Broadly neutralizing antibodies against highly variable pathogens have stimulated the design of novel vaccines and therapeutics. Here, we report on diverse camelid single-domain antibodies to influenza hemagglutinin from which we generated multi-domain antibodies with unprecedented breadth and potency. Multi-domain antibody MD3606 protects mice against influenza A and B infection when administered intravenously or expressed locally from a recombinant adeno-associated virus vector. Crystal and single-particle EM structures of these antibodies with hemagglutinins from influenza A and B viruses reveal binding to highly conserved epitopes. Collectively, our findings demonstrate that multi-domain antibodies targeting multiple epitopes exhibit enhanced virus cross-reactivity and potency. In combination with adeno-associated virus-mediated gene delivery, they may provide a groundbreaking new strategy to prevent infection with influenza virus and other highly variable pathogens.

Protein aggregation is of great concern to pharmaceutical formulations and has been implicated in several diseases. We engineered an anti-IL-13 monoclonal antibody CNTO607 for improved solubility. Three structure-based engineering approaches were employed in this study: (i) modifying the isoelectric point (pI), (ii) decreasing the overall surface hydrophobicity and (iii) re-introducing an N-linked carbohydrate moiety within a complementarity-determining region (CDR) sequence. A mutant was identified with a modified pI that had a 2-fold improvement in solubility while retaining the binding affinity to IL-13. Several mutants with decreased overall surface hydrophobicity also showed moderately improved solubility while maintaining a similar antigen affinity. Structural studies combined with mutagenesis data identified an aggregation 'hot spot' in heavy-chain CDR3 (H-CDR3) that contains three residues ((99)FHW(100a)). The same residues, however, were found to be essential for high affinity binding to IL-13. On the basis of the spatial proximity and germline sequence, we reintroduced the consensus N-glycosylation site in H-CDR2 which was found in the original antibody, anticipating that the carbohydrate moiety would shield the aggregation 'hot spot' in H-CDR3 while not interfering with antigen binding. Peptide mapping and mass spectrometric analysis revealed that the N-glycosylation site was generally occupied. This variant showed greatly improved solubility and bound to IL-13 with affinity similar to CNTO607 without the N-linked carbohydrate. All three engineering approaches led to improved solubility and adding an N-linked carbohydrate to the CDR was the most effective route for enhancing the solubility of CNTO607.

To assess the state of the art in antibody 3D modeling, 11 unpublished high-resolution x-ray Fab crystal structures from diverse species and covering a wide range of antigen-binding site conformations were used as a benchmark to compare Fv models generated by seven structure prediction methodologies. The participants included: Accerlys Inc, Chemical Computer Group (CCG), Schrodinger, Jeff Gray's lab at John Hopkins University, Macromoltek, Astellas Pharma/Osaka University and Prediction of ImmunoGlobulin Structure (PIGS). The sequences of benchmark structures were submitted to the modelers and PIGS, and a set of models were generated for each structure. We provide here an overview of the organization, participants and main results of this second antibody modeling assessment (AMA-II). Also, we compare the results with the first antibody assessment published in this journal (Almagro et al., 2011;79:3050).

Some antibodies have a tendency to self-associate leading to precipitation at relatively low concentrations. CNTO607, a monoclonal antibody, precipitates irreversibly in phosphate-buffered saline at concentrations above 13 mg/ml. Previous mutagenesis work based on the Fab crystal structure pinpointed a three residue fragment in the heavy chain CDR-3, (99)FHW(100a), as an aggregation epitope that is anchored by two salt bridges. Biophysical characterization of variants reveals that F99 and W100a, but not H100, contribute to the intermolecular interaction. A K210T/K215T mutant designed to disrupt the charge interactions in the aggregation model yielded an antibody that does not precipitate but forms reversible aggregates. An isotype change from IgG1 to IgG4 prevents the antibody from precipitating at low concentration yet the solution viscosity is elevated. To further understand the nature of the antibody self-association, studies on the Fab fragment found high solubility but significant self- and cross-interactions remain. Dynamic light scattering data provides evidence for higher order Fab structure at increased concentrations. Our results provide direct support for the aggregation model that CNTO607 precipitation results primarily from the specific interaction of the Fab arms of neighboring antibodies followed by the development of an extensive network of antibodies inducing large-scale aggregation and precipitation.

