Retrieval of HPV-18 E6 and E7 oncoproteins

Using BLASTP to search for similarities with target proteins found in HPV-18, the findings revealed that the E6 and E7 oncoproteins share similarities with the protein structures represented by PDB IDs 4GIZ and 6IWD, respectively. The protein alignment between E6 and 4GIZ resulted in a total score of 50.4 with 87% query cover and 24.09% identity. Similarly, the protein alignment between E7 and 6IWD showed a query cover, identity, and total score of 37%, 38.46%, and 35, respectively.

Physicochemical properties of HPV-18 E6 and E7 oncoproteins

The HPV-18 E6 oncoprotein is characterized by its length of 157 amino acids and molecular mass of 17920.75 Da. It has a theoretical isoelectric point (pI) value of 5.37 and consists of 16 strong basic (+) amino acids (K, R) and 20 strong acidic (−) amino acids (D, E). Its atomic composition is C804H1228N208O227S15, and it exhibits an instability index of 33.41, indicating its stability as a protein. The average hydrophilicity coefficient GRAVY is 0.053, suggesting it is a hydrophobic protein. The hydrophilic nature is further supported by the analysis of the Kyte and Doolittle hydropathy plot (Fig. 1a). On the other hand, the HPV-18 E7 oncoprotein consists of 103 amino acids with a molecular mass of 11699.37 Da. It has a theoretical pI value of 4.47 and contains 9 strong basic (+) amino acids (K, R) and 21 strong acidic (−) amino acids (D, E). The atomic composition is C510H814N136O162S8, and it exhibits an instability index of 52.66, classifying it as an unstable protein. The average hydrophilicity coefficient GRAVY is − 0.197, indicating it is a hydrophilic protein. This hydrophilic nature is also supported by the Kyte and Doolittle hydropathy plot (Fig. 1b). A comprehensive physiochemical profile of these proteins is presented in Table 1.

Fig. 1

Kyte and Doolittle hydropathy plot: a HPV-18 E6 oncoprotein, b HPV-18 E7 oncoprotein. A positive peak indicates a probability that the corresponding polypeptide fragment is hydrophobic

Table 1 Molecular and physiochemical properties of HPV-18 E6 and E7 oncoproteinsPrediction of secondary structures

To determine the secondary structures of the HPV-18 E6 and E7 oncoproteins, SOPMA was employed. The analysis unveiled that the HPV-18 E6 oncoprotein consisted of an alpha helix content of 57.32%, a random coil content of 26.11%, an extended strand content of 9.55%, and a beta turn content of 7.1%. In contrast, the HPV-18 E7 oncoprotein exhibited a random coil content of 49.51%, an alpha helix content of 32.04%, an extended strand content of 16.50%, and a beta turn content of 1.94%.

Prediction of tertiary structures

Using the MMDB program, the tertiary structures of the HPV-18 E6 and E7 oncoproteins were obtained and are illustrated in Fig. 2. The HPV-18 E6 oncoprotein, with a resolution of 2.55 Å, is composed of two chains, one molecule of alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranose, and two zinc ions. Meanwhile, the HPV-18 E7 oncoprotein, with a resolution of 1.80 Å, has four chains, four phosphate ions, two chloride ions, and two zinc ions.

Fig. 2

Three-dimensional structure of proteins: a HPV-18 E6 oncoprotein (PDB ID: 4GIZ), b HPV-18 E7 oncoprotein (PDB ID: 6IWD). The HPV-18 E6 protein consists of a complex structure, including multiple alpha-D-glucopyranosyl chains and two zinc ions, and the HPV-18 E7 protein is composed of four chains, four phosphate ions, two chloride ions, and two zinc ions

Figure 3 illustrates the connections between zinc ions (Image a in Fig. 3A) and alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranose (Image b in Fig. 3A) in HPV-18 E6 oncoprotein, along with their neighboring amino acids, as well as the associations of phosphate ions (Image a in Fig. 3B) and zinc ion (Image b in Fig. 3B) in HPV-18 E7 with their surrounding amino acids.

Fig. 3

Interactions of the ligands and/or metals in: A HPV-18 E6 oncoprotein with their surrounding amino acids; a zinc ions (Zn), b alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranosyl-(1- > 4)-alpha-D-glucopyranose. B HPV-18 E7 oncoprotein with their surrounding amino acids; a. phosphate ions (PO4), b. zinc ion (Zn)

Validation of predicted tertiary structures

The accuracy of the tertiary structures of the HPV-18 E6 and E7 oncoproteins (Figs. 4A and B, respectively) was assessed and validated using the Ramachandran plot (a), ProSA-web (b and c), and ERRAT (d) servers. The evaluation results, as depicted in image a in Fig. 4A, showed that 92.8% of the amino acids in the HPV-18 E6 oncoprotein were located in the highly favored region of the Ramachandran plot, while 7.2% were in the additional allowed region, with no atypical amino acids found in the generously allowed and disallowed regions. Similarly, in the case of the HPV-18 E7 oncoprotein (image a in Fig. 4B), 92.0% of the residues were situated in the favored region, 7.3% in the additional allowed region, and 0.7% in the generously allowed region.

