SARS-CoV-2 E protein: Pathogenesis and potential therapeutic development

Abstract

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has caused a devastating global pandemic, which has seriously affected human health worldwide. The discovery of therapeutic agents is extremely urgent, and the viral structural proteins are particularly important as potential drug targets. SARS-CoV-2 envelope (E) protein is one of the main structural proteins of the virus, which is involved in multiple processes of the virus life cycle and is directly related to pathogenesis process. In this review, we present the amino acid sequence of the E protein and compare it with other two human coronaviruses. We then explored the role of E protein in the viral life cycle and discussed the pathogenic mechanisms that E protein may be involved in. Next, we summarize the potential drugs against E protein discovered in the current studies. Finally, we described the possible effects of E protein mutation on virus and host. This established a knowledge system of E protein to date, aiming to provide theoretical insights for mitigating the current COVID-19 pandemic and potential future coronavirus outbreaks.

Graphical Abstract

1. Introduction

COVID-19, caused by SARS-CoV-2, broke out in December 2019 and has caused more than 518 million confirmed cases and 6 million deaths, according to the latest data from World Health Organization (WHO; https://www.who.int/emergencies/diseases/novel-coronavirus-2019). SARS-CoV-2 is the third coronavirus causing severe infection in humans after severe acute respiratory syndrome coronavirus (SARS-CoV-1) and middle east respiratory syndrome (MERS-CoV). All three of them are the enveloped, positive-sense, single-stranded RNA beta-coronavirus. SARS-CoV-2 contains a RNA genome and four structural proteins: Spike (S), Nucleocapsid (N), Membrane (M), and Envelope(E) protein. These four structural proteins ensure the formation of mature virions [1], [2]. S protein can interact with the human angiotensin converting enzyme II (ACE2) to mediate the invasion of the virus [3]. N protein is responsible for wrapping viral genomic RNA into helical structures in the cytoplasm and interacts with the M protein to direct the assembly and budding of mature virions [4]. M protein is the most abundant structural protein. It can interact with all other structural proteins and is thought to be central to virion assembly [5], [6]. E protein is the smallest structural protein, while still maintaining its mystique [7]. In infected cells, most E proteins are located at sites of intracellular protein trafficking, such as the endoplasmic reticulum (ER), Golgi apparatus, and endoplasmic reticulum-Golgi intermediate compartment (ERGIC), which have important roles in the viral life cycle [8]. In addition, E protein can also form an ion channel or viroporin, which may act as an independent virulence factor, inducing host death and triggering cytokine storm [9], [10]. This makes E protein a potential therapeutic target that may provide a promising therapeutic strategy for alleviating the current COVID-19 pandemic. For this purpose, we summarize the structure and targeted drugs of the SARS-CoV-2 E protein, and discuss its role in the viral life cycle and pathogenesis process as well as the development of its mutant strains.

2. Structure of E protein

E protein is a highly conserved structural protein present in all coronaviruses. It consists of three domains: a short hydrophilic N-terminal domain (NTD), a hydrophobic transmembrane domain (TMD), and a long hydrophilic C-terminal domain (CTD) [8]. In addition, the E protein of both SARS-CoV-1 and SARS-CoV-2 contain a binding motif called postsynaptic density protein 95 (PSD95)/Drosophila disc large tumor suppressor (Dlg1)/zonula occludens-1 protein (Zo-1) (PDZ) binding motif (PBM) [11], the last four amino acids at the CTD. The PDZ domain is a protein-protein interaction module that can bind to the CTD of target proteins such as cellular adaptor proteins of the host cell and appears to be involved in pathogenesis [12], [13], [14].

Comparative sequence analysis of the E protein of SARS-CoV-2 and other six known human coronaviruses did not reveal any large homologous/identical regions, with only the initial methionine, Leu39, Cys40 and Pro54 being generally conserved [15]. In terms of overall sequence similarity, the SARS-CoV-2 E protein had the highest similarity to SARS-CoV-1 (94.74%), followed by MERS-CoV (36.00%) [15]. Interestingly, the sequence similarity of coronaviruses that typically cause severe disease is significantly higher than that of coronaviruses that cause mild to moderate upper respiratory symptoms typical of the common cold. This illustrates the importance of E protein for disease development.

SARS-CoV-2 E protein consists of 75 amino acids and is a single-spanning membrane protein with a skewed distribution of charged residues on both sides of the membrane [16], [17]. Assembly of E proteins into the ERGIC membrane in the correct orientation (topology) is critical for their function. Studies show that topology expands the dimension of virus evolution. Topology can evolve by redistributing charged residues on both sides of the membrane. Alignment of MERS-CoV, SARS-CoV-1, and SARS-CoV-2 E protein revealed a tendency to accumulate a net positive charge balance in the CTD, suggesting stability of the topology from MERS-CoV to SARS-CoV-2 stronger [17]. This may also be one of the reasons why SARS-CoV-2 is more pathogenic and harder to control ( Fig. 1).

Fig. 1.

