Computational Evaluation of Philippine Vitex negundo Phytochemicals as Potential Inhibitors of Rhinovirus 3C Protease: Molecular Docking, Pharmacokinetic Analysis, and ADMET Studies ^†

Abstract

Human rhinoviruses (HRVs) are the primary cause of the common cold, a highly contagious upper respiratory tract infection characterized by nasal congestion, sneezing, and sore throat. HRV replication depends on its 3C protease (HRV-3Cpro), a key enzyme that cleaves the viral polyprotein into functional proteins essential for viral maturation. Currently, no FDA-approved inhibitors specifically target HRV-3Cpro. While rupintrivir, a synthetic inhibitor, advanced to clinical trials, it ultimately failed due to limited efficacy. This study investigated the potential of Vitex negundo (or lagundi)—a medicinal plant traditionally used in the Philippines for treating colds and respiratory ailments—as a source of natural HRV-3Cpro inhibitors through in silico molecular docking and pharmacokinetic (ADMET) evaluation. Fifteen phytochemicals were screened, with five compounds exhibiting strong binding affinities exceeding that of the reference inhibitor rupintrivir (−6.1 kcal/mol): agnuside (−6.9 kcal/mol), luteolin 7-O-glucoside (−6.7 kcal/mol), 2′-p-hydroxybenzoyl mussaenosidic acid (−6.5 kcal/mol), 6′-(p-hydroxybenzoyl) mussaenosidic acid (−6.5 kcal/mol), and luteolin (−6.2 kcal/mol). Among these, luteolin emerged as a particularly promising lead compound, forming stable hydrogen bonding and hydrophobic interactions with HRV-3Cpro. Luteolin also demonstrates a favorable ADMET and safety profile, predicted to be non-mutagenic and non-hepatotoxic. These findings position luteolin as a potential plant-based HRV-3Cpro inhibitor, warranting further in vitro and in vivo studies to validate its antiviral efficacy and pharmacokinetic properties.

1. Introduction

Viruses are well-established causative agents of respiratory infections, with human rhinoviruses (HRVs), members of the Picornaviridae family, accounting for the majority of common cold cases worldwide [1,2]. HRVs are the predominant infectious agents responsible for upper respiratory tract infections across all age groups and are increasingly linked to severe respiratory infections, particularly among children, older adults, and immunocompromised individuals [3,4,5]. Although HRV infections are generally self-limiting, their high transmissibility and frequency result in significant socio-economic and public health burdens [6].

The replication of HRVs relies on the viral 3C protease (HRV-3Cpro), a highly conserved enzyme responsible for the proteolytic cleavage of the viral polyprotein into functional units essential for viral maturation and replication [1]. Beyond its catalytic role, HRV-3Cpro also facilitates immune evasion by interfering with host antiviral signaling pathways [3,7]. These critical functions make HRV-3Cpro a key determinant of viral pathogenicity and an attractive molecular target for the development of antiviral therapies.

To date, there are currently no FDA-approved antiviral drugs specifically targeting HRV or other picornavirus infections [1,8]. Management of the common cold, therefore, remains largely supportive and symptom-based. Treatment focuses on relieving cough, fever, headache, nasal congestion, rhinorrhea, muscle aches, and sore throat rather than eliminating the causative virus [2,6]. Over-the-counter (OTC) medications such as non-steroidal anti-inflammatory drugs (NSAIDs) and analgesics, including ibuprofen and paracetamol, and decongestants like phenylephrine, are commonly used to alleviate symptoms but do not address the underlying viral infection. Moreover, prolonged or inappropriate use of these medications may lead to adverse effects such as skin rashes, gastrointestinal irritation, and hepatotoxicity [9].

Several antiviral candidates targeting HRV-3Cpro have been explored, including rupintrivir, a synthetic protease inhibitor developed by Pfizer. Rupintrivir demonstrated potent inhibition of HRV replication in experimental settings by binding to and inhibiting HRV-3Cpro; however, it failed to show sufficient clinical efficacy in naturally acquired infections, leading to the discontinuation of its development [1,10]. Thus, due to this limitation, the identification of novel compounds capable of effectively treating HRV infection remains warranted.

