Spiking Pandemic Potential: Structural and Immunological Aspects of SARS-CoV-2.

SARS-Coronavirus-2 (SARS-CoV-2) causes Coronavirus disease 2019 (COVID-19), an infectious respiratory disease causing thousands of deaths and overwhelming public health systems. The international spread of SARS-CoV-2 is associated with the ease of global travel, and societal dynamics, immunologic naiveté of the host population, and muted innate immune responses. Based on these factors and the expanding geographic scale of the disease, the World Health Organization (WHO) declared the COVID-19 outbreak a pandemic-the first caused by a coronavirus. In this review, we summarize the current epidemiological status of COVID-19 and consider the virological and immunological lessons, animal models, and tools developed in response to prior SARS-CoV and MERS-CoV outbreaks that can serve as resources for development of SARS-CoV-2 therapeutics and vaccines. In particular, we discuss structural insights into the SARS-CoV-2 spike protein, a major determinant of transmissibility, and discuss key molecular aspects that will aid in understanding and fighting this new global threat.

The COVID-19 Pandemic to Date

From Emergence in Wuhan to Global Pandemic

In December 2019, a novel human coronavirus (HCoV) was identified as the causative agent of clusters of pneumonia in China. The virus was named as SARS-CoV-2, based on its phylogenetic and taxonomic similarity to SARS-CoV [1], which caused an outbreak of severe acute respiratory syndrome (SARS) in 2002. The WHO named the disease caused by SARS-CoV-2 as coronavirus disease 2019 (COVID-19). The first confirmed SARS-CoV-2 case was found in Wuhan, China (Figure 1A). Subsequently, medical workers and family clusters who had not visited Wuhan tested positive for SARS-CoV-2, thus confirming human-to-human transmission [2]. SARS-CoV-2 rapidly spread to most countries and has resulted in thousands of fatalities (Figure 1B–D). WHO declared a pandemic on 11 March 2020. As of 5 May 2020, over 3 500 000 cases have been confirmed in over 185 countries, with over 243 000 deaths, suggesting a case fatality rate (CFR) of 6.9% (WHO report 106) (Figure 1B–D). However, nominal CFR are strongly influenced by the extent of testing of suspected cases and under-testing can result in higher apparent CFR. In this review, we summarize current progress in epidemiology and detection methods for SARS-CoV-2 and discuss findings concerning general characteristics of pathogenic coronaviruses. These studies form the foundation for future efforts to develop vaccines and pre- and postexposure therapeutics.

Figure 1

Timeline of Major Events for Severe Acute Respiratory Syndrome-Coronavirus-2 (SARS-CoV-2), Total Numbers of Coronavirus Disease 2019 (COVID-19) Cases, and Basic Reproduction Number (R0) and Mortality Rate of Select Viruses.

(A) Timeline of events, (B) cumulative confirmed cases, (C) reported deaths, and (D) countries with reported cases. Squares indicates the total number of COVID-19 cases reported worldwide. Circles and triangles indicate the number of COVID-19 cases reported in China and in other countries, respectively. Data from WHO as of May 5 2020. (E) R0 value indicates the number of people (colored in blue) infected from one contagious person (colored in red). (F) Case fatality rate (CFR) of select viruses. aCFR is based on patients not receiving therapy. bData on SARS-CoV-2 are derived from WHO. Abbreviations: Cryo-EM, Cryogenic electron microscopy; ICTV, International Committee on Taxonomy of Viruses; PHEIC, public health emergency of international concern; WHO, World Health Organization.

Transmissibility of SARS-CoV-2

The spread of an infectious disease is dependent on the transmissibility of the causative pathogen. The basic reproduction number, R 0, is used to measure the potential transmission of a disease and is defined as the average number of people who will catch a disease from one contagious person [3]. A higher CFR is generally associated with lower transmissibility (Figure 1E,F) [4]. Although COVID-19 has a lower mortality risk than SARS and Middle East respiratory syndrome (MERS), its rapid spread and higher R0 of ~3 contributed to the WHO designation of COVID-19 as a pandemic. The transmission of SARS-CoV-2 from infected, yet asymptomatic, carriers has been reported [5]. A metapopulation model simulation estimated that the transmission rate by asymptomatic persons or persons with mild symptoms is 55% [6], which adds to the difficulties in preventing transmission of SARS-CoV-2.

