key: cord-0964553-f7yeix7y
authors: Desantis, Jenny; Mercorelli, Beatrice; Celegato, Marta; Croci, Federico; Bazzacco, Alessandro; Baroni, Massimo; Siragusa, Lydia; Cruciani, Gabriele; Loregian, Arianna; Goracci, Laura
title: Indomethacin-based PROTACs as pan-coronavirus antiviral agents
date: 2021-09-04
journal: Eur J Med Chem
DOI: 10.1016/j.ejmech.2021.113814
sha: 1a44835c4fa56e3ce29de80be126f5417c0d96ac
doc_id: 964553
cord_uid: f7yeix7y

Indomethacin (INM), a well-known non-steroidal anti-inflammatory drug, has recently gained attention for its antiviral activity demonstrated in drug repurposing studies against severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2). Although the mechanism of action of INM is not yet fully understood, recent studies have indicated that it acts at an early stage of the coronaviruses (CoVs) replication cycle. In addition, a proteomic study reported that the anti-SARS-CoV-2 activity of INM could be also ascribed to its ability to inhibit human prostaglandin E synthase type 2 (PGES-2), a host protein which interacts with the SARS-CoV-2 NSP7 protein. Although INM does not potently inhibit SARS-CoV-2 replication in infected Vero E6 cells, here we have explored for the first time the application of the Proteolysis Targeting Chimeras (PROTACs) technology in order to develop more potent INM-derived PROTACs with anti-CoV activity. In this study, we report the design, synthesis, and biological evaluation of a series of INM-based PROTACs endowed with antiviral activity against a panel of human CoVs, including different SARS-CoV-2 strains. Two PROTACs showed a strong improvement in antiviral potency compared to INM. Molecular modelling studies support human PGES-2 as a potential target of INM-based antiviral PROTACs, thus paving the way toward the development of host-directed anti-CoVs strategies. To the best of our knowledge, these PROTACs represent the first-in-class INM-based PROTACs with antiviral activity and also the first example of the application of PROTACs to develop pan-coronavirus agents.

