Methods for prediction of HLA antigen presentation have improved substantially over the last years in particular for HLA class II. This achievement is the result of a dedicated development of immunoinformatics methods refined to mine and extract information from complex MS-immunopeptidome data sets. In my talk, I will describe this development, illustrate how analysis and interpretation of MS-immunopeptidome data is reliant on immunoinformatics and demonstrate how the reliability of protein drug immunogenicity assessment can be boosted by the integral use of in-silico and in-vivo experimental techniques.

Simona Saporiti,¹ Chiara Parravicini,¹ Carlo Pergola,² Uliano Guerrini,¹ Mara Rossi,³ Fabio Centola,³ and Ivano Eberini⁴

¹Dipartimento di Scienze Farmacologiche e Biomolecolari, Università degli Studi di Milano, Milano, Italy;
²Analytical Development Biotech and ³Global Analytical Pharmaceutical Science and Innovation, Merck Serono S.p.A., Rome, Italy;
⁴Dipartimento di Scienze Farmacologiche e Biomolecolari & DSRC, Università degli Studi di Milano, Milano, Italy

Currently, monoclonal antibodies (mAbs) are the most used biopharmaceuticals for human therapy. One of the key aspects in their development is the control of effector functions mediated by the interaction between fragment crystallizable (Fc) and Fcg receptors, which is a secondary mechanism of action of biotherapeutics. N-glycosylation at the Fc portion can regulate these mechanisms, and much experimental evidence suggests that modifications of glycosidic chains can affect antibody binding to FcgRIIIa, consequently impacting the immune response. In this work, we try to elucidate via in silico procedures the structural role exhibited by glycans, particularly fucose, in mAb conformational freedom that can potentially affect the receptor recognition. By using adalimumab, a marketed IgG1, as a general template, after rebuilding its three-dimensional (3D) structure through homology modeling approaches, we carried out molecular dynamics simulations of three differently glycosylated species: aglycosylated, afucosylated, and fucosylated antibody. First, the use of AMBER10:EHT forcefield provided by MOE was validated, being a good parametrization method for the glycans used in our study. Then, trajectory analysis showed different dynamical behaviors and pointed out that sugars can influence the overall 3D structure of the antibody. As a result, we propose a putative structural mechanism by which the presence of fucose introduces conformational constraints in the whole antibody and not only in the Fc domain, preventing a conformation suitable for the interaction with the receptor. As secondary evidence, we observed a high flexibility of the antibodies that is translated into an asymmetric behavior of Fab portions shown by all the simulated biopolymers, making the dynamical asymmetry a new, to our knowledge, molecular aspect that may be further investigated. In conclusion, these findings can help understand the contribution of sugars on the structural architecture of mAbs, paving the way to novel strategies of pharmaceutical development.

In this work, we present a method for modeling antibodies and performing pH-dependent conformational sampling, which can enhance property calculations. Structure-based charge descriptors are evaluated for their predictive performance on recently published antibody pI, viscosity, and clearance data. From this, we devised four rules for therapeutic antibody profiling which address developability issues arising from hydrophobicity and charged-based solution behavior, PK, and the ability to enrich for those that are approved by the U.S. Food and Drug Administration. Differences in strategy for optimizing the solution behavior of human IgG1 antibodies versus the IgG2 and IgG4 isotypes and the impact of pH alterations in formulation are discussed.

There is an increasing interest in developing computational methods to predict the developability of antibodies in order to reduce the cost of antibody discovery programmes as well as late stage failures. Among the major causes of attrition in these programmes, poor specificity stands out, in part because it is challenging to assess it experimentally in a high-throughput manner. The availability of large databases of antibody sequences and data about antibody specificity is making it possible to develop computational methods to predict the specificity of antibodies from their amino acid sequences. I will present a novel method to achieve this goal.

Our novel, automated high-throughput engineering platform enables the fast generation of large panels of multi-specific variants (up to 10.000) giving rise to large data sets (more than 100.000 data points). By mining our data sets we were able to extract engineering patterns and to develop AI-based virtual screening workflows to guide the exploration of huge design spaces for multi-specific biologics drug discovery.

During the lead identification and optimization phase, predicting and engineering biophysical properties of therapeutic antibody candidates is an important but challenging task. Properties such as aggregation propensity, folding stability, hydrophobicity, charge distribution, etc. are of particular interest because they may affect manufacturability, storage stability, pharmacokinetics and others. The presentation will focus on a comparison between different in-silico property calculation methods with MOE’s molecular dynamics-supported ensemble protein property method in terms of predictivity.

There is considerable enthusiasm in the pharmaceutical industry towards antibody-based biotherapeutics because they can bind diverse receptors highly selectively and often possess desirable pharmacology. Here, we have reviewed all approved antibody-based biotherapeutics from 1986 to Mid-2020 by gathering a wide range of public domain information on 89 marketed antibody-based biotherapeutics. Our analyses has documented ‘coming of age’ for therapeutic antibodies by revealing major trends in their establishment as the best-selling class of pharmaceuticals. For example, all therapeutic antibodies approved recently are either humanized or fully human, while the earlier ones were mostly chimeric. Before 2010, most therapeutic mAbs were developed to treat cancer and focused on merely fifteen receptors, with CD20 being the most common one. Over the next decade (2010-present), thanks to industrialization of antibody manufacturing technologies, their usage has blossomed to 38 additional targets distributed over 15 different therapeutic areas. Drug developers appear to be solidifying their choices in regard to source of antibodies, formulation, and routes of administration. The drivers for these trends are (a) minimize adverse effects such as immunogenicity; (b) improve the drug efficacy and pharmacology; and (c) improve patient convenience for greater compliance. However, humanized or fully human antibodies are merely 3% less likely to generate ADAs than chimeric ones. In spite of high concentration and greater likelihood for aggregation in vivo because of small volumes, the subcutaneous route of administration is ~4% less likely to generate ADAs than the intravenous one. These learnings should help us develop improved novel biotherapeutics.