Natural product-derived phytochemicals as potential agents against coronaviruses: A review

Graphical abstract

Open in a separate window

Highlights

Abstract

1. Introduction

The dramatic change of events with the recent unprecedented coronavirus pandemic declared by the World Health Organisation (WHO) has prompted an exponential increase of scientific interest in coronaviruses globally. As of April 22nd 2020, the pandemic has resulted in 2,553,112infections, with 177,286 deaths worldwide, which continues to drastically increase as we write (https://www.who.int/emergencies/diseases/novel-coronavirus-2019).

Coronaviruses (CoVs) belong to the family Coronaviridae, subfamily Coronavirinae and are large (genome size 26−32 kb; Wu et al., 2020a), enveloped, positive-sense single-stranded ribonucleic acid (RNA) viruses that can infect both animals and humans (Fig. 1 ). Based on their genotypic and serological characteristics, the viruses are subdivided into four genera: Alpha-, Beta-, Gamma-, and Deltacoronavirus (Chu et al., 2020; Lu et al., 2015). At present, all identified CoVs that are capable of infecting humans belong to the first two genera. These include the alphacoronaviruses (αCoVs) HCoV-NL63 (Human CoV-NL63) and HCoV-229E and the betacoronaviruses (βCoVs) HCoV-OC43 (Human CoV-OC43), HKU1 (Human CoV), SARS-CoV (Severe Acute Respiratory Syndrome CoV), and MERS-CoV (Middle Eastern Respiratory Syndrome CoV) (Lu et al., 2015). In the past two decades there have been three epidemics caused by the betaCoVs, namely SARS in 2002−03, MERS in 2012 and COVID-19, first identified in 2019 (Yang et al., 2020b).

Open in a separate window

Fig. 1

The general structure of a coronavirus (reproduced from Wikipedia under CC licence 4.0). E protein = envelope protein.

SARS-CoV emerged in 2002−03 in Southern China, causing a global threat and infecting more than 8000 people, with approximately 800 fatalities recorded, largely in China and the surrounding regions (Lu et al., 2015; Paraskevis et al., 2020). MERS-CoV emerged in the Middle East, spreading to several countries to infect close to 2300 individuals, resulting in 845 deaths as of July 2019 (World Health Organization, 2019). The present CoV pandemic resulting from SARS-CoV-2, which causes COVID-19 (coronavirus disease), was identified in Wuhan City, in the Hubei province of southern mainland China on the 31st December 2019 (Sohrabi et al., 2020). The genome of SARS-CoV-2 is approximately 70 % identical to that of SARS-CoV (Hui et al., 2020), hence leading to its current name.

The major druggable targets of SARS-CoV-2 include 3-chymotrypsin-like protease (3CLpro), papain like protease (PLpro), RNA-dependent RNA polymerase, and spike (S) proteins (Wu et al., 2020b). The S proteins interact directly with human angiotensin-converting enzyme (ACE) 2, allowing the virus to enter the cells. At present, no preventive vaccines or established antiviral therapies are available for coronaviruses (Sohrabi et al., 2020). However, several synthetic compounds have shown promise, including hydroxychloroquine and choloroquine phosphate (Cortegiani et al., 2020; Gao et al., 2020), which act through several mechanisms, including alkalisation of the host cell phagolysosomes. Newer antiviral medications such as lopinavir (Yao et al., 2020), remdesivir (Holshue et al., 2020; Wang et al., 2020), and arbidol (Khamitov et al., 2008) also show promise. Other suggested treatment options include lopinavir/ritonavir, nucleoside analogues, neuraminidase inhibitors, and peptide EK1 (Lu, 2020). A detailed list of current and planned clinical trials investigating various drugs for the treatment of SARS-CoV-2 was provided by Pang et al. (2020), with updated results available from ClinicalTrials.gov (2020).

