key: cord-326013-5i35zdmv authors: Carpinteiro, Alexander; Edwards, Michael J.; Hoffmann, Markus; Kochs, Georg; Gripp, Barbara; Weigang, Sebastian; Adams, Constantin; Carpinteiro, Elisa; Gulbins, Anne; Keitsch, Simone; Sehl, Carolin; Soddemann, Matthias; Wilker, Barbara; Kamler, Markus; Bertsch, Thomas; Lang, Karl S.; Patel, Sameer; Wilson, Gregory C.; Walter, Silke; Hengel, Hartmut; Pöhlmann, Stefan; Lang, Philipp; Kornhuber, Johannes; Becker, Katrin Anne; Ahmad, Syed A.; Fassbender, Klaus; Gulbins, Erich title: Pharmacological inhibition of acid sphingomyelinase prevents uptake of SARS-CoV-2 by epithelial cells date: 2020-10-29 journal: Cell Rep Med DOI: 10.1016/j.xcrm.2020.100142 sha: doc_id: 326013 cord_uid: 5i35zdmv The acid sphingomyelinase/ceramide system plays an important role in bacterial and viral infections. Here we report that either pharmacological inhibition of acid sphingomyelinase with amitriptyline, imipramine, fluoxetine, sertraline, escitalopram, or maprotiline, or genetic downregulation of the enzyme prevents infection of cultured cells or freshy isolated human nasal epithelial cells with SARS-CoV-2 or pseudoviral pp-VSV-SARS-CoV-2 particles expressing spike, a bona fide system mimicking SARS-CoV-2 infection. Infection activates acid sphingomyelinase and triggers a release of ceramide on the cell surface. Neutralization or consumption of surface ceramide reduces infection with pp-VSV-SARS-CoV-2 spike. Treating volunteers with a low dose of amitriptyline prevents infection of freshly isolated nasal epithelial cells with pp-VSV-SARS-CoV-2 spike. The data justify clinical studies investigating whether amitriptyline, a safe drug used clinically for almost 60 years, or other antidepressants that functionally block acid sphingomyelinase prevent SARS-CoV-2 infection. Infections with a novel member of the Coronaviridae family were reported in late 2019 in Wuhan, China 1 . The virus was named severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2). Subsequently, the virus spread globally and is responsible for the coronavirus disease 2019 (COVID- 19) pandemic. Infection with SARS-CoV-2 often results in mild respiratory tract disease, but a substantial number of patients also experience severe symptoms and pneumonia, and ∼70% of these critically ill patients require intensive care and ventilator treatment, with a mortality rate of 62% 2 . Even when the large number of only mildly affected patients are included, the mortality rates are higher than those associated with seasonal influenza 3,4 . Cellular infection with SARS-CoV-2 is initiated by the binding of the surface unit S1 of the viral spike glycoprotein to its cellular receptor angiotensin-converting enzyme 2 (ACE2), resulting in cleavage of the viral spike protein by the activity of transmembrane serine protease 2 (TMPRSS2) or cathepsin L, and in viral entry [5] [6] [7] [8] . Although the binding of the virus to its receptor has been elucidated in detail [6] [7] [8] , the changes that occur in the host cell membrane during viral processing and entry require definition. Membrane changes that mediate viral entry may be a very promising target for preventing the infection. Previous studies have used replication-deficient vesicular stomatitis virus (VSV) pseudoviral particles (pp-VSV) presenting coronavirus spike protein (pp-VSV-SARS-CoV-2 spike) on their surface. Studies have shown that these particles accurately reflect key aspects of the entry of coronavirus into host cells 5 . These particles were previously shown to bind to ACE2 for infectious entry, and entry was inhibited by anti-ACE2 antibodies 5 . Thus, these particles are a bona fide model for studying the events of SARS-CoV-2 entry. We have previously shown that acid sphingomyelinase and ceramide play an important role in receptor signaling and infection biology 9, 10 . Acid sphingomyelinase (EC 3.1.4.12, sphingomyelin phosphodiesterase; optimal pH 5.0) is a glycoprotein that functions as a lysosomal hydrolase, catalyzing the degradation of sphingomyelin to phosphorylcholine and ceramide. Acid sphingomyelinase is present in lysosomes, J o u r n a l