(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



Ethacridine inhibits SARS-CoV-2 by inactivating viral particles

['Xiaoquan Li', 'Department Of Pharmaceutical Chemistry', 'University Of California', 'San Francisco', 'California', 'United States Of America', 'Cardiovascular Research Institute', 'Peter V. Lidsky', 'Department Of Microbiology', 'Immunology']

Date: 2021-10

The respiratory disease COVID-19 is caused by the coronavirus SARS-CoV-2. Here we report the discovery of ethacridine as a potent drug against SARS-CoV-2 (EC 50 ~ 0.08 μM). Ethacridine was identified via high-throughput screening of an FDA-approved drug library in living cells using a fluorescence assay. Plaque assays, RT-PCR and immunofluorescence imaging at various stages of viral infection demonstrate that the main mode of action of ethacridine is through inactivation of viral particles, preventing their binding to the host cells. Consistently, ethacridine is effective in various cell types, including primary human nasal epithelial cells that are cultured in an air-liquid interface. Taken together, our work identifies a promising, potent, and new use of the old drug via a distinct mode of action for inhibiting SARS-CoV-2.

To identify inhibitors against SARS-CoV-2, here we have designed and demonstrated a high throughput screening system that works in living cells at BSL2. To repurpose existing drugs, we have screened an FDA-approved drug library and identified several drugs that inhibit Mpro and block SARS-CoV-2 in cells. Interestingly, one of the identified drugs, ethacridine, shows ~40-fold higher antiviral activity than its antiprotease activity. Further studies unveil that ethacridine inhibits the coronavirus by inactivating the viral particles. Our work demonstrated a powerful tool to screen protease inhibitors to inhibit coronavirus, as well as other diseases in the future.

Competing interests: I have read the journal’s policy and the authors of this manuscript have the following competing interests: X.L., P.V.L., Y.X., R.A., X.S. have filed a patent on a new use of the identified compounds.

Funding: This work was supported in part by NIH (R35 GM131766) and Program for Breakthrough Biomedical Research (which is partially supported by the Sandler Foundation) to X.S., NIH (R01 AI36178, AI40085, P01 AI091575), the Bill and Melinda Gates Foundation and the DARPA Intercept program (Contract No. HR0011-17-2-0027) to R.A.. Funders did not play any role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Copyright: © 2021 Li et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

To identify drugs that may inhibit the coronavirus, we redesigned the green fluorescent protein (GFP) into an activity reporter, which becomes fluorescent only upon cleavage by the active Mpro. Using this fluorescent assay, we screened a drug library in living cells and identified several drugs that inhibit Mpro activity. One highly effective drug, ethacridine, inhibit SARS-CoV-2 production by inactivating viral particles.

The remaining one-third of the genomic RNA is used by the RTC to synthesize subgenomic RNAs (sgRNAs) encoding four conserved structural proteins including spike protein (S), envelope protein (E), membrane protein (M), and nucleocapsid protein (N), and several accessory proteins. Eventually, the viral RNA-N complex and S, M, and E proteins are assembled in the ER-Golgi intermediate compartment (ERGIC) to form a mature virion that is then released via budding from the host cell. S protein is exposed on the surface of the virion and binds human angiotensin converting enzyme 2 (ACE2) on the host cell surface. Therefore, Mpro plays a central and critical role in the lifecycle of the coronavirus and is an attractive drug target [ 8 – 10 ], which also include other biological steps essential for viral replication and budding.

