Abstract
The retrospective analysis of clinical data of patients suffering from COVID-19 has indicated that statin therapy, used to lower plasma cholesterol levels, is associated with a better clinical outcome. We therefore investigated the effect of statins on SARS-CoV-2 infection and found that selective statins reduced SARS-CoV-2 cell entry and inhibited high and low pathogenic coronavirus infection in human cells. A retrospective study on hospitalized patients with COVID-19 implies that reduced high density lipoprotein levels, which are typically counteracted by statin therapy, are associated with aggravated disease outcome. These results suggest that statin therapy poses no additional risk to individuals exposed to SARS-CoV-2 and that some statins may have a mild beneficial effect on COVID-19 outcome. Placebo controlled trials are required to clarify the role of statins in COVID-19 infected patients.
Main
Statins are prescribed to treat hypercholesterolemia, cardiovascular disease and diabetes in a significant proportion of the population, in particular in the elderly1, which are at high risk of developing severe symptoms upon SARS-CoV-2 exposure2. Two retrospective cohort studies report a correlation of statin therapy and reduced all-cause mortality of COVID-19 patients3,4. On the other hand, statins may increase expression of the ACE2 receptor of SARS-CoV-25,6. Whether statins not only influence disease outcome but also directly impact SARS-CoV-2 entry and replication is unknown.
Here, we used coronavirus cell culture models to assess a possible direct impact of statins on coronavirus infection. At a concentration of 5 μM statins moderately reduced infection of human cells with the low pathogenic alphacoronavirus 229E and this reduction of susceptibility was most pronounced and statistically significant for fluvastatin (Fig. 1a). The most commonly prescribed statins simvastatin and atorvastatin as well as lovastatin, pravastatin and rosuvastatin slightly reduced coronavirus 229E infection, but the observations did not reach statistical significance in the cell culture assay. Chloroquine efficiently blocked coronavirus 229E infection as expected7,8. When performing dose-response analyses for fluvastatin, we found a dose-dependent reduction of coronavirus infection with a half-maximal inhibitory concentration of 9.8 μM. At the investigated fluvastatin doses, cell viability remained at 80% and higher indicating no major cytotoxic effects of the tested fluvastatin doses (Fig. 1b).
a, Selected statins reduce human coronavirus 229E susceptibility of human cells. b, Dose-response curve of the antiviral and cytotoxic effect of fluvastatin in human cells. The dotted lines indicate 50% and 100% of infection in solvent control treated cells. c, Fluvastatin reduces SARS-CoV-2 entry into human respiratory epithelial cells. Calu-3 cells were pretreated with the indicated compounds and transduced with VSV-pseudotypes harboring the glycoproteins of SARS-CoV-2, SARS-CoV, VSV or no envelope glycoproteins (No env). RLU, relative light units. d-e, Fluvastatin decreases SARS-CoV-2 susceptibility of human respiratory epithelial cells. Virus genome copy numbers in cell lysates and supernatants from cells pretreated with fluvastatin or solvent control and infected with SARS-CoV-2 shown. Results are normalized to viral copy numbers in DMSO treated cells (d). Number of infectious SARS-CoV-2 particles in cell culture supernatant from (d) was determined by titration on VeroE6 cells (e). Mean ± SEM of three independent biological replicates shown. (a,d) One-way ANOVA, followed by Dunnetts multiple comparison test. (c) Two-way ANOVA, followed by Sidak s multiple comparison test. * p<0.05, ** p<0.005,**** p<0.0001.
Statins inhibit the 3-hydroxy-3-methyl-glutaryl-coenzyme A (HMG-CoA) reductase and thereby de novo cholesterol synthesis9. Notably, at a cellular level the 2003 SARS-CoV requires intact cholesterol-rich membrane microdomains to efficiently engage its receptor angiotensin-converting enzyme 2 (ACE2)10. As the current pandemic SARS-CoV-2 engages the same receptor11, we sought to investigate if statins would affect SARS-CoV-2 cell entry. Hence, we infected human respiratory epithelial cells with vesicular stomatitis virus (VSV) pseudoparticles displaying the SARS-CoV-2 spike protein. Such pseudoparticles mimic binding to the ACE2 receptor and SARS-CoV-2 spike mediated uptake of virions into respiratory epithelial cells.
Fluvastatin pretreatment (24h, 10 μM) reduced SARS-CoV-2 spike mediated cell entry five-fold (Fig. 1c). As expected, the statin also reduced cell entry mediated by the 2003 SARS-CoV spike. Entry of pseudoparticles displaying the V SV glycoprotein, which does not bind to ACE2, was not affected by fluvastatin pretreatment of cells. Notably, chloroquine could not block entry of SARS coronaviruses into respiratory epithelial cells, which is in line with recent reports that in these cells priming of the spike protein occurs at the plasma membrane and not in endosomal compartments12,13. Cholesterol depletion using cyclodextrin, however, reduced entry of SARS-CoV and SARS-CoV-2 pseudoparticles, while VSV pseudoparticle entry remained unaffected by cholesterol depletion. Our data suggest that statin induced cholesterol depletion may reduce SARS-CoV-2 entry into the respiratory epithelium.