Medicinal plants have emerged as promising sources of novel antiviral agents. This is largely due to their rich content of phytochemicals, including bioactive compounds synthesized by plants for metabolic and defensive functions. These compounds have attracted increasing scientific interest as potential antiviral agents due to their diverse medicinal properties, relatively low toxicity, and minimal tendency to cause resistance. Numerous studies have demonstrated that phytochemicals can inhibit viral replication and transcription by interfering with viral protein synthesis, enzyme function, or other stages of the viral life cycle [11,12]. These findings support further research into plant-derived compounds for developing antiviral therapeutics.

Vitex negundo L., locally known in the Philippines as lagundi, is an officially endorsed herbal medicine approved by the Philippine Department of Health (DOH) for its proven therapeutic value. It has long been utilized in traditional medicine and is now widely recognized in evidence-based practice for the management of respiratory conditions, particularly coughs, colds, asthma, and other inflammatory airway disorders. In fact, standardized extracts of lagundi leaves are incorporated into several OTC formulations, typically in the form of tablets or syrups, for the relief of respiratory symptoms [13].

Several phytochemicals, including iridoids, phenolics, terpenoids, and flavonoids, have been identified in Philippine V. negundo leaves [14,15,16]. However, their potential antiviral activity against HRV-3Cpro has not been studied to date. To address this gap, we employed computer-aided drug discovery (CADD) approaches in this study to evaluate fifteen selected V. negundo phytochemicals as potential HRV-3Cpro inhibitors. Molecular docking and ADMET analyses were conducted to predict binding interactions, inhibitory potential, and pharmacokinetic profiles, with the aim of identifying promising plant-derived antiviral candidates.

2. Materials and Methods

2.1. Preparation of Target Protein

The three-dimensional crystal structure of human rhinovirus 3C protease (HRV-3Cpro) (PDB ID: 2XYA) in complex with the co-crystallized ligand 2-phenylquinolin-4-ol (Figure 1) was downloaded in PDB file format from the Research Collaboratory for Structural Bioinformatics Protein Data Bank (PDB) (https://www.rcsb.org, accessed on 7 May 2025). The structure has a resolution of 2.40 Å [8]. The protein structure was prepared using Discovery Studio Visualizer (DSV) software, version 21.1.0.20298. All bound substances, including the co-crystallized ligand and water molecules, were removed from the structure. AutoDockTools software, version 1.5.6, was then used to further prepare the protein by adding polar hydrogens and Kollman charges [17,18]. Finally, the protein was saved in PDBQT format.

2.2. Preparation of Ligands

A ligand library consisting of 15 phytochemicals (Figure 2) previously identified and isolated from Vitex negundo leaves was compiled based on phytochemical composition studies conducted in the Philippines [14,15,16]. When available, their 3D structures in SDF format were downloaded from the PubChem database (https://pubchem.ncbi.nlm.nih.gov, accessed on 7 May 2025); otherwise, their 2D counterparts were used, with compounds not yet added to the database manually drawn instead using BIOVIA Draw 2021, version 21.1. All the structures were then converted to PDB format using DSV. The structures were then optimized using Avogadro software, version 1.2.0, applying the universal force field (UFF) and steepest descent algorithm [19,20]. The ligands were further prepared using AutoDockTools, with the torsional degrees of freedom (TORSDOF) kept at default values [18]. Finally, the ligands were saved in PDBQT format.

