key: cord-296981-tded20ih
authors: Gilmore, Kerry; Zhou, Yuyong; Ramirez, Santseharay; Pham, Long V.; Fahnøe, Ulrik; Feng, Shan; Offersgaard, Anna; Trimpert, Jakob; Bukh, Jens; Osterrieder, Klaus; Gottwein, Judith M.; Seeberger, Peter H.
title: In vitro efficacy of Artemisinin-based treatments against SARS-CoV-2
date: 2020-10-05
journal: bioRxiv
DOI: 10.1101/2020.10.05.326637
sha: 
doc_id: 296981
cord_uid: tded20ih

Effective and affordable treatments for patients suffering from coronavirus disease 2019 (COVID-19), caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), are needed. We report in vitro efficacy of Artemisia annua extracts as well as artemisinin, artesunate, and artemether against SARS-CoV-2. The latter two are approved active pharmaceutical ingredients of anti-malarial drugs. Proof-of-concept for prophylactic efficacy of the extracts was obtained using a plaque-reduction assay in VeroE6 cells. Subsequent concentration-response studies using a high-throughput antiviral assay, based on immunostaining of SARS-CoV-2 spike glycoprotein, revealed that pretreatment and treatment with extracts, artemisinin, and artesunate inhibited SARS-CoV-2 infection of VeroE6 cells. In treatment assays, artesunate (50% effective concentration (EC50): 7 μg/mL) was more potent than the tested plant extracts (128-260 μg/mL) or artemisinin (151 μg/mL) and artemether (>179 μg/mL), while generally EC50 in pretreatment assays were slightly higher. The selectivity index (SI), calculated based on treatment and cell viability assays, was highest for artemisinin (54), and roughly equal for the extracts (5-10), artesunate (6) and artemether (<7). Similar results were obtained in human hepatoma Huh7.5 cells. Peak plasma concentrations of artesunate exceeding EC50 values can be achieved. Clinical studies are required to further evaluate the utility of these compounds as COVID-19 treatment.

The pandemic with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) 1,2 has worldwide been associated with over 1 million deaths from coronavirus disease 2019 (COVID- 19) . 3, 4, 5 This febrile respiratory and systemic illness is highly contagious and in many cases life-threatening. Remdesivir, the only antiviral drug with proven in vitro and clinical efficacy, was approved for treatment of COVID-19. 6 Still, COVID-19 treatment remains largely supportive with an urgent need to identify effective antivirals against SARS-CoV-2. An attractive approach is repurposing drugs already licensed for other diseases. Teas of A. annua plants have been employed to treat malaria in Traditional Chinese Medicine, as well as in clinical trials, 7, 8 and are used widely in many African countries, albeit against WHO recommendations. Artemisinin (Figure 1, 1) , a sesquiterpene lactone with a peroxide moiety and one of many bioactive compounds present in A. annua, is the active ingredient to treat malaria infections. 9, 10 The artemisinin derivatives artesunate (Figure 1 , 2) and artemether ( Figure 1, 3) exhibit improved pharmacokinetic properties and are the key active pharmaceutical ingredients (API) of WHO-recommended anti-malaria combination therapies used in millions of adults and children each year with few side effects. 11 A. annua extracts are active against different viruses, including SARS-CoV. 12, 13, 14 Therefore, we set out to determine whether A. annua extracts, as well as pure artemisinin, artesunate, and artemether are active against SARS-CoV-2 in vitro. Artemisinin-based drugs would be attractive repurposing candidates for treatment of COVID-19 considering their excellent safety profiles in humans, and since they are readily available for worldwide distribution at a relatively low cost.

