key: cord-0834579-fdrrbr6a
authors: Rajasekharan, Sreejith; Bonotto, Rafaela Milan; Kazungu, Yvette; Alves, Lais Nascimento; Poggianella, Monica; Orellana, Pamela Martinez; Skoko, Natasa; Polez, Sulena; Marcello, Alessandro
title: Repurposing of Miglustat to inhibit the coronavirus Severe Acquired Respiratory Syndrome SARS-CoV-2
date: 2020-05-18
journal: bioRxiv
DOI: 10.1101/2020.05.18.101691
sha: eb174c5bff21c6ca94b4c57d68efc42c9f10c2dd
doc_id: 834579
cord_uid: fdrrbr6a

Repurposing clinically available drugs to treat the new coronavirus disease COVID-19 is an urgent need in these early stages of the SARS-CoV-2 pandemic, when very few treatment options are available. The iminosugar Miglustat is a well-characterized drug for the treatment of rare genetic lysosome storage diseases such as Gaucher and Niemann-Pick type C, and has also been described to be active against a variety of enveloped viruses. The activity of Miglustat is here demonstrated for SARS-CoV-2 at concentrations achievable in the plasma by current clinical regimens without cytotoxicity. The drug acts at the post-entry level and leads to a marked decrease of viral proteins and release of infectious virus. The mechanism resides in the inhibitory activity towards α-glucosidases that are involved in early stages of glycoprotein N-linked oligosaccharide processing in the endoplasmic reticulum, leading to a marked decrease of the viral Spike protein. The wealth of available data on the clinical use of Miglustat for the treatment of lysosomal storage disorders and the antiviral properties against SARS-CoV-2 make it an ideal candidate for drug repurposing.

The novel Severe Acquired Respiratory Syndrome (SARS-CoV-2) coronavirus, the 2 etiologic agent of coronavirus disease 2019 , has now spread in everywhere 3 causing a global pandemic (1, 2). To date there have been more than 3.5 million reported 4 cases and 250.000 deaths worldwide urging for a global effort to tackle the disease (3). 5 SARS-CoV-2 belongs to the genus Betacoronavirus of the order/family/sub-family 6

Nidovirales/Coronaviridae/Coronavirinae. The virion is enveloped and contains a single 7 RNA genome of positive polarity. Morphologically, SARS-CoV-2 is about 120 nm in 8 diameter with large projections of the heavily glycosylated trimeric spike (S) proteins. 9

Other surface proteins include the membrane (M) and envelope (E) proteins, while inside 10 the envelope the helical nucleocapsid (N) wraps the viral RNA. The virus targets cells of 11 the upper and lower respiratory tract epithelia through the viral Spike that binds to the 12 angiotensin-converting enzyme 2 (ACE2) receptor, a process facilitated by the host type 2 13 transmembrane serine protease, TMPRSS2. Once inside the cell, viral polyproteins are 14 synthesized that encode for the replication machinery required to synthesize new RNA via 15 its RNA-dependent RNA polymerase activity. Replication is cytoplasmatic at the level of 16 the endoplasmic reticulum (ER) that is heavily rearranged. Structural proteins are then 17 synthesized leading to completion of assembly and release of viral particles (4, 5) . 18

Currently, no specific treatment against SARS-CoV-2 is available and the only antiviral 19 therapy comes from repurposing of drugs developed for other viral infections. Lopinavir 20 and ritonavir, remdesivir, (hydroxy)chloroquine, umifenovir and favipiravir are examples 21 that are currently being evaluated, but none have been conclusively shown to be effective 22 (6) . 23

The iminosugar Miglustat (Zavesca; N-butyl-1-deoxynojirimycin, NB-DNJ) inhibits α-24 glucosidases I and II that are involved in early stages of glycoprotein N-linked 25 oligosaccharide processing in the ER (7). Because most enveloped viruses require 26 glycosylation for surface protein folding and secretion, modulation of the oligosaccharide 27 to induce a reduction in infectivity became a strategy for treatment of the immune 28 deficiency virus type 1 (HIV-1), culminating in phase I/II clinical trials (8, 9) . The use of 29 iminosugars to misfold viral glycoprotein as a therapeutic approach has so far been 30 applied to several other viral infections including: hepatitis B and C viruses, Dengue and 31 other flaviviruses, and Ebola virus (10-12). An additional property of certain iminosugars is 32 the Glucosyltransferase inhibition activity, which is the basis for current therapy of rare 33 genetic lysosome storage diseases such as Gaucher and Niemann-Pick type C (13). This 34 activity of Miglustat could impact virus entry by modification of the plasma membrane. 35

