The SARS-CoV-2 spike protein may act as a superantigen kicking the immune system into overdrive as in toxic shock syndrome

A study from the University of Pittsburgh School of Medicine and Cedars-Sinai addresses a mystery first raised in March: Why do some people with COVID-19 develop severe inflammation?

The research shows how the molecular structure and sequence of the SARS-CoV-2 spike protein – part of the virus that causes COVID-19 – could be behind the inflammatory syndrome cropping up in infected patients.

The study, published this week in the Proceedings of the National Academy of Sciences, uses computational modeling to zero in on a part of the SARS-CoV-2 spike protein that may act as a “superantigen,” kicking the immune system into overdrive as in toxic shock syndrome – a rare, life-threatening complication of bacterial infections.

Symptoms of a newly identified condition in pediatric COVID-19 patients, known as Multisystem Inflammatory Syndrome in Children (MIS-C), include persistent fever and severe inflammation that can affect a host of bodily systems.

While rare, the syndrome can be serious or even fatal.

The first reports of this condition coming out of Europe caught the attention of study co-senior author Moshe Arditi, M.D., director of the Pediatric Infectious Diseases and Immunology Division at Cedars-Sinai and an expert on another pediatric inflammatory disease – Kawasaki disease.

Arditi contacted his long-time collaborator, Ivet Bahar, Ph.D., distinguished professor and John K. Vries Chair of computational and systems biology at Pitt School of Medicine, and the two started searching for features of the SARS-CoV-2 virus that might be responsible for MIS-C.

Bahar and her team created a computer model of the interaction between the SARS-CoV-2 viral spike protein and the receptors on human T cells, the foot soldiers of the immune system.

Under normal circumstances, T cells help the body fight off infection, but when these cells are activated in abnormally large quantities, as is the case with superantigens, they produce massive amounts of inflammatory cytokines – small proteins involved in immune system signaling – in what’s known as a “cytokine storm.”

Using this computer model, the team was able to see that a specific region on the spike protein with superantigenic features interacts with T cells.

Then, they compared this region to a bacterial protein that causes toxic shock syndrome and found striking similarities in both sequence and structure.

Importantly, the proposed SARS-CoV-2 superantigen showed a high affinity for binding T cell receptors – the first step toward touching off a runaway immune response.

“Everything came one after another, each time a huge surprise. The pieces of the puzzle ended up fitting extremely well,” said Bahar, co-senior author on the study.

By finding protein-level similarities between SARS-CoV-2 and the bacterial structure that causes toxic shock syndrome, the researchers said they may have opened up new avenues for treating not only MIS-C patients, but also adults with COVID-19 infection experiencing cytokine storm.

The researchers also collaborated with scientists studying adult COVID-19 patients in Germany and found that those who experienced severe symptoms had a T cell response similar to what is seen in people exposed to superantigens and very different from the T cell response in patients who had only mild symptoms.

“Our research finally begins to unravel the potential mechanisms involved and raises the possibility that therapeutic options for toxic shock syndrome, such as intravenous immunoglobulin and steroids, may be effective for managing and treating MIS-C in children and hyperinflammation in adult coronavirus patients,” said Arditi, professor of pediatrics and biomedical sciences at Cedars-Sinai.

Arditi’s and Bahar’s labs are now using the ideas generated by this study to search for and test antibodies specific to the SARS-CoV-2 superantigen, with the goal of developing therapies that specifically address MIS-C and cytokine storm in COVID-19 patients.

---

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which causes COVID-19, is a coronavirus closely related to SARS-CoV and Middle East Respiratory Syndrome (MERS) coronaviruses (1). COVID-19 can manifest in adults as a severe interstitial pneumonia with hyperinflammation while severe respiratory manifestations are rare in children (2–4). Recently, however, multisystem inflammatory system in children (MIS-C) has been recognized in patients that either tested positive for COVID-19 (by PCR or serology) or had epidemiological links to COVID-19 (5–7). After initial reports in UK (5), many cases have now been reported in Europe (6, 7), and New York (USA CDC). However, no such cases have been reported in China, Japan, or South Korea, which have also been severely impacted by the COVID-19 pandemic (ECDC).

