Subscriber access provided by University of Sunderland
Article
Developability assessment of engineered monoclonal antibody variants with a complex self-association behavior using complementary analytical and in silico tools Lu Shan, Neil Mody, Pietro Sormanni, Kim Rosenthal, Melissa Damschroder, and Reza Esfandiary Mol. Pharmaceutics, Just Accepted Manuscript • DOI: 10.1021/acs.molpharmaceut.8b00867 • Publication Date (Web): 05 Nov 2018 Downloaded from http://pubs.acs.org on November 6, 2018
Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.
Page 1 of 44 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
Developability assessment of engineered monoclonal antibody variants with a complex selfassociation behavior using complementary analytical and in silico tools Lu Shan1, Neil Mody2, Pietro Sormani3, Kim L. Rosenthal1, Melissa M. Damschroder1, Reza Esfandiary2 Departments of 1Antibody Discovery and Protein Engineering and 2Formulation Sciences, MedImmune, Gaithersburg, MD, USA; 3Centre for Misfolding Diseases, Department of Chemistry, University of Cambridge, Cambridge, UK KEYWORDS: monoclonal antibody, reversible self-association, antibody engineering, antibody developability, viscosity, AC-SINS, DLS, homology modeling
ACS Paragon Plus Environment
1
Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 2 of 44
Abstract Monoclonal antibodies (mAbs) are complex molecular structures. They are often prone to development challenges particularly at high concentrations due to undesired solution properties such as reversible self-association, high viscosity, and liquid-liquid phase separation. In addition to formulation optimization, applying protein engineering can provide an alternative mitigation strategy. Protein engineering during the discovery phase can provide great benefits to optimize molecular properties, resulting in improved developability profiles. Here, we present a case study utilizing complementary analytical and predictive in silico methods. We have systematically identified and reengineered problematic residues responsible for the self-association of a model mAb, driven by a complex combination of hydrophobic and electrostatic interactions. Noteworthy findings include a more dominant contribution of hydrophobic interactions to self-association and potential feasibility of mutations in the CDR regions to mitigate self-association. The engineered mutation panel enabled us to assess potential correlations among commonly utilized developability screening assays, including affinity capture self-interaction nanospectroscopy (AC-SINS), dynamic light scattering (DLS), and apparent solubility by PEG-precipitation. In addition, we evaluated the correlations between experimental measurements and computational predictions. CamSol, an in silico computational tool that accounts for complex molecular interactions and neighboring hotspots, was found to be an effective screening tool. Our work led to reengineered mAb variants, better suited for high-concentration liquid formulation development. The engineered mAbs exhibited enhanced in vitro and simulated in vivo solubility and reduced selfassociation propensity, while maintaining binding affinity and thermal stability.
ACS Paragon Plus Environment
2
Page 3 of 44 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
Introduction Monoclonal antibodies (mAbs) currently represent the fastest-growing class of biologics, with a rapid rate of six to nine first marketing approvals per year.1 With high specificity and affinity toward a wide range of disease targets and with a sustainable serum half-life, mAbs provide muchneeded therapeutic opportunities to patients in need.2 Although intravenous administration has been the most conventional route of delivery for mAbs, use of self-administered devices such as prefilled syringes and auto-injectors have become increasingly popular.3 This is particularly the case for treatment of chronic diseases. The limited subcutaneous injection volumes ( 0.99 (Figure S2 in Supporting Information). Overall, the results show that mAb-J parental antibody exhibits an apparent solubility of about 5 mg/mL. The Fab mutants, particularly those at VH positions W99 and K98, and VL position F52 resulted in improved apparent solubility in the range of 10-16 mg/mL. Analysis of the Fc CH3 mutants showed improvements in apparent solubility of mAb-J as well, although to a more moderate degree (Figure 5B and Table S1 in Supporting Information). More in-depth evaluation of the data further reveals the specificity of certain mutations with regard to their impact on apparent solubility. For example, mutation of VH W99 to asparagine improved the solubility by twofold, while the W99R mutant generated a fourfold improvement in solubility. Similarly, in the case of VL F52, mutation to aspartic acid improved the solubility by twofold, while mutation to serine resulted in fourfold improvements. Overall, the results from PEG solubility screening showed a range of impacts from the designed mutants to improve apparent solubility of mAb-J.
