Enhancing antibody affinity through experimental sampling of non-deleterious CDR mutations predicted by machine learning.

The application of machine learning (ML) models to optimize antibody affinity to an antigen is gaining prominence. Unfortunately, the small and biased nature of the publicly available antibody-antigen interaction datasets makes it challenging to build an ML model that can accurately predict binding affinity changes due to mutations (ΔΔG). Recognizing these inherent limitations, we reformulated the problem to ask whether an ML model capable of classifying deleterious vs non-deleterious mutations can guide antibody affinity maturation in a practical setting. To test this hypothesis, we developed a Random Forest classifier (Antibody Random Forest Classifier or AbRFC) with expert-guided features and integrated it into a computational-experimental workflow. AbRFC effectively predicted non-deleterious mutations on an in-house validation dataset that is free of biases seen in the publicly available training datasets. Furthermore, experimental screening of a limited number of predictions from the model (<10^2 designs) identified affinity-enhancing mutations in two unrelated SARS-CoV-2 antibodies, resulting in constructs with up to 1000-fold increased binding to the SARS-COV-2 RBD. Our findings indicate that accurate prediction and screening of non-deleterious mutations using machine learning offers a powerful approach to improving antibody affinity.

Optimizing Antibody Affinity and Developability Using a Framework-CDR Shuffling Approach-Application to an Anti-SARS-CoV-2 Antibody.

The computational methods used for engineering antibodies for clinical development have undergone a transformation from three-dimensional structure-guided approaches to artificial-intelligence- and machine-learning-based approaches that leverage the large sequence data space of hundreds of millions of antibodies generated by next-generation sequencing (NGS) studies. Building on the wealth of available sequence data, we implemented a computational shuffling approach to antibody components, using the complementarity-determining region (CDR) and the framework region (FWR) to optimize an antibody for improved affinity and developability. This approach uses a set of rules to suitably combine the CDRs and FWRs derived from naturally occurring antibody sequences to engineer an antibody with high affinity and specificity. To illustrate this approach, we selected a representative SARS-CoV-2-neutralizing antibody, H4, which was identified and isolated previously based on the predominant germlines that were employed in a human host to target the SARS-CoV-2-human ACE2 receptor interaction. Compared to screening vast CDR libraries for affinity enhancements, our approach identified fewer than 100 antibody framework-CDR combinations, from which we screened and selected an antibody (CB79) that showed a reduced dissociation rate and improved affinity against the SARS-CoV-2 spike protein (7-fold) when compared to H4. The improved affinity also translated into improved neutralization (>75-fold improvement) of SARS-CoV-2. Our rapid and robust approach for optimizing antibodies from parts without the need for tedious structure-guided CDR optimization will have broad utility for biotechnological applications.

Prediction of the structural interface between fibroblast growth factor23 and Burosumab using alanine scanning and molecular docking.

Burosumab, an FGF23 targeting monoclonal antibody, was approved by the FDA in 2018 for use in children and adults with X-linked hypophosphatemia (or XLH). While several clinical studies have demonstrated the long-term safety and efficacy of Burosumab, the molecular basis of FGF23-Burosumab interaction which underpins its mechanism of action remains unknown. In this study, we employed molecular docking combined with alanine scanning of epitope and paratope to predict a model of FGF23-Burosumab interaction. Then, we used the model to understand the species-species cross-reactivity of Burosumab and to reverse engineer mouse FGF23 with 'back to human' mutations to bind Burosumab. Finally, we redesigned the CDRs with two mutations to engineer an affinity enhanced variant of the antibody. Our study provides insights into the FGF23-Burosumab interaction and demonstrates that alanine-scanning coupled with molecular docking can be used to optimize antibody candidates (e.g., structure-guided affinity maturation) for therapeutic use.

Generation of bispecific antibodies by structure-guided redesign of IgG constant regions.

