haploid yeast strains of opposite mating type. These separate repertoires can then be combined by highly efficient yeast mating. Using this approach, we have rapidly generated a naive human Fab yeast display library of over 10(9) clones. In addition, utilizing error-prone polymerase chain reaction, we have diversified Fab sequences and generated combinatorial and hierarchical chain shuffled libraries with complexities of up to 5 x 10(9) clones. These libraries have been selected for higher affinity using a repeating process of mating-driven chain shuffling and flow cytometric sorting.

Affinity maturation of Fab antibody fragments by fluorescent-activated cell sorting of yeast-displayed libraries

We report for the first time the affinity maturation of Fab antibody fragments using fluorescent-activated cell sorting (FACS) of yeast-displayed repertoires. A single yeast display vector which enables the inducible expression of an anchored heavy chain and a soluble light chain has been constructed. The assembly and functional display on the yeast cell surface of Fab antibodies specific for different protein targets has been demonstrated by flow cytometry and immunofluorescence microscopy. We have affinity matured a Fab antibody specific for the tetravalent antigen streptavidin using FACS of yeast-displayed repertoires diversified by error-prone polymerase chain reaction. A panel of variants with up to 10.7-fold improvement in affinity was obtained after selection. Two leading clones, R2H10 (3.2 nM) and R3B1 (5.5 nM), had mutations in light chain complementarity determining region 1 LC-CDR1 (H34R) and LC-CDR3 (Y96H or Y96F) and gave a 10.7-fold and 6.3-fold affinity improvement over the starting antibody, respectively. The ability to efficiently affinity mature Fab antibodies is an important component of the antibody development pipeline and we have shown that yeast display is an efficient method for this purpose.

Target validation for genomics using peptide-specific phage antibodies: a study of five gene products overexpressed in colorectal cancer

Genomic approaches are providing a wealth of information on differential gene expression in cancer. To identify the most interesting genes amongst the many identified, high-throughput methods for analysis of genes at the translational level are required. We have used a rapid method for the in vitro selection of antibodies to peptide antigens for the generation of probes to 5 gene products that we have found to be overexpressed in colorectal cancer. The rationale of our study was to select a non-immune phage displayed human antibody library on peptides designed from the coding regions of the gene sequences and to verify whether such antibodies would be suitable probes for the parental protein in immunohistochemical and Western blot analysis. After the generation of a profile of genes overexpressed in primary colorectal cancer (CRC) we selected 5 genes, Ese-3b, Fls353, PBEF, SPARC and Smad5 for a more detailed analysis using phage display-derived antibodies. For these 5 antigens we designed 14-20 amino acid peptides predicted to be exposed on the surface of the parental protein. Selection of a large phage displayed antibody library resulted in specific antibodies for 6 of 8 different peptides with between 2 and 15 different antibodies isolated per peptide. Of 20 antibodies tested, 2 antibodies recognized the putative parental protein from primary CRC tissue. An antibody specific for a PBEF-derived peptide (Fab/PBEF-D4) was shown to recognize a protein product of the expected molecular weight in Western blotting and showed overexpression in n = 6/8 matched tumor/normal protein lysates. Furthermore, in immunohistochemistry this antibody showed restricted staining of the tumor stromal compartment with no detectable staining of epithelial cells. The discovery that PBEF is overexpressed in cancer is unexpected given that the normal function of PBEF is as a cytokine required for the maturation of B cell precursors. We also report on the isolation of an antibody (Fab/SMAD-50) specific for a Smad5-derived peptide that showed cytoplasmic staining of epithelial cells in both CRC tumor and matched normal mucosa. Fab/SMAD-50 also bound to a group of proteins in Western blotting with molecular weights consistent with belonging to the Smad family. These antibodies may be suitable probes for further investigation of the roles of PBEF and Smad5 in cancer. The amenability of phage display to automation suggests that this approach may be developed for implementation on a genomics scale. Indeed, the large-scale generation of antibody probes that can be used to study protein expression in situ would be of great value in target validation for functional genomics.