MIS-C manifests as persistent fever and hyperinflammation with multi organ system involvement including cardiac, gastrointestinal, renal, hematologic, dermatologic and neurologic symptoms (5–7) which are highly reminiscent of toxic shock syndrome (TSS) (8, 9) (Table 1), rather than Kawasaki disease due to marked demographic, clinical, and laboratory differences (6). The similarities to TSS and the association of MIS-C with COVID-19 led us to hypothesize that SARS-CoV-2 may possess superantigenic fragments that induce an inflammatory cascade and perhaps also contribute to the hyperinflammation and cytokine storm features observed in severe adult COVID-19 cases (3, 4). The question we raised is: does SARS-CoV-2 S possess superantigenic fragments that could elicit such reactions upon binding proteins involved in the host cell cytotoxic adaptive immune response? Such a reaction was not observed in the SARS-CoV pandemic of 2003 (shortly SARS1). What is unique to SARS-CoV-2, and how recent mutations in SARS-CoV-2 S may have promoted such an increased virulence?

Table 1:

Similarities between clinical and laboratory features of MIS-C and pediatric TSS

Clinical Features	MIS-Ca	Pediatric TSSb
High fever	+	+
Skin rash	+	+
Conjunctivitis	+	+
Oral mucosal involvement	+	+
Myalgia	+	+
Hypotension	+	+
Myocardial involvement (dysfunction)	+	+
Gastro-intestinal symptoms (vomiting, diarrhea, abdominal pain)	+	+
Renal involvement	+	+
CNS symptoms, altered mental state	+	+
Headache	+	+
High CRP	+	+
High Ferritin	+	+
High IL-6	+	+
High D-dimers	+	+
High Procalcitonin	+	+
Lymphopenia	+	+
Reduced Platelet count	+	+
Increased Neutrophil count	+	+
Increased AST and ALST	+	+
High Pro-BNP	+	NA
High Troponin	+	NA
Isolation of TSS inducing bacteria (Staphylococcus or Streptococcus)	−	+

TSS can be caused by two types of superantigens (SAgs): bacterial or viral. Bacterial SAgs have been broadly studied. They include proteins secreted by Staphylococcus aureus and Streptococcus pyogenes that induce inflammatory cytokine gene induction and toxic shock.

Typical examples are TSS toxin 1 (TSST1), and staphylococcal enterotoxins B (SEB) and H (SEH).

They are highly potent T cell activators that can bind to MHC class II (MHCII) molecules and/or to TCRs of both CD4+ and CD8+ T cells.

The ability of SAgs to bypass the antigen specificity of the TCRs results in broad activation of T cells and a cytokine storm, leading to toxic shock (12, 13).

Notably SAgs do not bind the major (antigenic) peptide binding groove of MHCII, but instead bind other regions as well as the αβTCRs, directly. While early studies showed that bacterial SAgs activate T cells by binding the β-chain of dimeric TCRs at their variable domain (V) (14–16), more recent studies revealed that they can bind to either α- or β-chains, or both (17). T

he question is then, does SARS-CoV-2 S possess any superantigenic fragments/domains that could bind to the αβTCRs?

Here, we used computational modelling to determine whether the SARS-CoV-2 S possesses SAg activity. We demonstrate that an insert present in SARS-CoV-2 S, which is absent from SARS1 and MERS, mediates high affinity, non-specific binding to the TCR.

Notably, a motif of ~20 amino acids enclosing this insert unique to SARS-CoV-2 among beta coronaviruses has sequence and structure features highly similar to those of the SEB toxin. Furthermore, our analysis indicates that a SARS-CoV-2 S mutation detected in a European strain may enhance TCR binding, suggesting such mutations may account for geographical differences in MIS-C occurrence. These finding have important implications for the management and treatment of COVID-19.

RESULTS AND DISCUSSION

SARS-CoV-2 spike harbors a high affinity site for TCR β-chain binding, which contains an insertion, P₆₈₁RRA₆₈₄, unique to SARS2.