Vero E6 cells (ATCC-1586) HEK 293T (ATCC CRL-3216), A549 (ATCC CCL-185), U2OS 4 (ATCC HTB-96) and human hepatocarcinoma Huh7 cells kindly provided by Ralf 5

Bartenschlager (University of Heidelberg, Germany) were cultured in Dulbecco's modified 6

Eagle's medium (DMEM) supplemented with 10% fetal bovine serum (FBS, Gibco) and 7

antibiotics. Cell cultures were maintained at 37 °C under 5% CO 2 . Cells were routinely 8 tested for mycoplasma contamination. 9

Working stocks of SARS-CoV-2 ICGEB-FVG_5 isolated in Trieste, Italy, were routinely 10 propagated and titrated on Vero E6 cells (14). Plaque assay was performed by incubating 11 dilutions of SARS-CoV-2 on Vero E6 monolayers at 37 °C for 1 hour, which were then 12 washed with phosphate buffered saline (PBS) and overlaid with DMEM 2% FBS 13 containing 1.5% carboxymethylcellulose for 3 days. Cells were then fixed with 3.7% 14 paraformaldehyde (PFA) and stained with crystal violet 1%. Cytotoxicity assay was 15 performed with Alamar Blue (ThermoFisher) according to manufacturer's instructions. 16 17

Miglustat (NB-DNJ) was purchased from Sigma (B8299). The drug was dissolved in 19

DMSO to obtain a stock solution, while following dilutions were made directly in growth 20

The SARS-CoV-2 Spike protein receptor-binding domain (RBD) was expressed from 22 pCAGGS using a construct generously provided by Florian Krammer (Mount Sinai, New 23 York) (15). Plasmid was transfected in 293T cells and cell extracts and supernatants were 24 harvested 24 hours post-transfection. Miglustat 200 µM was added after transfection and 25 maintained in the medium until the end of the experiment. 26

Sequence coding for the full length Spike protein was obtained from isolate Wuhan-Hu-1 27 (NCBI Reference Sequence: NC_045512.2). The nucleotide sequence, fused to an 28 immunoglobulin leader sequence (sec) at the N-terminus, codon optimized for expression 29 in mammalian cells, was obtained as synthetic DNA fragment from GenScript Biotech 30 (Netherlands) and cloned as HindIII/ApaI into a pCDNA3 vector. 31 32

Vero E6 cells were seeded at 6x10 4 cells/well density in a 48 wells' plate and incubated at 34 37 °C overnight. Cells were infected with 30 viral PFU/well, and incubated at 37 °C for 1 35 hour. Following incubation, the virus was removed and the wells washed with 1x PBS. The 1 infected cells were maintained with 800 µL of overlay medium containing 1.5% 2 carboxymethylcellulose (CMC) with DMEM + 2% heat-inactivated FBS, and Miglustat 3 dilutions. Cells were then incubated at 37 °C for 3 days. Finally, cells were fixed with 3.7% 4 PFA and stained with crystal violet. Plaques were counted and values were normalized to 5 vehicle (DMSO). The plaque reduction assays were conducted 6 in double replicates for three independent experiments. Inhibition was calculated with the 7 formula: (1-(average plaques Miglustat/average plaques Vehicle))*100 and plotted against 8 dilutions as antilog. For the cytotoxicity assay, fluorescence readings were normalized for 9 vehicle and percent plotted against dilutions. The half maximal effective concentration 10 (EC 50 ) and cytotoxic concentration (CC 50 ) were calculated using GraphPad Prism Version Healthcare) in binding buffer (20 mM sodium phosphate pH 7.0) and eluted with acetic 24 acid 50mM pH 2.7. Eluted antibody was immediately neutralized with 1M Tris pH 8, and 25 analyzed by RP-HPLC and by SDS PAGE to maintain the dimeric structure. Production 26 yield was 0.8 mg/ml. For immunofluorescence cells were fixed in 3.7% PFA, permeabilized 27 with 0.01% Triton and processed with mSIP-3022 as per standard procedure (19). Since 28 mSIP-3022 did not react with the denatured S protein a convalescent serum from a 29 COVID-19 patient was used for immunoblotting at a 1:200 dilution. Images were acquired 30 on a Zeiss LSM880 confocal microscope. For immunoblotting, whole-cell lysates were 31 resolved by 12% SDS-PAGE and blotted onto nitrocellulose membranes. The membranes 32