2.3. Molecular Docking

AutoDock Vina software, version 1.1.2, was used for the rigid protein–flexible ligand molecular docking simulations [21]. The search space was confined to a grid box encompassing the binding site of the co-crystallized ligand, 2-phenylquinolin-4-ol. The grid box (size: x = 14, y = 20, z = 20; spacing: 1 Å) was centered on the co-crystallized ligand’s coordinates (x = 8.811, y = 28.742, z = 20.990) [8]. Docking simulations for each ligand were performed in triplicate to ensure reproducibility. The resulting binding affinity values of the 15 V. negundo phytochemical ligands were recorded and compared with those of the reference HRV-3Cpro inhibitor, rupintrivir. For the top five ligands, ligand efficiency (LE) values were calculated using the formula: LE = −ΔG/N_HA. Here, ΔG is the mean binding affinity, and N_HA is the number of heavy atoms in each ligand. Protein–ligand interactions were subsequently visualized and analyzed using Discovery Studio Visualizer.

2.4. Drug-Likeness and ADMET Predictions

The drug-likeness and pharmacokinetic properties of the top V. negundo ligands were predicted from their SMILES strings, including absorption, distribution, metabolism, excretion, and toxicity (ADMET). These predictions were carried out using the web-based in silico tools SwissADME (http://www.swissadme.ch, accessed on 8 May 2025) and pkCSM (https://biosig.lab.uq.edu.au/pkcsm/prediction, accessed on 8 May 2025) [22,23].

3. Results

3.1. Validation of Docking Protocol

Docking validation is an important step in molecular docking studies, as re-docking the co-crystallized ligand ensures that the docking protocol can accurately reproduce the experimental binding conformation prior to reliable screening of a ligand library [24]. AutoDock Vina’s accuracy was evaluated by re-docking 2-phenylquinolin-4-ol into the active site of human rhinovirus 3C protease (HRV-3Cpro) and comparing its orientation with the co-crystallized structure using root-mean-square deviation (RMSD). The RMSD value obtained for the superimposed structures (Figure 3) was 0.5781 Å, which is below the accepted threshold of 2.0 Å [18,25]. This confirms the reliability and validity of AutoDock Vina for docking the selected phytochemicals into the active site of HRV-3Cpro.

3.2. Molecular Docking Results

In this study, molecular docking was employed to evaluate the binding affinities of fifteen Philippine Vitex negundo phytochemicals against HRV-3Cpro.

Table 1. Binding affinities of the 15 V. negundo phytochemicals against HRV-3Cpro.

Ligands	PubChem CID	Phytochemical Class	Binding Affinity (kcal/mol)
Agnuside	442416	Iridoid	−6.9 ± 0.0
Casticin	5315263	Flavonoid	−5.5 ± 0.0
Chrysoplenol D	5280699	Flavonoid	−5.9 ± 0.1
Isoorientin	114776	Flavonoid	−5.9 ± 0.0
Kaempferol 3-O-glucuronide	5318759	Flavonoid	−5.6 ± 0.1
Lagundinin	-	Iridoid	−4.5 ± 0.0
Luteolin	5280445	Flavonoid	−6.2 ± 0.0
Luteolin 4′-O-glucoside	5319116	Flavonoid	−5.8 ± 0.1
Luteolin 7-O-glucoside	5280637	Flavonoid	−6.7 ± 0.0
Negundoside	9935561	Iridoid	−5.7 ± 0.1
Protocatechuic acid	72	Phenolic	−4.7 ± 0.1
Vitexilactone	21636178	Terpenoid	−5.0 ± 0.0
2′-p-Hydroxybenzoyl mussaenosidic acid	73298898	Iridoid	−6.5 ± 0.1
6′-(p-Hydroxybenzoyl)mussaenosidic acid	133554308	Iridoid	−6.5 ± 0.0
(1r,3R,4s,5S)-4-{[(2E)-3-(3,4-Dihydroxyphenyl)-2-propenoyl]oxy}-1,3,5-trihydroxycyclohexanecarboxylic acid	5315600	Phenolic	−5.6 ± 0.1
Rupintrivir *	6440352	-	−6.1 ± 0.1

* Reference inhibitor. Values are means ± SD (n = 3).