Bispecific antibodies (BsAbs) form an exciting class of bio-therapeutics owing to their multispecificity. Although numerous formats have been developed, generation of hetero-tetrameric IgG1-like BsAbs having acceptable safety and pharmacokinetics profiles from a single cell culture system remains challenging due to the heterogeneous pairing between the four chains. Herein, we employed a structure-guided approach to engineer mutations in the constant domain interfaces (CH1-CL and CH3-CH3) of heavy and κ light chains to prevent heavy-light mispairing in the antigen binding fragment (Fab) region and heavy-heavy homodimerization in the Fc region. Transient co-transfection of mammalian cells with heavy and light chains of pre-existing antibodies carrying the engineered constant domains generates BsAbs with percentage purity ranging from 78% to 85%. The engineered BsAbs demonstrate simultaneous binding of both antigens, while retaining the thermal stability, Fc-mediated effector properties and FcRn binding properties of the parental antibodies. Importantly, since the variable domains were not modified, the mutations may enable BsAb formation from antibodies belonging to different germline origins and isotypes. The rationally designed mutations reported in this work could serve as a starting point for generating optimized solutions required for large scale production.

Orthogonal immunoassays for IgG antibodies to SARS-CoV-2 antigens reveal that immune response lasts beyond 4 mo post illness onset.

Immune response to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection during the current pandemic remains a field of immense interest and active research worldwide. Although the severity of acute infection may depend on the intensity of innate and adaptive immunity, leading to higher morbidity and mortality, the longevity of IgG antibodies, including neutralizing activity to SARS-CoV-2, is viewed as a key correlate of immune protection. Amid reports and concern that there is a rapid decay of IgG antibody levels within 1 mo to 2 mo after acute infection, we set out to study the pattern and duration of IgG antibody response to various SARS-CoV-2 antigens in asymptomatic and symptomatic patients in a community setting. Herein, we show the correlation of IgG anti-spike protein S1 subunit, receptor binding domain, nucleocapsid, and virus neutralizing antibody titers with each other and with clinical features such as length and severity of COVID-19 illness. More importantly, using orthogonal measurements, we found the IgG titers to persist for more than 4 mo post symptom onset, implying that long-lasting immunity to COVID-19 from infection or vaccination might be observed, as seen with other coronaviruses such as SARS and Middle East respiratory syndrome.

Modified recombinant human erythropoietin with potentially reduced immunogenicity.

Recombinant human erythropoietin (rHuEPO) is a biopharmaceutical drug given to patients who have a low hemoglobin related to chronic kidney disease, cancer or anemia. However, some patients repeatedly receiving rHuEPO develop anti-rHuEPO neutralizing antibodies leading to the development of pure red cell aplasia (PRCA). The immunogenic antibody response activated by rHuEPO is believed to be triggered by T-cells recognizing EPO epitopes bound to MHC molecules displayed on the cell surface of APCs. Previous studies have reported an association between the development of anti-rHuEpo-associated PRCA and the HLA-DRB1*09 gene, which is reported to be entrenched in the Thai population. In this study, we used computational design to screen for immunogenic hotspots recognized by HLA-DRB1*09, and predicted seventeen mutants having anywhere between one through four mutations that reduce affinity for the allele, without disrupting the structural integrity and bioactivity. Five out of seventeen mutants were less immunogenic in vitro while retaining similar or slightly reduced bioactivity than rHuEPO. These engineered proteins could be the potential candidates to treat patients who are rHuEpo-dependent and express the HLA-DRB1*09 allele.

Prediction of the binding interface between monoclonal antibody m102.4 and Nipah attachment glycoprotein using structure-guided alanine scanning and computational docking.

Nipah Virus (NiV) has been designated as a priority disease with an urgent need for therapeutic development by World Health Organization. The monoclonal antibody m102.4 binds to the immunodominant NiV receptor-binding glycoprotein (GP), and potently neutralizes NiV, indicating its potential as a therapeutic agent. Although the co-crystal structure of m102.3, an m102.4 derivative, in complex with the GP of the related Hendra Virus (HeV) has been solved, the structural interaction between m102.4 and NiV is uncharacterized. Herein, we used structure-guided alanine-scanning mutagenesis to map the functional epitope and paratope residues that govern the antigen-antibody interaction. Our results revealed that the binding of m102.4 is mediated predominantly by two residues in the HCDR3 region, which is unusually small for an antibody-antigen interaction. We performed computational docking to generate a structural model of m102.4-NiV interaction. Our model indicates that m102.4 targets the common hydrophobic central cavity and a hydrophilic rim on the GP, as observed for the m102.3-HeV co-crystal, albeit with Fv orientation differences. In summary, our study provides insight into the m102.4-NiV interaction, demonstrating that structure-guided alanine-scanning and computational modeling can serve as the starting point for additional antibody reengineering (e.g. affinity maturation) to generate potential therapeutic candidates.

