on 26-Nov-2024 (Tue)

Hydroxychloroquine, a less toxic derivative of chloroquine, is effective in inhibiting SARS-CoV-2 infection in vitro

As of March 3, 2020, over 80,000 cases have been confirmed in China, including 2946 deaths as well as over 10,566 confirmed cases in 72 other countries

We have recently reported that two drugs, remdesivir (GS-5734) and chloroquine (CQ) phosphate, efficiently inhibited SARS-CoV-2 infection in vitro 1 .

However, as an experimental drug, remdesivir is not expected to be largely available for treating a very large number of patients in a timely manner. Therefore, of the two potential drugs, CQ appears to be the drug of choice for large-scale use due to its availability, proven safety record, and a relatively low cost.

CQ (N4-(7-Chloro-4-quinolinyl)-N1,N1-diethyl-1,4- pentanediamine) has long been used to treat malaria and amebiasis. However, Plasmodium falciparum developed widespread resistance to it, and with the development of new antimalarials, it has become a choice for the pro- phylaxis of malaria. In addition, an overdose of CQ can cause acute poisoning and death 3

Hydroxychloroquine (HCQ) sulfate, a derivative of CQ, was first synthesized in 1946 by introducing a hydroxyl group into CQ and was demonstrated to be much less (~40%) toxic than CQ in animals 4 .

Since CQ and HCQ share similar chemical structures and mechanisms of acting as a weak base and immunomodulator, it is easy to conjure up the idea that HCQ may be a potent candidate to treat infection by SARS-CoV-2.

Whether HCQ is as efficacious as CQ in treating SARS-CoV-2 infection still lacks the experimental evidence

To this end, we evaluated the antiviral effect of HCQ against SARS-CoV-2 infection in comparison to CQ in vitro.

First, the cytotoxicity of HCQ and CQ in African green monkey kidney VeroE6 cells (ATCC-1586) was measured by standard CCK8 assay, and the result showed that the 50% cytotoxic concentration (CC 50 ) values of CQ and HCQ were 273.20 and 249.50 μM, respectively, which are not significantly different from each other (Fig. 1a).

Fig. 1 Comparative antiviral efficacy and mechanism of action of CQ and HCQ against SARS-CoV-2 infection in vitro. a Cytotoxicity and

antiviral activities of CQ and HCQ. The cytotoxicity of the two drugs in Vero E6 cells was determined by CCK-8 assays. Vero E6 cells were treated with

different doses of either compound or with PBS in the controls for 1 h and then infected with SARS-CoV-2 at MOIs of 0.01, 0.02, 0.2, and 0.8. The virus

yield in the cell supernatant was quantified by qRT-PCR at 48 h p.i. Y-axis represents the mean of percent inhibition normalized to the PBS group. The

experiments were repeated twice. b,cMechanism of CQ and HCQ in inhibiting virus entry. Vero E6 cells were treated with CQ or HCQ (50 μM) for 1 h,

followed by virus binding (MOI =10) at 4 °C for 1 h. Then the unbound virions were removed, and the cells were further supplemented with fresh

drug-containing medium at 37 °C for 90 min before being fixed and stained with IFA using anti-NP antibody for virions (red) and antibodies against

EEA1 for EEs (green) or LAMP1 for ELs (green). The nuclei (blue) were stained with Hoechst dye. The portion of virions that co-localized with EEs or ELs

in each group (n> 30 cells) was quantified and is shown in b. Representative confocal microscopic images of viral particles (red), EEA1

+

EEs (green),

or LAMP1

+

ELs (green) in each group are displayed in c. The enlarged images in the boxes indicate a single vesicle-containing virion. The arrows

indicated the abnormally enlarged vesicles. Bars, 5 μm. Statistical analysis was performed using a one-way analysis of variance (ANOVA) with

GraphPad Prism (F=102.8, df =5,182, ***P< 0.001)