A. annua plants grown from a cultivated seed line in Kentucky, USA, were extracted using either absolute ethanol or distilled water at 50 °C for 200 minutes, as described in Materials and Methods and Supplementary Information (Figure S1 ). For a third preparation, to protect artemisinin from degradation by reducing agents, ground coffeea natural source of polyphenols such as chlorogenic acid ( Figure 1 , (4) ) that also exhibit mild antiviral activities 15, 16 was coextracted with the plant material using ethanol (see supporting information). Solids were removed by filtration and the solvents were evaporated. The extracted materials were dissolved in dimethylsulfoxide (DMSO) (both ethanol extracts) or a DMSO:water mixture (3:1 for aqueous extract) and filtered (see supporting information for details). Artemisinin ( Figure 1 , (1)) was synthesized and purified following a published procedure, 17 

To initially screen whether extracts and pure artemisinin were active against SARS-CoV-2, their antiviral activity was tested by pretreating VeroE6 cells at different time points during 120 minutes with selected concentrations of the extracts or compounds prior to infection with the first European SARS-CoV-2 isolated in München (SARS-CoV-2/human/Germany/BavPat 1/2020). The virus-drug mixture was then removed and cells were overlaid with medium containing 1.3% carboxymethylcellulose to prevent virus release into the medium. DMSO was used as a negative control. Plaque numbers were determined either by indirect immunofluorescence using a mixture of antibodies to SARS-CoV N protein 18 or by staining with crystal violet. 19 The addition of either ethanolic or aqueous A. annua extracts prior to virus addition resulted in reduced plaque formation in a concentration dependent manner with median effective concentration (EC50) values estimated to range between 5 and 168 µg/mL (Supplemental Figure S2A -C). Artemisinin exhibited little antiviral activity with an EC50 >220 µg/mL (Supplemental Figure S2D ).

Concentration-response experiments using the Danish SARS-CoV-2 isolate SARS-CoV-2/human/Denmark/DK-AHH1/2020 were performed employing a 96-well plate based highthroughput antiviral assay, allowing for multiple replicates per concentration, as described in Materials and Methods and Supplementary Information (Figures S3 and S4 ). 20 Seven replicates were measured at each concentration and a range of concentrations was evaluated to increase data accuracy when compared to the plaque-reduction assay, which was carried out in duplicates. Extracts or compounds were added to VeroE6 cells either 1.5 h prior to (pretreatment (pt)) or 1 h post infection (treatment (t)), respectively, followed by a two-day incubation of virus with extracts or compounds. Both protocols yielded similar results, with slightly lower EC50 values observed for treatment assays.

The ethanolic extracts showed similar potency: for A. annua alone EC50 were 173 µg/mL (pt) and 142 µg/mL (t) and for A. annua with coffee EC50 were 176 µg/mL (pt) and 128 µg/mL (t) (Figures 2, 3 and Table 1 ). The aqueous extract was slightly less potent with EC50 being 390 µg/mL (pt) and 260 µg/mL (t) (Figures 2, 3 and Table 1 ). With all extracts, almost complete virus inhibition was achieved at high concentrations: For the A. annua ethanolic extract at 333 µg/mL (pt) and 444 µg/mL (t), for the A. annua + coffee ethanolic extract at 300 µg/mL (pt) and 267 µg/mL (t), and for the A. annua aqueous extract at 875 µg/mL (pt) and 1009 µg/mL (t) (Figures 2 and 3 ). The highest evaluated concentrations used in our assays were informed by the cytotoxicity of the extracts or compounds, as only concentrations resulting in cell viability greater than 90% were evaluated (Figures 2, 3, S5 and Table 1 ). Cell viability assays Table 1 ). Selectivity indexes (SI) were determined by dividing CC50 by EC50 and revealed similar results for the A. annua ethanolic extract being 6 (pt) and 7 (t), the A. annua + coffee ethanolic extract being 3 (pt) and 5 (t) as well as the A. annua aqueous extract being 7 (pt) and 10 (t) ( Table 1 ).

