Potent 3CLpro inhibitors effective against SARS-CoV-2 and MERS-CoV in animal models by therapeutic treatment

ABSTRACT Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and Middle East respiratory syndrome coronavirus (MERS-CoV) are zoonotic betacoronaviruses that continue to have a significant impact on public health. Timely development and introduction of vaccines and antivirals against SARS-CoV-2 into the clinic have substantially mitigated the burden of COVID-19. However, a limited or lacking therapeutic arsenal for SARS-CoV-2 and MERS-CoV infections, respectively, calls for an expanded and diversified portfolio of antivirals against these coronavirus infections. In this report, we examined the efficacy of two potent 3CLpro inhibitors, 5d and 11d, in fatal animal models of SARS-CoV-2 and MERS-CoV to demonstrate their broad-spectrum activity against both viral infections. These compounds significantly increased the survival of mice in both models when treatment started 1 day post infection compared to no treatment which led to 100% fatality. Especially, the treatment with compound 11d resulted in 80% and 90% survival in SARS-CoV-2 and MERS-CoV-infected mice, respectively. Amelioration of lung viral load and histopathological changes in treated mice correlated well with improved survival in both infection models. Furthermore, compound 11d exhibited significant antiviral activities in K18-hACE2 mice infected with SARS-CoV-2 Omicron subvariant XBB.1.16. The results suggest that these are promising candidates for further development as broad-spectrum direct-acting antivirals against highly virulent human coronaviruses. IMPORTANCE Human coronaviruses such as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and Middle East respiratory syndrome coronavirus (MERS-CoV) continue to have a significant impact on public health. A limited or lacking therapeutic arsenal for SARS-CoV-2 and MERS-CoV infections calls for an expanded and diversified portfolio of antivirals against these coronavirus infections. We have previously reported a series of small-molecule 3C-like protease (3CLpro) inhibitors against human coronaviruses. In this report, we demonstrated the in vivo efficacy of 3CLpro inhibitors for their broad-spectrum activity against both SARS-CoV-2 and MERS-CoV infections using the fatal animal models. The results suggest that these are promising candidates for further development as broad-spectrum direct-acting antivirals against highly virulent human coronaviruses.


Activity of 5c/d and 11c/d on SARS-CoV-2 pseudovirus entry assays in cells
The efficacy of 5c/d and 11c/d against 3CLpros of SARS-CoV-2 and MERS-CoV in enzyme assays and SARS-CoV-2 replication and pseudovirus entry assay in cells expressing hACE2 or hACE2 plus TMPRSS2 is summarized in Table 1.The potency of 5c/d and 11c/d against 3CLpros of SARS-CoV-2 and MERS-CoV or SARS-CoV-2 replicon was high with nanomolar IC 50 and EC 50 values and was reported in our previous publication (41).The EC 50 values of 5c/d and 11c/d against the entry of SARS-CoV-2 pseudoviruses, and all data from MDL28170 and Nafamostat were newly generated for this study.As expected, MDL28170 and Nafamostat did not have activity against the 3CLpros of SARS-CoV-2 and MERS-CoV.In terms of virus entry, 5c/d and 11c/d, and MDL28170 strongly inhibited the entry of pseudotyped viruses into cells expressing hACE2 but not in cells expressing hACE2 plus TMPRSS2.In contrast, Nafamostat was highly potent at inhibiting pseudovirus entry in cells expressing both hACE2 and TMPRSS2 with an EC 50 0.001 µM, but it had little effect in cells expressing hACE2 alone at up to 50 µM, as expected for a TMPRSS2 inhibitor (Table 1).

