2.1 Animal source

The COVID‐19 pandemic has gone through three phases: local outbreak, community transmission and large‐scale transmission.15 Up to this date, the first person who acquired the virus from an animal source (coined Patient Zero) still remains unknown. Based on phylogenetic analysis of available genome sequences, SARS‐CoV‐2 has vital sequence homology with MERS‐CoV (about 50%), SARS‐CoV (about 79%) and bat‐derived SARS‐like CoVs (bat‐SL‐CoVZC45 and bat‐SL‐CoVZXC21, about 90%).16–18 While bats are probably the natural reservoir of SARS‐CoV‐2 like its close relatives SARS‐CoV and MERS‐CoV, the Malaysian pangolin (Manis javanica) is suspected to be the intermediate host.19, 20 However, direct evidence suggesting pangolins as the animal source has not yet been reported as of this writing. Therefore, as to how and when SARS‐CoV‐2 became zoonotic and started infecting humans are still very unclear. From its natural host to intermediate host(s) then to Patient Zero, the transmission path of SARS‐CoV‐2 is largely ambiguous (Figure 1a).

2.2 Transmission

SARS‐CoV‐2 is mainly a respiratory virus. As such, the infected individuals present upper respiratory symptoms including coughing and sneezing. SARS‐CoV‐2 is mainly transmitted through contact with respiratory droplets from nose and throat of an infected individual or with contaminated fomites.21 Other modes of transmission include oral‐faecal route and aerosol transmission in healthcare facilities.22 The WHO had been extra careful in declaring the airborne transmission of the virus. On 9 July 2020, however, WHO finally acknowledged that SARS‐CoV‐2 is airborne and can stay in the air longer than previously thought. This is after several reports came out supporting airborne transmission.23–25 There is no evidence to support mother‐to‐child transmission so far.26 The transmission of SARS‐CoV‐2 has superseded its predecessors SARS‐CoV and MERS‐CoV. As of 23 August 2020, the number of confirmed cases is 23,025,622 with USA, India, Brazil and Russia accounting significant clusters (Figure 1b).

2.3 Clinical characteristics

The clinical characteristics of COVID‐19 have now been widely described due to open exchange of emerging data across different countries. Symptoms however are non‐specific and the disease presentation can range from absence of symptoms (asymptomatic) to severe pneumonia that can lead to death. The severity of the patient’s condition is confirmed mainly by clinical symptoms, laboratory tests and radiographic findings of the chest. Initial symptoms include fever, dry cough, muscle pain or fatigue.10, 21, 27 Some patients present dyspnoea, chest pain, diarrhoea or upper respiratory tract symptoms. The most common laboratory abnormalities were elevated C‐reactive protein, decreased lymphocyte count and increased lactate dehydrogenase.28 Similar to SARS‐CoV, SARS‐CoV‐2 can cause lung injury.29, 30 Other virologic and clinical features are summarized in Table 1.

2.4 Case fatality rate

While the virus can infect anyone regardless of age and sex, certain groups are at higher risk. These vulnerable groups include children, pregnant women, the elderly, healthcare providers, immunocompromised individual and others with comorbidities such as asthma, diabetes and heart diseases.31 In a large study analysing 72,314 case records from China, 87% of patients were between the age of 30 to 70 years old. Children below 10 years old only constituted 1% of the total cases, while young adults aged 20 to 29 comprised only 8%. In terms of the disease severity, majority of the cases (81%) were mild, while only 14% and 5% were severe and critical, respectively.

Although the global case fatality rate (CFR, the number of deaths divided by the number of confirmed) is about 4.01%, CFRs vary from country to country. Yemen has the highest CFR of 28.5% followed by European countries including Italy, UK, Belgium, Hungary and France with CFRs of 13.9%, 12.8%, 12.5%, 12.2% and 11.9%, respectively. Among the countries with lowest CFRs are Singapore (0.0%, 27 dead out of 56,031), Botswana (0.2%, 3 out of 1308) and Qatar (0.2%, 193 out of 115,956).32 This substantial cross‐country disparity of CFR can be attributed to several factors. In the UK, the risk of death is strongly associated with old age, lower socio‐economic status, gender with male having higher risk and comorbidities (diabetes, obesity, cancer, respiratory diseases, heart, kidney, liver, neurological and autoimmune conditions).33 In a worldwide cross‐sectional study analysing open access databases, mortality is associated with lower test number, lower government effectiveness, ageing population, fewer hospital beds and poor transport infrastructure.34 Aside from these demographic and socio‐economic factors, genomic variations of SARS‐CoV‐2 also skew the fatality rates. Two variants—ORF1ab 4715L and S protein 614G—had strong positive correlations with fatality rates. Certain HLA genotypes including HLA‐A*11:01 can also either increase susceptibility to SARS‐CoV‐2 infection or severity of COVID‐19.35 Fortunately, certain laboratory parameters such as increased levels of lactic dehydrogenase, lymphocyte and high‐sensitivity C‐reactive protein can be used to quickly predict which patients are at the highest risk of dying.36

