key: cord-0732904-etgqbhj3
authors: Garcia, Gustavo; Sharma, Arun; Ramaiah, Arunachalam; Sen, Chandani; Purkayastha, Arunima; Kohn, Donald B.; Parcells, Mark S.; Beck, Sebastian; Kim, Heeyoung; Bakowski, Malina A.; Kirkpatrick, Melanie G.; Riva, Laura; Wolff, Karen C.; Han, Brandon; Yuen, Constance; Ulmert, David; Purbey, Prabhat K.; Scumpia, Phillip; Beutler, Nathan; Rogers, Thomas F.; Chatterjee, Arnab K.; Gabriel, Gülsah; Bartenschlager, Ralf; Gomperts, Brigitte; Svendsen, Clive N.; Betz, Ulrich A.K.; Damoiseaux, Robert D.; Arumugaswami, Vaithilingaraja
title: Antiviral Drug Screen Identifies DNA-Damage Response Inhibitor as Potent Blocker of SARS-CoV-2 Replication
date: 2021-03-18
journal: Cell Rep
DOI: 10.1016/j.celrep.2021.108940
sha: 2d20c72384d095afd405e70d4822580dce6f8b65
doc_id: 732904
cord_uid: etgqbhj3

SARS-CoV-2 has currently precipitated the COVID-19 global health crisis. We developed a medium-throughput drug screening system and identified a small molecule library of 34 of 430 protein kinase inhibitors that were capable of inhibiting SARS-CoV-2 cytopathic effect in human epithelial cells. These drug inhibitors are in various stages of clinical trials. We detected key proteins involved in cellular signaling pathways mTOR-PI3K-AKT, ABL-BCR/MAPK, and DNA-Damage Response that are critical for SARS-CoV-2 infection. A drug-protein interaction based secondary screen confirmed compounds such as the ATR kinase inhibitor berzosertib and torin2 with anti SARS-CoV-2 activity. Berzosertib exhibited potent antiviral activity against SARS-CoV-2 in multiple cell types and blocked replication at post-entry step. Berzosertib inhibited replication of SARS-CoV-1 and MERS-CoV as well. Our study highlights key promising kinase inhibitors to constrain coronavirus replication as a host-directed therapy in the treatment of COVID-19 and beyond as well as provides an important mechanism of host-pathogen interactions.

The current pandemic is caused by a newly discovered coronavirus, severe acute respiratory 8 syndrome-related coronavirus 2 (SARS-CoV-2). As of today, the disease has spread to 215 the therapeutic potential of such drugs to combat viral infections. The high average cost (over 10 two billion dollars) and long timeline (8-12 years) to develop a new drug(Development, 2014), 11 limit the scalability of the DAA approach to drug development, particularly with respect to 12 emerging viruses. This approach is therefore not feasible for the short-term development of a 13 cure that is specific for SARS-CoV-2. The screening of approved drugs to identify therapeutics 14 for drug repurposing is an effective approach that has been used for many viral diseases, To shed light on an effective antiviral therapy for COVID-19 treatment, we established a SARS-28

CoV-2 infectious cell culture system and virological assays using Vero E6 cells. The SARS-29

CoV-2 isolate USA-WA1/2020 was obtained from BEI Resources of National Institute of Allergy 30

and Infectious Diseases (NIAID) and studies involving live virus were conducted in BSL3 high-31 containment facility. The SARS-CoV-2 was passaged once in Vero E6 cells and viral stocks 32 were aliquoted and stored at -80°C. Virus titer was measured in Vero E6 cells by TCID50 assay. presence of viral infection ( Figure S1B ). We also demonstrated that the drugs, 4 hydroxychloroquine (HQ; 10 µM), a known endosomal acidification inhibitor, as well as 5 interferon(IFN)-beta, IFN-alpha, and EIDD-2801 (Molnupiravir), effectively blocked SARS-CoV-2 6 infection (Liu et al., 2020) ( Figure S1C ). Therefore, we used this platform for subsequent drug 7 screening studies. 8 9