Modelling of the 11c binding mode with SARS-CoV-2 and MERS-CoV 3CLpro
Compound 11c adopts a conformation in the SARS-CoV-2 3CLpro active site where the side chains of the glutamine surrogate and Leu residues of the inhibitor are positioned with the S1 and S2 pockets, respectively, resulting in similar hydrogen bond interactions typically observed in the crystal structures (Fig. 1A).The orientation of the o-methoxyl phenyl ring is positioned near the hydrophobic residues in the S4 subsite with the methoxy group directed towards the surface (Fig. 1B).Superposition (43) of the 11c bound structure with the apo structure of SARS-CoV-2 3CLpro is similar overall with an RMSD deviation between Cα atoms of 0.75 Å (298 residues).The main differences are observed in the loop spanning D187-G195 near the S4 subsite which moves toward the inhibitor to form a hydrogen bond with Q189 and also engage in hydrophobic interactions with the inhibitor (Fig. 1C).The binding mode of 11c in the active site of MERS 3CLpro adopts a similar binding mode relative to SARS-CoV-2 3CLpro, as shown in Fig. 1D.The main difference is that the o-methoxyphenyl ring is positioned deep into  the S4 pocket although no hydrogen bonds are formed with the methoxy group (Fig. 1E).Although there are no apo MERS 3CLpro structures available to our knowledge, the crystal structure of the catalytically inactive mutant (C148A) has been determined in the apo form.Superposition indicated a high degree of structural similarity with RMSD deviation between Cα atoms of 0.75 Å (297 residues).Unlike SARS-CoV-2 3CLpro, the loop near the S4 subsite (F188-Q197) is similar in both structures (Fig. 1F).

Survival and morbidity in infected mice
Compounds 5d and 11d were tested in mouse models of SARS-CoV-2 or MERS-CoV infection because both compounds had high potency in the enzyme and cell-based assays.In mice infected with mouse-adapted (MA) SARS-CoV-2 (SARS2-N501Y MA30 ) or MA-MERS-CoV, treatment with vehicle led to 100% fatality by 6 or 9 days after virus infection, respectively (Fig. 2B and C).However, compound 11d treatment led to the survival of 80% or 90% of mice infected with MA-SARS-CoV-2 or MA-MERS-CoV, respectively (Fig. 2B and C).Treatment with compound 5d resulted in 30% or 50% survival of MA-SARS-CoV-2 or MA-MERS-CoV -infected mice, respectively.Mice treated with vehicle lost significant body weight following MA-SARS-CoV-2 or MA-MERS-CoV infection before they were euthanized (Fig. 2B and C).Less weight loss was observed in mice treated with compound 5d or 11d compared to vehicle-treated mice (Fig. 2B and C).
After reaching a nadir at 7-9 dpi, surviving mice treated with 5d or 11d gradually gained body weight, but weight gain was greater in those treated with 11d compared to 5d treatment (Fig. 2B and C).

Viral loads and lung pathology in animal studies
We  or MA-MERS-CoV infected mice showed significantly lower expression of these inflammatory genes compared to vehicle-treated mice (Fig. 4C and F).

