A structure-based database of antibody variable domain diversity

The diversity of natural antibodies is limited by the genetic mechanisms that engender diversity and the functional requirements of antigen binding. Using an in vitro-evolved autonomous heavy chain variable domain (V(H)H-RIG), we have investigated the limits of structurally-tolerated diversity in the three complementarity-determining regions and a fourth loop within the third framework region. We determined the X-ray crystal structure of the V(H)H-RIG domain at 1.9A resolution and used it to guide the design of phage-displayed libraries encompassing the four loops. The libraries were subjected to selections for structural stability, and a database of structurally-tolerated sequences was compiled from the sequences of approximately 1000 unique clones. The results reveal that all four loops accommodate significantly greater diversity than is observed in nature. Thus, it appears that most sequence biases in the natural immune repertoire arise from factors other than structural constraints and, consequently, it should be possible to enhance the functions of antibodies significantly through in vitro evolution.     

MoFvAb: Modeling the Fv region of antibodies

Knowledge of the 3-dimensional structure of the antigen-binding region of antibodies enables numerous useful applications regarding the design and development of antibody-based drugs. We present a knowledge-based antibody structure prediction methodology that incorporates concepts that have arisen from an applied antibody engineering environment. The protocol exploits the rich and continuously growing supply of experimentally derived antibody structures available to predict CDR loop conformations and the packing of heavy and light chain quickly and without user intervention. The homology models are refined by a novel antibody-specific approach to adapt and rearrange sidechains based on their chemical environment. The method achieves very competitive all-atom root mean square deviation values in the order of 1.5 Å on different evaluation datasets consisting of both known and previously unpublished antibody crystal structures.     

The effect of variable domain orientation and arrangement on the antigen-binding activity of a recombinant human bispecific diabody

In recent years a variety of recombinant methods have been developed for efficient production of bispecific antibodies (BsAb) in various formats. Bispecific diabody (bDAb), a 55-60 kDa molecule comprising two non-covalently associated cross-over single chain Fv (scFv) polypeptides, represents one of the most promising as well the most straightforward approaches to BsAb production. Here we constructed a bDAb, using two human scFv, 11F8 and A12, directed against the epidermal growth factor receptor (EGFR) and the insulin-like growth factor receptor (IGFR), respectively, as the building blocks. A total of 8 scFv and diabody constructs were prepared comprising the same two variable heavy (V(H)) and variable light (V(L)) chain domains but arranged in different orientations. V(H)/V(L) orientation, i.e., V(H)-linker-V(L) or V(L)-linker-V(H), showed significant effects on the expression and antigen-binding activity of scFv and monospecific diabody of both 11F8 and A12. Further, only 2 out of the 4 possible V(H)/V(L) orientations/arrangements in bDAb construction yielded active products that retain binding activity to both EGFR and IGFR. Both active bDAb preparations retained their original antigen-binding activity after incubation at 37 degrees C in mouse serum for up to 7 days, indicating excellent stability of the constructs. Taken together, our results underscore the importance of identifying/selecting optimal V(H)/V(L) orientation/arrangement for efficient production of active bDAb.     

VH-VL orientation prediction for antibody humanization candidate selection: A case study

Antibody humanization describes the procedure of grafting a non-human antibody's complementarity-determining regions, i.e., the variable loop regions that mediate specific interactions with the antigen, onto a β-sheet framework that is representative of the human variable region germline repertoire, thus reducing the number of potentially antigenic epitopes that might trigger an anti-antibody response. The selection criterion for the so-called acceptor frameworks (one for the heavy and one for the light chain variable region) is traditionally based on sequence similarity. Here, we propose a novel approach that selects acceptor frameworks such that the relative orientation of the 2 variable domains in 3D space, and thereby the geometry of the antigen-binding site, is conserved throughout the process of humanization. The methodology relies on a machine learning-based predictor of antibody variable domain orientation that has recently been shown to improve the quality of antibody homology models. Using data from 3 humanization campaigns, we demonstrate that preselecting humanization variants based on the predicted difference in variable domain orientation with regard to the original antibody leads to subsets of variants with a significant improvement in binding affinity.     