Fig. 1 Comparative antiviral efficacy and mechanism of action of CQ and HCQ against SARS-CoV-2 infection in vitro. a Cytotoxicity and

antiviral activities of CQ and HCQ. The cytotoxicity of the two drugs in Vero E6 cells was determined by CCK-8 assays. Vero E6 cells were treated with

different doses of either compound or with PBS in the controls for 1 h and then infected with SARS-CoV-2 at MOIs of 0.01, 0.02, 0.2, and 0.8. The virus

yield in the cell supernatant was quantified by qRT-PCR at 48 h p.i. Y-axis represents the mean of percent inhibition normalized to the PBS group. The

experiments were repeated twice. b,cMechanism of CQ and HCQ in inhibiting virus entry. Vero E6 cells were treated with CQ or HCQ (50 μM) for 1 h,

followed by virus binding (MOI =10) at 4 °C for 1 h. Then the unbound virions were removed, and the cells were further supplemented with fresh

drug-containing medium at 37 °C for 90 min before being fixed and stained with IFA using anti-NP antibody for virions (red) and antibodies against

EEA1 for EEs (green) or LAMP1 for ELs (green). The nuclei (blue) were stained with Hoechst dye. The portion of virions that co-localized with EEs or ELs

in each group (n> 30 cells) was quantified and is shown in b. Representative confocal microscopic images of viral particles (red), EEA1

+

EEs (green),

or LAMP1

+

ELs (green) in each group are displayed in c. The enlarged images in the boxes indicate a single vesicle-containing virion. The arrows

indicated the abnormally enlarged vesicles. Bars, 5 μm. Statistical analysis was performed using a one-way analysis of variance (ANOVA) with

GraphPad Prism (F=102.8, df =5,182, ***P< 0.001).

To better compare the antiviral activity of CQ versus HCQ, the dose–response curves of the two compounds against SARS-CoV-2 were determined at four different multi- plicities of infection (MOIs) by quantification of viral RNA copy numbers in the cell supernatant at 48 h post infection (p.i.). The data summarized in Fig. 1a and Supplementary Table S1 show that, at all MOIs (0.01, 0.02, 0.2, and 0.8), the 50% maximal effective concentra- tion (EC 50 ) for CQ (2.71, 3.81, 7.14, and 7.36 μM) was lower than that of HCQ (4.51, 4.06, 17.31, and 12.96 μM)

The differences in EC 50 values were statistically significant at an MOI of 0.01 (P< 0.05) and MOI of 0.2 (P< 0.001) (Supplementary Table S1). It is worth noting that the EC 50 values of CQ seemed to be a little higher than that in our previous report (1.13 μM at an MOI of 0.05) 1 , which is likely due to the adaptation of the virus in cell culture that significantly increased viral infectivity upon con- tinuous passaging.

Consequently, the selectivity index (SI =CC 50 /EC 50 ) of CQ (100.81, 71.71, 38.26, and 37.12) was higher than that of HCQ (55.32, 61.45, 14.41, 19.25) at MOIs of 0.01, 0.02, 0.2, and 0.8, respectively. These results were corroborated by immunofluorescence microscopy as evidenced by different expression levels of virus nucleoprotein (NP) at the indicated drug con- centrations at 48 h p.i. (Supplementary Fig. S1).

Taken together, the data suggest that the anti-SARS-CoV-2 activity of HCQ seems to be less potent compared to CQ, at least at certain MOIs

Both CQ and HCQ are weak bases that are known to elevate the pH of acidic intracellular organelles, such as endosomes/lysosomes, essential for membrane fusion 5 .In addition, CQ could inhibit SARS-CoV entry through changing the glycosylation of ACE2 receptor and spike protein 6 .