Post-infection treatment of 11d protects K18-hACE2 mice against Omicron XBB.1.16 infection
We tested the antiviral ability of 11d on SARS-CoV-2 Omicron subvariant XBB.1.16 in K18-hACE2 mice.In mice infected with the Omicron variant, treatment with vehicle led to 60% fatality by 14 days after virus infection, but compound 11d treatment reduced the fatality to 20%.Vehicle-treated mice lost over 20% of weight at 7 dpi to the end.Mice treated with 11d reduced weight losses compared to those treated with vehicle (Fig. 5B), showing significant difference at 5 to 9 dpi.The weight of 11d treated mice started to recover from 8 dpi, but those treated with vehicle did not recover weight losses to 14 dpi (Fig. 5B).When assessed of viral loads in nasal wash, nasal turbinate, and lungs, viral titers from all tissues were significantly lower with 11d treatment compared to those with vehicle treatment at 3 and 5 dpi (Fig. 5D).While severe edema (average score 3.5) and perivascular infiltrates were evident in the lungs from vehicle-treated animals, 11d treatment led to significantly reduced lung pathologies with little edema (average score <0.5) (Fig. 5E and F).
In our study, we determined the dual roles of 3CLpro inhibitors against SARS-CoV-2 3CLpro and cathepsin enzymes, as well as post-infection efficacy of select compounds in fatal mouse models of SARS-CoV-2 or MERS-CoV infections.Studies of entry of SARS-CoV-2 into cells have shown that virus entry routes through the cell membrane are dependent on the expression of the host cell surface protease TMPRSS2, and virus entering cells lacking TMPRSS2 via the endosomal pathway, which requires endosomal enzymes, such as cathepsin L and B (31,50).Therefore, compounds targeting multiple targets such as virus 3CLpro and cathepsins may be beneficial for antiviral activity and may limit antiviral resistance.One of our reported compounds, GC376, was previously shown to have dual activity against cathepsin (23,41), and we determined that 5c/d and 11c/d have anti-cathepsin B and L activity with nanomolar IC 50 values (Table 2).These compounds have potent activity against SARS-CoV-2 and MERS-CoV in enzyme assay and viral replication (SARS-CoV-2 replicon) (Table 1).The dual inhibitory effects of the compounds in cells are demonstrated in cells lacking expression of TMPRSS2, where virus entry is presumed to be dependent on the activity of cathepsins (Table 1).As most susceptible cells in the respiratory system express both hACE2 and TMPRSS2, inhibition of the viral endosomal route may not play a significant antiviral role in the major target tissue of SARS-CoV-2 in humans.However, SARS-CoV-2 has been reported to infect cells with no or little TMPRRS in other organs, including the intestines with clinical effects (51,52), and it remains to be determined whether this dual effect offers clinical benefits.SARS-CoV-2 and MERS-CoV, both betacoronaviruses, belong to different subgenera, Sarbecovirus and Merbecovirus, respectively.The homology of the amino acid sequences of SARS-CoV-2 and MERS-CoV 3CLpros used in this study is 50.83%, but the overall structures of 3CLpros are highly conserved among various coronaviruses with a unique primary substrate specificity for a P1 Gln residue.Molecular docking studies based on our previous co-crystallographic structures of SARS-CoV-2 and MERS-CoV 3CLpros bound with series of compounds related to 11c (41) showed 11c adopts a similar binding mode to SARS-CoV-2 and MERS-CoV 3CLpro with minor differences (Fig. 1), which may explain the similar IC 50 values against both coronavirus 3CLpro (Table 1).
Transgenic mice expressing hACE2 and non-transgenic mouse models for SARS-CoV-2 infection have widely been used to study SARS-CoV-2 infection and pathogenesis and evaluation of countermeasures against SARS-CoV-2 infection (53)(54)(55).SARS2-N501Y MA30 carries multiple adapted amino acid changes after being passaged in mice lungs and can infect non-transgenic mice, such as BALB/c, causing fatal infection (56).Transgenic hDPP4-KI mice have been excellent models for MERS-CoV infection with respiratory signs and fatality after virus inoculation (57)(58)(59).Further adaptation of MERS-CoV following serial passages in mice increased the virulence in hDPP4-KI mice (58).We have recently reported 3CLpro inhibitors that are highly effective against MERS-CoV and SARS-CoV-2 in animal models (28,38).In mice infected with MA-MERS-CoV, treatment with 6j starting one or two days after virus infection significantly increased survival and reduced lung viral titers and histopathology (28).In another report, we demonstrated in vivo efficacy of a deuterated derivative of GC376 against SARS-CoV-2 using K18-hACE2 mice (38).In this report, we tested the efficacy of 5d and 11d in mice to demonstrate their broad-spectrum activity against these two important coronavirus infections.While 5d and 11d show similar in vitro potency against both coronaviruses (Table 1), and both compounds significantly increased survival of mice when treatment started 1 day post infection compared to no-treatment, 11d had a greater impact on reducing fatality and body weight changes than 5d.Analyses of lung viral loads and histopathological changes as well as levels of pro-inflammatory cytokines including CXCL-10, IL-6, and TNF-α in infected mice treated with 11d revealed significantly reduced titers, fewer histopathological changes, and reduced proinflammatory cytokine levels compared to no-treatment, which correlate well with the observed improvement in survival.In vivo efficacy is influenced by many factors, including bioavailability, PK, metabolism, and the chemical stability of a compound, which may explain the relatively lower efficacy of 5d in mice.
Because current predominant circulating SARS-CoV-2 are Omicron variants (60), we examined 11d in K18-hACE2 mice infected with SARS-CoV-2 Omicron XBB.1.16.It has been shown that SARS-CoV-2 Omicron variants are known to cause less severe diseases in the mouse model (45,61,62).Because infections of Omicron subvariant XBB.1.16 in K18-hACE2 mice can lead to fatal outcome, we used this combination to evaluate in vivo efficacy of 11d in this study.Inoculation of Omicron XBB.1.16 in K18-hACE2 mice with 2 × 10 4 PFU resulted in up to 60% fatality by 14 dpi with vehicle-treatment.The treatment with 11d showed that similar results (Fig. 5) seen in BALB/c mice infected with MA-SARS-CoV-2 or hDPP4-KI mice infected with MA-MERS-CoV, which further confirmed in vivo potency of 11d in the models.
In summary, we show that the tested 3CLpro inhibitors have anti-coronavirus (SARS-CoV-2 and MERS-CoV) protease and cathepsin B/L activities and inhibitor 11d significantly enhances survival of mice fatally infected with SARS-CoV-2 or MERS-CoV when treatment started 1 day post infection.While there are several reports of 3CLpro inhibitors effective against human coronaviruses in animal models, to the best of our knowledge, this is the first report demonstrating a single compound being highly effective in both fatal animal models of SARS-CoV-2 (both MA strain and Omicron variant) and MERS-CoV treated after viral infections.While we demonstrated in vivo efficacy of compound 11d using the mouse models, clinical efficacy is influenced by many factors including host-specific components, which poses a major challenge in drug development, and further research is needed to establish whether these inhibitors can be effective therapeutics in humans.Nevertheless, our results add to the current arsenal of potential protease inhibitors of SARS-CoV-2 and MERS-CoV.