We first examined whether SARS-CoV-2 S could bind to the αβTCR. To this aim, we constructed a SARS-CoV-2 S structural model based on the cryo-EM structure resolved for the spike glycoprotein (¹⁸), and we used the X-ray structure of αβTCR resolved in a ternary complex with SEH and MHCII (¹⁷), and generated a series of structural models for possible SARS-CoV-2 S glycoprotein – TCR complex formation using ClusPro (¹⁹).

Our simulations led to the binding pose presented in Fig. 1A as one of the most probable mechanisms of complex formation, as described in detail in Supplemental Information (SI) and Supplementary Fig. S1.

Therein, the TCR binds at the interface between the S1 and S2 subunits of the spike protein, near the S1/S2 cleavage site. A closeup view of the interface between the spike and TCRVβ domain (Fig. 1B) reveals several strong interatomic interactions, involving residues S680-R683 on the spike, and R70-E74 and [Q52, D56] on the respective CDRs 3 and 2 on Vβ.

We note that the TCRVβ-binding epitope on SARS-CoV-2 S is centered around a sequence motif, P681RRA684 (or shortly PRRA, hereafter), and its sequential and spatial neighbors. Comparison of SARS-CoV-2 S to other beta-coronavirus spike protein sequences showed (20) that SARS-CoV-2 is distinguished by the existence of this four-residue insertion, PRRA, preceding the furin cleavage site (R685-S686 peptide bond) between the subunits S1 and S2 of each protomer (Fig. 2A).

Structural comparison of the trimeric S proteins between SARS-CoV and SARS-CoV-2 further shows their close structural similarity in general (except for the RBD which is engaged in specific interfacial interactions (18)), but the two spikes significantly differ near the PRRARS motif unique to SARS-CoV-2, which is exposed to the extracellular medium (Fig. 2B).

Further examination of the motif near PRRA reveals close structural similarity to the SEB superantigen as well as sequence similarities to neurotoxins and a viral SAg.

The insertion PRRA together with the sequentially preceding seven amino acids and succeeding Arg (fully conserved among β-coronaviruses) have been pointed out to form a motif, Y674QTQTNSPRRAR685, homologous to that of neurotoxins from Ophiophagus (cobra) and Bungarus genera, as well as neurotoxin-like regions from three RABV strains (21) (Fig. 2C).

We further noticed that the same segment bears close similarity to HIV-1 glycoprotein gp120 superantigenic motif F164-V164.

This close sequence similarity to both bacterial and viral SAgs, in support of the potential superantigenic character of the amino acid stretch Y674-R685 of SARS-CoV-2 S directed us to further analyze its local sequence and structural properties.

Our analysis led to an interesting sequence similarity between the partially overlapping fragment T678-Q690 of the spike and the SEB superantigenic peptide Y150NKKKATVQELD161 (Fig. 3A).

This dodecapeptide sequence within the SEB shows strong conservation among a broad range of staphylococcal and streptococcal SAgs (23, 24). We note that the sequentially aligned segment of SARS1 (S664-K672) bears minimal similarity to the SEB SAg (Fig. 3A left).

What is even more interesting is that SARS-Cov-2 motif showed a palindromic behavior with respect to the superantigenic SEB sequence in the sense that a broader stretch, from E661 to R685, could be aligned to the superantigen peptide in the reverse direction as well (Fig. 3A right).

This brings to our attention the versatility and high propensity of the SARS-CoV-2 S TCRVβ-binding site residues to potentially act as a superantigenic fragment.

Significantly, the structures of the two peptides exhibit a remarkable similarity (Fig. 3B–C), including a salt bridge stabilizing each structural motif (E159-K152 in SEB and E661-R685 in SARS-CoV-2 S), the relative orientations of three positively charged residues (K152-K153-K154 in SEB and R682-R683-R685 in SARS-CoV-2 S), a β-hairpin which apparently serves as a scaffold, and may be further supported in SARS coronavirus by a potential disulfide bond between C662-C671. The former two features are absent in SARS1 spike (Fig. 3D).

This analysis overall indicates that the segment T678NSPRRAR685 forms a putatively superantigenic core, consistently aligned against various bacterial or viral SAgs (Figs. 2C and ​and3A3A–C) with or without the participation of the adjoining amino acids.