ACS Paragon Plus Environment
21
Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 22 of 44
Figure 5. Apparent solubility of mAb-J variants in PEG precipitation assay. (A) Representative log plot for extrapolating solubility calculations. (B) Apparent solubility of most mutants gave moderate improvement compared to mAb-J’s apparent solubility (mg/mL), while H:K98A, H:W99A, H:W99R, and L:F52S resulted in up to four-fold improvement in apparent solubility.
In vitro bioavailability To better understand the impact of the mutations on RSA behavior of the engineered mAbs in the simulated subcutaneous space, analysis with the SCISSOR system was performed. The study was conducted at a high concentration of 75 mg/mL under experimental conditions consistent with those outlined by Kinnunen et al.36 This in vitro screening technique simulates the stress conditions and environmental transitions that a protein experiences upon injection into the subcutaneous environment, often at high concentrations. Therefore it may be used as a tool to assess the behavior of proteins immediately after administration. For the in vitro bioavailability screening studies described here, carbonate buffer and 10 mg/mL hyaluronic acid were identified as the optimal interstitial fluid and extracellular matrix components, respectively. This is in agreement with the previous published work.36
ACS Paragon Plus Environment
22
Page 23 of 44 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Molecular Pharmaceutics
The results yielded significant differentiation with regard to the visual appearance and physical stability of the wild-type mAb-J compared to its examined mutant, W99R in the VH region. Due to the larger material requirements for the SCISSOR experiment, only the W99R, one of the most effective mutants identified by DLS, PEG precipitation solubility studies, and AC-SINS, was examined. The results show that significant precipitation was observed upon injection of the wildtype mAb-J. The W99R mutant, however, resulted in a much clearer solution in the simulated subcutaneous environment (Figure 6A). In quantifying this observation, the transmission profile of the wild-type mAb-J cartridge (Figure 6B) showed a sharp decrease and approximately 86% change in transmission through the channel (channel 4). This channel corresponds to the bottom portion of the cartridge. This is consistent with the observed onset of turbidity and subsequent precipitation event post injection into the simulated subcutaneous space. The W99R mutant showed a much more moderate reduction (approximately 29%) in channel 4 transmission. This suggests that solution clarity was mostly maintained with no significant precipitation events observed. The total protein (measured as total diffused fraction percentage) versus time plot shows that the wild-type mAb-J resulted in a slower and reduced total diffused fraction compared to the W99R mutant (Figure 6C). This may suggest that in the parental antibody, there was incomplete diffusion out of the cartridge due to an aggregation and/or precipitation event perhaps caused by incompatibility with the interstitial space components, or immediate loss of excipient protection. This resulted in enhanced conditions to propagate protein-protein self-association upon injection into the simulated subcutaneous space. The SCISSOR results here highlight the potential correlation between the level of the native protein-protein interactions observed in vitro and the pharmacokinetic and pharmacodynamic properties in vivo upon clinical administration.
ACS Paragon Plus Environment
23
Molecular Pharmaceutics 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60
Page 24 of 44
Figure 6. In vitro bioavailability studies using a subcutaneous injection site simulator instrument (SCISSOR). The wild-type mAb-J and mutant W99R were injected into a subcutaneous simulation chamber at a concentration of 75 mg/mL. The visual inspection of the difference of the mAbs in the interstitial fluid is shown in (A). The transmission (%) readout at the bottom of the chambers (channel 4) was recorded over 25 hours (B), and the drug diffusion fraction (%) into the interstitial fluid was calculated over time (C).
In silico modeling and predictions of developability To gain more insight into the impact of mutations on mAb-J solubility, computational tools were utilized to dissect the possible contributing factors. Homology models of the parental mAb-J and of its 12 mutants were generated by MOE and Discovery Studios. In order to evaluate correlations over a wider solubility range, we also considered those variants that had precipitated in the experimental buffer (W99D, K100aA and K100aD), and assigned them an apparent solubility value of 0.9) when the pI of all variants considered was far from the buffer pH.29 This is in agreement with the very high correlation we find after excluding mutations to negatively charged residues (R = 0.93, Figure 7). Developability Index calculations yielded a poorer correlation with apparent solubility measurements of the Fab mutants (Pearson’s R = -0.5, Figure S4 in Supporting Information). However, the correlation can be increased to an R value of -0.84 (p
