Identification of potential inhibitors of the main protease from feline infectious peritonitis virus using molecular docking and dynamic simulation approaches

Introduction

Feline infectious peritonitis virus (FIPV) is one of the leading causes of death in naïve cats under three years of age (Thayer et al., 2022). However, some recent reports show older cats more than three years old could get severe FIPV infection, which may lead to the death of the infected animals (Shah et al., 2025). The FIPV infection in cats is usually associated with various clinical signs. Some cats show non-specific clinical signs such as inappetence, lethargy, and jaundice. Some cats develop a wet/effusive form, which is mainly characterized by asities, pale color of the mucus membranes, dyspnea, and pleural effusion (Felten & Hartmann, 2019). The dry form could also seen in some FIPV-infected cats in which the affected animal shows some neurological disorders such as seizer. Some ocular pathology could also seen in the cats infected with the dry form of FIPV, such as uveitis, keratitis, and retinal displacement (Felten & Hartmann, 2019).

Diagnosis of FIPV in cats is very tricky, and there are no gold standard clinical signs, techniques/assays for the diagnosis of FIPV in cats (Thayer et al., 2022). However, if some cats are presented with some effusive forms, testing aspirate of this fluid showed higher sensitivity in the confirmation of FIPV infection compared to other blood tests (Felten & Hartmann, 2019). The clinical pathological examination of various body fluids of the infected animals, including the blood, urine, sera, and effusion fluid, could assist with the FIPV diagnosis. The FIPV-infected cats are most likely to develop some types of anemia, lymphopenia, hyperbilirubinemia, hyperglobulinemia (Riemer et al., 2016). Diagnostic imaging play an important role in the diagnosis of FIPV and in the monitoring of the progression of the disease as well as in the success of the application of some antiviral compounds against FIPV infection in cats (Schreurs et al., 2008).

The histopathology and immunohistochemistry using some biopsy and tissue species from freshly dead cats would be an asset in the diagnosis of the FIPV infection (Stranieri et al., 2020). One of the highly effective approaches for FIPV diagnosis is the detection of the viral antigen in the tissues of infected animals with the immunohistochemistry (IHC), in association with the detection of the viral protein in the macrophages (Stranieri et al., 2020). Detection of antibodies in sera of cats against the feline coronavirus (FCoV) did not support the diagnosis of FIPV in cats due to the shared antigenic ship between those two forms of the virus (Hartmann et al., 2003). Application of the reverse transcription polymerase chain reaction (RT-PCR) using primers designed against various viral proteins, including the M, N, and S, of limited importance in the diagnosis of FIPV except the assay that could detect the mutated version of the FCoV that trigger the FIPV infection in cats (Meli et al., 2004).

There are several challenges associated with the design and development of FIPV drugs, including (1) the complex nature of the FIPV infection in cats, (2) the lack of the gold standard diagnostic assays/markers for the viral infection, (3) the high cost of the newly designed drugs, (4) the large number of animals require to validate the efficacy of some new drugs, the possibility of relapse of the treats cats hampered the design and success of the FIPV drugs. All these factors hampered the success of the newly developed FIPV drugs in cats. There is a high demand for the design, development, and validation of some novel drugs for the treatment of FIPV in cats, particularly in shelters, cats in crowded populations, and cats in the same household or catteries, in addition to immunocompromised cats.

FIPV belongs to the order Mononegavirales, family coronaviruses and genus alphacoronavirus, species alphacoronavirus-1 and subspecies feline coronavirus (FCoV).¹ The viral genome is a single molecule of a positive sense RNA. The FIPV genome has a typical coronavirus genome organization and encodes 11 proteins, including four structural and seven non-structural proteins. The major non-structural proteins encoded by Gene-1 (ORF1a and ORF1b with ribosomal frameshifting in between) reside at the 5′ two-thirds of the genome. Other non-structural proteins of the FCoV are encoded by the three ABC and 7/A/B genes, while the structural proteins (Spike (s), Envelope (E) Membrane (M), and the nucleocapsid (N)). The ORF1A/B is further processed by some virally encoded proteases into 16 non-structural proteins (NSP-1-NSP16) (Brierley, Digard & Inglis, 1989). The FIPV-3c protein is important for virus replication and contributes to viral virulence and tissue tropism (Jaimes & Whittaker, 2018). Although the function of the FIPV-7ab proteins is not fully understood, they might play important roles in viral immune evasion, particularly as an antagonist for the IFN-type I (Dedeurwaerder et al., 2013).