Homology modeling of the Abl-SH3 domain

A tertiary structure model of the Abl-SH3 domain is predicted by using homology modeling techniques coupled to molecular dynamics simulations. Two template proteins were used, Fyn-SH3 and Spc-SH3. The refined model was extensively checked for errors using criteria based on stereochemistry, packing, solvation free-energy, accessible surface areas, and contact analyses. The different checking methods do not totally agree, as each one evaluates a different characteristic of protein structures. Several zones of the protein are more susceptible to incorporating errors. These include residues 13, 15, 35, 39, 45, 46, 50, and 60. An interesting finding is that the measurement of the Cα chirality correlated well with the rest of the criteria, suggesting that this parameter might be a good indicator of correct local conformation. Deviations of more than 4 degrees may be indicative of poor local structure. © 1994 Wiley-Liss, Inc.     

Conserved amino acid networks involved in antibody variable domain interactions

Engineered antibodies are a large and growing class of protein therapeutics comprising both marketed products and many molecules in clinical trials in various disease indications. We investigated naturally conserved networks of amino acids that support antibody V_H and V_L function, with the goal of generating information to assist in the engineering of robust antibody or antibody‐like therapeutics. We generated a large and diverse sequence alignment of V‐class Ig‐folds, of which V_H and V_L domains are family members. To identify conserved amino acid networks, covariations between residues at all possible position pairs were quantified as correlation coefficients (?‐values). We provide rosters of the key conserved amino acid pairs in antibody V_H and V_L domains, for reference and use by the antibody research community. The majority of the most strongly conserved amino acid pairs in V_H and V_L are at or adjacent to the V_H–V_L interface suggesting that the ability to heterodimerize is a constraining feature of antibody evolution. For the V_H domain, but not the V_L domain, residue pairs at the variable‐constant domain interface (V_H–C_H1 interface) are also strongly conserved. The same network of conserved V_H positions involved in interactions with both the V_L and C_H1 domains is found in camelid V_HH domains, which have evolved to lack interactions with V_L and C_H1 domains in their mature structures; however, the amino acids at these positions are different, reflecting their different function. Overall, the data describe naturally occurring amino acid networks in antibody Fv regions that can be referenced when designing antibodies or antibody‐like fragments with the goal of improving their biophysical properties. Proteins 2009. © 2008 Wiley‐Liss, Inc.     

Structure of a shark IgNAR antibody variable domain and modeling of an early-developmental isotype

The new antigen receptor (IgNAR) antibodies from sharks are disulphide bonded dimers of two protein chains, each containing one variable and five constant domains. Three types of IgNAR variable domains have been discovered, with Type 3 appearing early in shark development and being overtaken by the antigen-driven affinity-matured Type 1 and 2 response. Here, we have determined the first structure of a naturally occurring Type 2 IgNAR variable domain, and identified the disulphide bond that links and stabilizes the CDR1 and CDR3 loops. This disulphide bridge locks the CDR3 loop in an "upright" conformation in contrast to other shark antibody structures, where a more lateral configuration is observed. Further, we sought to model the Type 3 isotype based on the crystallographic structure reported here. This modeling indicates (1) that internal Type 3-specific residues combine to pack into a compact immunoglobulin core that supports the CDR loop regions, and (2) that despite apparent low-sequence variability, there is sufficient plasticity in the CDR3 loop to form a conformationally diverse antigen-binding surface.     

Using homology modeling to interrogate binding affinity in neutralization of ricin toxin by a family of single domain antibodies

In this report we investigated, within a group of closely related single domain camelid antibodies (V_HHs), the relationship between binding affinity and neutralizing activity as it pertains to ricin, a fast‐acting toxin and biothreat agent. The V1C7‐like V_HHs (V1C7, V2B9, V2E8, and V5C1) are similar in amino acid sequence, but differ in their binding affinities and toxin‐neutralizing activities. Using the X‐ray crystal structure of V1C7 in complex with ricin's enzymatic subunit (RTA) as a template, Rosetta‐based homology modeling coupled with energetic decomposition led us to predict that a single pairwise interaction between Arg29 on V5C1 and Glu67 on RTA was responsible for the difference in ricin toxin binding affinity between V1C7, a weak neutralizer, and V5C1, a moderate neutralizer. This prediction was borne out experimentally: substitution of Arg for Gly at position 29 enhanced V1C7's binding affinity for ricin, whereas the reverse (ie, Gly for Arg at position 29) diminished V5C1's binding affinity by >10 fold. As expected, the V5C1_R29G mutant was largely devoid of toxin‐neutralizing activity (TNA). However, the TNA of the V1C7_G29R mutant was not correspondingly improved, indicating that in the V1C7 family binding affinity alone does not account for differences in antibody function. V1C7 and V5C1, as well as their respective point mutants, recognized indistinguishable epitopes on RTA, at least at the level of sensitivity afforded by hydrogen‐deuterium mass spectrometry. The results of this study have implications for engineering therapeutic antibodies because they demonstrate that even subtle differences in epitope specificity can account for important differences in antibody function.     

