Elucidating the microscopic and computational techniques to study the structure and pathology of SARS‐CoVs

Abstract Severe Acute Respiratory Syndrome Coronaviruses (SARS‐CoVs), causative of major outbreaks in the past two decades, has claimed many lives all over the world. The virus effectively spreads through saliva aerosols or nasal discharge from an infected person. Currently, no specific vaccines or treatments exist for coronavirus; however, several attempts are being made to develop possible treatments. Hence, it is important to study the viral structure and life cycle to understand its functionality, activity, and infectious nature. Further, such studies can aid in the development of vaccinations against this virus. Microscopy plays an important role in examining the structure and topology of the virus as well as pathogenesis in infected host cells. This review deals with different microscopy techniques including electron microscopy, atomic force microscopy, fluorescence microscopy as well as computational methods to elucidate various prospects of this life‐threatening virus.

γ-coronavirus and δ-coronavirus mainly infect birds, fishes, and few animals (Pillaiyar, Meenakshisundaram, & Manickam, 2020). Although the identification of these viruses dates back to about 60 years, their prominence intensified with the epidemic outbreak of SARS-CoV in 2002/2003 surfaced from southern China. With 8,096 SARS affected cases and 774 deaths were reported to the World Health Organization (WHO), the fatality rate was estimated to be 9.6% (Chang, Yan, & Wang, 2020;WHO, 2014). After that, another epidemic outbreak was recorded in 2012 due to an analogous virus MERS-CoV that emerged in Saudi Arabia. It was observantly more fatal than the previous outbreak, claiming the lives of 858 out of the 2,494 laboratory-confirmed cases globally reported to WHO, reaching the mortality rate of 34.4% (Guarner, 2020;WHO, 2019;Zaki, Van Boheemen, Bestebroer, Osterhaus, & Fouchier, 2012). Although the fatality rate was high, CDC, 2020) and condition termed as coronavirus disease-19 , this outbreak was announced as a "Public Health Emergency of International Concern" on January 30, 2020 and declared as a pandemic by WHO on March 11, 2020, due to the speed and scale of transmission.
Till March 20, 2020, a total of 234,073 confirmed cases and 9,840 deaths due to COVID-19 were reported to WHO worldwide (WHO, 2020). The rate of increase was rather abrupt, with 100,000 cases registered over the initial 3 months, and new 100,000 cases reported within the next 12 days. Although the death rate remained very low (about 2-3%), the R 0 is very high, between 2.5 and 3 (Guarner, 2020). Thus, precautionary measures are to be taken immediately and seriously.
Understanding the structure of a virus is an important aspect of virological studies. The word "Corona" comes from Latin, meaning "crown" or "halo." One of the defining features of these RNA viruses is the presence of clove-shaped, dumbbell-shaped, or pear-like spiky projections that looks like a crown (Li, 2016;Siddell et al., 1983). Coronaviruses, in general, have a nearly spherical structure and are moderately pleiomorphic (Masters, 2006). There are four important structural proteins associated with the virus namely, envelope (E) protein, spike (S) protein, nucleocapsid (N) protein, and membrane (M) protein ( Figure 1) (Spaan, Cavanagh, & Horzinek, 1988).
Previous studies have shown that the S protein (150 kDa) is a fusion glycoprotein that facilitates the viral attachment to the host cell. This structure gives the crown-like characteristic feature to the virus and is also responsible for the hemagglutinin activity. M protein (25-30 kDa) is one of the most abundant structural protein and provides definitive shape to the virion particles (Liu & Inglis, 1991). E protein (25-30 kDa) triggers the assembly and release of the virus, and also contribute to viral envelop formation by interacting with M proteins. The N protein binds to the genetic material in a string type conformation. β coronavirus has hemagglutinin-esterase (HE) protein that aids the S protein-mediated cell entry and virus spread through the mucosa (Spaan et al., 1988). Coronavirus replication is initiated with the binding of virion particles to the receptors of the cells, further directing the translation of the viral genome in the cytoplasm and synthesis of membrane-bound proteins. These structural proteins incorporate into the endoplasmic reticulum (ER) and are transported to the ER-Golgi intermediate compartment (ERGIC). Later on, the encapsulation process occurs, leading to the budding of the clustered particles into the ERGIC, and producing viroid. These viroids are carried to the plasma membrane of the cells through the formation of smoothwalled vesicles or Golgi sacs (Masters, 2006). Microscopic techniques are conventionally practiced for bacterial and viral identification and classification. They are feasible for observing and confirming the presence of viral particles based on morphological characteristics. Oshiro, Schieble, and Lennette (1971) Masters (2006) found spherical of diameter ranges from 80 to 160 nm (Oshiro et al., 1971).
