Engineering therapeutic monoclonal antibodies

Revolutionizing oncology: the role of Artificial Intelligence (AI) as an antibody design, and optimization tools

Antibodies play a crucial role in defending the human body against diseases, including life-threatening conditions like cancer. They mediate immune responses against foreign antigens and, in some cases, self-antigens. Over time, antibody-based technologies have evolved from monoclonal antibodies (mAbs) to chimeric antigen receptor T cells (CAR-T cells), significantly impacting biotechnology, diagnostics, and therapeutics. Although these advancements have enhanced therapeutic interventions, the integration of artificial intelligence (AI) is revolutionizing antibody design and optimization. This review explores recent AI advancements, including large language models (LLMs), diffusion models, and generative AI-based applications, which have transformed antibody discovery by accelerating de novo generation, enhancing immune response precision, and optimizing therapeutic efficacy. Through advanced data analysis, AI enables the prediction and design of antibody sequences, 3D structures, complementarity-determining regions (CDRs), paratopes, epitopes, and antigen-antibody interactions. These AI-powered innovations address longstanding challenges in antibody development, significantly improving speed, specificity, and accuracy in therapeutic design. By integrating computational advancements with biomedical applications, AI is driving next-generation cancer therapies, transforming precision medicine, and enhancing patient outcomes.

Addressing Clinical Challenges in Aberrant Pharmacokinetics of Biologic Therapeutic Drugs: Investigating Sample Processing Procedure in the Immunoassays

Bioanalytical Pharmacokinetics (PK) methods are designed for robust performance under rigorous regulatory compliance requirements to ensure the generated data is reliable and maintains integrity. In a phase 1 dose-finding clinical study, aberrant PK profiles of two co-administered biologics drugs were observed. Unexpectedly, we discovered high fill levels in collection tubes from the majority of samples. This led to the hypothesis that the highly filled tubes might cause difficulty in achieving complete sample thaw and thorough mixing at the time of sample analysis, potentially contributing to the abnormalities observed in the PK dataset. Evaluation of the impact of sample fill levels and processing procedure can be challenging since PK concentrations of study samples were unknown. Therefore, a systematic approach was employed to conduct a thorough examination using mock samples. The data illustrate a correlation between sample thawing and mixing process and the variability in the PK data. The concentrations from properly filled mock samples that underwent complete thawing and mixing showed 100% data reproducibility. In contrast, the concentrations from fully filled mock samples that did not follow the proper procedure showed sample recovery deviating by ± 30% from nominal value and exhibited considerable lack of precision. This data identified the root cause of aberrant PK, justifying revised sample preparation guidance and sample re-assay. Improved sample handling and subsequent reassay resolved the aberrant PK profile issues. In conclusion, this study reiterated that sample handling plays a crucial role in quality and reproducibility of PK data with immunoassays.

Amino acids characterization based on frequency and interaction analysis in human antigen-antibody complexes from Thera-SAbDab

BackgroundAntibodies are composed of light and heavy chains, both of which have constant and variable regions. The diversity, specific binding ability and therapeutic potential of antibodies are determined by hypervariable loops called complementarity-determining regions (CDRs), with the other regions being the framework regions.ObjectiveTo investigate the key amino acid patterns in various antibody regions in the human therapeutic antigen-antibody (Ag-Ab) complexes collected from the Thera-SAbDab database.MethodThe study focuses on identifying the amino acid frequency, diversity index in CDRs, paratope-epitope amino acid interactions, amino acid bond formation frequency, and bond types among selected therapeutic Ag-Ab complexes.ResultsThe results revealed that Ser is highly distributed in the overall light chain CDRs while Gly is highly distributed in the heavy chain CDRs. CDR profiling analysis indicated that the average amino acid diversity in heavy chain CDRs is 60% to 70%, while in the light chain, it is 50% to 60%. Aromatic residues such as Tyr, Trp and Phe are the top contributors to these paratope-epitope interactions in the light and heavy chains. Moreover, we examined the frequency of amino acids in light and heavy chains of Ag-Ab complexes. Importantly, the outcome of this study leverages the in depth analysis on single residues, dipeptides, and tripeptides for the therapeutic Ag-Ab complexes.ConclusionWe conclude that the amino acid frequency and interaction analysis centered on therapeutic Ag-Ab complexes will benefit antibody engineering parameters such as antibody design, optimization, affinity maturation, and overall antibody development.