Molecular basis for dengue virus broad cross-neutralization by humanized monoclonal antibody 513.

Dengue is a widespread viral disease with 3.6 billion people at risk worldwide. Humanized monoclonal antibody (mAb) 513, currently undergoing clinical trials in Singapore, targets an epitope on the envelope protein domain III exposed at the surface of the viral particle. This antibody potently neutralizes all four dengue virus serotypes in a humanized mouse model that recapitulates human dengue infection, without signs of antibody-mediated enhancement of the disease. The crystal structure of single-chain variable fragment (scFv) 513 bound to the envelope protein domain III from dengue virus serotype 4 was used as a template to explore the molecular origins of the broader cross-reactivity and increased in vivo potency of mAb 513, compared to the parent murine mAb 4E11, using molecular dynamics simulations and network analyses. These two methods are a powerful complement to existing structural and binding data and detail specific interactions that underpin the differential binding of the two antibodies. We found that a Glu at position H55 (GluH55) from the second Complementarity Determining Region of the Heavy chain (CDR-H2) which corresponds to Ala in 4E11, is a major contributor to the enhancement in the interactions of mAb 513 compared to 4E11. Importantly, we also validate the importance of GluH55 using site-directed mutagenesis followed by isothermal titration calorimetry measurements.

Rational Engineering and Characterization of an mAb that Neutralizes Zika Virus by Targeting a Mutationally Constrained Quaternary Epitope.

Following the recent emergence of Zika virus (ZIKV), many murine and human neutralizing anti-ZIKV antibodies have been reported. Given the risk of virus escape mutants, engineering antibodies that target mutationally constrained epitopes with therapeutically relevant potencies can be valuable for combating future outbreaks. Here, we applied computational methods to engineer an antibody, ZAb_FLEP, that targets a highly networked and therefore mutationally constrained surface formed by the envelope protein dimer. ZAb_FLEP neutralized a breadth of ZIKV strains and protected mice in distinct in vivo models, including resolving vertical transmission and fetal mortality in infected pregnant mice. Serial passaging of ZIKV in the presence of ZAb_FLEP failed to generate viral escape mutants, suggesting that its epitope is indeed mutationally constrained. A single-particle cryo-EM reconstruction of the Fab-ZIKV complex validated the structural model and revealed insights into ZAb_FLEP's neutralization mechanism. ZAb_FLEP has potential as a therapeutic in future outbreaks.

An inter-residue network model to identify mutational-constrained regions on the Ebola coat glycoprotein.

Recently, progress has been made in the development of vaccines and monoclonal antibody cocktails that target the Ebola coat glycoprotein (GP). Based on the mutation rates for Ebola virus given its natural sequence evolution, these treatment strategies are likely to impose additional selection pressure to drive acquisition of mutations in GP that escape neutralization. Given the high degree of sequence conservation among GP of Ebola viruses, it would be challenging to determine the propensity of acquiring mutations in response to vaccine or treatment with one or a cocktail of monoclonal antibodies. In this study, we analyzed the mutability of each residue using an approach that captures the structural constraints on mutability based on the extent of its inter-residue interaction network within the three-dimensional structure of the trimeric GP. This analysis showed two distinct clusters of highly networked residues along the GP1-GP2 interface, part of which overlapped with epitope surfaces of known neutralizing antibodies. This network approach also permitted us to identify additional residues in the network of the known hotspot residues of different anti-Ebola antibodies that would impact antibody-epitope interactions.

A Structural and Mathematical Modeling Analysis of the Likelihood of Antibody-Dependent Enhancement in Influenza.

Broadly neutralizing monoclonal antibodies (bNAbs) for viral infections, such as HIV, respiratory syncytial virus (RSV), and influenza, are increasingly entering clinical development. For influenza, most neutralizing antibodies target influenza virus hemagglutinin. These bNAbs represent an emerging, promising modality for treatment and prophylaxis of influenza due to their multiple mechanisms of antiviral action and generally safe profile. Preclinical work in other viral diseases, such as dengue, has demonstrated the potential for antibody-based therapies to enhance viral uptake, leading to enhanced viremia and worsening of disease. This phenomenon is referred to as antibody-dependent enhancement (ADE). In the context of influenza, ADE has been used to explain several preclinical and clinical phenomena. Using structural and viral kinetics modeling, we assess the role of ADE in the treatment of influenza with a bNAb.

