key: cord-0925560-jmifqi27
authors: Juang, Yu-Pu; Chou, Yu-Ting; Lin, Ru-Xian; Ma, Hsiu-Hua; Chao, Tai-Ling; Jan, Jia-Tsrong; Chang, Sui-Yuan; Liang, Pi-Hui
title: Design, synthesis and biological evaluation of niclosamide analogs against SARS-CoV-2
date: 2022-03-19
journal: Eur J Med Chem
DOI: 10.1016/j.ejmech.2022.114295
sha: 894a06152095f89fe7f176e4e9d13421eda01b27
doc_id: 925560
cord_uid: jmifqi27

Niclosamide, a widely-used anthelmintic drug, inhibits SARS-CoV-2 virus entry through TMEM16F inhibition and replication through autophagy induction, but the relatively high cytotoxicity and poor oral bioavailability limited its application. We synthesized 22 niclosamide analogs of which compound 5 was found to exhibit the best anti-SARS-CoV-2 efficacy (IC(50) = 0.057 [Formula: see text] M) and compounds 6, 10, and 11 (IC(50) = 0.39, 0.38, and 0.49 [Formula: see text] M, respectively) showed comparable efficacy to niclosamide. On the other hand, compounds 5, 6, 11 contained higher stability in human plasma and liver S9 enzymes assay than niclosamide, which could improve bioavailability and half-life when administered orally. Fluorescence microscopy revealed that compound 5 exhibited better activity in the reduction of phosphatidylserine externalization compared to niclosamide, which was related to TMEM16F inhibition. The AI-predicted protein structure of human TMEM16F protein was applied for molecular docking, revealing that 4′-NO(2) of 5 formed hydrogen bonding with Arg809, which was blocked by 2′-Cl in the case of niclosamide.

The global COVID-19 pandemic caused by SARS-CoV-2 has been an ongoing global economic and health burden since 2019. Different from the previous outbreaks related to other two coronavirus, SARS-CoV and Middle East respiratory syndrome coronavirus (MERS-CoV), SARS-CoV-2 exhibits higher transmission rate and lower fatality rate, resulting in a prolonged pandemic situation, extraordinary social and economic burdens and a big threat to public health [1] .

With extensive efforts from pharmaceutical industry around the world, effective SARS-CoV-2 vaccines were produced and made a great success in controlling the global pandemic [2, 3] . Currently available treatments for COVID-19 include J o u r n a l P r e -p r o o f Niclosamide (Fig. 1B) is an oral anthelmintic drug which has been approved by FDA for the treatment of tapeworm infection in humans for several decades. Recently, multifactorial roles of niclosamide in human disease have been discussed, and its broad antiviral activity was reported [9] [10] [11] . The anti-SARS-CoV-2 activity of niclosamide was identified in several drug-repurposing studies, and evaluated in clinical trials [12, J o u r n a l P r e -p r o o f 13] . Cellular experiments revealed that the inhibition mechanism is related to TMEM16F-mediated viral syncytia and SKP2 inhibition [14, 15] . These results suggest that niclosamide is a promising lead for structure-activity relationship study to identify novel structures against SARS-CoV-2 infection. Although structure derivatization of niclosamide have been applied for the inhibition of SARS-CoV-2, the ratio between cytotoxicity (CC50) and antiviral efficacy (IC50) were unable to exceed 5, indicating that the compound selectivity is beyond satisfying [16, 17] .

In this study, we focused on the structure-activity relationships of niclosamide core, which was divided into benzoic acid and aniline moiety. The strategy for modification included aromatic substitution with electron donating and withdrawing group, chain elongation, and glycosylation (Fig. 1B) . The antiviral efficacy of compound was primarily evaluated with cytopathic effect and antibody quantification in Vero E6 cells.

The membrane penetration activity and cytotoxicity of compounds were obtained by PAMPA, clogP calculation and the MTT assay to identify the compound with ideal biological activity. The in vitro human plasma and human liver S9 assay were applied to identify compounds with better stability. The TMEM16F inhibition was confirmed by the phosphatidylserine externalization assay using fluorescence microscopy, and the computer simulation revealed the essential residue for the TMEM16F protein binding.

Niclosamide analogs were synthesized through a two-step amide coupling reaction Similarly, compounds 18-22 were synthesized to evaluate the effect of substitution on the benzoic acid moiety. Benzoic acids with hydroxyl and methyl substitution were treated with thionyl chloride to give corresponding acyl chloride, followed by the addition of 2-chloro-4-nitroaniline in dichloromethane to afford niclosamide analogs

To increase water solubility of the niclosamide analogs and hence alter their cellular distribution, analog 14 was glycosylated with glucose (Scheme 2). Boron 

The antiviral efficacy of these niclosamide analogs was first evaluated by microscopic visualization of their cytopathic effect (CPE), as previously described [18] .