Structure of E Protein. Multi-alignment of amino acid sequences of the E protein of SARS-CoV-2 and two other large outbreaks of human coronaviruses. Conserved residues and PDZ-binding motif (PBM) are shown in green and orange, respectively. Differences between SARS-CoV-2 and SARS-CoV are shown in blue. (b) Phylogenetic data and (c) tree obtained with Clustal Omega (EMBL-EBI) using the default parameters.

3. Role of E protein in SARS-CoV-2 life cycle

SARS-CoV-2 uses ACE2 as a receptor for entry, which recognizes receptor-binding domain (RBD) of S protein [18]. Virus binding to ACE2, with other triggers, induces dramatic conformational changes in S protein that bring the viral and cellular membranes together, ultimately creating a fusion pore that allows the viral genome to reach the host cell cytoplasm.

The following SARS-CoV-2 life cycle inside the cell is similar to that of other coronaviruses [19]. The viral genome in the cytoplasm induces the formation of double-membrane vesicles (DMVs) derived from the ER, which ultimately integrate to form elaborate webs of convoluted membranes [20]. Here, the incoming positive-strand genome then serves as a template for full-length negative-sense genomic RNA (-gRNA) and subgenomic (sg)RNA. sgRNA further are translated into respective viral proteins, including N, S, M, and E protein. The viral structural proteins (S, E, and M) traffic through the ER to ERGIC. The N protein package subsequent positive-sense genomic RNA (+gRNA) into helical structures in the cytoplasm, and interact with hydrophobic M protein in the ERGIC that serve to direct assembly and budding of the mature virion [21], [22]. Finally, these +gRNA-encapsulated mature virions are transported to the cell surface in vesicles and then secreted through exocytosis from the plasma membrane [22].

The E protein plays a similar role in the life cycle of various coronaviruses. For all coronavirus, M and S proteins constitute the majority of the protein that is incorporated into the viral envelope of virions, while E protein is only a small portion [23]. Therefore, as the smallest of the major structural proteins, E protein has remained mysterious. But with the deepening of study, its mystery has gradually been revealed.( Fig. 2).

Fig. 2.

Role of E protein in SARS-CoV-2 Life Cycle.

3.1. E protein ensures progress of viral assembly

Coronaviruses assemble at intracellular membranes in the ERGIC, where E protein is mainly located in. E protein is oriented with its NTD in the lumen of the ERGIC and its CTD tail in the cytoplasm [23]. For SARS-CoV-1, the second structure of the β-hairpin of the E tail was speculated to be important for localization in the Golgi apparatus [24]. In fact, the Golgi position of E is disrupted when the motif responsible for the formation of this secondary structure is deleted [12]. When the mutation is engineered elsewhere in the tail, the location is preserved [24]. This is consistent with SARS-CoV-2 E localization in the ERGIC and Golgi apparatus [25].

In studying the interplay of SARS-CoV-2 structural proteins during the viral assembly process, Boson et al. [26], by combining biochemical and imaging assays in infected versus transfected cells, found that E protein induces S protein retention by regulating the cellular secretory pathway, whereas M protein-mediated S protein retention requires a CTD retrieval motif in the cytoplasmic tail of S protein. Interestingly, by inducing intracellular retention of SARS-CoV-2 S protein, the E and M proteins not only facilitate its targeting close to the virion assembly site, but also induce a specific maturation of N-glycosylation of S protein, limiting the fusion conformational activity and its cell surface expression ultimately prevent syncytia formation. This facilitates the next intrusion of the mature virions. Thus, E protein can regulate the localization and conformation of S protein by interacting with M protein, thereby ensuring the smooth progress of assembly.

3.2. E protein is critical for viral budding

E protein is known to be critical for the formation of viral protein-containing vesicles in SARS-CoV-1 and SARS-CoV-2 and occurs mainly in pentameric form. In fact, various recombinant coronaviruses lacking the E gene (ΔE) display an astonishingly aberrant morphology [8]. Deletion of the E proteins from SARS-CoV-1 (rSARS-CoV-1-ΔE) resulted in a 20- to 200-fold reduction in virus production [23]. Moreover, vesicles observed in rSARS-CoV-1-ΔE-infected cells contained dense, granular material interspersed between the virions, which is thought to be immature virions due to the aborted viral budding process [27]. In addition, when the C-terminal residue of mouse hepatitis virus (MHV) E protein was mutated to alanine, the virion became a contracted, elongated shape rather than the typical spherical particles observed in wild-type virions [28]. Virions of recombinant MHV-ΔE also exhibited a very similar aberrant morphology as small, irregularly shaped particles with jagged edges [29]. Clearly, viral budding can still proceed even in the absence of E, but the aberrant morphology of ΔE-virions strongly suggests that E is involved in the budding process. Most likely, instead of coordinating viral production, the function of the E protein is to regulate the fluidity of the viral envelope so that coronaviruses acquire their characteristic spherical shape and morphology. For SARS-CoV-2, also a member of β genus coronavirus like SARS and MHV, its role in the viral budding is conceivable.