In addition, traditional herbal medicines and purified natural products may guide the development of novel antiviral drugs. In other words, more efficient drugs can often be designed based on the structure of natural compounds that exhibit the desired activity. Classic examples of this drug discovery pathway include emetine, an isoquinoline alkaloid isolated from Cephaelis ipecacuanha and used as an amoebicidal drug; quinine, derived from the bark of Cinchona trees; and numerous other drugs modified from natural compounds, such aspirin, morphine and paclitaxel, an antineoplastic drug used for the treatment of cancer (Ganjhu et al., 2015). Indeed, half of all drugs approved between 1981 and 2014 were derived from or mimicked a natural compound (Newman and Cragg, 2016). Furthermore, in the current outbreak of COVID-19, many patients appear to be turning to complementary or traditional medicinal therapies, albeit using them almost exclusively in conjunction with western medicine. For example, one study suggested that almost 92 % of 135 hospitalised patients in northeast Chonqing (China) received traditional Chinese medicine in addition to western medicine (Wan et al., 2020). However, based on the many studies conducted on this topic, it is hard to separate the potential effects of, and interaction between, traditional Chinese herbal medicine and western medicine. Recent reviews have suggested that traditional Chinese medicine could be used for the prevention (Luo et al., 2020) or treatment (Yang et al., 2020a) of COVID-19; while still acknowledging that many studies involving clinical trials are poorly designed or controlled, and the choice of treatments is largely empirically based. As previous work has highlighted the potential of traditional Chinese medicines as a source of potential novel drugs (Ling, 2020), we have not included details on such studies investigating the antiviral activity of remedies comprising portions of numerous plant species in this review. Rather, our aim is to collate data on the broad spectrum of natural phytochemicals from individual plant species that may have therapeutic potential.

Naturally occurring antiviral agents acting against general coronaviruses were briefly reviewed by Lin et al. (2014) six years ago, while more recent reviews by Pang et al. (2020) and Lu (2020) on therapies for COVID-19 made only brief mention of natural therapeutics and did not explore the active compounds or their mechanism of action. In light of the current COVID-19 pandemic, this review aims to gather and consolidate information on extracts and compound(s) derived from natural products which show potential antiviral bioactivity for the inhibition of coronaviruses. It is hoped that the information presented may guide the naturally-derived drug discovery process in finding a treatment for SARS-CoV-2.

2. Methods

The PubMed database (www.ncbi.nlm.nih.gov/pubmed/) was used to locate articles including the following combination of terms: (coronavirus, SARS OR MERS) AND (traditional medicine, herbal, remedy OR plants). Papers primarily focussed on the antiviral activity of prepared Chinese traditional medicines, which typically comprise multiple plant species, were considered out of scope of this review. All articles up to and including 25 March 2020 were considered, yielding a total of 659 results. Two of the authors independently screened the results and identified relevant articles, yielding a total of 35 primary articles on human coronaviruses and 22 on animal coronaviruses were found to be pertinent and thus included in this review (total = 58 papers). Of these, two papers (6 %) were on SARS-CoV-2 (COVID-19). The majority of studies on human coronaviruses (69 %; n = 24) included SARS-CoV, with only 3 (9 %) including MERS-CoV and 8 (23 %) other human coronaviruses. It should be noted that one study included both SARS-CoV and MERS-CoV (O’Keefe et al., 2010) while another included both MERS-CoV and HCoV-229E (Müller et al., 2018), hence these percentages do not add to 100 %.

3. 

Table 1 summarises the studies reporting the inhibition of various human coronavirus strains using compounds derived from plant sources. The table is arranged by viral strain in order to better compare the bioactivity of compounds from different studies upon the same viral genotypes. Where identified, the key compounds responsible for the antiviral activity and their identified mechanisms of action are presented. It should be noted the term EC₅₀ (effective concentration) applies to cell-based assays, while IC₅₀ (inhibitory concentration) applies to enzyme- or biochemical-based assays.