Clinical Manifestations of COVID-19

The most common symptoms of COVID-19 are fever (89.5%), cough (73.4%), dyspnea (38.5%), and myalgia (31.3%), similar to SARS and MERS (Table 1 ). The median time from disease onset to dyspnea in COVID-19 is 8 days (range, 5–13 days) [7]. Another study of 291 patients with COVID-19 for whom the exposure date was known showed that the median incubation period was 4 days, with an interquartile range of 2–7 [8].

Table 1

Clinical Features of COVID-19, SARS, and MERS

Symptom	COVID-19							SARS	MERS
	Studya
	1(n = 62) [91]	2(n = 1099) [8]	3(n = 138) [92]	4(n = 140) [9]	5(n = 41) [7]	6(n = 99) [10]	Range (average)	7Range [93]	8Range [94]
Fever	77.4	88.7	98.6	91.7	97.6	82.8	77.4–98.6 (89.5)	99–100	98
Cough	80.6	67.8	59.4	75.0	75.6	81.8	67.8–86.2 (73.4)	62–100	83
Dyspnea	–	–	31.2	36.7	55.0	31.3	31.2–55.0 (38.5)	–	–
Myalgia	51.6	14.9	34.8	–	43.9	11.1	11.1–51.6 (31.3)	45–61	32
Panting	3.2	–	–	–	29.3	–	3.2–29.3 (16.2)	–	–
Headache	34.4	13.6	6.5	–	7.9	8.1	6.5–34.4 (14.1)	20–56	11
Sore throat	–	13.9	17.4	–	–	5.1	5.1–17.4 (12.1)	13–25	14
Nausea/vomiting	–	5.0	6.3-10	22.3	–	1.0	2.0–22.3 (9.4)	20–35	21
Diarrhea	4.8	3.8	10.1	12.9	2.6	2.0	2.0–12.9 (6.1)	20–25	26
Rhinorrhea	–	–	–	–	–	4.0	4.0 (4.0)	2–24	6
Abdominal pain	–	–	2.2	5.8	–	–	2.2–5.8 (4.0)	–	–

Percentage of study subjects that experienced the indicated symptom; for fields without values, this symptom was not evaluated.

In patients having confirmed SARS-CoV-2 infection, 15–42% have severe symptoms and some progress to acute respiratory distress syndrome (ARDS), which can be fatal [8,9]. COVID-19 is suggested to cause more severe illness in older people and patients with underlying diseases, such as hypertension, cardiovascular disease, or diabetes [7,10]. The CFR increased considerably among patients aged between 60 and 80 years (30%) and reached 36% in patients older than 80 years of age [11].

SARS-CoV-2 can also infect younger individuals, including children. Among 731 individuals younger than 15 years of age with confirmed SARS-CoV-2 infection, 12.9% were asymptomatic, 84% had mild to moderate disease symptoms, and <3% of cases had severe or life-threatening symptoms [12]. However, SARS-CoV-2 infection in children can still be fatal. [13].

Interspecies Transmission of CoVs

CoVs have been found to infect humans and a wide variety of domestic and wild vertebrates [14]. Most CoV infections in animals cause mild to severe gastrointestinal or respiratory infections [14]. In humans, CoV infections also cause respiratory illnesses. Four human CoVs, HCoV 229E, OC43, NL63, and HKU1 are associated with the common cold and have low morbidity [14].