To help elucidate budding, Collins et al. [30] performed atomistic molecular dynamics (MD) simulations using the Feig laboratory’s refined structural models of the SARS-CoV-2 E protein and M protein. This study shows that membrane curvature is mainly induced by the interaction of multiple M protein dimers, while E protein pentamers maintain the membrane planar [30]. Although E protein pentamer does not directly contribute to membrane curvature, a significant decrease in membrane curvature was detected in all simulations of E protein non-pentamer [30], [31]. A recent study emphasized the interaction between the various proteins of the virion and found that SARS-CoV-2 M and E protein interacted with each other through their CTD, and the E-M interaction also enhanced the M self-interaction [32]. These studies suggest that although the E protein may not induce membrane curvature directly, it may indirectly increase membrane fluidity by enhancing the self-interaction of the M protein, thereby promoting budding.

In contrast, Kuzmin et al. [33] have found that monomeric and pentameric E protein can directly generate membrane curvature, and this function can be ascribed to the amphiphilic CTD. The CTD of the E protein localizes to the convex region of the membrane, forming a potential budding site that facilitates virion budding [33].

Taken together, further studies are needed on whether the SARS-CoV-2 E protein can directly induce membrane curvature. However the contribution of E protein to budding is beyond doubt, especially its CTD.

3.3. What role does viroporin E play in viral release?

Viroporins are viral-encoded membrane pore-forming proteins that regulate cellular ion channels and are thought to play a role in multiple stages of the viral life cycle, from viral entry to assembly and release, and even pathogenesis [34]. The E protein of SARS-CoV-2 can self-oligomerize to form a pentameric ion channel, making it a viroporin [35]. And the E channel was shown to alter Ca²⁺ homeostasis in the cell and trigger inflammation processes, while calcium can modulate the transport properties of the E channel [36]. However, the jury is still out on whether SARS-CoV-2 E protein can trigger viral release through ion channel flux (direct) or membrane fission (indirect), or both. Study on the role of E protein in viral release is lacking.

Therefore, comparing the E protein of SARS-CoV-2 with the M2 protein of influenza virus may be beneficial to the understanding of the E protein of SARS-CoV-2. M2 is also a viroporin able to form ion channel and has been shown in many studies to affect the release of influenza viruses [34], [37], [38]. M2 can induce membrane scission and virion budding by altering membrane cholesterol [34]. We can reasonably speculate that the E protein of SARS-CoV-2 has a similar role. In fact, the M2 protein has an amphiphilic helix in its cytoplasmic tail that is complexly involved in membrane cleavage [38]. And this type of helix was also found to be present in the E protein of SARS-CoV-2 [16], strongly suggesting that promoting virion release may be one of the roles of E protein. Further studies are required to fully understand the role of SARS-CoV-2 E-formed ion channels on virion release, especially virion membrane scission.

4. E protein in host pathogenesis and COVID-19

In general, common coronaviruses tend to cause only mild upper respiratory symptoms and occasionally the gastrointestinal tract [22]. By contrast, infection with highly pathogenic coronaviruses, including SARS-CoV-2, causes severe "flu"-like symptoms that can progress to acute respiratory distress (ARDS), pneumonia, renal failure, and death [39], [40], [41], [42]. E protein, due to its properties such as the formation of ion channels, can act as a determinant for coronaviruses virulence and play an important role in its pathogenic process [9], [10], [43].

4.1. E protein affects host cell viability

E protein was mainly found in the ERGIC of cells transfected with a plasmid encoding E protein or infected with coronaviruses [44], where its NTD is translocated across the membrane of ERGIC, while the CTD is exposed to the cytoplasmic side [15]. In ERGIC membrane, E protein forms a cation-selective pentameric ion channel [9], [35], [45], which was lethal to host cells and even healthy surrounding cells [9].

Ion homeostasis is critical to the viability of cells, and the ion channels formed by E protein can disturb the electrochemical gradient of host cells and reduce their viability. In addition to monovalent potassium and sodium, E channels are also permeable to divalent calcium and magnesium. Calcium has been found to facilitate viral entry into host cells and is critical for Ebola and SARS-CoV-2 infections [46], [47]. Furthermore, after infection, E channels can immediately function to initiate the virulent factor and weaken the host cell defense by disturbing its ionic homeostasis [9]. These suggest that the E channel may facilitate further infection by SARS-CoV-2. In addition, the severity of COVID-19 patients was associated with electrolyte imbalances, including decreased serum potassium, sodium, and calcium concentrations [48]. Permeability of E channels to these ions may help explain the mechanism of electrolyte disturbance.