To investigate whether statins could modulate infection with highly pathogenic SARS-CoV-2 isolates, we pretreated human respiratory epithelial cells with fluvastatin (10 μM) and then infected the cells with the Human 2019-nCoV ex China strain BavPat1/2020 of SARS-CoV-211 in the presence of fluvastatin. SARS-CoV-2 genome copy numbers moderately decreased in cells and supernatants upon fluvastatin treatment (Fig. 1d). Furthermore, the titer of released infectious SARS-CoV-2 was reduced by 4-fold upon fluvastatin treatment compared to DMSO treatment, as determined by TCID50 assay on VeroE6 cells (Fig. 1e). Collectively, our data highlight that statins do not promote coronavirus infection in human cells as was speculated, but instead selected statins attenuate coronavirus entry and infection in vitro.
Finally, we retrospectively analyzed a German patient cohort of 65 hospitalized intensive care unit patients with COVID-19 to assess an association of serum lipoproteins and statin usage with clinical outcome. Six patients received statins (3, simvastatin; 3, atorvastatin), but none received fluvastatin. As statins modulate serum lipoprotein levels and typically raise high density lipoprotein (HDL) levels, we assessed a possible link between COVID-19 outcome and HDL levels. Eight patients in the cohort had documented low HDL plasma levels (<20 mg/dl). The median age in the whole cohort versus the eight low HDL patients was 63 and 73 years, respectively, 28% and 38%, respectively were female, and the median length of ICU stay was 12 days and 16 days, respectively. The incidence of death in the low HDL group (7/8; 5.1 cases per 100 person-days; death rate 87.5%) was markedly higher than that in the remainder patient group (14/57; 1.6 cases per 100 person-days; death rate 24.6%) or in the whole cohort (21/65; 2.1 cases per 100 person-days; death rate 32.3%). An age matched control group of eight patients without statin therapy and without reported low HDL levels (median age 74, 25% female) showed a similar trend with a median length of ICU stay of 10 versus 16 in the low HDL group and an incidence of death of 37.5% (3/5; 2.9 cases per 100 person-days) versus 87.5% (7/8; 5.1 cases per 100 person-days) in the low HDL group (Table 1). Additional confounder effects such as blood group and gender were not addressed.
Our data suggest that statin treatment is associated with a mild attenuation of SARS-CoV-2 infection in vitro. In accordance, previous studies reported an amelioration of COVID-19 outcome in patients3,4,10. Although our COVID-19 intensive care unit cohort was too small to confirm this beneficial effect of statins on infection outcome, we could detect an association of low HDL levels and increased all-cause mortality. Reduced HDL levels in critical care COVID-19 patients were previously reported14, however cause and consequence of this observation need to be addressed in larger cohorts. As statins reduce LDL and may increase HDL levels, statin modulation of serum lipoprotein homeostasis may -in addition to immunomodulatory and other yet unknown effects - contribute to the previous observations of beneficial effects of statins in patients.
Considering that a significant fraction of the population, e.g. 30-40% of all type 2 diabetes diagnosed in Germany15 and 26% of all Americans over the age of 40 years, is on statin therapy1, the primary goal of our study was to assess direct effects of statin treatment on SARS-CoV-2 infection. The observed moderate inhibitory effect of statins on an alpha- and a betacoronavirus infection in vitro are in keeping with the reduced risk of severe COVID disease noted in patients taking statins and could point to a direct antiviral effect of statins. Given the limitations of retrospective studies, placebo controlled trials are required to clarify the role of statins in COVID-19 infected patients. It is important to note that the inhibitory doses observed in our cell culture experiments are an order of magnitude higher than statin serum concentrations in patients, which are in the nM range16. Statin concentrations in the respiratory tract are likely even lower than in the serum. This suggests that the direct effects observed on SARS-CoV-2 entry and infection may not account for the observations in the retrospective cohort studies. Instead, immunomodulatory effects of statins as previously reported17-20 may ameliorate the outcome of hospitalized COVID-19 patients. This is in line with the reported immunopathology in severe cases of SARS-CoV-2 infection and a beneficial effect of statins on neutrophil function in infection21 and pneumonia22. Investigations on the effect of statins on neutrophil function in animal models of COVID-19 are ongoing and will clarify the clinical observations.
