key: cord-0784347-3fbxaeyf authors: Dampalla, Chamandi S.; Zhang, Jian; Perera, Krishani Dinali; Wong, Lok-Yin Roy; Meyerholz, David K.; Nguyen, Harry Nhat; Kashipathy, Maithri M.; Battaile, Kevin P.; Lovell, Scott; Kim, Yunjeong; Perlman, Stanley; Groutas, William C.; Chang, Kyeong-Ok title: Post-infection treatment with a protease inhibitor increases survival of mice with a fatal SARS-CoV-2 infection date: 2021-02-05 journal: bioRxiv DOI: 10.1101/2021.02.05.429937 sha: a38e70207fb6687f9ab337c1758ccfcc237038eb doc_id: 784347 cord_uid: 3fbxaeyf Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection continues to be a serious global public health threat. The 3C-like protease (3CLpro) is a virus protease encoded by SARS-CoV-2, which is essential for virus replication. We have previously reported a series of small molecule 3CLpro inhibitors effective for inhibiting replication of human coronaviruses including SARS-CoV-2 in cell culture and in animal models. Here we generated a series of deuterated variants of a 3CLpro inhibitor, GC376, and evaluated the antiviral effect against SARS-CoV-2. The deuterated GC376 displayed potent inhibitory activity against SARS-CoV-2 in the enzyme and the cell-based assays. The K18-hACE2 mice develop mild to lethal infection commensurate with SARS-CoV-2 challenge doses and was proposed as a model for efficacy testing of antiviral agents. We treated lethally infected mice with a deuterated derivative of GC376. Treatment of K18-hACE2 mice at 24 hr post infection with a derivative (compound 2) resulted in increased survival of mice compared to vehicle-treated mice. Lung virus titers were decreased, and histopathological changes were ameliorated in compound 2-treated mice compared to vehicle-treated mice. Structural investigation using high-resolution crystallography illuminated binding interactions of 3CLpro of SARS-CoV-2 and SARS-CoV with deuterated variants of GC376. Taken together, deuterated GC376 variants have excellent potential as antiviral agents against SARS-CoV-2. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection continues to be a serious global public health threat. The 3C-like protease (3CLpro) is a virus protease encoded by SARS-CoV-2, which is essential for virus replication. We have previously reported a series of small molecule 3CLpro inhibitors effective for inhibiting replication of human coronaviruses including SARS-CoV-2 in cell culture and in animal models. Here we generated a series of deuterated variants of a 3CLpro inhibitor, GC376, and evaluated the antiviral effect against SARS-CoV-2. The deuterated GC376 displayed potent inhibitory activity against SARS-CoV-2 in the enzyme and the cell-based assays. The K18-hACE2 mice develop mild to lethal infection commensurate with SARS-CoV-2 challenge doses and was proposed as a model for efficacy testing of antiviral agents. We treated lethally infected mice with a deuterated derivative of GC376. Treatment of K18-hACE2 mice at 24 hr post infection with a derivative (compound 2) resulted in increased survival of mice compared to vehicle-treated mice. Lung virus titers were decreased, and histopathological changes were ameliorated in compound 2-treated mice compared to vehicle-treated mice. Structural investigation using high-resolution crystallography illuminated binding interactions of 3CLpro of SARS-CoV-2 and SARS-CoV with deuterated variants of GC376. Taken together, deuterated GC376 variants have excellent potential as antiviral agents against SARS-CoV-2. Coronaviruses are a large group of viruses that can cause a wide variety of diseases in humans and animals (1) . They are single-stranded, positive-sense RNA viruses that belong to four genera, designated (4) and, most recently, SARS-CoV-2, the causative agent of COVID-19 (5, 6) . SARS-CoV-2 emerged in China in December 2019 and subsequently spread throughout the world. Ominously, the diversity of coronavirus strains in potential animal reservoirs suggests that emerging and reemerging pathogenic coronaviruses will continue to pose a significant threat to public health (7, 8) . Currently, vaccines using different platforms have been developed or under development, and two vaccines just became available in the US for COVID-19 licensed for emergency use with more others expected to be available soon. The specific therapeutic interventions that are currently licensed include a nucleoside analogue remdesivir (Veklury®), a combination of remdesivir and a JAK inhibitor baricitinib, and a cocktail of anti-SARS-CoV-2 monoclonal antibodies. These treatments may diminish disease progression, but such effects are not found in most studies, indicating the urgent necessity to develop additional antiviral therapies (9) (10) (11) (12) . The SARS-CoV-2 genome encodes two polyproteins which are processed by a 3C-like protease (3CLpro) and a papain-like protease. These viral proteases are essential for viral replication, making them attractive targets for drug development (13) (14) (15) (16) (17) (18) . It is furthermore acknowledged that, in addition to the development of effective vaccines, the concurrent identification of FDA-approved drugs that can be repurposed for use against SARS-CoV-2 may accelerate the development and implementation of effective countermeasures against the virus (19, 20) . We previously described a series of 3CLpro inhibitors (including GC376) with activities against multiple coronaviruses, including SARS-CoV(21), MERS-CoV (13, 22) and SARS-CoV-2 (13) . GC376 was recently demonstrated in clinical trials to