Since its identification in patients with severe pneumonia in Wuhan city, Hubei [1, 2] the 2019 novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has rapidly spread around the world causing the large-scale COVID-19 pandemic. As of 21 May 2021, SARS-CoV-2 infection has caused more than 165 million of confirmed positive cases, many hospitalizations, and more than 3.4 million deaths. [3] SARS-CoV-2 is an enveloped, positive-sense, single-stranded RNA virus characterized by the largest viral RNA genome found to date. [4] [5] [6] It belongs to the Coronavirinae subfamily (Coronaviridae family), which comprises four genera, i.e., α-, β-, γ-, and δ-coronavirus. Among the coronaviruses known to infect humans, the β-coronavirus genus includes SARS-CoV and SARS-CoV-2, MERS-CoV, HCoV-OC43 and HCoV-HKU1, while HCoV-229E and HCoV-NL63 belong to the α-coronavirus genus. [7] Currently, no specific antiviral drugs have been approved against SARS-CoV-2 and other human pathogenic coronaviruses, with the exception of remdesivir, which obtained an emergency use authorization in May 2020 for the treatment of patients with severe COVID-19 symptoms[8, 9] , but later in October 2020 WHO issued a recommendation against the use of remdesivir, given the controversial benefits in COVID-19 patients. [10] Given the severity of the SARS-CoV-2 outbreak, an intensive research effort from pharma companies, academic research labs, and other organizations is globally ongoing with the aim of identifying specific anti-coronavirus drugs. [11] [12] [13] [14] Until now, drug repurposing represents one of the most pursued strategies elected so far by the scientific community to identify effective drugs for combating the COVID-19 pandemic. Indeed, this strategy can rapidly lead to the identification of existing clinically approved drugs able to prevent, control, or eradicate the SARS-CoV-2 infection, thus representing possible therapeutics for . [13, [15] [16] [17] J o u r n a l P r e -p r o o f In this context, indomethacin (INM) (Figure 1a) , a non-steroidal anti-inflammatory drug (NSAID) with anti-inflammatory, analgesic, and antipyretic properties, has gained great attention from different research groups as potential treatment or adjunct for SARS-CoV-2/COVID-19. [18] [19] [20] [21] The pharmacological effect of INM as an NSAID drug is not yet fully understood, but it is commonly related to a potent and non-selective inhibition of the cyclooxygenase (COX)-1 and 2 enzymes, or prostaglandin G/H synthase. [22, 23] However, INM is also known to inhibit phospholipase A2 (PLA-2), [24] and microsomal prostaglandin E synthase type 2 (mPGES-2), [23] all involved in eicosanoid biosynthesis. Concerning its antiviral effect against coronaviruses (CoVs), INM was previously shown to possess the ability to inhibit the replication of canine coronavirus (CCoV) and SARS-CoV. [25, 26] [20] INM was also evaluated in vitro against SARS-CoV-2 alone or in combination with ketotifene. [28] Although in a 4-day cytopathic effect protection assay INM did not exhibit any antiviral activity (EC50 >400 µM, Vero E6 cells), its evaluation in a virus yield reduction assay under reduced rounds of virus replication (48 h) showed an inhibition of SARS-CoV-2 replication with an EC50 of 100.1 µM (in infected Vero E6 cells). [28] The mechanism of antiviral action of INM remains unknown, but different hypotheses have been postulated so far. In particular, results reported by Amici et al. indicate that INM does not affect J o u r n a l P r e -p r o o f virus infectivity, binding or entry into host cells, but acts at an early stage of coronaviruses replication cycle, causing a global repression of viral protein synthesis via a ds-RNA-dependent protein kinase R (PKR)-mediated pathway. [26] On the other hand, a recently published proteomic/chemoinformatic study on the SARS-CoV-2 interactome with human host proteins suggested that the anti-SARS-CoV-2 activity of INM could be ascribed to its ability to inhibit human prostaglandin E synthase type 2 (PGES-2, encoded by PTGES2). [21, 29, 30] In fact, INM has been previously reported as PGES-2 inhibitor in the low nanomolar range [25, 31] and PGES-2 has been found to interact with the SARS-CoV-2 NSP7 protein, [21, 29] which, along with NSP8, is a non-structural viral protein composing the viral primase complex that is part of the viral RNA polymerase machinery. Interestingly, PGES-2 resulted as a host proviral factor and the interaction between NSP7 and PGES-2 was found to be conserved across the three highly pathogenic CoVs SARS-CoV, SARS-CoV-2, and MERS-CoV, [29, 30] suggesting that PGES-2 may be a potential pan-CoV antiviral target, i.e., a target shared by all pathogenic CoVs that could be exploited for the development of universal anti-CoV strategies (from the Greek πᾶν, pan, meaning "involving all members" of a group).

Gordon and co-workers also found that COVID-19 patients treated with INM were less likely to require hospitalization than those treated with other anti-inflammatory drugs that do not target PGES-2, providing a connection between proteomics study and clinical data. [29] However, it is worth noting that the anti-SARS-CoV-2 potency of INM is limited in infected cells, being in the range of ~100 µM (this study and ref [28] ). Taking this into account, we hypothesized that the exploitation of INM for PROTAC design could result in an enhancement of antiviral activity, since the degradation of the target PGES-2 could be obtained in addition to enzymatic inhibition and potentiate the antiviral effect.

In the past few years, PROteolysis TArgeting Chimeras (PROTACs) have become an exciting new paradigm to target proteins by promoting and achieving target proteins degradation via the ubiquitin-proteasome system (UPS). [32] [33] [34] [35] [36] [37] PROTACs are hetero-bifunctional molecules J o u r n a l P r e -p r o o f composed of a ligand for a protein of interest (POI) and an E3 ligase binder connected through a linker. [38] They act by bringing the E3 ligase into close proximity of the POI, thus inducing its ubiquitination and subsequent UPS-dependent degradation. [32, 37] Unlike traditional inhibitors, the removal of all the POI functions at once through POI degradation can achieve a more profound pharmacological effect than that obtained by only inhibiting a functional binding site. [37] Moreover, PROTACs do not require long-term and high-affinity binding to the POI, since a potent target degradation can be even obtained with modest-affinity POI ligands if a stable ternary complex is formed, [37, 39, 40 ] thus permitting to reduce or even abrogate the development of drug resistance. [32] Despite the increasing interest in PROTAC technology, their application is still marginal in the field of antivirals, with only one well-characterized application reported against the hepatitis C virus NS3/4A protease. [41] This The present study not only provided the proof-ofconcept that targeted protein degradation may represent a valuable approach for antiviral design, but it also demonstrated that PROTACs are less prone to select viral mutants with impaired ligand binding and could thus be exploited in the case of viral resistance to conventional inhibitors. [41] So far, only one paper suggesting the possible use of PROTACs against SARS-CoV-2 has been published, [42] but chemical structures were not disclosed.