To evaluate the antiviral properties of cellular protein kinase inhibitors, we performed medium 10 throughput primary drug screening ( Figure 1A ) by selecting a drug compound library that 11 broadly covers 430 kinase inhibitors with the screening compound concentration that would 12 have potent anti-viral activity with low level of toxicity. There are only 518 human protein kinases 13 described with 478 kinases belonging to a single superfamily. The limitation in the number of 14 druggable kinases along with our criterion of selecting kinase inhibitors that are being evaluated 15

for clinical studies led to the use of 430 compounds. These kinase inhibitors have typically been 16 tested for oncologic and immunologic indications, for which we have clinical trial Phase 1/2/3 17 data (Table S1 ), but no data available on SARS-CoV-2. Since this kinase inhibitor library targets 18 cancer indications, we decided to avoid using human lung cancer epithelial cell lines. Drug 19 compounds were formulated into DMSO and pre-plated into media at a 2x concentration (final 20 drug concentration 250 nM). Compounds were added to the Vero E6 cells in the BSL-3 21 laboratory followed by SARS-CoV-2 at a Multiplicity of Infection (MOI) 0.1. After the 48-hour 22 incubation at 37°C, 5% CO2, viral CPE was scored and imaged ( Figure 1A ). The compounds 23 that prevented the viral CPE were identified (Table S2 and Figure S2 ) and subjected to pathway 24 analysis ( Figure 1B To rapidly confirm and prioritize the most promising compounds according to their anti-SARS-33

CoV-2 activity, we selected 34 compounds from the primary screening for secondary screening 34 with multiple drug doses (2.5, 25, 250 and 500 nM) in triplicate, using 96-well plates. We used 1 an immunofluorescent assay to quantify viral infected cells from each well. This comprehensive 2 screen in Vero E6 cells as well as HEK293-ACE2 cells verified many hits of the primary screen 3 ( Figure 1B -D and Figure S3 and S4). In Vero E6 cells, compounds berzosertib (M6620), torin2 4 and vistusertib (AZD2014) all demonstrated antiviral activity at IC50 below 25nM, while, 5 nilotinib, NVP-BHG712, VPS34-INI, URMC-099, and YM201636 showed IC50 ranges between 6 50nM -125nM ( Figure 1E and Figure 3 ). We also included an additional kinase inhibitor, 7 dactolisib, in the secondary validation step. These compounds act against SARS-CoV-2 by 8 limiting viral infection through inhibiting critical cellular enzymes needed for viral replication. In 9

contrast, nucleoside analogues such as remdesivir, EIDD-2801 (Molnupiravir), and ribavirin 10 have been shown to inhibit viral RNA dependent RNA polymerase (RdRp) enzyme and could 11 also induce compromising errors during viral genome replication. Our secondary screen 12 confirmed the antiviral activities of these kinase inhibitors, and the critical cellular pathways 13 subjected to activation, suppression, or some form of modulation by viral infection. Interestingly, 14

we observed that several antiviral compounds targeted the mTOR-PI3K-AKT pathway, including 15 dactolisib, AZD2014, and torin2 ( Figure 1B ). In general we observed that compounds blocking 16 this pathway had relatively increased cytotoxicity. The mammalian target of rapamycin (mTOR) To further confirm drug efficacy and to investigate mechanism of action, we tested drugs using 27 human cells. We focused on a class of antiviral compound from our screen, berzosertib, which 28 is a DNA damage response pathway inhibitor targeting the protein kinase ATR (ataxia 29 telangiectasia and Rad3-related protein) and already in Phase 2 clinical trials for solid tumors 30 Calu-3 ( Figure 3 ). Calu-3 cells were infected, treated with berzosertib and at 48 hpi, 6