Vero E6 cells, which are widely used in coronavirus assays due to their high expression levels of angiotensin-converting enzyme 2 (ACE2) receptor [19] , were infected with SARS-CoV-2 virus strain obtained from Taiwan Centers for Disease Control (hCoV19/Taiwan/4/2020). Of the 23 niclosamide analogs tested, nine (2, 3, 4, 5, 6, 10, 11, 12, 18) exhibited levels of antiviral efficacy that were superior to their cytotoxicity J o u r n a l P r e -p r o o f ( Fig. 2) , and were selected for determination of the concentration of each to inhibit half virus replication (IC50). Since niclosamide has been reported to inhibit the entry of viruses into cells, the pretreat experiment was applied for the determination of antiviral IC50 in Vero E6 cells [14] . Niclosamide and 9 analogs were separately added to the culture medium and incubated with cells. After 2 h, 100 TCID50 virus was added to the medium, incubated for 24 h, and the virus titer was quantified with antibodies targeting viral nucleocapsid proteins (Fig. 3 ). Under these conditions, the IC50 of niclosamide was found to be 0.4 μM, which is comparable with the literature (0.34 μM) [14] .

J o u r n a l P r e -p r o o f 

Niclosamide is well tolerated when administered orally to patients infected with tapeworm, but its high cytotoxicity, low solubility, and low oral bioavailability (F = 10%) limited its usage [9] . Accordingly, we investigated the clogP, PAMPA permeability, and cytotoxicity against Vero E6 cells of the analogs synthesized (Table   1 , Table 2 , Fig. S1 ). clogP values were calculated in silico using Molecular Operating Environment (MOE) software and, with the exception of compound 12, were all found to be within the range of 0 to 5, which is preferred for oral absorption (Fig. S1 ). Permeability in the gastrointestinal (GI) tract was evaluated using the parallel artificial membrane permeability assay (PAMPA). Verapamil (114.8 x 10 -6 cm/s) and ranitidine (1.6 x 10 -6 cm/s) were used as positive and negative control, respectively. Niclosamide exhibited a PAMPA permeability of 8.5 x 10 -6 cm/s ( Table 2 ). The permeability of three compounds 15, 16, and 19 was 5-fold higher, but none of them exhibited antiviral efficacy. The PAMPA permeabilities of all the compounds exhibiting anti-SARS-CoV-2 activity were similar to niclosamide suggesting that a prodrug strategy, such as J o u r n a l P r e -p r o o f The cytotoxicity of each niclosamide analog was firstly measured in conjunction with its antiviral efficacy over a 24h period, and most compounds exhibited an IC50 higher than 100 μM (Fig. 3) . To elucidate the difference in the cytotoxicity of all the analogs, the inhibition assay of Vero E6 proliferation was prolonged to 48 h ( Table 1) .

As a result, niclosamide exhibited the highest cytotoxicity against Vero E6 (IC50 = 1.03 μ M ), and the selectivity between antiviral efficacy and cytotoxicity was 2.6. Two compounds, 5 and 11, exhibited selectivity greater than ten folds. Compound 5, bearing 3′-Cl and 4′-NO2 substituents on the aniline moiety, exhibited the best antiviral activity (IC50 = 0.057 μM) and slightly less cytotoxic (IC50 = 1.51 μM) than niclosamide and improved the selectivity ratio to 26.5. Compound 11 was less cytotoxic than niclosamide as 6.23 μ M , improving the selectivity to 12.7, however, low PAMPA permeability resulting from trifluoromethyl group might diminish the GI absorption.

J o u r n a l P r e -p r o o f applied to broader use [20] . Therefore, the stability of compound was evaluated in human plasma and liver S9 enzymes to search for more stable analogs and the water solubility was also evaluated (Fig. 4A ). In human plasma, compounds 5, 6, 10, and 11

contained 78%, 71%, 77%, and 58% after 48 h under 37 ℃ while niclosamide (1) was only 40% remain (Fig. 4B ). On the other hand, liver metabolism plays an important role in orally administered drug, and compounds 5, 6, and 11 showed slightly better stability than niclosamide (1) (41%, 33%, 47% versus 27%) in liver S9 stability assay ( Fig. 4C ). The results showed that 5, 6, 11 were better anti-SARS-CoV-2 agents than niclosamide for oral delivery. 

