SARS-CoV-2 Omicron BA.2.86: less neutralization evasion compared to XBB sub-variants (preprint)

The continual emergence and circulation of new severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) variants have caused a great challenge for the coronavirus disease 2019 (COVID-19) pandemic control. Recently, Omicron BA.2.86 was identified with more than 30 amino acid changes on the spike (S) protein, compared to Omicron BA.2 or XBB.1.5. The immune evasion potential of BA.2.86 is of great concern. In this study, we evaluated the neutralizing activities of sera collected from participants and mice. Participants were divided into five groups according to their vaccination (inactivated vaccine, protein subunit vaccine ZF2001 or ZF2202-A) and infection (Omicron BF.7/BA.5.2) status. ZF2202-A is ZF2001 vaccine's next-generation COVID-19 vaccine with updated bivalent Delta-BA.5 RBD-heterodimer immunogen. BALB/c mice were immunized with XBB.1.5 RBD-homodimer, BA.5-BA.2, Delta-XBB.1.5 or BQ.1.1-XBB.1.5 RBD-heterodimers protein vaccine candidates for evaluating the neutralizing responses. We found that Omicron BA.2.86 shows stronger immune evasion than BA.2 due to >30 additional mutations on S protein. Compared to XBB sub-variants, BA.2.86 does not display more resistance to the neutralizing responses induced by ZF2001-vaccination, BF.7/BA.5.2 breakthrough infection or a booster dose of ZF2202-A-vaccination. In addition, the mouse experiment results showed that BQ.1.1-XBB.1.5 RBD-heterodimer and XBB.1.5 RBD-homodimer induced high neutralizing responses against XBB sub-variants and BA.2.86, indicating that next-generation COVID-19 vaccine should be developed to enhance the protection efficacy against the circulating strains in the future.

De novo design of site-specific protein interactions with learned surface fingerprints (preprint)

Physical interactions between proteins are essential for most biological processes governing life. However, the molecular determinants of such interactions have been challenging to understand, even as genomic, proteomic, and structural data grows. This knowledge gap has been a major obstacle for the comprehensive understanding of cellular protein-protein interaction (PPI) networks and for the de novo design of protein binders that are crucial for synthetic biology and translational applications. We exploit a geometric deep learning framework operating on protein surfaces that generates fingerprints to describe geometric and chemical features critical to drive PPIs. We hypothesized these fingerprints capture the key aspects of molecular recognition that represent a new paradigm in the computational design of novel protein interactions. As a proof-of-principle, we computationally designed four de novo protein binders to engage three protein targets: SARS-CoV-2 spike, PD-1, and PD-L1. The designs bound the target sites with nanomolar affinity upon experimental optimization, structural and mutational characterization showed highly accurate predictions. Overall, our surface-centric approach captures the physical and chemical determinants of molecular recognition, enabling a novel approach for the de novo design of protein interactions and, more broadly, of artificial proteins with function.