The first region, containing the first ORF (ORF1a/b), is about two-thirds of the genome. After coronavirus attach and entry into the host cell, the viral genomic RNA is released. The first region of the genomic RNA is translated to pp1a and pp1ab. The pp1a polyprotein is translated from ORF1a and the pp1ab polyprotein comes from a -1 ribosomal frameshift between ORF1a and ORF1b. Both pp1a and pp1ab are mainly processed by a 3-chymotrypsin-like protease (3CLpro, referred as the main protease, Mpro). Mpro cleaves the polyproteins in at least 11 conserved sites. The functional polypeptides including 16 non-structural proteins (nsp1–16) are released from the polyproteins after this extensive proteolytic processing. Among them, Nsp12 (i.e., RNA-dependent RNA polymerase (RdRp)), together with other nsps (e.g., nsp7 and nsp8), forms a multi-subunit replicase/transcriptase complex (RTC) that is associated with the formation of virus-induced double-membrane vesicles [ 4 , 6 , 7 ]. The membrane-bound RTC synthesizes a full-length negative-strand RNA template for the positive-strand viral genomic RNA.

The worldwide outbreak of the respiratory disease COVID-19 is caused by the coronavirus SARS-CoV-2 (severe acute respiratory syndrome coronavirus 2). SARS-CoV-2 is an RNA betacoronavirus of the family Coronaviridae. It contains a single-stranded positive-sense RNA genome encapsulated within a membrane envelope [ 1 – 4 ]. Its genomic RNA is approximately 30kb with a 5′-cap structure and 3′-poly-A tail [ 1 ]. The genome of SARS-CoV-2 can be split into two main regions that contain as many as 14 open reading frames (ORFs) [ 5 ].

Results

Rational design of a fluorogenic Mpro activity reporter FlipGFPMpro To develop an activity reporter of Mpro with a large dynamic range suitable for high-throughput screening (HTS), we applied the GFP-based protease reporter called FlipGFP [11], which was designed by flipping one of the 11 beta-strands of a split GFP. Briefly, the split GFP contains two parts: one part contains beta-strands 10 and 11 (i.e., GFP10 and 11), and the other contains nine other beta-strands and the central alpha helix (i.e., GFP1–9). GFP10–11 contains the highly conserved Glu222 that is essential for catalyzing chromophore maturation. GFP1-9 contains the three amino acids that form the chromophore via cyclization, dehydration and oxidation [12]. GFP10-11 spontaneously binds GFP1-9 and catalyzes the chromophore maturation, leading to green fluorescence. To design an Mpro activity reporter, we “flipped” GFP10-11 using heterodimeric coiled coils (E5 and K5) so that the flipped GFP10-11 cannot bind GFP1-9 when Mpro is inactive, and thus, no or little fluorescence is detected (Fig 1A). We incorporated an Mpro-specific cleavage sequence AVLQ↓SGFR (↓denotes the cleavage site) between GFP11 and K5. In this way, when Mpro cleaves the proteolytic site, GFP11 is flipped back, allowing GFP10-11 to bind GFP1-9, resulting in bright fluorescence (Fig 1A). We named this reporter FlipGFPMpro. To normalize the fluorescence, we added a red fluorescent protein mCherry within the construct via a “self-cleaving” 2A peptide [13] (Fig 1C). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 1. Design and demonstration of a GFP-based activity reporter of SARS-CoV-2 main protease Mpro. (a) Schematic of the reporter. (b) Sequence of the flipped GFP10-11. (c) Construct of the reporter FlipGFPMpro. (d) Fluorescence images (left) and quantitative analysis (right) of SARS-CoV-2 or mock-infected HEK293T cells that co-expressed hACE2. The images in the FlipGFP channel were brightened 30-fold compared to those in (e). (e) Fluorescence images of HEK293T cells expressing FlipGFPMpro and mCherry, together with the inactive Mpro mutant C145A (upper panels) or wild type Mpro (lower panels). (f) Normalized FlipGFP fluorescence by mCherry. The ratio of FlipGFP/mCherry for the Mpro/C145A is normalized to 1. Data are mean ± SD (n = 5). FlipGFPTEV is a TEV activity reporter containing TEV cleavage sequence in FlipGFP. Scale bar: 5 μm (d); 10 μm (f). https://doi.org/10.1371/journal.ppat.1009898.g001 To determine if FlipGFPMpro serves to report on Mpro activity of SARS-CoV-2 in living cells, we expressed ACE2, a SARS-CoV2 receptor, in HEK293-FlipGFPMpro cells. Next, we infected the cells with SARS-CoV-2, and at 24 hours post-infection, cells were analyzed by immunofluorescence using antibodies against double-stranded RNA (dsRNA) and FlipGFPMpro green fluorescence. The green fluorescence of the sensor, normalized by the co-expressed mCherry, was 63% greater in the coronavirus-infected cells than in mock-infected cells (Fig 1D). Infected cells also showed dsRNA fluorescence compared to non-infected (mock) cells without dsRNA staining (Fig 1D). These data demonstrate that the utility of FlipGFPMpro as a reporter of SARS-CoV-2 Mpro activity in human cells. Next, we established a system for screening Mpro inhibitors in living cells by exogenously expressing Mpro in HEK293 cells. Specifically, wild-type Mpro or an inactive Mpro mutant (with catalytic cysteine 145 mutated to alanine) were co-expressed in this cell line. The green fluorescence of FlipGFPMpro was barely detected in the cells expressing the inactive Mpro/C145A mutant, whereas the red fluorescence of mCherry was observed (Fig 1E, upper panels). On the other hand, strong green fluorescence was detected in the cells expressing Mpro with similar levels of mCherry fluorescence (Fig 1E, lower panels). The green fluorescence of FlipGFPMpro, normalized to the red fluorescence of mCherry, revealed an ~60-fold dynamic range between inactive and active Mpro (Fig 1F). Furthermore, based on these quantified data, we calculated a Z’-factor [14] which is ~0.8 with Mpro and its inactive mutant as positive (+) and negative (-) controls, respectively (here σ is standard deviation, μ is mean). This suggests that the assay is robust for HTS. The FlipGFPMpro sensor was not responsive to the TEV protease, and the FlipGFP-based TEV reporter (FlipGFPTEV) was only activated by TEV but not by Mpro (Fig 1F). Thus, FlipGFPMpro specifically detects Mpro activity with a large dynamic range. Therefore, our data show that we have established a robust HTS system for screening Mpro inhibitors at a BSL2 level with 60-fold fluorescence change and a robust z’-factor. Difference of the normalized FlipGFPMpro fluorescence in the SARS-CoV-2 infected cells (Fig 1D) and that in the cells expressing Mpro exogenously (Fig 1E and 1F) suggests that the active Mpro concentration in the cytoplasm of the coronavirus-infected cells is ~100-fold lower than that of the HEK293 cells exogenously expressing Mpro (under a EF1α promoter).