Niclosamide is a TMEM16F protein inhibitor able to prevent virus entry by reducing scramblase activity of TMEM16F, which can be quantified by staining the externalized phosphatidylserine with a fluorescence probe [14] . Subjection of either niclosamide or 5 for 1h significantly reduced the signal of phosphatidylserine on cell J o u r n a l P r e -p r o o f surface triggered by ionomycin. On the other hand, nelfinavir, a SARS-CoV-2 inhibitor targeting viral protease [6] , exhibited no inhibition toward phosphatidylserine externalization (Fig. 5 ). The results indicated that compound 5 was an inhibitor of TMEM16F scramblase activity in a similar way of niclosamide. To elucidate the TMEM16F inhibition mechanism of 5, the computer simulation was performed. Since the crystal structure of human TMEM16F (ANO6) was not available, predicted structure was obtained from AlphaFold protein structure database [21] , and the potential ligand binding pockets of TMEM16F were identified using the PrankWeb ( Fig. 6A ) [22] . The TMEM16K protein, an analog protein of TMEM16F, was chose for the validation of AlphaFold-predicted protein structure. The predicted (Uniprot: Q9NW15) and crystal (PDB: 5OC9) structure of human TMEM16K protein were superimposed and showed high similarity in quaternary protein structure, indicating the prediction was reliable (Fig. S4A ). To validate the pocket identification process, crystal structure of TMEM16F from mus musculus (PDB: 6P46), which shared 90% similarity in protein sequence to human's, was subjected to PrankWeb and the topranking pocket was at the same place to the predicted human TMEM16F protein ( Fig.   S5 ). Moreover, the top-ranking pocket, existed at extracellular domain, was close to the previously reported binding pocket of TMEM16A, which is also a TMEM16F analog [23] . Altogether, the binding pocket of predicted human TMEM16F protein was used for docking of niclosamide and compound 5.

J o u r n a l P r e -p r o o f interaction on the other side, including Asp596, Cys600, and Arg809 (Fig. 6B) .

Nevertheless, the alteration of 2′-Cl to 3′-Cl provided the exposure of chloride atom to the Lys575 and Gly574, enhancing the affinity of 5 to the pocket by the interaction of 4′-NO2 and carbonyl group to Arg809 (Fig. 6C ). The computer simulation provided a preliminary understanding of the TMEM16F inhibition, which was a potential target for virus entry inhibition.

In this study, 22 niclosamide analogs were synthesized and evaluated for the anti-SARS-CoV-2 activity. Compound 5 was found to exhibit the best anti-SARS-CoV-2 activity (IC50 = 0.057 μ M ), which was 7-fold higher than niclosamide, and the selectivity of cytotoxicity versus antiviral potency was improved from 2.6 (niclosamide)

to 26.5. The GI absorption efficacy of the analogs was evaluated by PAMPA assay, however, the effective compounds (2-6, 10-12, 18) contained low permeability in the range of 4 to 20 (x 10 -6 cm/s). Human plasma and liver S9 enzyme assay revealed that compounds 5, 6, and 11 were more stable than niclosamide. The inhibition activity toward TMEM16F was confirmed by phosphatidylserine externalization experiment through fluorescence microscopy, and 5 exhibited the same inhibition pattern to niclosamide. The extracellular binding pocket of AI-predicted human TMEM16F protein was revealed by computer simulation and docking results showed that Arg809 was an important residue for drug-protein interaction. Accordingly, compound 5, 6, 11

are proposed as a better anti-SARS-CoV-2 agents than niclosamide, which are worth for further study in the future.

J o u r n a l P r e -p r o o f

All reagents and solvents were reagent grade and used without further purification, ( Glucosyl imidate 23 was synthesized as previously reported [24] . 

Vero E6 cells were seeded into 96-well plates (1 × 10 4 per well) and incubated for 

To access water solubility of compound 1, 5, 6, 10, 11, niclosamide analogs were added to water continuously until the solid remain in the solution after sonication for 5 mins. After centrifugation at 3000 rpm for 20 min, supernatants were collected and 

To access human plasma stability, niclosamide analogs were dissolved in DMSO 

The predicted structure of human TMEM16F and TMEM16K proteins were obtained from Alphafold protein structure database (http://alphafold.ebi.ac.uk) searching by UniProt number Q4KMQ2 and Q9NW15 [21] . Docking was performed on the predicted binding pocket of TMEM16F, and 30 poses for London dG and 10 poses of GBVI/WSA dG were used for final docking. All the results were ordered by the binding energy using S Score function, and the interactions of compound in the pocket were visualized by ligand interaction function in MOE.