Based on the FIPV-S sequences, the virus is classified into two serotypes. The serotype-I is designated as FIPV, which is the most virulent serotype and causes a lethal infection in cats (Hohdatsu, Okada & Koyama, 1991). Although FIPV infection in cats is not contagious (it does not spread easily among cat populations) until now, the prognosis of the cats infected with the virus is always fatal in most cases (Kennedy, 2020). The absence of vaccines that could protect cats against FIPV infection makes antiviral therapy the only remedy to treat cats from FIPV infection. Several antiviral compounds, particularly nucleoside analogs and interferons, have been tried to inhibit or interrupt the FIPV replication cycle at various stages (Addie et al., 2020; Dickinson et al., 2020; Murphy et al., 2018; Pedersen et al., 2019). This approach includes the application of a single or a combination of different compounds together to ensure the robust inhibition of viral replication (Schmied et al., 2024).

The main protease (Mpro) is a potent enzyme in the life cycle of coronaviruses and plays a pivotal role in viral replication and transcription. It is responsible for proteolytically cleaving the large viral polyproteins conserved sites to release functional proteins required for the formation of the replication-transcription complex. Structurally, the Mpro protein consists of three domains and contains a catalytic dyad composed of His41 and Cys145, which are highly conserved among coronaviruses. Its crucial function in viral replication makes the Mpro an ideal target for antiviral drug development, as inhibition of its activity can effectively block viral replication without interfering with host cellular processes. Therefore, coronaviruses Mpro is one of the most important targets for the design of antiviral therapy for several coronaviruses, including the Middle East respiratory syndrome coronavirus (MERS-CoV) and the severe acute-2 (SARS-CoV-2), the porcine epidemic diarrhea virus (PEDV) as well as the FIPV (Wang et al., 2020).

The main rationale of this study is based on the inhibition of some key viral proteins that play essential roles in viral replication, particularly in the processing of the viral polyprotein, which proved to have strong inhibitory effects on the replication of many coronaviruses, including the FIPV. The major objective of the current study is to use virtual screening of a large number of antiviral compounds that most likely inhibit the FIPV-Mpro expression and, thus, inhibit viral replication.

The main reason for targeting CoVs-Mpro is their essential role in polyprotein processing during viral replications. In the current study, we used the in-silico drug design tools to predict the efficacy of 15 selected antiviral, anti-inflammatory, and immune-modulatory compounds on the inhibition of the FIPV-Mpro enzyme (standard inhibitor-N3) compound and nucleoside precursors/analog such as oxipurinol, favipiravir, pentoxifylline, baricitinib, methotrexate, gemcitabine, galidesivir, ribavirin, 6-Azauridine, GS441524, mizoribine, sofosbuvir, molnupiravir, and tenofovir. We used the crystal structure of the FIPV-Mpro available in the public domain and analyzed the binding sites of each compound to these in the FIPV-Mpro.

Although some other studies used molecular docking approaches to screen and design some other antiviral therapies for FIPV infection in cats, most of these studies have focused on limited compound libraries or lacked comprehensive validation through molecular dynamics simulations. The major novelty of this study is the integrative of the in silico approach, combining pharmacokinetic profiling, molecular docking, MM-GBSA free energy calculations, and extended molecular dynamics simulations to evaluate the inhibitory potential of a broad panel of nucleoside analogs against the FIPV main protease (Mpro). The inclusion of clinically relevant compounds such as GS-441524 and Sofosbuvir, along with rigorous computational validation, make this study more unique than the earlier studies (Theerawatanasirikul et al., 2020).

The research approach in this study will provide more insights into the inhibitory actions of the selected compounds on the FIPV-Mpro. These potential FIPV drugs could be used independently or in combination to inhibit FIPV replication, suppress the severe inflammatory conditions associated with viral infections, and enhance the immune response against the viral infection. It will also pave the way for more research to do more functional characterization of the potential inhibitory effects of these compounds on the FIPV replication in the in vitro and in vivo models. This approach will lead to the production of effective antiviral therapy against FIPV infection in cats.

Materials & Methods

Results

Mpro-N3 redocking validation of the molecular docking process

To validate the further docking results of selected ligands, the FIPV-Mpro receptor protein was re-docked with N3. The redocking’s binding site area was x: −47.488, y: −14.895, and z: −9.396, with site sphere radius of 13.53. The parameter of the validation method is RMSD. The RMSD analysis showed the degree of deviation from experimental ligand docking results to the crystallographic ligand attached with FIPV-Mpro protein at the same binding site. The higher the RMSD value, the greater the deviation, which indicates the higher prediction error of ligand-protein interactions. Conversely, the low RMSD value is attributed to better conformation because the redocking ligand position is closer to the ligand position resulting from crystallography. The redocking results indicated a 1.0683 Å RMSD value from the native ligand with the Mpro receptor (Fig. 1). Therefore, based on the low RMSD value between cocrystallized native and re-docked ligand, it can be said that the method used for redocking in this study is valid and can be used against tested ligands with the same binding site area.