However, combined broader sequence and structure analysis in Fig. 3 panels A (right) and ​andBB–C, reveals an even more compelling feature: this putative SAg core is structurally consolidated by spatial proximity to a conserved acidic segment, E661CD663, which forms a highly stable salt bridge with the polybasic segment PRRAR of SARS-CoV-2 S, much in the same way as to the salt bridge observed in SEB (but not in SARS1 S).

We note that the SEB superantigen peptide Y150NKKKATVQELD161 has been reported to bind CD28 (23), a T cell receptor that provides co-stimulatory signals required for T cell activation and survival. CD28 and TCRV domains share the same (immunoglobulin) fold (Fig. 3E), and the binding mechanism shown in Fig. 1B is likely to be adopted with minor rearrangements for enabling the binding of SEB to CD28.

Finally, because of the homologous superantigenic segment of SEB binding CD28, we also tested the potential binding of SARS2 spike E661-R685 onto CD28, considering the possibility that the target of SARS2 spike superantigenic segment might be CD28. Our simulations indicated that the same segment can equally bind to CD28, further supporting the strong propensity of the fragment to stimulate T cell activation.

SARS-CoV-2 spike retains other superantigenic or toxic fragments observed in SARS1 spike, including an ICAM-1 like motif engaged in stabilizing interactions with TCRVα.

The existence of potential superantigenic, toxic or intercellular-adhesion molecule (ICAM)-like sequence fragments in SARS1 was thoroughly examined by Li et al. following the 2003 pandemic (26).

This led to the identification of the nine sequence stretches including three Botulinum neurotoxin type D or G precursors, and two motifs that have a high similarity with the intercellular adhesion molecule 1 (ICAM-1). Comparative analysis with SARS-CoV-2 spike sequence revealed that seven of these sequence motifs are conserved between SARS-CoV and SARS-CoV-2 (Supplementary Fig. S2).

Among them, we note that Y279NENGTITDAVDCALDPLSETKC301, an ICAM-1 (CD54)-like motif, also participates in the association between the SARS-CoV-2 spike and the bound αβTCR (see Fig. 4).

Note that ICAM-1 involvement is critical to mediating immune and inflammatory responses. The observed interaction of the ICAM-1-like motif of SARS-CoV-2 S with TCRVα, in tandem with the interaction of the above discussed putative SAg motif (around the insert PRRA) with TCRVα, is likely to further strengthen the association of the virus with the T cell and the ensuing activation.

Precisely, N280-E281-N282 and T286 belonging to the ICAM-like fragment closely interact with the TCRVα CDRs; mainly T286 (spike) makes close contacts with S94 (CDR3), E281 (spike) forms a hydrogen bond with T51 (CDR2), and N280 and N282 (spike) closely associate with R69 (Fig. 4D).

A rare mutation, D839Y/E, recently observed in a SARS2 strain from Europe contributes to stabilizing the interaction with TCR.
Interestingly, the SARS-CoV-2 spike binding region harbors three residues that have been recently reported to have mutated in new strains from Europe and USA (27, 28): D614G, A831V and D839Y/N/E).

The former two may potentially interact with MHCII; while the latter (D839, European strain) is located close to TCRVβ and strongly interacts with N30; (Fig. 4A and ​andC,C, and Supplementary Figs. S3 and S4A).

Its substitution by glutamate in the mutant D839E increases the strength of the intermolecular (and thereby virus-T cell) association (Supplementary Fig. S4C). Even stronger interactions between spike and TCRVβ are observed upon replacing D839 with a tyrosine as illustrated in Supplementary Fig. S4D: The interfacial interactions in this case are further stabilized by a hydrogen bond between Y839 and D32; an aromatic (polar-π) interaction between Y839 and N30; as well as possible electrostatic interactions with K73 and S97.

Quantitative evaluation of the change in binding affinity between the spike and TCR upon mutating D839 to Y, E and N yields ΔΔGD→Y = −2.2kcal mol−1, ΔΔGD→E = −2.1kcal mol−1, and ΔΔGD→N = −1.3kcal mol−1 respectively (see Supplementary Table S1 for details). Thus, the D839Y/N/E mutations would be expected to strengthen/support the above described association between the superantigenic PRRA-containing segment and the TCRVβ.

