Broad-Spectrum Inhibitors against 3C-Like Proteases of Feline Coronaviruses and Feline Caliciviruses

Compounds.

To identify potential broad-spectrum inhibitors against FCoV and FCV, the 3CLpro inhibitor libraries generated by our group were evaluated. The synthesis of dipeptidyl compounds GC373, GC376, GC543, GC546, GC551, and GC554 (22, 24, 25) and tripeptidyl compounds NPI52 (compound 2), NPI59 (compound 6), NPI64 (compound 7), and NPI71 (compound 8) (23) was described previously (22–25). Compounds NPI58, NPI65, and NPI66 were synthesized by modification of a previously reported method (23) and have not been previously reported. Compound confirmation and purity assessment were performed by nuclear magnetic resonance, mass spectrometry, and high-pressure liquid chromatography in the laboratory of D. H. Hua (Department of Chemistry, Kansas State University) or W. C. Groutas (Department of Chemistry, Wichita State University). The structures of the compounds are shown in Fig. 1A and B.

Cells and viruses.

Crandell-Rees feline kidney (CRFK) cells were maintained in minimum essential medium (MEM) containing 2 to 5% fetal bovine serum and the antibiotics chlortetracycline (25 μg/ml), penicillin (250 U/ml), and streptomycin (250 μg/ml). FCoV WSU-79-1146, non-vs-FCV strains Urbana, 131, and F9, and vs-FCV strains 5, Ari, Deuce, and Jengo were propagated in CRFK cells. CRFK cells and WSU-79-1146 were obtained from ATCC (Manassas, VA). The FCVs were kind gifts from J. Parker at Cornell University. WSU-79-1146 is a cell culture-adapted group II FCoV strain which is reported to cause FIP in experimentally inoculated cats (26).

Antiviral effects of compounds in cell culture.

Serial dilutions of each compound were added to confluent monolayers of CRFK cells in 24-well plates or cells were mock treated, and the cells were then immediately inoculated with virus at a multiplicity of infection (MOI) of 0.05 to 0.1. The cells were then further incubated at 37°C until an extensive cytopathic effect was observed in the mock-treated (untreated) well (up to 24 h). After freezing and thawing of the viruses in cell culture, virus titers were determined by the 50% tissue culture infective dose (TCID₅₀) method (27). Stock solutions of test compounds (10 mM) were prepared in dimethyl sulfoxide (DMSO), and the concentration of DMSO in the cell culture did not exceed 0.5%. The 50% effective concentrations (EC₅₀) were determined by nonlinear regression analyses of dose-response curves of virus titers against log inhibitor concentrations (variable slope) using GraphPad Prism software (GraphPad Software, San Diego, CA).

Nonspecific cytotoxic effect.

CRFK cells in 96-well plates were incubated with each compound at various concentrations up to 150 μM for 24 h. Cell cytotoxicity was measured by use of a CytoTox96 nonradioactive cytotoxicity assay kit (Promega, Madison, WI) following the manufacturer's instructions. The 50% cytotoxic concentration (CC₅₀) of each compound was determined using GraphPad Prism software.

Western blot analysis.

CRFK cells were mock treated or treated with each compound and immediately infected with FCoV WSU-79-1146 or FCV Urbana at an MOI of 2. The cells were then further incubated at 37°C for 12 h. At 12 h postinfection, the cells were lysed with SDS-PAGE sample buffer containing 1% β-mercaptoethanol and the proteins were resolved on 10% Novex Tris-bis gels (Invitrogen, Carlsbad, CA) and transferred to nitrocellulose membranes. Viral proteins were probed by using an antibody specific for FCV VP1 (28) or the FCoV nucleocapsid protein (Biocompare, Windham, NH) and then with peroxidase-conjugated, goat anti-mouse IgG or rabbit anti-goat IgG. β-Actin was used as a loading control. Following incubation with a chemiluminescent substrate (Pierce Biotechnology, Rockford, IL), the chemiluminescent signals were detected using a Fotodyne transilluminator/digital camera system (FX; Fotodyne, Hartland, WI).

Multiple-sequence alignment and three-dimensional structural models for 3CLpro.