Mpro-Ligands interaction and binding energy

MPro in complex with N3 compound in X-ray crystal structure from RCSB PDB database (PDB ID:5eu8; 1.80 Å resolution) was chosen as the receptor for different ligand docking. The PubChem CID for all nucleoside analogs as ligands is given in Table 1. In this attempt to filter suitable compounds with the best binding to FIPV-Mpro, our molecular docking studies revealed a strong interaction of nearly all the selected compounds with binding energies (ΔG) from −7.4 to −4.8 kcal/mol and CDOCKER interaction energy score for binding affinities from 68 to 14 against the active pocket of FIPV-Mpro (Table 3). Among nucleoside precursors and analogs, nucleoside analog Sofosbuvir topped the binding energy and affinity score (ΔG: −7.38 kcal/mol and CDOCKER interaction energy score: 39.34) and interacted with His41, Cys144, Leu164, Glu165, Leu166, Pro188 residues of FIPV-Mpro active pocket (Fig. 2 and Table 3). The interaction of another top-scoring nucleoside analog, i.e., GS-441524 with FIPV-Mpro, was favored by ΔG of −7.2 kcal/mol, whereas binding affinity score (CDOCKER interaction energy score: 43) and interacting residues of Mpro with GS-441524 is given in Table 3. The interaction pattern of the top two ligands, sofosbuvir, GS-441524, and standard N3, is also given in 2D (Figs. S1–S3). The other top-scoring nucleoside precursors and analogs, except sofosbuvir and GS-441524, are methotrexate, pentoxifylline and molnupiravir, exhibits binding with some common residues of the active pocket of FIPV-Mpro i.e., Val26, His41, Thr47, Ala141, Cys144, His162, His163, Leu164, Glu165, Pro188 and Ser189 (Table 3 and Figs. S4–S6). In contrast the, the best docking pose of standard inhibitor (N3) exhibited a binding affinity to Mpro with low binding energy (ΔG= −7.11 Kcal/mol) (Table 3). However, also the highest -CDOCKER energy and -CDOCKER interaction energy score (positive value) referred to the most favorable binding of ligand to the protein. The -CDOCKER energy and -CDOCKER interaction energy score for the N3 best docking pose is 88.72 and 68.11 (Table 3). The interaction between Mpro and N3 involves key amino acid residues and bonds, including Asn25, His41, Ala141, Cys144 Leu164, Glu165, Leu166, Gly167, Pro188, Met190 shown in 2D image (Fig. 2 and Table 3).

Table 3:

Summary of the interacting patterns of selected nucleoside precursors and analogs with crystal structure of FIPV-Mpro with ligand N3 (PDBID:5EU8).

S.No	Compounds	-CDOCK Energy	CDOCK Int Energy	Binding energy	MM-GBSA	Ligand interacting residues of Mpro
1	*N3	88.72	68.11	−7.11	−90.64	Asn25, His41, Ala141, Cys144 Leu164, Glu165, Leu166, Gly167, Pro188, Met190
2	Sofosbuvir	39.56	39.34	−7.38	−106.82	His41, Cys144, Leu164, Glu165, Leu166, Pro188
3	GS-441524	41.12	43	−7.2	−91.02	His41, Thr47, His162, His163, Leu164, Glu165, Pro188
4	Methotrexate	18.71	24.68	−6.98	−88.76	Thr47, Ala141, Leu164, Glu165, Pro188
5	Pentoxifylline	11.40	39.67	−6.65	−67.42	Thr47, Cys144, His162, Leu164, Pro188, Ser189
6	Molnupiravir	25.78	25.81	−6.60	−63.24	His163, Leu164, Pro188, Ser189
7	Tenofovir	−3.03	29.49	−6.4	−62.66	Thr47, His163, Leu164, Glu165, Lgn187, Pro188
8	6-Azauridine	0.16	26	−6.5	−54.72	Thr47, Cys144, His162, Glu165
9	Mizoribine	−12.90	18.30	−5.83	−70.64	Ser48, His163, Glu165, Pro188
10	Oxipurinol	0.45	14	−5.7	−35.16	His41, Thr47, His163, Leu164, Glu165
11	Galidesivir	−1.87	27.53	−4.59	−46.67	Thr47, Leu164, Glu165, Asp186
12	Favipiravir	8.77	16.91	−5.32	−45.49	Leu164, Asp186, Pro188
13	Barcitinib	−29	28	−5.13	−59.06	Cys144, Leu164, Pro188
14	Ribavirin	−19	22	−5.11	60.24	Thr47, Cys144, His163, Leu164, Pro188
15	Gemcitabine	12	28	−5.86	−63.37	Thr47, His163, Leu164, Pro188

DOI: 10.7717/peerj.19744/table-3

Notes:

*The reference standard inhibitor of FIPV-Mpro activity.