Loopholes and missing links in protein modeling

This paper provides an unbiased comparison of four commercially available programs for loop sampling, Prime, Modeler, ICM, and Sybyl, each of which uses a different modeling protocol. The study assesses the quality of results and examines the relative strengths and weaknesses of each method. The set of loops to be modeled varied in length from 4-12 amino acids. The approaches used for loop modeling can be classified into two methodologies: ab initio loop generation (Modeler and Prime) and database searches (Sybyl and ICM). Comparison of the modeled loops to the native structures was used to determine the accuracy of each method. All of the protocols returned similar results for short loop lengths (four to six residues), but as loop length increased, the quality of the results varied among the programs. Prime generated loops with RMSDs <2.5 A for loops up to 10 residues, while the other three methods met the 2.5 A criteria at seven-residue loops. Additionally, the ability of the software to utilize disulfide bonds and X-ray crystal packing influenced the quality of the results. In the final analysis, the top-ranking loop from each program was rarely the loop with the lowest RMSD with respect to the native template, revealing a weakness in all programs to correctly rank the modeled loops.     

Toward high‐resolution homology modeling of antibody Fv regions and application to antibody–antigen docking

High‐resolution homology models are useful in structure‐based protein engineering applications, especially when a crystallographic structure is unavailable. Here, we report the development and implementation of RosettaAntibody, a protocol for homology modeling of antibody variable regions. The protocol combines comparative modeling of canonical complementarity determining region (CDR) loop conformations and de novo loop modeling of CDR H3 conformation with simultaneous optimization of V_L‐V_H rigid‐body orientation and CDR backbone and side‐chain conformations. The protocol was tested on a benchmark of 54 antibody crystal structures. The median root mean square deviation (rmsd) of the antigen binding pocket comprised of all the CDR residues was 1.5 Å with 80% of the targets having an rmsd lower than 2.0 Å. The median backbone heavy atom global rmsd of the CDR H3 loop prediction was 1.6, 1.9, 2.4, 3.1, and 6.0 Å for very short (4–6 residues), short (7–9), medium (10–11), long (12–14) and very long (17–22) loops, respectively. When the set of ten top‐scoring antibody homology models are used in local ensemble docking to antigen, a moderate‐to‐high accuracy docking prediction was achieved in seven of fifteen targets. This success in computational docking with high‐resolution homology models is encouraging, but challenges still remain in modeling antibody structures for sequences with long H3 loops. This first large‐scale antibody–antigen docking study using homology models reveals the level of “functional accuracy” of these structural models toward protein engineering applications. Proteins 2009; 74:497–514. © 2008 Wiley‐Liss, Inc.     

Biophysical properties of human antibody variable domains

There are great demands on the stability, expression yield and resistance to aggregation of antibody fragments. To untangle intrinsic domain effects from domain interactions, we present first a systematic evaluation of the isolated human immunoglobulin variable heavy (V(H)) and light (V(L)) germline family consensus domains and then a systematic series of V(H)-V(L) combinations in the scFv format. The constructs were evaluated in terms of their expression behavior, oligomeric state in solution and denaturant-induced unfolding equilibria under non-reducing conditions. The seven V(H) and seven V(L) domains represent the consensus sequences of the major human germline subclasses, derived from the Human Combinatorial Antibody Library (HuCAL). The isolated V(H) and V(L) domains with the highest thermodynamic stability and yield of soluble protein were V(H)3 and V(kappa)3, respectively. Similar measurements on all domain combinations in scFv fragments allowed the scFv fragments to be classified according to thermodynamic stability and in vivo folding yield. The scFv fragments containing the variable domain combinations H3kappa3, H1bkappa3, H5kappa3 and H3kappa1 show superior properties concerning yield and stability. Domain interactions diminish the intrinsic differences of the domains. ScFv fragments containing V(lambda) domains show high levels of stability, even though V(lambda) domains are surprisingly unstable by themselves. This is due to a strong interaction with the V(H) domain and depends on the amino acid sequence of the CDR-L3. On the basis of these analyses and model structures, we suggest possibilities for further improvement of the biophysical properties of individual frameworks and give recommendations for library design.     