Previous studies on other coronaviruses have found that the E protein is capable of triggering the ER stress response and inducing apoptosis. In cultured cell lines, transfection of MHV E and epitope-tagged SARS-CoV-1 E induces apoptosis[49], [50]. However, in a recent study on SARS-CoV-2, apoptosis was not observed after transfection with plasmid encoding SARS-CoV-2 E protein (2-E). This may be due to the insufficient understanding of "programmed cell death" before, which refers to all self-regulated cell death as "apoptosis". With the deepening of the understanding of "programmed cell death", the original so-called "apoptosis" has been subdivided into apoptosis, pyroptosis, ferroptosis, copper death and so on. After transfection with the 2-E plasmid, the surface of dying cells expressing 2-E began to swell, eventually exploding and releasing the cellular contents. This suggests that 2-E induces pyroptosis-like cell death [9]. However, biomarkers of pyroptosis were not elevated. Interestingly, in another study, a significant rise in biomarkers of pyroptosis was clearly detected after treatment with 2-E purified protein [10]. The possible reason for this contradiction is the difference in experimental methods. The former [9] uses 2-E plasmid for transfection, while the latter [10] is treated with 2-E purified protein plus ATP. Regardless of whether 2-E induces host cell death through the pyroptotic pathway, existing studies have shown that 2-E has the ability to disrupt host cells. Following cell rupture induced by 2-E, a large number of SARS-CoV-2 virions, 2-E and other damage associated molecular patterns (DAMPs) can be released simultaneously. These further damages surrounding healthy cells. As for whether 2-E can trigger ER stress, there is still a lack of relevant study. But SARS-CoV-2 ORF3a, also a viroporin like 2-E, has been found to trigger sequential ER stress during SARS-CoV-2 infection [51]. Moreover, after 2-E is translated, it is firstly transported into the ER [12]. Therefore, the regulation of ER stress by 2-E is an area worth exploring.

4.2. E protein activates immune response

Severe disease caused by SARS-CoV-2 infection is associated with an improperly regulated cytokine response, resulting in immune-mediated pathology in the lungs and other tissues. 2-E possesses two completely-conserved key functional features, namely ion channel and PBM. These features play key roles in inducing cytokine secretion and activating inflammasomes [52]. Therefore, 2-E is an important study target for the immune response induced by SARS-CoV-2.

A comparative pan-genomic analysis [52] of all sequenced reference β-coronaviruses revealed that among all the core gene clusters in which these viruses are present, the variant cluster of E-protein ion channel and PBM is shared by SARS-CoV-1 and SARS-CoV-2. This suggests that SARS-CoV-1 and SARS-CoV-2 have similar mechanisms in activating immune responses. A previous study found that SARS-CoV-1 strains lacking the E protein were unable to activate the NF-κB pathway, thereby significantly reducing the production of inflammatory cytokines after infection in mice [53]. Recent studies on SARS-CoV-2 have reached a similar conclusion that the 2-E can elicit a robust immune response in vitro and in vivo [9], [10]. Deletion of the 2-E or a dominant-negative mutation significantly reduces the virulence of the virus [9].

Zheng et al. [10] found that 2-E could induce the release of inflammatory cytokines such as TNF-α and IFN-γ and the activation of the NLRP3 inflammasome, and also identified Toll-like receptor 2 (TLR2) as an innate sensor of 2-E. TNF-α and IFN-γ are key factors in increased cytokines in covid-19 patients. These two specific cytokines act synergistically to activate potent inflammatory cell death (PANopotosis), which leads to tissue and organ damage and death [54]. Therefore, the role of 2-E in the immune response may be greater than we realize. The induction of inflammatory cytokines by 2-E is achieved by activating ERK and NF-κb signaling pathways [10]. The activation of the NLRP3 inflammasome may be due to the ion channel properties of 2-E. The ion channel formed by 2-E allows ion transport, which may provide an activation signal for NLRP3 inflammasome assembly [43]. Therefore, 2-E may be able to provide initiation and activation signals for the NLRP3 inflammasome. In addition, Yalcinkaya et al. [55] found that the modulation of NLRP3 inflammasome by 2-E was different in different infection stages. It was concluded that 2-E may initially suppress the host NLRP3 inflammasome response to viral RNA, while potentially increasing the NLRP3 inflammasome response in the later stages of infection.[55] This enriches our knowledge of the body's immune response throughout the course of viral infection.

4.3. E protein-host protein interactions

E protein can not only mediate host cell rupture by interacting with a host cell receptor (such as TLR2), but also disrupt cell polarity by interacting with some connexins [56]. The best-studied interaction site is the PBM, which consists of the last four carboxy-terminal amino acids (DLLV) of the E protein. PDZ domains are common protein interaction modules that recognize short amino acid residues at the C-terminus of target proteins [57]. The PDZ domain exists in a variety of connexins, including the recognition of human cell junction protein PALS1 [12], [14], [58], [59], tight junction protein ZO-1 [13], adhesion junction protein syntenin [60], etc. It has been reported that PALS1, syntenin and ZO-1 can all interact with E protein. Among them, the most studied is PALS1.

