Abstract
COVID-19 has infected more than 41 million people worldwide with over 1.1 million deaths and is caused by the severe acute respiratory syndrome coronavirus (CoV) 2 (SARS-CoV-2)(1). Currently there are no protective vaccinations available and the only antiviral therapy in active use in patients is remdesivir, which provides only limited benefit(2, 3). Hence, an urgent need for antiviral therapies against SARS-CoV-2 exists. SARS-CoV requires Ca2+ ions for host cell entry and based on the similarity between SARS-CoV and SARS-CoV-2 it is highly likely that the same requirements exist for both viruses(4, 5). Here, we tested whether FDA-approved calcium channel blocker (CCB) drugs can inhibit SARS-CoV-2 infection in cell culture. All the CCBs showed varying degrees of inhibition, with amlodipine and nifedipine strongly limiting SARS-CoV-2 entry and infection in epithelial lung cells at concentrations where cell toxicity was minimal. Further studies with pseudo-typed particles carrying the SARS CoV 2 Spike protein suggest that viral inhibition occurs at the level of viral host cell entry. Overall, our data suggest that CCBs have a high potential to treat SARS-CoV-2 infections and their current FDA approval would allow for a fast repurposing of these drugs.
Significance Covid-19 infections are still increasing around the globe and a number of countries are currently facing a second wave of infection resulting in re-instated lockdowns and dramatic consequences for the public health systems. Vaccine developments are in progress but as of now there are no efficacious drugs on the market to fight the pandemic. Here, we present the first case of a FDA-approved class of drugs that inhibit SARS CoV 2 growth in vitro. Their FDA approval may reduce the time for repurposing, allowing bypass of time-and cost-intensive animal models and to test them directly in clinical trials for human application for treatment of Covid-19.
Introduction
Coronaviruses (CoVs) are major zoonotic pathogens that cause respiratory and/or enteric tract infections in a variety of species, including humans. Most CoVs that are pathogenic to humans cause only mild cold-like disease symptoms(6). However, currently the worldwide COVID-19 pandemic is caused by the severe acute respiratory syndrome (SARS) CoV 2 and poses a dramatic risk to public health worldwide(7). The virus was first identified in December 2019 in Wuhan, China and since has spread all over the globe(8). To date, there are no vaccines or highly efficacious drugs available against SARS-CoV-2.
Early during the pandemic chloroquine and hydroxychloroquine were promoted as efficacious treatments against Covid-19. But a recent report dismissed both drugs showing that they suppress viral growth in the epithelial kidney cell line Vero E6 but not in the relevant epithelial lung Calu-3 cells (9). Remdesivir, initially developed for Ebola treatment, is currently the only antiviral therapy used to effectively treat COVID-19 patients and was shown to improve symptoms in 68% of treated patients in clinical trials(2, 3). However, with only one efficacious antiviral therapy on the market, there remains a need to identify and to provide additional drugs that can treat COVID-19 patients. Given the current pandemic and the need for expediency, these drugs should ideally be FDA approved so they can be rapidly repurposed for COVID-19 treatment.
A promising target for a potential drug is a crucial step in the viral life cycle, such as the entry of the virus into the host cells(5). Previous studies have revealed that SARS-CoV and MERS-CoV utilize calcium ions (Ca2+) for viral entry that are coordinated by amino acid residues within the fusion peptide of their Spike (S) proteins(4, 10). Depleting intracellular and/or extracellular Ca2+ resulted in full suppression of viral entry for SARS-CoV and partial reduction of viral fusion of MERS-CoV. Sequence comparisons show that the fusion domains of SARS-CoV and SARS-CoV-2 are virtually identical, strongly suggesting that the same Ca2+-dependency for viral entry exists for SARS-CoV-2(5).
The crucial role of Ca2+ in the viral entry of CoVs prompted us to explore whether FDA approved Ca2+ channel blocker (CCB) drugs have the potential to inhibit viral growth of SARS-CoV-2. We chose five drugs from different classes that all inhibit high voltage-activated Ca2+ channels of the L-type. We selected the dihydropyridines: amlodipine, nifedipine and felodipine; the phenylalkylamine: verapamil; and the benzothiazepine: diltiazem. These five drugs are primarily used to treat cardio-vascular diseases. In addition, we also tested the FDA-approved Ca2+ chelator drug: diethylenetriaminepentaacetic acid (DTPA), which is used to treat radioactive contamination of internal organs.