Fig. 4

Validation of predicted structures: A HPV-18 E6 oncoprotein and B HPV-18 E7 oncoprotein; a Ramachandran plot, b Local model quality at ProSA-web, c Overall model quality at ProSA-web, and d. ERRAT

To further assess the quality of the structures, the ProSA-web algorithm was utilized, and the obtained Z-scores were − 12.05 for HPV-18 E6 (Image b in Fig. 4A) and − 9.09 for HPV-18 E7 (Image b in Fig. 4B). Additionally, the ERRAT server was employed to evaluate the overall quality factor of the generated models. The HPV-18 E6 oncoprotein received a quality factor of 99.6032 (images d in Fig. 4A), while the HPV-18 E7 oncoprotein had a quality factor of 96.8652 (images d in Fig. 4B). These findings indicate that the modeled structures of the HPV-18 E6 and E7 oncoproteins are reliable and can be considered as potential targets for natural photosensitizers. According to the results of SignalP—5.0, HPV-18 E6 and E7 oncoproteins do not have signal peptides cleavage sites (Sec/SPI = 0.011 and Sec/SPI = 0.0014, respectively). So, these oncoproteins are not secreted through Sec-dependent pathways. Future experiments are required to identify these secretion pathways.

Functional classification

The refined models of the HPV-18 E6 and E7 oncoproteins were analyzed using C-I-TASSER, a structure-based method for determining the biological function of protein molecules. The analysis generated a comprehensive list of predicted Gene Ontology (GO) terms for each oncoprotein structure, which can be observed in Fig. 5. According to the analysis, the E6 oncoprotein (Fig. 5a) exhibited significant enrichment in GO terms associated with negative regulation of gene expression and regulation of transcription, DNA-templated, both with a CscoreGO value of 1.00, in the biological process (BP) category. As for the E7 oncoprotein (Fig. 5b), it demonstrated significant enrichment in a GO term related to the modulation of host morphology or physiology by the virus, with a CscoreGO value of 1.00. In terms of molecular function (MF), both E6 and E7 oncoproteins displayed significant enrichment in GO terms related to cation binding and transferase activity, transferring phosphorus-containing groups, with CscoreGO values of 0.59 and 0.44, respectively. Regarding cellular component (CC), the E6 oncoprotein was found to be significantly enriched in GO terms associated with the host cell nucleus, with a CscoreGO value of 1.00 (Fig. 5a). On the other hand, the E7 oncoprotein exhibited significant enrichment in GO terms related to both the host cell nucleus and cell part, both with a CscoreGO value of 1.00 (Fig. 5b).

Fig. 5

Gene Ontology (GO) enrichment analysis. The bar graphs indicate the enriched GO terms of the differentially expressed genes and the numbers of genes corresponding to each GO term. a HPV-18 E6 oncoprotein, b HPV-18 E7 oncoprotein

Molecular dynamics simulation

The results of the normal mode analysis (NMA) performed on the HPV-18 E6 and E7 oncoproteins are presented in Fig. 6A and B, respectively. Image a portrays the outcomes of the NMA, while image b highlights the regions in the proteins that exhibit high deformability. It is observed that the E7 protein (Image b in Figs. 6B) displays greater deformability compared to E6 (Image b in Figs. 6A), with multiple peaks indicating a deformability index of approximately 1.0. The B-factor, which indicates the flexibility of the proteins based on atomic displacement parameters, was calculated using NMA and is depicted in image c in Figs. 6A and B. Image d represents the eigenvalues of the proteins, which relate to the energy required to deform their structures. A lower eigenvalue suggests easier deformation. Specifically, the eigenvalues for HPV-18 E6 and E7 oncoproteins were 1.637941e-05 (Image d in Fig. 6A) and 2.751974e-04 (Image d in Fig. 6B), respectively. Image e showcases the variance plots, illustrating cumulative variances in green and individual variances in violet. In image f, the covariance matrix between pairs of residues is shown, where red indicates good correlation, white represents no correlation, and blue signifies anticorrelation. Furthermore, image g presents the elastic network model, which displays the connections between atom pairs and springs. Dark grey dots indicate stiffer springs, while lighter grey dots represent more flexible ones. The iMOD study conducted on the target proteins suggests that the proposed proteins are stable.

Fig. 6

Molecular dynamic simulation: A HPV-18 E6 oncoprotein, B HPV-18 E7 oncoprotein. a Protein–ligand complex, b Deformability, c. B-factor values, d Eigenvalues, e. Variance (violet: individual variances, green: cumulative variances), f Co-variance map (residues with correlated motions in red, uncorrelated motions in white, and anti-correlated motions in blue), and g. Elastic network (darker grays indicate stiffer springs) of the complex

Molecular docking

Table 2 and Fig. 7 present a summary of the free binding energy values and various interactions of natural flavonoid glycosides with HPV-18 E6 and E7 oncoproteins. The results indicate that Kaempferol exhibited a strong binding affinity to both E6 and E7 oncoproteins, with binding energies of − 9.2 kcal/mol and − 7.9 kcal/mol, respectively. In the case of E6, Kaempferol formed interactions with several amino acid residues, including GLU45, GLU46, PRO49, GLN50, ALA53, ASP66, ARG67, GLY70, TYR71, SER74, LEU76, ILE334, PRO335, SER338, SER80, TYR81, ASN127, ARG129, and GLY130. Similarly, in the case of E7, Kaempferol interacted with ALA975, GLU976, TRP977, HIS978, TYR979, GLN1097, ARG1100, ARG1101, ARG1113, HIS1114, PRO1115, PRO1116, ILE1117, TYR1139, and HIS1143.