In 2003, the SARS outbreak drew extensive attention to HCoVs. Fever is the most common clinical sign for SARS-CoV infection [14] and lower respiratory tract symptoms develop several days after disease onset. Of patients infected with SARS-CoV, 10–20% progressed to respiratory failure after initial symptoms failed to resolve [14]. During the SARS outbreak there were 8096 reported cases worldwide and the CFR was 10% (Figure 1F). The human-to-human transmission of SARS-CoV was controlled by public health measures shortly after the virus emerged and no infections in humans have been reported since 2004. Another HCoV-associated respiratory disease caused by MERS-CoV was identified in 2012. Similar to SARS-CoV, the clinical signs of MERS-CoV infection include fever, cough, and/or shortness of breath. To date, there have been 2494 laboratory-confirmed MERS-CoV cases, with a 40% CFR (Figure 1F).

Evolution in Host Species and Genetics of Virus–Host Shifts

RNA viruses accumulate substitutions in their genomes due to the low fidelity of viral RNA polymerases [15]. A typical mutation rate of 1 in 104 results in quasispecies diversity that can promote viral adaptation and potentially virulence [16]. CoVs undergo substitutions/mutations that drive CoV evolution [17]. Recombination events also provide another opportunity for the acquisition or modification of genes by CoVs. Thus, genomic RNA of CoVs can be modified via several pathways to promote rapid evolution and the ability to spill over into new host species [17].

Genetic studies revealed molecular evidence indicating that SARS-CoV likely originated from bats, with civet cats as an intermediate host [17]. MERS-CoV, however, was found to be transmitted to humans via camels [18]. Genetic studies indicate that SARS-CoV-2 likely originated from a bat CoV [19]. Current reports demonstrated that CoVs from pangolin showed the highest homology with SARS-CoV-2 in the receptor binding domain in S protein [20,21] and suggest that the pangolin could be an intermediate host.

Genomic Comparison of SARS-CoV-2 with SARS-CoV and MERS-CoV

The SARS-CoV-2 genome is 29 903 nucleotides, which contains 14 open reading frames (ORFs) [22] (Figure 2A). ORF1a and 1b encode the polyproteins pp1a and pp1ab, the latter through a ribosomal frameshifting mechanism at the 1a-1b gene boundary. These polyproteins are cleaved by viral proteases into 16 nonstructural proteins (nsp). Four ORFs encode structural proteins such as the spike (S), envelope (E), membrane (M), and nucleocapsid (N) genes. Between these structural genes, a series of accessory genes encode accessory proteins, which regulate infection but do not incorporate into the virion (ORFs 3a, 3b, 6, 7a, 7b, 8b, 9b, and 14) (Figure 2A).

Figure 2

Genomic Distribution of Severe Acute Respiratory Syndrome-Coronavirus-2 (SARS-CoV-2) and SARS-CoV.

(A) The genomes of SARS-CoV (upper panel) and SARS-CoV-2 (lower panel) are shown as lines and the open reading frames (ORFs) are represented by gray and colored boxes to indicate those that have similar and different lengths, respectively. The atomic structure for some of the SARS-CoV-2 proteins is shown in surface representation; main protease 3CLpro or nonstructural protein 5 (nsp5) with unliganded active site in pink, nsp9 RNA binding protein in cyan, nsp15 endoribonuclease in marine blue, nsp16–nsp10 complex in green, prefusion spike glycoprotein in gray, and nucleocapsid protein N terminal RNA binding domain in deep turquoise. For visual clarity, the length of the boxes is not proportional to the real sequence length and the atomic structures are not proportional to their molecular weight. (B) Percentage identity matrix for the alignment of SARS-CoV and SARS-CoV-2 amino acids. Abbreviations: PDB, Protein Data Bank.

