[论文解读] Review of the mechanisms of SARS-CoV-2 evolution and transmission

Jiahui Chen, Rui Wang+1 more|PubMed|Sep 15, 2021

SARS-CoV-2 and COVID-19 Research参考文献 42被引用 12

一句话总结

本文 identifies infectivity-based natural selection as the primary driver of SARS-CoV-2 evolution, with RBD mutations at residues 452 and 501 significantly enhancing transmissibility and immune evasion. It further reveals a new vaccine-resistant transmission pathway via co-mutations like [Y449S, N501Y], predicting that combinations at these sites could yield variants ten times more infectious and highly resistant to current vaccines and therapies.

ABSTRACT

The mechanism of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) evolution and transmission is elusive and its understanding, a prerequisite to forecast emerging variants, is of paramount importance. SARS-CoV-2 evolution is driven by the mechanisms at molecular and organism scales and regulated by the transmission pathways at the population scale. In this review, we show that infectivity-based natural selection was discovered as the mechanism for SARS-CoV-2 evolution and transmission in July 2020. In April 2021, we proved beyond all doubt that such a natural selection via infectivity-based transmission pathway remained the sole mechanism for SARS-CoV-2 evolution. However, we reveal that antibody-disruptive co-mutations [Y449S, N501Y] on the spike protein receptor-binding domain (RBD) debuted as a new vaccine-resistant transmission pathway of viral evolution in highly vaccinated populations a few months ago. Over one year ago, we foresaw that mutations on RBD residues, 452 and 501, would "both have high chances to mutate into significantly more infectious COVID-19 strains". Mutations on these residues underpin prevailing SARS-CoV-2 variants Alpha, Beta, Gamma, Delta, Epsilon, Theta, Kappa, Lambda, and Mu at present and are expected to be vital to emerging variants in the future. We anticipate that viral evolution will combine RBD co-mutations at these two sites, creating future variants that are about ten times more infectious than the original SARS-CoV-2. Additionally, two complementary transmission pathways of viral evolution, i.e., infectivity and vaccine resistance will prolong our battle with COVID-19 for years. We predict that RBD co-mutation sets [A411S, L452R, T478K], [L452R, T478K, N501Y], [L452R, T478K, E484K, N501Y], [K417N, L452R, T478K], and [P384L, K417N, E484K, N501Y] will have a high chance to grow into dominating variants due to their high infectivity and/or strong ability to break through current vaccines, calling for the development of new vaccines and antibody therapies.

研究动机与目标

• To identify the core mechanisms driving SARS-CoV-2 evolution and transmission at molecular, organismal, and population scales.
• To determine whether infectivity-based natural selection remains the sole evolutionary pathway for SARS-CoV-2, especially in vaccinated populations.
• To forecast high-risk RBD co-mutation sets that could lead to future variants with enhanced transmissibility and immune escape.
• To provide a predictive framework for next-generation vaccine and therapeutic design using deep learning and binding free energy modeling.
• To validate predictions against real-world genomic and experimental data from over 1.77 million SARS-CoV-2 genomes and antibody interaction studies.

提出的方法

• Integration of deep learning and algebraic topology to model binding free energy (BFE) changes induced by RBD mutations.
• Analysis of 506,768 SARS-CoV-2 genomic sequences to assess mutation frequency and BFE correlation.
• Use of experimental deep mutational scanning data to validate predicted BFE changes for ACE2 and antibody binding.
• Application of persistent homology and machine learning to identify co-mutation patterns with high infectivity and immune escape potential.
• Cross-validation of predicted BFE changes against experimental IC50 fold changes and luciferase assay data for key mutations like L452R and N501Y.
• Development of interactive web tools—Mutation Tracker and Mutation Analyzer—for real-time access to genomic and biophysical data.

实验结果

研究问题

• RQ1 What is the dominant evolutionary mechanism driving SARS-CoV-2 transmission, and does it remain consistent across different population immunity levels?
• RQ2 How do RBD mutations at residues 452 and 501 influence viral infectivity and immune escape, and what is their predictive power for emerging variants?
• RQ3 Can co-mutations such as [Y449S, N501Y] enable SARS-CoV-2 to evolve a new, vaccine-resistant transmission pathway in highly vaccinated populations?
• RQ4 Which specific RBD co-mutation sets are most likely to emerge as dominant variants due to enhanced infectivity and immune evasion?
• RQ5 To what extent do predicted binding free energy changes correlate with experimental data on ACE2 and antibody interactions?

主要发现

• Infectivity-based natural selection was confirmed as the sole evolutionary mechanism for SARS-CoV-2 in April 2021, with odds of accidental selection below 1 in 1.27 × 10^30.
• Mutations at RBD residues 452 and 501 were predicted over one year ago to significantly increase infectivity and are now central to Alpha, Beta, Gamma, Delta, Lambda, Mu, and other dominant variants.
• The co-mutation set [Y449S, N501Y] has emerged as a key driver of vaccine resistance in highly vaccinated populations, indicating a new transmission pathway.
• Combinations of RBD mutations at positions 452 and 501 are predicted to generate variants with approximately tenfold higher infectivity than the original SARS-CoV-2 strain.
• Five co-mutation sets—[A411S, L452R, T478K], [L452R, T478K, N501Y], [L452R, T478K, E484K, N501Y], [K417N, L452R, T478K], and [P384L, K417N, E484K, N501Y]—are forecast to have high potential to become dominant due to high infectivity and immune escape.
• Predicted BFE changes for key mutations showed high correlation with experimental data from luciferase assays and IC50 fold changes, validating the model’s predictive accuracy.