A possible direct inhibition of the SARS-CoV-2 main protease by statins was postulated in an in silico docking study23. While this could explain the effects we observed in vitro on low and high pathogenic coronavirus infection, it remains to be proven if doses achieved in individuals would efficiently block infection. Notably, the inhibition of SARS-CoV-2 entry described here must be mediated by a different mechanism as the pseudotypes used to study entry do not express the SARS-CoV-2 protease. Likely, cholesterol depletion in the plasma membrane affects ACE2 interactions with the SARS-CoV-2 spike protein as previously reported for SARS-CoV10. We are currently investigating the protein interactions of the ACE2 receptor and their cholesterol dependency to shed light on this phenomenon.
Despite the discussed limitations of this study, cumulative data from this and other studies strongly argue that individuals on statin therapy can continue statin regimen during the current SARS-CoV-2 pandemic.
Data Availability
No additional data reported or deposited
Contributions
G.G. and F.J.Z.B. conceptualized the study. F.J.Z.B., R.M., L.L., G.B. and G.G. developed the methodology. F.J.Z.B., R.M., L.L., G.B., A.P.G. and J.K carried out investigations. D.W., S.K. and G.G. collected and analyzed clinical cohort data. G.G., R.M., L.L., J.K., D.W. and L.L. wrote the original draft of the manuscript. Review and editing of the manuscript were carried out by G.G. and T.P. G.G., T.P. and S.K. acquired funds and provided resources and supervision.
Competing Interests statement
The authors declare no competing interests.
Methods
Cell lines and virus strains
HEK293T cells, Huh7.5 cells constitutively expressing a firefly luciferase (Huh7.5 FLuc) and Calu-3 cells were maintained in DMEM supplemented with 10% fetal bovine serum, 2 mM glutamine, 0.1 mM non-essential amino acids and 1% Penicillin/Streptomycin at 37°C and 5% CO2. Each cell line was tested regularly for Mycoplasma contamination. The recombinant hCoV229E encoding a Renilla luciferase gene was produced in Huh7.5 cells at 33°C and tittered on Huh7.5 cells. The SARS-CoV-2 isolate Human 2019-nCoV ex China strain BavPat1/2020 was kindly provided by Christian Drosten (Charité, Berlin) through the European Virus Archive - Global (EVAg), amplified in VeroE6 cells, and handled under BSL3 conditions.
Reagents and Inhibitors
Fluvastatin (sigma, order no. SML0038), simvastatin (sigma, order no. S6196), atorvastatin (sigma, order no. PZ0001), lovastatin (sigma, order no. M2147), pravastatin (biomol, Cay-10010342) and rosuvastatin (sigma, order no. SML1264) were dissolved in DMSO. Chloroquine (sigma, C6628), dissolved in H2O, served as a positive control. For entry assays methyl-β-cyclodextrin (MβCD, Sigma, order no. C4555) was used as an additional control.
Effect of statins on hCoV229E
Huh7.5 FLuc cells were seeded in 96-well plates at a density of 1X104 cells/well in complete DMEM and incubated for 24 hours at 37°C (5% CO2). The cells were pretreated for 24 hours with either 5 μM of each compound or serial dilutions of the test compounds in a final volume of 100 μl. Following the pretreatment, cells were inoculated with hCoV229E at an MOI of 0.05 in the presence of the test compounds. The inoculum was removed 4 hours later and fresh medium containing the respective compounds was added. Cells were lysed 48 hours post inoculation in 50 ul of detergent containing buffer and Renilla luciferase activity as a measurement for infection was determined by mixing 20 μl of lysate with 60 μl of luciferase substrate solution (Coelenterazine, 0.42 mg/ml in methanol). To determine cell viability, Firefly luciferase activity was measured from the same lysate by adding D-luciferin (0.2 mM D-luciferin in 25 mM glycyl-glycine) as a substrate. Luciferase activity was measured in a Centro XS3 LB960 (Berthold Technologies) microplate reader.
VSV-pseudotype entry assay
VSV pseudotypes harboring no envelope glycoprotein, the VSV-G, the SARS-1-S or the SARS-2-S protein were produced as described previously24. Briefly, 293T cells were transfected with respective plasmids carrying the glycoproteins or an empty vector control. 24 hours later, cells expressing the respective glycoproteins at their surface were inoculated with a replication-deficient VSV*ΔG complemented with G which expresses a GFP as well as a firefly reporter gene instead of the VSV-G open reading frame. The inoculum was removed 2 hours later and fresh medium was added. To block residual VSV-G on the cell surface, VSV-G antibody (mouse hybridoma supernatant from CRL-2700) was added to the cells except to cells expressing VSV-G. Supernatant was harvested 16 hours later, cellular debris cleared by centrifugation and VSV-pseudoparticles were stored at -80°C.
