Systems based investigations analyze the impact of SARS-CoV-2 Spike S2 subdomain N-glycans on protein stability and viral infection

Introduction:: The SARS-CoV-2 virus has caused the global pandemic COVID-19. Besides fears related to the appearance of new highlyinfective SARS-CoV-2 variants of concern (VOCs), it is also possible that additional zoonotic coronavirus may cross into humans. What mutations promote this cross-species jump, and what features are conserved among human infecting virions requires a systems-level investigation of coronavirus infection and disease. This understanding is necessary in the path to create new pan-coronavirus antivirals. To address this aspect, we have conducted several system-based studies that demonstrate a critical role for Spike protein N- and O-glycosylation in regulating SARS-CoV-2 viral infectivity [1], and identified molecular chaperones that regulate Spike maturation and viral infectivity [2]. Upon performing bioinformatics analysis of sequencing data, we observed that the SARS-CoV-2 Spike protein S2 subunit has several N-glycosylation sites in its stem region that are conserved in all other human-infecting β-coronavirus, including the original SARS-CoV and middle-eastern MERS virus. Additionally, in the GISAID human sequencing dataset, few mutations have been observed in this highly conserved tail region which includes the dense shield of five uniformly spaced N-glycans in the post-perfusion structure: N1098 (‘S1’), N1134 (‘S2’), N1158 (‘S3’), N1173 (‘S4’) and N1194 (‘S5’) ([3], Fig. 1A). As the role of these glycans is yet unknown, we performed an investigation on the functional roles of these glycosylation sites and determined if these represent critical vulnerabilities of the SARS-CoV-2 virion and other coronaviruses.


Materials and Methods:: The parent Spike containing a D614G mutation was used as the base for cloning 31 Spike N-glycan site-directed mutants that contained Asn (N) to Gln (Q) mutations. This includes: 5 single mutants (deleting glycans at ‘S1’, ‘S2’, ‘S3’, ‘S4’, ‘S5’), 10 double mutants (S12, S13 etc.), 10 triple mutants (S123 etc.), 5 quadruple mutants (containing only one glycan at either S1, S2, S3, S4 or S5) and 1 quintuple mutant lacking all five N-glycosylation sites (Fig. 1A). These Spike variants were transfected into 293T cells, and functional binding to ACE2-Fc recombinant protein and cell surface Spike expression was probed using anti-S1 and anti-S2 antibodies. Intracellular Spike mutant protein distribution was resolved by imaging cytometry. Syncytia formation was quantified with time using Incucyte live-cell imaging system. Viral entry efficiency into host cells was assessed using SARS-CoV-2 virus-like-particles (VLPs) [2]. Besides studies with SARS-CoV-2 virus, similar investigations were also performed with other β-coronavirus virus-like particles that express SARS-CoV and MERS-CoV Spike proteins. These last series of investigations resulted in a panel of additional Spike mutants for trans-complementation assays with the SARS-CoV-2 virus-like-particles.


Results, Conclusions, and Discussions:: Among the mutants, significant reduction of ACE2 binding, and Spike protein expression was observed in specific mutation combinations. These include the double mutants containing S1 mutation (i.e. S12, S13, S14, S15), with S14 and S15 displaying the more profound decrease in expression and function. Additionally, triple mutants containing S14 or S15 mutations (i.e. S124, S125, S134, S135, S145) also displayed weakened binding and protein expression. Finally, quadruple mutants composed of S14 or S15 mutations and the quintuple S12345 mutant lacking all 5 glycans showed an even greater reduction in ACE2-binding and protein expression. In contrast to these observations made with Spike protein containing more than one site mutations, the single mutant alone (S1, S2, S3, S4, S5) exhibited minimal loss of function (Fig. 1B). These data suggest that the loss of single glycans in the Spike stem region may be compensated by others. Similar observations related to glycosylation site-mutations were made in the Syncytia formation assays and viral entry studies performed with VLPs. Additional studies are currently in progress with SARS-CoV and MERS-CoV Spike protein virions to determine if the principles discovered using the SARS-CoV-2 virions are universal across the entire family of human β-coronavirus.

Taken together, this work explored the S2 tail N-glycans and determined their critical role in regulating β-coronavirus infection. Our results suggest that loss of any single glycan may not have an absolute role, but that these glycans act in synergy to stabilize the Spike protein and viral function. This may explain the conservation of the N-glycans in this region under natural selection over the entire β-coronavirus genus. Overall, countermeasures developed against Spike glycans and those directed against the stem region may be valuable drugs both for the ongoing Covid-19 pandemic and possible future emergent virial diseases.


Acknowledgements (Optional): :

References (Optional): : (1) Yang et al., eLife, article: e61552, 2020. (2) Yang et al., Science Advances, 8:eabq8678, 2021. (3) Cai et al., Science 369.6511 (2020):1586-1592.