CONCLUSION
An understanding of the immunopathology leading to severe manifestations of COVID-19, in both adults and children, is of critical importance for effective management and treatment of the disease. MIS-C shows remarkable similarity to pediatric TSS (5–9).

Using in silico modeling and analysis, we found that SARS-CoV-2 encodes a superantigen motif near its S1/S2 cleavage site.

This region is highly similar in structure to the SEB SAg motif that interacts with both the TCR and CD28 (23) and mediates TSS. SEB enables large-scale T cell activation and proliferation (13), resulting in massive production of pro-inflammatory cytokines including IFNγ, TNFα and IL-2 from T cells as well as IL-1 and TNFα from APCs (13).

This cytokine storm leads to multi-organ tissue damage similar to what is now observed in MIS-C. We therefore propose that MIS-C observed in COVID-19 patients may be mediated by superantigen activity of the SARS-CoV-2 S protein.

To date, MIS-C is mostly observed in Europe and East Coast of North America, and has not been described in Asia, despite sizeable out-breaks of COVID-19 (5–7) (CDC and ECDC). We show that a mutation at D839 found in a European strain of SARS-CoV-2 enhances the binding affinity of the SAg motif to the TCR.

This could (at least partly) explain the geographical skewing of MIS-C to areas where the European strain is endemic, and identification of other strain-specific mutations may help predict where future outbreak of MIS-C may occur.

Our findings raise the exciting possibility that immunomodulatory therapeutic options used for TSS may also be effective for MIS-C, including IVIG and steroids. Indeed, initial published and unpublished reports suggest that MIS-C patients respond well to IVIG with or without steroids (5–7).

IVIG recognizes SEB epitopes (29), and thus may function in part by neutralization of a superantigen. Given structural similarities between SEB and the S protein SAg motif, there is potential for cross-reactivity of these immunoglobins, explaining at least in part the response of MIS-C cases to IVIG.

Other FDA-approved anti-inflammatory drugs tested in models of SEB TSS may also be effective, including CTLA4-Ig which can inhibit CD28 co-stimulation (30), and the mTOR inhibitor rapamycin (31), which is already in use for COVID-19.

In addition, humanized monoclonal anti-SEB Abs have been described (32) that could also be of potential therapeutic benefit in MIS-C patients. Notably, it has been shown in the mouse model of TSS that lethal SEB superantigen challenge can be prevented by short peptide mimetics of its superantigen motif (23).

It would be interesting to examine whether short peptide mimetics of SARS-CoV-2 spike superantigen region might be employed to prevent/attenuate inflammatory cytokine gene induction and toxic shock in MIS-C patients.

At present, the majority of antibody therapies under investigation are designed to target the SARS-CoV-2 receptor binding domains (RBDs) (33, 34), and our simulations also indicated that RBD might potentially interact with TCRs.

However, compared with RBDs, relatively fewer mutations are found in the SAg region of SARS-CoV-2; notably, the “PRRA” insert is unique to SARS-CoV-2 and retained among all of its isolates sequenced to date (27, 28). It might be constructive to design antibodies or drugs targeting this SAg region, to not only block the cleavage essential to enabling viral entry (1, 20), but also modulate the SAg-induced inflammatory cytokine gene induction (13).

Fortunately, severe respiratory manifestations of COVID-19 in children as well as development of MIS-C are rare. This may be due to trained immunity (2) or cross-viral immunity to other coronavirus strains (35). T and B cells play an important role in the anti-viral response.

CD4+ and CD8+ T cells from convalescent COVID-19 patients can recognize a range of SARS2 epitopes, and the S protein represents a major target (35).

Interestingly, T cells from unexposed individuals can also respond to S protein epitopes from SARS-CoV-2, which supports the hypothesis of cross-viral immunity from other coronavirus strains (35). However, why only a fraction of infected children develop MIS-C is unclear.

It is possible that a poor initial antibody response to the virus fails to neutralize SAg leading to immune enhancement following re exposure or that certain HLA types are more permissive of binding SAg, and indeed HLA has been shown to play a role in COVID susceptibility (36).

Of the nine cases initially reported in the UK, six were of Afro-Caribbean descent, which also suggests a potential genetic component to susceptibility (5).