have efficacy against a fatal feline coronavirus infection, feline infectious peritonitis (FIP) (23, 24) , and is currently in clinical development for treating FIP in cats. Mice expressing human angiotensin I-converting enzyme 2 (ACE2) receptor under the cytokeratin-18 (K18) promoter, designated as K18-hACE2 mice, were previously proposed as a model for efficacy testing of antiviral agents (25, 26) . We report herein the results of our studies related to the synthesis and evaluation of deuterated GC376 variants which have enhanced antiviral activity and display efficacy in a fatal mouse model (K18-hACE2 mice) of SARS-CoV-2. SARS-CoV-2 in the enzyme and the cell-based assays. We synthesized deuterated variants based on GC376 ( Figure S1 , and Supporting information) and compared their inhibitory activities against SARS-CoV-2 to non-deuterated GC376 in the enzyme and the cell-based assays (Table 1) . Three different variants of deuterated aldehyde compounds (compounds 1, 6 and 9 with R 1 , R 2 and R 3 , respectively) as well as their bisulfite adducts (compounds 2, 7 and 10) were prepared for the testing. In addition, an α -ketoamide (compound 5) based on compound 1 and prodrug variations (compounds 3, 4, 8 and 11) of the bisulfite adducts of aldehydes (compounds 1, 6 and 9) were synthesized for the testing. In the enzyme assay, the bisulfite adducts showed similar 50% inhibitory concentration (IC 50 ) values as their aldehyde counterparts (Table 1C) . The (Table 1C ). The deuterated compounds that were more effective than GC376 in the enzyme assay were tested in the cell-based assay. The 50% effective concentration (EC 50 ) values of the tested deuterated compounds (compounds 1, 2, 6 and 7) (0.068 to 0.086 µM) were lower than GC376 by 2.67~3.38-fold in Vero E6 cells. All compounds, including GC376, did not show any cytotoxicity up to 100 µM (Table 1C) . variants of GC376. Compound 2 with a bisulfite adduct warhead and compound 5 with α -ketoamide were co-crystallized with the 3CLpro of SARS-CoV-2 and SARS-CoV and examined by X-ray crystallography. Examination of the active site of SARS-CoV-2 3CLpro revealed the presence of prominent difference electron density consistent with compound 2 covalently bound to the Sγ atom of Cys 145 in each subunit ( Figure 1A and B) . Interestingly, the electron density was most consistent with the S-enantiomer at the newly formed stereocenter. Although the electron density in subunit B did contain a small "bulge" that may be due to the R-enantiomer, only one configuration was modeled. Compound 2 adopts the same binding mode in each subunit and forms identical hydrogen bond interactions with residues Phe 140 , His 163 , His 164 , Glu 166 and Gln 189 ( Figure 1D and E). As we generally observed in studies of SARS-CoV 3CLpro, the electron density map was consistent with both the R and S-enantiomers of compound 2 at the new stereocenter formed by covalent attachment of the Sγ atom of Cys 145 in the cocrystal structure of SARS-CoV 3CLpro ( Figure 1C ). Overall, the hydrogen bond interactions are nearly identical relative to SARS-CoV-2 3CLpro. The main difference is that a hydrogen bond is formed between His 41 and the hydroxyl of compound 2 in the R-enantiomer and a long contact (3.29 Å) to the backbone N-atom of Ser 144 with the hydroxyl of the S-enantiomer ( Figure 1F) . Notably, the hydroxyl in compound 2 bound to SARS-CoV-2 3CLpro is 3.38 Å and 3.39 Å from the N-atom of Ser 144 , which would be a weak hydrogen bond contact. The benzyl ring in both structures is positioned outward from the hydrophobic S 4 subsite and are directed towards the surface as shown in Figure 1G , H and I. Notably, the structures of SARS-CoV-2 3CLpro in complex with nondeuterated G376 and its precursor aldehyde GC373 (PDB 6WTJ and 6WTK, respectively) adopts the same binding mode as that observed for compound 2 ( Figure S2 ). Superposition yielded root-mean-square deviation (RMSD) deviations of 0.59 Å (GC376) and 0.55 Å (GC373) between Cα atoms for 299 residues aligned (27) . The structures of SARS-CoV and SARS-CoV-2 3CLpro in complex with compound 5 also contained prominent difference in electron density consistent with the inhibitor covalently bound to the Sγ atom of Cys 145 ( Figure S3A and D). The entire inhibitor could be modeled in subunit A but was partially disordered in subunit B and the benzyl group in the S 4 subsite could not be modeled for SARS-CoV. The inhibitor forms direct hydrogen bond interactions similar to compound 2 as shown in Figure S3B Interestingly, the compound bound to subunit B of SARS-CoV-2 3CLpro adopts a conformation similar to that observed for compound 2 in which the benzyl group is directed away from the S 4 subsite and towards the surface (FigureS4). Therefore, it appears the structure of SARS-CoV-2 3CLpro in complex with compound 5 serendipitously contains two binding modes of the inhibitor. mice. Compound 2 was tested in SARS-CoV-2-infected K18-hACE2 mice for protective efficacy, because it potently inhibited SARS-CoV-2 in the cell-based assay described above. The dose curve of compound 2 against SARS-CoV-2 in cell culture is shown in Figure 2A . In the first experiment, infection with 2x10 3 pfu per mouse led to body weight loss in all vehicle-treated mice resulting in 50% survival by 9 dpi ( Figure 2B ). Mice treated with compound 2 (100 mg/kg/day, once a day) starting from 24 hr post infection (1 dpi) lost body weight, but loss was less severe compared to vehicle-treated mice with statistically significant differences (0.002