Raoul J. de Groot

Human coronavirus HKU1 comprises two distinct serotypes, A and B, whose spike proteins are substantially divergent. Here, we show that spikes from both serotypes preferentially bind 9-O-acetylated α2,8-linked disialosides. Cryo-electron microscopy of the B-type N5 spike reveals a conserved extended binding site in domain S1A that accommodates both the terminal and penultimate sialic acid residues, with interactions involving the penultimate residue substantially enhancing binding. Spike N-glycan processing modulates affinity and linkage selectivity; glycans flanking the binding pocket offer a plausible structural basis for these effects. In contrast to the HKU1-A spike, which adopts open S1B-up conformations upon ligand binding, the N5 apo structure showed that the ligand-binding site was already formed and the e1 relay element register-shifted in most protomers. Nevertheless, we detected neither spontaneous opening nor a transition to an S1B-up state following ligand binding. These findings, obtained with a minimally modified ectodomain, differ from recent reports of ligand-independent opening. Molecular dynamics simulations indicated that membrane-embedded GT3, but not GD3, presents its glycan chain in a geometry compatible with S1A-mediated engagement. Concordantly, in human nasal epithelial cultures cell surface GT3-like O-acetylated trisialoside glycotopes were detected in ciliated cells, linking their cell-type-specific presentation to HKU1 tropism.

The ternary complex, composed of eIF2, GTP and initiator methionyl-tRNA, delivers the first amino acid to the ribosome to initiate protein synthesis. Eukaryotic initiation factor 2B (eIF2B) catalyzes GDP to GTP exchange on eIF2, thereby setting the ternary complex level. Stress-induced phosphorylation converts eIF2 from the substrate of eIF2B into an inhibitor (eIF2-P). This conversion reduces ternary complex levels and induces the integrated stress response (ISR). Here we chart an allosteric axis running through eIF2B, revealing the importance of an α-helix in its β-subunit, the ‘latch-helix’, that hooks onto the α-subunit to induce eIF2B activity. eIF2-P binding promotes latch-helix unhooking, opening eIF2B, which inhibits its activity. Convergently evolved viral proteins stabilize this latch-helix-binding active state of eIF2B. Using these insights, we generated ISR-activating compounds that stabilize eIF2B in its inhibited, unlatched state. Our study thus highlights how long-range eIF2B allostery can be pharmacologically manipulated to sustain or attenuate the ISR.

Coronavirus spike proteins mediate receptor binding and membrane fusion, making them prime targets for neutralizing antibodies. In the cases of severe acute respiratory syndrome coronavirus, severe acute respiratory syndrome coronavirus 2 and Middle East respiratory syndrome coronavirus, spike proteins transition freely between open and closed conformations to balance host cell attachment and immune evasion^1–5. Spike opening exposes domain S1^B, allowing it to bind to proteinaceous receptors^6,7, and is also thought to enable protein refolding during membrane fusion^4,5. However, with a single exception, the pre-fusion spike proteins of all other coronaviruses studied so far have been observed exclusively in the closed state. This raises the possibility of regulation, with spike proteins more commonly transitioning to open states in response to specific cues, rather than spontaneously. Here, using cryogenic electron microscopy and molecular dynamics simulations, we show that the spike protein of the common cold human coronavirus HKU1 undergoes local and long-range conformational changes after binding a sialoglycan-based primary receptor to domain S1^A. This binding triggers the transition of S1^B domains to the open state through allosteric interdomain crosstalk. Our findings provide detailed insight into coronavirus attachment, with possibilities of dual receptor usage and priming of entry as a means of immune escape.