PALS1 is a tight junction-associated protein that plays a critical role in establishing and maintaining epithelial polarity in various organisms [57]. It has been suggested that the interaction of SARS-CoV-1 E protein with the PALS1 is involved in the destruction of lung epithelial cells in SARS patients [11]. Comparing the affinity of the E proteins from SARS-CoV-1 and SARS-CoV-2 for the PALS1 PDZ domain by equilibrium and kinetic binding experiments detected an increased affinity for the SARS-CoV-2 E protein, which may be one of the reasons for the increased virulence of SARS-CoV-2 [59]. Chai et al.[12] determined the complex structures of PALS1 and SARS-CoV-2 E protein using cryo-electron microscopy (cryo-EM). The reported structure indicates that the E protein C-terminal DLLV motif recognizes a pocket formed by hydrophobic residues from PDZ and SH3 domains of PALS1. E protein recognizes the pocket of PALS1 and subsequently breaks the apical cell polarity complex formed by PALS1, Crumbs, and Pals1-associated tight junction protein (PATJ) [58]. Similarly, E protein also disrupts cell polarity when it interacts with ZO-1 [13] and syntenin [60]. This can lead to loosening and leakage of lung epithelial junctions. Leaky junctions may promote local viral spread and multiple types of immune cells into lung alveolar spaces.

5. Pharmacology of SARS-CoV-2 E protein

SARS-CoV-2 E protein plays an important role in virial pathogenesis, making it an excellent target for drug therapy. Therefore, E protein inhibitors can be used as drug candidates for COVID-19. This review summarizes the SARS-CoV-2 E protein inhibitors screened by current studies and briefly describes their pharmacological mechanisms.( Table. 1).

Table 1.

Pharmacology of SARS-CoV-2 E Protein.

Class	Name	Potential Mechanism	Chemical Structure		Purely computational study	Reference
Natural SARS-CoV-2 Protein Inhibitors	Sinapic acid(SA)	fit in the N-terminal pore of E protein			Yes	[62]
	Withania somnifera	bind to the hydrophobic domain in the pore region of E protein, and interact with the D chain and E chain in the pentamer structure of E protein; bind to N-terminal domain（NTD） of E protein, block its channel activity	a series of phytochemicals		Yes	[64]
	Rutin	produces hydrogen bonds with E protein, inhibite SARS-CoV-2 envelope formation, virion assembly, and viral pathogenesis			Yes	[67]
	Glycyrrhizic acid(GA) and β-boswellic acid	interacts with E protein, destruct the structure of E protein			Yes	[70]
Synthetic SARS-CoV-2 Protein Inhibitors	Gliclazide and memantine	inhibit viral replication through potential interactions with E protein and lysosomal function; inhibit E protein to block its ion channel activity			No	[75]
	Hexamethylene-amiloride (HMA) and amantadine	interact with the amino acid residues of CTD and form hydrogen bonds with the N15 amide side chain of E protein, inhibit ion channel opening, viral replication and virus-mediated inflammatory responses.			No	[80]
	Retinoic acid	binds to the lumen of E channel and interacts with 11 protein residues to form stable interaction of hydrogen bond and high binding energy, block the channel and inhibit the viroporin function of E protein			Yes	[82]
	ZINC23221929 and ZINC06220062(ZINC6220062)	has hydrophobic interaction or hydrogen bonding with E protein, inhibit the ion channel activity of E protein by binding to the residues			Yes	[85]
	Ursodeoxycholate(UDC) and chenodeoxycholate(CDC)	bind to the transmembrane domain of E protein through hydrogen bond, destroy the hydrogen bond between adjacent chains to loose the structure of E protein pentamer, destroy the structure of E protein and promote the entry of inhibitors into virus-infected cells			Yes	[87]
Others	doxycycline	form hydrogen bonding interactions with Leu31, Thr35, Val52, and Ser55 of E proteins as E protein inhibitor			Yes	[67]
	nimbolin A	bind to E protein as E protein inhibitors			Yes	[88]

5.1. Natural SARS-CoV-2 Protein Inhibitors

5.1.1. Sinapic acid

Sinapic acid (SA), a natural phenolic acid compound, is a derivative of cinnamic acid [61]. It has antioxidant, antitumor, anti-inflammatory, antibacterial and neuroprotective effects [61].

Orfali et al. [62] screened the antiviral activities of 20 phenolic derivatives in vitro, among which SA showed the most effective inhibition of SARS-CoV-2. Subsequently, they virtually screened all currently available molecular targets using a multistep in silico protocol. Further in-depth molecular dynamic simulation-based investigation revealed the essential structural features of SA antiviral activity and its binding mode with E-protein—perfectly fitting in the N-terminal pore of E protein. The structural and experimental results presented in this study strongly recommend SA as a promising structural motif for anti-SARS CoV-2 agent development.

5.1.2. Withania somnifera

Withania somnifera, a medicinal plant, is widely used in the Indian medical system. It contains a series of different phytochemicals and has a wide range of biological significance. More than 12 alkaloids, 40 withanolides and several sitoindosides have been isolated and reported from the plant [63]. In preclinical studies, it has shown the characteristics of antimicrobial, anti-inflammation, anti-tumor, anti-stress, neuroprotection, cardioprotection and anti-diabetes [63].