Time-of-addition experiment confirmed that HCQ effectively inhibited the entry step, as well as the post-entry stages of SARS-CoV-2, which was also found upon CQ treatment (Supplementary Fig. S2)

explore the detailed mechanism of action of CQ and HCQ in inhibiting virus entry, co-localization of virions with early endosomes (EEs) or endolysosomes (ELs) was ana- lyzed by immunofluorescence analysis (IFA) and confocal microscopy. Quantification analysis showed that, at 90 min p.i. in untreated cells, 16.2% of internalized virions (anti-NP, red) were observed in early endosome antigen 1 (EEA1)-positive EEs (green), while more virions (34.3%) were transported into the late endosomal–lysosomal protein LAMP1 + ELs (green) (n> 30 cells for each group). By contrast, in the presence of CQ or HCQ, significantly more virions (35.3% for CQ and 29.2% for HCQ; P< 0.001) were detected in the EEs, while only very few vir- ions (2.4% for CQ and 0.03% for HCQ; P< 0.001) were found to be co-localized with LAMP1 + ELs (n> 30 cells) (Fig. 1b, c).

This suggested that both CQ and HCQ blocked the transport of SARS-CoV-2 from EEs to ELs, which appears to be a requirement to release the viral genome as in the case of SARS-CoV

Interestingly, we found that CQ and HCQ treatment caused noticeable changes in the number and size/mor- phology of EEs and ELs (Fig. 1c)

In CQ- and HCQ-treated cells, abnormally enlarged EE vesicles were observed (Fig. 1c, arrows in the upper panels), many of which are even larger than ELs in the untreated cells

Within the EE vesicles, virions (red) were loca- lized around the membrane (green) of the vesicle. CQ treatment did not cause obvious changes in the number and size of ELs; however, the regular vesicle structure seemed to be disrupted, at least partially. By contrast, in HCQ-treated cells, the size and number of ELs increased significantly (Fig. 1c, arrows in the lower panels)

Since acidification is crucial for endosome maturation and function, we surmise that endosome maturation might be blocked at intermediate stages of endocytosis, resulting in failure of further transport of virions to the ultimate releasing site.

To our knowledge, there is a lack of studies on the impact of HCQ on the morphology and pH values of endosomes/ lysosomes. Our observations suggested that the mode of actions of CQ and HCQ appear to be distinct in certain aspects

It has been reported that oral absorption of CQ and HCQ in humans is very efficient.

In animals, both drugs share similar tissue distribution patterns, with high con- centrations in the liver, spleen, kidney, and lung reaching levels of 200–700 times higher than those in the plasma 10

It was reported that safe dosage (6–6.5 mg/kg per day) of HCQ sulfate could generate serum levels of 1.4–1.5 μMin humans 11 . Therefore, with a safe dosage, HCQ con- centration in the above tissues is likely to be achieved to inhibit SARS-CoV-2 infection

Other than its direct antiviral activity, HCQ is a safe and successful anti-inflammatory agent that has been used extensively in autoimmune diseases and can significantly decrease the production of cytokines and, in particular, pro-inflammatory factors. Therefore, in COVID-19 patients, HCQ may also contribute to attenuating the inflammatory response.

In conclusion, our results show that HCQ can efficiently inhibit SARS-CoV-2 infection in vitro. In combination with its anti-inflammatory func- tion, we predict that the drug has a good potential to combat the disease.

This possibility awaits confirmation by clinical trials. We need to point out, although HCQ is less toxic than CQ, prolonged and overdose usage can still cause poisoning. And the relatively low SI of HCQ requires careful designing and conducting of clinical trials to achieve efficient and safe control of the SARS-CoV-2 infection

Chloroquine does not inhibit infection of human lung cells with SARS-CoV-2

Chloroquine is an anti-malaria drug that is used for the treatment of COVID-19 as it inhibits the spread of SARS-CoV-2 in the African green monkey kidney-derived cell line Vero1–3.

Here we show that engineered expression of TMPRSS2, a cellular protease that activates SARS-CoV-2 for entry into lung cells4, renders SARS-CoV-2 infection of Vero cells insensitive to chloroquine.

Moreover, we report that chloroquine does not block infection with SARS-CoV-2 in the TMPRSS2-expressing human lung cell line Calu-3.

These results indicate that chloroquine targets a pathway for viral activation that is not active in lung cells and is unlikely to protect against the spread of SARS-CoV-2 in and between patients.