A panel of candidate tumor antigens in colorectal cancer revealed by the serological selection of a phage displayed cDNA expression library

In the last few years it has been shown that the humoral immune response in cancer patients is a rich source of putative cancer vaccine candidates. To fully explore the complex information present within the Ab repertoire of cancer patients, we have applied a method, serological Ag selection, to molecularly define tumor Ags recognized by the humoral immune response in colorectal cancer (CRC). First, we built a cDNA display library by cloning a cDNA library from CRC cell line HT-29 for expression as a fusion protein with a filamentous phage minor coat protein, pVI. This cDNA display library was then enriched on pooled sera from CRC patients who had undergone active specific immunization with autologous tumor. We identified a panel of 19 clones reactive with the serum pool. Seventeen of 19 (89%) clones showed reactivity with one or more of the eight Ag-reactive sera, conversely six of eight (75%) sera were reactive with at least one of the 19 clones. Sequencing revealed that these 19 clones represented 13 different Ags. A detailed serological analysis of the 13 different Ags showed preferential reactivity to sera of cancer patients for six different Ags. Four of these Ags displayed increased serum reactivity after the active specific immunization procedure. Furthermore, one of the six Ags, a novel Ag homologous to HSPC218, showed restricted expression in normal testis, suggesting that it belongs to the cancer-testis Ag family. Some of the Ags we have identified may be candidates for tumor vaccination, for sero-diagnosis of cancer, as prognostic markers, or as probes for monitoring tumor cell-based vaccination trials.

Building novel binding ligands to B7.1 and B7.2 based on human antibody single variable light chain domains

Ligands specific for B7.1 (CD80) and B7.2 (CD86) have applications in disease indications that require inhibition of T-cell activity. As we observed significant sequence and structural similarity between the B7-binding ligand, cytotoxic T-lymphocyte associated protein-4 (CTLA-4), and antibody variable light chain domains (VLs), we have explored the possibilities of making novel B7 binding molecules based on single VL domains. We first describe the "rational" design and construction of a VL/CTLA-4 hybrid molecule in which we have grafted both the CDR1 and CDR3-like loops of CTLA-4 onto a single VL light chain, at sites determined by sequence and structure-based alignment. This molecule was secreted as a soluble product from Escherichia coli, but did not show any binding to B7.1 and B7.2. In a second approach we constructed a VL library in which human VL genes derived from B-cells were spiked with the CDR3-like loop of CTLA-4 and further diversified by DNA shuffling. This library was displayed on phage, and after selection gave B7.1 binding ligands which competed with CTLA-4. In order to evaluate the possible general utility of VL domains as binding ligands, we have constructed a non-biased VL library. From this DNA-shuffled human VL library we have selected single VL domains specific for B7.1, B7.2 or human IgG. Two B7.1-specific VL ligands and one B7.2-specific VL ligand showed competition with CTLA-4. One candidate VL domain-specific for B7.1 was affinity matured by simultaneous randomisation of all CDR loops using DNA shuffling with degenerate CDR-spiking oligonucleotides. From this library, a single VL domain with affinity of 191 nM for B7.1 was obtained, which also showed binding to B7.1 in situ. This VL had mutations in CDR1 and CDR3, indicating that antigen recognition for this single VL is most likely mediated by the same regions as in the VL domain of whole antibodies. The B7.1 and B7.2-specific VL domains described in this study may form the basis of a new family of immunomodulatory recombinant molecules. Furthermore, our studies suggest that it is feasible to create specific single VL domains to diverse targets as is the case for single VH domains.