All data were obtained at least in triplicate, and results are reported as mean ± mean of standard deviation (S.E.M.). Comparisons among groups were analyzed via student t tests, one-way ANOVA, and two-way ANOVA analysis using GraphPad Prism J o u r n a l P r e -p r o o f 8. The statistical significance was determined: n.s., nonsignificant difference; ***P < 0.001; *P < 0.05.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Compound 21 (35 mg, 15% yield), a white solid, was synthesized according to the general procedure using 4-methylsalicylic acid (120 mg, 0.787 mmol) and 2-chloro-4-nitroaniline (272 mg, 1.58 mmol)

DMSO-d6) δ 11.84 (s, 1H, NH), 8.87 (d, J = 9.3 Hz, 1H, H-6′)

(m, 2H), 2.30 (s, 3H, CH3) ppm; 13 C NMR (100 MHz, DMSO-d6) δ 169

HRMS (ESI-TOF) calcd. for C14H10ClN2O4-H [M-H] -: 305.0335, found: 305.0336; HPLC purity 95.9% (tR: 9.6 min, Hypersil BDS C18

Compound 22 (22 mg, 10% yield), a pale yellow solid, was synthesized according to the general procedure using 5-fluorolsalicylic acid (120 mg, 0.769 mmol) and 2-chloro-4-nitroaniline (265 mg, 1.54 mmol), and was recrystallized from DCM

118.8 (d, J = 6.8 Hz), 116.1 (d, J = 24.3 Hz) ppm

The suspension for the S9 assay was prepared as follows: 2.2 mg/mL of human liver S9 fraction (20 donor pool, mixed gender

After the suspension was preincubated at 37 °C for 10 min, the niclosamide analogs (1, 5, 6, 10, 11, 0.5 mM in 10 μL PBS) were added to the suspension and incubated for 0, 10, 30, 60, 120, and 360 min, after which experiments were halted by dilution with 200 μL of ACN. Followed by vortexing and centrifugation at 13000 rpm for 20 min, supernatants were collected and filtered through 0.22 μm PTFE membranes into 12 × 32 mm vials. An aliquot of 10 μL of the sample solution was injected into the HPLC each time and

Phosphatidylserine externalization assay The assay was performed as previously reported method with slight modification

TPP) at a density of 8,000 cells per well for 2 h, and the cells were treated with compounds for 1 h. After one wash with FBS-free medium, cells were incubated with or without 10 μM ionomycin for 30 s, and the medium was removed. Cells were incubated with 100 μL 1:100 annexin XII (ab129817, Abcam), and imaged by Olympus inverted microscope (CKX41) with reflected fluorescence system (U-RFLT50)

The COVID-19 Pandemic: A Comprehensive Review of Taxonomy, Genetics, Epidemiology, Diagnosis, Treatment, and Control

Efficacy and Safety of the mRNA-1273 SARS-CoV-2 Vaccine

Safety and Efficacy of the BNT162b2 mRNA Covid-19 Vaccine

COVID-19 Treatment Guidelines Panel. Coronavirus Disease

Remdesivir in adults with severe COVID-19: a randomised, doubleblind, placebo-controlled, multicentre trial

Identification of existing pharmaceuticals and herbal medicines as inhibitors of SARS-CoV-2 infection

Molnupiravir for Oral Treatment of Covid-19 in Nonhospitalized Patients

Covid-19: Pfizer's paxlovid is 89% effective in patients at risk of serious illness, company reports

Niclosamide: Beyond an antihelminthic drug

Broad Spectrum Antiviral Agent Niclosamide and Its Therapeutic Potential

hydroxybenzamide Analogues as Potent Human Adenovirus Inhibitors

A randomized, double-blind, placebo-controlled phase 1 trial of inhaled and intranasal niclosamide: A broad spectrum antiviral candidate for treatment of COVID-19

Identification of Antiviral Drug Candidates against SARS-CoV-2 from FDA-Approved Drugs

Drugs that inhibit TMEM16 proteins block SARS-CoV-2 spike-induced syncytia

SARS-CoV-2-mediated dysregulation of metabolism and autophagy uncovers host-targeting antivirals

Salicylanilides Reduce SARS-CoV-2 Replication and Suppress Induction of Inflammatory Cytokines in a Rodent Model

Application of niclosamide and analogs as small molecule inhibitors of Zika virus and SARS-CoV-2 infection

Small molecules targeting severe acute respiratory syndrome human coronavirus

Cell Entry Depends on ACE2 and TMPRSS2 and Is Blocked by a Clinically Proven Protease Inhibitor

Contributions of Hepatic and Intestinal Metabolism to the Disposition of Niclosamide, a Repurposed Drug with Poor Bioavailability

Highly accurate protein structure prediction with AlphaFold

PrankWeb: a web server for ligand binding site prediction and visualization

Arctigenin, a novel TMEM16A inhibitor for lung adenocarcinoma therapy

Synthesis of Oxaphosphinane-Based Pseudodisaccharides