Alharbi et al. [64] screened four compounds CID 10100411, CID 3035439, CID 101,281,364 and CID 44,423,097 from Withania somnifera. These four compounds feature aromatic nucleus, which can bind to the hydrophobic domain in the pore region of SARS-CoV-2 E protein, and interact with the D chain and E chain in the pentamer structure of E protein, namely Arg74 (D), Asn77 (D), Thr43 (E), Cys46 (E), Ile59 (E) and Val60 (E). They can directly bind to NTD of E protein to block its channel activity and then interfere with virus replication.

5.1.3. Rutin

Flavonoids are natural substances with different phenolic structures. They can be used as antioxidants to reduce cellular oxidative stress so they feature anti-cancer, anti-inflammation and anti-virus [65]. Studies have shown that a variety of flavonoids can be used as inhibitors of SARS-CoV E protein [66].

Bhowmik et al. [67] identified potential drug candidates from 548 (natural and synthetic) antiviral compounds against SARS-CoV-2 structural protein, characterized the affinity between the selected drug ligand and SARS-CoV-2 structural protein through high-end molecular docking analysis, and analyzed the stability of drug-protein interaction by high-level simulation. The results show that rutin (a bioflavonoid) produces hydrogen bonds with the Leu31, Thr35, Ser55, Arg69 and Pro71 of E protein, and the antibiotic doxycycline forms the interaction of hydrogen bond with the Leu31, Thr35, Val52 and Ser55 of E protein. Rutin can stably bind to E protein and are the most effective E protein inhibitors among the selected ligands. These lead compounds may be potential molecules for drug development by inhibiting SARS-CoV-2 envelope formation, virion assembly, and viral pathogenesis.

5.1.4. Glycyrrhizic acid and β-boswellic acid

Glycyrrhizic acid (GA), a triterpene glycoside isolated from licorice. It has a wide range of pharmacological and antiviral activities against enveloped viruses including SARS-CoV. Computer simulation data have proved that glycyrrhizic acid can interact with SARS-CoV-2 S protein to inhibit viral infection [68]. β-boswellic acid, a pentacyclic terpenoid molecule found in Boswellia serrata, is an active ingredient with anti-inflammatory, anti-rheumatic and anti-cancer effects [69].

Fatima et al.[70] The results show that glycyrrhizic acid interacts with E protein at Arg61, Phe23, Tyr57 and Val25, and so does β-boswellic acid with E protein at Leu34, Leu37, Arg38 and Leu39, thus resulting in the structural destruction of E protein. Glycyrrhizic acid and β-boswellic acid show their strong affinity with E protein in the screened compounds. Further research on its clinical application is necessary.

5.2. Synthetic SARS-CoV-2 Protein Inhibitors

5.2.1. Gliclazide and memantine

Gliclazide, a second generation sulfonylurea widely used in the treatment of non-insulin-dependent diabetes mellitus (NIDDM), [71] mainly induces channel closure by combining with sulfonylurea receptor 1(SUR1) subunit of potassium channel sensitive to ATP to stimulate the insulin secretion of pancreatic β cells [72]. Memantine, an antagonist of N-methyl-D-aspartic acid(NMDA) receptor, can decrease glutamate toxicity by reducing calcium concentration. It is used in the treatment of Alzheimer’s disease [73], and has also been shown to inhibit viral replication through potential interactions with E protein and lysosomal function [74].

Tomar et al. [75] screened 372 compounds in MedChemExpress “Membrane Transporter/Ion Channel” module. According to the results of positive and negative genetic tests, gliclazide and memantine can inhibit the growth of K⁺-dependent bacteria by blocking the ion channel function of E protein. In conclusion, memantine and gliclazide can be used as effective inhibitors of E protein to block its ion channel activity. However, the interaction mechanism of gliclazide and memantine with SARS-CoV-2 E protein still needs further study.

5.2.2. Hexamethylene-amiloride (HMA) and amantadine

As viroporin inhibitors, amantadine and rimantadine can inhibit the M2 viroporin of influenza A virus [76], [77], and hexamethylene-amiloride (HMA) can inhibit the M2 viroporin of influenza A virus [78], [79] and SARS-CoV-1 E viroporin [66].

Through solution nuclear magnetic resonance (NMR) spectroscopy in micelles and molecular dynamics (MD) simulation, Toft-Bertelson studied the binding characteristics of amantadine, amantadine, HMA and E protein. The electrophysiological results showed that amantadine and HMA could block the ion channel activity of E protein, while amantadine could not [80]. Amantadine and HMA interact with the amino acid residues of CTD and their polar groups can form hydrogen bonds with the N15 amide side chain of E protein, whose lipophilic group is stabilized near L18. By modifying the structure of E protein, both drugs can inhibit ion channel opening, which in turn inhibits viral replication and virus-mediated inflammatory responses.