Direct selection of a human antibody fragment directed against the tumor T-cell epitope HLA-A1-MAGE-A1 from a nonimmunized phage-Fab library

Antitumor antibodies with the same specificity as cytotoxic T lymphocytes that recognize antigenic peptides encoded by tumor-associated genes and presented by MHC class I molecules would be valuable tools to analyze the antigenicity or target tumor cells in vivo. To obtain a human antibody directed against a peptide encoded by gene melanoma-associated antigen (MAGE)-A1 and presented by HLA-A1 molecules, we selected a large phage Fab antibody repertoire on a recombinant version of the complex HLA-A1-MAGE-A1 produced by in vitro refolding. One of the selected phage antibodies shows binding to HLA-A1 complexed with the MAGE-A1 peptide, but does not show binding to HLA-A1 complexed with a peptide encoded by gene MAGE-A3 and differing from the MAGE-A1 peptide by only three residues. Phages carrying this recombinant antibody bind to HLA-A1(+) cells only after in vitro loading with MAGE-A1 peptide. These results indicate that nonimmunized phage Fab libraries are a source of antibodies with a T cell antigen receptor-like specificity. The human anti-HLA-A1-MAGE-A1 antibody described here may prove very useful for monitoring the cell surface expression of these complexes, and eventually, as a targeting reagent for the specific immunotherapy of HLA-A1 patients bearing a MAGE-A1-positive tumor.

Development and application of cytotoxic T lymphocyte-associated antigen 4 as a protein scaffold for the generation of novel binding ligands

We have explored the possibilities of using human cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) as a single immunoglobulin fold-based scaffold for the generation of novel binding ligands. To obtain a suitable protein library selection system, the extracellular domain of CTLA-4 was first displayed on the surface of a filamentous phage as a fusion product of the phage coat protein p3. CTLA-4 was shown to be functionally intact by binding to its natural ligands B7-1 (CD80) and B7-2 (CD86) both in vitro and in situ. Secondly, the complementarity determining region 3 (CDR3) loop of the CTLA-4 extracellular domain was evaluated as a permissive site. We replaced the nine amino acid CDR3-like loop of CTLA-4 with the sequence XXX-RGD-XXX (where X represents any amino acid). Using phage display we selected several CTLA-4-based variants capable of binding to human alphavbeta3 integrin, one of which showed binding to integrins in situ. To explore the construction of bispecific molecules we also evaluated one other potential permissive site diametrically opposite the natural CDR-like loops, which was found to be tolerant of peptide insertion. Our data suggest that CTLA-4 is a suitable human scaffold for engineering single-domain molecules with one or possibly more binding specificities.

A profile of differentially expressed genes in primary colorectal cancer using suppression subtractive hybridization

As a step towards understanding the complex differences between normal cells and cancer cells, we have used suppression subtractive hybridization (SSH) to generate a profile of genes overexpressed in primary colorectal cancer (CRC). From a 35¿ omitted¿000 clone SSH-cDNA repertoire, we have screened 400 random clones by reverse Northern blotting, of which 45 clones were scored as overexpressed in tumor compared to matched normal mucosa. Sequencing showed 37 different genes and of these, 16 genes corresponded to known genes in the public databases. Twelve genes, including Smad5 and Fls353, have previously been shown to be overexpressed in CRC. A series of known genes which have not previously been reported to be overexpressed in cancer were also recovered: Hsc70, PBEF, ribophorin II and Ese-3B. The remaining 21 genes have as yet no functional annotation. These results show that SSH in conjunction with high throughput screening provides a very efficient means to produce a broad profile of genes differentially expressed in cancer. Some of the genes identified may provide novel points of therapeutic intervention.