The SARS-CoV-2 genome shares 79% and 50% sequence identity with SARS-CoV and MERS-CoV genome, respectively [19]. The gene arrangement of SARS-CoV-2 is similar to SARS-CoV, with some variations (Figure 2A). The viroporin 8a protein is present in SARS-CoV but absent in SARS-CoV-2. This protein was also lost during the SARS pandemic in 2003, demonstrating that it is not essential for the virulence [23]. Meanwhile, the 8b protein is 17 amino acids longer in SARS-CoV-2 than in SARS-CoV and the 3b protein of SARS-CoV-2 is only 22 amino acids as compared with 154 amino acids for SARS-CoV.

Among the structural proteins E, M, N, and S, the E protein has the highest similarity between SARS-CoV-2 and SARS-CoV (96% identity) (Figure 2B). The sequence conservation of E could be due to its critical role in the virus life cycle and, as a transmembrane protein, E is relatively protected from immune surveillance [24]. The S glycoprotein of SARS-CoV-2 has the largest sequence divergence (76% identity with SARS-CoV) [25], which likely reflects increased immune pressure. Indeed, the SARS-CoV-2 S protein has 380 amino acid sequence substitutions compared with other SARS-like viruses [26].

CoV Life Cycle

The CoV life cycle (Figure 3 ) begins with interactions between S on the virion surface and specific virus receptors. HCoV-NL63 [27], SARS-CoV [28], and SARS-CoV-2 use angiotensin converting enzyme 2 (ACE2) as a receptor [29]. The entry of the virus is through receptor-mediated endocytosis, followed by fusion of viral and host cell membranes [30]. A fusion event also occurs between viral particles and the plasma membrane on the cell surface [31]. Exposure to low pH in the endosome activates host proteases, such as cathepsin L and TMPRSS2, that cleave the S protein [30]. This cleavage induces a conformational change in the S protein that promotes fusion between virus and cell membranes and subsequent release of viral genomic RNA into the cytoplasm. Viral genomic RNA serves as an mRNA for translation [32]. Two viral proteases, nsp3 (PLpro) and nsp5 (3CLpro), cleave the polyproteins that comprise mature nsps [32]. nsp3, nsp4, and nsp6 also modify the endoplasmic reticulum (ER) membrane to yield unique membrane structures termed double-membrane vesicles (DMVs) [33., 34., 35.]. Viral RNA transcription is carried out in the DMV, where viral RNAs are protected from host pattern recognition receptors [33]. The N protein interacts with genomic RNA to form the ribonucleoprotein complex [36], which is recruited to the viral assembly site, the ER–Golgi intermediate compartment where viral particles are formed [37]. Following viral assembly, the newly formed viral particles are transported to the cell surface in vesicles and released by exocytosis [30].

Figure 3

Coronavirus (CoV) Life Cycle and Host Immune Response.

The CoV life cycle initiates with the binding of spike proteins on the virion surface to their specific receptor [e.g., ACE2 for severe acute respiratory syndrome-coronavirus-2 (SARS-CoV-2)] on target cells. Following receptor binding and endocytosis, S protein undergoes a conformational change and viral genomic RNA is released into the cytoplasm. Viral genomic RNA functions as mRNA, which is translated into polyproteins (pp) pp1a and pp1ab, which are then proteolytically cleaved into mature nonstructural proteins (nsps). Many nsps synergize to modify the endoplasmic reticulum (ER) membrane to form double membrane vesicles (DMVs) where transcription of genomic and subgenomic RNA occurs. The truncated subgenomic RNA of the 5′-ends are used as templates to translate structural (S, E, M, and N) and accessory proteins. S, E, and M assemble together with nucleocapsid (one copy of viral genome encapsulated by N proteins) at the ER–Golgi intermediate compartment (ERGIC) and viral progeny are released by exocytosis. Host immune responses are triggered by danger signals from infected cells or free virions, which are recognized by innate immune cells. Elevated proinflammatory cytokines and chemokines can be seen in asymptomatic to mild cases. Exuberated cytokine production resulting in cytokine storm exacerbates the severity of coronavirus disease 2019 (COVID-19). Lymphopenia (T and B cells) and neutrophil infiltration into infected sites contribute to the pathogenesis of COVID-19. Several CoV proteins have been reported to be capable of inhibiting the type I IFN signaling pathway. Figure created with BioRender (biorender.com).