Chloroquine and hydroxychloroquine increase the endosomal pH of cells and inhibit viruses that depend on low pH for cell entry7.

We investigated whether these drugs could also block the cell entry by SARS-CoV-2 and whether entry inhibition accounted for the pre- vention of infection with SARS-CoV-2.

Moreover, we investigated whether entry inhibition is cell-type-dependent, as the virus can use pH-dependent and pH-independent pathways for entry into cells. The spike (S) protein of SARS-CoV-2, which mediates viral entry, is acti- vated by the endosomal-pH-dependent cysteine protease cathepsin L (CTSL) in some cell lines4. By contrast, entry into airway epithelial cells, which express low levels of CTSL 8 , depends on the pH-independent, plasma-membrane-resident serine protease TMPRSS2 4 . Notably, the use of CTSL by coronaviruses is restricted to cell lines8–10, whereas TMPRSS2 activity is essential for the spread and pathogenesis of the virus in the infected host11,12.

We compared the inhibition by chloroquine and hydroxychloro - quine of S-mediated entry into Vero (kidney), TMPRSS2-expressing Vero and Calu-3 (lung) cells.

Calu-3 cells, as with the airway epithe- lium, express low amounts of CTSL 8 and SARS-CoV-2 entry into these cells is dependent on TMPRSS24. By contrast, entry of SARS-CoV-2 into Vero cells is CTSL-dependent, and both CTSL and TMPRSS2 support entry into TMPRSS2-expressing Vero cells 4 . As a control, we used camostat mesylate, which inhibits TMPRSS2-dependent entry into cells4

Treatment with camostat mesylate did not interfere with cell viabil- ity, whereas chloroquine and hydroxychloroquine slightly reduced the viability of Vero, TMPRSS2-expressing Vero and Calu-3 cells when applied at the highest concentration (Fig. 1a)

Inhibition of S-driven entry by camostat mesylate was observed only in TMPRSS2 + cell lines, as expected (Fig. 1a and Table 1).

Moreover, chloroquine and hydroxy- chloroquine inhibited S-driven entry into TMPRSS2− Vero cells with high efficiency whereas the inhibition of entry into TMPRSS2 + Calu-3 and TMPRSS2 + Vero cells was inefficient and absent, respectively (Fig. 1a and Table 1). Therefore, chloroquine and hydroxychloroquine can block S-driven entry, but this inhibition is cell-line-dependent and efficient inhibition is not observed in TMPRSS2+ lung cells

We next investigated whether the cell-type-dependent differences in entry inhibition translated into differential inhibition of authen- tic SARS-CoV-2. Indeed, chloroquine efficiently blocked SARS-CoV-2 infection of Vero kidney cells, as expected1, but did not efficiently inhibit SARS-CoV-2 infection of Calu-3 lung cells (Fig. 1b, c)

A subtle reduction in SARS-CoV-2 infection was seen in the presence of 100 μM chloroquine, consistent with the modest inhibition of cellular entry of S-bearing pseudotypes under those conditions (Fig. 1a), but this effect was not statistically significant.

In summary, chloroquine did not effi- ciently block the infection of Calu-3 cells with S-bearing pseudotypes and authentic SARS-CoV-2, indicating that—in these cells—chloroquine does not appreciably interfere with viral entry or the subsequent steps of the viral replication cycle

Fig. 1 | Chloroquine does not block infection of human lung cells with

SARS-CoV-2. a, Vero, TMPRSS2-expressing Vero and Calu-3 cells were

preincubated for 2 h with the respective inhibitors (0 μM, 0.01 μM, 0.1 μM,

1 μM, 10 μM or 100 μM) and then inoculated with replication-defective

vesicular stomatitis virus reporter particles bearing the S protein. Top, the

transduction efficiency of the virus was assessed. Bottom, cells were not

inoculated with virus particles but cell viability after drug treatment was

instead assessed at the same time as transduction was quantified.

Transduction efficiency was quantified by measuring virus-encoded luciferase

activity in cell lysates. Cell viability was measured using the CellTiter-Glo assay.