supernatants were collected and titrated on Vero E6 cells to determine viral titers and IC 50 7 values. Berzosertib exhibited an IC 50 =0.48µM for SARS-CoV2 ( Figure 3A ) with a similar activity 8 against SARS-CoV1 ( Figure 3C ) and MERS-CoV ( Figure 3D ). In comparison remdesivir under 9 the same assay conditions showed an IC 50 =0.15µM ( Figure 3B ). In an assay conducted with 10 A549-ACE2 cells infected with SARS-CoV-2, berzosertib exhibited an IC50=0.22 ± 0.03 µM; 11 SI=204. Interestingly, it could be demonstrated that berzosertib is acting in a synergistic manner 12 in combination treatment with remdesivir, which showed an IC50 of 0.2 µM in the same system 13 ( Figure 3F ). The respective inhibition curves and the isobologram are shown in Figure Lastly, we tested berzosertib in a primary human lung tissue culture system consisting of 21 mucociliary air-liquid interface (ALI) cultures derived from primary human tissue (Purkayastha et 22 al., 2020) . Also in this ALI system, berzosertib was effective in inhibiting SARS-CoV-2 ( Figure  23 4A and B). Taken together, our results on a human primary cell system suggest that berzosertib 24 is a potent and safe class of antivirals against coronavirus infections with a low risk of cardiac 25 adverse events. 26 27 We further investigated a potential mode of action of berzosertib in modulating host cellular 28 signaling pathways. Berzosertib is a selective inhibitor of serine/threonine-protein kinase ATR, 29 which is an important member of DNA Damage Response pathway. Berzosertib has been 30 previously shown to block the ATR-CHK1 pathway in cancer cells, with no discernable effect on 31 normal cells (Mei et al., 2019) . In healthy cells, the ATR-CHK1 pathway is required for 32 maintaining cellular genome integrity. ATR kinase is also playing an important role as a Figure S5E ). We observed that Vero E6 cells had higher basal level of pCHK1, however 15 treatment with berzosertib reduced phosphorylation of CHK1 and inhibited SARS-CoV-2 16 infection ( Figure 4C and Figure S5E ). inhibitors are different. Moreover, while one may expect an IC50 on e.g. viral RNA replication in 23 the same range as the IC50 of the ATR kinase inhibition by berzosertib, this is by no means 24 assured. It has to be taken into consideration that the ATR pathway is activated by SARS-CoV-25 2, which will have its effects on the IC50. Our STRING analysis places the ATR pathway into 26 direct context with e.g. the mTOR axis (Figure 1 ), which is known to be vital for SARS CoV- 2 27 pathophysiology. Thus berzosertib activity would be consistent with an on-target mode of action. 28

However, it is possible that the berzosertib anti-SARS-CoV-2 activity might be directly targeted 29 towards viral factors. Differing antiviral activity could also be due to inhibition of additional other 30 cellular targets, as e.g. berzosertib has an inhibitory effect on DYRK2 and AXL. DYRK2 has 31 been shown previously to negatively regulate type I interferon induction and a knockdown of 32 in a cancer setting. Although toxicity considered acceptable here, that may not be acceptable in 25 other pathological conditions. Moreover, current NIH COVID-19 treatment guidelines discourage 26 the use of e.g. JAK inhibitors as they are immunosuppressive. In contrast, berzosertib is an 27 excellent candidate for rapid repurposing towards treating COVID-19 patients as it does not 28 have e.g. immunosuppressive/thrombotic side effect of JAK/STAT inhibitors, which is 29 contraindicated in an infectious disease context (Mehta et al., 2020) . In fact, berzosertib is 30 unique in its mode of action since it is intended as a potentiator of the therapeutic effects of 31 genotoxic drugs in oncology, i.e. it is not a standalone therapy. In addition, non-ATR kinase 32 inhibitors can have a much stronger direct effect on cell viability, thus one can speculate that 33 their adverse side effects in a severely affected COVID-19 patients may be much more 1 pronounced than berzosertib. 

This study did not generate new unique reagents. 

CoV-2/Germany/HPI06-n/2020) was isolated from a nasal swab of a SARS-CoV-2 infected 28 patient who was treated in an ICU at the University Medical Campus Hamburg-Eppendorf, 29