5.2.3. Retinoic acid

Retinoic acid, a synthetic or natural derivative of vitamin A, can regulate a variety of physiological functions of multiple organ systems. It is not only very important for normal immune ability but also participates in cell growth and differentiation. It has been successfully applied to various skin diseases, such as skin cancer, psoriasis, acne and ichthyosis [81].

Dey et al. studied four drugs, namely, retinoic acid, mefenamic acid, ondansetron and artemether as potential inhibitors of the 2-E channel by using molecular dynamic simulations, molecular docking and other methods [82] The experimental results show that retinoic acid binds to the lumen of 2-E channel and interacts with 11 protein residues to form stable interaction of hydrogen bond and high binding energy, indicating that it may block the channel and inhibit the viroporin function of E protein. The exploration of retinoic acid as a potential ion channel blocker of SARS-CoV-2 E protein may be a potential therapeutic strategy for the treatment of COVID-19.

5.2.4. ZINC database compounds

ZINC database [83] can be used to screen ligands of SARS-CoV-2 [84]. Mukherjee et al. found two ligands with high affinity for E protein with database screening, molecular docking and other methods, that is ZINC23221929 and ZINC06220062(ZINC6220062) [85].

The chemical name of ZINC23221929 is [2-[[(3 S)− 2,3-dihydro-1,4-benzodioxin-3-yl] methylamino]− 2-oxo-ethyl] (2 S)− 2-(1,3-dioxoisoindolin-2-yl)− 4-methylsulfanyl-butanoate. One aryl group in the structure has hydrophobic interaction with residues I46, L51 and P54 of E protein, and the other aryl group is involved in hydrogen bonding with S60 and hydrophobic interaction with C44 and V47. The chemical name of ZINC06220062 is 2-(2-amino-2-oxo-ethoxy)-N-benzyl-benzamide. The oxygen-containing functional group in its structure may participate in the key hydrogen bonding with S60 residue of E protein CTD, while phenyl group has hydrophobic interaction with residues V47 and C40. ZINC23221929 and ZINC06220062 can act as effective small molecule inhibitors of SARS-CoV-2 E protein, inhibiting the ion channel activity of E protein by binding to the residues.

5.2.5. Ursodeoxycholate and chenodeoxycholate

Ursodeoxycholate (UDC) and chenodeoxycholate (CDC) can reduce the cholesterol saturation of bile. They are commonly used in the treatment of gallstones [86].

Yadav et al. demonstrated by computational methods that CDC and UDC can stably bind to the transmembrane domain of SARS-CoV-2 E protein through hydrogen bond and other interactions to form a thermodynamically stable complex, in which T30 residue is the key residue to bind to CDC and UDC [87] CDC and UDC can also destroy the hydrogen bond between adjacent chains to loose the structure of E protein pentamer and allow a large number of CDC molecules to enter the membrane. In conclusion, these two drugs can inhibit the survival of SARS-CoV-2 virus by destroying the structure of E protein and promoting the entry of inhibitors into virus-infected cells.

5.3. Others

In silico approach can predict E protein inhibitors quickly and at low cost. Bhowmik et al. demonstrated that doxycycline can form hydrogen bonding interactions with Leu31, Thr35, Val52, and Ser55 of E proteins as E protein inhibitor [67]. In addition, nimbolin A, nimocin, 7-deacetyl-7-benzoylgedunin, 24-methylenecycloartanol, and cycloeucalenone are also predicted to bind to E protein, making them potential E protein inhibitors. Among them, the binding energy of nimbolin A with E protein is the highest [88].

6. Mutations of SARS-CoV-2 E protein

Detecting the evolutionary patterns of SARS-CoV-2 gene information is important for the diagnostic methods, vaccine design and drug development of COVID-19. A global investigation of SARS-CoV-2 genotypes revealed a large number of mutations in structural proteins, non-structural proteins, auxiliary proteins and untranslated regions [89]. Among the mutations in the SARS-CoV-2 genome, point mutations are the most common type, with other kinds of mutations occurring less frequently [90]. Point mutation can lead to missense mutations, causing changes in the encoded amino acids. Missense mutations on the coronavirus E protein may disturb channel activity and lead to viral attenuation. Introduction of small but critical attenuation mutations could be a strategy for the development of live attenuated vaccines.

The gene length of the E protein is 225 nucleotides. Wang et al. analyzed 15,140 SARS-CoV-2 genomes, of which only 52 mutations occurred in the E protein gene, indicating that its mutation probability is extremely rare [91]. In other studies of the SARS-CoV-2 genome, no mutations or very few mutations were found in the E protein gene, supporting this conclusion.[92] Missense mutations in E proteins is less than 0.5% of the SARS-CoV-2 genome [93].