The two ethanolic extracts were diluted with DMSO that by itself caused reduction of cell viability to <90% when used at a 1:28 dilution, but not at dilutions ≥1:42 ( Figure S6 ). Thus, the cytotoxicity observed when using the extracts at relatively high concentrations was most likely not caused by DMSO (Figures 2 and 3 ). DMSO at dilutions >1:152 including the dilutions used in antiviral assays did not have antiviral effects, defined as reduction of residual infectivity to <70% ( Figure S6 ). Thus, the observed antiviral effect of the tested extracts was most likely not caused by DMSO. A pure coffee extract estimated to contain 2.5-fold higher coffee concentrations than the A. annua + coffee ethanolic extract did not result in reduction of cell viability to <90% at dilutions ≥1:28 ( Figure S7 ). The cytotoxicity observed when using the A. annua + coffee extract at relatively high concentrations was most likely not caused by coffee (Figures 2 and 3) . Interestingly, coffee extract alone showed some antiviral activity at dilutions ≤1:273 ( Figure S7 ). Thus, the observed antiviral effect of the A. annua + coffee extract may be influenced by coffee.

VeroE6 cells. A. annua plants contain, in addition to many other bioactive compounds, artemisinin that is responsible for the potent anti-malarial activities of A. annua. To investigate whether artemisinin is the active component responsible for the antiviral activities of the plant extracts described above, the pure compound and synthetic derivatives were tested in pretreatment and treatment assays. Artemisinin was found to be active in SARS-CoV-2 assays with EC50 238 µg/mL (pt) and 151 µg/mL (t) (Figures 2, 3 , and Table 1 ). Close to complete virus inhibition was achieved in both assays at the highest concentration evaluated in the assays, 893 (pt) and 1208 µg/mL (t). The SI for artemisinin is relatively high, 34 (pt) and 54 (t), based on a CC50 of 8,216 µg/mL (Figures 2, 3 , S6, and Table 1 ). The observed cytotoxicity of artemisinin appeared to be at least partially caused by DMSO, as cytotoxicity was only observed at drug dilutions where DMSO was found to reduce cell viability (Figures 2, 3 , and S6). The antiviral effects observed when using artemisinin at relatively high concentrations were most likely not due to the diluent DMSO (Figures 2, 3 , and S6).

The synthetic artemisinin derivative artesunate, the API of WHO-recommended first-line malaria therapies with improved pharmacokinetic properties, showed the highest potency of all compounds tested, with EC50 being 12 µg/mL (pt) and 7 µg/mL (t) (Figures 2 and 3 ). In the treatment assay, close to complete virus inhibition was achieved at the highest evaluated concentration (15 µg/mL), as determined by cytotoxicity data, compared to 69% inhibition at this concentration in the pretreatment assay. Higher artesunate concentrations were not used considering its cytotoxicity in this assay (CC50: 41 µg/mL) (Figures 2, 3 , S5, and Table 1 ). SI of 3 (pt) and 6 (t) were calculated ( Table 1 ). The cytotoxicity and the antiviral effects observed when using artesunate at relatively high concentrations were most likely not due to the diluent DMSO (Figures 2, 3 , and S6).

Artemether, another artemisinin-derivative that is used globally as the active ingredient in malaria medications, did not show a significant antiviral effect at concentrations of up to 179 µg/mL (Figures 2 and 3 ). Considering artemether´s cytotoxicity (CC50 of 1,220 µg/mL), an SI < 7 was calculated (Figures 2, 3 , S5, and Table 1 ). The cytotoxicity observed when using artemether at relatively high concentrations was most likely not due to the diluent DMSO Table 1 ). In Huh7.5 cells, the EC50 for the ethanolic A. annua extract was 118 µg/mL, with 76% virus inhibition at the highest evaluated concentration (150 µg/mL), as determined by cytotoxicity data; the CC50 was 483 µg/mL and the SI was 4 (Figures 4, S8 and Table 1 ). Artemisinin showed no significant virus inhibition at the highest evaluated concentration (208 µg/mL) and an SI <24, based on a CC50 of 5,066 µg/mL (Figures 4, S8 and Table 1 ).