3.2. Inhibitors of SARS-CoV

Given the relatively large amount of research that has been performed searching for inhibitors of SARS-CoV, antiviral agents that successfully inhibit this viral strain may provide a good starting point for identifying compounds that are active against SARS-CoV-2.

3.2.1. Virtual screening

Several authors have utilised virtual computer docking models to screen for potential compounds that could bind to and inhibit key proteins present in SARS-CoV (Liu and Zhou, 2005; Toney et al., 2004; Wang et al., 2007), highlighting the potential antiviral activity of compounds such as sabadinine and aurantiamide acetate. Compounds may be screened against a number of binding sites in order to test for potential inhibition of coronaviruses; the main sites that are typically used are the chymotrypsin-like protease (3CLpro), papain like protease (PLpro), spike proteins and RNA-dependent RNA polymerase.

3.2.2. Compound library screening

Several large in vitro screening studies searching for inhibitory activity of naturally occurring compounds against SARS-CoV have been performed, mainly on Chinese medicinal herbs (Li et al., 2005; Wang et al., 2003). While the results highlight the potential of selected plant extracts against SARS-CoV, they also demonstrate that such work can be akin to searching for a ‘needle in a haystack’. For example, Li et al. (2005) screened over 200 ethanol/chloroform extracts of Chinese medicinal herbs and found only four (Lycoris radiata, Artemisia annua, Pyrrosia lingua and Lindera aggregata) with moderate to high antiviral activity using a CPE assay (EC₅₀ values ranging from 2.4 ± 0.2–88.2 ± 7.7 μg/mL). Of these, a single compound (lycorine) from one plant species (L. radiata) was earmarked as a potential drug candidate against SARS-CoV. The antiviral efficacy of lycorine was quite high (EC₅₀ of 15.7 ± 1.2 nM), with a selectivity index greater than 900. Although the authors did not make mention of this fact, lycorine can cause toxic effects at low dosage levels (around 1 mg/kg in dogs) (Kretzing et al., 2011), hence great caution would be required for further development of this compound as a drug candidate.

Yu et al. (2012) screened the activity of a library of 64 naturally occurring compounds against SARS-CoV helicase, which plays a key role in the viral genome replication, transcription, and translation. The polyphenolics myricetin and scutellarein (Fig. 2 ) were identified as the most promising candidates (IC₅₀ values of 2.71 ± 0.19 and 0.86 ± 0.48 μM, respectively). Although the antiviral activity of the compounds was not assessed in cell-based assays, the authors did report that neither compound was toxic to normal (non-tumorigenic) breast epithelial cells. Myricetin is found in reasonably high concentrations in fruits such as cranberry (Häkkinen et al., 1999) as well as in several vegetables such as Calamus scipionum and garlic (Miean and Mohamed, 2001). Scutellarein was isolated from Scutellaria baicalensis (Chinese Skullcap), which has been traditionally used in the treatment of inflammation and respiratory infections, amongst other uses (Zhao et al., 2016). Both compounds were found to inhibit SARS-CoV helicase (nsP13) through the inhibition of ATPase activity, but did not directly inhibit helicase activity. This appeared to be the only publication identifying naturally occurring compounds as inhibitors of SARS-CoV helicase.

Open in a separate window

Fig. 2

The structure of selected naturally occurring compounds that demonstrate promising anti-coronavirus activity. (1) Quercetin (2) Quercetin 7-rhamnoside (3) Myricetin (4) Psoralidin (5) Caffeic acid (6) Tryptanthrin (7) Lycorine (8) Scutellarein (9) Silvestrol (10) Saikosaponin B2 (11) Isobavachalcone (12) Griffithsin. As annotated on (4), note the aromatic rings and substituted fused rings present in most compounds.