HTS of drugs that inhibit Mpro activity in living cells Next, we conducted HTS of ~1600 FDA-approved drugs (20 μM final concentration, Fig 2A). The reporter construct (FlipGFPMpro and mCherry) was transfected into HEK293 cells, followed by addition and incubation of the drugs. Green fluorescence normalized to red fluorescence were then calculated. A volcano plot revealed ~120 drugs that showed ≥ 50% reduction of Mpro activity with a p-value < 0.001 (Fig 2A). To confirm this result, we re-screened the identified ~120 drugs under similar conditions (S1 Fig). We further assayed those top 15 drugs at a lower concentration (10 μM) and found that 12 drugs showed ≥50% reduction of FlipGFPMpro fluorescence (normalized by mCherry) at 10 μM concentration (Fig 2B). We finally calculated an IC 50 for each of the 12 drugs. IC 50 of six drugs were at 2–6 μM (Fig 2C), and the rest were above 6 μM (S2 Fig). Lastly, we also determined cellular viability of these identified drugs in HEK293T cells, which showed that they are not toxic at the concentrations that inhibit Mpro activity (S3 Fig). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. High-throughput screening and drug identification using FlipGFPMpro in living cells. (a) Volcano plot of 1622 FDA-approved drugs (20 μM) in inhibiting Mpro. HEK293T cells were transfected with FlipGFPMpro, mCherry and Mpro. FlipGFP fluorescence was normalized to co-expressed mCherry. (b) Normalized ratio of Mpro activity in drug (10 μM) vs DMSO-incubated cells for the third-round validation. The Mpro activity was determined as FlipGFP fluorescence normalized to mCherry. The ratio of Mpro activity was calculated by normalizing Mpro activity with that of cells treated with DMSO. Data are mean ± SD (n = 5). The 15 drugs were identified from a second-round imaging of the 120 identified drugs (20 μM, Extended data Fig 1). (c) Dose-response curve of top six drugs in inhibiting Mpro. Inhibition ratio was calculated as (1-(ratio of Mpro activity)) X100%. IC50 was represented as mean ± SEM (n = 5). See Extended data Fig 2 for the other six drugs. https://doi.org/10.1371/journal.ppat.1009898.g002