For transduction, Calu-3 cells were seeded at a density of 5×104 cells/well in complete DMEM in a 96-well culture plate. 24 post plating, cells were pretreated with 10 μM of fluvastatin or chloroquine and the respective solvent controls for 24 hours. As control, some wells were pretreated for 60 minutes with either serum free medium or 7.5 mM MβCD prior to transduction. The cells were washed once with PBS and transduced with VSV-based pseudoparticles for 90 minutes at 37°C. The wells were replenished with either DMEM complete or serum free medium (for MβCD and serum free control). 16 hours post transduction, firefly luciferase activity was determined in cell lysates as described above.
SARS-2 infection
For infection experiments with SARS-CoV-2, the SARS-CoV-2 isolate Human 2019-nCoV ex China strain BavPat1/202011, kindly provided by Christian Drosten (Charité, Berlin) through the European Virus Archive - Global (EVAg), was propagated in VeroE6 cells (3 passages) after primary isolation from patient material. Calu-3 cells were seeded in collagen coated P24 well plates at a density of 4.5×105 cells/well. Cells were pretreated with 10 μM fluvastatin or DMSO for 24 hours and then infected with SARS-CoV-2 isolate at an MOI of 2.0×10-5 based on titration in VeroE6 cells. Heat-inactivated virus was used as a control to determine unspecific binding of virions and RNA. 4 hours post infection, the inoculum was removed, cells were washed twice with PBS and fresh medium containing the respective compound was added. The infection was stopped 48 hours later, viral RNA was isolated from cell lysates and supernatants using NucleoSpin RNA kit (Macherey-Nagel) and QIAamp Viral RNA Mini Kit (Qiagen), respectively, following the manufacturer’s instructions. RT-qPCR was performed in duplicates as described previously25 using a LightCycler 480 (Roche). In addition, SARS-CoV-2 titer in cell culture supernatants was quantified by TCID50 assay on VeroE6 cells. Therefore, VeroE6 cells were seeded in 96 well plates and incubated with serial dilutions of cell culture supernatant. 72h later, cells were fixed with 10% formalin and stained with crystal violet.
Cell viability Assay
Cell viability was determined at the end of the experiment either by measuring firefly luciferase activity in cell lysates (Huh7.5 Fluc cells) or by MTT assay. For the latter, MTT reagent (3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyltetrazolium bromide) was added to the cells at 0.5 mg/ml diluted in medium. The reaction was stopped with DMSO after 1 hour and optical density was measured at 570 nm using a plate reader (Synergy 2, BioTek®).
Statistical Analysis
Experiments were conducted in three biological replicates each with three technical replicates unless otherwise stated. All statistical analyses were performed using one or two-way analysis of variance (ANOVA) followed by Dunnett or Sidak’s multiple comparison test using GraphPad Prism software version 8 (GraphPad). p values <0.05 were considered as statistically significant.
Clinical cohort
From March to June 2020, 65 patients with a PCR confirmed diagnosis of COVID-19 were treated at the Department of Intensive Care Medicine at the University Medical Center Hamburg-Eppendorf. The median age was 63 years, 78% were male, the median length of ICU stay was 12 days, 63.5% received mechanical ventilation (median duration 15 days), 39.2% renal replacement therapy and 16% veno-venous extracorporeal membrane oxygenation. Overall ICU mortality was 28.3% and 34% in patients with need for mechanical ventilation.
Ethics approval
The need for an informed consent was waived by the ethics committee at the Hamburg State Chamber of Physicians (registration no.: WF-053/20), because data were retrieved retrospectively from electronic health records.
Acknowledgement
The work was funded by the Knut and Alice Wallenberg Foundation to G.G.; the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) - Projektnummer 158989968 - SFB 900 project C7 to G.G. and the DFG project GE 2145/3-2 to G.G.; the German Academic Exchange Service (DAAD) to J.K.; the Infection Biology International PhD Program of Hannover Biomedical Research School to R.M., the German Centre for Infection Research (DZIF) to A.P.G. and T.P., the Helmholtz Alberta Initiative for Infectious Disease Research (HAI-IDR), and the Shandong University Helmholtz International Laboratory to T.P. This publication was supported by the European Virus Archive GLOBAL (EVA-GLOBAL) project that has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 871029. We thank Christian Drosten for providing the clinical SARS-CoV-2 isolate, Stefan Pöhlmann for the VSV pseudotype system, Gert Zimmer for VSV*ΔG and Volker Thiel for the hCoV 229E strain, Yannic Becker for assistance as well as Thomas Schulz, Tommy Olsson and Anders Sjöstedt for constant support.
Footnotes
TCID50 data added to figure 1 and modifications in the text