Evidence for intermolecular domain exchange in the Fab domains of dimer and oligomers of an IgG1 monoclonal antibody

Recombinant protein therapeutics have become increasingly useful in combating human diseases, such as cancer and those of genetic origin. One quality concern for protein therapeutics is the content and the structure of the aggregated proteins in the product, due to the potential immunogenicity of these aggregates. Collective efforts have led to a better understanding of some types of protein aggregates, and have revealed the diversity in the structure and cause of protein aggregation. In this work we used a broad range of analytical techniques to characterize the quinary structure (complexes in which each composing unit maintains native quaternary structure) of the stable non-covalent dimer and oligomers of a monoclonal IgG1λ antibody. The results supported a mechanism of intermolecular domain exchange involving the Fab domains of 2 or more IgG molecules. This mechanism can account for the native-like higher order (secondary, tertiary and disulfide bonding) structure, the stability of the non-covalent multimers, and the previously observed partial loss of the antigen-binding sites without changing the antigen-binding affinity and kinetics of the remaining sites (Luo et al., 2009, mAbs 1:491). Furthermore, the previously observed increase in the apparent affinity to various Fcγ receptors (ibid), which may potentially promote immunogenicity, was also explained by the quinary structure proposed here. Several lines of evidence indicated that the formation of multimers by the mechanism of intermolecular domain exchange took place mostly during expression, not in the purified materials. The findings in this work will advance our knowledge of the mechanisms for aggregation in therapeutic monoclonal antibodies.     

Finding biologically relevant protein domain interactions: conserved binding mode analysis

Proteins evolved through the shuffling of functional domains, and therefore, the same domain can be found in different proteins and species. Interactions between such conserved domains often involve specific, well-determined binding surfaces reflecting their important biological role in a cell. To find biologically relevant interactions we developed a method of systematically comparing and classifying protein domain interactions from the structural data. As a result, a set of conserved binding modes (CBMs) was created using the atomic detail of structure alignment data and the protein domain classification of the Conserved Domain Database. A conserved binding mode is inferred when different members of interacting domain families dock in the same way, such that their structural complexes superimpose well. Such domain interactions with recurring structural themes have greater significance to be biologically relevant, unlike spurious crystal packing interactions. Consequently, this study gives lower and upper bounds on the number of different types of interacting domain pairs in the structure database on the order of 1000-2000. We use CBMs to create domain interaction networks, which highlight functionally significant connections by avoiding many infrequent links between highly connected nodes. The CBMs also constitute a library of docking templates that may be used in molecular modeling to infer the characteristics of an unknown binding surface, just as conserved domains may be used to infer the structure of an unknown protein. The method's ability to sort through and classify large numbers of putative interacting domain pairs is demonstrated on the oligomeric interactions of globins.     

Evaluation of a noncanonical Cys40‐Cys55 disulfide linkage for stabilization of single‐domain antibodies

Incorporation of noncanonical disulfide linkages into single‐domain antibodies (sdAbs) has been shown to enhance thermostability and other properties. Here, we evaluated the effects of introducing a novel disulfide linkage formed between Cys residues at IMGT positions 40 and 55 on the melting temperatures (T _ms), reversibility of thermal unfolding, solubility, and antigen‐binding affinities of three types of sdAbs (V_HH, V_H, and V_L domains). The Cys40‐Cys55 disulfide linkage was tolerated by 9/9 V_HHs, 12/12 V_Hs, and 2/11 V_Ls tested and its formation was confirmed by mass spectrometry. Using circular dichroism, we found that the Cys40‐Cys55 disulfide linkage increased sdAb T _m by an average of 10.0°C (range: 0–21.8°C). However, enhanced thermostability came at the cost of a partial loss of refolding ability upon thermal denaturation as well as, for some sdAbs, significantly decreased solubility and antigen‐binding affinity. Thus, Cys40/Cys55 can be added to the panel of known locations for introducing stabilizing noncanonical disulfide linkages into antibody variable domains, although its effects should be tested empirically for individual sdAbs.     

Comparison of an antibody model with an X-ray structure: The variable fragment of BR96

A model of the BR96 antibody variable regions is compared to two X-ray structures of a BR96–carbohydrate complex, independently determined after the model was built and analyzed. The comparison illustrates the opportunities and limitations of antibody modeling. Encouraging results were obtained for the prediction of single CDR loop conformations and for the outline of the BR96 antigen binding site. The comparison of CDR loop conformations in the two X-ray structures provides a realistic reference frame for the CDR loop predictions. CDR loop prediction accuracy is lower when not only conformational, but also positional criteria are taken into account.     