CDOCK Int Energy

CDOCKER Interaction energy

Molecular dynamics simulation

Proteins play a crucial role in various structural and functional processes, including microbial pathogenesis, by mediating receptor-based internalization and replication. Therefore, investigating how ligands such as nucleoside analogs affect the structural stability of FIPV-Mpro is essential to understanding their potential to block viral internalization and replication. Molecular dynamics simulation (MD simulation) is a valuable tool for examining protein structures in silico. In this study, we selected the top-performing nucleoside analogs with the lowest ΔG values from CDOCKER-docking studies such as Sofosbuvir (Ligand 1), GS-441524 (Ligand 2) and N3 (standard inhibitor) for FIPV-Mpro, for 50,000 picosecond (ps) MD simulations. A comprehensive analysis of trajectories, including RMSD, RMSF, RG, number of hydrogen bonds established, and hydrogen bond distance, was measured to explore the stability and molecular interactions of the protein-ligand complexes compared to a standard inhibitor (N3).

RMSD calculations conferring stability of ligand–protein complexes

The RMSD value of < 2.5 Å is considered desirable and indicates similarity to the standard N3-Mpro complex. A lower RMSD value also suggests greater stability of the protein-ligand complexes. In this study, the RMSD of the ligand–protein complexes were significantly less fluctuating for GS-441524 and sofosbuvir as compared N3, during the simulations, indicating high stability of GS-441524 and sofosbuvir with Mpro (Fig. 2). The average RMSDs of the backbone atoms for the GS-441524-Mpro, sofosbuvir-Mpro and N3-Mpro ranged between 1 and 3 nm. Specifically, the average RMSDs of the backbone atoms for the GS-441524–Mpro complexes were 2.2 Å, of two MD runs respectively. The complex of ligand 2- GS-441524 showed stability after the initial 0–12,000 ps, while it showed slight fluctuation after the 8,000–11,000 ps (Fig. 3) with the FIPV-Mpro during MD, which quickly stabilized again to the average RMSD (2.2 Å). The complex of ligand 1-sofosbuvir-Mpro showed stability after the initial 0–17,000 ps, while it showed slight fluctuation after the 17,000–21,000 ps (Fig. 3) with the FIPV-Mpro during MD, which quickly stabilized again to the average RMSD (2.8 Å). Conversely, the standard compound- N3-Mpro complex also showed an average RMSD of 2.7 Å, MD runs, respectively.

Stability and mobility of complexes-RMSF

The predicted RMSF graphs of the C-α atom of all five complexes were plotted against the interacted residues based on the trajectory period of the MD simulation. All the residues in docking complexes have fluctuated around 0.5–3 Å in the simulation time scale. The RMSF values for the amino acids in the sofosbuvir as Ligand-1-Mpro complex were notably lower than those in the N3-Mpro complexes, while the compound GS441524 as Ligand-2-Mpro complexes suggesting near equal residue mobility Mpro protein with N3-inhibitor complex (Fig. 4). The significance of the low RMSF values from the RMSF analysis being an essential tool in the identification of the rigid and flexible sections of the protein structure. This conclusion is further validated after considering their low RG values, which indicated the low flexibility of both standard and inhibitor complexes.

Complex compactness analysis-radius of gyration (RG)

RG is used to determine whether the complexes are stably folded or unfolded during the MD simulation. The average RG value of two MD runs of N3-Mpro was calculated to be around 22.2 Å. Furthermore, the average RG values of sofosbuvir (Ligand-1) with Mpro complex and GS-441524 (Ligand 2) with Mpro were 21.9 Å and respectively, significantly similar to reference N3-Mpro complex which is 22.2 Å (Fig. 5). As a result, it can be observed, that GS-441524-Mpro and sofosbuvir-Mpro complex exhibited relatively similar behavior of compactness and consistent values of RG as compared to the reference standard N3-Mpro. It indicates that these are perfectively superimposed with each other and have high stability.