Antiviral activity of identified drugs We next investigated antiviral activity of selected drugs in Vero E6 cells. The cell monolayers were pretreated with the 12 selected drugs for 3 hours, and then infected with SARS-CoV-2. The cells were further cultivated in the presence of each respective compound at a concentration of 5 μM. After 16 hours of incubation, the culture media samples were collected, and the amount of infectious particles were estimated by plaque assay (Figs 3A and 3B and S4). Our data revealed that 9 of the 12 drugs showed significant antiviral activity at 5 μM. Strong inhibition was detected for ethacridine with 5–6 logs reduction in viral titer, simeprevir ~4-log reduction, ABT-199 ~2-log reduction, hydroxyprogesterone ~1-log reduction, cinacalcet ~1-log reduction. Two of the 12 drugs (ivermectin and verteporfin) were cytotoxic at 5–13 μM in Vero E6 cells and excluded from further analysis (S5 Fig). As a comparison, we tested the antiviral effect of a reported Mpro inhibitor, ebselen, which showed ~2-log reduction in viral titer. The RdRp inhibitor remdesivir showed ~4-log reduction. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. Antiviral activities of the identified drugs against SARS-CoV-2. (a) Schematic showing the experimental design, see Methods for details. (b) Quantitative analysis of viral titer from plaque assays on Vero E6 cells treated with each drug at 5 μM. Data are mean ± SEM (n = 3). *** p < 0.001. (c–h) Dose-response and cell-toxicity curve of each drug against SARS-CoV-2 by plaque assays. The percentage of relative infection was determined by the ratio of infection rate of SARS-CoV-2 treated with each drug divided by that of DMSO control. EC50 and CC50 are represented as mean ± SEM (n = 3). https://doi.org/10.1371/journal.ppat.1009898.g003 Next, we determined dose-response curves for the top 5 selected drugs. First, our data revealed that the EC 50 of four drugs (simeprevir, cinacalcet, ABT-199 and hydroxyprogesterone) was 1–3 μM (Fig 3C–3G), within a range similar to their IC 50 in inhibiting Mpro (Fig 2C). This was consistent with the expectation that SARS-CoV-2 replication is inhibited by restricting Mpro activity. Indeed, as we were finalizing this study, a preprint report showed that simeprevir inhibits Mpro activity and SARS-CoV-2 [15]. By contrast, ethacridine showed outstanding antiviral activity (EC 50 ~ 0.08 μM, Fig 3E), which is 40-fold lower (i.e. stronger) than its Mpro-inhibiting activity (IC 50 ~ 3.54 μM, Fig 2C). These data suggest that the antiviral activity of ethacridine is not mainly accounted for by its Mpro-inhibiting activity. Lastly, for comparison, we also determined the EC 50 of remdesivir ~0.52 μM in a side-by-side manner (Fig 3H), which indicates that ethacridine is more potent than remdesivir.