Hamburg, Germany. All the studies involving live virus was conducted in UCLA BSL3 high-30 containment facility. SARS-CoV-2 was passaged once in Vero E6 cells and viral stocks were 31 aliquoted and stored at -80 o C. Virus titer was measured in Vero E6 cells by established plaque 32 assay or TCID50 assay. were collected from infected cells and stored at -80 °C. Then, viral titers were determined by 1 plaque test on Vero E6 cells as described below. 2 3 A549 cells stably expressing ACE2 were seeded into 96-well plates. On the next day, 3-fold 4 serial dilutions of given drugs were added to the cells covering a drug concentration range of 2.5 5 nM -50 µM for Berzosertib, and 0.5 nM -10 µM for remdesivir. Thereafter, cells were inoculated 6 with SARS-CoV-2 (MOI = 1) and 24 h later, cells were fixed and viral nucleocapsid was 7 detected by immunostaining using a secondary antibody that was coupled to horseradish 8 peroxidase. Bound secondary antibody was detected by colorimetric assay and signal was 9 quantified by measuring absorbance at 405 nm. Values were normalized using solvent control 10 The qPCR primer pairs for mRNA transcript targets are provided in the Key Resources Table. . 9 SARS-CoV-2 RNA transcript levels were quantified by comparing them to a standard curve 10 generated using serial ten-fold dilutions (10 1 -10 9 copies) of a SARS-CoV-2 N gene containing 11 plasmid. SARS-CoV-2 RNA levels were expressed as SARS-CoV-2 genome copies per 1 µg of 12 RNA using the standard curve. Inc.) (Chou, 2006) . All testing was done at the two-sided alpha level of 0.05. Data were analyzed 29 for statistical significance using unpaired student's t-test to compare two groups (uninfected vs. Repurposed drug screen identifies cardiac glycosides as inhibitors of TGF-β-induced 35

cancer-associated fibroblast differentiation. Identification of Antiviral Drug Candidates against SARS-CoV-2 from FDA-Approved 9

Drugs. Antimicrobial Agents and Chemotherapy, 64, e00819-20. 

A 1 genome-wide genetic screen for host factors required for hepatitis C virus propagation

Marek's 4 Disease Virus Disables the ATR-Chk1 Pathway by Activating STAT3

Using or abusing: viruses and the 6 cellular DNA damage response

Hydroxychloroquine, a less toxic derivative of chloroquine, is effective in 9 inhibiting SARS-CoV-2 infection in vitro

Cerebral Micro-Structural Changes in COVID-19 Patients &#x2013

An MRI-based 3-month Follow-up Study

A systematic screen 17 of FDA-approved drugs for inhibitors of biological threat agents

JAK inhibitors in 20 COVID-19: the need for vigilance regarding increased inherent thrombotic risk. The 21 European respiratory journal

Ataxia telangiectasia and Rad3-related 23 inhibitors and cancer therapy: where we stand

Modulation of the cell growth regulator mTOR by Epstein-26 Barr virus-encoded LMP2A

Nonstructural Protein 5A 28 Impairs DNA Damage Repair: Implications for Hepatitis C Virus

Small-molecule protein kinase inhibitors and their effects on the 31 immune system: implications for cancer treatment

Persistent 34 Replication of Severe Acute Respiratory Syndrome Coronavirus in Human Tubular 35 Kidney Cells Selects for Adaptive Mutations in the Membrane Protein

Take your PIKK: tumour viruses 38 and DNA damage response pathways

The emerging spectrum of COVID-19 neurology: 1 clinical, radiological and laboratory findings

Multiorgan and Renal Tropism of SARS-CoV-2

Cigarette Smoke Increases Infection Severity and Alters the Stem Cell-Derived Airway 13 Repair Response

Insights into Cross-species Evolution of Novel 15 Human Coronavirus 2019-nCoV and Defining Immune Determinants for Vaccine 16 Development

Activation of the DNA Damage 25 Response by RNA Viruses

Cancer-Specific Synthetic Lethality between ATR and CHK1 Kinase Activities. 30 Cell Rep

Children's heart and COVID-19: Up-to-date evidence in the form of a systematic 33 review

Repurposing of Kinase Inhibitors as Broad-Spectrum Antiviral 35 Drugs

Derived Cardiomyocytes are Susceptible to SARS-CoV-2 Infection

40 Derivation of highly purified cardiomyocytes from human induced pluripotent stem cells 41 using small molecule-modulated differentiation and subsequent glucose starvation

Association of Cardiac Injury With 45 Mortality in Hospitalized Patients With COVID-19 in Wuhan