summarizes the predicted binding affinities, which ranged from −4.5 to −6.9 kcal/mol. The reference HRV-3Cpro inhibitor, rupintrivir, exhibited a binding affinity of −6.1 kcal/mol. The top five phytochemicals, agnuside (−6.9 kcal/mol), luteolin 7-O-glucoside (−6.7 kcal/mol), 2′-p-hydroxybenzoyl mussaenosidic acid (−6.5 kcal/mol), 6′-(p-hydroxybenzoyl)mussaenosidic acid (−6.5 kcal/mol), and luteolin (−6.2 kcal/mol) showed more favorable (more negative) binding affinities than rupintrivir. Ligand efficiency (LE) values of these top five ligands are presented in

Table 2. Ligand efficiency values of the top 5 ligands against HRV-3Cpro.

Ligands	Ligand Efficiency (LE) †
Agnuside	0.21
Luteolin	0.30
Luteolin 7-O-glucoside	0.21
2′-p-Hydroxybenzoyl mussaenosidic acid	0.19
6′-(p-Hydroxybenzoyl)mussaenosidic acid	0.19
Rupintrivir *	0.14

* Reference inhibitor. ^† In kcal/mol per heavy atom.

, with luteolin exhibiting the highest LE of 0.30 kcal/mol per heavy atom, agnuside and luteolin 7-O-glucoside both exhibiting 0.21, 2′-p-hydroxybenzoyl mussaenosidic acid and 6′-(p-hydroxybenzoyl)mussaenosidic acid each exhibiting 0.19, and rupintrivir displaying an LE of 0.14. Consequently, further analyses and interaction studies were focused on these top five ligands.

3.3. Protein–Ligand Interactions of the Top Ligands

Figure 4 illustrates the two-dimensional protein–ligand interaction profiles of the top five V. negundo phytochemicals and the reference inhibitor, rupintrivir, within the active site of HRV-3Cpro. The key intermolecular interactions formed by each ligand with HRV-3Cpro are systematically summarized in

Table 3. Summary of the types of interactions formed between the top five ligands and HRV-3Cpro.

Ligands	Interacting Amino Acid Residues
	Hydrogen Bonds	Hydrophobic Interactions	Other Interactions
Agnuside	Gly 164	Phe 170	None
Luteolin 7-O-glucoside	Thr 142, His 161, Gly 164, Asn 165	Phe 170	None
2′-p-Hydroxybenzoyl mussaenosidic acid	Gly 164, Val 162	Cys 147, Phe 170	None
6′-(p-Hydroxybenzoyl)mussaenosidic acid	Gly 164, Ser 144	Phe 170	None
Luteolin	Thr 142, His 161	Lys 143, Gly 163, Gly 164, Asn 165,	Cys 147
Rupintrivir *	Gly 164, Gly 166, Ser 144, Val 162	Gly 163, Phe 170	Thr 142

* Reference inhibitor.

, where they are categorized into hydrogen bonds and hydrophobic interactions. Among the top five phytochemicals, Gly 164 and Phe 170 were consistently identified as key interacting amino acid residues in the majority of protein–ligand complexes. Gly 164 primarily participated in hydrogen bond formation, while Phe 170 was predominantly involved in hydrophobic interactions. These residues were also present in the binding interactions of the reference inhibitor, rupintrivir.

3.4. Drug-Likeness and ADMET Profiles of the Top Ligands

Table 4. Physicochemical properties, drug-likeness, and ADMET characteristics of the top five ligands.

Property	(A)	(B)	(C)	(D)	(E)	(F)
Molecular Weight (g/mol)	466.44	448.38	496.46	496.46	286.24	598.66
MLogP	−1.20	−2.10	−0.93	−0.93	−0.03	1.51
No. of H-Bond Acceptors	11	11	12	12	6	9
No. of H-Bond Donors	6	7	6	6	4	3
No. of Violations to Lipinski’s Rule	2	2	2	2	0	2
Bioavailability Score	0.17	0.17	0.11	0.11	0.55	0.17
GI Absorption	Low	Low	Low	Low	High	Low
BBB Permeant	No	No	No	No	No	No
P-gp Substrate	No	Yes	Yes	No	No	Yes
CYP3A4 Inhibitor	No	No	No	No	Yes	Yes
CYP1A2 Inhibitor	No	No	No	No	Yes	No
CYP2C19 Inhibitor	No	No	No	No	No	No
CYP2C9 Inhibitor	No	No	No	No	No	No
CYP2D6 Inhibitor	No	No	No	No	Yes	No
Total Clearance (mL/min/kg)	6.081	3.006	6.592	5.420	3.126	7.603
Mutagenicity	No	No	No	No	No	No
Hepatotoxicity	No	No	No	No	No	Yes

