Applying Computational Systems Serology to Unravel the Heterogeneity of Antibody-Mediated Immunity in Infectious and Autoimmune Diseases

Abstract

Antibodies are a vital part of the immune system that provide protection against infectious diseases, but they have also been implicated in the development of autoimmune disease. While the fragment antigen-binding (Fab) domain of the antibody plays an important role in neutralization, the fragment crystallizable (Fc) region of the antibody binds to Fc receptors on innate immune cells to activate cellular effector functions, including antibody-dependent cellular toxicity (ADCC) and antibody-dependent cellular phagocytosis (ADCP). These functions are important for protection in infectious disease, but overactivation can lead to autoimmunity. Many different antibody and Fc receptor features contribute to these processes and thus it is difficult to unravel the role of each in ADCC and ADCP. As an additional layer of complexity, these features are highly variable across individuals, depending on personalized genetic and environmental factors. Newly developed systems serology approaches offer the opportunity to gain unique quantitative insight into these complex systems in both infectious and autoimmune diseases. In this work, we use systems serology to study the role of Fc effector functions in an autoimmune disease (Sjogren’s syndrome) and in two different infectious diseases (SARS-CoV-2 and HIV). We show how these approaches can identify novel biomarkers, enable better understanding of disease heterogeneity, and provide new insights that may guide the development of diagnostic tools and therapeutic strategies.

Using data-driven systems serology approaches in Sjogren’s syndrome, an autoimmune disease that causes dry mouth and dry eye, we found antibody/FcR signatures involving FcγRIIa and FcγRIIIa responses to classical antigens (such as Ro and La) were elevated in Sjogren’s patients. We also found that non-classical antigen signatures were elevated in non-Sjogren’s Sicca patients, as well as seronegative Sjogren’s patients who had lower levels of classical antibody responses. Overall, this suggests a new mechanism whereby FcR activation via both classical and non-classical autoantigen binding could play a role in disease pathogenesis. These results provide important insight into biomarkers that could be used for diagnosis. In SARS-CoV-2, we were able to identify cross-reactive SARS-CoV-2 antibody signatures that distinguished between pre-pandemic healthy children and elderly, providing the novel insight that elderly had more cross-reactive responses to human coronaviruses, while children had more targeted, SARS-CoV-2-specific responses. Furthermore, data-driven approaches identified an antibody signature that distinguished between a small sample size of COVID-19 infected individuals and healthy controls, driven by elevated FcγRIIIa responses. These results provided some of the first evidence to help understand differences in disease severity in younger children versus older adults early in the COVID-19 pandemic. Lastly, we employed mechanistic ordinary differential equation (ODE) models to gain insight into Fc effector function in HIV. A new model with two different FcR types was able to help us explore the complex dynamics between FcγRIIa and FcγRIIIa activation, depending on complex balances of antibody subclass concentrations and binding affinity related to host genetic background. The model revealed optimal combinations of IgG1 and IgG3 concentrations for maximizing ADCP and ADCC respectively. Surprisingly, the model also revealed a new mechanism whereby increases in IgG3 could decrease ADCP. Simulations specific to different tissues indicated that that host genetics (Fc receptor polymorphisms) may influence responses in the blood but are not expected to contribute to differences in mucosal tissues. Overall, these approaches provide new knowledge that could help guide future vaccines and therapeutic interventions, specific to desirable Fc effector functions.