It is interesting to note that approximately a third or fewer of MIS-C patients tested positive for the SARS-CoV-2, but the majority (but not all) have serologic evidence of infection or a history of exposure to COVID-19 (5–7). This may suggest that the SARS-CoV-2 SAg causes a delayed hyperinflammation response in certain children.

SAgs have been implicated in autoimmunity by triggering self-reactive T cells (12). Antibody-mediated enhancement upon re-exposure to the virus may also contribute to uncontrolled infection and inflammation (37).

It is also possible that despite a negative nasopharyngeal PCR test, the virus may still be present in the gastrointestinal tract (38). MIS-C patients demonstrate unusually severe GI symptoms, abdominal pain, vomiting and diarrhea, in addition to severe myocardial dysfunction and cardiac shock (5–7) and such severe GI symptoms are also frequently associated with the SAg (9).

In summary we made three major observations: (a) PRRAR and sequential neighbors interact with TCRVβ residues D56, R70 and E74 at the CDRs, and this association closely resembles that of SEB SAg with TCRVβ; (b) nearby D839 participates in this interaction and its mutation to tyrosine would further strengthen the association with TCRVβ; and (c) a sequence motif (N280-T286) typical of ICAM-1 further interacts with the TCRVα further stabilizing/enhancing the association between the viral spike and host cell TCR.

So, the binding pose of TCR with respect to the spike allows simultaneously for all three associations. The first is the most important. The 2nd is significant because it is the European strain.

The 3rd may be imparting a stronger host-virus adhesion. Therefore, strategies used for the treatment of SEB-mediated Toxic Shock or approaches to block the interaction of S protein with the TCR may reduce hyperinflammatory manifestations of COVID-19 in both adults and children.

REFERENCES

1. Walls A. C. et al., Structure, Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein. Cell 180, 281–292 (2020). [PMC free article] [PubMed] [Google Scholar]

2. Cristiani L. et al., Will children reveal their secret? The coronavirus dilemma. Eur. Respir. J., in press (2020). [PMC free article] [PubMed] [Google Scholar]

3. Tay M. Z., Poh C. M., Rénia L., MacAry P. A., Ng L. F. P., The trinity of COVID-19: immunity, inflammation and intervention. Nat. Rev. Immunol., in press (2020). [PMC free article] [PubMed] [Google Scholar]

4. Vabret N. et al., Immunology of COVID-19: current state of the science. Immunity, in press (2020). [PMC free article] [PubMed] [Google Scholar]

5. Riphagen S., Gomez X., Gonzalez-Martinez C., Wilkinson N., Theocharis P., Hyperinflammatory shock in children during COVID-19 pandemic. Lancet, in press (2020). [PMC free article] [PubMed] [Google Scholar]

6. Verdoni L. et al., An outbreak of severe Kawasaki-like disease at the Italian epicentre of the SARS-CoV-2 epidemic: an observational cohort study. Lancet, in press (2020). [PMC free article] [PubMed] [Google Scholar]

7. Belhadjer Z. et al., Acute heart failure in multisystem inflammatory syndrome in children (MIS-C) in the context of global SARS-CoV-2 pandemic. Circulation, in press (2020). [PubMed] [Google Scholar]

8. Low D. E., Toxic shock syndrome: major advances in pathogenesis, but not treatment. Crit. Care Clin. 29, 651–675 (2013). [PubMed] [Google Scholar]

9. Cook A., Janse S., Watson J., Erdem G., Manifestations of Toxic Shock Syndrome in Children, Columbus, Ohio, USA, 2010–2017. Emerg. Infect. Dis. 26, 1077–1083 (2020). [PMC free article] [PubMed] [Google Scholar]

10. Young A. E., Thornton K. L., Toxic shock syndrome in burns: diagnosis and management. Arch. Dis. Child. Educ. Pract. 92, ep97–ep100 (2007). [PubMed] [Google Scholar]11. Matsuda Y. et al., Early and definitive diagnosis of toxic shock syndrome by detection of marked expansion of T-cell-receptor Vβ2-positive T cells. Emerg. Infect. Dis. 9, 387 (2003). [PMC free article] [PubMed] [Google Scholar]

