Revealing the Threat of Emerging SARS-CoV-2 Mutations to Antibody Therapies

The coronavirus disease (COVID-19) is caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Since its onset, the virus rapidly spread worldwide, causing a global pandemic that has devastated the global economy and claimed over 6 million lives. Besides the stringent measures enacted to control and manage the spread of the virus, developing vaccines and antibody therapies are considered the most effective and sustainable means of ending the COVID-19 pandemic. Indeed, COVID-19 vaccines have been developed and are currently in use with billions of doses already administered. Unfortunately, the emerging SARS-CoV-2 variants present a significant threat to the efficacy of the existing monoclonal antibody (mAb) therapies and vaccines.

Most COVID-19 drugs and vaccines are designed to improve host immune response to the invading virus by enhancing the binding between the angiotensin-converting enzyme 2 (ACE2) receptor of the host and the spike (S) protein receptor-binding domain (RBD) of the virus. However, mutations of the S protein might compromise the binding and further regrade the efficiency of the drugs and vaccines. To date, SARS-CoV-2 S protein has produced tens of thousands of unique mutations with over 95% of the spike protein sequence has undergone the mutations since the outbreak of the COVID-19 pandemic disease, raising concerns about their potential impact on existing medical interventions.

Recently, the binding free energy (BFE) between S protein RBD and the ACE2 has been reported to be proportional to the infectivity of the various COVID-19 variants. Thus, the infectivity of various SARS-CoV-2 variants and the impacts of the mutations on antibodies and vaccines can be estimated by determining their BFE changes. Despite numerous experimental studies on the threats of emerging COVID-19 strains, the studies are mostly limited to a few well-known and frequently observed mutations and mAbs therapies like Regeneron and Eli Lilly. Therefore, extending the current studies to a globally vast number of other mutations and many mAbs at different stages of drug development is of great importance. Predicting the protein-protein interaction BFE via machine learning methods can provide reliable and consistent results.

Herein, scientists at Michigan State University:Jiahui Chen, Kaifu Gao, Rui Wang and Guo-Wei Wei studied the potential threat of emerging SARS-CoV-2 mutations to different antibody drug candidates by using topology-based deep learning models to determine the mutation-induced BFE changes. A total of 796,759 genome isolates from coronavirus patients were analyzed to identify 606 non-degenerate RBD mutations. The impact of the identified mutations on 16 mAbs currently in clinical trials was investigated. The obtained results were validated with thousands of existing experimental data. Their research work is currently published in the Journal of Molecular Biology.

The authors’ findings revealed that most observed mutations like F490L/S, V483F/A, F486L, S494P, R346K/S, L455F, Q493L, N439K and G446V may compromise the efficiency and efficacy of some mAbs in clinical trials. Despite the lack of experimental data, the authors predicted that the efficacy of Celltrion antibody therapy regdanvimab (now FDA emergency approved) could be degraded by variants P.1, B1.351, B.1.427, and B.1.526 and mutations L455F, E484A, F490L/S, and S494P/L. Additionally, the Rockefeller University antibody C144 was highly susceptible to mutation E484Q/A and variants P.1, B1.351, and B.1.526, while antibody C135 was less prone to mutations R346K/S. Furthermore, the low-frequency RBD mutations emerged as potential candidates for future vaccines or antibody escape variants because they can weaken the binding between European Medicines agency (EMA) and most antibodies while enhancing RBD binding to ACE2. The findings were in good agreement with existing experimental data on well-known SARS-CoV-2 variants like Alpha, Beta, Delta, Epsilon and Kappa variants.

In summary, the study provided an in-depth look into the potential threats of 100 frequent mutations to mAbs, including Regeneron, Eli Lilly and many others currently on clinical trials. The variants and mutations can reduce antibody neutralization effects of the immune systems or antibody treatments and increase the difficulty of predicting the transmissibility or diagnosing the virus. It is worth noting that many RBD mutations globally pose threats to current and future antibody therapies and vaccines. “Our  topology-based deep learning models have been applied to the accurate prediction of the new Omicron variant’s infectivity, vaccine breakthrough and antibody resistance and  the development of more effective vaccines and antibody drugs for treating coronavirus disease’’, said Professor Guo-Wei Wei.