The 48 TSC2/mTOR pathway drives endothelial cell transformation induced by the Kaposi's 49 sarcoma-associated herpesvirus G protein-coupled receptor

Axl Promotes Zika Virus Entry and 2 Modulates the Antiviral State of

Inhibition of ATR downregulates PD-L1 5 and sensitizes tumor cells to T cell-mediated killing

Patient-specific induced pluripotent stem cells as 10 a model for familial dilated cardiomyopathy

Identification of human kinases involved in hepatitis C virus 14 replication by small interference RNA library screening

Myocardial localization of coronavirus in COVID-19 cardiogenic shock

Phase I 24 Study of ATR Inhibitor M6620 in Combination With Topotecan in Patients With 25 Advanced Solid Tumors

& 28 MOCH, H. 2020. Endothelial cell infection and endotheliitis in COVID-19

Viral DNA replication-31 dependent DNA damage response activation during BK polyomavirus infection

Mammalian target of rapamycin: a signaling kinase for every aspect of 34 cellular life

Repurposing of Kinase 37 Inhibitors for Treatment of COVID-19

Broad Anti-coronavirus Activity of Food and 40 Drug Administration-Approved Drugs against SARS-CoV-2 <em>In Vitro</em> and 41 SARS-CoV <em>In Vivo</em>

Single-cell RNA-45 seq data analysis on the receptor ACE2 expression reveals the potential risk of different 46 human organs vulnerable to 2019-nCoV infection

High 48 expression of ACE2 receptor of 2019-nCoV on the epithelial cells of oral mucosa. 49

Coronavirus infection induces DNA 1 replication stress partly through interaction of its nonstructural protein 13 with the p125 2 subunit of DNA polymerase δ

Pathological findings of COVID-19 associated with acute respiratory distress 6 syndrome. The Lancet Respiratory Medicine

Coronavirus Disease 2019 (COVID-19) and the

Heart-Is Heart Failure the Next Chapter

Phase I Trial of First-in-Class ATR 14 Inhibitor M6620 (VX-970) as Monotherapy or in Combination With Carboplatin

Patients With Advanced Solid Tumors

20 Inhibition of ATM Increases Interferon Signaling and Sensitizes Pancreatic Cancer to 21 Immune Checkpoint Blockade Therapy

These drugs have been shown to modulate cellular signaling cascades including mTOR-PI3K-AKT and DNA-damage response (DDR) pathways. A highly effective drug candidate, berzosertib, blocked multiple coronaviruses such as SARS-CoV-1, SARS-CoV-2, and MERS-CoV, thus providing a potential therapeutic against COVID-19. Highlights: • Kinase inhibitor screen identified 34 compounds with anti-SARS-CoV-2 activity • Inhibitors targeted mTOR-PI3K-AKT and DNA-damage response (DDR) signaling pathways • ATR kinase inhibitor berzosertib blocked SARS-CoV-1, SARS-CoV-2, and MERS-CoV Figure S5. Berzosertib inhibits SARS-CoV-2 replication in human HeLa-ACE2 cells and affects DDR pathway. (A) Graph shows antiviral activity measured with a SARS-CoV-2 immunofluorescence signal leading to identification of infected cells with 0% activity equals 100% infected cells. (B) total cells per well in SARS-CoV2 infected cell test with 0% activity equaling no change vs. control (C) total cells per well in HeLa-ACE2

Western blot analysis shows phosphorylation of key ATR kinase downstream target protein CHK1

Evaluating additional DNA-Damage Pathway ATR Kinase Inhibitor. Vero E6 cells pre-treated (24 hours) with AZD6738 (10µM) were infected with SARS-CoV-2 48 hours postinfection virus replication was visualized with immunostaining. Scale Bar=100 µm. IFA images show SARS-CoV-2 (red) infection in vehicle or AZD6738 treated cells. No inhibition of SARS-CoV-2 infection in AZD6738 (10µM) treated cells is noted

J o u r n a l P r e -p r o o f SUPPLEMENTARY TABLE   Table S2 . Drugs compounds selected from primary screen having antiviral activity at 250 nM concentration (Related to Figure 1 ).