Results and Discussion
We infected Vero E6 cells at a MOI of 0.1 and added four different concentrations of each compound immediately post infection (10 μM, 50 μM, 100 μM and 500 μM). Viral titers were analyzed 24 hours post infection. For amlodipine, we observed a 5-log reduction of viral growth at 50 μM and we were not able to detect any viral titer at a concentration of 100 μM (Figure 1A). However, amlodipine also exhibited a 25% reduction of cell viability at a concentration of 50 μM, which increased to 90% at 100 μM and 500 μM (Figure 1A). Nifedipine reduced viral titers by 5 logs at 500 μM and expressed about 15% cytotoxicity (Figure 1B). Felodipine completely suppressed growth of SARS-CoV-2 at a concentration of 50 μM with no statistically significant cytotoxicity (Figure 1C). 100 μM verapamil reduced the viral titers by 50% compared to the untreated control without a cytotoxic effect (Figure 1D). Diltiazem reduced viral growth at 500 μM but also compromised cell viability at this concentration and DTPA showed no effect Figure 1E and 1F).
Vero E6 cells are a good model system for SARS-CoV-2 infections, but the cell line is derived from epithelial kidney cells and may not represent the virus-drug-cell interactions in the lung. Therefore, we tested the CCBs and DTPA in the epithelial lung cell line Calu-3 (Figure 2). The results for amlodipine were comparable to what we found in Vero E6 cells (Figure 1A and 2A). Nifedipine and felodipine, however, significantly inhibited viral growth at lower concentrations compared to Vero E6 cells (Figure 2B and 2C). Nifedipine reduced the viral titers by 1.5 logs at a concentration of 100 μM and no virus was detectable at 500 μM while cytotoxicity was moderate (Figure 2B). Felodipine diminished SARS-CoV-2 growth by half at 10 μM and at 50 μM no virus was detected with no cytotoxic effect on the cells (Figure 2C). In contrast, verapamil had a weakened inhibitory effect at 100 μM compared to Vero E6 cells (Figure 2D). As in Vero E6 cells, 100 μM verapamil fully suppressed viral growth but also compromised cell viability by about 90%. We found very modest infection inhibition for diltiazem and DTPA at non-cytotoxic concentrations of 500 μM and 100 μM, respectively (Figure 2E and 2F).
With Amlodipine, Felodipine and Nifedipine being the most promising candidates that suppressed SARS CoV 2 growth in epithelial lung cells we aimed to determine their selectivity index (SI). The SI provides a measure how safe a given drug would be for the treatment of SARS CoV 2 infections in vivo. To calculate the SI we determined for each of the three drugs at which concentration they inhibit 50% of the viral growth (EC50) and the concentration at which each drug exerts 50% cytotoxicity (CC50). Dividing CC50/EC50 results in the SI and the higher the SI the more selective the drug is against the pathogen. However, it has to be emphasized that the SI determined here serves as an indicator and provides important information to consider for human application but the final efficacy in humans may differ and potential side effects of the drugs cannot be taken into account. Calu-3 cells as the more relevant cell line compared to Vero cells were infected with SARS CoV 2 and different concentrations of the respective drugs were applied (SI Figure 1). The obtained EC50 values were 10.36 μM for Amlodipine, 0.01255 μM for Felodipine and 20.47 μM for Nifedipine (SI Figure 1A - C, Table 1). We determined the CC50s in a similar fashion using a cytotoxicity assay as described above resulting in CC50 values of 27.85 μM for Amlodipine and 122.8 μM for Felodipine (SI Figure 1A and B). For Nifedipine the highest tested concentration of 2 mM did not significantly impair cell viability (SI Figure 1C, Table 1). The obtained SI values were 2.69 for Amlodipine and 9784.86 for Felodipine (Table 1). Because Nifedipine did not show a significant cytotoxic effect at approx. 7x the concentration of the most efficacious antiviral concentration of 300 μM a SI value could not be determined. Hence, all three drugs seem to be very selective against SARS CoV 2 with Felodipine and Nifedipine being the most promising candidates because of their high SI score and very low cytotoxicity, respectively.
It is important to emphasize that felodipine and nifedipine work with higher efficiency in Calu-3 cells compared to Vero E6 cells (Figure 1 and 2). In this context, it is noteworthy that CCBs are believed to affect airway smooth muscle cells, which rely on L-type calcium channels for their contraction(11–13). Recently, the same CCBs tested here were shown to have a beneficial effect on the lung function of asthma patients, providing evidence that CCBs act in the lungs(14). More evidence that interference with the cellular Ca2+ balance suppresses viral growth comes from nebulizers used for asthma medications. Nebulizers are supplemented with EDTA as a preservative, which may help to deplete extracellular Ca2+ (15) and provide a clue as to why asthma is not on the top ten list of chronic health problems of people who died of COVID-19.