2.5 Control strategies

Many countries adopted aggressive quarantine measures including massive lockdowns as their major control strategy to either stop the virus from spreading to other areas or to slow down the transmission such that the healthcare system can cope up. However, because this approach has severe impact to the economy, most countries were unable to sustain months‐long lockdown. The most effective strategy so far, as also advocated by WHO, is to ‘test, isolate and trace’. This strategy involves testing the highly susceptible member of the population and those with increased risk of exposure, followed by isolating the infected from the rest of the population and finally tracing other possible infections from the close contacts of a confirmed index case. When public health measures are effectively and timely instituted, the basic reproduction number (R₀) can be <1. An R₀ of <1 can indicate that a communicable disease like COVID‐19 is expected to die out eventually in the population.37 In late January 2020, the first estimated R₀ was 2.2 (95% CI, 1.4 to 3.9).38 The WHO is currently conducting a global serologic study, dubbed as Solidarity 2, to grasp a global understanding of the epidemiology of SARS‐CoV‐2. Among the expected results from this global cooperation are to determine the R₀ as well as other key epidemiological parameters including secondary infection rate, asymptomatic fraction of infection and incubation period.39

3.1 Mechanism of SARS‐CoV‐2 entry and infection

The binding of CoVs to host cell surface receptors and membrane fusion processes are mediated by the Spike (S) protein composed of two subunits (S1 and S2).3 In the case of SARS‐CoV‐2, the cleavage and activation of S protein are governed by the intracellular protease TMPRSS2 to generate unlocked, fusion‐catalysed forms on the cell surface. This promotes the early entry of the virus.40 Albeit some differences in the aa sequences of SARS‐CoV‐2 and SARS‐CoV S protein, the receptor binding domain region of the S1 subunit is highly similar and both use the same cellular receptor Angiotensin I converting enzyme 2 (ACE2) to enter the target cell.17, 41, 42 SARS‐CoV‐2 S protein shows affinity to full‐length ACE2 which can bind to SARS‐CoV‐2 in either open or closed conformation.43 The spike’s RBD and ACE2 complex further suggests that the binding capacity of SARS‐CoV‐2 S protein with ACE2 is significantly higher than SARS‐CoV conferring its higher infectivity.44, 45 Furthermore, the S2 subunit is highly conserved that contains a fusion peptide, a transmembrane domain and cytoplasmic domain.46 Particularly, the hepeptide repeat 1 (HR1) and hepeptide repeat 2 (HR2) at the S2 subunit play some eminent roles in fusion regulation between the virus and host cell membrane. The HR1 and HR2 interact to form a six‐helix bundle, making the virus and cell membrane approach and fuse (Figure 2).47, 48

It is reported that there are potential recombination of RBD and insertion of furin or TMPRSS2 cleavage sites with unique four aa insertion. This polybasic cleavage site is inserted at S1/S2 domains of S protein, which leads to the ‘RRAR’ furin cleavage site, making it different from bat coronavirus RaTG13 (with 96.2% genome sequence identity) and SARS‐CoV.17, 18, 49, 50 This allows effective cleavage by furin and other proteases, enabling SARS‐CoV‐2 to be more infectious.51 The extended structural loop containing basic amino acids between S1 and S2 domains could also endow SARS‐CoV‐2 fusion activation and entry properties.52 Acquisition of these cleavage sites may have conferred the virus higher capacity to infect new host cells and expand host range. However, studies using cell or animal models are needed to determine the functional consequences of furin cleavage site.

Theoretically, all organs with high expression of ACE2 are susceptible to SARS‐CoV‐2 infection. Liver damage, intestinal inflammation, renal and testis insufficiency and pancreatitis in COVID‐19 patients were reported to be related with high expression of ACE2 in cholangiocytes, gastrointestinal tract (small intestine and duodenum), urinary organs (kidney) and testis and pancreas.53–56 ACE2 is also highly expressed in the heart which may contribute to acute myocardial injury and chronic damage to the cardiovascular system.10, 57 Additionally, the ubiquitous expression of furin in many tissues potentially permits the virus to attack multiple organs.49 Therefore, a small molecule that inhibits or blocks the function of furin can be a potential treatment for COVID‐19.