The nucleocapsid protein N of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) enwraps and condenses the viral genome for packaging but is also an antagonist of the innate antiviral defense. It suppresses the integrated stress response (ISR), purportedly by interacting with stress granule (SG) assembly factors G3BP1 and 2, and inhibits type I interferon responses. To elucidate its mode of action, we systematically deleted and over-expressed distinct regions and domains. We show that N via domain N2b blocks PKR-mediated ISR activation, as measured by suppression of ISR-induced translational arrest and SG formation. N2b mutations that prevent dsRNA binding abrogate these activities also when introduced in the intact N protein. Substitutions reported to block post-translation modifications of N or its interaction with G3BP1/2 did not have a detectable additive effect. In an encephalomyocarditis virus-based infection model, N2b - but not a derivative defective in RNA binding—prevented PKR activation, inhibited β-interferon expression and promoted virus replication. Apparently, SARS-CoV-2 N inhibits innate immunity by sequestering dsRNA to prevent activation of PKR and RIG-I-like receptors. Similar observations were made for the N protein of human coronavirus 229E, suggesting that this may be a general trait conserved among members of other orthocoronavirus (sub)genera.

Coronavirus (CoV) spikes mediate receptor binding and membrane fusion, making them prime targets for neutralising antibodies. In the cases of SARS-CoV, SARS-CoV-2, and MERS-CoV, spikes transition freely between open and closed conformations to balance host cell attachment and immune evasion. The open conformation exposes domain S1 ^B , allowing it to bind to proteinaceous cell surface receptors. It also facilitates protein refolding during spike-mediated membrane fusion. However, with a single exception, the pre-fusion spikes of all other CoVs studied so far have been observed exclusively in the closed state. This raises the possibility of regulation, where spikes more commonly transition to open states in response to specific cues, rather than spontaneously. In our study, using cryo-EM and molecular dynamics simulations, we show that the spike protein of the common cold human coronavirus HKU1 undergoes local and long-range conformational changes upon binding a sialoglycan-based primary receptor to domain S1 ^A . This binding triggers the transition of S1 ^B domains to the open state via allosteric inter-domain cross-talk. Our findings paint a more elaborate picture of CoV attachment, with possibilities of dual receptor usage and priming of entry as a means of immune escape.

Human coronavirus OC43 is a globally circulating common cold virus sustained by recurrent reinfections. How it persists in the population and defies existing herd immunity is unknown. Here we focus on viral glycoprotein S, the target for neutralizing antibodies, and provide an in-depth analysis of its antigenic structure. Neutralizing antibodies are directed to the sialoglycan-receptor binding site in S1_A domain, but, remarkably, also to S1_B. The latter block infection yet do not prevent sialoglycan binding. While two distinct neutralizing S1_B epitopes are readily accessible in the prefusion S trimer, other sites are occluded such that their accessibility must be subject to conformational changes in S during cell-entry. While non-neutralizing antibodies were broadly reactive against a collection of natural OC43 variants, neutralizing antibodies generally displayed restricted binding breadth. Our data provide a structure-based understanding of protective immunity and adaptive evolution for this endemic coronavirus which emerged in humans long before SARS-CoV-2.

The transmission of viruses from animal reservoirs to humans poses major threats to public health. Preparedness for future zoonotic outbreaks requires a fundamental understanding of how viruses of animal origin have adapted to binding to a cell surface component and/or receptor of the new host. Here we report on the specificities of human and animal viruses that engage with O-acetylated sialic acid, which include betacoronaviruses, toroviruses and influenza C and D viruses. Key to these studies was the development of a chemoenzymatic methodology that can provide almost any sialate-acetylation pattern. A collection of O-acetylated sialoglycans was printed as a microarray for the determination of receptor specificity. These studies showed host-specific patterns of receptor recognition and revealed that three distinct human respiratory viruses uniquely bind 9-O-acetylated α2,8-linked disialoside. Immunofluorescence and cell entry studies support that such a glycotope as part of a ganglioside is a functional receptor for human coronaviruses.

Significance: Human coronavirus OC43 arose relatively recently, presumably from a bovine coronavirus spillover. Both viruses use 9- O -acetylated sialoglycans as receptors to which they attach via spike protein S. Another envelope protein, hemagglutinin-esterase (HE), serves as a receptor-destroying enzyme. We demonstrate that HE and S are functionally intertwined and that receptor destruction and receptor binding need to be carefully balanced for efficient (pre)attachment. During early emergence of OC43 this balance was reset, presumably as an adaptation to the human host. Such a two-protein mechanism for dynamic virion attachment is unique among coronaviruses, but reminiscent of that of influenza A viruses. Apparently, general principles fundamental to virion–sialoglycan interactions prompted convergent evolution of two zoonotically-relevant groups of pathogens.