12. Li H., Llera A., Malchiodi E. L., Mariuzza R. A., The structural basis of T cell activation by superantigens. Annu. Rev. Immunol. 17, 435–466 (1999). [PubMed] [Google Scholar]

13. Krakauer T., Staphylococcal superantigens: pyrogenic toxins induce toxic shock. Toxins 11, 178 (2019). [PMC free article] [PubMed] [Google Scholar]

14. Scherer M. T., Ignatowicz L., Winslow G. M., Kappler J. W., Marrack P., Superantigens: bacterial and viral proteins that manipulate the immune system. Annu. Rev. Cell Biol. 9, 101–128 (1993). [PubMed] [Google Scholar]

15. Choi Y. W. et al., Interaction of Staphylococcus aureus toxin “superantigens” with human T cells. Proc. Natl. Acad. Sci. U.S.A. 86, 8941–8945 (1989). [PMC free article] [PubMed] [Google Scholar]

16. Fraser J. D., Proft T., The bacterial superantigen and superantigen-like proteins. Immunol. Rev. 225, 226–243 (2008). [PubMed] [Google Scholar]

17. Saline M. et al., The structure of superantigen complexed with TCR and MHC reveals novel insights into superantigenic T cell activation. Nat. Commun. 1, 119 (2010). [PubMed] [Google Scholar]

18. Wrapp D. et al., Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation. Science 367, 1260–1263 (2020). [PMC free article] [PubMed] [Google Scholar]

19. Kozakov D. et al., The ClusPro web server for protein–protein docking. Nat. Protoc. 12, 255 (2017). [PMC free article] [PubMed] [Google Scholar]

20. Hoffmann M. et al., SARS-CoV-2 Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor. Cell 181, 271–280 (2020). [PMC free article] [PubMed] [Google Scholar]

21. Changeux J.-P., Amoura Z., Rey F., Miyara M., A nicotinic hypothesis for Covid-19 with preventive and therapeutic implications. Qeios doi:10.32388/FXGQSB (2020). [CrossRef] [Google Scholar]

22. Bracci L., Ballas S. K., Spreafico A., Neri P., Molecular mimicry between the rabies virus glycoprotein and human immunodeficiency virus-1 GP120: cross-reacting antibodies induced by rabies vaccination. Blood, Am. J. Hematol. 90, 3623–3628 (1997). [PubMed] [Google Scholar]

23. Arad G. et al., Binding of superantigen toxins into the CD28 homodimer interface is essential for induction of cytokine genes that mediate lethal shock. PLoS Biol. 9, e1001149 (2011). [PMC free article] [PubMed] [Google Scholar]

24. Popugailo A., Rotfogel Z., Supper E., Hillman D., Kaempfer R., Staphylococcal and streptococcal superantigens trigger B7/CD28 costimulatory receptor engagement to hyperinduce inflammatory cytokines. Front Immunol. 10, 942 (2019). [PMC free article] [PubMed] [Google Scholar]

25. Papageorgiou A. C., Tranter H. S., Acharya K. R., Crystal structure of microbial superantigen staphylococcal enterotoxin B at 1.5 A resolution: implications for superantigen recognition by MHC class II molecules and T-cell receptors. J. Mol. Biol. 277, 61–79 (1998). [PubMed] [Google Scholar]

26. Li Y. et al., Structure-based preliminary analysis of immunity and virulence of SARS coronavirus. Viral Immunol. 17, 528–534 (2004). [PubMed] [Google Scholar]

27. Zhan S. H., Deverman B. E., Chan Y. A., SARS-CoV-2 is well adapted for humans. What does this mean for re-emergence? bioRxiv, 2020.2005.2001.073262 (2020). [Google Scholar]

28. Korber B. et al., Spike mutation pipeline reveals the emergence of a more transmissible form of SARS-CoV-2. bioRxiv, 2020.2004.2029.069054 (2020). [Google Scholar]

29. Nishi J.-I. et al., B cell epitope mapping of the bacterial superantigen staphylococcal enterotoxin B: the dominant epitope region recognized by intravenous IgG. J. Immunol. 158, 247–254 (1997). [PubMed] [Google Scholar]