Previous studies have shown the use of a scanning electron microscope (SEM) for obtaining surface information and TEM for revealing inner components of the SARS-CoV particle. Also, their surface irregularities were investigated using atomic force microscopy (AFM).
Recently, it was reported that 2019-nCoV (diameter of 120 nm) can be easily identified under a TEM with its crown-like appearance, a feature unique to coronaviruses (Monteil et al., 2020;Prasad, Potdar, Cherian, Abraham, & Basu, 2020). Accurate reconstruction of the viral particles can help to improve the understanding of science behind these viruses. Table 1 compares various microscopy techniques for understanding the structure of SARS-CoV and its effect in host cells.
This review intends to provide an outline of various microscopic techniques used for investigating the structure and pathophysiology of coronavirus, as well as computational methods for the same.
F I G U R E 2 Thin-section electron microscopy images show the various development stages of coronaviruses in the human fetal diploid lung (HDFL) cells 24 hr after infection. Part figure (a) shows that coronaviruses are in the cisternae of the endoplasmic reticulum of a cell (part of cytoplasm). Part figure (b) shows that coronaviruses are in the perinuclear spaces in a cell. The virus particles are in spherical shape and its diameter ranges from 80 to 160 nm. Part figure (c) shows the formation of six particles inside a vacuole in the cytoplasm in various stages of budding process formation. Part figure (d) shows a tubular structure containing a dense material in cytoplasmic inclusion. Part figure (e) illustrates the relationship of virus particles to a cytoplasmic inclusion composed of tubular structures. Arrows in the figure point to structures of developing virus particles which also resemble the tubular structures of the inclusion. Part figure (f) shows that the cytoplasmic inclusion is composed of densely staining material around the tubular membrane. Source: This figure is adapted with permission from Oshiro et al. (1971) 2 | ELECTRON MICROSCOPY VISUALIZATION OF SARS-COVS Electron microscope (EM) uses an accelerated beam of electrons as the source of illumination to attain higher magnification and resolution up to a few nanometers of particle size, making them favorable for viral studies and diagnosis. The visualization and characterization of viruses have become easier with the intervention of EM in biology. EM played a crucial role in characterizing the causative virus during the SARS outbreak in 2002/2003 (Goldsmith & Miller, 2009;Lin et al., 2004) and the ultra-high resolution SEM was able to visualize the SARS-CoV in threedimensional (3D) with 10-20 nm spikes on the virion surface.
The initial identification of coronavirus during the 2003 outbreak has relied on tissue culture isolation followed by EM visualization. The  Lichtman & Conchello, 2005;Wang et al., 2008;Lu, Liu, & Tam, 2008;Manopo et al., 2005;Yuan et al., 2006;Yuan et al., 2005;Knoops et al., 2008 novel coronavirus that was successfully isolated from patients with SARS and identified them using TEM. The oropharyngeal sample from the patient was inoculated into Vero E6 mammalian cell lines (Ksiazek et al., 2003). The E6 cell lines were subjected to thin-layer electron microscopy and the images revealed typical coronavirus particles within the rough endoplasmic reticulum, specifically in cisternae, as well as in vesicles and several large clusters of extracellular particles were found attached to the surface of the plasma membrane.
Negative-stain electron microscopy images of these particles have elucidated that the virus particle dimensions were within the diameter

| VISUALIZATION OF SARS-COVS BY ATOMIC FORCE MICROSCOPY
The AFM is primarily used in the field of material sciences after its invention in 1986 (Eaton & West, 2010). The emphasis of AFM in biology surged about a decade later, due to its efficacy in studying wide ranges of biological materials such as cells, macromolecules, proteins, and nucleic acids (Parot et al., 2007). It has become an important tool to investigate surface topologies in life sciences research including cell morphology, tissue heterogeneity, bacterial and viral characteristics, and structure identification (Yang, 2004). One of the advantages of using AFM is the atomic and molecular resolution achieved which is close to that obtained by an EM (Lin et al., 2005). It can also be used for revealing the inner architecture by scanning the internal layers using chemical, physical, or enzymatic methods to expose the interior of the viruses. These applications have proved AFM to be an important tool in virology.
AFM has been used to study morphology as well as surface struc- The virus was initially thought to enter the cell either by fusing with the plasma membrane but later studies showed that viral entry also involves endocytosis (Sieczkarski & Whittaker, 2002). In a study, Due to this endocytic entry, the cellular rate of infection of the virus has shown to be increased. This finding also contributes to understand the pathogenesis of the virus and may lead to antiviral drug research.