Hydrogen bonds and its distance-Complex stability

Molecular dynamics simulation had been carried out through 50,000 ps simulation time for all new compounds against Mpro protein. To investigate the stability of the ligand interaction with protein, the measurement of intermolecular hydrogen bond development and its distance between ligand-protein complex were evaluated. The complexes of sofosbuvir (Ligand-1) and N3 (Standard) with Mpro protein protease maintained two hydrogen bonds throughout the entire simulation time. However, the GS-441524 (Ligand-2) was able to maintain three hydrogen bonds throughout the simulation time confers the stability of ligand interaction with Mpro protein (Figs. 6B and 6D). Furthermore, all the three complexes in two runs of MD were able to maintain four hydrogen bonds through most of the conformations of simulation (Figs. 6A and 6F).

On the other hand, the estimation of hydrogen bond distance monitor between the ligands and the interacted residues of the Mpro-protein, the sofosbuvir (Ligand 1) showed four hydrogen bonding interactions during simulation time with the Mpro protein. In the depicted hydrogen bond distance graph of two MD simulation runs for sofosbuvir, residue Cys144, Gly167, THR and Glu165 interaction shown weak hydrogen bonding, on the basis of distance of hydrogen bond interaction of Ligand-1 (sofosbuvir) with Mpro. Mpro Cys144 residue showed average bond distance at 2.9 Å (Fig. 7A). All four bonds shown average bond distance fluctuation 1.9–8.9 Å, during the whole simulation time. While in the second run of MD, the interaction of Ligand 1 and Cys144, Gly167 and Glu165 appeared at about average bond distance of 2.9 Å, 3.1 Å and 3.2 Å and average bond distance fluctuation is 1.9–8 Å, shown less stable interaction between sofosbuvir with Mpro (Fig. 7B).

In contrast, GS-441524 (Ligand 2) exhibited stable binding with Mpro, showed interaction of Glu165 with average hydrogen bond distance of 2.2 Å throughout the simulation time. The other residues like Ser189, Gly167, and Pro188 showed an average hydrogen bond distance between 2.9 Å, while average bond distance fluctuation is 1.9−4.7 Å, conferring stable interaction between GS-441524 and Mpro (Figs. 8A and 8B). The standard reference ligand, N3 compound against Mpro protein during simulation time, showed interaction with Glu165 and Ser189 with an average bond distance of 2.5 Å, conferring stable interaction between standard N3 and Mpro. However, the residue Pro188 and Asn25 also exhibited an average hydrogen bond distance of 2.8 Å (Figs. 9A and 9B).

Discussion

Feline infectious peritonitis (FIP) is a severe disease in cats caused by the feline infectious peritonitis virus (FIPV), a variant of the feline coronavirus (FCoV). Normally causing mild enteritis, FCoV transforms into FIPV, affecting multiple organs in cats (Sherding, 2006). FIPV is an enveloped, positive-sense, single-stranded RNA virus classified within the Coronaviridae family, which also includes viruses like SARS-CoV, MERS-CoV, and SARS-CoV-2. Among humans and animals, coronaviruses exhibit the presence of similar structurally related functional proteins, such as the Mpro, also known as 3CL protease (3CLpro), which plays a crucial role in viral replication (Theerawatanasirikul et al., 2020). In this study, the initial phases of a drug screening for Mpro inhibitors, imply and involve drug-likeness assessment of different chemical compounds with the help of various factors. In this context, various computer-aided approaches, such as Lipinski’s Rule of Five, are commonly applied (Ahmad et al., 2024; Jia et al., 2020).

The LogP values of all the compounds provided in Table 1, suggested that these compounds are likely to be more hydrophilic and may have better solubility in water, which can be beneficial for certain therapeutic applications where high aqueous solubility is required. This low LogP also indicates favorable hydrophobicity, contributing to increased persistence and bioavailability by reducing renal excretion (Alvi et al., 2017).

In drug discovery, an MPSA value of ≤ 140 Ų is often considered an optimal threshold for oral bioavailability (Veber et al., 2002). Drugs with higher MPSA values are less lipid-soluble and distributed less extensively attributed to its less extensive and slow absorption rate, than drugs with lower TPSA values. All the nucleoside analogs and precursors showed MPSA values below the threshold. As a result, these compounds exhibit enhanced bioavailability and improved pharmacological efficiency. Based on these findings, all the selected nucleoside precursors and analogs qualifying the criteria for MPSA. Compounds showing slight increased MPSA values (near threshold ≤ 140 Ų) can be considered for further drug screening evaluations.