Ethacridine inhibits SARS-CoV-2 by inactivating viral particles To determine how ethacridine inhibits SARS-CoV-2, we tested infectivity of the virus particles after ethacridine treatment with plaque assay, and we also measured viral RNA levels using qRT-PCR. We examined the antiviral effect of ethacridine on different stages of the lifecycle of SARS-CoV-2, including virus-cell binding, RNA replication, and budding. To test overall effect of ethacridine on virus replication, we pre-incubated SARS-CoV-2 particles with ethacridine (5 μM) or DMSO control for 1 hour. The mixture was then added to Vero E6 cells for viral adsorption at a multiplicity of infection (MOI) at 0.5. Next, we removed the solution and added fresh medium containing ethacridine (5 μM) or DMSO control. Sixteen hours later, we collected supernatant and conducted plaque assay with overlaid agar without ethacridine or DMSO to measure viral titer. We also conducted qRT-PCR and measured viral RNA levels in the supernatant and within cells. In this way, we developed three conditions (Fig 4A): 1) Control (DMSO + DMSO): the virus and cells were exposed to DMSO and not the drug; 2) The virus and cells were exposed to the drug at all stages, including 1 hour before infection, during replication, and after viral budding (i.e. Eth. + Eth.); 3) The virus and cells were exposed to the drug only after viral entry, during replication, and after budding (i.e. DMSO + Eth.). Lastly, we used a fourth condition (Fig 4B): we conducted plaque assay right after pre-treatment of SARS-CoV-2 with ethacridine for 1 hr (i.e. Eth. [1 hr]), which determines direct effect of the drug on viral particles. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. Ethacridine bocks SARS-CoV-2 by inactivating viral particles. (a) Upper panel: schematic showing the experimental design for plaque assay and qRT-PCR. The virus was pre-incubated with ethacridine or DMSO for 1 hr. The mixture was added to Vero cells for adsorption at 37°C for 1 hr. Details can be found in the text. Lower panel: quantitative analysis of viral titer from plaque assay (left), and viral RNA (vRNA) copies by qRT-PCR in Vero cells (middle) and supernatant (right). (b) Quantitative analysis of viral titer by plaque assay. (c) Proposed mode of action of ethacridine by mainly inactivating viral particles of the coronavirus with no or little effect on viral RNA replication. (d) Schematic showing the experimental design for immunostaining. (e, f) Representative images of immunostaining against nucleocapsid protein (N) and spike (S) in Vero E6 cells after infection with the virus that was pre-treated with ethacridine (5 μM) or DMSO (control), or no infection (mock). (g) Quantitative analysis of the immunofluorescence. Data are mean±SEM (n = 3 or 4 biological replicas). **: p value < 0.01. *: p value < 0.05. Scale bar, 20 μm. https://doi.org/10.1371/journal.ppat.1009898.g004 When ethacridine was present continuous prior to plaque assay (Eth. + Eth.), viral titers were reduced 3–4 logs, compared with the control (DMSO + DMSO) (Fig 4A, lower left panel). When ethacridine was added to the Vero cells after viral entry (DMSO + Eth.), similar level of reduced infectivity (3–4 logs reduction in viral titer) was observed (Fig 4A). Importantly, when SARS-CoV2 was pre-incubated with ethacridine for 1 hr (Eth. [1 hr]) and followed by plaque assay without drug, we also observed 3–4 logs reduction in infectivity (Fig 4B). This result suggests that the drug directly inactivates SARS-CoV2 viral particles. Because of similar-level reduction in infectivity in all of the three conditions, our data strongly suggests that ethacridine inhibits SARS-CoV-2 mainly by inactivating viral particles. We next examined viral RNA accumulation in infected cells. qRT-PCR measurement revealed no change of viral RNA (vRNA) levels when the drug was added after viral binding and cell entry (DMSO + Eth.) in both the supernatant and the cells (Fig 4A, lower middle and right panels), compared with the control (DMSO + DMSO). This indicates that the drug has no effect on vRNA replication. As the infectivity of supernatant from “DMSO + Eth” treated samples showed 3–4 logs reduction in plaque assay (Fig 4A, lower left panel), these