Data are mean ± s.e.m. of three biological replicates, each of which consisted of

quadruplicate samples. Data were normalized as the relative entry efficiency

or cell viability of inhibitor-treated cells compared with those of untreated

cells (set to 100%). The calculated 50% inhibitory concentration (IC50) values

are summarized in Table 1. b, Untreated or chloroquine-preincubated Vero

and Calu-3 cells were inoculated with SARS-CoV-2 Munich isolate (patient

isolate 929, BetaCoV/Munich/BavPat1/2020|EPI_ISL_406862) at a multiplicity

of infection (MOI) of 0.001. After inoculation for 24 h, viral RNA was isolated

from the culture supernatant (extracellular virus) (dark blue) and the infected

cells (intracellular virus) (light blue), and SARS-CoV-2 genome equivalents (GE)

were determined by quantitative PCR with reverse transcription. Data are

mean ± s.e.m. of three biological replicates, each of which consisted of single

samples. c, The experiment was conducted as described in b, but the number

of infectious SARS-CoV-2 particles in culture supernatants was determined

by plaque titration using Vero E6 cells. PFU, plaque-forming units.

Statistical significance was analysed by two-way analysis of variance (ANOVA)

with Dunnett’s post hoc test. NS, not significant (P > 0.05); *P ≤ 0.05;

**P ≤ 0.01; ***P ≤ 0.001. P values (from left to right) are as follows. a, Entry

efficiency (camostat mesylate/chloroquine/hydroxychloroquine), Vero

(0.9999/0.8587/0.9997, 0.9842/0.9846/0.3904, 0.6860/0.0991/0.0223,

0.9968/0.0001/0.0001, 0.9997/0.0001/0.0001), TMPRSS2-expressing

Vero (0.9999/0.9968/0.9795, 0.1251/0.9962/0.9998, 0.0004/0.9997/

0.9999, 0.0001/0.9967/0.9982, 0.0001/0.9981/0.9986; Calu-3 (0.9900/

0.9999/0.9986, 0.0003/0.9999/0.9983, 0.0001/0.9988/0.9929, 0.0001/

0.1291/0.9938, 0.0001/0.0005/0.0045); cell viability (camostat mesylate/

chloroquine/hydroxychloroquine), Vero (0.9273/0.9999/0.9999, 0.9999/

0.8710/0.9642, 0.9999/0.9996/0.9999, 0.9999/0.8958/0.4818, 0.9998/

0.0838/0.0161), TMPRSS2-expressing Vero (0.9998/0.9999/0.9959,

0.9811/0.9985/0.9362, 0.9998/0.9985/0.9997, 0.9997/0.8835/0.9998,

0.9999/0.0315/0.1422), Calu-3 (0.9986/0.9999/0.9999, 0.9999/0.9997/0.9999,

0.9986/0.9999/0.8134, 0.9924/0.9275/0.7125, 0.9983/0.0492/0.0002).

b, (extracellular/intracellular), Vero (0.6844/0.6989, 0.0121/0.0002, 0.0002/

0.0001), Calu-3 (0.9434/0.8800, 0.9999/0.8830, 0.0517/0.3924). c, (extracellular/

intracellular), Vero (0.9561, 0.0001, 0.0001), Calu-3 (0.1184, 0.9997, 0.0987)

Research to confirm our results in primary respiratory epithelium is ongoing. Moreover, virus production in Calu-3 cells relative to Vero E6 cells was more robust in the present study compared with a previously published study 13 , potentially due to the use of the Calu-3 subclone 2B4 in the previous but not the present study. Nevertheless, our results sug- gest that chloroquine and hydroxychloroquine will exert no antiviral activity in human lung tissue and will not be effective against COVID-19, in keeping with the results of recent clinical trials 14,15 .

Moreover, our results highlight the fact that cell lines that mimic important aspects of respiratory epithelial cells should be used when analysing the antiviral activity of compounds that target host cell functions.