In ADMET analysis, aqueous solubility refers to the ability of a compound to dissolve in water, which is crucial for the compound’s absorption, bioavailability, and pharmacokinetic properties. The aqueous solubility of a drug influences how well it can be absorbed in the gastrointestinal tract and how effectively it can be distributed throughout the body. However, the compound mizoribine, 6-Azauridine, and ribavirin shows solubility level 5 (Table 1), indicates compounds are extremely soluble, conferring unsuitable for drug likeliness of these compounds on the basis of solubility parameter of ADMET. On the other hand, the blood brain barrier (BBB) model predicts blood–brain penetration after oral administration. In our investigation, all 14 selected compounds showed low penetration values for BBB. Plasma protein binding of drug molecules can affect the efficiency of a drug, because the bound fraction is temporarily shielded from metabolism. On the other hand, only the unbound fraction exhibits pharmacological effects. Drugs with extreme PPB reflect a low volume of distribution (Vd), long plasma half-lives (T1/2), and may incur lower hepatic and renal clearance (Gurevich, 2013; Roberts, Pea & Lipman, 2013).

On the basis of ADMET model predicted PPB, the model predicted false PPB for all the compounds, suggests that, according to the model, these compounds are not predicted to have significant plasma protein binding (Table 2). This implies that the compounds do not strongly interact with plasma proteins, remain largely in their free form, which could lead to greater bioavailability, easily distribute to tissues and reach its site of action and enhanced therapeutic effects. From intestinal absorption analysis prediction results, compounds with intestinal absorption levels of 0–2 indicate good to moderate absorption and enhanced bioavailability, linked to the pharmacological actions of nucleoside analogs (Table 2). In contrast, inhibitors like N3, methotrexate, mizoribine, ribavirin, and 6-Azauridine, with an absorption level of 3, suggest poor intestinal absorption. However, a compound with less absorption is not directly correlated with its pharmacological potential.

Based on the CYP2D6 values and corresponding predictions, out of 14 compounds, none of these drugs are expected to strongly interact with the CYP2D6 enzyme. Similarly, Mpro-standard inhibitor N3 also did not inhibit the CYP2D6 activity. Therefore, they are likely neither inhibitors nor substrates for CYP2D6, which may reduce the risk of metabolic interference. These results imply that the selected nucleoside analogs are likely to undergo rapid metabolism and excretion, potentially contributing to their potent pharmacological effects and reduced toxicity. CYP2D6 is involved in the metabolism of a wide range of substrates in the liver, and its inhibition by a drug constitutes a majority of cases of drug-drug interaction, mediates around 25% of total metabolism and clearance of the drugs after administration (Ahmad et al., 2024). Furthermore, the toxicity assessment of nucleoside precursor and analogs results concluded that Ames test predicts 6-Azauridine, galidesivir, and tenofovir as mutagenic, while the remaining compounds are non-mutagenic. Additionally, the negative hepatotoxicity and Ame’s test values suggest that most compounds are less toxic and safe from mutagenic effects (Table 2). In conclusion, the drug likeliness and ADMET properties prediction results suggested that the selected nucleoside precursors and analogs may exhibits instant metabolism and excretion which might be attributed to their substantial pharmacological effects as well as low toxicity.

Further, the molecular docking analysis of all selected compounds was conducted to identify potential anti-FIPV agents from a pool of nucleoside precursors and analogs. Using CDOCKER for molecular docking, we evaluated and ranked the binding affinities of promising compounds with the catalytic dyad residues (His41 and Cys144) in the binding pocket of FIPV-Mpro. The binding affinity (-CDOCKER interaction energy) scores of the all the compounds with Mpro were ranged from 68.11 to 28 (Table 3). These compounds were further selected to calculate ΔG through calculate binding energy method. The binding energy scores of all interactions between ligands and the Mpro were ranged between −7.11 to −4.59 kcal/mol (Table 3). Out of 14 selected compounds, top two compounds which exhibited better binding affinity in catalytic dyad pocket, are sofosbuvir (Ligand-1) and GS-441524 (Ligand-2) (Table 3). The binding score and energy of all the compounds in including reference ligand (N3) with FIPV-Mpro is shown in Table 3. The interaction of N3 establishes interaction with Mpro through binding with specific residues Cys-144 and His-41 of catalytic dyad of Mpro protein. Most interestingly, the residues stabilizing the binding of top scoring nucleoside analog (sofosbuvir and GS-441524) against FIPV-Mpro also exhibited interaction with residues of catalytic dyad. Therefore, due to high negative binding energy and interaction pattern, GS-441524 and sofosbuvir were slightly different than that of other top-scoring nucleoside analogs and precursors (methotrexate, pentoxifylline and molnupiravir) suggesting that these potent inhibitors of FIPV-Mpro can be used synergistically to produce enhanced synergistic protective effects against FIPV infection via restricting the proteolysis of polyproteins and thereby restricting the expression of FIPV viral proteins/enzymes. The catalytic dyad residues Cys-144 and His-41, plays crucial role in Mpro enzymatic activity, facilitating the cleavage of viral polyproteins into functional proteins essential for viral replication (Jin et al., 2020). This catalytic dyad is pivotal for the enzymatic activity of Mpro, facilitating the cleavage of viral polyproteins into functional proteins essential for viral replication and assembly (Galasiti Kankanamalage et al., 2018; Jin et al., 2020). After takeover of the host’s transcriptional machinery, coronaviruses makes host cell synthesize two overlapping polyproteins, pp1a and pp1ab, which cleaves through coronavirus-encoded proteases—papain-like protease (PL^pro) and Mpro. The cleavage of pp1a and pp1ab forms 16 nonstructural proteins play roles in viral replication (Lu et al., 2022). Owing to the determining role in the proteolysis of viral polyproteins, the Mpro of the FIPV has been established as the preferred target in combating its virulence. Similarly, we also used FIPV-Mpro as the major target via selected nucleoside analogs for the management of this highly lethal infectious virus of domestic cats.