The above study also pointed out that lipid rafts play an important role in endocytosis of the SARS-CoV. Further, the involvement of lipid rafts in SARS-CoV entry into the Vero E6 cell was studied. Lipid rafts are domains that concentrate membrane-associated proteins including receptors and signaling molecules including ACE2 receptors.
The confocal fluorescence images (as shown in Figure 8b) showed the co-localization of ectodomain (S1188HA) spike protein with raftresident ganglioside GM1 and hence lipid rafts play a significant role.
Thus, proving that SARS-CoV requires intact lipid raft for entering into the host cells (Lu et al., 2008). Apart from studying how the virus enters the cell, immunofluorescence was also used to investigate antibody response to the SARS-CoV and use it as an efficient detection method. A 441-700 amino acid domain (called the protein C) representing the S1 of the spike protein was identified to be responsible for the immune response.

Protein C domain expressing baculovirus was allowed to infect
Spodoptera frugiperda (Sf-9) cells and the cells were fixed on coverslips (forming the antigen part of immunoassay). It was found that there was fluorescence ring around the Sf-9 cells which were treated with virus-infected serum but was absent when treated with normal human serum as shown in Figure 8c. This proved that, protein C gives rise to an immune reaction. There was a clear difference between the spike protein based-immunofluorescence assay (IFA) and commercial IFA in BrdUrd than the 7a-GFP negative cells indicating that cells are being obstructed from entering the S-phase. It was also established that G0/G1 arrest is not an apoptotic inducer in SERS-CoV infected cells.
The study also concluded that the cell cycle arrest was via cyclin D3/Rb pathway (Yuan et al., 2006). SARS-CoV has shown the presence of various non-structural proteins between the S and E genes or This indicated that the virus affecting the host immune proteins may also affect the normal immune process (Kim, Cho, Lee, Kim, & Son, 2019). Further, it was presented that nCoV-2019 and the closest Bat relative exhibited identities of more than 85% along with the fully conserved genome (30 kb) (Ceraolo & Giorgi, 2020). It was also reported that humans possessed two T and B cell epitopes from Betacoronaviruses and 398 from SARS-CoV. For B cell epitopes, Discotope prediction was used and algorithms were used for T cell epitopes. B and T cell epitopes were observed to be conserved between 2019-nCoV and SARS-CoV based on phylogenetic analysis (Grifoni et al., 2020).
To obtain information based on the molecular level of structures of 2019-nCoV proteins and their function, computational predictions were used in addition to various experimentally solved structures. In 2016, Kirchdoerfer et al. reported that N-terminal domain of the spike glycoprotein subunit 1 of SARS-CoV was modeled using Modeller, a homology tool presents in UCSF Chimera, where bovine CoV NTD with PDB ID-4H14 was used as a template. The modeled structure was docked with Human Coronavirus I (HKU1) and refined using Rosetta. Clustering was performed on refined models dependent on pairwise RMSD. A model with the least energy was chosen for further refinement (Kirchdoerfer et al., 2016). Along with this model, C-Terminal Domain from SARSCoV with PDB ID-2AJF was used to build and refine the model. The available structure of the target gene Mpro (Main protease) was used to identify potential drugs for 2019-nCoV using molecular docking and the results of this study confirmed the earlier preliminary reports that some of the drugs approved for other viral infections can be used to treat 2019-nCoV infections (Talluri). Another study also focused on the same drug target, the main protease (Mpro, 3CLpro), and structures of 2019-nCoVMpro and its complex with an α-ketoamide inhibitor ( Figure 9a) were used in this study. The characterization of α-ketoamide based on its drug-likeness, ADME and toxicity properties reveals that it can be administered by inhaling .