The need to isolate and synthesise more active agents as part of the development pipeline was highlighted in a study by Runfeng et al. (2020). The authors reported that the Chinese herbal medicine Lianhuaqingwen (comprised of a mixture of plant species) showed antiviral activity against SARS-CoV-2; however, the EC₅₀ was quite high (∼411 μg/mL). For comparison, the commercial drug remdesivir showed an EC₅₀ of 0.651 μM (approx. 0.39 μg/mL) using the same assay (Runfeng et al., 2020). This also underscores the potential lack of potency in some traditional medicines promoted for the treatment of coronavirus symptoms.

3.2.3. Polyphenols

One family of compounds that demonstrate antiviral activity across a number of studies is the polyphenols. For example, quercetin (Fig. 3 ) showed an IC₅₀ of 8.6 ± 3.2 μM against SARS-CoV PLpro (Park et al., 2017). No cell-based assay of antiviral activity was performed. Quercetin (Fig. 2) is a flavonoid found in many foods, but in particularly high levels in certain berries and herbs (Justesen and Knuthsen, 2001; Kaack and Austed, 1998). As previously mentioned (section 3.2.2), the structurally similar polyphenolics myricetin and scutellarein (Fig. 2) display reasonable levels of inhibitory activity against SARS-CoV helicase (Yu et al., 2012).

Open in a separate window

Fig. 3

Possible binding sites of quercetin in SARS-CoV-2 3CLpro and tryptanthrin in PLpro. Docking was performed in AutoDock Vina 1.1.2 (Trott and Olson, 2010) against target proteins generated by SWISS-MODEL (https://swissmodel.expasy.org/repository/species/2697049).

Bioassay-guided fractionation of the ethanolic extract obtained from Psoralea corylifolia seeds has also identified polyphenolics as the bioactive compounds responsible for the activity of this plant species against SARS-CoV PLpro (Kim et al., 2014). Furthermore, six phenolic phytochemicals were isolated from the ethanolic extracts - identified as bavachinin, neobavaisoflavone, isobavachalcone, 4′-O-methylbavachalcone, psoralidin and corylifol A - with their antiviral activity varying widely (IC₅₀ values between 4.2–38.4 μM). Again, no cell-based antiviral assays were performed. Isobavachalcone and psoralidin (Fig. 2) showed the greatest antiviral activity, with both found to be mixed, reversible inhibitors of PLpro through a type I mechanism (i.e. preferentially bind to the free enzyme, rather than the enzyme substrate complex) (Kim et al., 2014).

3.2.4. Lectins

Plant lectins, which are proteins that can bind specifically and reversibly to carbohydrate groups (Mitchell et al., 2017), are another group of naturally occurring compounds that may inhibit SARS-CoV. Lectins have shown promise as antiviral agents against viruses such as influenza and herpes simplex virus (Hwang et al., 2020), as well as Ebola (Covés-Datson et al., 2019; Michelow et al., 2011). Remarkably, elevating the plasma levels of recombinant human mannose-binding lectin in mice allowed them to survive otherwise fatal Ebola infections (Michelow et al., 2011). Keyaerts et al. (2007) screened the activity of a broad range of plant lectins (33 in total) against SARS-CoV using a cytopathicity effect (CPE) assay, finding EC₅₀ values as low as 0.45 ± 0.08 μg/mL for Lycoris radiata agglutinin. Although the exact mechanism of action was not determined, activity at the stage of viral attachment or the end of the infectious viral cycle were deemed to be the most likely targets. In clinical trials, other lectins have demonstrated reasonable to good tolerability (Petersen et al., 2006), hence with further testing, they may prove to be one of the more promising classes of naturally derived compound(s) for the treatment of SARS-CoV-2 and other coronavirus infections.