Multiple-amino-acid-sequence alignment of the 3CLpro enzymes from FCV Urbana (GenBank accession number L40021.1), vs-FCV strains Jango (GenBank accession number DQ910793.1), Ari (GenBank accession number DQ910794.1), and Deuce (GenBank accession number DQ910789.1), FCoV strains WSU-79-1146 (GenBank accession number DQ010921.1), Black (GenBank accession number EU186072.1), and DF-2 (GenBank accession number JQ408981.1), and MHV A59 (GenBank accession number NC_001846.1) was performed using the ClustalW multiple-sequence-alignment program. FCoV strains WSU-79-1146, Black, and DF-2 are FIP-causing FCoVs. The three-dimensional structure of FCoV 3CLpro was built by use of the EasyModeller (version 4.0) program (29) and the 3CLpro structure of transmissible gastroenteritis virus (TGEV), a porcine coronavirus (Protein Data Bank accession number 2AMP), as the template. The FCV 3CLpro three-dimensional structure was built by use of the EasyModeller program and rhinovirus 3Cpro, poliovirus 3Cpro, human norovirus 3CLpro, and hepatitis A virus 3Cpro (Protein Data Bank accession numbers 1CQQ, 1L1N, 2LNC, and 1QA7, respectively) (30) as the templates. The quality of the models was assessed using the Verify 3D program (31).

Animal experiments.

The animal study was performed in accordance with a protocol approved by the Institutional Animal Care and Use Committee (IACUC) at Kansas State University. BALB/c mice were purchased from Charles River Labs (Wilmington, MA). Prior to the animal experiments, the EC₅₀ of GC376 and NPI52 against MHV A59 were determined to be 0.2 to 1 μM in CCL9.1 mouse liver cells. To confirm that MHV A59 infection induces consistent and high levels of virus replication in the livers of the infected mice, we inoculated 4- to 5-week-old female BALB/c mice intraperitoneally with MHV A59 at 7.2 × 10⁴ or 5.2 × 10⁵ TCID₅₀/mouse. At 2 and 4 days postinfection (p.i.), the mice were sacrificed (4 to 6 mice/group), and the livers were collected and processed for virus titration by the TCID₅₀ method. For the in vivo efficacy study, 4- to 5-week-old female BALB/c mice were inoculated intraperitoneally with MHV A59 at 7.2 × 10⁴ or 5.2 × 10⁵ TCID₅₀/mouse. Mice were intraperitoneally given 50 μl of drug vehicle (10% ethanol, 70% polyethylene glycol 400, 20% phosphate-buffered saline), GC376 (10, 50, or 100 mg/kg of body weight/day), or NPI52 (10 or 100 mg/kg/day) divided into two doses per day. Compound administration started 4 h prior to virus infection and continued daily until the mice were euthanized. At 2 and 4 days p.i., the mice were sacrificed and the livers were collected and processed for virus titration. Virus titers were determined by the TCID₅₀ method, and the liver virus titers were compared by two-tailed Student's t test. Fold changes in the geometric mean liver virus titers in each group were calculated by dividing the virus titers in the control group by those in the treated group.

Liver histopathology.

At 4 days p.i., the left lateral lobes were collected from NPI52-treated (10 and 100 mg/kg/day) mice, fixed with formalin, embedded in paraffin, sectioned, and stained with hematoxylin and eosin for histopathological examination by a board-certified pathologist. Five views were examined per mouse liver, and a score of from 0 to 5 was assigned to each lesion contained in the view on the basis of the severity of the histopathological changes. Each score in each sample was added to give a final total score, and then the mean of the total score per sample was calculated for each group. The mean number of lesions per sample was also calculated for each group. The mean total score per sample and the mean number of lesions per sample were compared among the different experimental groups using two-tailed Student's t test.

Antiviral effects of dipeptidyl and tripeptidyl compounds on the replication of FCoV and FCV.

We evaluated the activities of dipeptidyl and tripeptidyl compounds with various R1 and R2 side chains against FCoV WSU-79-1146 and FCV Urbana in cell culture (Fig. 1A and B). For dipeptidyl compounds, replacing the R2 isobutyl (Leu side chain) with benzyl (Phe side chain) or cyclohexylmethyl (Cha side chain) on a representative dipeptidyl compound, GC373, increased the anti-FCV activity, while it decreased its potency toward FCoV. GC373 was previously shown to have potent anti-FCoV activity but minimal activity against FCV in cell culture (22). The tripeptidyl compound NPI52 has an additional residue of 1-napththylalanine compared to the structure of GC373 at the P3 position, and its activity against FCoV or FCV has not been previously tested (23). In this study, we found that NPI52 exhibited potent anti-FCoV and anti-FCV activity, with its EC₅₀ being in the nanomolar range (Fig. 1A), which indicates that the presence of the additional residue at the P3 subsite dramatically increased its activity against FCV. When the R2 isobutyl was replaced with a benzyl or cyclohexylmethyl in NPI52, the benzyl substitution decreased the anti-FCoV activity more than the cyclohexylmethyl substitution, but the reduction in anti-FCV activity was similar between benzyl and cyclohexylmethyl substitutions. Replacement of the aldehyde warhead in NPI52 with ketoamides [(C=O)(C=O)NHCH(CH₃)₂ or (C=O)(C=O)NHC(CH₃)₃] greatly decreased the anti-FCV activity, but the activity of these compounds against FCoV was only moderately decreased. Similarly, replacement of aldehyde with α-hydroxy phosphonate [CH(OH)P(O)(OCH₂CH₃)₂] greatly decreased the anti-FCV activity but had only a minor effect on anti-FCoV activity. NPI64, GC376, GC551, and GC554 are bisulfite adducts of NPI52, GC373, GC543, and GC546, respectively, and they showed antiviral activities against FCoV and FCV comparable to those of their aldehyde counterparts in cell culture. The CC₅₀ values of all compounds ranged from 21.96 μM to greater than 150 μM in CRFK cells (Fig. 1A and B). Western blot analysis confirmed the effects of our compounds on the expression of the FCoV nucleocapsid protein or FCV VP1 (Fig. 2).