Structural Insights into SARS-CoV-2 S-Receptor and -Antibody Interactions

Unique Furin Cleavage Site in the SARS-CoV-2 S Protein

The SARS-CoV-2 S protein has 1273 amino acids and can be divided in two domains, S1 and S2 (Figure 4A). S1 contains the receptor-binding domain (RBD) and facilitates attachment of the virus to host cells. S2 drives fusion of viral and host membrane and contains the fusion peptide [38], two heptad repeat regions, and the transmembrane domain that anchors S in the viral membrane [39]. The SARS-CoV-2 S protein trimerizes to form a metastable prefusion spike [Protein Data Bank (PDB): 6vsb] [40]; Figure 4A) in which S1 stabilizes S2, including the fusion peptide. Under this conformation, SARS-CoV-2 S1 and S2 domains are separated by a flexible loop containing a cleavage site that is exposed and accessible to host proteases (Figure 4A). Cleavage triggers irreversible conformational changes that are required for membrane fusion to occur [31,41,42].

Figure 4

Atomic Structure of Severe Acute Respiratory Syndrome-Coronavirus-2 (SARS-CoV-2) Receptor-Binding Domain (RBD).

(A) Schematic architecture of the SARS-CoV-2 spike glycoprotein. Degree of protein surface conservation between trimeric SARS-CoV-2 and SARS-CoV spike protein. A color range is shown with green and magenta representing not conserved and highly conserved, respectively. The atomic position of the cleavage site is indicated by an arrow. (B) Cartoon representation of a structural alignment of the SARS-CoV (gray) and SARS-CoV-2 (green) RBDs interacting with ACE2 receptor (orange) shown in surface representation. Corresponding footprints of SARS-CoV RBD and SARS-CoV-2 RBD overlaid on the ACE2 receptor are colored in gray and green, respectively, to illustrate overlap between both interaction sites. A differential loop between SARS-CoV-2 and SARS-CoV is indicated with a dashed circle. (C) Closed and opened conformation of the SARS-CoV-2 S with one of the RBD domains buried or exposed, respectively (yellow). Abbreviations: FP, Fusion peptide; HR, heptad repeat regions; RRAR, unique furin cleavage site; SP, signal peptide; TM, transmembrane domain.

Both SARS-CoV-2 S and SARS-CoV S are activated by enzymatic cleavage at two sites: S1/S2 and S2′. The S2′ cleavage sites are similar, but the S1/S2 cleavage sites differ. The earlier SARS-CoV encodes a single arginine [43] while SARS-CoV-2 encodes PRRAR [44]. Both viral S proteins are primed by the serine protease TMPRSS2 in human cells [29], while SARS-CoV-2 can additionally be cleaved by furin at its unique RRAR site. It has been hypothesized that furin-mediated precleavage at the S1/S2 site is important for subsequent S activation by TMPRSS2, as demonstrated previously for other CoVs [29,31]. The proline inserted before the RRAR cleavage site could promote addition of O-linked glycans [45]. However, the importance of this proline and any associated glycans are not yet understood. The S2′ cleavage site for SARS-CoV and SARS-CoV-2 (Arg 797 and Arg 815, respectively) is further required to mediate membrane fusion and entry [41]. Replication-defective vesicular stomatitis virus particles displaying either SARS-CoV-2 or SARS-CoV S proteins can infect the same range of cell lines, suggesting that the addition of a new cleavage site and proline do not change virus tropism [29].