In Huh7.5 cells, DMSO caused reduction of cell viability to <90% when used at a 1:28 dilution, but not at dilutions ≥1:56 ( Figure S9 ). Thus, the cytotoxicity observed when using the ethanolic extract or the pure compounds at relatively high concentrations was most likely not caused by DMSO ( Figure 4 ). DMSO at dilutions >1:179 including dilutions used in antiviral assays did not have any antiviral effects ( Figure S9 ). Thus, the observed antiviral effect of the ethanolic A. annua extract and the pure compounds was most likely not caused by DMSO.

Here, we demonstrate the in vitro efficacy of artemisinin-based treatments against SARS-CoV-2. Initially, several A. annua extracts, as well as artemisinin, were screened for antiviral activity using a plaque-reduction assay in a pretreatment setting using a German SARS-CoV-2 strain from Munich. Based on these findings, three A. annua extracts and pure, synthetic artemisinin, artesunate, and artemether were studied in detail to establish concentration-response curves for extracts and compounds for pretreatment and treatment settings using a Danish SARS-CoV-2 strain from Copenhagen.

High-throughput antiviral assays facilitated testing of drug concentrations in multiple replicates resulting in accurate EC50 values. The EC50 values in the pretreatment setting were slightly higher than EC50 values determined in the treatment setting possibly because preincubation may have a negative impact on the stability of the extracts and pure compounds.

Generally, EC50 values depend on the specific assay employed. While the type of assay we used with a single treatment and subsequent incubation of virus and drug is state of the art for antiviral efficacy measurements, assay modifications, such as repeated administration of treatment, might result in slightly different EC50 values. Since the active antiviral substance may be an artemisinin metabolite, such that the artemisinin derivatives and extracts can be considered prodrugs, we used the human Huh7.5 cell line to confirm the EC50 determined in VeroE6 cells.

While A. annua extracts have been considered "natural combination therapies" as they contain several bioactive compounds, 21 the WHO discourages the use of non-pharmaceutical forms of artemisinin as a therapeutic option for malaria due to lack of standardization with its sourcing and preparation, implying risks of suboptimal efficacy and resistance development. 22 In this context, it is important to note that the extracts used in this study were prepared from plants grown under optimized and standardized conditions, in a manner where concentrations of the extracted material are reproducible.

Interestingly, we found that coffee extracts exhibited in vitro efficacy against SARS-CoV-2.

While modelling studies suggested that ingredients in coffee such as chlorogenic acid, caffeic acid, and tannins show activity against SARS-CoV-2, 23 When the typically used doses of 2 to 2.4 mg/kg intravenously were administered, reported peak plasma concentrations (Cmax) were between 19.4 and 29.7 µg/mL in patients. 26 Based on these observations and our treatment data in VeroE6 and Huh7.5 cells, the calculated Cmax/EC50 values are between 2.5 and 4.2. In animal studies following administration of a single dose of artesunate, tissue concentrations including lung, kidney, intestine, and spleen concentrations were several-fold higher than plasma concentrations. 27 In contrast, following administration of artemisinin, artemether, and A. annua teas, Cmax values between 311−776 ng/mL were reported, which is close to three orders of magnitude below EC50 values for SARS-CoV-2. Plasma and tissue concentrations that can be achieved with standardized A. annua extracts with high artemisinin content used in this study still have to be determined. In vivo, immunomodulatory effects of artemisinin-based treatments have been reported for this class of drugs. 28 Such effects that may involve cytokine signaling cannot be monitored in in vitro assays performed here and will have to be carefully studied in subsequent clinical evaluations. infected and nontreated as well as 12 noninfected and nontreated control wells were included in the assay. After 72±2 hours incubation at 37 °C and 5% CO2, cultures were immunostained for SARS-CoV-2 spike glycoprotein and evaluated as described below.

Cells were fixed and virus was inactivated by immersion of plates in methanol (J.T.Baker, Gliwice, Poland) for 20 min. Unless specified, immunostaining was done at room temperature.