P r e -p r o o f 5 but because these compartments are constantly recycling to the plasma membrane it can also be found on the cell surface 9, 10 . The activity of acid sphingomyelinase on the cell surface results in the formation of ceramide in the outer leaflet of the cell membrane. The generation of ceramide molecules within the outer leaflet alters the biophysical properties of the plasma membrane, because the very hydrophobic ceramide molecules spontaneously associate with each other to form small ceramide-enriched membrane domains that fuse and form large, highly hydrophobic, tightly packed, gel-like ceramide-enriched membrane domains [10] [11] [12] [13] . In addition, ceramide has been shown to directly bind to a variety of proteins, including cathepsin D 14 , phospholipase A 2 15 , ceramide-activated protein serine/threonine phosphatases (CAPP) 16 , protein kinase C isoforms 17, 18 , and LC3B-II 19 . Many antidepressants functionally inhibit acid sphingomyelinase activity [20] [21] [22] [23] [24] [25] . These cationic amphiphilic drugs indirectly inhibit acid sphingomyelinase activity by displacing the enzyme from lysosomal membranes, in particular intralysosomal vesicles, thereby releasing the enzyme into the lysosomal lumen and causing its partial degradation [20] [21] [22] [23] [24] [25] . We have previously shown that rhinovirus infections activate acid sphingomyelinase and lead to the formation of ceramide and ceramide-enriched membrane domains. Amitriptyline, sertraline, and other functional inhibitors of acid sphingomyelinase activity (FIASMAs) inhibit cellular infection with rhinovirus 26 . Similar observations have been made regarding infections with Ebola virus 27 demonstrating that amitriptyline and other FIASMAs inhibit infection with Ebola virus in vitro 27 . Because some antidepressants are widely used in clinical practice and have a very favorable safety profile, we investigated whether these drugs could be repurposed to treat or prevent infections with SARS-CoV-2. Our results show that acid sphingomyelinase is activated within 20 to 30 minutes To determine whether acid sphingomyelinase can be targeted to prevent the infection of cells with SARS-CoV-2, we treated Vero cells with the FIASMA amitriptyline [22] [23] [24] or left the cells untreated. We then exposed the cells to pp-VSV-SARS-CoV-2 spike Next, we determined whether amitriptyline also affects the infection of human Caco-2 cells with authentic SARS-CoV-2. Amitriptyline-pretreated Caco-2 cells were infected with SARS-CoV-2, and the replication of SARS-CoV-2 was monitored by reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR) using two separate primer sets and a plaque assay. The results showed that amitriptyline also reduced the infection of human cells with SARS-CoV-2 by approximately 90% (Fig. 2 ). J o u r n a l P r e -p r o o f 8 The results of these studies suggested that acid sphingomyelinase serves an important function in SARS-CoV-2 infection. To confirm the role of acid sphingomyelinase in infection with SARS-CoV-2 independent of pharmacological inhibition, we used shRNA-mediated suppression of the expression of acid sphingomyelinase in Caco-2 cells. We confirmed the downregulation of acid sphingomyelinase by measuring the activity of the enzyme in lysates of uninfected cells (Fig. 3A) . The results showed that genetic downregulation of acid sphingomyelinase prevented infection with pp-VSV-SARS-CoV-2 spike (Fig. 3A ). Control shRNA constructs had no effect on cellular infection with pp-VSV-SARS-CoV-2 spike. We further tested whether neutral sphingomyelinase may also be involved in cellular infection with SARS-CoV-2. To this end, we transfected Caco-2 cells with shRNA targeting neutral sphingomyelinase 2 and control shRNA. Transfection reduced neutral sphingomyelinase activity by approximately 60% but had no effect on cellular infection with pp-VSV-SARS-CoV-2 spike (Fig. 3A) . We next investigated whether infection with pp-VSV-SARS-CoV-2 spike activates the acid sphingomyelinase/ceramide system. The results showed that treating Vero cells with pp-VSV-SARS-CoV-2 spike resulted in rapid