Multiplexed Deep Antibody Profiling for Antibody Discovery

Antibodies via binding to Fc receptors or complement, mediate several non-neutralizing functions that can drive disease outcome or vaccine responses. The contribution of the antibody biophysical properties and functions, via binding to Fc receptors and complement, to heterogenous individual outcomes in most diseases is still poorly defined due to current assays being either too sample-intensive or cost-intensive to generate a comprehensive antibody profile across a large number of antigen-specificities. Earlier studies of the humoral response have focused on measuring antibody titer and neutralizing functions of only the main antibody isotypes against only a few well-studied antigens in most infectious disease contexts. We developed a sample-sparing, highly multiplexed antibody profiling platform for deep characterization of a broad set of antigen-specific antibodies. We probed both the Fab and Fc profiles of the antigen-specific antibodies as well as the isotype, subtype, Fc receptor binding, and glycosylation. We used a multiplexed bead-based assay to generate high dimensional datasets which are analyzed with bivariate and multivariate methods to predict disease specific biomarkers. We applied this platform in different disease contexts including COVID-19 and Tuberculosis. We profiled antibodies directed against a broad set of SARS-CoV-2 antigens going beyond the well-studied Spike antigen and its associated proteins. We used multivariate machine learning models that predicted biomarkers associated with recovery in severe disease, and discriminated mortality outcomes in an ICU cohort. This experiment also recognized ‘cross-reactive’ responses with other endemic CoVs in a pre-pandemic healthy cohort and gave an insight into the importance of antibodies targeting non-canonical proteins of SARS-CoV-2 for favorable outcomes in severe COVID-19. Similar experimental approach was also undertaken in the TB context to identify biomarkers to differentiate between active TB and latent TB infection. Potential biomarkers and trends were identified in the overall IgG response as well as specific Fc receptor binding and glycosylation. Varying differences among the actively infected and individuals with a latent infection were seen in the antibody glycosylation, specifically in galactosylation and sialylation. To reduce the cost and increase the automation and applicability of this antibody profiling platform, we also designed an alternative microfluidic approach for sample- and cost-effective immunoassays for biomarker discovery. We demonstrated the potential of a high degree of multiplexing using a combination of bio-printed antigens and alternating PDMS channels of flowing sample and probe volumes.

Date

2022-05-23

Resource Type

Text

Resource Subtype

Thesis

Full item page