Study design
The primary objective of this study was to further evaluate the antiviral activity of compounds 5d and 11d (41) in fatal mouse models of SARS-CoV-2 (56) or MERS-CoV infection (57,58).Compounds 5d and 11d were also evaluated for efficacy against several human proteases, including cathepsin L, using lentivirus-based pseudotyped viruses containing SARS-CoV-2 S in cells expressing human angiotensin-converting enzyme-2 (hACE2) or hACE2 plus TMPRSS2.

Activity of 5c/d and 11c/d on SARS-CoV-2 pseudovirus entry assays in cells
To elucidate the potential dual roles of 5c/d and 11c/d in the entry of SARS-CoV-2, we used lentivirus-based pseudotyped viruses expressing coronavirus S proteins (42).In addition to 5c/d and 11c/d, well-known cathepsin L aldehyde inhibitors, including MDL28170 and a trypsin inhibitor, Nafamostat were tested in the virus entry assay.The virus entry assay was performed in 293T cells expressing hACE2 alone or hACE2 plus TMPRSS2 was previously established in our lab (42).Briefly, cells were incubated with serial dilutions of 5c/d, 11c/d, MDL28170, Nafamostat, or DMSO (mock) and immedi ately transduced with pseudotyped virus in the presence of polybrene (10 µL/mL).The dose-dependent inhibition curve for each compound was prepared and the EC 50 values were determined by GraphPad Prism software using a variable slope (GraphPad, La Jolla, CA).

Modeling of the 11c binding mode with SARS-CoV-2 and MERS-CoV 3CLpro
The binding modes of 11c (aldehyde form of 11d) in the 3CLpro active sites of SARS-CoV-2 and MERS-CoV were modeled using previously determined crystal structures (41) of each 3CLpro.The binding mode of 11c in the active sites was modeled using the coordinates of previously determined crystal structures of SARS-CoV-2 3CLpro [PDB 7TQ5 (41)] and MERS-CoV 3CLpro [PDB 5WKK (63)] superimposing the inhibitor in the active sites.Superposed models of 11c were prepared for docking by adding the covalent bond between the inhibitor and the Sγ atom of the active site Cys, specifying His163 (SARS-CoV-2) and His166 (MERS-CoV) as the HIE tautomer with protonation of the Nε atom.The protein preparation wizard in Schrodinger was used to optimize hydrogen bonding and minimize the structure using Schrodinger's OSPL4 energy function (64) and 11c was prepared for docking using LigPrep (64).These models were subsequently used for covalent docking using CovDock, also from Schrodinger (64,65), selecting the "Nucleophilic Addition to a Double Bond" reaction, adding a core constraint to the glutamine surrogate side chain, performing MM-GBSA scoring, and outputting five poses per ligand reaction site.The hydroxyl formed after covalent attachment to Cys145 was oriented toward His41 and the complex was minimized allowing for flexibility within 6 Å of the ligand while constraining the key hydrogen bonds that are typically observed in the P1 and P2 sites.