Discovery and investigation of the truncation of the (GGGGS)n linker and its effect on the productivity of bispecific antibodies expressed in mammalian cells

Protein engineering is a powerful tool for designing or modifying therapeutic proteins for enhanced efficacy, increased safety, reduced immunogenicity, and improved delivery. Fusion proteins are an important group of therapeutic compounds that often require an ideal linker to combine diverse domains to fulfill the desired function. GGGGS [(G4S)n] linkers are commonly used during the engineering of proteins because of their flexibility and resistance to proteases. However, unexpected truncation was observed in the linker of a bispecific antibody, which presented challenges in terms of production and quality. In this work, a bispecific antibody containing 5*G4S was investigated, and the truncation position of the linkers was confirmed. Our investigation revealed that codon optimization, which can overcome the negative influence of a high repetition rate and high GC content in the (G4S)n linker, may reduce the truncation rate from 5-10% to 1-5%. Moreover, the probability of truncation when a shortened 3* or 4*G4S linker was used was much lower than that when a 5*G4S linker was used in mammalian cells. In the case of expressing a bispecific antibody, the bioactivity and purity of the product containing a shorter G4S linker were further investigated and are discussed.

Humanization of Pan-HLA-DR mAb 44H10 Hinges on Critical Residues in the Antibody Framework

Reducing the immunogenicity of animal-derived monoclonal antibodies (mAbs) for use in humans is critical to maximize therapeutic effectiveness and preclude potential adverse events. While traditional humanization methods have primarily focused on grafting antibody Complementarity-Determining Regions (CDRs) on homologous human antibody scaffolds, framework regions can also play essential roles in antigen binding. Here, we describe the humanization of the pan-HLA-DR mAb 44H10, a murine antibody displaying significant involvement of the framework region in antigen binding. Using a structure-guided approach, we identify and restore framework residues that directly interact with the antigen or indirectly modulate antigen binding by shaping the antibody paratope and engineer a humanized antibody with affinity, biophysical profile, and molecular binding basis comparable to that of the parental 44H10 mAb. As a humanized molecule, this antibody holds promise as a scaffold for the development of MHC class II-targeting therapeutics and vaccines.

A Detailed Protocol for Constructing a Human Single-Chain Variable Fragment (scFv) Library and Downstream Screening via Phage Display

The development of monoclonal antibodies (mAbs) represents a significant milestone in both basic research and clinical applications due to their target specificity and versatility in therapeutic and diagnostic applications. The innovative strategy of mAb screening, utilizing phage display, facilitates the in vitro screening of antibodies with high affinity to target antigens. The single-chain variable fragment (scFv) is a subset of mAb derivatives, known for its high binding affinity and smaller size-just one-third of that of human IgG. This report outlines a detailed and comprehensive procedure for constructing a scFv phagemid library derived from human patients, followed by screening via phage display affinity selection. The protocol utilizes 348 primer combinations spanning the entire human antibody repertoire to minimize sequence bias and maintain library diversity during polymerase chain reaction (PCR) for scFv generation, resulting in a library size greater than 1 × 108. Furthermore, we describe a high-throughput phage display screening protocol using enzyme-linked immunosorbent assay (ELISA) to evaluate more than 1200 scFv candidates. The generation of a highly diverse scFv library, coupled with the implementation of a phage display screening methodology, is expected to provide a valuable resource for researchers in pursuit of scFvs with high affinity for target antigens, thus advancing both research and clinical endeavors.