For the analysis of the stable interaction of ligands with Mpro docking complexes were further investigated through MD simulation parameters. The RMSD was analyzed to assess the variations in the backbone Cα atoms of target proteins resulting from the binding of top-ranking nucleoside precursors and analogs along with their respective reference standards (N3). Ideally, RMSD values would be zero; however, due to statistical uncertainties, it is not possible for a protein to have an RMSD of zero (Ahmad et al., 2024). The RMSD plots of each complex clearly demonstrate that the all compounds–Mpro complexes showing low RMSD values (RMSD < 2.5 Å), confirming the stability of complexes during the simulation period when compared to the standard N3-Mpro complex (Fig. 3). RMSD played a significant role in protein stability (Elengoe, Naser & Hamdan, 2014).

Unlike RMSD, the residual fluctuations in the MD simulation complexes were observed by RMSF (Ashraf et al., 2016). The RMSF is also defined as the standard measure of deviation of a molecule from its initial position (Swetha, Ramaiah & Anbarasu, 2016). The RMSF quantifies the overall structural deviations over time and measures the root mean square fluctuations of individual amino acid residues, providing insight into the dynamic behavior and flexibility of these residues during simulations. This analysis enables the identification of critical residues contributing to conformational flexibility within the protein. In this study, a comprehensive RMSF analysis of the top ligand-protein complexes was performed to evaluate the mobility of residues within the protein’s active site. The fluctuations observed in the RMSF profiles throughout the simulation runs pinpointed regions of the protein undergoing significant motion. Investigation of RMSF showed low or equal RMSF for all the complexes as compared to standard N3-Mpro complex (Fig. 4). In conclusion, it indicated that RMSF of all complexes is significantly similar compared to reference standrad N3-mpro complex residues, exhibiting less fluctuation which is substantial enough to be considered as good stability of complexes (Mathpal et al., 2022).

However, the compactness of the target protein-ligand complex was measured through RG (Abbas et al., 2017). The stability of the FIPV- Mpro with ligands exhibiting the best binding energies was further evaluated using the RG, which measures the protein’s size and compactness (Arnittali, Rissanou & Harmandaris, 2019). The magnitude of RG inversely correlates with protein stability, where a larger RG indicates a less stable, more expanded structure. The RG value GS-441524 -Mpro complex was 0.2−0.4 Å low comparable to that of the N3-Mpro complex. However, the RG value of the sofosbuvir-Mpro complex was similar to N3-Mpro complex over the 50,000 ps simulation, suggesting that the protein is less compact and denser upon complex formation (Fig. 5). In conclusion, the higher RG value attributed to low compactness of ligand-protein complex. This indicates that GS-441524-Mpro and sofosbuvir-Mpro complexes demonstrated significant structural compactness and stability. If the protein likely to maintain a relatively steady value of RG over the MD simulation, it is regarded as s folded and more stable, and if its RG changed over time, it would be considered unfolded (Ghasemi et al., 2016).