Though ACE2 is mainly expressed in human type II alveolar cells (AT2), the area ratio of type I and type II alveolar cells in the alveoli is about 9:1. Type II alveolar cells expressing ACE2 account for only 1.4% of all AT2 cells.58 Given this low percentage, it is possible that other receptor‐assisted or unknown molecular interactions govern SARS‐CoV‐2 infection that explain its high infectivity. Other potential receptors, for example, DC‐SIGN (genotypes CD209), L‐SIGN (CLEC4M) and CD147 can also allow the virus to either capture and anchor onto host cells or bind to S protein,59, 60 but how these receptors are used during infection and whether other host factors are involved remain to be investigated.

3.2 Inflammatory and immune response of SARS‐CoV‐2 infection

After the virus enters the host cell, the process of SARS‐CoV‐2 replication, transcription, assembly and release is similar to that of SARS‐CoV (Figure 2). The rapid virus replication and cell damage are important factors that lead to acute inflammatory and immune responses. This rapid replication in the early stages of infection results in epithelial and endothelial cell apoptosis and increased vascular permeability, triggering the release of pro‐inflammatory cytokines and chemokines in large quantity.61 In severe cases requiring intensive care, the plasma levels of IL‐2, IL‐7, IL‐10, G‐CSF, IP‐10, MCP‐1, MIP‐1A and TNF‐α were higher.10 As with other viral infections, cytokine storm may have a major role in the pathogenesis of SARS‐CoV‐2 infection and is associated with disease severity.62, 63 The granulocyte‐macrophage colony stimulating factor is of special concern for it further activates CD14⁺CD16⁺ inflammatory monocytes and generates more IL‐6 and other inflammatory factors through a positive feedback mechanism.64 This eventually leads to severe immune damage to the lungs and other organs. Hence, finding the key cytokines induced by SARS‐CoV‐2 infection and blocking its signal transduction will greatly reduce the damage of hyper‐inflammatory response.

In a case report of a non‐severe patient with no respiratory failure and acute respiratory distress syndrome, the recruitment of immune cell populations was documented to occur in the patient’s blood before resolution of symptoms. Antibody‐secreting cells, follicular helper T cells, activated CD4⁺ and CD8⁺ T cells and IgM and IgG were detected in blood before symptomatic recovery and persisted for at least 7 days. This detection of IgM and IgG, the SARS‐CoV‐2‐binding antibodies, may prove to be beneficial in developing protective vaccine candidates.65 Contrary to previous reports, however, pro‐inflammatory cytokines and chemokines in this patient even when symptomatic were found to be minimal. This warrants further investigation involving larger patient cohort to determine whether these immune parameters could be used to predict disease outcome.

SARS‐CoV S protein was also previously shown to down‐regulate ACE2 and induce shedding of the ACE2 extracellular region.66 This decrease of ACE2 on the cell surface affects its related normal physiological functions in the lung, causing dysfunction of the renin‐angiotensin system (RAS) that leads to further enhancement of the inflammatory response.67 As for SARS‐CoV‐2, the binding with ACE2 might result in ACE2 depletion, inhibiting ACE2/Ang (1‐7)/Mas receptor pathway. This eventually results to the RAS being out of balance, leading to multi‐system inflammation.68, 69 Besides, the underlying disease mechanism may be related to antibody‐dependent enhancement (ADE). ADE occurs in some patients with early, sub‐optimal antibody activity that cannot completely clear the virus, but instead leads to persistent viral replication and inflammation.70 Previous studies on animal models have confirmed ADE in SARS‐CoV infection.71 ADE can promote taking in infectious virus‐antibody complexes, which is mediated by the engagement of Fc receptors expressed on different immune cells or other receptors.72, 73 This process is independent of ACE2 expression and can eventually enhance infection of target cells in COVID‐19.74 Therefore, the ADE in SARS‐CoV‐2 infection should not be neglected.