data suggests that ethacridine inhibits SARS-CoV-2 by inactivating the viral particles without effect on vRNA replication. This is consistent with the results of plaque assays for the supernatant samples with 3 different treatments in that showed 3~4 log reduction in infectivity (Fig 4A, lower left panel and Fig 4B) after virions in the supernatant were exposed to the drug before plaque assay. Next, when ethacridine was present continuously (i.e. Eth. + Eth.), 4–5 fold reductions were observed in vRNA copies in the supernatant and within cells (Fig 4A, middle and right panels). Because plaque assay-based measurement of the same conditioned sample (Eth. + Eth.) showed 2400-fold reduction in viral titer, the effect of ethacridine on viral replication (4–5 fold reduction) is about 500-fold smaller than its effect on viral infectivity. This further supports the conclusion that ethacridine inhibits SARS-CoV-2 by inactivating the viral particles. The 4–5 fold reduction of vRNA copies is likely due to reduced viral copy numbers that may bind to the cells (see below), because here the additional step is that the virus were pre-incubated with the drug. Thus, our plaque assay and qRT-PCR data suggests that ethacridine inhibits SARS-CoV-2 mainly by inactivating viral particles, including the virus before binding to cells and in the supernatant after budding from host cells, with no or little effect on vRNA replication (Fig 4C). To further investigate the mechanism of ethacridine-based inactivation of the viral particles, we conducted immunofluorescence staining and imaging to determine whether the ethacridine-treated SARS-CoV-2 can bind to the cells. We treated SARS-CoV-2 with ethacridine (5 μM) or DMSO for 1 hour at 37°C. Then the virus was added to cells for adsorption (4°C, 1 hour) at a MOI = 100. Cells were then quickly washed and fixed with 4% PFA (Fig 4D). Immunostaining with antibodies against the nucleocapsid protein (N) of SARS-CoV-2 showed strong anti-N fluorescence on the plasma membrane of the cells infected with control virus, but little anti-N fluorescence in cells exposed to ethacridine treated virus (Fig 4E and 4G). Immunostaining against the Spike protein (S) of SARS-CoV-2 showed the same results (Fig 4F and 4G). Furthermore, to examine whether ethacridine blocks viral binding to cells by perturbing the cellular factors for viral binding such as cellular receptors, we pre-treated cells with ethacridine for 3 hours. This was followed by washing and drug removal, immediately prior to addition of SARS-CoV-2. The data showed that viral infection was not affected by these procedures (S6 Fig), suggesting that the main effect of the drug is not on the cells, but on the viral particles. These results indicate that ethacridine-treated SARS-CoV-2 cannot bind cells to initiate infection. To further support our model, we conducted an additional experiment. We mixed ethacridine with SARS-CoV-2 and immediately added the drug/virus mixture to the Vero cells (i.e. no preincubation of the drug with the virus). After adsorption for 1 hour (37°C), we overlayed the cells with agar and media for plaque assay. Under this condition, first, the drug will be able to inhibit potential cellular factors since the drug is not removed after the adsorption step. Second, the drug is not preincubated with the virus, and thus according to our model, we expect much smaller effect of the drug on the virus than when the drug was preincubated with the virus. Indeed, we observed dramatically smaller effect of the drug: ~2.7-fold inhibition (S7 Fig) versus 3–4 logs inhibition when the drug was preincubated with the virus (Figs 4B and S6). These results further support our model that ethacridine inhibits SARS-CoV-2 by mainly inactivating the viral particles. We also tested the dependency of the viral-inactivation effect of ethacridine on dose, incubation time and incubation temperature with a plaque assay. For the conditions tested, the viral-inactivation effect showed dose-dependency but was comparable to a 1- or 2-hour incubation at room temperature or at 37°C (S8 Fig).

[END]

[1] Url: https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1009898

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.