Phage display of cDNA repertoires: the pVI display system and its applications for the selection of immunogenic ligands

The selection of phage displayed cDNA repertoires on an immobilised target has been reported to be an efficient way to rapidly identify interacting partners. To date, however, only a few successful applications have been reported. Here, we present a review of the current status of the display and selection of cDNA libraries using phage. As an example, we report the construction of a set of phage display vectors suitable for cDNA display based on fusion to the minor bacteriophage coat protein 6 (pVI) of filamentous phage. We have evaluated these vectors through the display of the C(H)3 domain of human IgG and of the Escherichia coli alkaline phosphatase (PhoA) gene. Both the C(H)3 domain of IgG and PhoA are shown to be displayed on pVI, and PhoA is also shown to be enzymatically active. We have constructed primary colorectal tumor cDNA repertoires in these vectors and performed selections on both a monoclonal antibody to beta2 microglobulin (beta2M) and polyclonal antibody sera to human IgG. In both cases, relevant ligands were recovered from the phage displayed cDNA repertoire. These vectors may be used for selection of phage displayed cDNA libraries with polyclonal sera from patients. This will allow the identifying antigenic cDNA products in such diseases as cancer, viral/bacterial infections or autoimmune disease. Furthermore, by selections with other specific biomolecules, this display system may aid the identification of interacting partners in functional genomics.

Design and expression of soluble CTLA-4 variable domain as a scaffold for the display of functional polypeptides

We have designed and engineered the human cytotoxic T-lymphocyte associated protein-4 (CTLA-4) variable (V-like) domain to produce a human-based protein scaffold for peptide display. First, to test whether the CTLA-4 CDR-like loops were permissive to loop replacement/insertion we substituted either the CDR1 or CDR3 loop with somatostatin, a 14-residue intra-disulfide-linked neuropeptide. Upon expression as periplasmic-targeted proteins in Escherichia coli, molecules with superior solubility characteristics to the wild-type V-domain were produced. These mutations in CTLA-4 ablated binding to its natural ligands CD80 and CD86, whereas binding to a conformation-dependent anti-CTLA-4 monoclonal antibody showed that the V-domain framework remained correctly folded. Secondly, to develop a system for library selection, we displayed both wild-type and mutated CTLA-4 proteins on the surface of fd-bacteriophage as fusions with the geneIII protein. CTLA-4 displayed on phage bound specifically to immobilized CD80-Ig and CD86-Ig and in one-step panning enriched 5,000 to 2,600-fold respectively over wild-type phage. Bacteriophage displaying CTLA-4 with somatostatin in CDR3 (CTLA-4R-Som3) specifically bound somatostatin receptors on transfected CHO-K1 cells pre-incubated with 1 microg/ml tunicamycin to remove receptor glycosylation. Binding was specific, as 1 microM somatostatin successfully competed with CTLA-4R-Som3. CTLA-4R-Som3 also activated as well as binding preferentially to non-glycosylated receptor subtype Sst4. The ability to substitute CDR-like loops within CTLA-4 will enable design and construction of more complex libraries of single V-like domain binding molecules. Proteins 1999;36:217-227.

A large non-immunized human Fab fragment phage library that permits rapid isolation and kinetic analysis of high affinity antibodies

We report the design, construction, and use of the first very large non-immunized phage antibody library in Fab format, which allows the rapid isolation and affinity analysis of antigen-specific human antibody fragments. Individually cloned heavy and light chain variable region libraries were combined in an efficient two-step cloning procedure, permitting the cloning of a total of 3.7 x 10(10) independent Fab clones. The performance of the library was determined by the successful selection of on average 14 different Fabs against 6 antigens tested. These include tetanus toxoid, the hapten phenyl-oxazolone, the breast cancer-associated MUC1 antigen, and three highly related glycoprotein hormones: human chorionic gonadotropin, human luteinizing hormone, and human follicle-stimulating hormone. In the latter category, a panel of either homone-specific or cross-reactive antibodies were identified. The design of the library permits the monitoring of selections with polyclonal phage preparations and to carry out large scale screening of antibody off-rates with unpurified Fab fragments on BIAcore. Antibodies with off-rates in the order of 10(-2) to 10(-4) s-1 and affinities up to 2.7 nM were recovered. The kinetics of these phage antibodies are of the same order of magnitude as antibodies associated with a secondary immune response. This new phage antibody library is set to become a valuable source of antibodies to many different targets, and to play a vital role in target discovery and validation in the area of functional genomics.