SARS-CoV-2-S and ACE2 Interaction

The ability of SARS-CoV to infect a variety of species is linked to changes in the RBDs that affect ACE2 binding activity [46]. The RBDs of SARS-CoV-2 and SARS-CoV are 76% identical [47] and structurally very similar (Figure 4B), with similar binding interfaces between ACE2 and SARS-CoV (PDB: 2AJF) [46] and SARS-CoV-2 (Figure 4B) (PDB: 6M17) [48], however, the binding affinity of SARS-CoV-2 for ACE2 is 10–20-fold higher than SARS-CoV [40]. Structural studies captured the SARS-CoV-2 RBD in two different conformations: ‘opened’ when the RBDs are exposed and ready for the interaction with the receptor and ‘closed’ when the RBDs are buried in the interacting interface of the protomers and not accessible to the receptor [49] (Figure 4C).

Neutralizing Antibodies Compete for Binding at the RBD

The S glycoprotein is the main target for the protective humoral immune response (Figure 3). Antibodies against S are predicted to neutralize infection by blocking ACE2 binding to the RBD [50]. The crystal structure of SARS-CoV S RBD in complex with the human monoclonal antibody 396 (m396) illustrated that the antibody footprint overlaps with that of the receptor on the RBD [51]. m396 does not crossreact to SARS-CoV-2 since two critical residues, Ile 489 and Tyr 491, that are key to m396 binding, are not conserved. However, crossreactivity is noted for antibody CR3022 [52] and antibody 47D11 that target highly conserved epitopes in the RBD [53].

Host Immune Responses to SARS-CoV and MERS-CoV

Innate Immunity

Innate immunity serves as the first line of virus clearance and initiates adaptive immunity via cytokine/chemokine secretion. Dysregulation of cytokine production resulting in a cytokine storm is believed to be associated with disease severity. Elevated secretion of cytokines, particularly interleukin (IL)-2, IL-6, IL-10, IP-10, G-CSF, MCP-1, MIP1α, and TNFα were found in severe COVID-19 patient (Figure 3) [7,54,55]. Similarly, increased level of serum cytokines was found in SARS [56,57] and MERS patients [58]. Meanwhile, both signal transduction and production of type I interferon (IFN), the cytokine that limits viral spread by elevating neighboring cells to antiviral status, were delayed in SARS [57] and MERS patients [58]. Delayed IFN responses permit robust viral replication, accumulation of cytokine/chemokine-producing monocytes/macrophages, and increase in disease severity in SARS-CoV [59] and MERS-CoV [60] infected mice. Similarly, MERS-CoV M, ORF 4a, ORF 4b, and ORF 5 antagonized IFN pathways and, in turn, diminished type I IFN production [61]. Impaired type I IFN response and uncontrolled inflammatory response subsequently contribute to adverse disease outcomes. In fact, different type I IFNs have been shown to have antiviral effects against both SARS-CoV [62] and MERS-CoV [63] in vitro. However, there is no clear evidence that type I IFN therapies had direct benefit for SARS and MERS patients [62,64].

Adaptive Immunity

The second arm of host immunity against viral infection is adaptive immunity that involves T cell and B cell responses. CD4 T cells promote development of antibody responses, whereas CD8 T cells can directly kill virus-infected cells. Immunogenic CD4 and CD8 T cell epitopes in SARS and MERS patients were found to localize mainly to structural proteins, particularly the S protein [65,66]. Several T cell epitopes for SARS-CoV-2 have been predicted by computational analysis [25,67], although these predictions require additional validation in terms of their immunodominance in human populations. A summary of the experimentally confirmed immunodominant epitopes and their HLA-restrictions identified for SARS-CoV that could be used for eliciting crossreactive T cell responses against SARS-CoV-2 is shown in Table 2 . These epitopes are highly conserved between SARS-CoV-2 isolates. We aligned 93 S protein sequences and 103 N protein sequences from SARS-CoV-2 isolates. Among the 20 epitopes (Table 2), only two viral isolates contained a single amino acid substitution at two different epitopes (see the supplemental information online). Polyfunctional virus-specific CD8 T cells can be sustained in SARS patients for more than 1 year after recovery [68]. However, in terms of clinical features, COVID-19 (63%), SARS (80%), and MERS (34%) patients often exhibit lymphopenia with reduced numbers of CD4 and CD8 T cells (Figure 3) [7,69]. MERS-CoV-infected T cells undergo apoptosis mediated by both intrinsic and extrinsic pathways [70]. Further investigation is required to determine whether lymphopenia seen in severe COVID-19 patients is correlated with lymphocyte apoptosis. Moreover, MERS-CoV, but not SARS-CoV, can infect both CD4 and CD8 T cells from human blood and lymphoid organs via DPP4 receptor binding [71].