3.2.5. Increasing potency via chemical modification

Whilst many natural derived compounds show considerable promise for the inhibition of SARS-CoV and other human coronaviruses, few approach the level of efficacy and/or selectivity required for commercial drugs. For example, the calpain inhibitor MDL28170 has been shown to have an IC₅₀ value of just 2.5 nM (∼1 ng/mL) in its activity against the SARS-CoV enzyme cathepsin-L (Simmons et al., 2005). Cathepsin-L is a cysteine protease which is an important lysosomal endopeptidase enzyme involved in the initiation of protein degradation (Sudhan and Siemann, 2015) and mediates entry and infection of SARS-CoV in its host cells (Huang et al., 2006). Chemical modification of naturally-derived compounds may be required to increase the potency of their antiviral activity to levels suitable for therapeutic application. For the drug discovery process, beginning with the understanding of the structural conformity (i.e. structure including potential isomers) of the naturally derived compound(s) can reduce timelines and greatly increase the chance of finding effective viral inhibitors.

Working on this principle, Hoever et al. (2005) found that modification of glycyrrhizin, naturally found in liquorice and previously used for the treatment of SARS-CoV (Haiying et al., 2003), could increase its viral inhibition activity. For example, adding 2-acetamido-β-D-glucopyranosylamine to the glycoside chain of glycyrrhizin increased its antiviral activity in a CPE assay by 10-fold (decrease in EC₅₀ from 365 ± 12 μM to 40 ± 13 μM), through increased attraction to the S proteins. Amides and conjugates with amino acid residues and free COOH increased activity of glycyrrhizin by up to 70-fold (EC₅₀ for CPE assay ranging from 5 ± 3 μM to 139 ± 20 μM), albeit with significantly reduced selectivity. Cho et al. (2013) demonstrated that tomentins A–E (all naturally occurring compounds) showed viral inhibition activity against SARS-CoV greater as compared to their non-geranylated precursor compounds. Similarly, quercetin-7-rhamnoside (Fig. 2) demonstrates over 100 times higher antiviral activity against an animal coronavirus strain compared to its parent compound, quercetin (Choi et al., 2009).

While the examples of modifications presented here vary, it should not be considered that substitutions are added in a random fashion, but rather performed in the context of targeting a specific receptor or biochemical pathway. For example, the addition of glycosides and specific amino acid residues to glycyrrhizin was performed with the intent of increasing its affinity for the highly glycosylated spike proteins of SARS-CoV. With this in mind, future studies that identify an effective natural inhibitor of coronavirus should also consider the design and testing of potential modifications or synthetic derivatives that could increase its desired activity.

3.2.6. The significance of viral strains

A difference in the efficacy of natural therapeutics, mirroring that observed for commercially/synthetically available drugs, has been observed between coronavirus species/strains. For example, Park et al. (2017) found that the compound 3′-(3-methylbut-2-enyl)-3′,4,7-trihydroxyflavane, isolated from Broussonetia papyrifera, showed good inhibition of PLpro from SARS-CoV (IC₅₀ of 3.7 ± 1.6 μM), but not against PLpro from MERS-CoV (IC₅₀ of 112.5 ± 7.3 μM). In contrast, the polyphenolics kazinol F and broussochalcone A isolated from the same species showed better efficacy against MERS-CoV PLpro. Similarly, O'Keefe et al. (2010) found that the efficacy of the compound griffithsin against SARS-CoV was highest for the Urbani and Tor-II strains (EC₅₀ of 0.61 μg/mL using a CPE inhibition assay) and lowest for the Frank strain (EC₅₀ of 1.19 μg/mL). Griffithsin, isolated from the red alga Griffithsia sp., is a lectin possessing three large identical carbohydrate-binding domains orientated as an equatorial triangle, which enable multivalent binding (O’Keefe et al., 2010). Although the source of these observed differences in the inhibition activity of griffithsin against different SARS-CoV strains was not explored by O’Keefe et al. (2010), it may be attributable to genomic differences in the amino acid sequences of the spike proteins between SARS-CoV strains, leading to varying multivalent interactions with the carbohydrates-binding domains and thus differing affinity for binding to the spike proteins. Nevertheless, these studies highlight the importance of conducting in vitro tests of potential naturally-derived therapeutics on SARS-CoV-2 prior to animal or clinical trials.