The observed inhibition differences in these two cell lines suggest that inhibition may occur at the level of host cell entry. In Vero cells, CoVs preferentially enter the host cells via the endocytic pathway (5, 16). Calu-3 cells, on the contrary, are predominantly infected via plasma membrane fusion(5, 17). To further investigate whether CCB-mediated inhibition of SARS-CoV-2 infectivity affects viral host cell entry, we utilized murine leukemia virus (MLV)-based pseudo particles (PP) that were decorated with the SARS-CoV-2 S protein(18). These virion surrogates allow for only one infection cycle without intracellular replication and thus, inhibition of PP infection would suggest that the CCBs affect host cell entry.
We pretreated all cells with the CCBs and DTPA for one hour before adding PPs carrying the SARS-CoV-2 S protein (Figure 3). Amlodipine suppressed PP infection at 50 μM in both, Vero E6 and Calu-3 cells (Figure 3). Nifedipine inhibited PP infection in Vero E6 and Calu-3 cells with 500 μM having the strongest effect. Felodipine inhibited PP entry in both cell lines at 50 μM. Consistent with what we found in the live virus infection study above, verapamil suppressed PP entry at 50 μM in Calu-3 and at 100 μM in Vero E6 cells. In contrast to our observations with SARS-CoV-2 live virus infections, diltiazem and DTPA also inhibited PP infections (Figure 1, 2 and 3). Diltiazem inhibited PP infection in Calu-3 starting at 50 μM, and in Vero E6, at 500 μM. DTPA had a much stronger effect, suggesting that it might be able to chelate Ca2+ ions in the extracellular space, and thus, prevent fusion of S with the membrane as previously described for SARS-CoV PPs in the presence of EGTA(4). The applied DTPA concentrations, however, seem sufficient to prevent a single host cell entry, but incapable of inhibiting SARS-CoV-2 spread over several rounds of viral replication under the conditions used here.
There is evidence from previous reports that CCBs inhibit viral infections at the cell entry level. Influenza A infections, for example, trigger an influx of Ca2+ which assists the endocytic uptake of the virus, while treatment with verapamil or diltiazem inhibits viral infection(19–21). New World hemorrhagic fever arenaviruses, such as the Junin virus, are reportedly sensitive to CCBs and treatment with CCBs blocks the entry of the virus into the host cell(22). However, the exact mechanism of how CCBs suppress SARS-CoV-2 infection needs to be addressed in a follow up study. CCBs were also shown to interfere with viral replication of several other viruses (e.g., Japanese Encephalitis virus, Zika virus, Dengue virus)(23). Hence, while our results point to an interference of SARS-CoV-2 at host cell entry, it is conceivable that CCB-mediated inhibition of viral spread occurs at other stages as well, including viral release. It will require further studies to determine whether CCBs interfere directly with virus-cell fusion, how these drugs may affect the different fusion pathways and to explore whether CCBs may inhibit other cell functions that subsequently lead to inhibition of host cell entry. Regardless of the exact mechanism, these examples demonstrate that CCBs exhibit an antiviral efficacy against a broad range of viral pathogens relevant for public health. They also show the broad requirement of Ca2+ ions for viral propagation, supported by other studies that report a Ca2+ requirement for major human pathogens like Ebola virus and Rubella virus(24, 25). Therefore, CCBs may represent a novel class of antiviral therapeutics against a broad range of major viral diseases that warrant further clinical studies, but especially to address the current crisis of Covid-19.
In summary, the results described above are promising and particularly Felodipine and Nifedipine may represent a viable treatment option against Covid-19. However, it is unclear how the efficacious doses found here and the SI score would translate into clinical use in human patients to treat Covid-19 infections. As a next step, a meta-analysis of patient data taking CBBs medication and the relation to the severity of Covid-19 (e.g. hospitalizations, intubations etc.) in these patients may provide further insights about the efficacy of CBBs to inhibit SARS CoV 2 in humans.