activation of acid sphingomyelinase and a concomitant release of ceramide (Fig. 3B, Fig. 4A ). In contrast, treating Vero cells with pp-VSV lacking SARS-CoV-2 spike or with pp-VSV-G particles did not result in the activation of acid sphingomyelinase or in a release of ceramide (Fig. 3B, Fig. 4A ), a finding indicating that the activation of acid sphingomyelinase and the release of ceramide are specifically induced by the spike protein. Pretreatment of the cells with 5, 10, 20, or 25 µM amitriptyline prevented the activation of acid sphingomyelinase and the release of ceramide upon infection with pp-VSV-SARS-CoV-2 spike for 30 min (Fig. 3B, Fig. 4A ). Controls showed that amitriptyline reduced both the constitutive and the viral-induced activity of acid sphingomyelinase (Fig. 3B ). Treating Vero cells with neutralizing antibodies to spike or with recombinant ACE2 protein prevented the activation of acid sphingomyelinase and the release of ceramide upon infection with pp-VSV-SARS-CoV-2 spike (Fig. 3B, Fig. 4A Amitriptyline and other drugs with similar structure and properties have been clinically used for many years (since 1962) to treat patients with depressive disorder. In the volunteer studies we used a low dose of amitriptyline, i.e., 0.5 mg/kg, which is lower than the dose usually used to treat patients with major depression (75-150 mg/day, i.e., approximately 1-2 mg/kg). At this dose, the drug has only mild effects, such as dry mucosa and tiredness. However, at higher dosages (higher than 100 mg/day, twice the dose we used or higher), amitriptyline and many other antidepressants may induce a cardiac QT elongation as a possible (but rare) adverse effect. QT elongation can result in cardiac arrhythmia and serious cardiac adverse effects 29 . Thus, it may be advisable to perform an electrocardiogram (ECG) before the first administration of amitriptyline or any of the other FIASMAs used here to exclude a pre-existing QT elongation. Amitriptyline and other antidepressants acting as FIASMAs exert no known adverse effects on the immune system. Our studies demonstrate that even low dosages of amitriptyline provide long-lasting and very efficient protection against infection. Our findings also indicate that other drugs, such as imipramine, desipramine, fluoxetine, sertraline, escitalopram, and maprotiline, with concentrations similar to that of amitriptyline 22 Our studies are strongly supported by a recently pre-printed study 30 that showed a marked beneficial effect of antidepressants on the clinical course of COVID-19. Hospitalized patients receiving antidepressants had a much better outcome, determined as decreased medical necessity of intubation or death, than patients receiving no antidepressants. Best results were obtained with venlafaxin, fluoxetine, escitalopram and mirtazipine, drugs that were also shown in the present study to inhibit acid sphingomyelinase and ceramide release upon pp-VSV-SARS-CoV-2 spike infection. It is important to note that the dose of antidepressants used in this J o u r n a l P r e -p r o o f 13 retrospective study was not optimized. Amitriptyline was used at a medium dose of 26 mg/day, which is a rather low dose. Amitriptyline is often used at these low dosages for sedation and not as an antidepressant, while other drugs such as fluoxetine and escitalopram are always used as antidepressants and applied at higher dosages. Further, fluoxetine also inhibits SARS-CoV-2 replication 31 , an effect that was not investigated in the present study, since the pseudoviral particles do not replicate. Amitriptyline and many other antidepressants are weak bases that are protonated in lysosomes and thereby trapped in lysosomes, but they also localize in acidic subcompartments of the cell membrane. Because of their physicochemical properties (for instance for amitriptyline: high lipophilicity and weak basicity; logP = 4.92, pKa = 9.4, www.drugbank.ca), these drugs accumulate in acidic intracellular compartments 32 . This accumulation leads to a high tissue concentration and accordingly to a high volume of distribution (for instance for amitriptyline: Vd = 16 