were blocked in 5% nonfat milk in Tris-buffered saline (TBS) plus 0.1% Tween 20 (TBST), 33 followed by incubation with the human serum diluted 1:200 in the same solution for 1 hour 34 at room temperature. After washing three times with TBST, secondary horseradish 35 peroxidase (HRP)-conjugated antibodies were incubated for 1 hour at room temperature. 1

The blots were developed using a chemioluminescent HRP substrate (Millipore). The anti-2 his antibody (monoclonal #8722 Sigma) was used at 1/2000 concentration for immunoblot. the asterisks above the graphs (**p < 0.01 highly significant; *p < 0.05 significant). Where 17 asterisks are missing the differences are calculated as non-significant (n.s). 18

19

The antiviral properties of Miglustat were assessed by performing a plaque assay. SARS-2

CoV-2 strain ICGEB-FVG_5 was used to infect Vero E6 for 1 hour. After removal of the 3 inoculum and wash in PBS, the cells were overlaid with medium containing 1.5% CMC and 4 dilutions of drug as indicated in Figure 1A . 72 hours post-infection cells were fixed and 5 stained to reveal plaques, which were visually counted. Data are from three independent 6 experiments each in duplicate biological replicates. In parallel, cytotoxicity was assessed 7 by the Alamar blue method at the indicated dilutions of drug. The effective concentration 8 for 50% inhibition (EC 50 ) of Miglustat was 41±22 µM with no apparent cyototoxicity until 1 9 mM (CC 50 > 1 mM). The plaque assay was performed in Vero E6 cells, which are standard 10 for the growth of SARS-CoV-2. However, further analysis would be better performed in 11 cells of human origin. To better characterize this hypothesis, a time-of-addition (TOA) experiment was 1 performed. Different conditions were used: pre-treatment, co-treatment and post-2 treatment. Huh7 cells were pre-treated with 200 µM Miglustat for 3h and then infected for 3 1h in the absence of drug (moi = 0.1). Afterwards, the virus was removed and the cells 4

were cultured in drug-free medium until the end of the experiment. For co-treatment, the 5 drug was added together with the virus during infection and then cells were maintained in 6 drug-free medium. For post-entry experiment, drug was added at 3 h post-infection and 7 maintained until the end of the experiment. As shown in Figure 2A , drug didn't affect viral 8 entry, neither was virucidal when administered concomitant with infection. Replication 9

(intracellular viral RNA) was slightly affected at 48hpi and significantly at 72hpi consistent 10 with the idea that Miglustat was effective at the post-entry level. This was reflected by the 11 reduction of intracellular nucleocapsid N protein observed at both time points ( Figure 2B) . detected by the conformation-dependent mSIP-3022 antibody. As shown in Figure 2J and 29 2K, Miglustat reduce the amount of protein on the cell surface. Host directed antiviral therapy is a strategy of inhibiting virus infection by targeting host 2 factors that are essential for viral replication (23). Currently there is a pressing need for 3 antiviral drugs for immediate use in the context of SARS-CoV-2 infection. Miglustat is a 4 drug that is in current clinical use for the treatment of certain genetic disorders and has 5 shown to be active against a variety of viral infections, making it a suitable candidate for 6 drug repurposing towards SARS-CoV-2. In this work the activity of Miglustat against 7 SARS-CoV-2 has been demonstrated in vitro with EC 50 in the micromolar range. The 8 standard dosage for lysosomal storage diseases such as Gaucher or Niemann Pick is 100 9 mg/3 times a day, with a maximum daily dose of 600 mg/day. A single dose of 100 mg 10