Further, the number of hydrogen bond establishments and changes in the distance of the compounds to the amino acid residues contributing to substantial hydrogen bonds were assessed by studying the variation of distance between Ca atoms of the binding cavity and the compounds throughout the MD simulations. Based on the simulation result, the average number of hydrogen bond establishment and hydrogen bond distance between the compounds and the contributing residues was shown by GS-441524-Mpro as compared to standard N3-Mpro, which is below 2.5 Å. During the MD simulations of the compound sofosbuvir (Ligand-1) and the distance between the compound and Mpro, both Ser189 and Val42 amino acid residues showed high fluctuations compared to Standard N3 and GS-441524. While, compound GS-441524 (Ligand-2) and Mpro interacted residues showed less fluctuation for hydrogen bond distance (Fig. 8), as compared to standard N3-Mpro and (Fig. 9) and Ligand-1 sofosbuvir-Mpro interacted residues (Fig. 7). More the fluctuation in hydrogen bond distance attributed to the less effective binding of ligand with residues of active pocket of protein (Rasyid, Purwono & Pranowo, 2020). While, the average distance between the compounds and the hydrogen bond contributing residues should be near or below 2.5 Å which falls within the accepted range (Nada, Elkamhawy & Lee, 2022). The residues stabilizing the binding of compounds exhibiting low hydrogen bond distance particulary the GS-441524 and sofosbuvir against FIPV-Mpro were different than that of top-scoring residues of reference standard N3, suggesting that these potent inhibitors can be used synergistically to produce enhanced synergistic protective effects against FIPV infection via restricting the proteolysis of polyproteins and related expression of viral proteins. Restricting the expression of viral proteins/enzymes, as a combinatorial effect of compounds is reported for SARS-CoV-2 inhibition (Keretsu, Bhujbal & Cho, 2020; Ullrich & Nitsche, 2020). Therefore, the results of findings such as docking analysis of different MD simulation parameters including hydrogen bond distance analysis compared with N-Mpro complex, validate the GS-441524 as FIPV-Mpro inhibitor.

Gilead Sciences (GS)-441524 is a 10-cyano adenosine analog, which is the main plasma metabolite of the more famous antiviral drug remdesivir (Amirian & Levy, 2020). Several cellular studies conducted on GS-441524 indicated an anti-SARS-CoV-2 activity comparable when not higher than remdesivir (Yan & Muller, 2020), with some studies pointing out that GS-441524 would be even more convenient than remdesivir for the COVID-19 therapy (Yan & Muller, 2020), GS-441524 advantages over remdesivir include ease of synthetic preparation, lower hepatic toxicity, as well as oral administration route (not suitable for remdesivir due to its poor liver stability) (Yan & Muller, 2020). However, Sofosbuvir is also a nucleoside analog used to treat HCV infection, which can inhibit the SARS-CoV-1 RNA-dependent RNA polymerase, affecting the viral life cycle (Jacome et al., 2020). Sofosbuvir-based treatment regimens may help to reduce the mortality of patients with SARS-CoV-2 and improve associated complications (Hsu et al., 2022). The Sofosbuvir is a prodrug that is hydrolyzed by liver enzymes after absorption to form the monophosphate uridine analog, which is further phosphorylated to form the active triphosphate form (Alrehaily et al., 2023).

In our docking study, GS-441524 interaction could inhibit its binding with polyproteins, pp1a and pp1ab. Our docking and molecular dynamic simulation study parameters confirm that developing a single antiviral agent targeting Mpro, suggesting GS-441524 is either used as a drug alone in treatment or in combination with other potential therapies. This could serve as a potential system that could pave the way to defense against diseases associated with animal coronaviruses. Additionally, our findings provide valuable insights for advancing the development of antiviral agents and broaden the range of anti-CoV agents available for combating feline coronavirus and other related coronaviruses.

The current study shows the promising results of using a newly designed drug targeting the FIPV-Mpro (GS-441524) as well as the repurposing of the Sofosbuvir (FDA-approved drug for the hepatitis-C virus (Summers, 2018). This study supports the strategy of the design and repurposing of some antiviral drugs for one of the most serious viral diseases of cats that cause high mortality rates among the affected cats. The outcomes of this strategy will improve the health care of cats and improve FIPV management in crowded cat populations as shelters, as well as increase the cat survival rate due to FIPV infection in the future. This approach will be an ideal example of the success of the design/repurposing of some drugs for some viral diseases in cats.

The in silico findings presented here offer valuable insights into the potential activity of the studied compounds, the absence of experimental validation represents a key limitation. The computational predictions, while useful for hypothesis generation, must be supported by in vitro and in vivo studies to confirm their biological and therapeutic relevance. Future work will focus on experimentally validating these predictions to establish their translational significance.

Conclusions

The results from this study support the development of a single antiviral agent GS-441524, Mpro targeting agent, and its combination with other potential therapies could provide an effective first line of defense against diseases associated with feline coronaviruses. Nevertheless, this is an initial in-silico study and should be validated through in-vitro and in-vivo experiments in infection models. Additionally, the predicted binding affinity profiles, stability analysis ADME, and toxicity of the selected nucleoside precursors and analogs based on algorithm-based tools and approaches may vary in experimental settings. While the in-silico findings are promising, experimental validation through in vitro and in vivo studies remains essential to confirm their biological and therapeutic relevance. Further, our findings offer essential insights for the further development of antiviral agents and expand the reservoir of anti-FIPV agents for targeting feline coronavirus and other related.

Supplemental Information