Methods
Cells and reagents
Vero E6 and Calu-3 cells were obtained from the American Type Culture Collection (ATCC). Cells were maintained in Dulbecco’s modified Eagle medium (DMEM) (Cellgro) supplemented with 25 mM HEPES (Cellgro) and 10% HyClone FetalClone II (GE) at 37° C and 5% CO2. For virus infections, cells were grown in Eagle’s Minimum Essential Medium (EMEM) (Cellgro) supplemented with 4% heat inactivated fetal bovine serum (FBS) (Gibco) at 37° C and 5% CO2.
The SARS-CoV-2 isolate USA-WA1/2020 was obtained from the Biological and Emerging Infections Resources Program (BEI Resources). Amlodipine, nifedipine, felodipine, verapamil, diltiazem and DTPA were purchased from Sigma Aldrich. 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MMT) was obtained from Thermo Fisher. Crystal violet was purchased from VWR. pCDNA3.1/SARS-CoV-2 S Wuhan Hu-1 was generously provided by Dr. David Veesler.
SARS-CoV-2 infections and TCID50 assays
Vero E6 and Calu-3 cells were grown to confluency under biosafety level-2 (BSL-2) conditions. Cells were then transferred to the BSL-3 lab, washed with DBPS and a volume of 200 μL infection media with SARS-CoV-2 at a MOI of 0.1 was added to the cells. Cells were then incubated at 37° C and 5% CO2 for 1 hour on a rocker. Cells were then supplemented with 800 μL infection media and appropriate concentrations of each drug were added. Amlodipine, nifedipine, felodipine and verapamil were dissolved in DMSO. Therefore, DMSO was used as a control at the same volume that was applied for the highest drug concentration. After 24 hours, the supernatants were harvested and stored at −80° C.
For the TCID50, Vero E6 cells were grown to confluency in 96 well plates and serial dilution of the samples were prepared. Undiluted and diluted samples were then added to the Vero E6 cells and grown for 72 hours at 37° C and 5% CO2. After 72 hours supernatants were aspirated, and cells were fixed with 4% paraformaldehyde. Cells were then stained with 0.5% crystal violet and subsequently washed with dH2O. Wells were scored and analyzed for living or dead cells according to the Reed-Muench method(26).
Cytotoxicity assay
The cytotoxicity of the calcium ion blocking drugs (amlodipine, nifedipine, felodipine, diltiazem) and calcium chelator (DTPA) on VeroE6 and Calu-3 cells were determined by an MTT Assay. A total of 5 × 105 VeroE6 cells/well, and 6.7 × 105 Calu-3 cells/well were incubated in the presence of the six calcium blocking drugs and calcium chelator at the concentrations of 10, 50, 100, and 500 μM for 24 hour at 37°C C and 5% CO2 in a CO2 incubator. After 24 hours incubation, cells were treated with MTT solution (5 mg/mL) and incubated for 4 hours at 37°C and 5% CO2 with rocking to allow purple formazan crystals to form. 50 μL DMSO was added into each well to dissolve the crystals. The absorbance was measured at 540 nm in a microplate reader (Bio-Tek Instrument Co., WA, USA).
Pseudo particle production and infections
Pseudo particles production and infections were performed as previously described with minor modification(10). For PP production, HEK 293T cells were transfected with pCDNA3.1/SARS-CoV-2 S Wuhan Hu-1 to generate SARS-CoV-2 S carrying PPs. Supernatants were harvested 72 hours post transfection and stored at −80°C. For infections, cells were seeded in 96-well plates and DMEM containing the different CCBs and DTPA was added and incubated for 1 hour. After one hour 50 μL PPs were added. After 24 hours, cells were replenished with DMEM containing the individual drugs at the appropriate concentration and incubated for another 48 hours. After lysis, luciferase activity was measured using a GloMax Navigator Microplate Luminometer (Promega). Data was analyzed using Microsoft Excel and GraphPad Prism 8.
Author contributions
MRS: study design, experimental execution, data analysis, writing and editing of the manuscript. MB: experimental execution and data analysis. TT: experimental execution and data analysis. GRW: study design, writing and editing of the manuscript. SD: study design, writing and editing of the manuscript, funding acquisition.
Acknowledgements
This projected was funded by Fast Grant, Mercatus Center. We would like to thank Paul Jeannette for his support for all the BSL-3 work and Dr. Luis Schang, Dr. Nihal Altan-Bonnet, Dr. Fernando Martinez, Dr. Bruce Kornreich and Dr. Hanno Andreas Ludewig for their critical input. TT is supported by the National Science Foundation Graduate Research Fellowship Program under Grant No. DGE-1650441.
Footnotes
Format changes in figures and text. Data added.