Assessing developability early in the discovery process for novel biologics

Abstract

Beyond potency, a good developability profile is a key attribute of a biological drug. Selecting and screening for such attributes early in the drug development process can save resources and avoid costly late-stage failures. Here, we review some of the most important developability properties that can be assessed early on for biologics. These include the influence of the source of the biologic, its biophysical and pharmacokinetic properties, and how well it can be expressed recombinantly. We furthermore present in silico, in vitro, and in vivo methods and techniques that can be exploited at different stages of the discovery process to identify molecules with liabilities and thereby facilitate the selection of the most optimal drug leads. Finally, we reflect on the most relevant developability parameters for injectable versus orally delivered biologics and provide an outlook toward what general trends are expected to rise in the development of biologics.

Keywords: Biologics; antibodies; biotherapeutics; developability; drug development; drug discovery; drug properties; half-life; immunogenicity; manufacturability.

Figure 1.

Schematic overview of antibody sources, discovery strategies, in vitro assays, and in silico methods used to assess developability properties during the development of biologics. Antibodies can be of either natural or synthetic origin and are typically discovered using various display technologies or immunization strategies, or a combination of these. In addition, antibodies can be derived from patient populations. During the development of antibodies, various in vitro assays in combination with in silico methods for machine learning and molecular dynamics can be applied to assess the antibodies’ developability properties and select candidates with the best developability profile.

Figure 2.

Knowledge and physics-based approaches for characterizing biophysical properties of antibodies. Left panel, knowledge-based: Overview of critical steps for sequence-based in silico prediction of biophysical properties from several thousands of potential hit sequences. Right panel, physics-based: A) The antibody binding interface exists as an ensemble of conformations, which includes binding competent as well as non-binding states. Partially unfolded conformations also exist with a lower probability. B) Different conformations exhibit different properties, where partially unfolded conformations may aggregate which leads to further unfolding. In C) and D), the hydrophobicity profile of two different conformations of the TNF-α binding antibody golimumab is mapped on its molecular surface using localized free energy of hydration. The two conformations show a significantly altered hydrophobicity profile and will therefore most likely interact differently with other hydrophobic molecules.

Figure 3.

Schematic illustration of a Peripheral Blood Mononuclear Cell (PBMC) immunogenicity assay. A) PBMCs are isolated from healthy donors. B) Isolated cells are cultured in cell media with added candidate biologics. C) Candidate biologics are taken up by antigen presenting cells (APCs) and presented to T cells in culture. If the biologic is immunogenic, this may lead to T cell activation and concurrent cytokine secretion. D) Immunogenicity assessment can then be performed by phenotyping the T cells following stimulation with candidate biologics and measuring cytokine levels in culture supernatant.

Figure 4.

Human Endothelial Recycling Assay (HERA) as a tool for in vitro pharmacokinetic assessment and addressing FcRn-targeting strategies. A) Generalized HERA protocol. (1) Stably FcRn-transfected human microvascular endothelial cells (HMEC)-1 are seeded, prior to (2) adding FcRn-binding candidate biologics to two parallel cell plates. Following an incubation period, (3) cells from one plate are lysed to obtain an uptake sample. For the other plate, the media is exchanged to recycling medium, and after another incubation period, (4) the medium is harvested as a recycling sample, and (5) the cells are lysed to obtain a residual sample. Candidate biologics in all samples are quantified by an ELISA tailored for specific detection of the assessed biologic. B) Variations of the HERA protocol, enabling analysis of both FcRn-dependent and -independent uptake, cellular accumulation, and FcRn-dependent recycling. Variations include (1) performing the uptake step at mildly acidic extracellular pH, effectively forcing intracellular accumulation of biologics by preventing FcRn-mediated recycling, (2) manipulating FcRn-expression or blocking binding to FcRn to analyze FcRn-dependent and -independent cellular accumulation, and (3) introducing competition for FcRn binding to mirror endogenous competition on the ligand-binding sites of FcRn and its effects on the cellular transport for FcRn-binding biologics. C) HERA data can be used to address the impact of structural design of candidate biologics on FcRn-mediated cellular transport and unspecific cellular accumulation. For some candidate biologics, HERA data may allow for calculation of a score that correlates with plasma half-life in human FcRn-transgenic mice.