In the present work, based on the experimental and clinical evidence accumulated to date, which suggests a potential therapeutic use of INM against SARS-CoV-2 infection, we decided to exploit INM to design PROTAC derivatives (Figure 1 ) in order to investigate whether this emerging new technology could represent a valid approach also in the search for anti-coronavirus agents. PROTACs. The IC50 value represents the compound concentration that inhibits 50% of viral replication. The CC50 value represents the compound concentration that inhibits 50% of cell viability.

Considering the INM carboxylic acid group as a suitable and easy site for linker attachment to exploit for amidation reaction, also due to the exposition of this group to the solvent ( Figure S1 ), herein we designed and synthesized four INM-based PROTACs (2-5, Table 1 

The antiviral activity of the synthesized PROTAC compounds, along with INM, was first evaluated by plaque reduction assays (PRA) in Vero E6 cells infected with SARS-CoV-2/NL/2020 ( Table 1) .

The viral RNA polymerase inhibitor remdesivir (RMV) was included as a positive control of inhibition. [45] As reported in Table 1 Table 1 ). In contrast, compounds 2 and 4 did not show dose-dependent inhibitory activity against any of the tested SARS-CoV-2 strains up to 50 µM ( Table 1) . To exclude that the observed antiviral activity could be due to toxic effects in the target cells, the cytotoxicity of the compounds was tested in parallel by MTT assays in Vero E6 cells. None of the compounds showed cytotoxicity up to a concentration of 200-250 µM ( Table 1) . Interestingly, as reported in Table 2 , both compound 3 and 5 not only inhibited SARS-CoV-2

replication, but they also showed antiviral activity against another β-coronavirus, i.e., HCoV-OC43 (EC50 = 4.7 and 2.5 µM, respectively). Moreover, we observed antiviral activity for 3 and 5 also against the α-coronavirus HCoV-229E, although compound 3 resulted less active against HCoV-229E compared to HCoV-OC43 ( Table 2) . Importantly, both compounds 3 and 5 did not show significant cytotoxicity also in MRC-5 cells, thus excluding that the possibility the antiviral activity of these PROTAC derivatives of INM might be due to cytotoxic effects. Thus, both compounds 3 and 5 exhibited specific antiviral activity against pandemic and epidemic CoVs belonging to different genera of Coronaviridae family. To note, the fact that INM-derived PROTACs 3 and 5 resulted more active against HCoVs compared to SARS-CoV-2 could be ascribed to the different cell lines used in the experiments. In fact, for the HCoVs experiments, a human fibroblast cell line J o u r n a l P r e -p r o o f (MRC-5) was used and possible differences in E3 ligase VHL expression with respect to that in the simian Vero E6 cell line used for SARS-CoV-2 experiments could exist. Alternatively, also differences in the expression and/or activity of the PGES-2 target in the two cell lines cannot be excluded. Further studies will be aimed at investigating these aspects. 

Although the limited number of compounds investigated in this study does not allow determination of a structure-activity relationship, which will be a topic for future studies, we were intrigued by the significant increased activity of compound 3 compared to 2, as they only differ by two methylene units. We hypothesized that the linker in compound 2 might be too short to allow the ternary (Figure 2-a,b) , and VHL ligand moieties were superimposed onto the x-ray VHL ligand pose.