Table 2

Immunodominant T Cell Epitopes Identified in SARS-CoV

T cell epitopes	Protein	Peptide position	Sequencea	HLA-restriction	Refs
CD4 T cell immunodominant epitopes
	Spike	159–171	CTFEYISDAFSLD	HLA-DRB1*0401 and HLA-DRB1*0701	[95]
	Spike	166–178	DAFSLDVSEKSGN	HLA-DRB1*0401	[95]
	Spike	358–374	STFFSTFKCYGVSATKL	HLA-DR	[96]
	Spike	427–444	NIDATSTGNYNYKYRYLR	HLA-DR	[96]
	Spike	449–461	RPFERDISNVPFS	HLA-DRB1*0401	[95]
	Spike	729–745	TECANLLLQYGSFCTQL	HLA-DR	[96]
	Spike	1083–1097	SWFITQRNFFSPQII	HLA-DRB1*0401	[95]
	Nucleocapsid	346–362	N.A.b	N.A.	[97]
CD8 T cell immunodominant epitopes
	Spike	411–420	KLPDDFMGCV	HLA-A*02:01	[98]
	Spike	787–795	ILPDPLKPT	HLA-A*02:01	[99]
	Spike	940–948	ALNTLVKQL	HLA-A*02:01	[100]
	Spike	958–966	VLNDILSRL	HLA-A*02:01	[101]
	Spike	978–986	LITGRLQSL	HLA-A*02:01	[102]
	Spike	1042–1050	VVFLHVTYV	HLA-A*02:01	[99]
	Spike	1167–1175	RLNEVAKNL	HLA-A*02:01	[103]
	Spike	1174–1182	NLNESLIDL	HLA-A*02:01	[100]
	Spike	1203–1211	FIAGLIAIV	HLA-A*02:01	[102]
	Nucleocapsid	216–225	GETALALLLL	HLA-B*40:01	[104]
	Nucleocapsid	223–231	LLLDRLNQL	HLA-A*02:01	[99]
	Nucleocapsid	227–235	RLNQLESKV	HLA-A*02:01	[99]
	Nucleocapsid	317–325	GMSRIGMEV	HLA-A*02:01	[99]
	Nucleocapsid	331–347	N.A.	N.A.	[97]
	Nucleocapsid	346–362	N.A.	N.A.	[97]

Underlined sequences indicate identical amino acids between SARS-CoV (GenBank accession number: NC_004718.3) and SARS-CoV-2 (GenBank accession number: MN908947.3).

Peptide sequence was not included in the original article.

Neutralizing antibodies against SARS-CoV S can prevent viral entry and protect against SARS-CoV challenge [72]. In addition, passive transfer of convalescent sera from recovered SARS-CoV patients decreased viral burden in recipient SARS patients [73] and is being advanced as a potential treatment for COVID-19 [74,75]. The mean time for seroconversion in SARS patients was around 2 weeks after disease onset [76]. Whether neutralizing antibodies can offer protection from or limit the spread of SARS-CoV-2 infection is currently unclear. In rhesus macaques reinfected with SARS-CoV-2 28 days after prior challenge, no viral replication was observed and the animals exhibited no clinical signs, but the animals showed increasing titers of neutralizing antibodies after rechallenge [77]. These results suggest that prior SARS-CoV-2 infection could elicit protective immunity against subsequent virus exposure, although the long-term protection offered by neutralizing antibodies requires further study.