(A) Agnuside; (B) Luteolin 7-O-glucoside; (C) 2′-p-Hydroxybenzoyl mussaenosidic acid; (D) 6′-(p-Hydroxybenzoyl)mussaenosidic acid; (E) Luteolin; (F) Rupintrivir.

presents the predicted physicochemical properties, drug-likeness parameters, and ADMET profiles of the top five phytochemicals, with rupintrivir included for comparison. Four of the phytochemicals (agnuside, luteolin 7-O-glucoside, 2′-p-hydroxybenzoyl mussaenosidic acid, and 6′-(p-hydroxybenzoyl)mussaenosidic acid) exhibited two violations of Lipinski’s rule of five, primarily due to the number of hydrogen bond donors and acceptors. Rupintrivir also showed two violations, attributed to its molecular weight and number of hydrogen bond acceptors. Luteolin was the only phytochemical to fully comply with Lipinski’s rule, highlighting its favorable drug-likeness profile relative to the others.

Additionally, most ligands, including rupintrivir, were predicted to have low gastrointestinal absorption, whereas luteolin was the only compound with high GI absorption and a bioavailability score of 0.55. All ligands were predicted to be non-permeant to the blood–brain barrier. Luteolin was also predicted as a non-substrate of P-glycoprotein, positioning it as a standout among the top five phytochemicals. Regarding cytochrome P450 inhibition, luteolin was predicted to inhibit CYP3A4, CYP1A2, and CYP2D6, but not CYP2C19 or CYP2C9. The predicted total clearance values for the phytochemicals ranged from 3.006 to 6.592 mL/min/kg, with luteolin showing a lower clearance (3.126 mL/min/kg) than rupintrivir (7.603 mL/min/kg). In terms of toxicity, luteolin, along with the other phytochemicals, showed no mutagenicity, and hepatotoxicity was not predicted for luteolin, whereas rupintrivir exhibited predicted hepatotoxic potential.

4. Discussion

Molecular docking is an in silico technique used to predict the binding orientation and affinity of small molecules to a target protein. This technique was employed in this study to evaluate the potential interactions of Vitex negundo phytochemicals with human rhinovirus 3C protease (HRV-3Cpro). Notably, three iridoids (a class of monoterpenoids) and two flavonoids—namely, agnuside, 2′-p-hydroxybenzoyl mussaenosidic acid, 6′-(p-hydroxybenzoyl) mussaenosidic acid, luteolin 7-O-glucoside, and luteolin—exhibited stronger binding affinities than the reference inhibitor rupintrivir, suggesting a potentially stronger interaction with, and inhibitory effect on, the viral protease. These observed binding affinities of the top five V. negundo phytochemicals align with the plant’s established use in respiratory health, supporting its traditional and evidence-based applications for coughs, colds, and airway inflammation [13]. These results also provide in silico evidence supporting natural products as potential anti-rhinovirus agents, alongside previously reported computational findings [26,27].

Ligand efficiency (LE) provides a complementary perspective by assessing the binding affinity of a ligand relative to its molecular size. A threshold value of LE ≥ 0.3 kcal/mol per heavy atom is generally considered indicative of efficient binding [28]. Among the top five ligands, only luteolin achieved an LE of 0.30 kcal/mol per heavy atom, surpassing the efficiency of the reference inhibitor rupintrivir (0.14) and demonstrating its ability to form strong binding interactions while optimally utilizing its molecular structure.