Plates were washed twice with PBS (Sigma, Gillingham, UK) containing 0.1% Tween-20 (Sigma, Saint Louis, Missouri, USA). Endogenous peroxidase activity was blocked by incubation with 3% H2O2 for ten minutes followed by two washes with PBS containing 0.1% Tween-20 and blocking with PBS containing 1% bovine serum albumin (Roche, Mannheim, Germany) and 0.2% skim milk powder (Easis, Aarhus, Denmark) for 30 minutes. Figure S3 and representative images of single wells are show in Figure S4 . which an antiviral effect (<70% residual infectivity) / cytotoxic effect (<90% cell viability) due to DMSO is expected according to Figure S9 . 

The continuing 2019-nCoV epidemic threat of novel coronaviruses to global health-the latest 2019 novel coronavirus outbreak in Wuhan, China

Importation and human-to-human transmission of a novel coronavirus in Vietnam

Clinical features of patients infected with 2019 novel coronavirus in Wuhan, China. Lancet. 2020

Coronavirus infections-more than just the common cold

Outbreak of pneumonia of unknown etiology in Wuhan, China: the mystery and the miracle

Remdesivir for the Treatment of Covid-19 -Preliminary Report N Engl

Artemisia annua and Artemisia afra tea infusions vs. artesunateamodiaquine (ASAQ) in treating Plasmodium falciparum malaria in a large scale, double blind, randomized clinical trial

Artemisia annua L. infusion consumed once a week reduces risk of multiple episodes of malaria: a randomized trial in a Ugandan community

Qinghaosu (artemisinin): an antimalarial drug from China

Flavonoids from Artemisia annua L. as antioxidants and their potential synergism with artemisinin against malaria and cancer

Effect of the affordable medicines facility -malaria (AMFm) on the availability, price, and market share of quality-assured artemisinin-based combination therapies in seven countries: a before-and-after analysis of outlet survey data

Natural products as promising drug candidates for the treatment of hepatitis B and C

Valacyclovir combined with artesunate or rapamycin improves the outcome of herpes simplex virus encephalitis in mice compared to antiviral therapy alone

Antiviral activities of coffee extracts in vitro

Anti-hepatitis B virus activity of chlorogenic acid, quinic acid and caffeic acid in vivo and in vitro

Recovery of artemisinin from a complex reaction mixture using continuous chromatography and crystallization

Antigenic and cellular localization analysis of the severe acute respiratory syndrome coronavirus nucleocapsid protein using monoclonal antibodies

The α-TIF (VP16) homologue (ETIF) of equine herpesvirus 1 is essential for secondary envelopment and virus egress

Efficient culture of SARS-CoV-2 in human hepatoma cells enhances viability of the virus in human lung cancer cell lines permitting the screening of antiviral compounds

From ancient herb to modern drug: Artemisia annua and Artemisinin for cancer therapy. Sem. Cancer Bio

A Natural products may interfere with SARS-CoV-2 attachment to the host cell

Anti-SARS-CoV-2 potential of artemisinins in vitro

Drug therapy in dental practice: general principles part 2 -pharmacodynamic considerations

Review of the clinical pharmacokinetics of artesunate and its active metabolite dihydroartemisinin following intravenous, intramuscular, oral or rectal administration

The distribution pattern of intravenous [14C] artesunate in rat tissues by quantitative whole-body autoradiography and tissue dissection techniques

Artemisinins and immune system

Highly permissive cell lines for subgenomic and genomic hepatitis C virus RNA replication

Differential efficacy of protease inhibitors against HCV genotypes 2a, 3a, 5a, and 6a NS3/4A protease recombinant viruses

Recombinant HCV variants with NS5A from genotypes 1-7 have different sensitivities to an NS5A inhibitor but not interferon-. Gastroent

Novel infectious cDNA clones of hepatitis C virus genotype 3a (strain S52) and 4a (strain ED43): genetic analyses and in vivo pathogenesis studies