Glycan-protein interactions in viral pathogenesis.

The surfaces of host cells and viruses are decorated by complex glycans, which play multifaceted roles in the dynamic interplay between the virus and the host including viral entry into host cell, modulation of proteolytic cleavage of viral proteins, recognition and neutralization of virus by host immune system. These roles are mediated by specific multivalent interactions of glycans with their cognate proteins (generally termed as glycan-binding proteins or GBPs or lectins). The advances in tools and technologies to chemically synthesize and structurally characterize glycans and glycan-GBP interactions have offered several insights into the role of glycan-GBP interactions in viral pathogenesis and have presented opportunities to target these interactions for novel antiviral therapeutic or vaccine strategies. This review covers aspects of role of host cell surface glycan receptors and viral surface glycans in viral pathogenesis and offers perspectives on how to employ various analytical tools to target glycan-GBP interactions.

A broadly neutralizing human monoclonal antibody is effective against H7N9.

Emerging strains of influenza represent a significant public health threat with potential pandemic consequences. Of particular concern are the recently emerged H7N9 strains which cause pneumonia with acute respiratory distress syndrome. Estimates are that nearly 80% of hospitalized patients with H7N9 have received intensive care unit support. VIS410, a human antibody, targets a unique conserved epitope on influenza A. We evaluated the efficacy of VIS410 for neutralization of group 2 influenza strains, including H3N2 and H7N9 strains in vitro and in vivo. VIS410, administered at 50 mg/kg, protected DBA mice infected with A/Anhui/2013 (H7N9), resulting in significant survival benefit upon single-dose (-24 h) or double-dose (-12 h, +48 h) administration (P < 0.001). A single dose of VIS410 at 50 mg/kg (-12 h) combined with oseltamivir at 50 mg/kg (-12 h, twice daily for 7 d) in C57BL/6 mice infected with A/Shanghai 2/2013 (H7N9) resulted in significant decreased lung viral load (P = 0.002) and decreased lung cytokine responses for nine of the 11 cytokines measured. Based on these results, we find that VIS410 may be effective either as monotherapy or combined with antivirals in treating H7N9 disease, as well as disease from other influenza strains.

Structure-Guided Design of an Anti-dengue Antibody Directed to a Non-immunodominant Epitope.

Dengue is the most common vector-borne viral disease, causing nearly 400 million infections yearly. Currently there are no approved therapies. Antibody epitopes that elicit weak humoral responses may not be accessible by conventional B cell panning methods. To demonstrate an alternative strategy to generating a therapeutic antibody, we employed a non-immunodominant, but functionally relevant, epitope in domain III of the E protein, and engineered by structure-guided methods an antibody directed to it. The resulting antibody, Ab513, exhibits high-affinity binding to, and broadly neutralizes, multiple genotypes within all four serotypes. To assess therapeutic relevance of Ab513, activity against important human clinical features of dengue was investigated. Ab513 mitigates thrombocytopenia in a humanized mouse model, resolves vascular leakage, reduces viremia to nearly undetectable levels, and protects mice in a maternal transfer model of lethal antibody-mediated enhancement. The results demonstrate that Ab513 may reduce the public health burden from dengue.

Influenza surveillance: 2014-2015 H1N1 "swine"-derived influenza viruses from India.

The 2014-15 H1N1 outbreak in India has reportedly led to 800 fatalities. The reported influenza hemagglutinin sequences from India indicate that these viruses contain amino acid changes linked to enhanced virulence and are potentially antigenically distinct from the current vaccine containing 2009 (Cal0709) H1N1 viral hemagglutinin.

Glycan receptor specificity as a useful tool for characterization and surveillance of influenza A virus.