Cells Vero 76, Vero 76 stably expressing TMPRSS2 (both used for pseudotype experiments)4, the Vero 76 subclone Vero E6 (used for SARS-CoV-2 experiments), HEK293T and Calu-3 cells 16 were cultured in Dulbecco’s modified Eagle’s medium (DMEM) or minimum essential medium (MEM, Calu-3) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin. In case of Calu-3 cells, the medium was also supplemented with 1% non-essential amino acids and 1% sodium pyruvate. All cell lines were incubated at 37 °C and 5% CO 2 and were obtained from repositories (Vero E6 and HEK293T) or collaborators (Calu-3 and Vero 76). Cell lines were free of mycoplasma, authenticated on the basis of morphology and growth properties and confirmed by PCR to be of the correct species. The cell lines used were not listed as commonly misidentified cell lines by the ICLAC register

Cells Vero 76, Vero 76 stably expressing TMPRSS2 (both used for pseudotype experiments)4, the Vero 76 subclone Vero E6 (used for SARS-CoV-2 experiments), HEK293T and Calu-3 cells 16 were cultured in Dulbecco’s modified Eagle’s medium (DMEM) or minimum essential medium (MEM, Calu-3) supplemented with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin. In case of Calu-3 cells, the medium was also supplemented with 1% non-essential amino acids and 1% sodium pyruvate. All cell lines were incubated at 37 °C and 5% CO 2 and were obtained from repositories (Vero E6 and HEK293T) or collaborators (Calu-3 and Vero 76). Cell lines were free of mycoplasma, authenticated on the basis of morphology and growth properties and confirmed by PCR to be of the correct species. The cell lines used were not listed as commonly misidentified cell lines by the ICLAC register.

Production of pseudotyped particles Vesicular stomatitis virus (VSV) particles pseudotyped with SARS-CoV-2 S were generated according to published protocols 4,17 . At 24 h after transfection, HEK293T cells expressing the S protein were inoculated with a replication-restricted, VSV-G-trans-complemented VSV, which lacks the genetic information for VSV-G but instead encodes the reporter genes eGFP (enhanced green fluorescent protein) and FLuc (firefly luciferase), VSV*ΔG-FLuc 18 (provided by G. Zimmer). After 1 h of incubation at 37 °C and 5% CO 2 , the inoculum was aspirated and the cells were washed with phosphate-buffered saline (PBS) before culture medium was added. The culture medium was further supplemented with the culture supernatant from I1-hybridoma cells (CRL-2700 cells, ATCC) containing anti-VSV-G antibody (1:1,000) to inactivate residual input virus. After an incubation period of 18 h at 37 °C and 5% CO2, the culture supernatant was collected, centrifuged to pellet cellular debris, and the clarified supernatant was aliquoted and stored at −80 °C until further use

Transduction of target cells with pseudotypes and its inhibition For transduction experiments, Vero, TMPRSS2-expressing Vero and Calu-3 cells were grown in 96-well plates and allowed to reach about 50–70% confluency. Then, cells were preincubated with medium containing different concentrations (10 nM, 100 nM, 1 μM, 10 μM or 100 μM) of camostat mesylate (Sigma-Aldrich), chloroquine or hydroxy- chloroquine (both Tocris) or DMSO (Roth, solvent control) for 2 h at 37 °C and 5% CO 2 , before they were inoculated with S-bearing VSV. At 18 h after transduction, culture supernatants were aspirated and cells were lysed by incubation (30 min, room temperature) with Cell Culture Lysis Reagent (Promega). Cell lysates were subsequently transferred to white, opaque-walled 96-well plates and FLuc activity was quantified as an indicator of transduction efficiency, using the Beetle-Juice substrate (PJK) and a Hidex Sense plate reader (Hidex) operated with Hidex plate reader software (version 0.5.41.0, Hidex). Raw luminescence values (indicating luciferase activity) were recorded as counts per second. For normalization, transduction of DMSO-treated cells was set to 100% and the relative transduction efficiencies in the presence of camostat mesylate, chloroquine or hydroxychloroquine were calculated. Trans- duction experiments were performed in technical quadruplicates using three separate pseudotype preparations.