30. Whitfield S. J. et al., Interference of the T cell and antigen-presenting cell costimulatory pathway using CTLA4-Ig (abatacept) prevents Staphylococcal enterotoxin B pathology. J. Immunol. 198, 3989–3998 (2017). [PMC free article] [PubMed] [Google Scholar]

31. Krakauer T., Buckley M., Issaq H. J., Fox S. D., Rapamycin protects mice from staphylococcal enterotoxin B-induced toxic shock and blocks cytokine release in vitro and in vivo. Antimicrob. Agents Chemother. 54, 1125–1131 (2010). [PMC free article] [PubMed] [Google Scholar]

32. Larkin E. A., Stiles B. G., Ulrich R. G., Inhibition of toxic shock by human monoclonal antibodies against staphylococcal enterotoxin B. PLoS One 5, e13253 (2010). [PMC free article] [PubMed] [Google Scholar]

33. Yuan M. et al., A highly conserved cryptic epitope in the receptor binding domains of SARS-CoV-2 and SARS-CoV. Science 368, 630–633 (2020). [PMC free article] [PubMed] [Google Scholar]

34. Chi X. et al., A potent neutralizing human antibody reveals the N-terminal domain of the Spike protein of SARS-CoV-2 as a site of vulnerability. bioRxiv, 2020.2005.2008.083964 (2020). [Google Scholar]

35. Grifoni A., Weiskopf D., Ramirez S.I., Mateus J., Dan J.M., Moderbacher R. C.R. S.A., Sutherland A., Premkumar L., Jadi R.S., Marrama D., de Silva A.M., Frazier A., Carlin A., Greenbaum J.A., Peters B., Krammer F., Smith D.M., Crotty S., Sette A.,, Targets of T cell responses to SARS-CoV-2 coronavirus in humans with COVID-19 disease and unexposed individuals. Cell in press (2020). [PMC free article] [PubMed] [Google Scholar]

36. Nguyen A. et al., Human leukocyte antigen susceptibility map for SARS-CoV-2. J. Virol., JVI.00510–00520 (2020). [PMC free article] [PubMed] [Google Scholar]

37. Tirado S. M. C., Yoon K.-J., Antibody-dependent enhancement of virus infection and disease. Viral Immunol. 16, 69–86 (2003). [PubMed] [Google Scholar]

38. Xu Y. et al., Characteristics of pediatric SARS-CoV-2 infection and potential evidence for persistent fecal viral shedding. Nat. Med. 26, 502–505 (2020). [PMC free article] [PubMed] [Google Scholar]

39. Bordoli L., Schwede T., “Automated protein structure modeling with SWISS-MODEL Workspace and the Protein Model Portal” in Homology Modeling. (Springer, 2011), pp. 107–136. [PMC free article] [PubMed] [Google Scholar]

40. Song W., Gui M., Wang X., Xiang Y., Cryo-EM structure of the SARS coronavirus spike glycoprotein in complex with its host cell receptor ACE2. PLoS Pathog. 14, e1007236 (2018). [PMC free article] [PubMed] [Google Scholar]

41. Zhang Y., Skolnick J., The protein structure prediction problem could be solved using the current PDB library. Proc. Natl. Acad. Sci. U.S.A. 102, 1029–1034 (2005). [PMC free article] [PubMed] [Google Scholar]

42. Peitsch M. C., Protein modeling by E-mail. Bio/technology 13, 658–660 (1995). [Google Scholar]

43. Jo S., Kim T., Iyer V. G., Im W., CHARMM-GUI: a web-based graphical user interface for CHARMM. J Comput. Chem. 29, 1859–1865 (2008). [PubMed] [Google Scholar]

44. Xue L. C., Rodrigues J. P., Kastritis P. L., Bonvin A. M., Vangone A., PRODIGY: a web server for predicting the binding affinity of protein–protein complexes. Bioinformatics 32, 3676–3678 (2016). [PubMed] [Google Scholar]

---

More information: Mary Hongying Cheng et al. Superantigenic character of an insert unique to SARS-CoV-2 spike supported by skewed TCR repertoire in patients with hyperinflammation, Proceedings of the National Academy of Sciences (2020). DOI: 10.1073/pnas.2010722117