All structural proteins of SARS-CoV-2 are essential for the production of intact virion. Among them, E protein and M protein are relatively conserved, indicating that mutations in them affect the integrity and life cycle of the virus, while more mutations are generated in S protein and N protein. Current studies suggest that the E protein has the lowest mutation frequency of the four structural proteins and has the strongest mutagenic resistance [94], [95], [96], [97], [98], [99]. Troyano et al. analyzed 101376 complete sequences of E protein obtained from GISAID, and the analysis results showed that the conservation of E protein amino acid was 99.98% and 142 types of amino acid mutations occurred at 65 positions among 75 amino acid residues, followed by L73F（122 sequences）, R69I（92）, P71L（68）, T9I（56）and V62F（52）[100]. In addition, no significant increase in mutation frequency over time was observed.

The highest probability of missense mutations in SARS-CoV-2 E protein is CTD, followed by TMD [101]. Some mutations in CTD can lead to changes in the properties of R group of amino acids and thus affect the interaction of E protein with host cell protein. L73F and D72Y mutations change the DLLV motif in the CTD to DFLV or YLLV, interfering with binding to the host cell PALS1 protein and thereby inhibiting SARS-CoV-2 infection [93]. In addition, any missense mutation in the TMD of E protein will destroy the functions of E protein [102]. F26L, L39M, A36V, and L37H mutations can block E protein ion channel activity and induce viral attenuation. Mutations in cysteine residues at positions 40, 43 and 44 of the TMD interfere with the ion transduction function of the SARS-CoV-1 E protein [103]. Transmembrane α-helical domain generates ion channel activity by forming transmembrane α-helices. F23A mutation not only kinks α-helices, but also weakens inter-helix interactions, which partially inactivates the SARS-CoV-1 E protein ion channels [104]. C34R and C44Y mutations disrupt the disulfide bond between E protein and S protein and weaken their structural basis [8], [104]. In November 2021, a new variant of SARS-CoV-2, the Omicron variant, emerged and spread in more than 26 countries around the world. The Omicron variant has higher transmissibility and incidence. Xia et al. analyzed the effect of the T9I mutation of 2-E channel in the protein sequence of this variant. The results show that T9I mutation causes loss of ion selectivity of 2-E channel and reduced its pH-sensitivity, while reducing cell death and decreasing cytokine expression.[105] The altered 2-E channel properties caused by T9I mutation may be the reason for less efficient release and lower cellular damage of Omicron variant.

As a relatively stable conservative protein, E protein has important biological functions in terms of structural integrity and toxicity to the host. In the future, its mutations should be continuously monitored so as to avoid the risks caused thereby. Meanwhile, its relative conservatism suggests that the protein has potential as a therapeutic intervention that can be further investigated.

7. Conclusion and perspectives

From the end of 2019 to the present, COVID-19 has been a global pandemic for more than two years. During this period of study on the culprit SARS-CoV-2, the multiple important functions of E protein were successively discovered, ensuring the production of mature and complete virions: assembly, budding and release, participation in the pathogenesis process: inducing host cell rupture, triggering immune response and disrupting cell polarity (protein-protein interactions). Currently, there are limited studies on the mechanism of the SARS-CoV-2 E protein in the pathogenesis. However, since the amino acid sequence (94.74%) and protein structure are highly similar to the SARS-CoV-1 E protein, previous studies on the SARS-CoV-1 E protein can serve as an excellent guide for study on SARS-CoV-2 E protein. In the future, the specific ways of E protein-induced cell rupture and the specific effects of E protein on ER and other organelles in host cells are all areas worthy of study.

With the discovery of its function, the E protein has become a potential drug target. Through today's efficient and high-throughput in silico method, many drugs or compounds (natural and synthetic) have been found to interact with the E protein and are expected to be potential therapeutics for COVID-19. However, in this struggle between humans and viruses, not only humans are constantly updating treatment methods, but SARS-CoV-2 are also fighting for their chances of survival through various mutations. E protein is a highly conserved structural protein, so its mutation frequency is extremely low. Because mutations in highly conserved proteins are associated with great risks for both the virus and the host (human). But for the few E protein mutants that can remain active, humans may not have enough tools to deal with it. For example, the recent Omicron variant has the characteristics of higher transmissibility and lower toxicity, and the T9I mutation of E protein may be one of the main reasons for its change. The outbreak of cases in various countries proves the lack of human response to Omicron. Therefore, the development of drugs with multiple key sites or conserved binding sites may be a better option. And we should be focusing more on how to prevent the pandemic than on looking for new treatments after the mutation happened.

In conclusion, the current study on E protein is still relatively lacking, especially the pathogenic mechanism involved in E protein and the impact of E protein mutation on the virus and the host. A more complete knowledge system of E protein is needed in further studies to deal with the current COVID-19 pandemic and potential future coronavirus outbreaks.

CRediT authorship contribution statement

Shilin Zhou: Conceptualization, Writing – original draft. Panpan Lv: Writing –revision. Mingxue Li: Writing – original draft. Zihui Chen and Hong Xin: Drawing the Figures and Tables. Svetlana Reilly and Xuemei Zhang: Writing – review & editing, Funding acquisition, Project administration.

Conflict of interest statement

The authors declare that there are no conflicts of interest.

Data Availability

Data will be made available on request.