Vaccine Strategies

Vaccination with adenovirus-delivered SARS-CoV proteins, including N and the S protein S1 fragment can induce virus-specific neutralizing antibodies and nucleocapsid-specific T cell responses in rhesus macaques [78]. Immunization with vaccinia virus carrying full-length SARS-CoV-S reduced viral titer in BALB/c mice after SARS-CoV challenge [79]. For MERS-CoV, immunization with full-length MERS-CoV S protein delivered using recombinant vaccinia virus induced high levels of neutralizing antibodies and virus-specific IFNγ-producing CD8+ T cell responses [80]. The MERS-CoV RBD protein, specifically the S358-588 fragment, induced high immunogenicity and elicited a strong neutralizing antibody response in vaccinated mice and rabbits [81,82]. A recent review summarizes current vaccine studies for SARS-CoV and MERS-CoV [83]. Future vaccine efforts could be focused on enhancing mucosal immunity in the respiratory tract using optimized administration routes, antigens, and adjuvants to evaluate how vaccine-induced immune responses in the lungs correlate with protection.

Animal Models for Antiviral Discovery and Vaccine Development

To date, the most commonly used animal models for SARS are older (i.e., 12–14-month-old) BALB/c mice [84] and transgenic mice (K18-hACE2) that express the SARS-CoV receptor human ACE2 under the control of an epithelial cell-specific promotor on a C57BL/6 background. SARS infections in these mice are lethal [85]. Several other mouse strains, including C57BL/6, 129S Sv/Ev, and STAT1–/– mice have been reported to be susceptible to SARS-CoV infection [86]. Additionally, the use of a virus adapted to mice (SARS-MA15, SARS-CoV passaged 15 in BALB/c mice) produces clinical disease in young (6–8-week-old) BALB/c mice [87] that is similar to ARDS observed in humans [88]. For MERS, transgenic mice encoding human DPP4 showed viral replication with interstitial pneumonia [89]. Thus, older BALB/c mice, hACE2 transgenic mice, mice lacking one or more components of the IFN system, and mouse-adapted viruses will likely be important tools for developing mouse models of SARS-CoV-2 infection and disease. In fact, a recent study has already shown that hACE2 transgenic mice with SARS-CoV-2 infection reproduce the clinical symptoms of disease, supporting virus replication in lung tissue [90].

Concluding Remarks

SARS-CoV-2 represents the third HCoV, after MERS-CoV and SARS-CoV, to emerge in the 21st century. Although these viruses have had a marked impact on public health and the economy, no effective vaccine or treatment is available. The virological and immunological lessons from prior CoV outbreaks can guide us in understanding, treating, and eventually preventing COVID-19. In particular, evaluation of S protein molecular structures, neutralizing antibody responses, and immunodominant epitopes that elicit strong T cell responses will all be critical for the development of comprehensive vaccine strategies to fight emerging CoVs (see Outstanding Questions).

Which viral and host factors contribute to SARS-CoV-2 transmissibility and infectivity?

Which cells are permissive to SARS-CoV-2 infection?

Does prior SARS-CoV-2 infection protect against subsequent infection?

What are the magnitude and quality of innate immune responses to SARS-CoV-2?

What are the magnitude and quality of adaptive immune responses to SARS-CoV-2?

Which host factors allow for durable, protective immunity following SARS-CoV-2 infection?

Is there antibody-dependent enhancement of SARS-CoV-2 infection?

Which viral and host factors contribute to disease severity?

What are the immune response profiles in COVID-19 patients and do these profiles correlate to disease severity?

What are the mechanisms and sources of SARS-CoV-2 cytokine storm?

What are the most immunocompetent and relevant mouse models of COVID-19?

Can a peptide vaccine for SARS-CoV-2 be designed based on identification of immunodominant epitopes?

How do the differences in amino acid sequence, such as furin cleavage site, make differences between SARS-CoV-2 and other human coronaviruses?

Alt-text: Outstanding Questions