Antibody phage display technology and its applications

In recent years, the use of display vectors and in vitro selection technologies has transformed the way in which we generate ligands, such as antibodies and peptides, for a given target. Using this technology, we are now able to design repertoires of ligands from scratch and use the power of phage selection to select those ligands having the desired (biological) properties. With phage display, tailor-made antibodies may be synthesized and selected to acquire the desired affinity of binding and specificity for in vitro and in vivo diagnosis, or for immunotherapy of human disease. This review addresses recent progress in the construction of, and selection from phage antibody libraries, together with novel approaches for screening phage antibodies. As the quality of large naïve and synthetic antibody repertoires improves and libraries becomes more generally available, new and exciting applications are pioneered such as the identification of novel antigens using differential selection and the generation of receptor a(nta)gonists. A combination of the design and generation of millions to billions of different ligands, together with phage display for the isolation of binding ligands and with functional assays for identifying (and possibly selecting) bio-active ligands, will open even more challenging applications of this inspiring technology, and provide a powerful tool for drug and target discovery well into the next decade.

Structure/function analysis of the domains required for the multimerisation of phenylalanine hydroxylase

Phenylalanine hydroxylase (PAH) exists as an equilibrium of dimers and tetramers. However, there is little information concerning the inter- or intra-molecular interactions required for enzyme quaternary structure. It is predicted that the formation of a PAH tetramer will require at least two points of contact per enzyme subunit. Sequence analysis has suggested the existence of a C-terminal domain with characteristics of a leucine zipper or a variant of this called a coiled-coil. By deletion of 24 amino acids from the C-terminus or conversion of leucine 448 to an alanine residue, we have shown that this putative leucine zipper/coiled-coil domain is involved in the assembly of an active enzyme tetramer from dimers. The removal of this C-terminal domain of PAH reduces enzyme activity but does not abolish it. Furthermore, we report that an alanine 447 to aspartate mutation associated with phenylketonuria may affect subunit assembly which suggests the formation of enzyme tetramers is physiologically relevant. Our analysis of subunit interactions in vivo, show that in the absence of the C-terminal coiled-coil domain, dimers can form and this is only possible when the N-terminal domain is present. This provides the first evidence that N-terminal domain is required for multimerisation. We propose that the N-terminal regulatory domain in conjunction with the C-terminal coiled-coil domain, mediates the formation of fully active enzyme tetramers.

Structure and function of the aromatic amino acid hydroxylases

Images

Structure-function analysis of the vitamin B12 receptor of Escherichia coli by means of informational suppression

We describe a genetic analysis of the vitamin B12 receptor of Escherichia coli. Through the use of informational suppression, we have been able to generate a family of receptor variants, each identical save for a single, known substitution (Ser, Gln, Lys, Tyr, Leu, Cys, Phe) at a known site. We have studied 22 different mutants, 14 in detail, distributed throughout the length of the btuB gene. Most amino acid substitutions have a pleiotropic effect with respect to all ligands tested, the two colicins E1 and E3, the T5-like bacteriophage BF23, and vitamin B12. (The dramatic effect of a single amino acid substitution is also well exemplified by the G142A missense change which renders the receptor completely non-functional.) In some instances, however, we have been able to modify a subset of receptor functions (viz. Q62, Q150 and Q299 and the response to phage BF23). These data are summarized on a two-dimensional folding model for the BtuB protein in the outer membrane (devised using both amphipathic beta-strand analysis and sequence conservation amongst the TonB-dependent receptors). In addition, we report that the extreme C-terminus of BtuB is vital for receptor localization and provide evidence for it being a membrane-spanning beta-sheet with residue L588 situated on its hydrophobic surface. Two of the C-terminal btuB mutations are located within the region of overlap with the recently identified dga (murl) gene.