A blinded study to assess the state of the art in three-dimensional structure modeling of the variable region (Fv) of antibodies was conducted. Nine unpublished high-resolution x-ray Fab crystal structures covering a wide range of antigen-binding site conformations were used as benchmark to compare Fv models generated by four structure prediction methodologies. The methodologies included two homology modeling strategies independently developed by CCG (Chemical Computer Group) and Accerlys Inc, and two fully automated antibody modeling servers: PIGS (Prediction of ImmunoGlobulin Structure), based on the canonical structure model, and Rosetta Antibody Modeling, based on homology modeling and Rosetta structure prediction methodology. The benchmark structure sequences were submitted to Accelrys and CCG and a set of models for each of the nine antibody structures were generated. PIGS and Rosetta models were obtained using the default parameters of the servers. In most cases, we found good agreement between the models and x-ray structures. The average rmsd (root mean square deviation) values calculated over the backbone atoms between the models and structures were fairly consistent, around 1.2 Å. Average rmsd values of the framework and hypervariable loops with canonical structures (L1, L2, L3, H1, and H2) were close to 1.0 Å. H3 prediction yielded rmsd values around 3.0 Å for most of the models. Quality assessment of the models and the relative strengths and weaknesses of the methods are discussed. We hope this initiative will serve as a model of scientific partnership and look forward to future antibody modeling assessments.

To assess the state-of-the-art in antibody structure modeling, a blinded study was conducted. Eleven unpublished Fab crystal structures were used as a benchmark to compare Fv models generated by seven structure prediction methodologies. In the first round, each participant submitted three non-ranked complete Fv models for each target. In the second round, CDR-H3 modeling was performed in the context of the correct environment provided by the crystal structures with CDR-H3 removed. In this report we describe the reference structures and present our assessment of the models. Some of the essential sources of errors in the predictions were traced to the selection of the structure template, both in terms of the CDR canonical structures and VL/VH packing. On top of this, the errors present in the Protein Data Bank structures were sometimes propagated in the current models, which emphasized the need for the curated structural database devoid of errors. Modeling non-canonical structures, including CDR-H3, remains the biggest challenge for antibody structure prediction.

The use of consensus design to produce stable proteins has been applied to numerous structures and classes of proteins. Here, we describe the engineering of novel FN3 domains from two different proteins, namely human fibronectin and human tenascin-C, as potential alternative scaffold biotherapeutics. The resulting FN3 domains were found to be robustly expressed in Escherichia coli, soluble and highly stable, with melting temperatures of 89 and 78°C, respectively. X-ray crystallography was used to confirm that the consensus approach led to a structure consistent with the FN3 design despite having only low-sequence identity to natural FN3 domains. The ability of the Tenascin consensus domain to withstand mutations in the loop regions connecting the β-strands was investigated using alanine scanning mutagenesis demonstrating the potential for randomization in these regions. Finally, rational design was used to produce point mutations that significantly increase the stability of one of the consensus domains. Together our data suggest that consensus FN3 domains have potential utility as alternative scaffold therapeutics.

Microtubule‐associated protein tau becomes abnormally phosphorylated in Alzheimer's disease and other tauopathies and forms aggregates of paired helical filaments (PHF‐tau). AT8 is a PHF‐tau‐specific monoclonal antibody that is a commonly used marker of neuropathology because of its recognition of abnormally phosphorylated tau. Previous reports described the AT8 epitope to include pS202/pT205. Our studies support and extend previous findings by also identifying pS208 as part of the binding epitope. We characterized the phosphoepitope of AT8 through both peptide binding studies and costructures with phosphopeptides. From the cocrystal structure of AT8 Fab with the diphosphorylated (pS202/pT205) peptide, it appeared that an additional phosphorylation at S208 would also be accommodated by AT8. Phosphopeptide binding studies showed that AT8 bound to the triply phosphorylated tau peptide (pS202/pT205/pS208) 30‐fold stronger than to the pS202/pT205 peptide, supporting the role of pS208 in AT8 recognition. We also show that the binding kinetics of the triply phosphorylated peptide pS202/pT205/pS208 was remarkably similar to that of PHF‐tau. The costructure of AT8 Fab with a pS202/pT205/pS208 peptide shows that the interaction interface involves all six CDRs and tau residues 202–209. All three phosphorylation sites are recognized by AT8, with pT205 acting as the anchor. Crystallization of the Fab/peptide complex under acidic conditions shows that CDR‐L2 is prone to unfolding and precludes peptide binding, and may suggest a general instability in the antibody. Proteins 2016; 84:427–434. © 2016 The Authors. Proteins: Structure, Function, and Bioinformatics Published by Wiley Periodicals, Inc.