Based on the previous coronavirus infections, innate immune response can serve either protective or destructive roles. The protective innate immune response to viral infection mainly depends on the type I IFN response and its downstream cascade, which ultimately controls viral replication and later induces an effective adaptive immune response.75 SARS‐CoV‐2 infection can result in pyroptosis of macrophages and lymphocytes, causing a decrease in the number of lymphocytes.63 This is consistent with the clinical data of two patient cohorts (n = 99 and 83.2% in n = 1099) who showed reduced total lymphocytes accompanied with an increase of total neutrophils, serum IL‐6 and C‐reactive protein.76, 77 The levels of IL‐6, IL‐10 and TNF‐α are also higher, and are inversely correlated with CD4⁺ and CD8⁺ T cell count.78, 79 It is therefore logical to speculate that the parallel deviations in total lymphocytes and neutrophils count during SARS‐CoV‐2 infection is probably because SARS‐CoV‐2 induces an overdue type I IFN and loss of viral control at the beginning of infection.75 The recent research also suggested that ACE2 can function as an interferon‐stimulated gene (ISG) in human epithelial cells and is upregulated by type I IFN, and to a lesser extent type II IFN.80 IFN responses that induce ISGs are essential for host antiviral defence. Thus, SARS‐CoV‐2 can enhance its infectivity by inducing ACE2 expression. This finding is important for drugs that can reduce ACE2 expression such that siRNA nucleic acid drugs may be used to prevent and treat COVID‐19.

In the activation of type I IFN signalling upon RNA virus infection, the pattern recognition receptor retinoic‐acid inducible gene I (RIG‐I) and melanoma differentiation‐associated gene 5 (MDA‐5) play a key role. Previous studies revealed that post‐translational modifications (PTMs) including ubiquitylation, phosphorylation, deamidation, and SUMOylation can modulate RIG‐I function.81 Whether SARS‐CoV‐2 infection down‐regulates type I IFN pathway by tuning RIG‐I PTMs is worth pursuing as this may reveal crucial regulations of innate immune response. The down‐regulation of type I IFN response may exacerbate the inflammatory response, which may lead to immune failure and immunosuppression as a feedback regulating mechanism.

Taken together, the recognition and binding of S protein with ACE2 or other receptors and the recombination and insertion of RBD are the key factors for SARS‐CoV‐2 infection. Blocking these processes is an effective antiviral strategy. The phenomena of cytokine storm, RAS imbalance and ADE bring many challenges to the treatment of COVID‐19. We do not fully understand how SARS‐CoV‐2 invades human cells to replicate, infect and influence the immune system at present. The mechanism of virus‐host interaction and how to maintain the favourable balance among antiviral, anti‐inflammatory and immune responses need further clarifications.

4.1 Drug screening for SARS‐CoV‐2

Initial drug screening is a primary stage of drug development. The first and most intuitive approach is to use available broad‐spectrum antiviral drugs used to treat other viral infections whose efficacy and safety are already known.82 Interferon was endorsed officially to be used for clinical treatment at the early stage of COVID‐19 but it turned out to be unsatisfactory because of no specific antiviral effect and its potential to upregulate ACE2.80, 83

The second approach is to find effective inhibitors or drugs based on the available genome and the molecular model of related viral proteins. The catalytic sites of four enzymes of SARS‐CoV‐2 are highly conserved and homological with the sequences of SARS‐CoV and MERS‐CoV.84 Thus, applying the inhibitors for SARS‐CoV and MERS‐CoV might be viable options considering these viruses share similar biology. In this case, the inhibitor of the main protease (M^pro, also called 3CL^pro) Lopinavir/Ritonavir was considered but unfortunately no treatment benefit was observed beyond standard care in severe COVID‐19 patients.85 Among drug candidates that received much attention is Remdesivir (GS‐5734), a drug used to treat Ebola.86, 87 Remdesivir competes with the RNA‐dependent RNA polymerase for ATP substrate and interfere with the synthesis of viral RNA. It is also predicted to bind with the replication enzyme complex (nsp12‐nsp7‐nsp8) of SARS‐CoV‐2.88 However, the efficacy of remdesivir in vitro and in clinical trials is controversial and not yet widely accepted.89, 90 In a clinical trial in China, remdesivir failed citing no statistically significant clinical benefits but the US trial showed a promising result citing faster recovery time by 31%. In that randomised, double‐blind, placebo‐controlled, multicentre trial in China with 237 participants (158 to remdesivir and 79 to placebo), remdesivir use was not associated with a difference in time to clinical improvement.91 In the US trial and its subsites in other countries with 1059 participants (538 to remdesivir and 521 to placebo), those who received remdesivir had a median recovery time of 11 days as compared with 15 days in those who received placebo.92 Similar to the report from China, serious adverse events were also noted. Even among severe patients, remdesivir also appears to be ineffective. In a randomized, open‐label, phase 3 trial with 397 severe patients from 55 hospitals in different countries, no significant difference between a 5‐day course and a 10‐day course of remdesivir was found.93