L/kg, www.drugbank.ca). Thus, high tissue concentrations of these drugs can be detected in all organs. In fact, the highest uptake of amitriptyline among all tissues with a lung/blood concentration gradient of approximately 50 is found in the lungs of mice and rats 33, 34 . In humans, even higher concentration gradients have been found between lung and blood 35 . Because of the particularly high accumulation in the lung, the antiviral in vitro concentrations (up to 25 µM) that we found to be effective are likely to be achieved with conventional oral therapy of amitriptyline. We also expect a considerably longer elimination half-life of amitriptyline in lung tissue than in blood because of the high concentration of these drugs in deep compartments, such as lysosomes. The organic ring system of amitriptyline and other FIASMAs may bind to the lipid membrane, whereas the protonated tertiary amine displaces acid sphingomyelinase from lysosomal membranes or from the plasma membrane within acidic subdomains. Thus, these weak bases do not directly inhibit acid sphingomyelinase activity but rather functionally inhibit it. This mode of action also explains why antidepressants do not completely inhibit acid sphingomyelinase activity and do not induce severe adverse effects in clinical use. In addition, our studies indicate that the administration of anti-ceramide antibodies or neutral ceramidase also protects against SARS-CoV-2 infections. Anti-ceramide J o u r n a l P r e -p r o o f 14 antibodies and neutral ceramidase are not approved for clinical use, but could be easily tested for systemic or local adverse effects before a potential treatment of SARS-CoV-2 infections. Furthermore, both anti-ceramide antibodies and neutral ceramidase could potentially be administered via inhalation or as a nasal spray to prevent SARS-CoV-2 infections. The molecular mechanisms how ceramide mediates infection of epithelial cells with SARS-CoV-2 remain to be determined. Ceramide has been shown to directly activate several enzymes and to form large, highly hydrophobic ceramide-enriched membrane domains that serve to re-organize receptor and signalling molecules [10] [11] [12] [13] . It remains to be determined whether SARS-CoV-2 also induces these platforms and whether ceramide and/or these platforms are required for internalization or activation of TMPRSS2 and cathepsin L that mediate processing of spike require for fusion with the cell membrane 36 . High ceramide levels have been linked with hypertension and obesity 37, 38 The authors declare no competing financial interests. Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Erich Gulbins (erich.gulbins@uni-due.de). This study did not generate new unique reagents. The published article includes all datasets generated or analyzed during the study. Human nasal epithelial cells were obtained from 6 healthy volunteers, 4 women and 2 men. Their ages were 20, 21, 43, 48, 54, and 56. We did not observe any differences in infection between cells from women or men. Sample size was planned for the continuous variable difference in viral uptake and was based on two-sided We obtained nasal epithelial cells from these volunteers immediately before oral administration of 0.5 mg/kg amitriptyline (Neuraxpharm, Germany) and again 1. Caco-2 colon epithelial cells (ATCC HTB-37) were cultured in DMEM supplemented as described above. Pseudotyped viral particles based on a replication-deficient VSV that codes for eGFP and firefly luciferase instead of parental VSV-G (VSV*∆G-FLuc) 39 A further aliquot of the nasal epithelial cells was immediately shock frozen in liquid nitrogen after removal for determination of acid sphingomyelinase activity (see below). Nasal epithelial cells were also obtained from untreated volunteers as described above and were treated with 10 µM amitriptyline (Sigma; # A 8404) dissolved in 0.9% NaCl, anti-ceramide or control antibodies, neutral ceramidase (see below) or left untreated for 60 min ex vivo and then infected with pp-VSV-SARS-CoV-2 spike particles for 60 min. Cells were then washed and further incubated for 24 h to allow expression of eGFP. Infection