Mapping of antibody epitopes based on docking and homology modeling

Antibodies are key proteins produced by the immune system to target pathogen proteins termed antigens via specific binding to surface regions called epitopes. Given an antigen and the sequence of an antibody the knowledge of the epitope is critical for the discovery and development of antibody based therapeutics. In this work, we present a computational protocol that uses template-based modeling and docking to predict epitope residues. This protocol is implemented in three major steps. First, a template-based modeling approach is used to build the antibody structures. We tested several options, including generation of models using AlphaFold2. Second, each antibody model is docked to the antigen using the fast Fourier transform (FFT) based docking program PIPER. Attention is given to optimally selecting the docking energy parameters depending on the input data. In particular, the van der Waals energy terms are reduced for modeled antibodies relative to x-ray structures. Finally, ranking of antigen surface residues is produced. The ranking relies on the docking results, that is, how often the residue appears in the docking poses' interface, and also on the energy favorability of the docking pose in question. The method, called PIPER-Map, has been tested on a widely used antibody–antigen docking benchmark. The results show that PIPER-Map improves upon the existing epitope prediction methods. An interesting observation is that epitope prediction accuracy starting from antibody sequence alone does not significantly differ from that of starting from unbound (i.e., separately crystallized) antibody structure.     

Analysis of six protein structures predicted by comparative modeling techniques

The protein structures of six comparative modeling targets were predicted in a procedure that relied on improved energy minimization, without empirical rules, to position all new atoms. The structures of human nucleoside diphosphate kinase NM23-H2, HPr from Mycoplasma capricolum, 2Fe-2S ferredoxin from Haloarcula marismortui, eosinophil-derived neurotoxin (EDN), mouse cellular retinoic acid protein I (CRABP1), and P450eryf were predicted with root mean square deviations on Cα atoms of 0.69, 0.73, 1.11, 1.48, 1.69, and 1.73 Å, respectively, compared to the target crystal structures. These differences increased as the sequence similarity between the target and parent proteins decreased from about 60 to 20% identity. More residues were predicted than form the common region shared by the two crystal structures. In most cases insertions or deletions between the target and the related protein of known structure were not correctly positioned. One two residue insertion in CRABP1 was predicted in the correct conformation, while a nine residue insertion in EDN was predicted in the correct spatial region, although not in the correct conformation. The positions of common cofactors and their binding sites were predicted correctly, even when overall sequence similarity was low. © 1995 Wiley-Liss, Inc.     

Structural modeling of human cardiac sodium channel pore domain

The pore domain of human voltage-dependent cardiac sodium channel Na_v1.5 (hNa_v1.5) is the crucial binding targets for anti-arrhythmics drugs and some local anesthetic drugs but its three-dimensional structure is still lacking. This has affected the detailed studies of the binding features and mechanism of these drugs. In this paper, we present a structural model for open-state pore domain of hNa_v1.5 built using single template ROSETTA-membrane homology modeling with the crystal structure of Na_vMs. The assembled structural models are evaluated by rosettaMP energy and locations of binding sites. The modeled structures of the pore domain of hNa_v1.5 in open state will be helpful to explore molecular mechanism of a state-dependent drug binding and help designing new drugs.     

Effects of variable domain orientation on anti-HER2 single-chain variable fragment antibody expressed in the Escherichia coli cytoplasm

Single-chain variable fragment (scFv) antibodies have great potential for a range of applications including as diagnostic and therapeutic agents. However, production of scFvs is challenging because proper folding and activity depend on the formation of two intrachain disulfide bonds that do not readily form in the cytoplasm of living cells. Functional expression in bacteria therefore involves targeting to the more oxidizing periplasm, but yields in this compartment can be limiting due to secretion bottlenecks and the relatively small volume compared to the cytoplasm. In the present study, we evaluated an anti-HER2 scFv, which is specific for human epidermal growth receptor 2 (HER2) overexpressed in breast cancer, for functional expression in the cytoplasm of Escherichia coli strains BL21(DE3) and SHuffle T7 Express, the latter of which is genetically engineered for cytoplasmic disulfide bond formation. Specifically, we observed much greater solubility and binding activity with SHuffle T7 Express cells, which likely resulted from the more oxidative cytoplasm in this strain background. We also found that SHuffle T7 Express cells were capable of supporting high-level soluble production of anti-HER2 scFvs with intact disulfide bonds independent of variable domain orientation, providing further evidence that SHuffle T7 Express is a promising host for laboratory and preparative expression of functional scFv antibodies.