The FDA‐approved immunomodulator chloroquine and its derivative hydroxychloroquine (HCQ) also received attention in the early months of the pandemic. These two drugs have been used for decades as treatment for malaria and autoimmune diseases. They were reported to both inhibit SARS‐CoV‐2 in vitro with HCQ being more potent and effective.89, 94 A docking study reveals that chloroquine works by combining with E‐channels and nsp3b.95 The first report claiming the effectiveness of HCQ supplemented with azithromycin for virus elimination was received with criticisms due to poor study design.96 Later a multinational registry analysis (n = 96,032) suggested that the use of HCQ was associated with decreased in‐hospital survival and increased frequency of ventricular arrhythmias when used for treatment of COVID‐19.97 However, The Lancet released a retraction notice citing that they cannot perform independent third party peer review and cannot therefore guarantee the reliability of the report. In a multicenter, randomized, open‐label, three‐group, controlled trial in Brazil involving 667 hospitalized patients with mild‐to‐moderate forms of COVID‐19, the use of HCQ alone or in combination with azithromycin did not show clinical improvement at 15th day as compared with the standard care.98 Even as postexposure prophylaxis, HCQ has no clear benefit too.99 On 17 June 2020, WHO finally announced that the solidarity trial for hydroxychloroquine is discontinued citing no reduction of mortality of hospitalised patients when compared with standard of care. For other potential drugs for SARS‐CoV‐2 previously used to treat other viral infections (HBV, HCV and HIV), they are described in details in another review.100

Screening small molecule drug candidates from existing compound library through high throughput virtual screening, molecular docking and in vitro experiments is also a powerful method. Of the possible drug targets, M^pro of SARS‐CoV‐2 has become an attractive choice. Several compounds and inhibitors were identified using computer‐aided, structure‐based and cell‐based drug design strategies. Among these include a Michael acceptor inhibitor N3, synthetic compounds 11a and 11b, the selenium‐based organic complex Ebselen, and peptidomimetic α‐ketoamides.101–103 Artificial intelligence and machine learning have also aided in accelerating the search of potent compounds. With BenevolentAI, Baricitinib was identified as potential treatment since it inhibits the key regulatory factor of cellular endocytosis AP2‐associated protein kinase 1 (AAK1). Blocking AAK1 can prevent cell entry and self‐assembly of viral particles within cells.104 From the traditional Chinese medicines (TCMs) compound library, several potentially effective compounds were also identified, including baicalin, Scutellarin, Hesperetin, Nicotianamine and glycyrrhizin.105 Compound screening with both computational methods and compound library can be done in a fast and high‐throughput way while the safety and efficacy still needs further investigation.

4.2 Other potential therapeutic options

The rapid development of biomedical sciences offers multiple strategies including vaccine, mAb, nucleic acid oligomer, peptide and small molecule drugs to control or even prevent SARS‐CoV‐2 infection.100 Because it usually takes time to develop new specific vaccines, alternative strategies should be considered. Due to the high similarity of viral protein RBD of SARS‐CoV‐2 and SARS‐CoV, the possibility of SARS‐CoV specific human mAb CR3022 being able to bind SARS‐CoV‐2 RBD should be explored as potential treatment for COVID‐19.106 A new recombinant protein with human IgG1 may also prove effective since SARS‐CoV‐2 uses the receptor ACE2.17 This recombinant protein exhibits cross reaction activity to SARS‐CoV‐2 in vitro experiments and can be a potential candidate for the diagnosis, prevention and treatment.106 Additionally, the clinical grade human recombinant soluble ACE2 (hrsACE2) can inhibit the viral load of SARS‐CoV‐2 in Vero cells by 1000–5000 times and in artificial cultured human tissues including human blood vessel and kidney organoids.107 The use of restorative whole blood and restorative plasma to treat infectious diseases is also an effective choice.108 In the current guideline of China, the plasma therapy is mainly used for patients with rapid disease progression or critically ill patients.109

TCMs have also performed satisfactorily in the treatment of COVID‐19. The flu drug ShufengJiedu Capsule has shown clinical efficacy along with Lopinavir/Ritonavir and arbidol.83, 110 Another TCM called Lianhuaqingwen can inhibit SARS‐CoV‐2 replication in Vero E6 cells and significantly reduce the production of pro‐inflammatory cytokines (e.g., TNF‐α and IL‐6) at the mRNA level.111 At present, TCMs adjuvant therapy showed potential efficacy by improving pathological state to some extent and immune function, enhancing the body’s ability to clear viruses and also reducing the damage caused by infection. However, mechanism‐based studies are required to fully characterize the mode of action of these TCMs as treatment of COVID‐19.