was analyzed as above. Vero cells were grown on glass coverslips, treated with 10 or 25 µM amitriptyline as Ceramide amounts were determined by comparison with a standard curve using C16 to C24 ceramides as substrates. To Investigators were blinded to the identity of the samples in all microscopy experiments. All data are available upon request from the authors. Authors confirm that all data are included in the manuscript. A pneumonia outbreak associated with a new coronavirus of probable bat origin Clinical course and outcomes of critically ill patients with SARS-CoV-2 pneumonia in Wuhan, China: a single-centered, retrospective, observational study SARS-CoV-2 cell entry depends on ACE2 and TMPRSS2 and is blocked by a clinically proven protease inhibitor Cryo-EM structure of the 2019-nCoV spike in the prefusion conformation Structure of the SARS-CoV-2 spike receptor-binding domain bound to the ACE2 receptor Structural and functional basis of SARS-CoV-2 entry by using human ACE2 CD95 signaling via ceramide-rich membrane rafts Host defense against Pseudomonas aeruginosa requires ceramide-rich membrane rafts Raft ceramide in molecular medicine Compartmentalization of ceramide signaling: physical foundations and biological effects Observation of topical catalysis by sphingomyelinase coupled to microspheres Cathepsin D targeted by acid sphingomyelinase-derived ceramide Ceramide binds to the CaLB domain of cytosolic phospholipase A2 and facilitates its membrane docking and arachidonic acid release Ceramide-activated protein phosphatase: partial purification and relationship to protein phosphatase 2A PKC zeta is a molecular switch in signal transduction of TNF-alpha, bifunctionally regulated by ceramide and arachidonic acid Selective ceramide binding to protein kinase C-alpha and -delta isoenzymes in renal mesangial cells Targeting FLT3-ITD signaling mediates ceramide-dependent mitophagy and attenuates drug resistance in AML The tricyclic antidepressant desipramine causes proteolytic degradation of lysosomal sphingomyelinase in human fibroblasts Interactions of acid sphingomyelinase and lipid bilayers in the presence of the tricyclic antidepressant desipramine Identification of new functional inhibitors of acid sphingomyelinase using a structure-property-activity relation model Acid sphingomyelinase/ceramide system mediates effects of antidepressant drugs Antidepressants act by inducing autophagy controlled by sphingomyelin-ceramide Emerging mechanisms of drug-induced phospholipidosis Rhinoviruses infect human epithelial cells via ceramide-enriched membrane platforms Ebolavirus requires acid sphingomyelinase activity and plasma membrane sphingomyelin for infection SARS-CoV-2 receptor ACE2 is an Interferonstimulated gene in human airway epithelial cells and is detected in specific cell subsets across tissues Increase of heart rate and QTc by Amitripytline, but not by Venlafaxine, is correleated to serum concentration Association between SSRI antidepressant use and reduced risk of intubation or death in hospitalized patients with coronavirus disease 2019: a multicenter retrospective observational study The serotonin reuptake inhibitor Fluoxetine inhibits SARS-CoV-2 Quantitative modeling of selective lysosomal targeting for drug design Distribution and fate of C14-amitriptyline in mice and rats Postmortem release of amitriptyline from the lungs; a mechanism of postmortem drug redistribution Postmortem distribution of tramadol, amitriptyline, and their metabolites in a suicidal overdose Efficient activation of the severe acute respiratory syndrome coronavirus spike protein by the transmembrane protease TMPRSS2 Ceramide is upregulated and associated with mortality in patients with chronic heart failure The unexpected role of acid sphingomyelinase in cell death and the pathophysiology of common diseases A Vesicular Stomatitis Virus replicon-based bioassay for the rapid and sensitive determination of multispecies type I Interferon Mutations in the spike protein of Middle East respiratory syndrome coronavirus transmitted in Korea increase resistance to antibody-mediated neutralization