Qualification of a Biolayer Interferometry Assay to Support AZD7442 Resistance Monitoring

Abstract

AZD7442, a combination of two long-acting monoclonal antibodies (tixagevimab [AZD8895] and cilgavimab [AZD1061]), has been authorized for the prevention and treatment of coronavirus disease 2019 (COVID-19). The rapid emergence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants requires methods capable of quickly characterizing resistance to AZD7442. To support AZD7442 resistance monitoring, a biolayer interferometry (BLI) assay was developed to screen the binding of tixagevimab and cilgavimab to SARS-CoV-2 spike proteins to reduce the number of viral variants for neutralization susceptibility verification. Six spike variants were chosen to assess the assay's performance: four with decreased affinity for tixagevimab (F486S:D614G and F486W:D614G proteins) or cilgavimab (S494L:D614G and K444R:D614G proteins) and two reference proteins (wild-type HexaPro and D614G protein). Equilibrium dissociation constant (K_D) values from each spike protein were used to determine shifts in binding affinity. The assay's precision, range, linearity, and limits of quantitation were established. Qualification acceptance criteria determined whether the assay was fit for purpose. By bypassing protein purification, the BLI assay provided increased screening throughput. Although limited correlation between pseudotype neutralization and BLI data (50% inhibitory concentration versus K_D) was observed for full immunoglobulins (IgGs), the correlations for antibody fragments (Fabs) were stronger and reflected a better comparison of antibody binding kinetics with neutralization potency. Therefore, despite strong assay performance characteristics, the use of full IgGs limited the screening utility of the assay; however, the Fab approach warrants further exploration as a rapid, high-throughput variant-screening method for future resistance-monitoring programs. IMPORTANCE SARS-CoV-2 variants harbor multiple substitutions in their spike trimers, potentially leading to breakthrough infections and clinical resistance to immune therapies. For this reason, a BLI assay was developed and qualified to evaluate the reliability of screening SARS-CoV-2 spike trimer variants against anti-SARS-CoV-2 monoclonal antibodies (MAbs) tixagevimab and cilgavimab, the components of AZD7442, prior to in vitro pseudovirus neutralization susceptibility verification testing. The assay bypasses protein purification with rapid assessment of the binding affinity of each MAb for each recombinant protein, potentially providing an efficient preliminary selection step, thus allowing a reduced testing burden in the more technically complex viral neutralization assays. Despite precise and specific measures, an avidity effect associated with MAb binding to the trimer confounded correlation with neutralization potency, negating the assay's utility as a surrogate for neutralizing antibody potency. Improved correlation with Fabs suggests that assay optimization could overcome any avidity limitation, warranting further exploration to support future resistance-monitoring programs.

Keywords: coronavirus disease 2019; monoclonal antibody; receptor binding domain; severe acute respiratory syndrome coronavirus 2; spike protein.

FIG 1

HexaPro spike trimer construct and spike trimer expression assay flow. Light blue shows proline substitutions in trimer construct, and gold shows GSAS site. ACE2, angiotensin-converting enzyme 2; dpt, days posttransfection. Figure created with BioRender.com.

FIG 2

Binding traces for each spike protein in the qualification panel against the two test antibodies. Spike proteins were loaded on the biosensor tips. Indicated concentrations refer to either tixagevimab (left) or cilgavimab (right) in the association phase for each spike protein. The association phase was 180 s, followed by a 600-s dissociation phase. Binding is represented by wavelength shift (Δλ; measured in nanometers [nm]) as detected by the Octet instrument.

FIG 3

Precision represented by K_D values against variant replicates for tixagevimab and cilgavimab. K_D values of <1E−12 M were considered below the LOD. K_D, equilibrium dissociation constant; LOD, limit of detection.

FIG 4

Range and linearity by analyte for tixagevimab and cilgavimab. Dotted lines show lower limit of quantitation and upper limit of quantitation. The ULOQs were 0.3497 nm for tixagevimab and 0.3811 nm for cilgavimab.

FIG 5

Assay specificity. (A) SARS-CoV-2 spike trimers versus non-SARS-CoV-2 anti-RSV 1g7-TM-YTE antibody. RSV antibody used in the association phase with either the wild-type trimer or D614G. (B) SARS-CoV-2 spike trimers versus ICAM-Fc. ICAM-Fc used in the association phase with either the wild-type trimer or D614G. ICAM-Fc started at 1 μM to match concentrations of ACE2-Fc. (C) Nonspecific binding with mock trimer. Mock transfection was used as trimer in the load phase of the assay. Association and dissociation are shown for the four analytes that did not show any evidence of binding. (D) Nonspecific binding with RSV A2 trimer. RSV trimer was used in the load phase of the assay. Association and dissociation are shown for the four analytes that did not show any evidence of binding. (E) CV7 versus mock trimer and RSV A2 trimer. Left, nonspecific binding of CV7 with mock trimer with the corresponding concentrations listed. Right, nonspecific binding of CV7 with RSV trimer with the corresponding concentrations listed. Mock trimer is buffer-exchanged transfection supernatant. Panels show the association (0 to 180 s) and dissociation (180 to 780 s) phase for each test. ICAM, intercellular adhesion molecule; RSV, respiratory syncytial virus; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

FIG 6

Correlation between pseudotyped-virus neutralization assay and BLI assay for tixagevimab and cilgavimab and for tixagevimab Fab and cilgavimab Fab. BLI, biolayer interferometry; Fab, antibody fragment; IC₅₀, 50% inhibitory concentration; K_D, equilibrium dissociation constant; k_d, dissociation constant.