Post-infection treatment in MA-SARS-CoV-2 infected mice
Compounds 5d and 11d were examined for efficacy using 10-weeks-old female BALB/c mice infected with MA-SARS-CoV-2 (SARS2-N501Y MA30 ) (56).For evaluating body weight and survival rates, animals were divided into three groups (N = 3 for vehicle or N = 5 for compound 5d or 11d) and were lightly anesthetized with ketamine/xylazine prior to infection with 50 µL of 1,000 PFU MA-SARS-CoV-2 via intranasal inoculation.Both compounds were formulated in 10% ethanol and 90% PEG400 and given to mice from 1 (24 h post infection) to 10 days post infection (dpi) at 100 mg/kg/day (once per day) via intraperitoneal administration.Control mice received vehicle.Animals were weighed daily and monitored for 14 days.The same experiment was repeated, and results were combined (thus, total N = 6 for vehicle or N = 10 for compound 5d or 11d).To evaluate virus replication in lungs, 10-weeks-old BALB/c mice were divided into 4 groups (N = 5 for vehicle or 11d at 3 or 5 dpi) and infected with 50 µL of 1 × 10 3 PFU MA-SARS-CoV-2 via intranasal inoculation.Compound 11d was given to mice from 1 (24 h post infection) to 3 or 5 dpi at 100 mg/kg/day (once per day) via intraperitoneal administration.Animals were euthanized when an animal lost 30% of its' initial weight or at 14 dpi, and the lungs were harvested to evaluate for virus titration.One animal in the vehicle group died at 5 dpi.SARS-CoV-2 virus titration was done in Vero E6 cells using a plaque assay, and mRNA levels of viral N gene were quantified by RT-qPCR.
An additional experiment was conducted to evaluate histopathological changes in the lungs of mice treated with vehicle or compound 11d (N = 5).Animals were euthan ized at 5 dpi, and the lungs were fixed with 10% formalin.Hematoxylin and eosin (HE) stained lung tissues were examined by a veterinary pathologist (DKM) using the post-examination method of masking (66).Lung tissues were evaluated for edema (0-4) Mouse HPRT reverse: 5′-CTCGAGCAAGTCTTTCAGTCC-3′.

Statistical analysis
Statistical differences between indicated groups for mice body weight changes, viral loads, and histopathological score were analyzed using two-tailed unpaired t test with Welch's correction.Log-rank (Mantel-Cox) test was used for the analysis of survival curves between groups.All statistical analyses were performed in GraphPad Prism 8.0 software (San Diego, CA).P < 0.05 was considered statistically significant.*P < 0.05, **P < 0.01, ***P < 0.001.

2 FIG 3
FIG 3 Lung virus titers of BALB/c or hDPP4-KI mice infected with MA-SARS-CoV-2 or MA-MERS-CoV, respectively.(A) Schematic of experimental design for Fig.
3 and 4. (B) The mice were infected with the MA-SARS-CoV-2 or MA-MERS-CoV at 0 dpi and treated with vehicle or compound 11d starting at 1 dpi.Lungs were collected for virus titration (B, left and C, left) or viral N gene quantitation (B, right and C, right) at 3 and 5 dpi of SARS-CoV-2 or MERS-CoV infection, respectively.Data are presented as mean ± SEM.Asterisks indicate statistical significance (*P < 0.05, **P < 0.01, ***P < 0.001).

All
in vitro studies with live SARS-CoV-2 were performed in biosafety level 3 facilities at the University of Iowa, and the studies with SARS-CoV-2 replicon and lentivirus-based pseudoviruses were performed in biosafety level 2 facilities at Kansas State University under protocols approved by the Institutional Biosafety Committee at the University of Iowa and Kansas State University, respectively, according to the guidelines set by the Biosafety in Microbiological and Biomedical Laboratories, the U.S. Department of Health and Human Services, the U.S. Public Health Service, the U.S. Centers for Disease Control and Prevention, and the National Institutes of Health.In vivo studies were performed in animal biosafety level 3 facilities at the University of Iowa.All mouse experiments were conducted under protocols approved by the Institutional Animal Care and Use Committee at the University of Iowa according to the guidelines set by the Association for the Assessment and Accreditation of Laboratory Animal Care and the U.S. Depart ment of Agriculture.

Compounds IC 50 (µM) a
a Values are averages ± standard deviations.