Compound NPI52, which possesses potent antiviral activities against both FCoV WSU-79-1146 and FCV Urbana, was also tested for its activity against other non-vs-FCV and vs-FCV strains in cell culture to determine whether this compound is effective against various FCV strains. The EC₅₀ in Table 1 show that NPI52 potently inhibited the replication of various non-vs-FCV and vs-FCV strains in cell culture.

Multiple-sequence alignment and three-dimensional structural models for 3CLpro.

The amino acid sequences of the 3CLpro enzymes have high sequence homology of >95% within strains of FCV or FCoV (Fig. 3A and C). However, there are substantial differences in the 3CLpro sequences (19.72% homology) between FCV and FCoV strains. Although the sequence homology between FCV 3CLpro and FCoV 3CLpro is low, the catalytic residues are well conserved (Fig. 3A to D). MHV A59 3CLpro shares an amino acid sequence homology of 47.35% with FCoV 3CLpro, and the locations of the catalytic residues (red arrowheads in Fig. 3C) correspond well to those of FCoV strains. The residues in the catalytic dyad or triad are shown in the blue boxes in Fig. 3B and D.

In vivo efficacy of compounds in coronavirus-infected mice.

Intraperitoneal inoculation of MHV A59 at 7.2 × 10⁴ or 5.2 × 10⁵ TCID₅₀ per mouse led to high levels of virus replication in the liver, and the levels of virus replication were not significantly different between the two virus inocula, as determined by two-tailed Student's t test (P < 0.05) (data not shown). In the in vivo efficacy study, NPI52 and GC376 were tested in mice infected with MHV A59. In two separate experiments where mice were treated with GC376 or mock treated, the liver virus titers in mice treated with GC376 at 50 or 100 mg/kg were significantly lower than those in the no-treatment control mice at 4 days p.i. but not at 2 days p.i. (P < 0.01) (Fig. 4A and B). The fold reductions of the geometric mean virus titers in mice that received GC376 at 50 or 100 mg/kg at 4 days p.i. were 9.86 and 21.99, respectively, compared to the titers in the control mice (Fig. 4A and B, bar graphs). In contrast, GC376 at 10 mg/kg did not consistently lead to a significant reduction in virus titers at the two time points.

In two independent NPI52 treatment experiments, treatment with NPI52 at 100 mg/kg resulted in a significant reduction in liver virus titers at 4 days p.i. (fold reductions, 19.63 to 40.27) and at 2 days p.i. (fold reductions, 3.46 to 12.3) compared to the titers in the controls (Fig. 4C and D). However, NPI52 at 10 mg/kg failed to significantly reduce the virus titers compared to those in the controls at 2 days p.i. or 4 days p.i. (Fig. 4C and D), although a significant reduction in viral titer was observed at 4 days p.i. in one of the experiments (Fig. 4C and D). The mock-infected mice did not show any signs of illness during the duration of the experiments, and no gross pathological lesions were observed on necropsy.

Histopathology of liver.

Panels 0 through 5 in Fig. 5A show microscopic lesions for increasing histopathology severity scores of from 0 through 5, respectively. On the basis of the lesion scoring, the group treated with NPI52 at 100 mg/kg had a significantly lower mean number of lesions and a significantly lower mean total histopathology scores per mouse liver than the controls (Fig. 5B and D). There was no statistically significant difference in the mean total histopathology scores and the mean number of lesions between the control group and the group treated with NPI52 at 10 mg/kg. However, the lesions in all drug-treated groups were scored 3 or lower, which is in contrast to the presence of lesions scored 4 or 5 in the no-treatment control group (Fig. 5C). Of note, the liver section from a mouse in the group treated with NPI52 at 10 mg/kg did not contain any histopathology lesion, and the data for that liver section were excluded from the statistical analysis for Fig. 5B to D. Examination of the liver samples from mock-infected mice revealed no significant microscopic lesions associated with compound toxicity.