SARS-CoV-2 infection of target cells and its inhibition Virus infections were done with SARS-CoV-2 Munich isolate 929. Vero E6 or Calu-3 cells were seeded at densities of 3.5 × 105 cells per ml or 6 × 105 cells per ml in 12-well plates, respectively. After 24 h, cells were incu- bated with chloroquine (1 μM, 10 μM or 100 μM) or left untreated (con- trol) for 1 h at 37 °C. Subsequently, cells were infected with an MOI of 0.001 in serum-free OPTIpro medium containing the above-mentioned chloroquine concentrations at 4 °C for 30 min to enable virus attach- ment. Afterwards, infection medium was removed and the wells were washed twice with PBS and DMEM supplemented with chloroquine was added as described above and the plates were incubated at 37 °C. Samples were taken at 24 h after infection. Infection experiments were conducted with biological triplicates in a biosafety level 3 laboratory

Viral RNA extraction and quantitative RT–PCR For viral RNA extraction from supernatants, 50 μl of cell culture superna- tant was mixed with RAV1 lysis buffer (Macherey-Nagel) followed by an incubation at 70 °C for 10 min. RNA extraction was performed as recom- mended by the manufacturer (Macherey-Nagel). For intracellular viral RNA extraction, cells were washed with PBS and lysed with TRIzol (Zymo). SARS-CoV-2 genome equivalents were detected by quantitative RT–PCR targeting the SARS-CoV-2 E gene as previously reported 19 , using the following primers: E_Sarbeco_F, ACAGGTACGTTAATAGTTAATAGCGT; E_Sarbeco_P1, FAM-ACACTAGCCATCCTTACTGCGCTTCG-BBQ; E_Sarbeco_R, ATATTGCAGCAGTACGCACACA. The quantitative RT–PCR experiment and data processing were carried out using the LightCy- cler 480 Real-Time PCR System (Roche) and LightCycler 480 Software (version 1.5, Roche Molecular Systems). Absolute quantification was performed using SARS-CoV-2-specific in vitro-transcribed RNA stand- ards, as previously described19.

Plaque assay Infectious SARS-CoV-2 plaque-forming units were quantified by plaque titration on Vero E6 cells, as previously described20, with minor modi- fications. Vero E6 monolayers were seeded in 24-well plates, washed with PBS, incubated with serial dilutions of SARS-CoV-2-containing cell culture supernatants in duplicates, and overlaid with 1.2% Avicel in DMEM, supplemented as described above. After 72 h, cells were fixed with 6% formaline and visualized by crystal violet staining.

Cell viability assay The cell viability was quantified using the CellTiter-Glo assay (Promega) and using the same experimental conditions as described above for transduction experiments with the exception that cells were not inocu- lated with virus particles. In brief, cells were preincubated for 2 h at 37 °C and 5% CO2 with medium containing different concentrations (10 nM, 100 nM, 1 μM, 10 μM or 100 μM) of camostat mesylate, chloroquine or hydroxychloroquine, or DMSO (solvent control), before culture medium was added (instead of medium containing VSV pseudotyped with SARS-CoV-2 S) and cells were further incubated for 18 h. Next, intra- cellular ATP levels were quantified as an indicator of cell viability. For this, culture supernatants were aspirated and cells were lysed by incuba- tion with CellTiter-Glo substrate for 30 min at room temperature. Cell lysates were subsequently transferred to white, opaque-walled 96-well plates and luminescence was measured using a Hidex Sense plate reader (Hidex). Luminescence values (indicating cell viability) were recorded as absolute counts over a period of 200 ms per well. For normalization, cell viability of control-treated cells was set to 100% and the relative viability of cells incubated in the presence of camostat mesylate, chloroquine or hydroxychloroquine was calculated. Cell viability experiments were performed in technical quadruplicates and repeated with three sepa- rately prepared dilution series of the inhibitors