Thus, the linker moiety was allowed to freely move and poses were inspected. Figure 2 -a,b clearly shows that the VHL ligand-protein interaction occurs in a very superficial region, and therefore both linkers are not buried but potentially available for complex generation. On the contrary, when the overall procedure was repeated by building the INM-linker moiety and removing the VHL ligand (Figure 2-c,f) , the linker in compound 2 resulted as buried into the target protein (the human mPGES-2 here as a homology model), and thus less prone to form the ternary complex. In the case of compound 3, the linker resulted as more able to expose its terminal group out of the target cavity, although by a limited amount. Compounds 4 and 5, whose linkers have a length that is comparable to the one in 3 despite the different chemical nature, resulted as able to form the ternary complex (data not shown). (Figure 3-c,d and g,h) also after optimization of the overall complex. 

PROTAC technology has been applied to a variety of targets and PROTACs have been recently as possible next-generation anti-coronavirus drugs. Here, we present the first PROTAC application in the context of coronavirus inhibition, in which a human host protein rather than a viral one was exploited as a target. Indeed, host-directed antivirals should have an advantage in overcoming resistance issues related to the emergence of mutations in the viral genome that might make the virus less susceptible to some inhibitors without affecting viral fitness. [48, 49] This is an important 

Unless otherwise noted, starting materials, reagents, and solvents were purchased from commercial suppliers and were used as received without further purification. Indomethacin (1) eq) was added and the reaction mixture was stirred at room temperature (5-18h). The mixture was poured in ice-water yielding a precipitate collected by filtration. The crude was purified as described below.

General Procedure A (16h) was followed by using 1 (0.030 g, 0.084 mmol) and VHL-linker 

General Procedure A (5h) was followed by using 1 (0.050 g, 0.139 mmol) and VHL-linker 

General Procedure A (16h) was followed by using 

A Committee on Microbiological Safety.

The effect on cell viability of test compounds was determined in Vero E6 and MRC-5 cells at 72 h by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT; Sigma-Aldrich) method as described previously. [50] J o u r n a l P r e -p r o o f

The crystallographic structure of human PGES-2 is not available in the Protein Data Bank. [51] Therefore, in order to model its interaction with indomethacin, we used a homology model obtained from the Macaca fascicularis PGES-2 crystallographic structure, since the two proteins show a sequence homology of 98.94% (a focus on the overlap of the INM binding site between the two proteins is showed in Figure S2) . selected, that is, those that allowed optimal contacts, without repulsive steric interactions between the INM-linker-VHL complex with the respective protein structures. To hypothesize the formation of the PGES-2/PROTAC/E3 ternary complex, GRID Molecular Interaction Fields (MIFs) [46, 47, 56] were calculated at the PGES-2 and VHL-E3 ligase protein surfaces in the proximity of the binding site, and the best complementarity of the hydrophobic and polar MIFs was used as the driving force of the ternary complex formation. Once the best orientation was found, the PROTAC was docked by overlapping the ligands onto the x-ray poses and adjusting the linker to get rid of clashes and to optimize interactions. To generate the hydrophobic MIF the CRY probe was used, [57] while polar interactions were evaluated using the N1 and the O probes, all within the GRID 2021 package.

[56]

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.

funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 871029, which provided free access to the SARS-CoV-2/NL/2020 virus. The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication. The authors thank Dr. Simon Cross for English revision. 

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

China Novel Coronavirus, T. Research, A novel coronavirus from patients with pneumonia in China

COVID-19) Dashboard

Genomic characterisation and epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding

Emerging coronaviruses: genome structure, replication, and pathogenesis

A new coronavirus associated with human respiratory disease in China

Genetic Recombination, and Pathogenesis of Coronaviruses

Clinical efficacy and safety of remdesivir in patients with COVID-19: a systematic review and network meta-analysis of randomized controlled trials

Solidarity" clinical trial for COVID-19 treatments

Racing to immunity: Journey to a COVID-19 vaccine and lessons for the future

Combating the Pandemic COVID-19: Clinical Trials, Therapies and Perspectives, Frontiers in molecular biosciences

Drug repurposing approach to combating coronavirus: Potential drugs and drug targets

Targeting SARS-CoV-2 proteases and polymerase for COVID-19 treatment: state of the art and future opportunities

Drug repurposing for the treatment of COVID-19: Pharmacological aspects and synthetic approaches

Drug repurposing approaches to combating viral infections

Drug repurposing for viral infectious diseases: how far are we?