4. Animal coronaviruses

Animal coronavirus strains are responsible for severe morbidity events across a wide range of domestic animals and livestock, incurring major economic demise worldwide (Jackwood et al., 2010; Lelesius et al., 2019; McCutcheon et al., 1995). The genomic diversity, coupled with the ability of coronaviruses to rapidly adapt and mutate, presents unique challenges in the development of novel antiviral agents, hence exploring alternative methods of controlling these viruses could potentially be effective across many or all serotypes. As natural phytochemicals have been shown to have activity across a wide range of viral pathogens (Chen et al., 2014), they form the basis of the studies reviewed here. The majority of plant-derived compounds considered in this review are direct-targeting antivirals, acting through direct inhibition of some part of the virus, such as proteases or spike proteins. For example, silvestrol prevents viral replication occurring through the specific inhibition of RNA helicase eIF4A (Müller et al., 2018), while griffithsin binds directly to the S protein, preventing viral entry to the host cell (O’Keefe et al., 2010). However, host-targeting antivirals form another important group of antiviral compounds. For example, extracts of Cinnamomi sp. Inhibit the clathrin-dependent endocytosis pathway, thus preventing viral entry to the host cells (Zhuang et al., 2009). Other host-targeting antiviral compounds may stimulate the immune response (Lau et al., 2008).

Table 2 presents a summary of studies reporting anti-viral activity of plant-derived isolates against a range of animal coronavirus strains. Where available, the key compounds responsible for the antiviral activity and their mechanism of action are provided.

5. The significance of chemical polarity

Across both human and animal CoV strains, a clear trend toward the use of chemically polar compounds is evident (Fig. 2). Of the 30 studies that specified the solvent extraction protocols used, ethanol or an ethanol/water combination was the most commonly used (50 % of all studies), followed by methanol or a methanol/water combination (27 %). A further 17 % of relevant literature used a water-based extraction protocol, with only three studies using relatively non-polar solvents (diaminopropane, chloroform and acetone). This is consistent with a wide body of research indicating that relatively polar extracted fractions generally contain a greater level of bioactive and antimicrobial compounds compared to their nonpolar counterparts (Han et al., 2007; Tian et al., 2009; Wigmore et al., 2016). This may also be indicative of increased bioactivity of highly polar glycosylated compounds, similar to the vast difference in antiviral activity observed between quercetin and quercetin-7-rhamnoside (Choi et al., 2009).

In the case of compounds being administered orally, more polar compounds are subject to compartmentalisation within the body, reducing their rate of elimination. However, the potential chemical changes that could occur as a result of highly acidic stomach conditions and the level of absorption in the intestine would need to be considered individually for each compound.

6. Conclusions

Naturally occurring phytochemicals provide a valuable and powerful resource of chemical compounds displaying antiviral properties. Further chemical modification of these structures, guided by computer-based docking simulations, may also increase their potency and/or selectivity. Some of the key compounds that show promise for the treatment of coronavirus in humans include scutellarein, silvestrol, tryptanthrin, saikosaponin B₂, lectins such as griffithsin, lycorine and polyphenolics – including quercetin, myricetin, caffeic acid, psoralidin and isobavachalcone. Needless to mention, these compounds may be toxic at certain levels, and hence in vitro and in vivo testing is required to determine safe and therapeutic levels for each compound before clinical trials in humans can be performed. Initial studies could focus on compounds which have previously been approved for drug use, or are generally regarded as safe by the FDA or other national organisations, as is the case for some polyphenolic compounds. It is hoped that researchers will be guided by this information presented here in the process of developing safe, effective anti-coronavirus therapeutic agents from naturally derived compounds.

Funding sources

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of Competing Interest

The authors declare that no conflict of interest exists.

References