Table 2 Properties of molecular docking analysis between the HPV-18 E6 and E7 oncoproteins (target receptors) and natural flavonoid glycosides (ligands) including Fisetin, Kaempferol, Morin, Myricetin, and QuercetinFig. 7

Depiction of docked ligand–protein complex along with the interaction of the amino acid residues of the protein with ligand: a HPV-18 E6 oncoprotein-Fisetin, b HPV-18 E7 oncoprotein-Fisetin, c. HPV-18 E6 oncoprotein-Kaempferol, d HPV-18 E7 oncoprotein-Kaempferol, e. HPV-18 E6 oncoprotein-Morin, f HPV-18 E7 oncoprotein-Morin, g HPV-18 E6 oncoprotein-Myricetin, h HPV-18 E7 oncoprotein-Myricetin, i HPV-18 E6 oncoprotein- Quercetin, and j HPV-18 E7 oncoprotein- Quercetin

Evaluation of drug-likeness properties

Table 3 presents the drug-likeness properties of natural flavonoid glycosides. The results indicate that all of the natural compounds, except for Myricetin, fulfilled Lipinski's rule of five without any violations. Lipinski's rule states that for a compound to have good drug-like properties, it should have a molecular weight less than 500 g/mol, an octanol/water partition coefficient (LogP) less than 5, fewer than 5 hydrogen bond donors, fewer than 10 hydrogen bond acceptors, and a total polar surface area (TPSA) less than 140 Å2.

Table 3 Drug-likeness characteristics of natural flavonoid glycosides including Fisetin, Kaempferol, Morin, Myricetin, and QuercetinEvaluation of ADMET properties

Table 4 provides information on the ADMET (absorption, distribution, metabolism, excretion, and toxicity) properties of natural flavonoid glycosides. The absorption level of these compounds was determined using parameters such as intestinal absorption (human), skin permeability, Caco-2 permeability, and their interactions with P-glycoprotein. A Papp coefficient greater than 8 × 10–6 indicates high Caco-2 permeability and easy absorption. Notably, none of the tested compounds showed poor Caco-2 permeability, and all were predicted to have high absorption levels for intestinal absorption (human). Skin permeability, assessed by the log Kp value, indicated that none of the compounds had low skin permeability.

Table 4 Predicted ADMET properties of natural flavonoid glycosides including Fisetin, Kaempferol, Morin, Myricetin, and Quercetin

In terms of P-glycoprotein interactions, Fisetin, Kaempferol, Morin, Myricetin, and Quercetin were identified as substrates, suggesting that they may be actively excreted by P-glycoprotein. However, none of the compounds were predicted to be P-glycoprotein inhibitors.

The distribution of compounds was evaluated using parameters such as distribution volume (VDss), blood–brain barrier (BBB) permeability (log BB), fraction unbound (human), and CNS permeability. A lower VDss value indicates relatively low distribution volume, while a higher VDss value suggests relatively high distribution volume. The results revealed that all natural flavonoid glycosides had high distribution volumes.

For BBB permeability, a log BB greater than 0.3 indicates easy crossing of the BBB, while a log BB less than − 1 suggests difficulty. Among the compounds, Fisetin (− 1.039), Kaempferol (− 0.939), Morin (− 1.18), Myricetin (− 1.493), and Quercetin (− 1.098) were predicted to have difficulty crossing the BBB. Regarding CNS permeability, Morin (− 3.389), Myricetin (− 3.709), and Quercetin (− 3.065) were predicted to be unable to penetrate the CNS, while Fisetin (− 2.282) and Kaempferol (− 2.228) were predicted to have the ability to do so.

The compounds were also evaluated for their interactions with cytochrome P450 enzymes, specifically CYP2D6, CYP3A4, CYP1A2, CYP2C9, and CYP2C19. None of the compounds were substrates for CYP2D6 and CYP3A4, but all were predicted to be inhibitors of CYP1A2. Furthermore, it was predicted that all the compounds would function as inhibitors for CYP2D6, CYP3A4, and CYP2C19. However, out of these naturally occurring flavonoid glycosides, only Fisetin displayed inhibitory effects on CYP2C9 (Table 4).

The clearance of compounds was assessed based on their molecular weight and hydrophilicity, with Morin having the highest total clearance, followed by Kaempferol, Myricetin, Fisetin, and Quercetin. None of the compounds showed toxicity in the AMES test or hepatotoxicity, according to the prediction results. Additionally, they were not found to inhibit the hERG channel, suggesting a lack of cardiotoxicity, nor did they show any skin sensitization. In summary, the predicted results suggest that all of the tested compounds exhibit similar ADMET characteristics.