Indomethacin and resveratrol as potential treatment adjuncts for SARS-CoV

Indomethacin has a potent antiviral activity against SARS CoV-2 in vitro and canine coronavirus in vivo, bioRxiv

Repurpose open data to discover therapeutics for COVID-19 using deep learning

A SARS-CoV-2 protein interaction map reveals targets for drug repurposing

Nonsteroidal anti-inflammatory drugs (NSAIDs): progress in small molecule drug development

Crystal structure and possible catalytic mechanism of microsomal prostaglandin E synthase type 2 (mPGES-2)

Low concentrations of indomethacin inhibit phospholipase A2 of rabbit polymorphonuclear leukocytes

Potential anti-SARS-CoV-2 therapeutics that target the post-entry stages of the viral life cycle: a comprehensive

Indomethacin has a potent antiviral activity against SARS coronavirus

In vitro assessment of the antiviral activity of ketotifen, indomethacin and naproxen, alone and in combination, against SARS-CoV-2

Comparative host-coronavirus protein interaction networks reveal pan-viral disease mechanisms

Mapping the SARS-CoV-2-host protein-protein interactome by affinity purification mass spectrometry and proximity-dependent biotin labeling: a rational and straightforward route to discover host-directed anti-SARS-CoV-2 therapeutics

Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs

PROTACs to address the challenges facing small molecule inhibitors

Prey for the proteasome: targeted protein degradation-A medicinal chemist's perspective

PROTACs: an emerging therapeutic modality in precision medicine

Proteolysis-targeting chimeras as therapeutics and tools for biological discovery

PROteolysis TArgeting Chimeras (PROTACs) -Past, present and future, Drug discovery today

Proteolysis targeting chimeras (PROTACs) come of age: entering the third decade of targeted protein degradation

Targeted protein degradation: elements of PROTAC design

Differential PROTAC substrate specificity dictated by orientation of recruited E3 ligase

Lessons in PROTAC design from selective degradation with a promiscuous warhead

Small molecule degraders of the hepatitis C virus protease reduce susceptibility to resistance mutations

The development of Coronavirus 3C-Like protease

PROTAC-induced BET protein degradation as a therapy for castration-resistant prostate cancer

Bi-functional compounds and methods for targeted ubiquitination of androgen receptor

A comparative analysis of remdesivir and other repurposed antivirals against SARS-CoV-2

Hydrogen bonding interactions of covalently bonded fluorine atoms: from crystallographic data to a new angular function in the GRID force field

A computational procedure for determining energetically favorable binding sites on biologically important macromolecules

Curing a viral infection by targeting the host: the example of cyclophilin inhibitors

Novel broad-spectrum antiviral inhibitors targeting host factors essential for replication of pathogenic RNA viruses

Selective anti-cytomegalovirus compounds discovered by screening for inhibitors of subunit interactions of the viral polymerase

The Protein Data Bank

The SWISS-MODEL Repository-new features and functionality

PGH2 degradation pathway catalyzed by GSH-heme complex bound microsomal prostaglandin E2 synthase type 2: the first example of a dual-function enzyme

Structural basis of PROTAC cooperative recognition for selective protein degradation

High end GPCR design: crafted ligand design and druggability analysis using protein structure

400 MHz, CDCl3) δ 8.67 (s, 1H)

Hz, 1H), 6.69 (dd, J = 9.2, 2.0 Hz, 1H), 6.27 (d, J = 8.3 Hz, 1H), 5.77-5.68 (m, 1H)

96 (s, 1H), 6.81 (s, 1H), 6.72 (d, J = 8.8 Hz, 1H), 6.59 (d

Hz, 2H), 7.43-7.31 (m, 4H), 6.88 (dd, J = 11.5, 5.9 Hz, 3H), 6.69 (dd, J = 9

400 MHz, MeOD) δ 3.62 -3.55 (m, 4H)