Influenza A viruses are rapidly evolving pathogens with the potential for novel strains to emerge and result in pandemic outbreaks in humans. Some avian-adapted subtypes have acquired the ability to bind to human glycan receptors and cause severe infections in humans but have yet to adapt to and transmit between humans. The emergence of new avian strains and their ability to infect humans has confounded their distinction from circulating human virus strains through linking receptor specificity to human adaptation. Herein we review the various structural and biochemical analyses of influenza hemagglutinin-glycan receptor interactions. We provide our perspectives on how receptor specificity can be used to monitor evolution of the virus to adapt to human hosts so as to facilitate improved surveillance and pandemic preparedness.

Broadly neutralizing influenza hemagglutinin stem-specific antibody CR8020 targets residues that are prone to escape due to host selection pressure.

Broadly neutralizing antibodies (bNAb) that target a conserved region of a viral antigen hold significant therapeutic promise. CR8020 is a bNAb that targets the stem region of influenza A virus (IAV) hemagglutinin (HA). CR8020 is currently being evaluated for prophylactic use against group 2 IAVs in phase II studies. Structural and computational analyses reported here indicate that CR8020 targets HA residues that are prone to antigenic drift and host selection pressure. Critically, CR8020 escape mutation is seen in certain H7N9 viruses from recent outbreaks. Furthermore, the ability of the bNAb Fc region to effectively engage activating Fcγ receptors (FCγR) is essential for antibody efficacy. In this regard, our data indicate that the membrane could sterically hinder the formation of HA-CR8020-FcγRIIa/HA-IgG-FcγRIIIa ternary complexes. Altogether, our analyses suggest that epitope mutability and accessibility to immune complex assembly are important attributes to consider when evaluating bNAb candidates for clinical development.

Structural determinants for naturally evolving H5N1 hemagglutinin to switch its receptor specificity.

Of the factors governing human-to-human transmission of the highly pathogenic avian-adapted H5N1 virus, the most critical is the acquisition of mutations on the viral hemagglutinin (HA) to "quantitatively switch" its binding from avian to human glycan receptors. Here, we describe a structural framework that outlines a necessary set of H5 HA receptor-binding site (RBS) features required for the H5 HA to quantitatively switch its preference to human receptors. We show here that the same RBS HA mutations that lead to aerosol transmission of A/Vietnam/1203/04 and A/Indonesia/5/05 viruses, when introduced in currently circulating H5N1, do not lead to a quantitative switch in receptor preference. We demonstrate that HAs from circulating clades require as few as a single base pair mutation to quantitatively switch their binding to human receptors. The mutations identified by this study can be used to monitor the emergence of strains having human-to-human transmission potential.

Glycan receptor binding of the influenza A virus H7N9 hemagglutinin.

The advent of H7N9 in early 2013 is of concern for a number of reasons, including its capability to infect humans, the lack of clarity in the etiology of infection, and because the human population does not have pre-existing immunity to the H7 subtype. Earlier sequence analyses of H7N9 hemagglutinin (HA) point to amino acid changes that predicted human receptor binding and impinge on the antigenic characteristics of the HA. Here, we report that the H7N9 HA shows limited binding to human receptors; however, should a single amino acid mutation occur, this would result in structural changes within the receptor binding site that allow for extensive binding to human receptors present in the upper respiratory tract. Furthermore, a subset of the H7N9 HA sequences demarcating coevolving amino acids appears to be in the antigenic regions of H7, which, in turn, could impact effectiveness of the current WHO-recommended prepandemic H7 vaccines.

Antigenically intact hemagglutinin in circulating avian and swine influenza viruses and potential for H3N2 pandemic.

The 2009 swine-origin H1N1 influenza, though antigenically novel to the population at the time, was antigenically similar to the 1918 H1N1 pandemic influenza, and consequently was considered to be "archived" in the swine species before reemerging in humans. Given that the H3N2 is another subtype that currently circulates in the human population and is high on WHO pandemic preparedness list, we assessed the likelihood of reemergence of H3N2 from a non-human host. Using HA sequence features relevant to immune recognition, receptor binding and transmission we have identified several recent H3 strains in avian and swine that present hallmarks of a reemerging virus. IgG polyclonal raised in rabbit with recent seasonal vaccine H3 fail to recognize these swine H3 strains suggesting that existing vaccines may not be effective in protecting against these strains. Vaccine strategies can mitigate risks associated with a potential H3N2 pandemic in humans.