The predominance of hydrogen bonding and hydrophobic interactions among the top phytochemical ligands suggests that both polar and nonpolar forces contribute to the formation of stable complexes with HRV-3Cpro [29]. Key amino acid residues Gly 164 and Phe 170 were consistently involved across most ligand interactions, indicating that these compounds engage a similar region of the HRV-3Cpro active site. Agnuside, which exhibited the strongest binding affinity among the top ligands—even surpassing rupintrivir—interacted only with Gly 164 and Phe 170. In contrast, luteolin displayed a unique interaction profile, forming additional contacts with Lys 143, Gly 163, and Cys 147, distinguishing it from the other phytochemicals and suggesting a broader engagement within the binding site. This observation highlights that the number and nature of intermolecular interactions do not always correlate directly with binding strength; rather, a ligand’s efficacy depends largely on the complementarity of its shape and size to the enzyme’s active site, which can inherently enhance binding affinity [30].

In the early stages of drug development, Lipinski’s Rule of Five serves as a widely accepted guideline for evaluating the drug-likeness and potential oral bioavailability of compounds. According to this rule, an orally active compound should generally not violate more than one of the following criteria: fewer than 5 hydrogen bond donors (HBD), fewer than 10 hydrogen bond acceptors (HBA), a molecular weight (MW) under 500 g/mol, and an octanol–water partition coefficient (MlogP) below 5 [31]. Among the top phytochemicals, luteolin was the only compound to fully comply with these criteria, suggesting a higher likelihood of favorable oral bioavailability. Its non-substrate profile for P-glycoprotein indicates that it may avoid active cellular efflux, enhancing systemic retention [32], while its non-permeant profile for the blood–brain barrier, combined with predicted non-mutagenicity and non-hepatotoxicity, highlights a favorable safety profile.

Luteolin was predicted to inhibit CYP3A4, CYP1A2, and CYP2D6, while not affecting CYP2C19 or CYP2C9. These CYP450 enzymes are essential for drug metabolism, and such inhibition indicates that luteolin may interact with the metabolism of other co-administered drugs, which is important to consider in drug development. Nevertheless, these interactions do not diminish its value as a lead compound; rather, they provide useful insights into its metabolic profile. Taken together, luteolin’s full compliance with Lipinski’s rule and favorable ADMET characteristics position it as a particularly promising lead compound for further investigation as a plant-derived antiviral agent.

Despite the promising in silico findings, this study has certain limitations. Molecular docking provides static predictions of binding affinity and orientation but does not account for the dynamic flexibility of proteins and ligands under physiological conditions. We therefore recommend that future investigations use molecular dynamics simulations to provide a more accurate assessment of the stability and conformational behavior of the phytochemical–HRV-3Cpro complexes over time, including luteolin, which has emerged as a particularly promising lead compound. Moreover, while the ADMET predictions offer preliminary insight into pharmacokinetics and safety, experimental validation is necessary. Subsequent in vitro studies should evaluate the antiviral efficacy of luteolin and the other top phytochemicals against HRV replication, and in vivo studies could assess their pharmacokinetic profiles, toxicity, and therapeutic potential in relevant models. Integrating these computational and experimental approaches would strengthen the translational potential of V. negundo phytochemicals, particularly luteolin, as plant-based HRV-3Cpro inhibitors.

5. Conclusions

This study reports for the first time that Philippine Vitex negundo phytochemicals, with a particular focus on luteolin, exhibit strong and stable hydrogen bonding and hydrophobic interactions with HRV-3Cpro, surpassing the reference inhibitor rupintrivir. Importantly, luteolin complies with Lipinski’s Rule of Five, shows good bioavailability, and demonstrates a favorable ADMET and safety profile. These properties position luteolin as a particularly promising lead compound, warranting future in vitro and in vivo studies to validate its antiviral efficacy and pharmacokinetic potential for the development of plant-based HRV-3Cpro inhibitors.