The human betacoronaviruses HKU1 and OC43 (subgenus Embecovirus) arose from separate zoonotic introductions, OC43 relatively recently and HKU1 apparently much longer ago. Embecovirus particles contain two surface projections called spike (S) and haemagglutinin-esterase (HE), with S mediating receptor binding and membrane fusion, and HE acting as a receptor-destroying enzyme. Together, they promote dynamic virion attachment to glycan-based receptors, specifically 9-O-acetylated sialic acid. Here we present the cryo-EM structure of the ~80 kDa, heavily glycosylated HKU1 HE at 3.4 Å resolution. Comparison with existing HE structures reveals a drastically truncated lectin domain, incompatible with sialic acid binding, but with the structure and function of the esterase domain left intact. Cryo-EM and mass spectrometry analysis reveals a putative glycan shield on the now redundant lectin domain. The findings further our insight into the evolution and host adaptation of human embecoviruses, and demonstrate the utility of cryo-EM for studying small, heavily glycosylated proteins.

The type I interferon response is an important innate antiviral pathway. Recognition of viral RNA by RIG-I-like receptors (RLRs) activates a signaling cascade that leads to type I interferon (IFN-α/β) gene transcription. Multiple proteins in this signaling pathway (e.g. RIG-I, MDA5, MAVS, TBK1, IRF3) are regulated by (de)ubiquitination events. Most viruses have evolved mechanisms to counter this antiviral response. The leader protease (L^pro) of foot-and-mouth-disease virus (FMDV) has been recognized to reduce IFN-α/β gene transcription; however, the exact mechanism is unknown. The proteolytic activity of L^pro is vital for releasing itself from the viral polyprotein and for cleaving and degrading specific host cell proteins, such as eIF4G and NF-κB. In addition, L^pro has been demonstrated to have deubiquitination/deISGylation activity. L^pro’s deubiquitination/deISGylation activity and the cleavage/degradation of signaling proteins have both been postulated to be important for reduced IFN-α/β gene transcription. Here, we demonstrate that TBK1, the kinase that phosphorylates and activates the transcription factor IRF3, is cleaved by L^pro in FMDV-infected cells as well as in cells infected with a recombinant EMCV expressing L^pro. In vitro cleavage experiments revealed that L^pro cleaves TBK1 at residues 692–694. We also observed cleavage of MAVS in HeLa cells infected with EMCV-L^pro, but only observed decreasing levels of MAVS in FMDV-infected porcine LFPK αVβ6 cells. We set out to dissect L^pro’s ability to cleave RLR signaling proteins from its deubiquitination/deISGylation activity to determine their relative contributions to the reduction of IFN-α/β gene transcription. The introduction of specific mutations, of which several were based on the recently published structure of L^pro in complex with ISG15, allowed us to identify specific amino acid substitutions that separate the different proteolytic activities of L^pro. Characterization of the effects of these mutations revealed that L^pro’s ability to cleave RLR signaling proteins but not its deubiquitination/deISGylation activity correlates with the reduced IFN-β gene transcription. Outbreaks of the picornavirus foot-and-mouth disease virus (FMDV) have significant consequences for animal health and product safety and place a major economic burden on the global livestock industry. Understanding how this notorious animal pathogen suppresses the antiviral type I interferon (IFN-α/β) response may help to develop countermeasures to control FMDV infections. FMDV suppresses the IFN-α/β response through the activity of its Leader protein (L^pro), a protease that can cleave host cell proteins. L^pro was also shown to have deubiquitinase and deISGylase activity, raising the possibility that L^pro suppresses IFN-α/β by removing ubiquitin and/or ISG15, two posttranslational modifications that can regulate the activation, interactions and localization of (signaling) proteins. Here, we show that TBK1 and MAVS, two signaling proteins that are important for activation of IFN-α/β gene transcription, are cleaved by L^pro. By generating L^pro mutants lacking either of these two activities, we demonstrate that L^pro’s ability to cleave signaling proteins, but not its deubiquitination/deISGylase activity, correlates with suppression of IFN-β gene transcription.