Miglustat reaches a peak in plasma concentration of around 3-5 µM within 4 hours, while 11 half-life is approximately 8 hours. If this dose is administered every 4 hours/6 times per 12 day, the plasma concentration of Miglustat would stabilize around 10 µM (24). If 200 mg 13

Miglustat is administered every 8 hours/3 times a day, the plasma concentration could be 14 also higher than 10 µM in 24 hours. However, increased dosage could lead to well-15 described adverse reactions that include tremors, diarrhea, numbness and 16

Miglustat has been shown to act through two different mechanisms: at the level of virus 18 entry, by perturbing the plasma membrane, and at the level of folding and secretion of 19 virion proteins, by affecting essential glycosylation steps in the ER. The latter mechanism 20 is supported by the data presented in this work, where a strong antiviral activity is detected 21 only when the drug is added post-infection. Correct folding and secretion of glycoproteins 22 is a process tightly controlled in the ER by chaperons such as Calnexin that recognize 23 specific glycosylation intermediates (25). Miglustat interferes with this process resulting in 24 the accumulation of misfolded proteins and a defect in secretion. Consistently, the Spike 25 protein of SARS-CoV-1 has been shown to bind Calnexin and disruption of this function 26 caused decrease of virus infectivity (22). 27

In conclusion this work provides in vitro evidence for the use of Miglustat as inhibitor of 28 SARS-CoV-2 and proposes its use in clinical trials for COVID-19 patients. monolayers infected with SARS-CoV-2. Following incubation for three days cells were 5 fixed and stained to count viral plaques against vehicle control, which were plotted as 6 percent inhibitory activity (black dots). Cytotoxicity was measured by the Alamar blue 7 method and data plotted as percent viability (red squares). 

A pneumonia outbreak associated with a new 6 coronavirus of probable bat origin

11 3. WHO. 2020. World Health Organization -Coronavirus disease 2019 (COVID-19). 12 Situation Report -106

Insight into 2019 novel coronavirus -An updated interim 15 review and lessons from SARS-CoV and MERS-CoV

The trinity of COVID-19: 17 immunity, inflammation and intervention

Pharmacologic 20 Treatments for Coronavirus Disease 2019 (COVID-19): A Review

Glycosidase inhibitors: inhibitors of N-linked oligosaccharide 23 processing

The safety and efficacy 26 of combination N-butyl-deoxynojirimycin (SC-48334) and zidovudine in patients with 27 HIV-1 infection and 200-500 CD4 cells/mm3

The tolerability and pharmacokinetics 31 of N-butyl-deoxynojirimycin in patients with advanced HIV disease (ACTG 100). The 32 AIDS Clinical Trials Group (ACTG) of the National Institute of Allergy and Infectious 33 Diseases

Targeting glycosylation as a 35 therapeutic approach

Small molecule inhibitors of ER alpha-glucosidases are 39 active against multiple hemorrhagic fever viruses

Antiviral effects of 41 an iminosugar derivative on flavivirus infections

Lysosomal storage 43 diseases

Isolation and full-length genome characterization of 46 SARS-CoV-2 from COVID-19 cases in Northern Italy

SARS-CoV-2 Seroconversion in

Human monoclonal antibody 5 combination against SARS coronavirus: synergy and coverage of escape mutants

Potent binding of 2019 novel coronavirus spike protein by a SARS 9 coronavirus-specific human monoclonal antibody

New tags for recombinant 12 protein detection and O-glycosylation reporters

Viral priming of cell intrinsic innate antiviral signaling by the unfolded protein 15 response

Multitarget synthetic RNA for SARS-Cov-2 17 detection and quantification

Identification of N-linked carbohydrates from severe acute 20 respiratory syndrome (SARS) spike glycoprotein

Monitoring of S protein maturation in the 23 endoplasmic reticulum by calnexin is important for the infectivity of severe acute 24 respiratory syndrome coronavirus

Influence of food intake on the 28 pharmacokinetics of miglustat, an inhibitor of glucosylceramide synthase

Intracellular functions of N-linked glycans