Recent studies have identified Human Angiotensin-Converting
Enzyme 2 (hACE2) as a potential receptor for interacting with spike Free web-based server Fahmi et al., 2020 MEGA X Analyze evolution and build phylogenetic trees Freeware Fahmi et al., 2020;Zhou et al., 2020;Chan et al., 2020;Wu, Liu, et al., 2020;Wu, Zhao, et al., 2020 MAFFT Multiple alignment program Free web-based server Fahmi et al., 2020 CLUSTAL omega Multiple sequence alignment tool Free web-based server Chan et al., 2020;Ibrahim et al., 2020 Clustal W Progressive multiple sequence alignment Free web-based server Wu, Liu, et al., 2020;Wu, Zhao, et al., 2020 C-I-TASSER Server for protein structure prediction using multiple threading method LOMETS JPred Secondary structure prediction server Free web-based server Fahmi et al., 2020;Wu, Liu, et al., 2020;Wu, Zhao, et al., 2020 PyMOl Molecular visualization software Open source Ibrahim et al., 2020;Talluri, ;Kirchdoerfer Rasmol Molecular visualization software Open source Talluri, 2020 Enrichr Performs gene set enrichment analysis Free web-based server Zhou et al., 2020 TMHMM Transmembrane helices prediction server Free web-based server Chan et al., 2020;Wu, Liu, et al., 2020;Wu, Zhao, et al., 2020 ProtScale of murine monoclonal and polyclonal antibodies that interacted with the SARS-CoV spike protein did not interact with that of 2019-nCoV, implying distinct peculiarity of antigenicity in 2019-nCoV. Such distinctions provide profound insight for being responsible for the lifethreatening nature of 2019-nCoV and therefore, must be targeted to develop a therapeutic solution thereby inhibiting the pathogen . Spike glycoprotein interacting with human glucoserelated protein, GRP78 was demonstrated and four distinct regions of the spike glycoprotein were identified to be highly involved in an interaction with the candidate receptor as depicted in Figure 9b ( Ibrahim et al., 2020). Host cell surface binding takes place through the S1 subunit followed by viral fusion. Two domains of S1 subunit of various coronaviruses accept different host receptors. S1-CTDs recognize DPP4, APN, and ACE2 protein receptors. S1-CTD, in turn, consists of two subdomains: a core domain and a receptor-binding motif (RBM). RBM has a concave surface to mediate binding to ACE2.
RBMs in HCoV-NL63, PRCV, and SARS-CoV may have diverged into ACE1-binding RBDs. The spike protein is present in two different conformations, that is, prefusion and postfusion which undergo transition for membrane fusion. The prefusion structure was determined which was found similar to the influenza virus hemagglutinin. This provided information about the evolution of coronavirus S1. Betacoronavirus S1-NTDs consist of galectin folds which indicate coronavirus S1-NTDs as host origin. Hence, the evolutionary relationship between S1-CTD and host galectins is possible (Li, 2016).
The 2019-nCoV Docking Server was established to determine the binding affinity between the targets and small molecules, peptides, or antibodies using AutoDockVina and CoDockPP software for docking analysis. The server also helps to visualize the docked complexes and hence, an effective tool for 2019-nCoV drug discovery (Kong et al., 2020). Jin et al., 2020 claimed that a combination of structure-based drug design and high-throughput virtual screening can be effective in discovering new drugs to treat 2019-nCoV. Candidate drugs were found to bind at the binding site of the target protein Mpro. N3 inhibitor bound to the target protein Mpro at its binding cavity is depicted in Figure 9c. The binding pockets were observed to be situated between domains I and II (Figure 9d), which seemed to be highly conserved. Figure 9e depicts the key residues of Mpro involved in the interaction with N3 inhibitor and Figure 9f highlights the C-S covalent bond among other interactions. Therefore, binding pockets located between the domains I and II show good binding with antiviral inhibitors (Jin et al., 2020).
Wu et al. (Wu, Liu, et al., 2020;Wu, Zhao, et al., 2020)  however, it requires further validation. Anti-AIDS drugs, ritonavir, and lopinavir did not bind to the identified targets, thus, they cannot be used for the treatment of coronavirus infections. Future work will focus on the further validation of the activities of screened drugs, drug design, and in vivo and in vitro tests (Wu, Liu, et al., 2020;Wu, Zhao, et al., 2020). To evaluate receptor and receptor-ligand interaction, simulation mechanisms must be emphasized (Robson, 2020).

Molecular dynamics (MD) simulations conducted by Zhou et al.
suggested changing the intermolecular dynamics in protein-substrate complexes eliminates the mechanism underlying the protease activity.
The discovery of novel crucial residues for enzyme activity in the binding pocket could potentially provide more druggable sites for the design of protease inhibitors (Zhou et al., 2019). Recently, it is reported that hydroxychloroquine (HCQ) can efficiently inhibit SARS-CoV-2 infection in vitro . HCQ is safe and less toxic compared with chloroquine; however, an overdose of HCQ can cause poisoning and death.
In conclusion, the process of understanding the underlying mechanism of 2019-nCoV protein targets interacting with receptor and drug candidates at the molecular level is easier due to computational approaches. Table 2  Also, cryo-EM was used to determine the 3D structure of the HCoV-NL63 spike protein and its cellular receptor during infection. Along with these techniques, computational biology contributes to understanding the viral mechanism at the molecular and atomic levels paving a path for an alternative therapeutic strategy. Therefore, careful design and clinical trials are required to achieve efficient control of 2019-nCoV. guidance. We would also express our sincere gratitude to MAHE, Manipal for providing the infrastructure needed for the study.