Study proposes structural explanation for immune evasion effect of SARS-CoV-2 BA.2 variant mutations S371L/F

By Neha MathurReviewed by Aimee MolineuxMar 1 2022

In a recent study posted to the bioRxiv* pre-print server, researchers proposed a structural mechanism to explain immune evasion effects of the two-point mutations, S371L and S371F, in the receptor-binding domain (RBD) of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Omicron variant of concern (VOC).

*Important notice: bioRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.

While assaying individual RBD point mutations of Omicron sub-variants BA.1 and BA.2, Liu et al. and Iketani et al. observed that due to S371L and S371F mutations, these variants escaped from the majority of antibodies targeting all four RBD epitope classes, including those on farther RBD surfaces. 

The existing computational and experimental studies performed on monomeric RBD contradict these observations; more importantly, the known escape mechanisms of the SARS-CoV-2 RBD cannot explain these results.

Further, what instigated the authors to examine these findings was the observation that mutations at the S371 site could produce broad antibody escape without compensating for the SARS-CoV-2 fitness. Thus, they keenly explored this fitness tradeoff that has constrained evolution at this site before the emergence of Omicron.

About the study

In the current study, researchers examined the above-mentioned conflicting datasets and structurally analyzed antibodies assayed by Liu et al. and Iketani et al. 

Based on a perturbation of spike (S) trimer conformational dynamics, not yet used to describe any SARS-CoV-2 escape mutation, they proposed a mechanism to explain the observed broad escape effect for both S371L and S371F point mutations.

The immune evasion effect of S371L/F mutation across antibodies targeting all four epitope classes was observed by Liu et al. and Iketani et al. when assayed in the context of the S trimer. Therefore, the researchers also computed the ratio of antibody epitope accessibility in the S-closed conformation vis-à-vis the S-open conformation. 

Finally, they plotted this metric against the antibody escape across BA.1 and BA.2 mutations.

Findings

They observed that S371L/F-mediated escape was strongly associated with epitope accessibility in the S-closed (3 RBD_down) vs. S-open (1-3 RBD_up) conformational states. Subsequently, Class 1 and 4 antibody epitopes were not accessible in the closed-S state, whereas epitopes of Class 2 and 3 antibodies were accessible. 

Further, the relative closed-state accessibility for each antibody epitope helped predict whether a given antibody was moderately or strongly escaped by S371L/F mutations.

The authors noted that Class 2 and 3 and Class 1 and 4 antibodies were escaped weakly (moderately) and strongly by S371L/F, respectively. Overall, antibodies whose epitopes were concealed in S closed (Class 1 and 4) suffered from greater S371L/F-mediated escape than antibodies whose epitopes were exposed in S-closed conformation.

They also observed that for S371L and S371F, epitope inaccessibility in the RBD_down conformation was associated with greater antibody escape; however, this was not the case with other BA.1 or BA.1/2 escape mutations such as K417N or E484A or the full suite of BA.1/2 mutations. 

More specifically, all Class 1 and 4 antibodies whose epitopes were not accessible in RBD_down, were strongly escaped by S371 mutations. In contrast, Class 2 and 3 antibodies whose epitopes were partially or fully accessible in RBD_down were escaped to a lesser extent by S371 mutations. 

Although the degree of Class 2/3 antibody accessibility in the closed state correlated with escape for S371L, the association was weaker for S371F, suggesting that other factors also contributing to escape were at play, particularly across Class 3 antibodies.

Class 3 antibody epitopes were present in proximity to the N343 glycan (5.4 Å between N-glycan and nearest REGN10987 light chain heavy atom); in the case of S309, these epitopes involved direct contacts with this N-glycan. Additionally, the REGN10987 antibody featured steric constraints on binding in the closed state resulting from clashes between the light chain and the adjacent RBD despite an epitope that was highly accessible. This hypothesis by Liu et al. explains the variance for Class 3 antibody escape via modulation of the N343 glycan.

Even molecular dynamic (MD) simulations have previously identified the N343 glycan as a “gate” or a control element that drives S-opening. 

Further, Liu et al. had previously suggested that S371 mutations interacted with the N343 glycan to confer escape from Class 3 antibodies. The authors also hypothesized that interactions between L/F371 and the N343 glycan further resulted in altered S-closed versus S-open conformational dynamics to provide a broad escape from Class 1 and 4 antibodies whose epitopes were inaccessible in the S-closed conformation. S371L/F may even confer a predominantly “locked” S-closed state and reduce occupancy of the S-open state.

Conclusions 

Overall, the authors of the present study proposed three mechanisms to explain which S371 mutations mediate disturbance of RBD dynamics, including - i) RBD_down-RBD_down interfaces interactions; ii) the N343 glycan, or iii) RBD stability. 

Although the authors strongly supported the structural mechanism involving the N343 glycan; nevertheless, in all these cases, the effect may bear a significant fitness cost rationalizing why SARS-CoV-2 variants with S371 mutations in isolation do not exist. 

Although the S371L/F escape effect exceeds the escape effects of the BA.1 and BA.2 variants against several antibodies, the study findings demonstrated that the observed global dominance and fitness of Omicron is due to a combination of other mutations, such as G339D, S37P, S375F, and T376A S371L/F mutations. 

In the future, further experimental and structural investigations could offer a plausible biological explanation for the profound escape conferred by S371L/F mutations. In addition, these studies should explore mutations in Omicron/BA.1/BA.2 variants, especially those compensating for the S371L/F fitness cost, could be particularly valuable to provide insights into the evolutionary landscape of Omicron.

Previously, Sztain et al. have identified a few compensatory sites, which may be functionally constrained to mutate, as participating in the “glycan gate”-mediated RBD opening process. If therapeutic antibodies target these compensatory sites, such as D405N and R408S, they could prove effective against Omicron and other new SARS-CoV-2 VOCs. 

*Important notice: bioRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.